JPS5847906B2 - solid-state image sensor - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a solid-state image sensor in which a large number of photoelectric conversion elements and a scanning circuit for extracting optical information from each element are integrated on a semiconductor substrate.

The solid-state image sensor must have a resolution comparable to that of the imaging electron tube used in current television broadcasting, and for this reason, 500 x 500 nine-density conversion elements and equivalent (X, y) coordinate selection switches or switches are required. 500 stages of X-scanning circuits and 500-stage y-scanning circuits are required.

Therefore, it is usually manufactured using MOS large-scale integrated circuit technology, which is relatively easy to achieve high integration, and allows the photoelectric conversion element and the switching element to be manufactured in an integrated structure, as shown in FIG.

1 is a y-coordinate selection switch MOS field effect transistor (hereinafter abbreviated as MOST) equipped with a gate 2 that is opened and closed by a y-scanning circuit; 3 is an insulating film; 4 is a pn junction photodiode using the source junction of switch 1; Reference numeral 5 designates the drain of the switch 1 which is connected to the X coordinate selection switch MOST (not shown) through the signal output line 6.

Further, 7 is an N-type silicon semiconductor substrate on which these elements are integrated.

The solid-state imaging device configured as described above can be manufactured with a photoelectric conversion device and a switch in an integrated structure, and is therefore suitable for integration.

Since it uses MOST, it has advantages such as low power consumption.

Therefore, although the number of photodiodes that can be integrated is relatively large, it is still only about IOOXIOO, and this type of solid-state image sensor has the following important problems in realizing an image sensor. .

1. Charges generated by incident light are accumulated in the capacitance of the diode, but a predetermined diode capacitance is required to obtain the necessary signal-to-noise ratio.

The capacitance of a pn junction diode is generally 19F per 100 μm square, and in order to achieve the above signal-to-noise ratio, a 50 μm square is essential.

When 500 x 500 diodes are integrated, the IC chip size of the image sensor is 25i, ya square, and currently the standard Q size is 2 to 3 wafers, and even the smallest one is 7 squares. Considering this, it is extremely difficult to manufacture a 25-angle IC.

2. In order to manufacture an imaging IC, an IC photomask is required as is well known, but the size of the chip that can be photographed by one-shot imaging is currently 10 squares.

Therefore, in the case of a 25 mm square, it must be manufactured by dividing into two blocks in the x and y directions, a total of four blocks, and then connecting each block to each other.

Not only does this make it difficult to manufacture masks, but the image obtained by the image sensor created by the split mask has white (or black) lines reflecting the seams running vertically and horizontally, degrading the image quality.

Therefore, an object of the present invention is to simplify the structure of a diode used for photoelectric conversion and reduce the area occupied by the diode, that is, the size of the image sensor, and to achieve this, a perforated diode is manufactured. .

Hereinafter, the details of the present invention will be explained using examples.

FIG. 2 is a diagram showing the gist of the present invention.

First, for example, as shown in a, an insulating film 9 (S+02 is often used) serving as an etching mask is formed on a silicon semiconductor substrate 8 having a {110} plane, and a part of this insulating film is Etching holes 10 are formed by photo-etching.

Thereafter, pores 11 are formed by orientation-dependent etching.

Orientation-dependent etching can be achieved using an alkaline solution, such as an aqueous KOH solution.

In addition to KOH, examples include hydrazine and ethylenediamine.

Taking silicon and KOH aqueous solution as an example, the etching rate of the {111} plane is particularly slow compared to the {110} plane, and if appropriate conditions are chosen, the etching rate can be several hundred times lower than that of planes other than the {111} plane. It is possible to obtain.

By utilizing this difference in etching rate, the pores can be sufficiently shaped.

As shown in b, the insulating film in the part that will become the drain region of the MOST is removed by photo-etching. A substrate of the first conductivity type and a second conductivity type opposite to the first conductivity type substrate are
A conductive type region 12 is formed.

Here, 12-1 is a junction type photodiode that plays a photosensitive role, and 12-2 is a drain.

Thereafter, as shown in C, the insulating film 1 is formed by a thermal oxidation method or the like.
3, and an electrode connection hole 14 is formed on the drain 12-2 by photoetching or the like, and then (H
The gate electrode 15 and drain electrode 16 are shaped as shown.

Here, 15 is a gate that is opened and closed by a scanning circuit (not shown) provided at the periphery, and 16 is a photodiode 12.
This becomes a signal output line that takes out the -1 signal.

Gate Electrode As is clear from this figure, the photodiode of the present invention has a vertical structure, unlike the conventional planar structure (FIG. 1), so that the occupied area can be greatly reduced.

When orientation-dependent etching is used, as shown in the example above, the etching rate in the X direction and the etching rate in the two directions also depend on the plane orientation of the pores 11 relative to the substrate, but can have a difference of several hundred times at most. It is relatively easy to select the etching rate in the two directions to be 100 times that in the X direction.

Figure 3a shows a plan view of the pore 11, and the pore has four side walls L, , L2, L3 as shown in the figure.
, has L4.

Even if the length of one piece is designed to be 6 μm, a depth of 100 μm in two directions can be formed by orientation-dependent etching, so the surface area of the entire sidewall is 2400 μm2 (
6X4X10. O).

Therefore, in the present invention, it is possible to provide a diode capacitance of 6 μm square corresponding to that of a conventional planar diode of 50 μm square.

This means that the overall size of the image sensor can be reduced to 1/10 of the conventional size, and 500 x 500 pixels can be reduced to 7
It will fit within about an angle.

In addition, when the photodiode is made into a pore type photodiode, as shown in Figure 3b, the light 30 incident on the photodiode is not reflected on the top surface as in a conventional planar diode, but instead is reflected by the arrow in the figure. Since it is repeatedly reflected toward the inside of the substrate, it has the effect of absorbing all of the incident light without loss, and is also useful for improving photosensitivity.

FIG. 4 shows another embodiment of a nine-density conversion element which is a unit of the solid-state image sensor of the present invention.

First, an insulating film 1 that serves as an etching mask is formed on the substrate 17.
8, and etching holes 19 are formed in this film by the λ-etching method.

After that, the pores 20 are formed by orientation-dependent etching.

Thereafter, a self-aligned electrode 21 is formed on a predetermined insulating film 18, and using this as a mask, a second conductivity type region 22 is formed by known ion implantation or thermal diffusion method.

Here, 22-1 is a photodiode, and 22-2 is a drain for extracting signal charges from the photodiode.

The self-aligned electrode 21 may be made of a material that can withstand ion implantation or thermal diffusion.
, W, and other high melting point metals are often used.

Furthermore, on top of that, CVD (Chemical Vapor
PSG (Phospho-Si) film made by adding phosphorus or boron
licate Glass) and BSG (Borosil
The second layer insulating film 2 is represented by
3, an electrode connection hole 24 is formed on the drain 22-2, and a drain electrode 25 is provided.

In this embodiment, since the photodiode region, the train region, and the gate are formed in self-alignment, miniaturization of the device can be achieved.

As described above in detail using the examples, a perforated photodiode is formed on a semiconductor substrate by using orientation-dependent etching using a KOH solution, and this is arranged in a two-dimensional manner. The solid-state image sensor of the present invention can reduce the occupied area to 1/10 of that of a conventional solid-state image sensor, and the manufacturing of the image sensor becomes significantly easier.

It also has extremely great practical value, such as being able to utilize incident light without loss.

In addition, in the above explanation, MOST is used as a constituent unit of the image sensor.
We took the combination of photodiode and photodiode as an example, but
It is possible to consider the use of bipolar transistors, phototransistors, and charge transfer elements without departing from the spirit of the invention.

,,Brief explanation of the drawings Fig. 1 shows the structure of a conventional solid-state image sensor, Fig. 2 shows the cross-sectional structure of the solid-state image sensor according to the present invention, and Fig. 3 shows the characteristics of the solid-state image sensor according to the present invention. , FIG. 4 is a diagram showing another embodiment of the solid-state imaging device according to the present invention.

Claims (1)

[Claims] 1. Photoelectric conversion elements each having one or more switch MOS field effect transistors connected thereto are arranged one-dimensionally or two-dimensionally on the same semiconductor substrate, and have a scanning circuit that selects the photoelectric conversion elements time-sequentially. In the solid-state imaging device, a pore is formed at a predetermined position of the semiconductor substrate of a first conductivity type, a region of a second conductivity type is formed on the entire surface of the pore, and this is used as a junction photodiode for photoelectric conversion; A second conductivity type region located apart from the second conductivity type region of the pore is used as a drain for reading out signal charges of the photodiode, and an insulated region is formed on the semiconductor substrate between the junction diode and the drain. A gate electrode is disposed through the film, and a switch consisting of the diode, the drain, and the gate is sequentially opened and closed by the scanning circuit, and an optical signal accumulated by the incident light is transferred to the junction capacitance of the junction photodiode. A solid-state imaging device characterized in that electric charges are taken out to the drain.