JPH07105482B2 - Solid-state infrared image sensor - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a solid-state infrared image sensor that converts information of an infrared image into a time-series electric signal.

(Prior Art) A conventional solid-state infrared image sensor of this type is disclosed in Japanese Patent Application Laid-Open No. 63-237583 (Japanese Patent Application No. 62-73240), which is related to an infrared sensor forming a light receiving portion of the image sensor. 2 is presented as an example.

First, the operation principle of the infrared sensor that constitutes the light receiving portion of the image sensor will be described.

The energy band structure of the infrared sensor is shown in FIG.
It is as shown in (b). In this figure, (a) is the case where the first conductivity type is p-type and the second conductivity type is n type, and (b) is the case where the first conductivity type is n-type and the second conductivity type is p-type. . Although infrared rays are incident from the carrier injection region side, they may be incident from the photoelectric conversion region side. Potential barriers that hinder the majority carriers in the photoelectric conversion region 19 or 25 and the carrier injection region 21 or 27 are formed at the interface between the photoelectric conversion region and the potential barrier region and the interface between the carrier injection region and the potential barrier region. It In particular, the potential barriers for majority carriers in the photoelectric conversion regions 19 and 25 are Φ _H and Φ _E , respectively. The height of the potential barrier is from 0 to 0 depending on the conductivity type of the potential barrier region, the thickness of the potential barrier region, the balance of the impurity concentrations of the three regions, and the bias voltage applied between the photoelectric conversion region and the carrier injection region. Normal
It can be arbitrarily set with the upper limit being about the energy corresponding to the diffusion potential formed by the pn junction. If the dark current is large at room temperature due to the small potential barrier, cooling is necessary.

Infrared light whose energy is smaller than the semiconductor band gap E _g 35
(Or 36) is hardly absorbed in the carrier injection region 21 (or 27) and the potential barrier region 20 (or 26), is transmitted, and is incident on the photoelectric conversion region 19 (or 25).

When the first conductivity type in (a) is p-type and the second conductivity type is n-type, in the photoelectric conversion region 19, electrons below the Fermi level E _f in the valence band 22 are incident infrared rays. absorbs energy h _v of light 35, a transition from under the Fermi level E _f to the empty level between the Fermi level E _f and the valence band edge E _V, form a hot electrons 31 and hot holes 32 To do. These hot electrons 31 and hot holes 32 move in the photoelectric conversion region 19 until they are recombined. The hot hole 32 in motion is the potential barrier region.
20 and has more energy than the potential barrier Φ _H , the potential barrier region 20
There is a probability of being injected into the carrier injection region 21 after passing through. When the hot holes 32 are injected into the carrier injection region 21, the hot electrons 31 left in the photoelectric conversion region 19 and the hot holes 32 injected into the carrier injection region 21 become signal charges.

On the other hand, when (b) the first conductivity type is n-type and the second conductivity type is p-type, in the photoelectric conversion region 25, the Fermi level E _f in the conduction band 30 and the conduction band edge E _C are The electrons in between absorb the energy h _v of the incident infrared light 36 and make a transition from below the Fermi level to the vacant level above the Fermi level, forming hot electrons 33 and hot holes 34. These hot electrons 33 and hot holes 34 move in the photoelectric conversion region 25 until they are recombined, as in the case described above. The moving hot electrons 33 reach the potential barrier region 26, and if they have energy larger than the potential barrier Φ _E, they pass through the potential barrier region 26 and are injected into the carrier injection region 27. There is a probability. Hot electron 33 is carrier injection region 27
Injected into the photoelectric conversion region 25, hot holes 34 left in the photoelectric conversion region 25 and hot electrons injected into the carrier injection region 27.
33 and become the signal charge.

In FIGS. 2A and 2B, the photoelectric conversion region is drawn as a degenerate state, but even if this is a non-degenerate state, if the impurity concentration is significantly higher than that of the carrier injection region, it is used as a sensor. Operate.

Next, the solid-state infrared image sensor presented as the second embodiment in the specification of this infrared sensor will be described.

The structure of the solid-state infrared image sensor is shown in FIGS. 3 and 4. These are examples of a two-dimensional interline transfer CCD system. FIG. 3 is a sectional view of a unit pixel and FIG. 4 is an overall configuration diagram.

The unit pixel is composed of an infrared sensor 52, a transfer gate 53, and a vertical CCD 54. p-type single crystal Si substrate 40
A region of the infrared sensor portion in the state in which the substrate is produced is a carrier injection region 39, and a carrier injection region 39, that is, a p-type photoelectric conversion region 37 containing impurities at a higher concentration than the substrate is provided on the surface. . These photoelectric conversion areas
Potential barrier region between 37 and carrier injection region 39
38 has been formed. The potential barrier region 38 is a p-type having a lower impurity concentration than the carrier injection region 39, an intrinsic type, or an n-type having a fully depleted state at least under operating conditions. In order to reduce electric field concentration around the photoelectric conversion region 37 and the potential barrier region 38, an n-type card ring 41 is provided.
Has been applied. On the transfer gate 53 side of the n-type card ring 41, an n ⁺ -type region 42 containing a high concentration of donor impurities and serving as a source region when reading out optical signal charges is formed. Since the photoelectric conversion region 37 and the n ⁺ type region 42 are short-circuited,
A silicide film 43 is formed so as to make ohmic contact with both of them. Silicide is used because it has the advantage of excellent stability of characteristics,
You do not absolutely have to use this. It can also be replaced by a simple metal film or metal wiring. Further, if the contact portion of the photoelectric conversion region 37 with the n ⁺ type region 42 and the n ⁺ type region 42 both contain a high concentration of impurities so that a tunnel current can easily flow, the silicide film 43 is not provided. However, it can be short-circuited. The photoelectric conversion region 37 and the potential barrier region 38 are finally connected to an electric drop by the silicide film 43, the n ⁺ -type region 42, and the n-type guard ring 41, and transition between energy bands occurs to conduct the potential barrier region 38. Even when electrons are accumulated in the band, they are quickly removed.

The structure of the vertical CCD 54 is a buried channel type. The poly-Si electrode 47, which is the transfer electrode, is
A thermal oxide film 46 is provided from the Si substrate to the n-type channel region 45. In order to limit the channel width in the lateral direction with respect to the charge transfer direction, the thermal oxide film 46 is thickened at both end portions of the vertical CCD 54, and the p ⁺ type channel block region 44 is provided. Since the cross section of the portion including the transfer gate 53 is shown in FIG. 3, the transfer gate 53 occupies the thick oxide film and the p ⁺ -type channel block region at the end of the vertical CCD 54 on the infrared sensor 52 side. There is only a small area, and a thick oxide film and a p ⁺ type channel blocking area are provided at both ends of the vertical CCD 54 in the area without it.

The entire surface of this image sensor is made of silicon oxide (SiO, SiO ₂ ) or silicon nitride SiN, which is formed by the CVD method.
It is covered with an insulating film 48 made of Si ₃ N ₄ ). A metal reflection film 49 made of aluminum or the like is provided on a portion of the insulating film 48 facing the photoelectric conversion region 37. This is for reusing the infrared light that has passed through the photoelectric conversion region 37, and is the part of the carrier injection region 39 and the potential barrier region 38 of the single crystal Si substrate 40 / the photoelectric conversion region 37 / insulating film 48 / metal. Since the optical resonance state occurs in the multilayer structure of the reflection film 49, the insulating film 48 is arranged so that the antinode of the standing wave generated by the infrared light of the center wavelength in the wavelength band used by the sensor is located near the photoelectric conversion region 37. If you set the thickness of, you will get high performance.

Further, the entire front surface thereof is covered with the protective film 50, and the entire rear surface thereof is covered with the antireflection film 51.

The overall structure of the image sensor is as shown in FIG.
This is an interline transfer system, in which unit pixels are arranged in a two-dimensional array, and vertical and horizontal scanning are performed with vertical CCD54 and horizontal CCD56, respectively.
It is supposed to be done with and. Vertical CCD54 is 4 phase drive,
The horizontal CCD 56 is a two-phase drive and has a structure in which a signal is output from the horizontal CCD via an output unit 57 to the outside.

The operation of this two-dimensional infrared CCD image sensor is as follows.

The photoelectric conversion is performed by setting the infrared sensor to the accumulation mode. In the accumulation mode, the transfer gate 53 is in the OFF state and the infrared sensor 52 is in the biased state as shown in FIG. 2 (a). When the infrared light 55 incident from the back surface of the infrared sensor 52 reaches the photoelectric conversion area 37 and is photoelectrically converted, the electrons which are the signal charges generated are converted into the photoelectric conversion area 3
7, accumulated in the n-type guard link 41, the n ⁺ -type region 42, and the silicide film 43. The vertical CCD 54 reads out signals during a period in which signal charges are accumulated in the photoelectric conversion region 37 and the other three portions. After accumulating the signal charges for a certain accumulation time, the transfer gate 53 is turned on, and the signal charges accumulated in the photoelectric conversion region and the other three parts are read out to the vertical CCD 54. After that, the transfer gate 53 is turned off, and the photoelectric conversion region and the other three parts start to accumulate the signal charges again.

The signal charge read to the vertical CCD 54 is transferred to the horizontal CCD 56 by one horizontal line in one horizontal period, and the horizontal C
It is read out from the CD 56 via the output unit 57. Output part 57
Is composed of a capacitor and a source follower amplifier, and performs signal potential conversion and impedance conversion on the signal charge and outputs it in the form of voltage change. This readout of signals for one horizontal line is repeated every horizontal period, and the signals of all pixels are read out during the signal charge accumulation period of the photoelectric conversion region 37 and the other three parts.

In the above example, the first conductivity type is described as p-type and the second conductivity type is described as n-type, but the same holds true, and in that case, pn in FIG. 3 are all replaced. This example is a front-illuminated image sensor, but a front-illuminated image sensor can also be realized, in which case the metal reflection film 49 and the antireflection film 51 are not required. Also, here, only the two-dimensional interline transfer CCD system has been described in detail,
As a conventional technique, there is another signal reading method such as a MOS method or a one-dimensional array method. Furthermore, this type of solid-state infrared image sensor is also feasible if semiconductor materials other than Si are used as materials.

(Problems to be Solved by the Invention) In the photoelectric conversion region of the infrared sensor that constitutes the light receiving unit, a high concentration and thin impurity distribution is required in consideration of infrared absorption and emission efficiency of hot carriers to the carrier injection region. To be

The reason is as follows.

Since the infrared absorption coefficient of free carriers increases in proportion to the carrier concentration in the semiconductor, the absorption of infrared rays can be increased by increasing the impurity concentration. On the other hand, regarding the emission efficiency of hot carriers to the carrier injection region, by making the thickness of the photoelectric conversion region thin, multiple reflection of hot carriers occurs between the insulator interface and the potential barrier region interface, and the life ends. By increasing the number of times hot carriers are incident on the resurface of the potential barrier region, the number will increase. Therefore, if the photoelectric conversion region is provided with a high concentration and thin impurity distribution, a sensor with high sensitivity can be obtained. (However, if the impurity concentration is extremely high, the crystallinity will be impaired and the life of the hot carriers will be shortened, so there is an optimum concentration even if it is said to be high concentration. Therefore, there is an optimum thickness in consideration of the infrared absorption and the efficiency of hot carrier emission to the carrier injection region.) An impurity that is easy to form such an impurity distribution as a property peculiar to semiconductor materials. Alternatively, the conductivity type is fixed. Si and InP (Indium-Phosphorus) are typical examples of semiconductors whose conductivity type is n-type and which easily form an impurity region with a high concentration and a shallow junction depth. Impurities suitable for this application are:
The case of Si is As (arsenic), and the case of InP is Se (selenium). When the above-mentioned infrared image sensor is formed using such a semiconductor material as a substrate, the infrared sensor of the light receiving portion can have higher performance when the first conductivity type is n-type and the second conductivity type is p-type. However, in the case of the infrared sensor with this structure, the transferred optical signal charges become holes, which causes the transfer efficiency to deteriorate in the CCD method and the response speed in the MOS method compared to the case where the transferred charges are electrons. Will lead to a decrease in That is, in the conventional solid-state infrared image sensor, when a semiconductor of n-type conductivity is used as a material, it is easy to form an impurity region having a high concentration and a shallow junction depth,
There is a drawback in that it is difficult to form a light receiving part having high performance and excellent charge transfer capability.

(Means for Solving the Problem) In order to solve the above-mentioned problems, a solid-state infrared image sensor of the present invention includes a photoelectric conversion region made of a first conductivity type semiconductor, and a first impurity concentration lower than that of the photoelectric conversion region. A carrier injection region made of a conductivity type semiconductor and a first conductivity type semiconductor existing between the photoelectric conversion region and the carrier injection region and having an impurity concentration lower than that of the carrier injection region, an intrinsic semiconductor, or at least an operating condition. An infrared sensor having a potential barrier region made of a second conductivity type semiconductor that is in a fully depleted state at 1, and a photoelectric conversion region, a potential barrier region, and a carrier injection region that form a homojunction structure. Alternatively, a two-dimensionally arranged infrared sensor array and a charge generated by photoelectric conversion in the infrared sensor array are time-based. In a solid-state infrared image sensor having an electronic scanning circuit for reading out as a column signal to the outside, the first conductivity type is n type, the second conductivity type is p type, and the carrier injection region has an n well structure. It is characterized in that the optical signal charge in the infrared sensor is read out from the carrier injection region.

(Operation) In the solid-state infrared image sensor of the present invention, the first conductivity type is n
Type and the second conductivity type are p-type, the infrared sensor of the light receiving part has high performance when a semiconductor of n-type conductivity is used as a material, which easily forms an impurity region having a high concentration and a shallow junction depth. Become. However, the carrier injection region has an n-well structure,
Since the optical signal charge in each infrared sensor is read from the carrier injection region, the transfer charge is an electron. Therefore, since the charge transfer capability is excellent, the above-mentioned problems can be solved.

(Example) Next, the Example of this invention is described with reference to drawings.

FIG. 1 is a vertical sectional structural view of a unit pixel according to an embodiment of the present invention.

The unit pixel is composed of the infrared sensor 15, the transfer gate 16, and the vertical CCD 17, as in the conventional case. The region of the well structure formed in the p-type single crystal Si substrate 4 of the infrared sensor 15 is the carrier injection region 3. An n-type photoelectric conversion region 1 containing impurities at a higher concentration than the carrier injection region 3 is provided on the surface. A potential barrier region 2 is formed between the photoelectric conversion region 1 and the carrier injection region 3. The potential barrier region is n-type having an impurity concentration lower than that of the carrier injection region 3, is intrinsic, or is at least fully depleted under operating conditions p
It is a type. Photoelectric conversion region 1 and potential barrier region 2
Since the periphery of mitigating electric field concentration, but p ⁺ -type guard ring 5 is applied, the p ⁺ -type guard ring 5, the vertical
The p ⁺ -type channel block region is provided to limit the channel width of the n-type channel region, which is in the well of the carrier injection region 3 on the side with the CCD 17 and reaches the outside of the carrier injection region 3 on the side without the vertical CCD 17. It overlaps with 7. The transfer gate 16 side of the carrier injection region 3 becomes a source region when reading out optical signal charges. Since the photoelectric conversion region 1 and the p ⁺ -type guard ring 5 are short-circuited, the silicide film 6 is formed so as to make ohmic contact with both of them. The reason why the silicide is used is the same as the conventional one, and it can be replaced by a simple metal film or metal wiring. Further, the contact portion of the photoelectric conversion region 1 with the p ⁺ -type guard ring 5 and the p ⁺ -type guard ring 5 are
If both of them contain a high concentration of impurities so that the tunnel current can easily flow, a short-circuit state can be achieved without providing the silicide film 6. The photoelectric conversion region 1 and the potential barrier region 2 are connected to the p-type Si substrate 4 by the silicide film 6, the p ⁺ -type guard ring 5, and the p ⁺ -type channel blocking region 7.

Structure of vertical CCD 17, insulating film 11, metal reflective film 12, protective film 1
Since 3 and the antireflection film 14 are the same as the conventional ones, the description thereof will be omitted.

The overall structure of the image sensor is the same as that shown in FIG. That is, the entire configuration is an interline transfer system, in which unit pixels are two-dimensionally arranged and arranged, and vertical / horizontal scanning is performed by the vertical CCD 17 and the horizontal CCD 56, respectively. The vertical CCD 17 is a four-phase drive, the horizontal CCD 56 is a two-phase drive, and the horizontal CCD 56 has a structure in which a signal is output to the outside via an output unit 57.

The difference in operation between this two-dimensional infrared CCD image sensor and the conventional one is shown below.

The infrared sensor 15 is operated in the accumulation mode as in the conventional case, but at that time, the infrared sensor 15 is in a biased state as shown in FIG. 2 (b). The infrared light 18 incident from the back surface of the infrared sensor 15 reaches the photoelectric conversion region 1, and when photoelectrically converted, hot electrons are injected into the carrier injection region 3.
And are left in the photoelectric conversion region 1. However, in the present invention, the photoelectric conversion region 1 has the silicide film 6,
Since the p ⁺ type guard ring 5 and the p ⁺ type channel blocking region 7 are connected to the p type Si substrate 4, the holes escape to the p type Si substrate 4. On the other hand, the hot electrons emitted to the carrier injection region 3 are accumulated therein because the transfer gate 16 is in the OFF state and the carrier injection region 3 is in the float state. After accumulating the signal charges for a fixed accumulation time, the transfer gate 16 is turned on, and the signal charges (electrons) accumulated in the carrier injection region 3 are read out to the vertical CCD 17. The read signal charge is transferred by the same operation as that of the conventional infrared image sensor, and is output from the output unit 57 as a time series signal.

As described above, in the infrared image sensor of the present invention, since the transfer charge can be converted into an electron in the n-type photoelectric conversion region, a light receiving unit having high performance and excellent charge transfer capability can be used. Can be provided.

In this embodiment, the short circuit between the photoelectric conversion region 1 and the potential barrier region 2 is due to the silicide film 6 and the p ⁺ type card ring 5, and the photoelectric conversion region 1 and the potential barrier region 2 and the p type Si. The connection to the substrate 4 (ground) is made by the structure in which the p ⁺ type guard ring 5 and the p ⁺ type channel block region 7 overlap each other, but another structure is also possible. As another structural example, (i) the p ⁺ type guard ring 5 and the p ⁺ type channel blocking region 7 are separated from each other, and metal wiring is provided from the p ⁺ type guard ring 5 to the ground electrode, or The p ⁺ -type guard ring 5 and the p ⁺ -type channel block region 7 are connected by a metal wiring, and the other portions have the same structure. (ii) The guard ring is a p-type that does not have a high impurity concentration, and There is a high impurity concentration p ⁺ -type region at the position where it contacts the guard ring, and this p ⁺ -type region and the photoelectric conversion region are connected by metal wiring, and the metal wiring is led to the ground electrode or p Examples include a structure connected to the ⁺ type channel block region.

Further, also in the infrared image sensor of the present invention, it is possible to realize a front-illuminated image sensor,
It is the same as the prior art that it includes another signal reading method such as MOS method and one-dimensional array.

(Effects of the Invention) As described above, according to the present invention, when a semiconductor of n-type conductivity type, such as Si or InP, which easily forms an impurity region having a high concentration and a shallow junction depth, is used as a substrate material. In addition, there is an effect that it is possible to form an infrared image sensor having a high performance light receiving portion and an excellent charge transfer capability.

[Brief description of drawings]

FIG. 1 is a vertical sectional structural view of a unit pixel according to an embodiment of the present invention. FIGS. 2 (a) and 2 (b) are explanatory views of the energy band structure of the light receiving portion and the photoelectric conversion mechanism in the infrared image sensor of the related art or the present invention, and FIG. When the conductivity type is n-type, (b) is the case where the first conductivity type is n-type and the second conductivity type is p-type. FIG. 3 is a vertical sectional structural view of a unit pixel of a conventional infrared image sensor of this type. FIG. 4 is an overall configuration diagram of an infrared image sensor of the related art or the present invention. 1,25 …… Photoelectric conversion region (n type), 2,20,26,38 …… Potential barrier region, 3,27 …… Carrier injection region (n type),
4,40 …… p type single crystal Si substrate, 5 …… P ⁺ type guard ring,
6,43 …… Silicide film, 7,44 …… p ⁺ type channel blocking region, 8,45 …… n type channel region, 9,46 …… Thermal oxide film, 1
0,47 …… Poly-Si electrode, 11,48 …… Insulating film, 12,49 …… Metal reflective film, 13,50 …… Protective film, 14,51 …… Anti-reflection film, 15,5
2 …… Infrared sensor, 16,53 …… Transfer gate, 1
7,54 …… Vertical CCD, 18,35,36,55 …… Infrared light, 19,37 ……
Photoelectric conversion region (p-type), 21,39 ... Carrier injection region (p-type), 22,28 ... Valence band, 23,29 ... Forbidden band, 24,3
0 …… conduction band, 31,33 …… hot electron, 32,34 …… hot hole, 41 …… n-type guard ring, 42 …… n ⁺ type region, 56
…… Horizontal CCD, 57 …… Output section.

Claims (1)

[Claims] 1. A photoelectric conversion region made of a first conductivity type semiconductor, a carrier injection region made of a first conductivity type semiconductor having an impurity concentration lower than that of the photoelectric conversion region, and between the photoelectric conversion region and the carrier injection region. A first conductivity type semiconductor which is present and has an impurity concentration lower than that of the carrier injection region, an intrinsic semiconductor, or at least a potential barrier region composed of a second conductivity type semiconductor which is completely depleted under operating conditions, An infrared sensor constituted by the photoelectric conversion region, the potential barrier region, and the carrier injection region forming a homojunction structure is generated by photoelectric conversion in the infrared sensor array arranged one-dimensionally or two-dimensionally. In a solid-state infrared image sensor equipped with an electronic scanning circuit for reading out electric charges as a time series signal to the outside A first conductivity type is n-type, a second conductivity type is p-type, the carrier injection region has an n-well structure, and optical signal charges in each infrared sensor are read from the carrier injection region. Infrared image sensor.