JP2009140996A - Solid-state imaging element and its manufacturing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To secure an insulating breakdown voltage between a transfer electrode and a light shielding film without inducing the deterioration of picture quality caused by a smear or a dark current. <P>SOLUTION: A photoelectric conversion part 41 for generating a signal charge corresponding to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion part 41 are formed on a semiconductor substrate. Furthermore, a transfer electrode formed opposite to the charge transfer channel and a light shielding film covering the upper side of the transfer electrode are accumulated in the solid-state imaging element 100. A first silicon nitride film having a hydrogen shielding property is formed between the light shielding film and the transfer electrode, and a second silicon nitride film having hydrogen permeability is formed on the photoelectric conversion part. The first silicon nitride film can be formed of a CVD thermal nitride film, and the second silicon nitride film can be formed of a low-temperature plasma CVD film. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a solid-state imaging device and a method for manufacturing the same, and more particularly to an improved technique for sufficiently maintaining a withstand voltage between an electrode and a light shielding film without increasing a dark current level and a smear level.

FIG. 9 is a cross-sectional view of the vicinity of a light receiving portion in a conventional solid-state imaging device.
The gate insulating film under the transfer electrode 1 of the solid-state imaging device has a so-called ONO structure in which a silicon nitride (SiN) film 3 having a high breakdown voltage is sandwiched between silicon oxide (SiO) films 5 and 7. In recent solid-state imaging devices that have been miniaturized, a gate insulating film having such a structure has been widely adopted in order to reduce the thickness of the gate and provide an insulating strength. A light shielding film 9 is formed above the transfer electrode 1. An antireflection thermal nitride film 11 is formed on the oxide film 5 on the surface of the photoelectric conversion portion (PD) in order to reduce the reflection loss of incident light.

In addition, the noise of the solid-state imaging device includes a dark current generated through a Si / SiO ₂ interface state (dangling bond or the like). As the effective means to erase the dark current, there is a method of treating the hydrogen termination of dangling bonds of hydrogen annealing to stabilize reduces the Si / SiO ₂ interface state (dangling bonds). In this dangling bond termination method, since the low-pressure thermal nitride film (LPCVD nitride film) does not pass hydrogen, in order to secure a hydrogen path, for example, as shown in FIG. The passage 13 must be opened.

In addition, there is a solid-state image sensor disclosed in Patent Documents 1 to 3 as a means for eliminating smear and dark current.
In the solid-state imaging device disclosed in Patent Document 1, an Al film as a light-shielding film is covered with a SiO film that is a passivation film, and an SiN film is formed on the SiO film so that an optical path caused by light refraction at a stepped portion. A passivation film having a two-layer structure is formed by reducing the film thickness to such an extent that the deviation can be ignored. Since the SiO film has a low refractive index, the lens action for refracting the light incident on the stepped portion toward the vertical transfer register is small, and the smear is small. Further, the SiN film contains a large amount of hydrogen, and this hydrogen reduces the interface state, thereby suppressing dark current.

  In the solid-state imaging device disclosed in Patent Document 2, the upper oxide film of the three-layer gate insulating film of oxide film-nitride film-oxide film is etched by etching using a resist film as a mask before forming the transfer electrode. And the nitride film is removed. Thus, when heat treatment is performed in a hydrogen atmosphere, the surface state of the charge storage portion is terminated with hydrogen, and the dark current of the charge storage portion is reduced.

The solid-state imaging device described in Patent Document 3 includes a silicon nitride film formed by plasma chemical vapor deposition in an atmosphere containing hydrogen-containing gas molecules, and a silicon oxide film having a refractive index lower than that of the silicon nitride film. The laminated multilayer film is formed as a protective film for supplying hydrogen to the silicon substrate. That is, the protective film is used as a hydrogen supply source for suppressing dark current.
JP-A-6-132515 JP-A-6-77457 JP 2000-252451 A

  In general, polysilicon is often used for the read and transfer electrodes 1, and tungsten is often used for the light shielding film 9. Since tungsten is a metal and acts as an electrode, it is generally fixed at 0 V (GND) to avoid an unstable floating state. The read and transfer electrodes 1 are pulsed with VH, VM, VL and ternary voltage, and typical values are VH = 15V, VM = 0V, VL = -8V, and the like. Therefore, a high withstand voltage (specifically, a potential difference greater than or equal to the VH-GND potential difference) is required between the polysilicon electrode 1 and the light shielding film 9. A high withstand voltage between the polysilicon electrode 1 and the light shielding film 9 is ensured by the insulating film (oxide film 7) between the polysilicon electrode 1 and the light shielding film 9.

On the other hand, the process tends to be lowered in order to suppress the spread of the impurity diffusion layer by miniaturizing the element and reduce the bird's beak of the polysilicon electrode itself. When the temperature of the process is lowered, the viscous flow at the time of oxidation decreases. Therefore, as shown in part A of FIG. 10, the shape of the transfer electrode 1 tends to have a convex corner 1 a and the insulation distance from the light shielding film 9 is small. Shorter and lower withstand voltage. Furthermore, since the polysilicon electrode itself is required to have a low resistance, if silicidation is performed in the future, it is difficult to oxidize the polysilicon itself, which causes a serious problem due to the withstand voltage.
As described above, sensitivity and saturation, which are basic characteristics of the sensor, have a trade-off relationship in principle, but miniaturization is progressing while maintaining or improving the basic characteristics such as sensitivity and saturation. . As described above, there are several techniques for avoiding the fundamental trade-off, but generally, constant voltage scaling that does not reduce the operating voltage is adopted for miniaturization. As a result, the problem that the electric field becomes large in various places is included.

  As another attempt to ensure withstand voltage, there is a method of making the interlayer insulating film a thermal CVD oxide film instead of thermal oxidation, but this method may cause insulation leakage due to the occurrence of pinholes and the like. Get higher. In addition to the polysilicon thermal oxide film, an antireflection nitride film or a CVD oxide film is sandwiched between the polysilicon electrode and the light shielding tungsten electrode. Of these, the antireflection nitride film is usually a low-pressure thermal nitride film (LPCVD nitride film) having high insulation and high uniformity that can also be used as a gate insulating film. However, since the low-pressure thermal nitride film hardly passes hydrogen for terminating the dangling hydrogen at the interface between silicon and the oxide film as a dark current source, it is necessary to open a hydrogen passage 13 in the pixel portion. For this reason, it cannot be expected to function as an insulating film that ensures the dielectric strength between the polysilicon electrode and the light-shielding tungsten electrode.

Further, a CVD oxide film has been conventionally used as an insulating film for ensuring a dielectric breakdown voltage between a polysilicon electrode and a light-shielding tungsten electrode. However, a thermal oxide film can be used as a thermal CVD film or a thermal TEOS-CVD film. It is inferior in terms of generation of dielectric strength and pinholes. Thus, conventionally, the withstand voltage between the polysilicon electrode and the light-shielding tungsten electrode has been maintained by the combination of the polysilicon thermal oxide film and the interlayer insulating CVD oxide film, but it has become difficult to maintain. Although it is an easy solution to thicken the CVD oxide film as much as the insulation property of the polysilicon thermal oxide film is weakened, this method increases the gap distance (height) between the silicon substrate and the tungsten light-shielding film. Therefore, smear which is noise peculiar to CCD becomes large and is not realistic.
Furthermore, although the above-mentioned patent document has a configuration in which a silicon nitride film is used as a protective film for supplying hydrogen to a silicon substrate, that is, a hydrogen supply source for dark current suppression, Since the silicon substrate is covered with the nitride film, good hydrogen termination cannot be expected.
The present invention has been made in view of the above situation, and does not form a special hole for hydrogen termination, and does not cause image quality degradation due to smear or dark current, so that reliable withstand voltage between the transfer electrode and the light shielding film can be ensured. Is to provide a solid-state imaging device and a method for manufacturing the same.

The above object of the present invention is achieved by the following configuration.
(1) The charge transfer channel is formed on a semiconductor substrate on which a photoelectric conversion unit that generates a signal charge corresponding to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit are formed. A solid-state imaging device in which a transfer electrode formed to face and a light-shielding film covering the upper part of the transfer electrode are stacked,
A first silicon nitride film having a hydrogen shielding property is formed between the light shielding film and the transfer electrode;
A solid-state imaging device in which a second silicon nitride film having hydrogen permeability is formed on the photoelectric conversion unit.

  According to this solid-state imaging device, the first silicon nitride film and the second silicon nitride film are interposed between the light shielding film and the transfer electrode, so that the withstand voltage is secured and the second silicon nitride film Hydrogen permeability is also secured and good hydrogen termination is realized.

(2) The solid-state imaging device according to (1),
The first silicon nitride film is a solid-state imaging device which is a CVD thermal nitride film.

  According to this solid-state imaging device, it is easy to control the film thickness when determining the trade-off between the transfer electrode and light-shielding film dielectric strength and the smear suppression effect.

(3) The solid-state imaging device according to (1) or (2),
The first silicon nitride film is a solid-state imaging device having a thickness of 20 to 100 nm.

  According to this solid-state imaging device, a decrease in inter-electrode dielectric breakdown voltage that occurs when the film thickness is less than 20 nm is suppressed, and a decrease in smear suppression effect that occurs when the film thickness exceeds 100 nm is avoided.

(4) The solid-state imaging device according to any one of (1) to (3),
The solid-state imaging device, wherein the second silicon nitride film is a low temperature plasma CVD film.

  According to this solid-state imaging device, hydrogen permeability can be imparted to the nitride film while ensuring the inter-electrode dielectric strength.

(5) The solid-state imaging device according to (4),
The second silicon nitride film is a solid-state imaging device having a film thickness of at least 5 nm.

  According to this solid-state imaging device, the opening of the first silicon nitride film is filled with the second silicon nitride film having a thickness of 5 nm or more, and a predetermined inter-electrode dielectric strength is ensured.

(6) The solid-state imaging device according to any one of (1) to (5),
A solid-state imaging device in which the first silicon nitride film and the second silicon nitride film are bonded to each other, and the entire pixel region including the photoelectric conversion unit and the charge transfer channel is covered with a silicon nitride film.

  According to this solid-state imaging device, the withstand voltage is further increased by the continuous nitride film between the light shielding film and the transfer electrode. Also, optically no reflection occurs in the extra direction, and the uniformity is improved, which also improves smear. Further, it is possible to control the oxide film in a good state as optical characteristics.

(7) The solid-state imaging device according to any one of (1) to (6),
The transfer electrode is a solid-state imaging device formed of polysilicon.

  According to this solid-state imaging device, a vertical charge transfer path with high transfer efficiency can be easily manufactured.

(8) The solid-state imaging device according to any one of claims 1 to 7,
The transfer electrode is a solid-state imaging device formed of silicide.

  According to this solid-state imaging device, since the transfer electrode is made of silicide generated by the reaction of metal and Si, the transfer electrode can be easily reduced in resistance by high-temperature annealing or the like.

(9) The solid-state imaging device according to any one of (1) to (8),
A solid-state imaging device in which a gate insulating film having a laminated structure of a silicon oxide film, a silicon nitride film, and a silicon oxide film is formed on the surface of the semiconductor substrate below the first and second silicon nitride films.

  According to this solid-state imaging device, it is possible to reduce the thickness in miniaturization.

(10) The solid-state imaging device according to any one of (1) to (8),
A solid-state imaging device in which a gate insulating film made of a silicon oxide film is formed on the surface of the semiconductor substrate below the first and second silicon nitride films.

  According to this solid-state imaging device, the manufacturing process can be shortened compared to the formation of the ONO film.

(11) A silicon oxide film, silicon on a semiconductor substrate on which a photoelectric conversion unit that generates a signal charge corresponding to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit are formed A first step of forming a gate insulating film made of a laminate of a nitride film and a silicon oxide film;
A second step of forming the transfer electrode at a position on the gate insulating film facing the charge transfer channel;
A third step of forming an interelectrode insulating film surrounding the electrode;
A fourth step of forming a first silicon nitride film made of a CVD thermal nitride film so as to cover the interelectrode insulating film;
A fifth step of forming a light-shielding film having an opening in the photoelectric conversion unit;
A sixth step of etching and removing the light receiving side of the photoelectric conversion unit using the silicon oxide film on the photoelectric conversion unit as an etching stop material;
A seventh step of forming a second silicon nitride film, which is a low-temperature plasma CVD film, in a pixel region including the photoelectric conversion unit and the charge transfer channel;
A method for manufacturing a solid-state imaging device including:

  According to this method for manufacturing a solid-state imaging device, the first silicon nitride film is formed so as to cover the interelectrode insulating film, and the first silicon nitride film is formed using the silicon oxide film on the photoelectric conversion portion as an etching stop material. Etching is removed, and then the second silicon nitride film is formed in the pixel region including the photoelectric conversion portion and the charge transfer channel, so that the first silicon nitride film removed by etching has a hydrogen permeability. A continuous nitride film is formed by being filled with the silicon nitride film.

(12) A single silicon oxide film is formed on a semiconductor substrate on which a photoelectric conversion unit that generates a signal charge corresponding to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit are formed. A first step of forming a gate insulating film composed of layers;
A second step of forming the transfer electrode at a position on the gate insulating film facing the charge transfer channel;
A third step of forming an interelectrode insulating film surrounding the electrode;
A fourth step of forming a first silicon nitride film made of a CVD thermal nitride film so as to cover the interelectrode insulating film;
A fifth step of forming a light-shielding film having an opening in the photoelectric conversion unit;
A sixth step of etching and removing the light receiving side of the photoelectric conversion unit using the silicon oxide film on the photoelectric conversion unit as an etching stop material;
A seventh step of forming a second silicon nitride film, which is a low-temperature plasma CVD film, in a pixel region including the photoelectric conversion unit and the charge transfer channel;
A method for manufacturing a solid-state imaging device including:

  According to this method for manufacturing a solid-state imaging device, after the first silicon nitride film is formed so as to cover the interelectrode insulating film and the light-shielding film having an opening is formed in the photoelectric conversion unit, The first silicon nitride film is etched away using the silicon oxide film as an etching stop material. Thereafter, the second silicon nitride film is formed in the pixel region including the photoelectric conversion portion and the charge transfer channel, so that the second silicon nitride film in which the first silicon nitride film removed by etching has hydrogen permeability. As a result, a continuous nitride film is formed. Further, the second silicon nitride film is formed on the surface where the end portion of the light-shielding film removed by etching is exposed, so that the end portion of the light-shielding film is connected to the second silicon nitride film.

(13) A method of manufacturing a solid-state imaging device according to (11) or (12),
The method of manufacturing a solid-state imaging device, wherein the opening of the light shielding film in the fifth step is formed by etching in the sixth step.

  According to this method for manufacturing a solid-state imaging device, the light-shielding film and the first silicon nitride film are etched on the same surface and connected together to a second silicon nitride film to be formed later.

(14) A method for producing a solid-state imaging device according to (11) or (12),
A method for manufacturing a solid-state imaging device, comprising an eighth step of reflowing the silicon oxide film between the sixth step and the seventh step.

  According to this method for manufacturing a solid-state imaging device, the light receiving side of the photoelectric conversion unit is removed by etching using the silicon oxide film as an etching stop material, and the silicon oxide film formed thereon is subjected to reflow treatment, so that the silicon oxide in the opening is obtained. The corner of the film has an R shape, and optical characteristics are imparted to a second silicon nitride film to be formed later on the silicon oxide film.

  According to the solid-state imaging device of the present invention, the first silicon nitride film having hydrogen shielding properties is formed between the light shielding film and the transfer electrode, and the second silicon having hydrogen permeability on the photoelectric conversion portion. Since the nitride film is formed, it is possible to ensure a reliable withstand voltage between the transfer electrode and the light shielding film without causing deterioration of image quality due to smear or dark current.

  According to the method for manufacturing a solid-state imaging device according to the present invention, the first silicon nitride film made of the CVD thermal nitride film is formed so as to cover the interelectrode insulating film, and the light shielding film having an opening in the photoelectric conversion portion is formed. Then, the silicon oxide film on the light receiving side of the photoelectric conversion unit is removed by etching as an etching stop material, and a second silicon nitride film that is a low temperature plasma CVD film is formed in the pixel region including the photoelectric conversion unit and the charge transfer channel. A hydrogen-shielding first silicon nitride film is provided on the inter-layer insulating film, and a hydrogen-permeable second silicon nitride film is provided in the opening of the photoelectric conversion unit to reduce image quality due to smear and dark current. Without incurring, a solid-state imaging device in which a certain withstand voltage is ensured between the transfer electrode and the light shielding film can be obtained.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of a solid-state imaging device and a manufacturing method thereof according to the invention will be described with reference to the drawings.
FIG. 1 is a cross-sectional view of an imaging unit of a solid-state imaging device according to the present invention, and FIG. 2 is a plan view showing a configuration example of a main part of the solid-state imaging device shown in FIG.
As shown in FIG. 1, the solid-state imaging device 100 is provided with a light receiving region 33, a horizontal charge transfer unit (HCCD) 35, an amplifier 37, and a signal output terminal 39 in the imaging device formation region 31. In the light receiving region 33, a photoelectric conversion unit 41, a charge readout unit 43, and a vertical charge transfer unit 45 are provided.

  More specifically, as shown in FIG. 2, in the solid-state imaging device 100, a large number of photoelectric conversion units 41 are arranged on the plane along the row direction (the direction of the arrow X) and the column direction (the direction of the arrow Y). The light receiving areas 33 are arranged in a dimension. Each photoelectric conversion unit 41 includes a photodiode (PD) made of a semiconductor.

  A plurality of vertical charge transfer units 45 and a line memory 47 are used to extract signal charges output from each of a large number of two-dimensionally arranged photoelectric conversion units 41 from the output terminals of the solid-state imaging device as signals for each time-series frame. And a horizontal charge transfer unit 35.

  The signal charges for one row accumulated in the line memory 47 are transferred from the line memory 47 toward the horizontal charge transfer unit 35, and as a result, the signal charges for one row are taken into the horizontal charge transfer unit 35. . The horizontal charge transfer unit 35 sequentially transfers the signal charges for one row held by itself in the horizontal direction (arrow X direction) in units of one pixel. The signal charge transferred through the horizontal charge transfer unit 35 is amplified by the amplifier 37 and output from the output terminal OUT.

  Control signals necessary to realize such a read operation, that is, a vertical transfer control signal φV (usually a signal of a plurality of phases), a transfer control signal φLM, and a horizontal transfer control signal φH (usually a signal of a plurality of phases) ) Are generated by a timing signal generation circuit (not shown) and applied to each vertical charge transfer unit 45, line memory 47, and horizontal charge transfer unit 35 of the solid-state imaging device. Note that the line memory 47 may be omitted.

  As shown in FIG. 2, the solid-state imaging device 100 is arranged so that a large number of photoelectric conversion units 41 form a honeycomb-like pattern (a pattern in which the arrangement positions of the photoelectric conversion units are shifted by a half pitch in the horizontal direction for each row). Has been. In addition, as indicated by “G1”, “G2”, “B”, and “R” in FIG. 2, the color components to be detected by each photoelectric conversion unit 41 are determined in advance. That is, each photoelectric conversion unit 41 of “G1” and “G2” detects the brightness of the green component, each photoelectric conversion unit 41 of “B” detects the brightness of the blue component, and each photoelectric conversion unit 41 of “R” detects the brightness of the red component. .

  These detection colors are set according to the spectral characteristics of the optical filter disposed in front of the light receiving surface of each photoelectric conversion unit 41. In the example illustrated in FIG. 2, four types of filter columns FC1, FC2, FC3, and FC4 are arranged so as to be divided for each column of the photoelectric conversion unit 41. Each optical filter has a so-called Bayer arrangement inclined by 45 °.

  The vertical charge transfer unit 45 is formed in a shape meandering at a position adjacent to each column for each column of the photoelectric conversion units 41. Each of the vertical charge transfer units 45 includes a vertical charge transfer channel 51 formed on the semiconductor substrate 49 and a number of charge transfer channels arranged on the semiconductor substrate 49 via an electrical insulating film (not shown). A first vertical transfer electrode 53, a second vertical transfer electrode 55, a first auxiliary transfer electrode 57, a second auxiliary transfer electrode 59, and a third auxiliary transfer electrode 61 are provided. The horizontal charge transfer unit 35 has one horizontal charge transfer channel 63 extending in a band shape in the direction of the arrow X.

  That is, by applying a predetermined voltage to each electrode 53, 55, 57, 59, 61 to form a predetermined potential distribution on each vertical charge transfer channel 51, and sequentially switching the voltage applied to each electrode, The vertical charge transfer unit (VCCD) 45 can sequentially transfer the signal charge of each pixel in a target direction.

  One first vertical transfer electrode 53 and one second vertical transfer electrode 55 are formed for each row of the photoelectric conversion units 41. Each first vertical transfer electrode 53 also functions as a read gate for controlling the transfer of signal charges from the photoelectric conversion unit 41 to the vertical charge transfer channel 51 of the vertical charge transfer unit 45.

  As shown in FIG. 2, four-phase vertical transfer control signals (also referred to as drive pulses) φV1 and φV2 are respectively applied to the second vertical transfer electrodes 55 and the first vertical transfer electrodes 53 that are alternately arranged in the arrow Y direction. , ΦV3, φV4 is applied according to the positional relationship between the second vertical transfer electrode 55 and the first vertical transfer electrode 53. Similarly, a vertical transfer control signal φV2 is applied to the first auxiliary transfer electrode 57, a vertical transfer control signal φV3 is applied to the second auxiliary transfer electrode 59, and a vertical transfer control signal φV4 is applied to the third auxiliary transfer electrode 61. Is applied. Further, a transfer control signal φLM is applied to the transfer control electrodes LM1 and LM2 in order to control the transfer of signal charges in the line memory 47.

3 is a cross-sectional view taken along line A1-A2 of FIG.
In the cross section of one photoelectric conversion unit 41 of the solid-state imaging device 100, a photodiode (PD) 65 that is a photoelectric conversion unit, a vertical charge transfer unit 45, and a charge readout unit 43 are formed in a semiconductor substrate 49. ing. A thin P-type region 67 is formed on the surface side of the photodiode 65. Between the vertical charge transfer portion 45 and the transfer electrode 69, an ONO structure is employed in which a silicon nitride (SiN) film 71 having a high breakdown voltage is sandwiched between silicon oxide (SiO) films 73 and 75. A light shielding film 77 covering the upper portion of the transfer electrode 69 is formed above the transfer electrode 69.

  The transfer electrode 69 is made of polysilicon. Thereby, the vertical charge transfer part 45 with high transfer efficiency can be easily manufactured. Note that the transfer electrode 69 may be formed of silicide. Since the transfer electrode 69 is made of silicide generated by the reaction between the metal and Si, the resistance can be easily reduced by high-temperature annealing or the like.

  A first silicon nitride film 79 is formed between the light shielding film 77 and the transfer electrode 69, and the first silicon nitride film 79 has a hydrogen shielding property. A second silicon nitride film 81 is formed on the photodiode 65, and the second silicon nitride film 81 has hydrogen permeability. The first silicon nitride film 79 and the second silicon nitride film 81 are connected below both ends of the light shielding film 77. That is, the semiconductor substrate 49 is covered with the continuous first silicon nitride film 79 and second silicon nitride film 81. In the figure, reference numeral 83 denotes a microlens that is independently arranged for each cell on the upper surface of the filter rows FC2, FC3, FC4.

  The first silicon nitride film 79 can be a CVD thermal nitride film. This facilitates film thickness control when determining the trade-off between the transfer electrode 69 and the light blocking dielectric strength and the smear suppression effect. The first silicon nitride film 79 has a thickness of 20 to 100 nm. This suppresses a decrease in inter-electrode dielectric breakdown voltage that occurs when the film thickness is less than 20 nm, and avoids a decrease in smear suppression effect that occurs when the film thickness exceeds 100 nm.

  By forming the second silicon nitride film 81 as a low-temperature plasma CVD film, it is possible to impart hydrogen permeability to the nitride film while ensuring an inter-electrode breakdown voltage. The second silicon nitride film 81 has a film thickness of at least 5 nm, whereby the opening of the first silicon nitride film 79 is filled with the second silicon nitride film 81 having a film thickness of 5 nm or more. Thus, a predetermined inter-electrode dielectric strength is ensured.

  In the present embodiment, the first silicon nitride film 79 and the second silicon nitride film 81 are bonded to each other, and the entire surface of the pixel region including the photodiode 65 and the charge transfer channel is covered with the silicon nitride films 79 and 81. Since the silicon nitride films 79 and 81 are continuous between the light shielding tungsten film 77 and the transfer electrode 69, the withstand voltage is further increased. Also, optically no reflection occurs in the extra direction, and the uniformity is improved, which also improves smear. Further, it is possible to control the oxide film in a good state as optical characteristics.

  In the present embodiment, a gate insulating film having a laminated structure of a silicon oxide film 73, a silicon nitride film 71, and a silicon oxide film 75 is formed on the surface of the semiconductor substrate below the first and second silicon nitride films 79 and 81. Is formed. Thereby, the thinning in miniaturization is possible.

  As a modification of the solid-state imaging device 100, as shown in FIG. 4, a solid-state imaging device 100A in which the silicon nitride film 71 is omitted may be used. Thereby, the manufacturing process can be shortened as compared with the formation of the ONO film.

  Therefore, according to the solid-state imaging device 100, the first silicon nitride film 79 having hydrogen shielding properties is formed between the light shielding film 77 and the transfer electrode 69, and the hydrogen permeable first film is formed on the photodiode 65. Since the second silicon nitride film 81 is formed, the withstand voltage is secured, and the hydrogen permeability is secured in the second silicon nitride film 81, thereby realizing a good hydrogen termination. As a result, it is possible to ensure a reliable withstand voltage between the transfer electrode 69 and the light shielding film 77 without deteriorating image quality due to smear or dark current.

Next, a method for manufacturing the above-described solid-state imaging device 100 will be described.
FIG. 5 is a process explanatory view showing the manufacturing procedure of the solid-state imaging device according to the embodiment by (a) to (f).
On a Nepi / N-type semiconductor substrate 49 in which an N-type epi layer is stacked on an N-type substrate, a Pwell region 91, a PD (P ⁺ / N-type) 65, a VCCD (N-type) 45, and a readout channel (P-type) 43 Then, an element isolation (P-type) region is formed, gate insulating films (ONO films) 71, 73, and 75, a transfer electrode 69 are formed, and an interlayer insulating oxide film is formed.
Note that the N-type and P-type notations are examples of electron storage and transfer CCDs that have been generally used in the past. In the case of hole storage and transfer CCDs, the opposite conductive semiconductor layers are used. .
Up to this point, this is basically the same as the conventional pixel manufacturing method.

  The difference is that the manufacturing method and film thickness of the oxide film 73 on the conventional photodiode 65 are the following in order to obtain interface stability for dark current suppression and gettering effect that can incorporate impurities such as heavy metals. In order to increase the optical sensitivity in combination with a thermal nitride film that is a deposited layer, it was determined mainly from two effects, but in this embodiment, any one of these may be considered, and as a result, It is possible to make it as thin as possible. For example, a thickness of about 20 to 50 nm is conventionally used, but in this embodiment, an oxide film 73 having a thickness of about 5 nm can be used.

  Next, a CVD thermal nitride film 79 is deposited. This process itself is the same as the conventional pixel manufacturing method, but the nitride film thickness at this time does not need to be a film having an antireflection effect for increasing the optical sensitivity as in the prior art. This nitride film thickness is determined by a trade-off between the dielectric breakdown voltage between the transfer electrode 69 and the light-shielding tungsten electrode and the smear suppressing effect. The greater the thickness, the more effective the dielectric strength between electrodes, and the higher the effect, the thinner the smear suppression effect.

  From these trade-offs, for example, about 20 to 100 nm is deposited. In this embodiment, the nitride film is not patterned and etched. That is, it is not necessary to pattern and etch only the antireflection thermal nitride film of the prior art which mainly aims to secure the hydrogen path.

  Next, an insulating film 95 such as thermal CVD or TEOS-CVD is deposited. Further, a light shielding tungsten film 77 is deposited by CVD or PVD, and patterning and etching are performed on the light shielding tungsten film. Form.

  The light shielding tungsten film 77 may have a TiN / W laminated structure, or any other conductive film having a light shielding property such as AL, and may be another film as long as the light shielding property and the electrical conductivity are satisfied. .

  Next, an oxide film 99 having good filling properties and flatness such as BPSG, thermal TEOS, HDP-SiO, and SOG is deposited. This oxide film layer may be a single layer, a stacked layer, or a combination of several deposition methods. If necessary, in consideration of the next patterning and optical characteristics, planarization processing such as height adjustment and CMP or etch back for maintaining patterning accuracy may be performed. On the contrary, it is not always necessary to make it flat, and a downward convex shape may be used. Thus, the structure of FIG. 5A is completed.

  Thereafter, the resist 101 is patterned (FIG. 5B), and an oxide film anisotropic etching is performed using the silicon nitride film 79 as an etching stop material (FIG. 5C). Here, in order to maintain the uniformity of the optical characteristics of the image sensor with good control, the silicon nitride film 79 is used as an etching stop material so as to stabilize the etching shape (that is, the shape of the lens in the lower convex layer), and uniform etching is performed. Create a hole shape.

  Next, anisotropic or isotropic etching of the silicon nitride film 79 is performed using the oxide film 73 on the PD 65 as an etching stop material (FIG. 5D), and the resist 101 is peeled off (FIG. 5E). ). Note that, when the silicon nitride film 79 is etched, conditions with a high selection ratio with respect to the oxide film 73 are used. Therefore, in the example shown in FIG. Post-etching may be performed.

  Thereafter, a low-temperature nitride film (plasma CVD nitride film) 81 is deposited (FIG. 5F). In order to improve the embedding property, the low temperature nitride film 81 may be formed, for example, by dividing the deposition several times or repeating the deposition and etch back several times. Thereafter, another film 103 other than the nitride film is deposited to complete the downward convex in-layer lens.

  Here, for example, in an ultrafine pixel structure of 2 μm square or less, the embedding hole is a square having a diameter of about 1 μm to a circle less than that and a side of about 1 μm to a square of less than that. There arises a problem that air holes (air gaps) are formed. In addition to electrical characteristics, the variation in optical characteristics of the aperture of the image sensor becomes fatal, so unintentional void formation must be avoided. In order to avoid the formation of such voids, a low-temperature nitride film 81 is deposited thinly as shown in FIG. 5F, and then another film 103 with good coverage may be deposited or filled.

  For example, a structure that is covered with a coating-type low-temperature oxide film (SOG or the like), a resist agent, or the like that is optically transparent and has good filling properties and good flatness is completed. Thereafter, the planarization layer stack 105 (see FIG. 3), color filter (CF) formation, and microlens 83 are formed, whereby the structure according to the present embodiment shown in FIG. 3 is completed.

  These flattening layer stack, color filter CF formation, microlens 83 formation, and upper convex in-layer lens (not shown) are determined by the application of the image sensor that is required. It is not necessarily essential to the invention. Note that the gate insulating film is not limited to the ONO film. FIG. 4 shows the structure in the case where the gate insulating film is made of only the silicon oxide film 73.

  In this manufacturing method, a first silicon nitride film 79 is formed so as to cover the interelectrode insulating film, and the first silicon nitride film 79 is removed by etching using the silicon oxide film 73 on the photodiode 65 as an etching stop material. Thereafter, the second silicon nitride film 81 is formed in the pixel region including the photodiode 65 and the charge transfer channel, so that the first silicon nitride film 79 removed by etching has a hydrogen permeability. A continuous nitride film is formed by being filled with the silicon nitride film 81. As a result, it is possible to obtain the solid-state imaging device 100 in which a reliable withstand voltage is ensured between the transfer electrode 69 and the light shielding tungsten film 77 without causing image quality degradation due to smear or dark current.

<Modification 1>
Next, a modified example of the manufacturing method will be described.
FIG. 6 is a process explanatory diagram showing the manufacturing procedure of the solid-state imaging device according to the modified example 1 by (a) to (f).
FIG. 5 and the manufacturing procedure are different in (a) to (c). A Pwell region 91, a PD (P ⁺ / N type) 65, a VCCD (N type) 45, a read channel (P type) 43, and an element isolation (P type) region are formed on a Nepi / N type semiconductor substrate 49, and a gate Insulating films (ONO films) 71, 73, and 75 and gate electrodes (N ⁺ type polysilicon) 69 are formed, and an insulating oxide film is formed between the polysilicon electrode layers.

  Next, a CVD thermal nitride film 79 is deposited. Next, an insulating film 95 such as thermal CVD or TEOS-CVD is deposited, and further a light shielding tungsten film 77 is deposited by CVD or PVD. Up to this point, the process is the same as in FIG. Next, a silicon oxide film 99 with good embeddability and flatness is deposited. Thus, the structure of FIG. 6A is completed. The difference from FIG. 5A is that the light shielding tungsten film 77 is not patterned and etched.

  Thereafter, the resist 101 is patterned (FIG. 6B), and anisotropic etching is performed using the silicon nitride film 79 as an etching stop material (FIG. 6C). Here, the material to be etched has a structure of a silicon oxide film 99, a light shielding tungsten film 77, and a silicon oxide film from the top. In this etching, it is desirable to use the silicon nitride film 79 as an etching stop material and finish it by batch etching. However, since the laminated structure of the film is complicated, an etching hole may be formed by dividing the etching several times. . The subsequent steps are the same as those in FIG.

  A first silicon nitride film 79 is formed so as to cover the interelectrode insulating film, a light shielding tungsten film 77 having an opening is formed in the photodiode 65, and then the silicon oxide film 73 on the photodiode 65 is etched away. As a result, the first silicon nitride film 79 is removed by etching. After that, the second silicon nitride film 81 is formed in the pixel region including the photodiode 65 and the charge transfer channel, so that the first silicon nitride film 79 removed by etching has the second silicon having hydrogen permeability. A continuous nitride film is formed by being filled with the nitride film 81. Further, the second silicon nitride film 81 is formed on the surface where the end of the light-shielding tungsten film 77 removed by etching is exposed, so that the end of the light-shielding tungsten film 77 is connected to the second silicon nitride film 81. Is done.

<Modification 2>
7A to 7F are process explanatory views showing the manufacturing procedure of the solid-state imaging device according to the modified example 2 by (a) to (f).
The last procedure shown in FIG. 7F is different from the procedure shown in FIG. In FIG. 7F, deposition / filling is performed only with the low-temperature nitride film 81. In order to improve the embedding property, the low temperature nitride film 81 may be formed, for example, by dividing the deposition several times or repeating the deposition and etch bag several times. After filling, if necessary, a planarization process such as CMP or etch back may be performed for height adjustment or the like in consideration of optical characteristics or the like.

<Modification 3>
FIG. 8 is a process explanatory diagram showing the manufacturing procedure of the solid-state imaging device according to the modified example 3 by (a) to (f).
The second half of FIGS. 8E and 8F is different from the procedure shown in FIG. 7 in this modification. In this modification, after the silicon nitride film 79 is etched, the silicon oxide film 99 is reflowed (melted) as shown in FIG. 8 (e), and then as shown in FIG. 8 (f). The low-temperature nitride film 81 to be deposited is formed in the shape of a convex convex lens under the nitride film.

  In this manufacturing method, the light receiving side of the photodiode 65 is removed by etching using the silicon oxide film 73 as an etching stop material, and the silicon oxide film 99 formed thereon is subjected to a reflow process, whereby the corners of the silicon oxide film 99 in the opening are formed. The portion becomes a curved surface, and the optical characteristics of the convex lens shape can be imparted to the second silicon nitride film 81 formed later on the silicon oxide film 99.

It is a top view in the image pick-up part of the solid-state image sensing device concerning the present invention. It is a top view which shows the structural example of the principal part of the solid-state image sensor shown in FIG. It is sectional drawing in the A1-A2 part of FIG. It is sectional drawing of the modification of the structure shown in FIG. It is process explanatory drawing which represented the manufacturing procedure of the solid-state image sensor by embodiment by (a)-(f). It is process explanatory drawing which represented the manufacturing procedure of the solid-state image sensor by the modification 1 with (a)-(f). It is process explanatory drawing which represented the manufacturing procedure of the solid-state image sensor by the modification 2 with (a)-(f). It is process explanatory drawing which represented the manufacturing procedure of the solid-state image sensor by the modification 3 with (a)-(f). It is sectional drawing of the light-receiving part vicinity in the conventional solid-state image sensor. FIG. 10 is an enlarged cross-sectional view of a proximity portion between a transfer electrode and a light shielding film illustrated in FIG. 9.

Explanation of symbols

49 Semiconductor substrate 65 Photodiode (photoelectric converter)
69 Transfer electrode 77 Light-shielding tungsten film 79 First silicon nitride film 81 Second silicon nitride film 97 Opening 100 Solid-state imaging device

Claims (14)

Opposite the charge transfer channel on a semiconductor substrate on which a photoelectric conversion unit that generates a signal charge according to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit are formed. A solid-state imaging device in which a formed transfer electrode and a light-shielding film covering the transfer electrode are stacked,
A first silicon nitride film having a hydrogen shielding property is formed between the light shielding film and the transfer electrode;
A solid-state imaging device in which a second silicon nitride film having hydrogen permeability is formed on the photoelectric conversion unit. The solid-state imaging device according to claim 1,
The first silicon nitride film is a solid-state imaging device which is a CVD thermal nitride film. The solid-state imaging device according to claim 1 or 2,
The first silicon nitride film is a solid-state imaging device having a thickness of 20 to 100 nm. The solid-state imaging device according to any one of claims 1 to 3,
The solid-state imaging device, wherein the second silicon nitride film is a low temperature plasma CVD film. The solid-state imaging device according to claim 4,
The second silicon nitride film is a solid-state imaging device having a film thickness of at least 5 nm. It is a solid-state image sensing device according to any one of claims 1 to 5,
A solid-state imaging device in which the first silicon nitride film and the second silicon nitride film are bonded to each other, and the entire pixel region including the photoelectric conversion unit and the charge transfer channel is covered with a silicon nitride film. The solid-state image sensor according to any one of claims 1 to 6,
The transfer electrode is a solid-state imaging device formed of polysilicon. The solid-state imaging device according to any one of claims 1 to 7,
The transfer electrode is a solid-state imaging device formed of silicide. The solid-state imaging device according to any one of claims 1 to 8,
A solid-state imaging device in which a gate insulating film having a laminated structure of a silicon oxide film, a silicon nitride film, and a silicon oxide film is formed on the surface of the semiconductor substrate below the first and second silicon nitride films. The solid-state imaging device according to any one of claims 1 to 8,
A solid-state imaging device in which a gate insulating film made of a silicon oxide film is formed on the surface of the semiconductor substrate below the first and second silicon nitride films. A silicon oxide film, a silicon nitride film, a photoelectric conversion unit that generates a signal charge according to incident light, and a semiconductor substrate on which a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit is formed, A first step of forming a gate insulating film made of a stack of silicon oxide films;
A second step of forming the transfer electrode at a position on the gate insulating film facing the charge transfer channel;
A third step of forming an interelectrode insulating film surrounding the electrode;
A fourth step of forming a first silicon nitride film made of a CVD thermal nitride film so as to cover the interelectrode insulating film;
A fifth step of forming a light-shielding film having an opening in the photoelectric conversion unit;
A sixth step of etching and removing the light receiving side of the photoelectric conversion unit using the silicon oxide film on the photoelectric conversion unit as an etching stop material;
A seventh step of forming a second silicon nitride film, which is a low-temperature plasma CVD film, in a pixel region including the photoelectric conversion unit and the charge transfer channel;
A method for manufacturing a solid-state imaging device including: A single layer of a silicon oxide film is formed on a semiconductor substrate on which a photoelectric conversion unit that generates a signal charge according to incident light and a charge transfer channel for transferring the signal charge accumulated in the photoelectric conversion unit are formed. A first step of forming a gate insulating film;
A second step of forming the transfer electrode at a position on the gate insulating film facing the charge transfer channel;
A third step of forming an interelectrode insulating film surrounding the electrode;
A fourth step of forming a first silicon nitride film made of a CVD thermal nitride film so as to cover the interelectrode insulating film;
A fifth step of forming a light-shielding film having an opening in the photoelectric conversion unit;
A sixth step of etching and removing the light receiving side of the photoelectric conversion unit using the silicon oxide film on the photoelectric conversion unit as an etching stop material;
A seventh step of forming a second silicon nitride film, which is a low-temperature plasma CVD film, in a pixel region including the photoelectric conversion unit and the charge transfer channel;
A method for manufacturing a solid-state imaging device including: It is a manufacturing method of the solid-state image sensing device according to claim 11 or 12,
The method of manufacturing a solid-state imaging device, wherein the opening of the light shielding film in the fifth step is formed by etching in the sixth step. It is a manufacturing method of the solid-state image sensing device according to claim 11 or 12,
A method for manufacturing a solid-state imaging device, comprising an eighth step of reflowing the silicon oxide film between the sixth step and the seventh step.

Images