JPS60257676A - Solid-state image pickup device - Google Patents

Abstract

PURPOSE:To decrease the reduction in signal electric charge due to the decrease in BCCD channel width by separating a BCCD for signal transfer and a photodiode of an adjacent picture element string by means of an insulator even at the deep depth of the substrate to suppress generation of smear noise. CONSTITUTION:A light incident in a P-channel silicon substrate 1 through an N region 2 forming a PN junction photodiode generates an electric charge. After the electric charge is stored in the PN junction capacitance, the charge is transited to the BCCD3 by a switching gate 4 and then transferred as a signal current. The BCCD3 is separated from the N region 2 forming the photodiode of adjacent picture elements at the depth of the P-channel silicon substrate by means of a silicon dioxide film 8. Thus, a part of electric charges is not invaded to the said N region 2 as shown in the arrow B' to reduce remarkably smear noise.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a solid-state imaging device used in a video camera or the like.

Conventional Structures and Problems There has been a remarkable improvement in the performance of solid-state imaging devices in recent years, and they have already reached the stage of practical use, mainly for the purpose of downsizing home video cameras and the like. In particular, photoelectric conversion is performed using a PN junction photodiode, and the charge transfer section is
The interline transfer type CCD image sensor is formed using OD.

It is becoming popular because it produces less noise and can easily be made highly sensitive. The conventional CCD imaging device as described above will be explained below with reference to one drawing.

FIG. 2 is a schematic cross-sectional view perpendicular to the CCD transfer direction of a unit pixel of a conventional interline transfer type CCD image sensor. In Figure 2, (1) is a P-type silicon substrate, (2) is an N-type region that cooperates with the substrate (1) to form a PN junction photodiode, and (3) is a buried channel CCD (BCCD). N shadow area, (4) is the oxide film formed on the substrate surface of the final plate tJd between N shadow areas (2) and (3), (5) is the photodiode to BCCD (3) on the oxide film (4). (6) is a polycrystalline silicon film used to form a switching gate for transferring signal charges to the adjacent pixel, and (6) is a selective oxide film (LOCO8
), and (7) is a channel stopper made of a P region formed under the selective oxide film (6). Note that the circles in the figure indicate charges generated by incident light, and the arrows indicate the movement of charges within the substrate.

The operation of the unit pixel configured as described above will be explained below. 'First, the P-type silicon substrate (1) passes through the N shadow region (21) that forms the P-N junction photodiode.
Light incident on generates an electric charge.

Most of it moves as shown by arrow IA and accumulates PNN customer service capacity. After these charges are accumulated for a certain period of time, the switching gate;
(It is transferred to 31, and after that transfer, it becomes an output signal current.

However, the above structure has two major drawbacks: generation of smear noise and decrease in the amount of +BCCD transfer charge. That is, smear noise is generated because a portion of the charge directly flows into the N shadow region (3) as shown by arrow B, and it becomes a false signal and significantly deteriorates the image quality.

On the other hand, the decrease in the transferred charge amount of %B CCD is due to the N shadow area (3
), that is, the width of the transfer channel, is narrowed due to the bird's beak in the LOCO8 separation and the lateral diffusion of high concentration boron, which is a channel stopper.

These drawbacks become even more serious when an image pickup device is made smaller (Is) and more highly integrated, and this has become a major problem in reducing the size and improving the performance of video cameras.

OBJECTS OF THE INVENTION In view of the above-mentioned drawbacks, the present invention provides a solid-state imaging device that generates less smear noise and has a transfer channel width that can handle sufficient r3 charge even when downsized.

Structure of the Invention In order to achieve this object, the solid-state imaging device of the present invention has a photoelectric conversion section and a signal generated in the photoelectric conversion section. A unit pixel including a COD that transfers charge is separated from an adjacent unit pixel by an insulator having a substantial depth compared to the COD channel portion. There is.

With this configuration, the rate at which charges generated within the substrate directly flow into the BCOD transfer section of an adjacent pixel is reduced, making it possible to significantly reduce smear noise. In addition, since the thickness of the insulator is thin, CCD as in LOGO8 separation
There is no decrease in channel width, and the amount of signal charge can be significantly increased.

EMBODIMENT OF THE INVENTION Hereinafter, one embodiment of the present invention will be described with reference to the drawings. FIG. 3 is a schematic cross-sectional view of a unit pixel of a CCD solid-state image sensor according to an embodiment of the present invention. In FIG. 1, (8) is a silicon dioxide film forming a separation region from adjacent unit pixels. Also, -fil is a P-type silicon substrate, (2) and (31(i) are N shadow regions forming the photodiode and BCCD, respectively, and (4) is a polycrystalline silicon gate. It is.

The silicon dioxide film/+st can be easily formed by conventional techniques such as etching the silicon substrate by reactive ion etching using chlorine gas and then CVD using silane gas and oxygen gas.

The operation of the unit pixel configured as above is as follows.

First, the N shadow region (
The light incident on the P-type silicon substrate (1) through 2) generates charges. After the charge is accumulated in the PN junction capacitance, it is transferred to the switching gate (4) and then to the BCCD (3), and the operation of transferring it as a signal current is the same as in the conventional structure example shown in FIG.

However, according to this embodiment, BCCDf3) is the N shadow area (2) forming the photodiode of the adjacent pixel.
Since it is completely separated from the silicon dioxide film (7) in the deep part 1 of the pPP type silicon substrate, part of the charge does not mix into the N shadow region (3) as shown by arrow B/. , smear noise can be significantly reduced.

Further in -LOCO8 separation. Since a so-called 11 bird's beag '7 or a high concentration region of boron as a channel stopper is not required, the amount of transferred charge does not decrease at all due to a decrease in the number of 113 COD channels.

In this practical example, the silicon substrate is of P type, but the same effect can be obtained with a P well structure in which a P well is formed on an N type substrate and elements are configured therein. Further, although a -PN junction photodiode is used, a similar effect can be obtained with a structure in which compound semiconductors, amorphous silicon films, or the like are laminated.

Effects of the Invention As described above, the present invention provides B COD for signal transfer.
Since the F1 diodes of adjacent pixel rows are separated by an insulator even deep within the substrate, the generation of smear noise is greatly suppressed, and the signal charge is also reduced due to the reduction of the BCCD channel width. Although it is small, its practical effects are great.

[Brief explanation of the drawing]

FIG. 1 is a schematic cross-sectional diagram of a unit pixel of a solid-state imaging device according to an embodiment of the present invention, and FIG. 2 is a schematic cross-sectional diagram of a unit pixel of a conventional solid-state imaging device. fl)---P-type silicon substrate+21. f, 3+ ---N shadow region (4)--m-polycrystalline silicon film f5L (6), (8) ---: silicon oxide film f7)
------High concentration P-type region patent author Kenji Arata Matsushita Electronics Co., Ltd. Agent (1 other person) Figure 1 Figure 2

Claims (1)

[Claims] A unit configured by including a photoelectric conversion section formed on the main surface of a semiconductor substrate and a CCD channel section formed on the main surface in proximity to the photoelectric conversion section for transferring signal charges generated in the same section. A solid-state imaging device characterized in that a pixel is separated from an adjacent unit pixel by an insulation (4) having a substantial depth compared to a CCD channel portion.