JP2008112907A - Image sensor, and manufacturing method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image sensor capable of improving crosstalk, an after-image and a dark current, and to provide a manufacturing method related to the same. <P>SOLUTION: The image sensor is provided with a semiconductor substrate, and a pixel array disposed on the substrate and consisting of a plurality of pixels each having one pixel electrode. The image sensor has a photoelectric conversion layer and a transparent electrode successively disposed on the pixel electrodes. The image sensor has a shield electrode formed between adjacent pixels. The shield electrode is provided in a lattice shape so as to surround the pixel electrode. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image sensor and a manufacturing method thereof, and more particularly to an image sensor including a shield electrode for eliminating carrier crosstalk and a manufacturing method thereof.

  Image sensors such as CMOS (Complementary Metal Oxide Semiconductor) and CCD (Charge Coupled Device) take photons (light) and convert them to electrons, and then convert the electrons into voltages that can be measured and converted as digital data. It is a silicon semiconductor. In order to further improve the functions of a conventional CCD or CMOS image sensor, an image sensor using hydrogenated amorphous silicon (α-Si: H) laminated on a CCD or CMOS device has been developed. Hydrogenated amorphous silicon (α-Si: H) has a direct transition type photoelectric conversion mechanism, so the quantum efficiency is remarkably high, and the high fill factor brought about by the laminated structure makes the entire pixel area photosensitive, High sensitivity is expected.

However, as pointed out in the past, this type of image sensor has problems such as crosstalk, afterimages, and dark current signals. In particular, carrier crosstalk between adjacent pixels decreases the resolution of photoresponse and the uniformity of photoresponse, causes color crosstalk, and reduces color fidelity. Furthermore, the afterimage is a phenomenon caused by a decrease in carrier velocity in an α-Si: H-based carrier transport structure using deep traps or field emission. In the α-Si: H-based carrier transport structure, the image is tailed from the bright spot of each frame of the moving image and becomes an afterimage. When an afterimage occurs, it becomes impossible to faithfully reproduce colors in a low signal level state. This is because all the required signals cannot be read out in one frame. In addition, the dark current signal is generated by tunneling electrons and holes from the metal electrode to the p-type layer and the n-type layer of the photoelectric conversion layer, and causes noise on the dark screen. Therefore, in order to produce a high-quality image sensor superior to conventional silicon-based CCD / CMOS image sensors, it is indispensable to solve the above problems.

  The conventional structure based on α-Si: H includes the following elements. (1) A transparent metal layer such as ITO (indium tin oxide). (2) A boron ion heavily doped p-type layer made of hydrogenated amorphous silicon carbide (α-SiC: H) that collects photogenerated holes formed in the intrinsic layer and contacts ITO. (3) An α-Si: H intrinsic layer (i layer) used as a photoelectron / hole pair generation layer. (4) A phosphorus ion heavily doped n-type layer made of hydrogenated carbon-doped amorphous silicon for receiving electrons from the i layer and sending them to the metal pixel electrode. (5) A metal pixel electrode laid under an n-type layer so as to be connected to a transistor and vertically stacked on a CMOS circuit on a silicon substrate. Please refer to FIG. FIG. 1 is an energy band diagram of a p-i-n heterojunction having the i-layer / n-type layer interface. Among them, the conversion gain from charge to voltage is determined by the sensing capacitance, and the sensing capacitance can be minimized by increasing the thickness of the i layer.

  To improve the quantum efficiency of α-Si: H-based i-layers, optimizing the hydrogen content, extending the lifetime of minority carriers and increasing carrier mobility, photoelectric conversion and light absorption It is desirable to improve. Further, as an alternative to the boron ion heavily doped p-type layer provided under the ITO layer, a CH4 based α-SiC: H layer that forms a heterojunction with an α-Si: H based i layer can also be used. This is because the large optical band gap (Eopt) of SiC improves the transparency, and the CH4-based α-SiC: H layer widens the energy band gap and prevents electron emission from the ITO layer to the p-type layer due to the tunnel effect. This is because the problem of the dark current signal can be suppressed. α-SiC: H is also suitable for n-type layers. This is because α-SiC: H lowers the conductivity of the n-type layer between pixel electrodes and prevents lateral carrier crosstalk between pixels. Further, it is possible to prevent the emission of holes from the pixel electrode made of titanium nitride (TiN) similar to the electron tunnel injection to the p-type layer to the n-type layer. However, since α-SiC has a high density of deep traps, dark current signals and afterimages become serious problems if they are used.

  Further, a more serious problem may occur at the corner of the pixel electrode adjacent to the n-type layer. This is because, as shown in FIG. 2, the electric field strength concentrated locally on the sharp edge of the pixel electrode bends the energy band, raises the hole tunnel transition probability in the reverse bias state, and increases the dark current. That is. The α-SiC-based n-type layer can be formed without peeling off from the pixel electrode made of titanium nitride (generally occurs at the sharpened portion of the pixel electrode due to tension), but it is pulled as shown in FIG. In addition, the trap density is high in the portion of the α-SiC film corresponding to the corner of the pixel electrode. This contributes to pixel defects and afterimages.

  Please refer to FIG. 3 to FIG. 3 is a cross-sectional view of a conventional image sensor including a pin stacked structure, FIG. 4 is an equivalent circuit diagram of the image sensor shown in FIG. 3, and FIG. 5 is an energy of the pixel electrode and the pixel electrode gap shown in FIG. It is a band diagram. The conventional image sensor 10 includes a plurality of pixel circuits (non-display) and a separation film 24 provided on a substrate (non-display), a plurality of pixel electrodes 12 provided on the pixel circuit and the separation film 24, and a pixel. A photoelectric conversion layer 14 provided on the electrode 12 and a transparent electrode 16 provided on the photoelectric conversion layer 14 are included. Among them, the photoelectric conversion layer 14 has a structure in which an n-type layer 18, an i-layer (intrinsic layer) 20, and a p-type layer 22 are sequentially laminated from the bottom to the top, and this structure is referred to as a p-i-n laminated structure.

  Next, the photoelectric conversion layer 14 and the four capacitive elements Cpd, Csub, C1, and C2 will be described. As shown in FIG. 3, these capacitive elements are arranged with the n / i layer interface of the pixel electrode gap as a node. Cpd is a capacitive element toward the transparent electrode 16 made of ITO, Csub is a capacitive element in the p-type silicon substrate (not shown) and the separation film 24 made of silicon oxide (SiO2), and C1 and C2 are The metal pixel electrodes 12 are adjacent to each other. The structure of the conventional image sensor 10 shown above can be regarded as an n-type channel MISFET (metal-insulated semiconductor field effect transistor) 30 in which pixels are reversed. As shown in FIG. 4, the source and drain are connected to both pixel electrodes 12, the substrate bias is supplied from the transparent electrode 16 to the p-type layer 22, and the grounded silicon substrate is passed through the thick dielectric film. This is because it is considered to correspond to the gate of the MISFET 30 having the gate capacitance Csub.

  Here, compared with C1 and C2, Cpd and Csub cannot secure a sufficiently large capacity ratio from the aspect ratio. Therefore, due to the two-dimensional spatial effect between adjacent electrodes, the potential distribution between the pixel electrodes 12 is affected by the supply bias voltage applied to both electrodes, and the potential rises. Therefore, the coupling between C1 and C2 raises the channel potential profile of the flip MISFET 30 high, but in the pixel electrode gap region, the potential that Cpd and Csub lower the level cannot be maintained. As a result, the electrostatic barrier height of electrons is reduced as compared with the one-dimensional approximation, and the crosstalk current between the pixels shown in FIG. 5 is easily generated.

  Please refer to FIG. FIG. 6 is a vertical energy band diagram of the structure including the pixel electrode and the electrode gap region shown in FIG. In FIG. 6, the electrode gap refers to an interval between adjacent pixel electrodes 12, and an electron channel layer exists at the interface of the i layer 20 / n-type layer 18. Since the materials constituting the i-layer 20 and the n-type layer 18 form a heterojunction band having individual energy conduction band levels, electrons accumulate at the interface between them to form an electron channel layer. Conceivable. Then, the photogenerated electrons flow into the pixel electrode 12 through the conduction band of the n-type layer 18. At the interface of the i layer 20 / n-type layer 18, a carrier path is formed between the adjacent pixel electrodes 12 in the horizontal direction, resulting in crosstalk as described with reference to FIGS. Further, as shown in FIGS. 5 and 6, the potential barrier in the pixel electrode gap region is lowered. Considering a MISFET model as shown in FIG. 4, crosstalk can be suppressed by making the i layer 20 thin or applying a high voltage to the pixel electrode 12 to enhance the substrate effect. However, a thick i-layer is excellent in quantum efficiency. However, in an image sensor with a thick i layer, the substrate effect due to the potential distribution between the pixel electrode gap regions is weak, so that the surface potential cannot be lowered and crosstalk is likely to occur. Since the conventional image sensor uses the thin i-layer 20, sensitivity and color balance must be sacrificed, and obtaining a high pixel electrode voltage also places a heavy burden on the power supply apparatus.

  An object of the present invention is to provide an image sensor and a manufacturing method thereof for solving the above-mentioned problems.

The present invention provides an image sensor. The image sensor includes a semiconductor substrate and a pixel array including a plurality of pixels disposed on the substrate. Each pixel includes a pixel circuit formed on the substrate and a pixel electrode connected to the pixel circuit. The image sensor further includes a shield electrode formed between adjacent pixels, a photoelectric conversion layer provided on the shield electrode and the pixel electrode, and a transparent electrode that covers the photoelectric conversion layer. Among them, the shield electrode is provided in a lattice shape so as to surround the pixel electrode.

The present invention further provides a method for manufacturing an image sensor. The method provides a substrate including a plurality of pixel circuits each corresponding to a pixel defined on the substrate, forms a conductive layer on the substrate, and forms a conductive layer in a first photolithography etching process (PEP). Each pixel electrode is formed in each of the pixels so as to be connected to a corresponding pixel circuit, and a shield electrode is formed between any two adjacent pixel electrodes. An insulating film is coated on the electrode. Further, the insulating film is removed by leaving a part of the pixel electrode and the shield electrode by a second photolithography / etching process (PEP), an electrode photoelectric conversion layer is formed thereon, and the transparent conductive layer covering the photoelectric conversion layer is formed. It comprises the steps of forming a layer.

The present invention provides an image sensor that suppresses crosstalk by increasing a potential barrier between pixel electrodes by a shield electrode. Further, the insulating layer covering the end portions of the shield electrode and the pixel electrode can be expected to have a tunnel effect, an effect of suppressing afterimage and dark current, and a high-quality image sensor can be realized.

In order to elaborate on the features of such an apparatus and method, specific examples are given and described below with reference to the figures.

Please refer to FIG. 7 and FIG. FIG. 7 is a sectional view of an image sensor according to the present invention, and FIG. 8 is a plan view showing a part of the image sensor 100 shown in FIG. As shown in the figure, a POAP (photoelectric conversion body on active pixel) type image sensor 100 is provided on a semiconductor chip 102 including a silicon substrate 104. The image sensor 100 includes a dielectric layer 106 provided on the substrate 104 and a plurality of pixels 108 defined on the substrate 104. Among them, the plurality of pixels 108 are arranged in the pixel array 110 shown in FIG. 8. Each pixel 108 is provided in the dielectric layer 106, and includes a pixel circuit 112 including at least one MOSFET, and a pixel electrode 114. Including. The pixel electrode 114 made of a metal material such as titanium nitride (TiN) is electrically connected to the corresponding pixel circuit 112 via the via plug 136. As another example, the pixel electrode 114 can be made of a conductive metal such as tungsten, aluminum, or copper. In addition, the electrode gap area | region G in a figure points out the space | interval between the edges of the adjacent pixel electrode 114. FIG.

In the electrode gap region G between any two adjacent pixel electrodes 114, one shield electrode 116 is always provided. Therefore, as shown in FIG. 7, the shield electrode 116 has a lattice shape surrounding the pixel electrode 114. According to a preferred embodiment of the present invention, the shield electrode 116 provided at the center of the electrode gap region G forms a boundary between the pixels 108 while maintaining an equal distance d from the adjacent pixel electrode 114. The potential is supplied from a potential supply circuit formed around the pixel array. The shield electrode 116 and the pixel electrode 114 can be made of the same material (for example, titanium nitride), or can be made by the same process. The shield electrode 116 to which the ground voltage is applied does not form a current path in the pixel region. The shield electrode 116 electrically isolates the plurality of pixels 108 by lowering the potential level close to the surface of the pixel electrode gap region G. On the other hand, the conventional image sensor has a strong mutual electric field effect between the pixels, the inter-pixel barrier is lowered, and a crosstalk current is generated.

In the present invention, an insulating layer 118 is further disposed on the ends of the dielectric layer 106, the shield electrode 116, and the pixel electrode 114. The insulating layer 118 made of a thin oxide film (for example, silicon oxide SiO 2) exposes the central portion of the pixel electrode 114, and the pixel electrode 114 is connected to the photoelectric conversion layer 120 through the central portion.

The image sensor 100 further includes a photoelectric conversion layer 120 and a transparent conductive layer 122 that cover the insulating layer 118 and the pixel electrode 114. The photoelectric conversion layer 120 includes an n-type layer 130, an i-layer 132, and a p-type layer 134 from bottom to top, of which the i-layer 132 is made of α-Si: H, and the p-type layer 134 and the n-type layer 130 are α -Made of SiC: H. In order to improve the sensitivity and color balance of the image sensor 100, the thickness H of the i layer 132 must be about 5000 mm or more. The transparent conductive layer 122 forming the upper electrode plate is made of ITO. The image sensor 100 further includes a first planarization layer 124, a color filter layer 126, and a second planarization layer 128 provided on the photoelectric conversion layer 120, which are sequentially stacked. Among them, the color filter layer 126 includes different color filters such as red, green, and blue according to the pixel 108 to which the color filter layer 126 belongs.

The effect of the present invention can also be explained in the equivalent circuit shown in FIG. Of the capacitances formed with reference to the node of the interface of the i layer / n-type layer 130 corresponding to the center of the electrode gap region G, Csub is the capacitance of the shield electrode 116, and Cpd is the capacitance of the transparent electrode (ITO) 122. C1 and C2 are coupled to adjacent metal pixel electrodes 114, respectively. In such a structure, Csub is much larger than Csub of a conventional image sensor that does not include a shield electrode. As a result, the potential near the surface can be suppressed to a low level by the shield electrode 116. As a result, a one-dimensional barrier is formed in the pixel electrode gap region. Therefore, carrier crosstalk between adjacent pixels generated in the conventional image sensor as shown in FIG. 5 is suppressed.
Please refer to FIG. FIG. 9 is an energy band diagram for the adjacent pixel electrode 114 and shield electrode 116 of the image sensor 100 shown in FIG. Since the potential under the shield electrode 116 is kept low through the thin insulating layer 118 and the α-SiC n-type layer 130, the crosstalk generated at the interface of the i layer 132 / n-type layer 130 is caused by the electrode gap region G High potential barrier (for example, a one-dimensional barrier). Therefore, even in the conventional bias state, it is possible to improve the quantum efficiency by using the thick i layer 132.

On the other hand, the thickness of the thin insulating layer 118 on the shield electrode 116 is determined based on the potential level at the interface between the α-Si: H base i layer 132 and the α-SiC: H base n-type layer 130. The gate capacitance (Csub) is most preferably adjusted to the extent that it can be maximized. The adjustment of the thickness of the insulating layer 118 protects the sharp end of the pixel electrode 114, prevents hole tunnel injection caused by the concentration of the electric field, relaxes the dark current, and forms an α-SiC: H-based n-type. This has the effect of reducing deep traps caused by tension in the layer 130 and solving the afterimage problem. As shown in FIG. 8, the end portions of the shield electrode 116 and the pixel electrode 114 are covered with an insulating layer 118, and the tension and strong electric field in the contact region between the α-SiC: H n-type layer 130 and the pixel electrode 114 are covered. And the problem of leakage current that tends to occur at the end of the pixel electrode 114 of the image sensor 100 can be solved.

Please refer to FIG. FIG. 10 is a potential diagram when the thickness of the i layer 20 is 5000 mm, 7000 mm, and 10,000 mm in the conventional image sensor 10 shown in FIG. As shown in FIG. 10, the adjacent pixel electrodes 12 have potentials of 1.2 V and 2.6 V, respectively, and there is almost no potential barrier in the gap region between them. Then, the electrons generated in the i layer 20 move from the high-potential right pixel electrode 12 to the low-potential left pixel electrode 12 and become a crosstalk current. In contrast, FIG. 11 is referred to. FIG. 11 is a potential diagram when the thickness of the i layer 132 is 5000 mm, 7000 mm, and 10,000 mm in the image sensor 100 according to the present invention shown in FIG. As shown in FIG. 11, adjacent pixel electrodes 114 have potentials of 1.2 V and 2.6 V, respectively, and the potential barrier in the gap region G between them is relatively high. Therefore, the lateral electric field formed between adjacent pixel electrodes 114 having a potential difference is not so strong, and the shield electrode 116 of the image sensor 100 forms a high potential barrier between the pixel electrodes 114, and optical crosstalk is reduced. The problem is also improved. Therefore, the image sensor according to the present invention can be applied to an i layer having a thickness of 5000 mm or more.

Reference is now made to FIGS. FIG. 12 to FIG. 16 are explanatory diagrams showing the manufacturing process of the image sensor 100 according to the present invention. First, as shown in FIG. 12, a semiconductor chip 102 made of a silicon substrate 104 is provided. Next, a plurality of electrical elements are formed on the substrate 104. These electric elements constitute a pixel circuit 112 in the dielectric layer 106. Thereafter, a conductive layer 138 having a thickness of about 300 mm made of a metal material (preferably titanium nitride) is formed on the dielectric layer 106, that is, on the pixel circuit 112. Please refer to FIG. Next, a part of the conductive layer 138 is removed by a photolithography etching process, a pixel electrode 114 is formed for each pixel 108, and a shield electrode 116 is formed between the pixel electrodes 114. In other words, the shield electrode 116 is formed in the same plane as the pixel electrode 114 and is formed so as to be equidistant from both adjacent pixel electrodes 114. In this embodiment, the width of the shield electrode 116 is set to 0.2 μm, and the distance from the pixel electrode 114 to the shield electrode 116 is also set to 0.2 μm.

Thereafter, as shown in FIG. 14, a thin insulating layer 118 made of an insulating material such as silicon oxide and having a thickness of about 200 mm is laminated on the substrate 104 so as to cover the pixel electrode 114 and the shield electrode 116. Refer to FIG. Next, a part of the insulating layer 118 is removed by another photolithography etching process, and most of the pixel electrode 114 is exposed except for the end portion of the pixel electrode 114 and the shield electrode 116. Subsequently, an α-SiC: H n-type layer 130, an α-Si: H i-layer 132, and an α-SiC: H p-type layer 134 constituting the photoelectric conversion layer 120 are formed on the substrate 104. The n-type layer 130 is formed to be electrically connected to the pixel electrode 114, and the pixel electrode 114 is formed to be coupled to the corresponding pixel circuit 112 through the via plug 136. Here, as an embodiment of the present invention, the thickness of the p-type layer 134 can be set to 50 mm, the thickness of the i layer 132 can be set to 5000 mm, and the n-type layer 130 can be set to 100 mm. Thereafter, as shown in FIG. 15, a transparent conductive layer 122 is formed on the photoelectric conversion layer 120, and the first planarization layer 124, the color filter layer 126, and the second planarization layer 128 are further formed on the transparent conductive layer 122. And the fabrication of the image sensor 100 is completed.

The image sensor manufactured by the above method suppresses crosstalk by increasing the potential barrier between the pixel electrodes. The insulating layer covering the end portions of the shield electrode and the pixel electrode is effective in preventing a tunnel effect and improving afterimage and dark current. Therefore, the image sensor according to the present invention can express good image quality.

The above is a preferred embodiment of the present invention and does not limit the scope of the present invention. Accordingly, any modification or change that can be made by the right holder, which is made under the intention of the present invention and has an equivalent effect on the present invention, belongs to the scope of the claims of the present invention. And

The present invention is characterized in that a shield electrode is added to a conventional image sensor. Such a structure is feasible.

It is an energy band diagram of a conventional p-i-n heterojunction having an α-Si: H i layer / α-SiC: H n-type layer interface. It is explanatory drawing showing the problem by the tension | tensile_strength which arises in the edge part of a pixel electrode, and a tunnel effect. It is sectional drawing of the pixel of the image sensor containing the conventional p-i-n laminated structure. FIG. 4 is an equivalent circuit diagram of the image sensor shown in FIG. 3. FIG. 4 is an energy band diagram of a pixel electrode and a pixel electrode gap shown in FIG. 3. FIG. 4 is a vertical energy band diagram of a structure including a pixel electrode and an electrode gap region shown in FIG. 3. It is sectional drawing of the image sensor by this invention. FIG. 8 is a plan view illustrating a part of the image sensor illustrated in FIG. 7. FIG. 8 is an energy band diagram of an image sensor including adjacent pixel electrodes and shield electrodes shown in FIG. 7. FIG. 4 is a potential diagram when the thickness of the i layer is 5000 mm, 7000 mm, and 10,000 mm in the conventional image sensor 10 shown in FIG. 3. FIG. 8 is a potential diagram when the thickness of the i layer is 5000 mm, 7000 mm, and 10,000 mm in the image sensor 100 according to the present invention shown in FIG. 7. It is 1st explanatory drawing showing the manufacture process of the image sensor by this invention. It is 2nd explanatory drawing showing the manufacture process of the image sensor by this invention. It is a 3rd explanatory drawing showing the manufacture process of the image sensor by this invention. It is the 4th explanatory view showing the manufacture process of the image sensor by the present invention. It is a 5th explanatory view showing the manufacture process of the image sensor by the present invention.

Explanation of symbols

10, 100 Image sensor 12, 112, 114 Pixel electrode 14, 120 Photoelectric conversion layer 16 Transparent electrode 18, 130 n-type layer 20, 132 i-layer 22, 134 p-type layer 24 Separation film 30 MISFET
102 Semiconductor chip 104 Silicon substrate 106 Dielectric layer 108 Pixel 112 Pixel circuit 116 Shield electrode 118 Insulating layer 122 Transparent conductive layer 124 First planarization layer 126 Color filter layer 128 Second planarization layer 136 Via plug Cpd, Csub, C1, C2 Capacitor

Claims (6)

An image sensor including a semiconductor substrate, a pixel array including a plurality of pixels each having one pixel electrode disposed on the substrate, and a photoelectric conversion layer and a transparent electrode sequentially disposed on the pixel electrodes And a shield electrode formed between adjacent pixels, wherein the shield electrode is provided so as to surround the pixel electrode in a grid pattern. The image sensor according to claim 1, wherein the shield electrode is provided on the dielectric layer and in the same plane as the pixel electrode. 2. The image sensor according to claim 1, wherein an insulating layer is formed on end portions of the shield electrode and the pixel electrode, and other pixel electrode portions are directly formed under the photoelectric conversion layer. Image sensor. 2. The image sensor according to claim 1, wherein the potential of the shield electrode is supplied from a potential supply circuit formed in a peripheral portion of the pixel array. An image sensor manufacturing method,
Providing a substrate including a plurality of pixel circuits each corresponding to a pixel defined on the substrate;
Forming a conductive layer on the substrate;
A part of the conductive layer is removed by a first photolithography etching process (PEP), and one pixel electrode is formed in each of the pixels so as to be connected to the corresponding pixel circuit. Forming a shield electrode between two pixel electrodes,
Forming a photoelectric conversion layer on the pixel electrode and the shield electrode;
A method for producing an image sensor, comprising the step of forming a transparent conductive layer covering a photoelectric conversion layer. The manufacturing method further includes
Forming an insulating layer covering the pixel electrode and the shield electrode;
The method includes a step of removing a part of the insulating layer in the second PEP process, leaving at least the insulating layer covering the end portion of the pixel electrode and the shield electrode, and exposing the central portion of the pixel electrode. The image sensor manufacturing method according to claim 5.