JPH0230183A - Solid-state image sensing element - Google Patents

Abstract

PURPOSE:To prevent an increase in a residual image caused by an imperfect readout operation of a signal charge by a method wherein the surface of a semiconductor substrate in a readout region under a readout electrode is made deeper than the surface of the semiconductor substrate in a region of a photodetector in order to prevent that a potential barrier is caused at an end on the side of a readout part of a photodiode. CONSTITUTION:A solid-state image sensing element is constituted in the following way: a buried photodiode which is composed of a region 13, of the other conductivity type, formed inside a semiconductor layer 12, of one conductivity type, on the surface of a semiconductor substrate 11 and composed of a shallow semiconductor layer 16, of one conductivity type, on the surface of the region 13 of the other conductivity type is used as a photodetector; a readout electrode 18 used to read out a signal charge from the photodetector to a signal- charge transfer means 14 is formed on the semiconductor substrate via an insulating film 17. In this solid-state image sensing element, the surface of the semiconductor substrate in a readout region under the readout electrode 18 is made deeper than the surface of the semiconductor substrate in a region of the photodetector. For example, the surface of a semiconductor substrate in a readout region of a signal charge between an N-type region 13 constituting a photodiode and an N-type region 14 constituting a charge transfer means is made deeper than the surface of the semiconductor substrate in the photodiode part.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a solid-state image sensor using a buried photodiode as a light receiving element, and particularly to the structure of a signal charge readout portion.

[Conventional technology]

Generally, a PN junction photodiode is used as a light receiving element of a solid-state image sensor. In a PN junction photodiode, in order to reduce afterimages, the N-type region of the photodiode is depleted so that the N-type region of the photodiode is completely depleted when the signal charge is read out by applying a readout pulse voltage to the signal readout electrode. The impurity concentration is low. but,
When the surface of the N-type region is depleted, charges are generated by the pairing centers existing on the substrate surface, which increases a noise component called dark current that is not caused by photoelectric conversion, and reduces the S/N ratio of the solid-state image sensor. It has the disadvantage that it decreases.

As a method to reduce this dark current, a shallow, highly concentrated P-type layer is formed on the surface of the N-type region of the photodiode, and this potential is fixed to a reference potential to completely deplete the N-type region of the photodiode. Even in this case, it is very effective to use a buried photodiode as a light receiving element, which prevents depletion of the P-type layer on the surface of the substrate. FIG. 5 is a cross-sectional view of a unit cell showing an example of a solid-state imaging device using a conventional buried photodiode as a light receiving element.

In a P-type well 42 on an N-type semiconductor substrate 41, an N-type region 43 forming a photodiode, an N-type region 44 forming a charge transfer means, and a channel stop region 42 are formed. A polysilicon electrode 48 for reading and transferring signal charges is formed with an insulating film 47 interposed therebetween. Furthermore, a shallow P-type layer 46 is formed on the surface of the photodiode. Further, an aluminum light shielding film 49 is formed on the surface other than the photodiode.

[Problem to be solved by the invention]

In the above-mentioned solid-state imaging device using the conventional buried photodiode as a light receiving element, when a readout pulse voltage is applied to the polysilicon electrode 48 at the readout electrode, the channel of the photodiode and the channel under the readout electrode of the signal charge are A low potential area occurs between the two. therefore,
There is a problem in that when reading signal charges from the photodiode, it becomes a potential barrier and prevents reading of the signal charges, resulting in an increase in afterimages.

That is, FIG. 6(a) is an enlarged view of the vicinity of the charge readout section in FIG. 5, and the broken line in the figure indicates the polysilicon electrode 5.
4 is connected to the part of the channel where the potential is highest when a pulse voltage for reading signal charges is applied. Further, FIG. 6(b) illustrates the potential along the broken line in FIG. 6(a). As shown in FIG. 6, the N-type region 52 forming the photodiode is narrowed at the end on the readout side side due to the shallow P-type layer 55 on the photodiode surface, and the voltage of the polysilicon electrode 54 is reduced. A potential barrier 57 is generated in the bulk portion where it is difficult to be influenced by the signal charges 56, which prevents the signal charges 56 from being read.

[Means to solve the problem]

In the solid-state imaging device of the present invention, which uses an embedded photodiode as a light receiving element, the surface of the semiconductor substrate in the readout region under the readout electrode is deeper than the surface of the semiconductor substrate in the region of the light receiving element. The semiconductor surface of the readout region is characterized by having a depth that is approximately half the junction depth of the impurity region forming the photodiode.

In a conventional buried photodiode, when the signal charge is read out and the photodiode is depleted, the highest potential within the photodiode is at a depth near the middle of the impurity region that makes up the photodiode. A channel is generated on the surface of the semiconductor substrate in the signal readout region when a readout pulse voltage is applied to the electrode. However, according to the present invention, the depth of the semiconductor substrate surface in the signal readout region can be made approximately the same depth as the region where the potential of the depleted photodiode is highest. Therefore, the potential that is generated at the end of the photodetector on the readout side during signal readout, which causes image formation due to incomplete readout of signal charges, which has been a problem with conventional solid-state image sensors using embedded photodiodes as photodetectors. It becomes possible to prevent barriers from occurring.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a sectional view of a unit cell according to a first embodiment of the present invention, and corresponds to FIG. 4 showing a conventional example. In the P-type well 12 on the N-type semiconductor substrate 11, an N-type region 13 forming a photodiode and an N-type region 1 forming a charge transfer means.
4. Channel stop regions 15 are formed respectively. The surface of the semiconductor substrate in the signal charge readout region between the N-type region 13 forming the photodiode and the N-type region 140 forming the charge transfer means is formed to be deeper than the surface of the semiconductor substrate in the photodiode section. ing.

A shallow P-type layer 1 on the surface of the photodiode is formed by ion implantation using a polysilicon electrode 18 as a mask for reading and transferring signal charges.
6 is formed. The wood-flow side is a solid-state image sensor that uses a BCCD (buried channel type C0D) as a signal charge transfer means.

FIG. 2(a) is an enlarged view of the vicinity of the charge readout section in FIG. The high points are connected, and FIG. 2(b) illustrates the potential along the broken line in FIG. 2(a). As shown in Figure 2, the surface of the semiconductor substrate in the signal charge readout section is made deeper than the surface of the semiconductor substrate in the photodiode section, and the channel of the signal charge readout section is approximately the same as the highest potential part of the photodiode section. Make it deep. As a result, unlike the case in FIG. 6, the dashed line connecting the parts with the highest potential does not pass through the part where the N-type region forming the photodiode is narrowed at the end of the photodiode on the readout part side. Therefore, the potential barrier as shown in FIG. 6(b) is not generated, and as shown in FIG. 2(b), the signal charge 26 is completely read out to the N-type region 23 which forms the charge transfer means. Therefore, image retention due to incomplete reading of signal charges, which has been a problem in the past, does not occur.

It should be noted that if the surface of the semiconductor substrate in the readout section is made even slightly deeper, a corresponding effect can be obtained, but normally the depth is 1/3 to 2 of the junction depth of the N-type region forming the photodiode section.
A preferable effect can be obtained by setting the depth to /3.

Next, an example of a method for manufacturing the solid-state image sensor of the first embodiment described above is shown in FIGS. 3(a) to 3(i). Note that this example shows an example in which a silicon substrate is used as the semiconductor substrate.

First, as shown in FIG. 3(a), a P-type well 62 is formed on an N-type silicon substrate 61.

Next, as shown in FIG. 3(b), the silicon oxide film 63
After growing a silicon nitride film 64 in this order, using the photoresist 65 as a mask, a portion of the silicon nitride film that will become a signal charge readout region is etched and removed.

Subsequently, as shown in FIG. 3C, by performing thermal oxidation using the silicon nitride film 64 as an oxidation mask, the signal charge readout region is selectively oxidized, and the silicon substrate in the signal charge readout region is Make the surface deeper than other parts.

Next, as shown in FIG. 3(d), the silicon nitride film is etched using the photoresist pattern 66 as a mask, and then boron ions are implanted. Subsequently, thermal oxidation is performed to form a channel stop region 67 as shown in FIG. 3(e). After removing all the silicon nitride film, each N-type impurity is ion-implanted and impurities are thermally diffused using the thick silicon oxide film formed by selective oxidation as a mask, as shown in Figure 3 (f).
), an N-type region 68 forming a photodiode
Then, an N-type region 69 serving as a charge transfer means is formed.

Next, as shown in FIG. 3(g), using the photoresist pattern 70 as a mask, the silicon oxide film is removed only from the signal charge readout region. Subsequently, as shown in FIG. 3(h), the silicon oxide film is refreshed and oxidized to form a gate oxide film 71, and then a polysilicon electrode 72 for reading and transferring signal charges is formed. Thereafter, using this polysilicon electrode as a mask, ions of P type impurity are implanted to form a shallow P type layer 73 on the surface of the photodiode. Then, as shown in FIG. 3(i), an aluminum light-shielding film 75 is formed to form a solid-state image sensor using the embedded photodiode as a light-receiving element.

FIG. 4 is a sectional view of a unit cell according to a second embodiment of the present invention. This example is for an MO3 type solid-state image sensor, and N
In a P-type well 32 on a type semiconductor substrate 31, an N-type region 33 forming a photodiode and an N-type region 3 forming a drain are formed.
4. Channel stop regions 35 are formed respectively. The semiconductor surface 2 of the signal charge readout region between the N-type region 33 forming the photodiode and the N-type region 340 forming the drain is formed to be deeper than the semiconductor substrate surface of the photodiode portion.

A shallow P-type layer 36 is formed on the surface of the photodiode by ion implantation using a polysilicon electrode 38 as a mask for reading signal charges. In this embodiment, as in the first embodiment, no potential barrier is generated when the readout voltage is applied to the readout polysilicon electrode, and no afterimage is generated due to incomplete transfer.

Note that the method for manufacturing the solid-state image sensor of the second example is as follows:
Instead of forming an N-type region that serves as a charge transfer means in the first embodiment, an N-type region that serves as a drain is formed, and after forming a polysilicon electrode that reads signal charges, an N-type region that serves as a drain is formed. The device can be manufactured in the same manner as the first embodiment except for adding the step of forming a contact hole on the region and the step of forming an electrode forming a signal line.

〔Effect of the invention〕

As explained above, the present invention makes the surface of the semiconductor substrate in the readout region under the readout electrode deeper than the surface of the semiconductor substrate in the light receiving element region, and when reading signal charges, the position of the surface channel formed in the readout region is adjusted. By making the buried photodiode as deep as the highest potential part of the photodiode when it is depleted,
Preventing a potential barrier from occurring at the end of the photodiode on the readout side has the effect of preventing an increase in afterimages due to incomplete readout of signal charges.

[Brief explanation of the drawing]

FIG. 1 is a cross-sectional view of a unit cell of the first embodiment of the solid-state image sensing device of the present invention, FIG. 2(a) is an enlarged view of the vicinity of the signal charge readout section of FIG. 1, and FIG. Channel potential diagram of the part along the broken line in Fig. 2(a), Fig. 3(a) to (1)
is a sectional view showing the manufacturing method of the first embodiment, FIG. 4 is a sectional view of a unit cell of the second embodiment of the solid-state image sensor of the present invention,
FIG. 5 is a cross-sectional view of a unit cell of an example of a conventional solid-state image sensor, FIG. 6(a) is an enlarged view of the vicinity of the signal charge readout section in FIG. ) is a channel potential diagram of a portion along the broken line. 11.31, 41...N-type semiconductor substrate, 12°21.32,42,51...P-type well, 13
゜22.33,43,52...N-type region forming a photodiode, 14,23,44.53...
・N-type region forming charge transfer means, 15, 35.45...
...Channel stop area, 16, 25, 36, 4
6゜55...Shallow P-type layer on photodiode surface, 17.37.47...Insulating film, 18,24
゜3g, 48.54...Polysilicon electrode, 1
9゜40.49・・・Aluminum light shielding film, 26,58
...Signal charge, 57...Potential barrier, 3
4... N-type region forming the drain, 39...
...Electrodes forming signal lines. Agent Patent Attorney Otosuke Uchihara Page 1 No. 2121 Figure Figure Figure Figure

Claims (1)

[Claims] A buried photodiode comprising a region of another conductivity type formed in a semiconductor layer of one conductivity type on the surface of a semiconductor substrate and a shallow semiconductor layer of one conductivity type on the surface of the region of other conductivity type is used as a light receiving element,
In a solid-state image sensor in which a readout electrode for reading signal charges from the light receiving element to the signal charge transfer means is provided on a semiconductor substrate with an insulating film interposed therebetween, the surface of the semiconductor substrate in the readout area under the readout electrode is A solid-state imaging device characterized in that the area of the light receiving element is deeper than the surface of the semiconductor substrate.