JPH0657056B2 - Solid-state image sensor - Google Patents

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a solid-state image sensor used for a visual sensor in a broad sense such as a video camera.

Configuration of conventional example and its problems In recent years, with the progress of semiconductor integrated circuit technology, a photoelectric conversion function, a storage function, and a reading function are provided on a single semiconductor substrate for an imaging tube that has been used in a conventional video camera or the like. With the integrated semiconductor solid-state image sensor being put to practical use, it is rapidly applied not only for the video camera but also as an image signal input device for industrial measurement or a visual sensor by taking advantage of its reliability and stability. The field is expanding.

Several types of configuration methods have been developed so far for the solid-state image pickup device in relation to the photoelectric conversion unit and the horizontal and vertical readout methods. That is, the photoelectric conversion unit can be a photodiode having a PN junction, a MOS diode, or a composite configuration thereof. Further, the horizontal shift register includes a CCD shift register or a MOS dynamic shift register. For vertical reading, a vertical CCD shift register, a so-called MOS type vertical reading by a combination of a vertical signal transmission line and a dynamic shift register, and a switch element. The configurations are combined. However, if the horizontal shift register is a MOS shift register,
Only examples are known in which the type read configuration is applied.

In any of such conventional solid-state image pickup devices, the occurrence of an abnormal signal called a vertical smear signal is widely known.
This vertical smear signal is a so-called stray charge excited in the area other than the charge storage section or in the inter-pixel area due to a strong optical input during the non-readout period, and leaks to the vertical read-out mechanism, which causes vertical vertical movement of the high-luminance image on the image plane. This is a phenomenon in which a signal appears in the direction.

That is, the phenomenon of vertical smear in a solid-state image sensor is
In the region where there is no internal electric field between many photoelectric conversion elements,
Due to undesired diffusion of the optically excited signal charge. In particular, for excitation of long-wavelength light, the thickness of the semiconductor required to absorb it becomes significantly larger for short-wavelength light, so it can be ignored compared to the spacing of the photoelectric conversion elements. The excitation component in the deep portion of the substrate at a non-existent depth becomes dominant, and those signal charges are not accommodated only in the storage portion of the closest photoelectric conversion element, and reach the neighboring photoelectric conversion elements, Not only the smear signal but also the deterioration of the resolution characteristic.

Therefore, this phenomenon occurs even when the charge storage unit is in an unsaturated signal level state. An optical surface light-shielding structure of an element should be adopted in which a region where an internal electric field does not exist between photoelectric conversion elements is reduced and light input to the region is suppressed as much as possible. Generally, the region without electric field in the photoelectric conversion two-dimensional plane and the deep region in the thickness direction of the semiconductor substrate are the so-called, Overflow due to blooming phenomenon, or a mechanism for discharging overflowed charges-an overflow drain is provided, and is usually minimized. It goes without saying that the signal charge, which causes the smear phenomenon, is also reduced by the "discharging mechanism". However, in general, it is equivalent to discarding the signal charges generated between the photoelectric conversion elements having no electric field, and does not fundamentally suppress the smear phenomenon.

FIG. 1 shows that the horizontal shift register 1 is composed of a CCD (charge coupled device) in the above solid-state image sensor configuration example.
The vertical shift register 2 is composed of a dynamic shift register, and the photoelectric conversion element is formed of a switching transistor 5 corresponding to the photodiode 4 and a horizontal selection line 6
Thus, the signal charge of the selected photoelectric conversion element 4 is drawn out to the vertical transmission line 7, passes through the vertical conversion unit 3, and passes through the horizontal CC.
An example of the configuration in which the data is transferred to D1 and output by the charge detector 8 is shown.

When a signal charge exceeding the storage capacity thereof is applied to the photodiode 4 of the photoelectric conversion element by strong photoexcitation, the signal charge overflowing before the reading level of the corresponding reading transistor 5 is exceeded, The structure of the transistor 9 to be overflowed and discharged at a level higher than the cut-off level on the reading side is attached as shown in FIG. 2, and the structure for discharging the overflowed charges through the discharge line 10 is a two-dimensional surface of imaging. May be configured within. In addition, the direction perpendicular to the two-dimensional plane, that is,
It is also well known that the surface of the main part of the image pickup is reverse to the conductivity polarity of the semiconductor substrate, and a so-called “floor raising structure” is used to form a three-dimensional structure to discharge the overflowing charges.

FIG. 3 shows a cross section of the photoelectric conversion main part of the solid-state image pickup device in the above-mentioned so-called MOS type. However, there is no so-called overflow signal charge discharging mechanism due to the excessive light input. This is because the smear phenomenon at the unsaturated light input level is not suppressed without cutting off all the excitation signal charges.

In the photoelectric conversion main part according to the conventional example, the drain diffusion layer 14 of FIG. 3 corresponding to the drain of the readout transistor 5 of FIG. 2 is not gated with the diffusion layer 13 of the photoelectric conversion photodiode during the non-reading period. Is in the "off" state and is electrically cut off. However, the non-readout period,
Since the potential of the diffusion layer is at a high level, the signal charges excited in the deep portion of the semiconductor substrate 11 and the signal charges 20 generated by the input light 19 obliquely incident on the element surface are stored in the diffusion layer 13. The probability of being absorbed is almost equal to the probability of being absorbed by the drain diffusion layer 14 at a high level and forming a smear signal. In order to prevent this, it is effective to increase the light shielding area 15 and reduce the opening window portion 18, but it is not compatible with maintaining the sensitivity of the image sensor.

SUMMARY OF THE INVENTION In view of the above-mentioned drawbacks, the present invention provides a solid-state imaging device that does not sacrifice smear signals in terms of sensitivity.

Configuration of the Invention The solid-state imaging device of the present invention includes a diffusion region of the opposite conductivity type to the substrate formed in at least one region in the signal charge storage portion in the photoelectric conversion main portion, and a channel or diffusion region forming the readout portion. The diffusion region of the signal charge accumulating portion is characterized in that the gate portion and the drain diffusion region of the reading portion are surrounded by at least a part of the semiconductor substrate inside a region of the same conductivity type as the semiconductor substrate. is doing.

Description of Embodiments FIG. 4 shows an embodiment of the present invention. In the conventional example as shown in FIG. 3, the signal charge 20 by photoexcitation in the deep portion of the semiconductor substrate is obtained.
However, in order to prevent the so-called vertical smear signal from being absorbed by the drain 14 of the read transistor during non-reading, in the present invention, the storage section 13 is provided between the substrate section 22 for the read section transistor and the substrate 11. A potential region 21 that is connected to one end and has the same potential as the signal charge storage portion 13 or higher potential than the substrate is provided. This potential area 2
1 may be a carrier concentration that is depleted at the time when the signal charges in the signal charge storage unit are read out. Further, the region 21 enclosing the reading section does not necessarily need to completely enclose the reading section drain 14 with respect to the substrate 11, and the substrate section 22 of the reading section transistor is partially connected to the substrate 11, as shown in FIG. As shown, the same purpose can be achieved when connected to the channelization blocking region 12. Furthermore, in general, the areas 2 of the substrates 11 and 22 are
1 does not necessarily have to be of the opposite conductivity type to them,
When the signal charge 20 due to the excitation is an electron, the conductivity type and the impurity concentration which give a higher potential than the substrate 22 and lead to the signal charge accumulating portion 13 are sufficient. For example, the substrate 11 is P ^+.
Alternatively, in P, the substrate 22 may be P, and 21 may be P ⁻ or a linear region. It is only necessary that the impurity concentration of the region 21 is higher than that of the substrate portion 22 and is equal to or lower than that of the signal charge storage portion 13 after the signal charge is read out.

The semiconductor substrate 11 is usually P-type (100) of 5 to 10 Ω · cm.
A crystal surface is used. However, it suffices that a portion of the surface of the crystal having a thickness of 5 to 15 μm has this characteristic, and as a semiconductor substrate body, a high concentration of P
^It may be ⁺ or the opposite conductivity type. In that case,
Appropriate bias voltage is often applied. The thickness of the substrate 22 of the read-out transistor has an average carrier concentration of 1
When it is set to about 0 Ω · cm, it is desirable to adjust the photoelectric conversion operation of the signal charge storage unit to about 12 μm if a maximum of about 12 V is required.

According to the present invention, in the photoelectric conversion main part of the solid-state imaging device, the charges generated between the photoelectric conversion parts, in the vicinity of the reading part adjacent to the photoelectric conversion part, and in the deep part of the substrate directly enter the reading part and generate a false signal. It has an effect of remarkably preventing the formation. Even if the charges are generated in the lower part of the reading part due to the oblique incident light or the charges are generated between the photoelectric conversion elements, they are accommodated in any one of the signal charge accumulating parts and thus are directly absorbed in the reading part. The effect is significantly reduced.

Therefore, the present invention is extremely useful as it provides a structure that fundamentally solves the vertical smear phenomenon, which is the greatest drawback of the solid-state imaging device.

[Brief description of drawings]

FIG. 1 is a circuit diagram showing a typical configuration example of a solid-state image pickup device, FIG. 2 is an equivalent circuit diagram of a photoelectric conversion element, FIG. 3 is a configuration sectional view of a photoelectric conversion main part in a conventional example, and FIG. FIG. 5 is a schematic view of another photoelectric conversion main part according to the present invention. 21 ... Read-out transistor envelope diffusion area, 22 ... Read-out transistor substrate part.

Claims (1)

[Claims] 1. A signal charge storage portion is formed on a surface of a semiconductor substrate by PN.
A signal composed of a pair of a gate structure and a drain diffusion region, which forms a switch part adjacent to the photoelectric conversion part formed of a junction or a metal monoxide-semiconductor structure or a combination thereof, and the signal charge storage part. In a solid-state imaging device having a two-dimensional array of image pickup elements with a charge reading unit, a potential higher than that of the reading unit is given to a signal charge between the switch unit and a semiconductor substrate, and one end of the signal charge storage unit is provided. A solid-state image pickup device having a region connected to.