JPS5822899B2 - solid-state imaging device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to improvement of a light receiving device or a solid-state imaging device manufactured on a semiconductor single crystal substrate.

The imaging device conventionally used is an imaging tube of the type that operates in an accumulation mode and scans a photoconductive target with an electron beam.

In this case, since an electron beam is used, there are drawbacks such as the need for high voltage and the difficulty of miniaturization.

Solid-state imaging devices or imaging plates have been devised to overcome these difficulties.

FIG. 1 shows the principle of a solid-state imaging device.

Each picture element 4 is arranged in a grid pattern and read out one by one using the XY addressing method.

Selection of picture elements is performed by a horizontal scanning signal generator 1 and a vertical scanning signal generator 2.

3 is a switch section connected to each picture element, and 5 is an output end.

Specific configurations of the photoelectric conversion section of each picture element include an example in which a diffusion region is directly formed on the SL substrate, and an example in which a photoconductive thin film is used.

However, the example in which the photoelectric conversion section is formed by forming a diffusion region on the 81 substrate has the following serious disadvantages in practical use.

In order for a solid-state imaging device to have a resolution comparable to that of an image pickup tube, it is required to have 500×400 or more picture elements.

It is difficult to reduce the size of the picture element to 20 μm x 20 μm or less even at its smallest size, so the size of the Si substrate is 1c.
Super large IC of about rrt×lCrl1 or more
becomes.

As the chip size increases, the running distance decreases.

The disadvantage in terms of cost is just a puppy.

Furthermore, each picture element is generally composed of a photoelectric conversion section (corresponding to the diffusion region to the Si substrate) using a source region of a MOS switch, and this two-dimensionally arranged M
The OS switch occupies a considerable area, which is not a good idea as a configuration of the light receiving surface.

Furthermore, since the wiring lines running vertically and horizontally also run on the diode surface, the incident area of light decreases.

These lead to a decrease in optical sensitivity and a decrease in signal output, resulting in a decrease in signal-to-noise ratio (S/N ratio).

On the other hand, in an example using a photoconductive thin film, a scanning circuit for performing XY addressing such as the MOS switch is formed on a Si substrate, and a photoconductive thin film is arranged three-dimensionally on the upper layer to form a light receiving section. It is something to do.

An example of such a solid-state imaging device is disclosed in Japanese Patent Laid-Open No. 51-95720.

FIG. 2 shows a sectional view for explaining the principle.

Diffusion regions 7 and 8 are provided in the 81 substrate 6 to serve as the source and drain of the MOS switch.

10 is a gate electrode of a MOS switch, 5 is a drain electrode for taking out a signal, and 16 is a source electrode.

A photoconductive thin film 17 and a transparent electrode 18 are formed on the switch circuit thus configured.

Note that 13 is an insulating layer.

The electrode 16 is, for example, Sb2S3 J CdS +As2
A capacitance C corresponding to the area S of the electrode is formed between the photoconductive thin film 1γ and the transparent conductive thin film 18 made of a photoconductive substance such as Se3+polycrystalline Si.

Since the electrode patterns are separated and placed in a matrix, the capacitors are arranged in a matrix.

Since this capacitor has a photoconductive thin film in between, it functions as a photosensitive element and forms a picture element.

If a photosensitive element is represented by an equivalent circuit, it can be represented by a parallel connection of a variable resistor R and a capacitor C whose electrical resistance changes depending on the intensity of light.

The size of the capacitance C is determined by the electrode area S, the thickness t of the photoconductive thin film 11, and the dielectric constant ε, and is expressed as C=5°8.

Further, the magnitude of the resistance is inversely proportional to the intensity of light incident on the electrode surface at that position, and if no light is applied at all, it will generally be R-■, although it depends on the type of photoconductive thin film.

Since a target voltage (VT) is applied to the transparent electrode 18, the capacitance of the portion that is not exposed to light during one field maintains the same voltage VT.

Since the resistance R of the portion exposed to light decreases in accordance with the intensity of the light, the charge stored in the capacitor C is discharged, and the voltage held in the capacitor decreases in proportion to the amount of light.

(2) If the voltage remaining after discharging during the field period is UT, a charging current corresponding to the voltage VU-UT flows, and when charging is completed, the voltage is raised to the target voltage again.

The charging current at this time becomes a video signal corresponding to this field.

In such a solid-state imaging device, imaging characteristics such as spectral sensitivity characteristics, resolution, signal-to-noise ratio, and afterimage characteristics are naturally important, but for this reason, the heat resistance, mechanical strength, etc. of the photoconductive thin film are extremely important.

In other words, it is necessary to attach a transparent electrode after attaching a photoconductive thin film on the Si substrate, but as a transparent electrode, 5n02
(Suzunesa) as a transparent electrode at 400-500°C.

Even when using NESA, it was necessary to heat the substrate to around 250°C.

This is the reason why the photoconductive film is required to have heat resistance.

A translucent metal thin film can also be used instead of a transparent electrode.
In this case, there is no need to heat the substrate.

However, due to the reflection and absorption by the metal thin film, the photosensitivity, which is an important characteristic for imaging, decreases significantly, which is undesirable.

This is particularly a problem in the imaging device having the structure shown in FIG.

That is, in the case of an imaging target for a normal image pickup tube, first a NESA electrode is attached on a glass face plate, and then a photoconductive thin film is attached, so the presence or absence of heat resistance does not matter, at least in the manufacturing process.

Mechanical strength is also important.

After applying the photoconductive thin film, operations such as attaching a Nesa electrode and, in the case of a color imaging plate, attaching a filter are required, and mechanical strength is required from the viewpoint of ease of handling.

Further, the photoconductive thin film has a specific resistance of 1010Ω, Crn.

or more is required.

This is because it is necessary that the charge pattern does not diffuse and disappear within the time interval during which a specific picture element is scanned, that is, the accumulation time.

When polycrystalline Si is used for a photoconductive thin film, it has a particularly low resistivity and must be divided into mosaic shapes, which complicates the process and lowers the yield.

Furthermore, photoconductive thin films of Sb2S3 and A8□Se have problems in mechanical strength and heat resistance, and are not suitable for practical use in the imaging device having the structure shown in FIG.

The present invention solves the drawbacks of the prior art described above.

The basic structure is similar to that illustrated in FIG. 2, in which a four-scanning circuit and the like are formed on a Si substrate, and a photoconductive thin film is placed on top of this.

Particularly, the present invention is characterized by the use of an amorphous material mainly composed of silicon containing hydrogen as the photoconductive thin film.

Particularly 50 atomic % or more of silicon. Preferred is an amorphous material containing carbon and 10 atomic % to 50 atomic % hydrogen.

In this case, a portion of silicon in the amorphous material can be replaced with at least a portion of germanium or carbon, which are homologous elements.

The amount of replacement is limited to 60% of silicon.

A film thickness of 0.05 μm or more is used.

In practice, it is common to select from the range of 0.2 μm to 4 μm.

Furthermore, the thin film may be multilayered, or the composition may be continuously changed.

In this way, an amorphous film containing silicon and hydrogen at the same time is an excellent material that can easily be made to have a high specific resistance of 1010 Ω, crrL or more, and has very few trap levels that hinder the movement of carriers.

The photoconductive material of the present invention can be manufactured by various methods.

Typical examples will be explained below. The first method is a reactive sputtering method.

FIG. 3 shows a schematic diagram of an apparatus used in the reactive sputtering method.

The device itself is a general sputtering device.

101 is a container that can be evacuated, 102 is a spuck target, 103 is a sample substrate, 104 is a shutter, 10
5 is an input from a high-frequency oscillator for sputtering, 106 is a heater for heating the substrate, 10γ is a water-cooled tube for cooling the substrate, 108 is a high-purity hydrogen inlet, 109 is an argon or other gas inlet, 1
10 is a gas reservoir, 111 is a pressure gauge, 112 is a vacuum gauge, 11
3 is a connection port to the exhaust system.

A target for sputtering may be one cut from molten silicon.

Further, in the case of an amorphous material containing silicon, germanium, or carbon, a target that is a combination of these three group (I) elements is used.

In this case, it is convenient to use a thin piece of graphite, germanium, or the like as a target on a silicon substrate, for example.

The composition of the amorphous material can be controlled by appropriately selecting the area ratio of silicon to germanium or carbon.

Of course, conversely, for example, a silicon thin piece may be provided on a carbon substrate.

Furthermore, both materials may be juxtaposed to form a target, or a melt of the same composition may be used.

Also, as a target for sputtering, for example,
Si containing phosphorus (P), arsenic (As), borosilicate (B), etc.
By using these elements, it is possible to introduce these elements as impurities.

By this method, an amorphous material of any conductivity type such as n-type or p-type can be obtained.

Further, by doping with such impurities, the resistance value of the material can be changed.

It is possible to achieve a high resistance of ~1013Ω in one step.

Note that such impurity doping can also be carried out by mixing diborane or phosphine into the rare gas.

Using the apparatus as described above, 30 mol% of hydrogen (H,2)
Spackling Si and graphite by generating a high single frequency discharge in an Ar atmosphere containing the following various mixing ratios,
Thin layers can be obtained by depositing this on a substrate.

In this case, the pressure of the Ar atmosphere containing hydrogen may be within any range that can maintain glow discharge, and is generally 0.0
01~) Approximately 1.0 Torr is used.

0.0~1. OTorr is particularly stable.

The temperature of the sample substrate is preferably selected between room temperature and 300°C.

150-250°C is most practical.

This is because at too low a temperature it is difficult to introduce hydrogen into the amorphous material, and at too high a temperature hydrogen tends to be released from the amorphous material.

The amount of hydrogen contained is controlled by controlling the hydrogen partial pressure in the Ar atmosphere.

When the amount of hydrogen in the atmosphere is 5 to 7 mol %, a content of about 30 atomic % can be realized in the amorphous material.

For other compositions as well, the hydrogen partial pressure may be set using approximately this ratio as a guide.

The hydrogen component in the material was determined by mass spectrometry using hydrogen gas generated by heating.

Note that Ar in the atmosphere can be replaced with another rare gas such as Kr.

Further, in order to obtain a film with high resistance, a magnetron-type low-temperature, high-speed sputtering apparatus is preferable.

A second method of manufacturing the amorphous material of the present invention is a method using glow discharge.

It is formed by performing glow discharge of SiH4 to decompose organic matter and depositing it on the substrate.

In addition, in the case of an amorphous material containing Si and C, 5IH4
A mixed gas of and CH4 may be used.

In this case, the pressure of the mixed gas of 5iI (4 and CH4 is 0.1
Maintain between ~5 Torr.

The glow discharge may be a direct current bias method or a high frequency discharge method.

Further, by changing the ratio of the mixed gas of 5tH4 and CH4, the ratio of Si and C can be controlled.

In order to obtain good quality amorphous material, the substrate temperature should be 200℃.
It is necessary to maintain the temperature at ~400°C.

In order to create p-plastic or n-type amorphous material, it is carried out by mixing 0.1 to 1% (volume ratio) of B2H6゜PH3, etc. to a mixed gas of 5IH4 and CH4. I can do it.

Furthermore, the amorphous film of the present invention can also be produced by electron beam evaporation in an atmosphere containing H2.

Embodiment 1 FIG. 4 to FIG. 10 are device cross-sectional views showing a method of manufacturing a solid-state imaging device of the present invention.

A scanning circuit section such as a switch circuit formed on a semiconductor substrate is manufactured using normal semiconductor device processes.

As shown in FIG. 4, an 800λ
A relatively thin SiO2 film 21 is formed, and this S i 02
Approximately 1,400 Si3N4 films 22 are placed at predetermined positions on the film.
form.

The SiO2 film is made using the normal CVD method, and the Si3N4 film is made using the Si3
A CVD method using N,,NH4,'N2 was performed.

A p diffusion region 23 is formed from above the silicon substrate by ion implantation.

FIG. 5 shows this state. This diffusion region 23 was provided to better isolate each element.

Next, the silicon is locally oxidized in a H2:0221:8 atmosphere to form a S r 02 layer 24 (FIG. 6).

This method is generally called LOGO8 and is a local oxidation method of silicon for element isolation.

- The 513N4 film 22 and the Sio2 film 21 are removed, and the gate insulating film 25 of the MOS transistor is replaced with SiO2.
Formed with two films.

Next, a gate portion 26 and diffusion regions 27 and 28 made of polysilicon are formed (FIG. 1), and the upper O
io2 film 29 is formed.

Then, 8000 Al is vapor-deposited as a drain electrode 31, which is opened in this film by etching to take out the electrodes of the source 21 and drain 28 (FIG. 8).

Furthermore, a total of 3,200 SiO2 films 32 are formed, and then Al is deposited to a thickness of 1 μm as a source electrode 33.

FIG. 9 is a sectional view showing this state.

Note that the electrode 33 was formed widely so as to cover the regions 27 and 28 and the gate portion.

This is because if light enters the signal processing region between the element isolation diffusion layers 23, it will cause blooming, which is undesirable.

Also, some of the shift registers placed around the periphery may have a general configuration as shown in FIG. Stable operation can be obtained regardless of the phase of the clock pulse that shifts the scanning pulse.

When the start pulse VIN is input, each bit terminal sequentially generates shift pulses o1. vo2... is output.

FIG. 12 shows the timing of this operation.

Of course, the specific circuit configuration of the shift register is not limited to this.

In this way, the scanning circuit MOS transistor section is completed.

A light receiving section is formed above this MOS transistor section.

FIG. 13 shows a plan view of the base portion. 41 is a contact hole for an electrode.

Next, the semiconductor substrate prepared through the steps up to this point is mounted on a magnetron type sputtering device.

The apparatus is as shown in FIG. The atmosphere was a mixed gas of Ar and hydrogen at 0.2 Torr.

The hydrogen content is 6 moles.

Silicon is used as the sputter target.

Frequency: 13.56 MHz. Reactive sputtering is performed with an input power of 300 W to deposit an amorphous silicon thin film 35 containing hydrogen to a thickness of 500 um on the semiconductor substrate 40 (FIG. 10).

The hydrogen content in this amorphous thin film was 20 atomic %, and the specific resistance was 5×10 13 Ω·.

A first electrode for bias voltage is provided on the top of the amorphous silicon thin film 35.
It is necessary to attach electrodes.

This electrode is a transparent electrode because all the light needs to enter from the top.

Since the heat resistance of amorphous silicon is 300° C., a Nesa electrode made of In2O3 was used.

Cr-Au is placed on a part of the top of the Nesa electrode that is not the light receiving area.
was vapor-deposited using a mask, and wire bonded thereto to form a bias electrode.

As in a normal semiconductor device, A
A second electrode is formed using a U film or the like.

In this way, the solid-state imaging device is completed. Note that 31 in FIG. 10 indicates incident light.

The solid-state imaging device manufactured by the method described above makes it possible to obtain a good screen without blooming.

Example 2 A solid-state imaging device was manufactured using the materials shown in Table 1 as photoconductive thin films.

The manufacturing procedure is similar to that described in Example 1.

Furthermore, the spectral sensitivity characteristics can be improved by adopting the following configuration as the photoconductive thin film.

First, 1 amorphous silicon film containing 25 atoms of hydrogen was deposited.
μ, then 20 at% hydrogen, 20 at% germanium
, an amorphous material consisting of 60 atomic percent silicon, and 2 atomic percent hydrogen.
Layers of amorphous materials each having a thickness of 0.5 μm are formed, each having a thickness of 0 atomic %, 30 atomic % carbon, and 50 atomic % silicon.

The formation method was based on the above-mentioned reactive spack method.

Furthermore, it is placed in a vacuum evaporation apparatus and Ce 02 is evaporated to a thickness of 10 nm using a resistance heating method.

Finally, gold was deposited to a thickness of 25 nm. If gold has a thickness of this level, the light transmittance can be 60% or more, and sufficient light intensity can be obtained.

Instead of CeO2 in the above example, 5in2. Good results were also obtained by depositing TiO□ or the like.

A film thickness of 100 to 300 was used.

Example 3 Shift register and switching MO8F using MOS transistors on an n-type silicon substrate as in Example 1
Manufacture ET.

The basic structure is the same as that of the first embodiment.

However, since the substrate is n-type, it has a p-type channel configuration.

This may be done in accordance with a well-known semiconductor IC manufacturing method.

Amorphous silicon containing hydrogen was deposited by glow discharge on the Si substrate on which the scanning circuit was prepared in this way.

The discharge atmosphere is S + H4, 1.5 Torr.

The substrate is heated to 500℃, and the high frequency input is set to a frequency of 0.5.
The amorphous material was deposited at MHz, pressure of 1.0 Torr, and substrate temperature of 300°C.

The thickness of the amorphous material is 21 tm, and the specific resistance is I x 10'i.
It was Ω・dark.

A NESA electrode was formed using In2O3 on this amorphous material, and a solid-state imaging device was manufactured.

Embodiment 4 In this example, a CCD (Charge Couple) is used as a scanning circuit.
pl-ed Device) uses the transfer layer area.

FIG. 14 shows a plan view of the arrangement of each element.

50 is a horizontal clock terminal, 51 is a vertical clock terminal,
52 is an output horizontal transfer register, 53 is a vertical transfer gate,
54 is a vertical analog transfer register, and 55 is a picture element portion in which a diffusion region and an MO8 type switch with this as 'north' are combined.

FIG. 15 is a cross-sectional view of the CCD transfer area, and FIG. 16 is a cross-sectional view of the picture element portion.

In FIG. 15, an electrode 62 is placed on a Si substrate 61 via an insulating layer.
．． 63 is formed and a two-phase clock voltage is applied through 64.65.

This moves the potential well in the Si substrate,
Charge transfer takes place.

FIG. 16 is a cross-sectional view of the light-receiving region, that is, the picture element, in which 11 is a diffusion layer, 12 is an insulating layer, 13 is a metal electrode, γ4 is a gate, γ5 is a photoconductive film, and γ6 is a transparent electrode.

The CCD transfer area shown in FIG. 15 is formed in succession to this light receiving area.

The transparent electrode γ6, the photoconductive film γ5, and the metal electrode γ3 form a light-receiving area, and the switch area that moves the carriers induced here to the transfer area is G! 14 forms a substantial MOS switch.

Si forming the CCD transfer area and MOS switch section
A substrate is prepared and attached to a magnetron type sputtering device.

The atmosphere was set to 0.2 T (+rr) with a mixed gas of Ar and hydrogen.

The hydrogen content is 6 mol%.

The spuck target was silicon.

Note that each element constituting the Si substrate, ie, the CCD transfer area, MOS switch section, etc., may be manufactured by a method known up to now.

Reactive sputtering is performed at a frequency of 13.56 MHz and an input of 300 W to deposit a hydrogen-containing amorphous material thin film γ5 to a thickness of 500 nm on the light receiving portion of the Si substrate.

The hydrogen content in this amorphous material was 20 atomic %, and the specific resistance was 5×10 13 Ω, crIL.

An In2O3 nesa electrode is formed on this amorphous material.

Cr--Au was mask-deposited from a part of the Nesa electrode, and wire-bonded thereto to form a bias electrode.

The operation will be briefly explained using FIG. 14.

When light hits the light receiving section through the transparent electrode, the carriers induced by this optical signal are vertically transferred by applying a voltage to the gate electrode between the diffusion region in the light receiving section 55 and the vertical analog transfer register 54. It is moved to register 54.

The vertical transfer register is a CCD driven through a two-phase vertical clock terminal 51, and is sent column by column through a vertical transfer gate 53 to an output horizontal transfer register 52.

This horizontal transfer register also has two-phase horizontal clock terminal 5.
It is a CCD driven through 0, and the charge corresponding to the signal C is sent to the output and taken out as a signal output.

The frequency of the 2-phase 1 drive may be selected so that the transfer of the horizontal transfer register is completed within the period of the voltage pulse applied to the vertical transfer gate.

The imaging device of the present invention described using the above embodiments has good spectral sensitivity matching with visual sensitivity, and good sensitivity and noise characteristics.

In addition to features such as high resolution and no blooming, it also has features such as low power consumption, small size, light weight, and high reliability, and has extremely large industrial effects.

[Brief explanation of the drawing]

Figure 1 is a diagram showing the principle of a solid-state imaging device, Figure 2 is a cross-sectional view of a pixel part of a solid-state imaging device using a photoconductive thin film, and Figure 3 is a diagram showing the principle of a solid-state imaging device.
The figure is an explanatory diagram of a reactive sputtering device, FIG. 4 to FIG. 10 are main part sectional views showing the manufacturing process of the solid-state imaging device of the present invention, FIG. 11 is a diagram showing an example of a shift register, and FIG. 12 13 is a plan view of the solid-state imaging device of the embodiment, FIG. 14 is an explanatory diagram of an embodiment using a CCD as a scanning circuit, and FIG. 15 is a diagram showing the operation timing of the shift register.
A cross-sectional view of the D transfer region, and FIG. 16 is a cross-sectional view of the light receiving section. In the figure, 1...horizontal scanning signal generator, 2...
... Vertical scanning signal generator, 3 ... Switch section, 4 ... Picture element, 5 ... Output end, 6
, 20... Semiconductor substrate, ? , 8,2γ, 28・
...Diffusion region, 10,26...Gate electrode, 13,29.32...Insulating layer, 16.33
...First conductive layer, 17,35...Photoconductive thin film, 18.36...Transparent electrode.

Claims (1)

[Scope of Claims] 1. A semiconductor substrate having at least a scanning means for sequentially selecting a plurality of photoelectric conversion parts, at least a photoelectric conversion material layer on the top of the semiconductor substrate, and a light-transmitting electrode on the top of the photoelectric conversion material layer. What is claimed is: 1. A solid-state imaging device having a photoelectric conversion section in which a photoelectric conversion section is arranged, wherein the photoelectric conversion material is an amorphous material mainly composed of silicon and containing hydrogen. 2. The solid-state imaging device according to claim 1, wherein the photoelectric conversion material layer is formed by a reactive sputtering method in an atmosphere containing at least hydrogen. 3. The solid-state imaging device according to claim 1, wherein the photoelectric conversion material layer is formed by a glow discharge method in an atmosphere containing at least silane. 4. The solid-state imaging device according to claim 1, wherein the scanning means is constituted by at least a field-effect transistor. 5. The solid-state imaging device according to claim 1, wherein the scanning means is constituted by at least a charge transfer element.