JPH0969649A - Semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the generation of an interface level and that of a defect caused by the stress due to ground recess and projection and hence to improve element characteristics by suppressing the recess and projection in an interlayer interface in a semiconductor device using hydrogenated amorphous silicon. SOLUTION: In a semiconductor device with a photoelectric transducing diode which is formed by successively forming an i-type hydrogenated amorphous silicon carbide layer 21, an i-type hydrogenated amorphous silicon layer 22, and a p-type hydrogenated amorphous silicon carbide layer 23, carbon is contained at a lower concentration than the p-type hydrogenated amorphous silicon carbide layer 23 near the interface with the p-type hydrogenated amorphous silicon carbide layer 23 of the hydrogenated amorphous silicon layer 22.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor device using an amorphous semiconductor, and more particularly to a semiconductor device with improved characteristics of an amorphous semiconductor layer.

[0002]

2. Description of the Related Art In recent years, amorphous semiconductors have attracted attention in the development of semiconductor materials. A typical example of the material is amorphous silicon. Amorphous silicon has a feature that a large-area film can be obtained at low cost, and thus has been widely tried for device application. Examples of applications include solar cells, thin film transistors, photosensitive drums, solid-state imaging devices, and the like.

As an example, FIG. 1 shows a typical element structure of a photoconductive film stack type solid-state imaging device using amorphous silicon. On the surface of the p-type silicon substrate 10, n ⁺
Charge accumulation diode 11, n consisting of layers and n ^{+ +} layers
^- a layer CCD buried channel (vertical CCD) 1
2, the element separation layer 13 including the p ⁺ layer is formed, and the CCD
Transfer gates 14 and 15 are formed on the buried channel 12. A part of the transfer gate 14 extends to above the storage diode 11 and also serves as a signal charge reading section. An interlayer insulating film 16 such as SiO ₂ is formed on these,
An extraction electrode 17 such as WSi connected to the storage diode 11 is formed thereon. And BP on the whole surface
A flattening insulating film 18 of SG or the like is formed, and a pixel electrode 19 of Ti or the like connected to the extraction electrode 17 is formed thereon.

A photoconductive film 20 is formed on the solid-state image pickup device chip constructed as described above. Photoconductive film 2
Reference numeral 0 denotes an i-type amorphous silicon carbide layer (hereinafter referred to as an a-SiC: H (i) layer) 21, in order from the solid-state imaging device chip side,
An i-type amorphous silicon layer (hereinafter referred to as an a-Si: H (i) layer) 22 and a p-type amorphous silicon carbide layer (hereinafter referred to as an a-SiC: H (p) layer) 23 are laminated. And forms a diode. Then, the IT is formed on the photoconductive film 20.
A transparent electrode 30 such as O is formed.

In the solid-state image pickup device having such a structure,
Photoelectric conversion is performed in the i: H (i) layer 22, and the electrons generated therein are read out to the vertical CCD 12 via the storage diode 11 to obtain an image signal. In addition, the a-SiC: H (i) layer 21 acts as a layer that blocks the injection of carriers (holes) from the external electrode.
The SiC: H (p) layer 23 is a window that allows light to enter from the outside, and at the same time blocks the injection of electrons from the external electrode. As shown in FIG. 1, since the photoelectric conversion unit can be stacked on the charge transfer unit, the aperture ratio is 100% and the light does not enter the underlying CCD. Are high sensitivity and low smear.

As described above, the photoelectric conversion section (photoelectric conversion diode) using amorphous silicon has a well-known laminated structure of i / i / p, or more generally n / i / p. In such a structure, device performance is determined by the bulk properties and interface properties of each layer. With respect to a-Si: H and its alloys, their bulk properties have been widely discussed, and their correlation with the film forming process has also been obtained with considerable information. On the other hand, the information about the interface is scarce. As a factor of characteristic deterioration in that region at the dissimilar junction interface, for example, generation of an interface level due to lattice mismatch, generation of an interface level due to the influence of surface roughness at the interface, and addition of stress to the film near the interface The bulk of the film itself may be deteriorated.

As for the former, a countermeasure is considered by gradually changing the composition near the bonding interface. On the other hand, the surface roughness at the interface is expected to be affected, but no measures have been taken.
The generation of the interface state and the generation of defects in the bulk due to the stress have caused an increase in the leak dark current in the diode and a deterioration in the afterimage characteristics, which is a problem particularly in the case of a photoelectric conversion element of an image sensor.

[0008]

As described above, the influence of the interface appears significantly in the element characteristics of the photoconductive film laminated type solid-state image pickup device. Concavities and convexities at the bonding interface are (1) an increase in the interface level accompanying an increase in the interface area.

(2) Increase in film defects due to stress applied to the film formed on the uneven base. As a result, the element characteristics are deteriorated. On the other hand, if the material of the bulk or the film forming conditions are changed, the bulk characteristics are affected at the same time as the surface, so it is difficult to find the compatibility condition of the bulk characteristics and the surface morphology. Therefore, it is desired to maintain the conventional bulk properties and selectively improve only the surface morphology.

The present invention has been made in view of the above circumstances, and an object thereof is to suppress unevenness at a laminated interface when a photoelectric conversion diode is formed by using hydrogenated amorphous silicon. By doing so, it is possible to suppress the generation of interface states and the generation of defects caused by stress due to the unevenness of the underlying layer, and to provide a semiconductor device capable of improving element characteristics.

[0011]

[Means for Solving the Problems]

(Summary) In order to solve the above problems, the present invention employs the following configuration.

That is, according to the present invention, a semiconductor device having a structure in which a hydrogenated amorphous silicon layer forms a photoelectric conversion diode by forming an interface with an amorphous compound layer containing silicon, hydrogen and an element X other than that as a main component. In the above, the hydrogenated amorphous silicon layer near the interface contains the element X at a concentration lower than that of the amorphous compound layer.

The preferred embodiments of the present invention are as follows. (1) The element X of the amorphous compound layer is carbon, nitrogen or oxygen. (2) Element X contained in hydrogenated amorphous silicon layer
However, the concentration gradually increases from the inside of the hydrogenated amorphous silicon layer toward the interface with the amorphous compound layer.

The present invention can also employ the following configurations. (1) In a semiconductor device including a photoelectric conversion diode formed by laminating an n-type or i-type hydrogenated amorphous silicon carbide layer, an i-type hydrogenated amorphous silicon layer, and a p-type hydrogenated amorphous silicon carbide in the above order, It is characterized in that carbon is contained in a concentration lower than that of the p-type hydrogenated amorphous silicon carbide layer in the vicinity of an interface between the hydrogenated amorphous silicon layer and the p-type hydrogenated amorphous silicon carbide layer. (2) A solid-state imaging device in which a signal charge storage part, a signal charge reading part, and a signal charge transfer part are formed on a semiconductor substrate, and a pixel electrode electrically connected to the signal charge storage part is formed in the uppermost layer. In a semiconductor device including a chip, a photoconductive film deposited on the solid-state imaging device chip, and a transparent electrode formed on the photoconductive film, an n-type or i-type hydrogenated film is used as the photoconductive film. Amorphous silicon carbide layer, i
-Type hydrogenated amorphous silicon layer and p-type hydrogenated amorphous silicon carbide are stacked in the above order, and a p-type hydrogenated amorphous layer is formed near the interface between the hydrogenated amorphous silicon layer and the p-type hydrogenated amorphous silicon carbide layer. It is characterized in that carbon is contained at a concentration lower than that of the silicon carbide layer. (Operation) When forming diodes such as i / i / p and n / i / p, desired films are sequentially laminated. Therefore, when considering the interface states due to surface roughness,
The surface morphology of the film formed immediately before the film to be formed poses a problem. 1 μm thick film formed on Si substrate
The surface morphology of the a-Si: H (i) film is shown in FIG. Thus, the characteristic surface roughness was observed, and the average roughness Ra was 1.
4 nm. By the way, the average roughness Ra is

[0015]

[Equation 1] However, L: measurement area f (x): amount given by deviation from the average value at each point. A-Si as a carrier injection blocking layer
The reverse of the case where an n / i / p diode is formed by using C: H (n) and a-SiC: H (p) and using the a-Si: H (i) film having the surface structure shown here. The current-voltage characteristic in the direction is shown by A in FIG.

On the other hand, similar to FIG. 2, FIG. 4 shows the surface morphology of an a-SiC: H (i) film having a thickness of 1 μm formed on a Si substrate. It can be seen that a flat surface is obtained, which is clearly different from the case of a-Si: H. It is considered that this is because the influence of carbon-based radicals flying during the formation of the a-SiC: H film suppressed the condensation of silicon-based radicals on the growth surface to form a flat surface. By utilizing such surface reaction control, it becomes possible to flatten the surface irregularities of the a-Si: H film. That is, after forming the a-Si: H film with the desired film thickness,
By forming a-SiC: H with a small film thickness, the bulk characteristics are
It is possible to improve the surface structure while maintaining the a-Si: H characteristics.

As described above, the surface reaction is suppressed by the mixing of the carbon-based radicals, and the radicals can move to the site having a lower potential. However, when surface irregularities are present, the surrounding atoms are present in the valley. Since it will be many, it will be a site with low potential. Therefore,
The film is grown so as to fill the valley region, so that a flat film is formed.

Further, the formation of a-SiC: H aiming at the smoothing of the surface produces an a-Si: H / a-SiC: H interface, and the improvement effect is suppressed, so that the carbon-based radical It is desirable that the introduction of the is done not gradually but gradually while increasing its amount. FIG. 5 shows the surface after forming a-Si: H to a thickness of 1 μm and forming a-SiC: H to a thickness of 5 nm while gradually increasing the number of carbon-based radicals. A clear improvement in the surface structure is observed as compared to FIG.

In this way, the surface was smoothed a
Regarding the n / i / p diode using -Si: H (the carrier injection blocking layer is the same as described above), comparing with FIG.
It can be seen that the leak dark current is suppressed, as shown in B of FIG. This is because the flattening of the interface reduces the density of defects formed in the vicinity of the interface, resulting in a decrease in the number of generated carriers.

Here, if the composition and the discharge state are continuously performed during the formation of a-Si: H (i) and a-SiC: H (p), the problems described above can be avoided. Becomes In that case, a dope layer for drawing out carrier injection blocking performance is formed in the same reaction chamber as the intrinsic layer, but it is not desirable in consideration of the effect of autodoping and the effect of throughput. It is desirable to form each layer independently with the apparatus having the multi-film forming chamber configuration. Therefore, considering productivity, it is necessary to stop the film growth on the way, and therefore the improvement of the surface morphology shown in the present invention has a great significance.

The film forming method is not particularly limited as long as it is a commonly used film forming method such as plasma CVD or photo CVD. Also, although only a-Si: H and a-SiC: H were discussed as the different types of junctions, the combination is not limited to these, and any combination that has an effect of suppressing the surface reaction may be used. -SiN, a-SiO, etc. are also possible.

As described above, according to the present invention, in a semiconductor device using an amorphous semiconductor, by flattening the a-Si: H surface, an interface of a film forming an interface with the surface and a defect near the interface are generated. By suppressing the above, it is possible to suppress a leak dark current of the device and an afterimage which is a problem when used as a photoelectric conversion device of the image pickup device.

[0023]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. (Embodiment 1) Here, a-Si formed on a Si substrate
Consider the surface morphology and film composition of C: H films. The film formation was performed by a high frequency plasma CVD method using SiH ₄ , CH ₄ , and H ₂ as raw material gases, the substrate temperature was 230 ° C., and the discharge power was 170 mW / cm ² . As parameters, [H ₂ ] / ([SiH ₄ ] + [CH ₄ ]) and [SiH ₄ ] + [CH ₄ ] (where [] is the flow rate (sc
cm)) is constant, and [SiH ₄ ] / [CH ₄ ]
Changed the ratio.

By doing so, Si /
It is possible to control the (Si + C) composition ratio. Figure 6
Is the average roughness Ra on the vertical axis and C / (Si + in the film on the horizontal axis).
C) The composition ratio is shown. Thus, even if a small amount of carbon is mixed in, the flatness is remarkably improved. Further, it can be seen that as the amount of carbon increases, the roughness is promoted again when C / (Si + C) is around 0.5. (Embodiment 2) In this embodiment, an n / i / p diode (however, the n and p layers are a-SiC: H layers doped with phosphorus and boron, respectively, and the i layer is an a-Si: H layer) The a-Si: H (i) layer in the vicinity of the interface with the p-layer in is discussed. aS
The i: H (i) layer is formed by a plasma CVD method using SiH ₄ and H ₂ as raw material gases. After forming a desired film thickness, CH ₄ gas is mixed to achieve surface flattening. ing. Here, the mixing ratio of CH ₄ gas was changed to form the film under three conditions. FIG. 7 shows the distribution of carbon concentration (C / (Si + C)) in the film thickness direction measured by Rutherford backscattering method.
(A) to (c) are shown. Here, the carbon concentration of the a-SiC: H (p) film was about 20%.

7A shows the carbon concentration in the surface flattening layer near the interface of the a-Si: H (i) layer continuously changed, and FIG. 7B shows the surface flattening layer. The carbon concentration at the interface is changed by continuously changing the carbon concentration at a-SiC: H
Higher than the carbon concentration of (p), (c) the carbon concentration of the surface flattening layer is larger than that of the a-Si: H (i) layer, and a-SiC: H
It is a constant value smaller than that of the (p) layer.

In these three configurations, an ITO transparent electrode was formed on the a-SiC: H (p) film, and afterimages were measured in a state where light could be incident. a-Si: H (i) film thickness 2μm, a-SiC: H
^Photocurrent 2 × 10 ⁻⁷ when (i, p) film thickness is 20 nm
FIG. 8 shows the applied voltage dependence of the afterimage of three fields with respect to A / cm ² . The symbols A, B, and C in FIG. 8 correspond to (a), (b), and (c) in FIG. 7, respectively.

In this way, the carbon concentration for flattening is a
It can be seen that when the value is larger than that of the -SiC: H (p) layer, the afterimage value increases. As expected from the band profile, this is a manifestation of the influence on the running characteristics of the carrier due to the sandwiching of the wide gap layer. Also,
It can also be seen that when the introduction of CH ₄ is made steep and the carbon concentration change is made steep, the afterimage increases. This is because the introduction of carbon for planarization causes a new interface level, which leads to an increase in trap level density. In other words, in order to reduce the afterimage, as shown in FIG. 7 (a), the carbon concentration for flattening is made smaller than that of the a-SiC: H (p) layer, and the carbon concentration change is made gentle. It is better to do. (Third Embodiment) In the present embodiment, the C / (Si + C) ratio is set to 0.2 corresponding to the first embodiment,
[H ₂ ] / ([SiH ₄ ] + [CH ₄ ]) flow rate ratio dependencies are compared (the same flow rate ratio was 10 in the first embodiment). In FIG. 9, the vertical axis represents the average roughness Ra and the horizontal axis represents [H ₂ ] /
([SiH ₄ ] + [CH ₄ ]) Flow rate ratio is shown. It can be seen that Ra decreases until the flow rate ratio is about 10, and Ra is almost saturated at that value.

From the above, [H ₂ ] during the formation of the a-Si: H layer
/ [SiH _4] when the flow ratio is formed in less than 10 conditions, for planarization is carried out also introduced at the same time as the introduction of H _{2 CH 4, [H 2] /} ([SiH 4 ] + [C
H ₄ ]) It is effective to increase the flow rate ratio.

As described above with reference to the embodiments, in a photoelectric conversion diode using an amorphous semiconductor, particularly hydrogenated amorphous silicon, a radical species other than silicon-based radicals, such as carbon-based radicals, is formed during formation of the hydrogenated amorphous silicon film. By suppressing the surface reaction of silicon-based radicals, a flat surface of the hydrogenated amorphous silicon layer is obtained, and the generation of defects at and near the interface of the film forming the interface with the surface is suppressed. You can Therefore, it is possible to suppress the leak dark current of the photoelectric conversion diode, and it is possible to suppress an increase in afterimage which is a problem particularly when used in the photoelectric conversion unit of the stacked solid-state imaging device.

The present invention is not limited to the above embodiments. Although the embodiment has been described by taking the example applied to the stacked solid-state imaging device, the present invention is not limited to the imaging device and may be applied to various semiconductor devices in which an amorphous silicon layer and an amorphous compound layer are stacked to form a photoelectric conversion diode. You can The element other than silicon and hydrogen in the amorphous compound layer is not necessarily limited to carbon, but may be nitrogen or oxygen. In addition, various modifications can be made without departing from the scope of the present invention.

[0031]

As described above in detail, according to the present invention, the hydrogenated amorphous silicon layer forms an interface with the amorphous compound layer containing silicon, hydrogen and the other element X as the main components, and photoelectric conversion is performed. When a diode is constructed, in the hydrogenated amorphous silicon layer near the interface,
By including the element X at a concentration equal to or lower than that of the amorphous compound layer, it is possible to suppress unevenness at the stacking interface and suppress the generation of interface states and the generation of defects caused by stress due to the unevenness of the underlying layer. It is possible to improve the device characteristics.

[Brief description of drawings]

FIG. 1 is a cross-sectional view showing an element structure of a photoconductive film stack type solid-state imaging device.

FIG. 2 is a micrograph showing the surface structure of an a-Si: H (i) film.

FIG. 3 is a diagram showing reverse current-voltage characteristics of an n / i / p diode with and without interface flattening.

FIG. 4 is a micrograph showing the surface structure of an a-SiC: H (i) film.

FIG. 5 is a micrograph showing the surface structure of a film in which a-SiC: H is formed on a-Si: H.

FIG. 6 is a diagram showing the relationship between the average roughness Ra and the C / (Si + C) composition ratio in the film.

FIG. 7 is a diagram showing a distribution of a C / (Si + C) composition ratio in a film thickness direction.

FIG. 8 is a diagram showing a reverse applied voltage dependency of a three-field afterimage with respect to standard light.

FIG. 9: Average roughness Ra and [H2] / ([SiH4] + [CH
4]) Diagram showing the flow ratio dependency.

[Explanation of symbols]

10 ... Si substrate 11 ... Signal charge storage diode 12 ... CCD embedded channel (vertical CCD) 13 ... Element isolation layer 14, 15 ... Poly Si transfer gate 16 ... SiO ₂ interlayer insulating film 17 ... WSi extraction electrode 18 ... BPSG planarization insulation Film 19 ... Ti pixel electrode 20 ... Photoconductive film 21 ... a-SiC: H (i) film 22 ... a-Si: H (i) film 23 ... a-SiC: H (p) film 30 ... ITO transparent electrode

Claims (2)

[Claims] 1. A semiconductor device having a structure in which a hydrogenated amorphous silicon layer forms a photoelectric conversion diode by forming an interface with an amorphous compound layer containing silicon, hydrogen and an element X other than silicon as a main component, A semiconductor device, wherein the hydrogenated amorphous silicon layer near the interface contains the element X at a concentration lower than that of the amorphous compound layer. 2. A semiconductor provided with a photoelectric conversion diode formed by laminating an n-type or i-type hydrogenated amorphous silicon carbide layer, an i-type hydrogenated amorphous silicon layer, and a p-type hydrogenated amorphous silicon carbide layer in the above order. In the device, carbon is contained near the interface of the hydrogenated amorphous silicon layer with the p-type hydrogenated amorphous silicon carbide layer in a concentration lower than that of the p-type hydrogenated amorphous silicon carbide layer. Semiconductor device.