JPS5819146B2 - charge transfer device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an improvement in a charge transfer device, particularly a charge transfer device having a meandering charge path.

Various types of charge transfer devices have been proposed.
Generally, electrodes for transferring charge are divided into bits and arranged in a straight line, but in a meandering type charge transfer device, charge is transferred in a meandering manner using two electrodes regardless of the number of bits. is possible.

Regarding this, the present applicant previously filed Japanese Patent Application No. 52-80264.
proposed by.

As shown in FIG. 1, this serpentine charge transfer device has an area sandwiched between band-shaped charge layers 1 and 2 extending parallel to each other on one main surface of a semiconductor substrate as a charge transfer region, and each band-shaped Short charge layers 3a, 3b', 3c, . . . extend in a comb-like shape from the charge layer toward the center of the transfer region.
......and 4a, 4b.

4c......... is provided.

Charge layer 3a below. 3b, 3c・・・・・・・・・・・・
・is the first group, similarly 4a+4bt4c...
... will be called the second group.

As is clear from the figure, the two groups of charge layers are arranged so as to protrude alternately from above and below, and this arrangement creates a meandering charge path R within the charge transfer region.

However, in order to transfer charges along the direction indicated by the arrowed curve 5 within the charge path R, a well (Potenti 1 w'e I
l').

One way to achieve this is to locally introduce impurities into the charge transfer region.

This point will be explained in detail next.

two short charge layers adjacent to each other in the same group within the charge transfer region;
For example, the area sandwiched between 3a and 3b will be called a cell for convenience of explanation.

Therefore, for example, a region sandwiched by short charge layers 4a and 4b is a separate cell from a region sandwiched by the same <3a and 3b.

Further, each cell is numbered 11, 12, 13, . . . in order in the direction indicated by the arrowed curve 5.

For example, suppose that charge is currently accumulated in cell 12,
At the time of the next charge transfer, the transfer voltage is simultaneously applied to cell 11 and cell 13, so the charge is not necessarily transferred only to cell 13, but may also flow back toward cell 11.

Therefore, in the device shown in FIG. 1, impurities are doped into the surface of the substrate corresponding to the region 128 in the cell 12 to form a barrier region that prevents the reverse flow of charges, and the depth of the potential well generated under this region is is shallower than that under the undoped region.

By making a difference in the impurity concentration in the surface layer of both regions in this way, all of the charges that have moved from cell 11 to the shallow well below the impurity doped region 12a of cell 12 are immediately transferred to the deep well immediately below the undoped region 12b. It flows into the interior and is accumulated here.

Therefore, no backflow of charge occurs.

Similarly, in cell 13, a shallow well is formed under the region 13a and a deep well is formed under the region 13b, so 12
The charges from b move once under the impurity doped region 13a, and the charges immediately move to the storage section below the undoped region 13b, and do not flow back toward the cell 12 again.

In this way, each cell is made up of an accumulation region for storing charge and an impurity-doped region, that is, a barrier region, for preventing charge transfer and reverse flow.

Note that the barrier area is marked with intersecting diagonal lines for easy understanding.

As is clear from the above explanation, in order to provide directionality to charge transfer, conventionally, barrier regions within cells have been formed. A stepped oxide structure that creates a difference in surface potential by creating a difference in thickness, or a structure that creates a difference in surface potential by creating a difference in the ζζ impurity concentration in the channel of both regions by impurity diffusion, ion implantation, etc. However, in either case, the selective etching or diffusion process becomes complicated, and the number of steps increases, resulting in many problems in terms of process yield and degree of integration.

In view of the above-mentioned points, the present invention aims to reduce the number of steps without deteriorating the function of the meandering type charge transfer device. This method does not use processes such as changing the impurity concentration or creating a difference in impurity concentration within the channel by ion implantation.

The principle of the invention is shown in FIG.

FIG. 2 shows the potential distribution of the charge transfer section of the charge transfer device in which a gate electrode 22 is disposed through an insulating layer 21,
In this example, it is assumed that the gate voltage and substrate concentration are the same.

When the width of the channel 20 in FIG. 2A, that is, between the charge layers 30 and 4A is wide, the surface potential ξS generated therebetween is large as shown by the dotted line, and a deep potential well is created. The example shown in Figure 2A shows a case where the width of the channel 20 between the charge layers 30 and 4A is narrowed, and the surface potential ξS generated therebetween becomes small as shown by the dotted line, creating a shallow potential well.

so-called narrow channel
) effect occurs, and the present invention utilizes this effect,
It is an object of the present invention to provide a charge transfer device in which a barrier region of a charge transfer path can be formed by a structure in which the distance between charge layers is narrowed at positions where a barrier region is to be formed.

In order to achieve this object, the present invention employs a structure in which the channel width between adjacent short charge layers arranged to form the meandering charge transfer path is narrowed, thereby producing a narrow channel effect. It is characterized by comprising a region having a barrier function.

An embodiment of the present invention will be described below with reference to FIG.

FIG. 3 shows an embodiment of the charge transfer device according to the present invention as a top view, and the same parts as in FIG. 1 are given the same reference numerals.

In the embodiment of FIG. 3, each short charge layer 3a.

3b...... and 4a, 4b...
An island-shaped charge layer 31 with a width of 2 to 5 μm and a depth of approximately 5 to 8 μm is located between the protruding ends beyond the center line X-x.
1a, 312i, 313i, and similarly charge layers 4a and 3a.
An island-shaped charge layer 411a similar to that described above is formed between b and b.

412a and 413a are arranged along the center line xx' with an interval of about 3 to 5 μm.

In this case, the island-shaped charge layers 411a, 412a, 413i, etc. are the charge layers 3a, 3b... and 4a,
4b, -0, etc. can be formed simultaneously by the same process.

In this way, so-called barrier regions 401A, 4010401C, 40 or 402A, 4020, 402H, 4022 are formed on both sides of each island-like charge layer, each of which is a barrier due to the narrow channel effect. It becomes possible to functionally transfer charges along the arrowed curve 5.

Therefore, according to the present invention, a high concentration layer, which is the same as a charge weir, is provided continuously or discontinuously between adjacent short charge layers so as to narrow the width of the region where the barrier region is to be formed. This makes it possible to create a barrier region simultaneously with the charge weir.

In this case, the charge weirs that provide the narrow channel effect are arranged in the form of islands, so even if the cell capacitance increases, the conductance of the entire channel can be extremely reduced by dividing the island-like charge layer appropriately. Barrier function can be achieved without

From the above, the charge transfer device of the present invention utilizes the narrow channel effect to narrow the channel width at the position where the barrier region of the charge transfer path is to be formed by the shape of the charge weir, thereby achieving the same effect as the barrier region. I am getting .

Therefore, if a charge weir is formed, the result is equivalent to forming a barrier region at the same time, so the step of forming a barrier region can be omitted, and the number of steps can be significantly reduced, which is an excellent advantage.

[Brief explanation of drawings]

FIG. 1 is a top view showing the charge weir and barrier region of a conventional charge transfer device having a meandering charge path; FIG. FIG. 3 is a top view showing the shapes of the charge weir and barrier region in one embodiment of the charge transfer device according to the present invention. 1 and 2: band-shaped charge layer, 3a, 3bt3ct...
...: 1st group of short charge weirs, 4a, 4b, 4c doors ......: Also 2nd group, 5: Charge path, 1
1゜12.13・・・・・・・・・：Cell, 12a,
13a, 14a......: barrier area, 3
11i, 312i. 313i, 411i, 412i, 413i: island-shaped charge layer, 401i, 4010, 401c. 140 and 402a, 4020, 402c. 4022: Barrier area.

Claims (1)

[Claims] 1. A charge layer having at least a pair of strip-shaped portions extending parallel to each other on the surface of a semiconductor substrate, a charge transfer region sandwiched between the strip-shaped portions, and a charge transfer region extending from the edge of the region beyond the center line within the region. comprising a plurality of short charge layers that alternately protrude from each other and terminate at opposite edges and at a certain interval;
In a charge transfer device in which a meandering charge path is formed in a charge transfer region by the short charge layer, a short charge is generated from the tip of the short charge layer beyond the center line and the opposing edge adjacent to the tip of the short charge layer. At least one island-like charge layer is provided along the center line between the charge island layer and the charge island layer, and charge backflow prevention regions utilizing a narrow channel effect are formed on both sides of the island-like charge layer. charge transfer device.