CN100452414C - Method and apparatus for acquiring physical information, method for manufacturing the apparatus - Google Patents

Abstract

Method and apparatus for acquiring physical information, method for manufacturing semiconductor device including array of a plurality of unit components for detecting physical quantity distribution, light-receiving device and manufacturing method therefor, and solid-state imaging device and manufacturing method therefore are provided. The method for acquiring physical information uses a device for detecting a physical distribution, the device including a detecting part for detecting an electromagnetic wave and a unit signal generating part for generating a corresponding unit signal on the basis of the quantity of the detected electromagnetic wave. The detecting part includes a stacked member having a structure in which a plurality of layers having different refractive indexes between the adjacent ones and each having a predetermined thickness is stacked, the stacked member being provided on the incident surface side to which the electromagnetic wave is incident and having the characteristic that a predetermined wavelength region component of the electromagnetic wave is reflected, and the remainder is transmitted.

Description

Obtain the manufacture method of method, device and the device of physical message

The cross reference of related application

The application comprises the Japanese patent application JP2004-358139 that submits in Japan Patent office with on December 10th, 2004, on July 20th, 2005 at the Japanese patent application JP 2005-209409 of Japan Patent office submission and the relevant theme of submitting in Japan Patent office on December 22nd, 2004 of Japanese patent application JP 2004-371602, incorporates its full content here into as a reference.

Technical field

The present invention relates to a kind of manufacture method of obtaining the side and the device of physical message and comprising the semiconductor device of a plurality of arrays that are used to survey the unit assembly that physical quantity distributes.More specifically, the present invention relates to a kind of signal and obtain technology, it is applicable to and uses semiconductor device to survey the solid imaging element that physical quantity distributes, this semiconductor device comprises a plurality of to the array from the input of the electromagnetic wave of outside, for example light and radiosensitive unit assembly, can be used as the signal of telecommunication and reads thereby distributed by the physical quantity that the unit assembly is converted to the signal of telecommunication.Especially, the present invention relates to a kind of image device, it can make wavelength component (for example, the infrared light) imaging except visible light.

The invention still further relates to a kind of light receiving element and solid imaging element, be included in the optical-electrical converter that forms in the semiconductor layers such as silicones, compound semiconductor and the manufacture method of various devices separately.

Background technology

The semiconductor device application of surveying the physical quantity distribution is in various fields, and semiconductor device comprises line or the matrix array such as the responsive a plurality of units assembly (for example pixel) of the electromagnetic physical quantity variation of light and radiation to for example importing from the outside separately.

For example, in the video-unit field, use the variation of CCD (charge coupled device), MOS (metal-oxide semiconductor (MOS)) or CMOS (complementary metal oxide semiconductors (CMOS)) solid imaging element detection as the light (an electromagnetic example) of an example of physical quantity.In these devices, the physical quantity distribution that is converted to the signal of telecommunication by unit assembly (for example pixel in the solid imaging element) is read as the signal of telecommunication.

For example, solid imaging element is surveyed from for example light or the radiation of electromagnetic wave of outside input, uses photodiode as the optical-electrical converter (light receiving element that is provided with in the imaging moiety (pixel portion) of device; Light-sensitive device), thus produce and collect signal charge.The signal charge of collecting (photoelectron) is read as image information.

In recent years, proposed to be used for visual light imaging and infrared light imaging structure (referring to, for example open No.2004-103964,10-210486 of Japanese Unexamined Patent Application, 2002-369049,6-121325,9-166493,9-130678 and 2002-142228).For example, prepare infrared bright spot in advance, make and to survey the position of the infrared bright spot in the visible images by following the tracks of infrared bright spot.In addition, for example,, can obtain distinct image by the infrared radiation imaging even in the evening that does not have visible light.And, can improve sensitivity by adopting the infrared light except that visible light.

Disclosed structure is a monolithic type among the open No.2004-103964 of Japanese Unexamined Patent Application, and it is applied in the semi-conductive depth direction absorption coefficient with wavelength change.

The disclosed structure multi-disc type of respectively doing for oneself among Japanese Unexamined Patent Application open No.10-210486,2002-369049 and the 6-121325, its application comprises the wavelength resolution optical system of wavelength separated mirror (separation mirror) and prism as fore optics system, makes to receive visible light and infrared light by each image device.

Disclosed structure is a monolithic type among the open No.9-166493 of Japanese Unexamined Patent Application, and it uses rotation wavelength resolution optical system as fore optics system, makes to receive visible light and infrared light by same image device.For example, when inserting/extracting infrared absorption filter by the rotating machinery device except that filter (cut filter), output is not by the visible coloured image of near infrared light and infrared light influence, yet when extracting infrared absorption filter, will export image with the luminous intensity that comprises visual intensity and near infrared light intensity except that filter.

The aperture optical system (diaphragm optical system) that disclosed structure applications has the wavelength decomposition function among the open No.9-130678 of Japanese Unexamined Patent Application makes to receive visible light and infrared light by same image device as fore optics system.

Disclosed structure comprises visible light and the light activated image device of near-infrared among the open No.2002-142228 of Japanese Unexamined Patent Application, wherein on pixel, arranged regularly and had four types colored filter of light-filtering characteristic separately, and calculated independent determine visible coloured image and near infrared light image by the output matrix of each pixel of having arranged four types of colored filters on it.

Solid imaging element comprises the optical-electrical converter that is formed in the semiconductor layer.

Therefore, solid imaging element has the problem of the so-called dark current that is produced by the surface level (surface level) of the semiconductor layer that has wherein formed optical-electrical converter.

Shown in the potential diagram of Figure 60 A, the electronics of catching mainly due to the surface level place is discharged into conduction band by heat, and is therefore moved to the n N-type semiconductor N district of the photodiode that constitutes each optical-electrical converter by the electric field of surface depletion layer, and produces dark current.

For example, in the semiconductor layer that silicon constitutes, band gap (band gap) is 1.1eV, and surface level (Fermi (Fermi) energy level) is positioned at because Ba Ding restriction (Bardeem limit) band gap is divided into 2: 1 position.

Therefore, prevent that the potential barrier that electronics is caught by surface level from being 0.7eV.

Therefore, in order to reduce the dark current that produces by surface level, be applied in the method that forms the p+ layer on the surface of photodiode (referring to, the open No.2002-252342 of Japanese Unexamined Patent Application for example, Fig. 5).

This method has suppressed dark current to a certain extent.

That is, shown in the potential energy diagram of Figure 60 B, because the existence of p+ layer prevents that the potential barrier that electronics is caught by surface level from becoming 1.0eV.In other words, compare potential barrier with the situation that does not wherein have the p+ layer and increased about 0.3eV, and can reduce the number of the electronics of heat release thus, to reduce dark current.

When the p+ layer is set on surface of silicon, to compare with the situation that does not wherein have the p+ layer, the amount of the dark current of being estimated by Fermi-Di Lake (Fermi-Dirac) distribution function under the room temperature (T=300K) has reduced by four.

Fermi-Di Lake distribution function is represented by following equation 10:

Equation 10

$f(E,T)=\frac{1}{1+e^{\frac{E-E_P}{\mathrm{kT}}}}$

Wherein E is an energy, E _FBe Fermi energy, T is an absolute temperature, and k is a Maxwell constant, and e is a natural logrithm, and E-E _FValue corresponding to potential barrier.

Summary of the invention

Figure 53 A and 53B are the topology example figure of disclosed transducer among the open No.2004-103964 of Japanese Unexamined Patent Application, and wherein Figure 53 A is the accompanying drawing of the optical absorption spectra characteristic of semiconductor layer, and Figure 53 B is the schematic diagram of the combining structure of device.

In this structure, the semi-conductive absorption coefficient of light of Si (silicon) reduces according to the order of blue, green, red and infrared light, shown in Figure 53 A.Promptly, for the blue light that contains among the incident light L1, green glow, ruddiness and infrared light, by the position correlation (dependency) that is applied in semiconductor depth direction medium wavelength, on depth direction, be provided for surveying the layer of visible light (blue, green, red) and infrared light respectively according to the order shown in Figure 53 B from the semi-conductive surface of Si.

Yet, in the open No.2004-103964 of Japanese Unexamined Patent Application in the disclosed structure, it utilizes absorption coefficient with wavelength change, when ruddiness and green glow when being used to survey the layer of blue light to a certain extent by this layer absorption, though do not have in theory thus to reduce the light quantity that detects, be detected as blue light.Therefore, even initialize signal does not have blue light, and green glow and red signal light meet and form the blue light signal, thereby produce glitch and therefore can not reach enough colorrendering qualitys.

In order to address this problem, preferably to handle and carry out the correction of three primary colors integral body, and therefore counting circuit is set separately by calculated signals.Therefore, circuit layout becomes complicated and increases on scale, and cost has also increased.And, for example, when one of three primary colors are saturated, can not determine saturated light initial value and cause mistake in computation.The result is that signal processing causes the color of generation to be different from priming color.

Shown in Figure 53 A, most of semiconductors have the absorption sensitiveness to infrared light.Therefore, for example in using the semi-conductive solid imaging element of Si (graphical sensory device), as the example of subtractive filter, preferably the infrared absorption filter of making at the front of transducer insertion glass removes filter.

Therefore, for by only accepting infrared light or visible light and infrared light, preferably remove the filtering ratio that infrared absorption filter removes filter or reduces infrared light as the signal imaging.

Yet in this case, infrared light mixes with visible light and incides on the optical-electrical converter, thereby produces the visible images that shade of color (tone) is different from initial tone.Therefore producing correct visible images and correct infrared light image self (or mixing of infrared light and visible light) separately simultaneously may be very difficult.

Except that above-mentioned problem, in conventional solid imaging element, use infrared absorption filter and also weakened visible light to a certain extent, thereby lowered sensitivity except that filter.Quote infrared absorption filter and also increased cost except that filter.

In Japanese Unexamined Patent Application open No.10-210486,2002-369049 and the disclosed structure of 6-121325, increased the scale of fore optics system owing to the wavelength resolution optical system that comprises the speculum that is used for wavelength separated and prism.

In the structure of the open No.9-166493 of Japanese Unexamined Patent Application, insert/extract the scale that mechanical device has increased device owing to infrared absorption filter removes filter, and infrared absorption filter can not automation mechanized operation except that filter.

In the structure of the open No.9-130678 of Japanese Unexamined Patent Application, increased the scale of device owing to aperture optical system with wavelength decomposition function.In addition, though can obtain infrared light image and visible images simultaneously, the electric composite signal of visible images and infrared light image only exported from imageing sensor, can not only export visible images or infrared light image thus.

On the other hand, in the structure of the open No.2002-142228 of Japanese Unexamined Patent Application, the colored filter of quoting four types carries out wavelength separated.Therefore, this structure has the problem of algorithm process but does not have the problem that the fore optics system scale increases among Japanese Unexamined Patent Application open No.10-210486,2002-369049,6-121325,9-166493 and the 9-130678.Promptly, in the structure of the open No.2002-142228 of Japanese Unexamined Patent Application, visible coloured image and near infrared light image that the output matrix operation institute of the pixel by having arranged four types colored filter with each light-filtering characteristic on it respectively determines separately, and can separate visible images and infrared light image thus and export simultaneously.Yet,, significantly increased arithmetic processing thus on the whole even, between visible light and infrared light component, carry out arithmetic processing when having obtained visible images.

Expectation provides a kind of manufacture method of the device that can solve the new mechanical device of at least one the problems referred to above and be adopted in this mechanical device.

According to one embodiment of present invention, provide a kind of image device, wherein used same imageing sensor and independently obtain visible coloured image and near infrared light image with new mechanical device.

According to another embodiment of the invention, a kind of mechanical device is provided, wherein when using same imageing sensor while imaging visible images and infrared light image, solved owing to remove infrared absorption filter and removed the problem that shade of color that filter produces is different from initial tone, made visual light imaging and the infrared and ultraviolet light imaging that has correct shade of color simultaneously.

According to another embodiment of the invention, provide a kind of mechanical device, its thick infrared absorption filter that is used for solving owing to making at the imageing sensor application glass of routine removes the problem that filter increases cost.

On the other hand, when pixel becomes fine gradually, reduced the light quantity that photodiode received of each pixel, therefore also reduced semaphore, thereby reduced the S/N ratio relatively.

Therefore, even, when being used to reduce by four dark current, can not obtain gratifying S/N ratio on surface of silicon when the p+ layer is set.For example, even in night sky imaging, the noise of similarity will appear in the image that obtains.

In this case, when the amount of incident light is very little,, increase the gain of figure signal to increase signal strength signal intensity by amplifier etc. usually in order to remedy muting sensitivity.Yet, also increased simultaneously the intensity of noise as signal strength signal intensity, cause occurring in the image significant noise.

Owing to signal strength signal intensity will reduce along with miniaturization (refining) in future, still be not enough by the noise that dark current produces only by the reduction of p+ layer is set on surface of silicon substrate.

Therefore, expectation is provided with a kind of new type mechanical device that is used to guarantee enough S/N ratios.

In the light receiving element that comprises the photodiode that is formed on the optical-electrical converter in the semiconductor layer, and in figure attitude image device, the miniaturization that develops owing to light receiving element has reduced the S/N ratio, just can not detect the signal that is obtained by optical-electrical converter thus satisfactorily.

According to another embodiment of the invention, be provided with a kind of can be by reducing the noise that produces by dark current with the light receiving element of guaranteeing gratifying S/N ratio, manufacturing process, solid imaging element and the manufacturing process thereof of light receiving element.

The method and apparatus that obtains physical message according to an embodiment of the invention is used a kind of stacked film, it has wherein stacked a plurality of structure with layer of different refractivity, being used for wavelength separated is transmission peak wavelength ranges of components and reflected wavelength range component, makes the exploring block that separates obtain the signal of two kinds of components independently or side by side.

In other words, the method application of obtaining physical message is the device that the basis is used for the detection physical quantity distribution of predetermined purpose with the unit signal, this device comprises, as the unit assembly, being used to survey electromagnetic exploring block and being used for the electromagnetic wave amount that detects is the unit signal production part that the basis produces corresponding unit signal and output unit signal, and the unit assembly is arranged on the same substrate with predetermined order.Exploring block comprises the stacked member (member) on the incidence surface side of incident electromagnetic wave disposed thereon, stacked member has the predetermined wavelength range component of reflection electromagnetic wave and the characteristic of transmission residual term (remainder), also has the wherein stacked adjacent structure that has different refractivity between the two and have the multilayer of predetermined thickness separately.

Exploring block is surveyed the transmission peak wavelength ranges of components through stacked member transmission, obtains the physical message that is used for predetermined purpose on the basis of the transmission peak wavelength ranges of components unit signal that obtains from the unit signal production part.

Term " residual term " refers in fact not contain at least the component of reflected wavelength range component, but not gets rid of all wavelengths component of reflected wavelength range component.Statement " does not in fact contain the reflected wavelength range component " and refers to not exist basically the influence of reflected wavelength range component and the influence of reflected wavelength range component minute quantity to exist.This is because for the transmission peak wavelength side, is enough to obtain the signal that the wherein influence of reflected wavelength range is ignored.In addition, for reflected wavelength range, be enough to obtain the signal that the wherein influence of transmission peak wavelength scope is ignored.

The device that obtains physical message is applicable to the method for obtaining physical message of carrying out.This device comprises the stacked member on the incidence surface side of arranging the exploring block of incident electromagnetic wave thereon, with in the information processing unit that detects by sensing element and on the basis of the unit signal of the transmission peak wavelength ranges of components of stacked member transmission, obtain the physical message that is used for specific purpose, on the basis of transmission peak wavelength ranges of components, obtain unit signal from the unit signal production part.Stacked member has the wherein stacked adjacent structure that has different refractivity between the two and have the multilayer of predetermined thickness separately, also has the predetermined wavelength range component of reflection electromagnetic wave and the characteristic of transmission residual term.Stacked member can be separated with exploring block, but preferred and exploring block integrates.

The manufacture method of semiconductor device is applicable to and makes above-mentioned device according to an embodiment of the invention.The method comprising the steps of: form the semiconductor element layer with exploring block and unit signal production part on Semiconductor substrate, on semiconductor element layer, be formed for forming the wiring layer of holding wire, be used for reading unit signal from the unit signal production part, with the stacked film of formation on wiring layer, this stacked film have wherein stacked adjacent have different refractivity between the two and have separately predetermined thickness multilayer structure and have the predetermined wavelength range component of reflection electromagnetic wave and the characteristic of transmission residual term.

In order to make the detection of reflected wavelength range component become possibility, this method also comprise remove regularly with corresponding to the step of the stacked membrane portions of a plurality of exploring block position alignment of wavelength separately.In this case, one of a plurality of exploring blocks are surveyed the transmission peak wavelength ranges of components through stacked film transmission, and another surveys the reflected wavelength range component without stacked film transmission in a plurality of exploring block.From remove regularly with corresponding to the viewpoint of the stacked membrane portions of a plurality of exploring block position alignment of wavelength separately, preferred stacked film and exploring block integrate but not are separated with exploring block.

In the application of colour imaging, this method also is included on the stacked film of aiming at the location of pixels of corresponding wavelength separately and is formed for the step of the optical component of wavelength separately, is used for the predetermined wavelength component of transmission transmission peak wavelength ranges of components.From with the exploring block position alignment of corresponding wavelength separately be formed for the viewpoint of the optical component of wavelength separately, be preferred for the optics of wavelength separately and stacked film and exploring block and integrate but not be separated with stacked film and exploring block.

Other characteristic of various details.

For example, in order to obtain image about the reflected wavelength range component, for example with respect to infrared light such as the transmission peak wavelength ranges of components of visible light, be used for the reflected wavelength range component, electromagnetic wave is not provided with stacked member on the incidence surface sidepiece of the exploring block of its incident, thereby survey the reflected wavelength range component by exploring block, and therefore can obtain to be used for the physical message of the second predetermined purpose on the basis of the unit signal that from the unit signal production part, obtains the reflected wavelength range component.

In addition, can select and export based on first physical message of the unit signal of transmission peak wavelength ranges of components with based on second physical message of the unit signal of reflected wavelength range component, or can export both simultaneously.

And, can on each light incident side of a plurality of exploring blocks that are used to survey the transmission peak wavelength ranges of components, optical component be set, be used for the transmission peak wavelength ranges of components is separated into each wave-length coverage component, and each transmission peak wavelength ranges of components can be surveyed by a plurality of exploring blocks respectively.In this case, can be by the unit signal of each transmission peak wavelength ranges of components of combination combination, it obtains from the unit signal production part, obtains other physical message on the transmission peak wavelength ranges of components.For example, can use, as optics, wherein the transmitted light in visible-range have the three primary colors filter of trichromatic wavelength component or wherein the transmitted light in the visible-range have the complementary colours filter of trichromatic complementary colours separately, carry out the imaging coloured image.

In order to obtain the signal of reflected wavelength range component, exploring block can load (for example wavelength component of one of three primary colors) in whole or in part of (loaded) reflected wavelength range component and transmission peak wavelength ranges of components simultaneously, and only wherein the signal of the negligible reflected wavelength range component of the influence of transmission peak wavelength ranges of components can be obtained by different algorithm operatings.Replacedly, can on the incidence surface side of the exploring block that is used for the reflected wavelength range component, be provided for the optical component of Transflective wave-length coverage component and filtering transmission peak wavelength ranges of components, to prevent the influence of transmission peak wavelength ranges of components.

When the signal that obtains the reflected wavelength range component equally when forming image, the exploring block that the common aligning part that is preferred for surveying the exploring block of transmission peak wavelength ranges of components is used to survey the reflected wavelength range component replaces.In this case, can influence resolution by the arrangement of each exploring block.

From this viewpoint, for example, when the conventional color image resolution based on the transmission peak wavelength ranges of components is awarded bigger attention, can in alternating graph (checked pattern), arrange the detecting element (being typically the pixel of color G) that a plurality of detecting elements that are used for being of value to each color that forms conventional coloured image are used to survey the predetermined wavelength component.On the other hand, when the image resolution ratio based on the reflected wavelength range component is awarded bigger attention, can in alternating graph, be formed with the exploring block that benefits this image of formation.

When arranging pixel with the two-dimensional lattice form, with respect to the square lattice of wherein on the direction that is parallel and perpendicular to vertical and horizontal read direction respectively, arranging pixel, the oblique lattice that (is typically about 45 degree) more preferably at a predetermined angle and rotates.This is because the increase of picture element density has further increased resolution on these directions vertically and on the horizontal direction.

Light receiving element according to an embodiment of the invention comprises the optical-electrical converter that is formed in the semiconductor layer, with the single crystalline layer that is formed on the semiconductor layer part that has wherein formed optical-electrical converter at least, single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.

In above-mentioned light receiving element, single crystalline layer is formed on the semiconductor layer part that has wherein formed optical-electrical converter at least, and single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.Therefore, single crystalline layer has wideer band gap, has increased thus to prevent that electronics is in the potential barrier of surface level, thereby has reduced the dark current by electron production.

The manufacture method that is included in the light receiving element that is formed with optical-electrical converter in the semiconductor layer comprises, at least the semiconductor layer that has formed optical-electrical converter is therein partly gone up the step that forms single crystalline layer, and single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.

In the manufacture method of light receiving element, the method comprising the steps of: the semiconductor layer that has formed optical-electrical converter is therein at least partly gone up the formation single crystalline layer, single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer, increased by single crystalline layer and to have prevented that electronics is in the potential barrier of surface level, thereby reduced dark current by electron production.

Solid imaging element according to an embodiment of the invention comprises the optical-electrical converter that is formed in the semiconductor layer, with the single crystalline layer that is formed on the semiconductor layer part that has wherein formed optical-electrical converter at least, single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.

In above-mentioned solid imaging element, single crystalline layer is formed on the semiconductor layer part that has wherein formed optical-electrical converter at least, and single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.Therefore, single crystalline layer has wideer band gap, has increased thus to prevent that electronics is in the potential barrier of surface level, thereby has reduced the dark current by electron production.

Be included in the manufacture method of the solid imaging element that is formed with optical-electrical converter in the semiconductor layer, comprise step: the semiconductor layer that has formed optical-electrical converter is therein at least partly gone up the step that forms single crystalline layer, and single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer.

In the manufacture method of solid imaging element, the method comprising the steps of: the semiconductor layer that has formed optical-electrical converter is therein at least partly gone up the formation single crystalline layer, single crystalline layer is made of the material with band gap wideer than the band gap of semiconductor layer, increased by single crystalline layer and to have prevented that electronics is in the potential barrier of surface level, thereby reduced dark current by electron production.

According to one embodiment of present invention, application has wherein stacked structure with multilayer of different refractivity transmission peak wavelength ranges of components and reflected wavelength range component is separated into each wavelength, and is surveyed the signal of two kinds of components by each exploring block.

Therefore, (for example, imageing sensor) can obtain the physical message on the transmission peak wavelength ranges of components in single semiconductor device, and wherein the influence of reflected wavelength range component can be ignored.In this case, for example, the optics of the costliness that is used for the filtering infrared light that glass makes can be set, it is used for filtering with respect to the reflected wavelength range component infrared light for the visible light of transmission peak wavelength ranges of components example.Therefore, be not applied on the semiconductor depth direction absorption coefficient, and therefore can not occur because the problem of the colorrendering quality that this variation causes with wavelength change.

When surveying transmission peak wavelength ranges of components and reflected wavelength range component respectively to obtain signal when output of two kinds of components simultaneously, with respect to the transmission peak wavelength ranges of components, the reflected wavelength range component is removed by stacked membrane filtration earlier.Therefore, it is such to be different among the open No.2002-142228 of Japanese Unexamined Patent Application disclosed structure, can not carry out the transmission peak wavelength ranges of components of transmission peak wavelength ranges of components signal of the wave-length coverage components influence that is used to obtain at all not to be reflected and the arithmetical operation between the reflected wavelength range component.

Certainly, can be discretely or side by side survey the signal of transmission peak wavelength ranges of components and reflected wavelength range component, can for example use the structure of surveying red-ultraviolet light and visible light discretely thus, carry out visual light imaging and infrared and ultraviolet light imaging simultaneously.In this case, when further visible light being divided into the signal component of primary colors and surveying, can obtain having the visible images and the infrared-ultraviolet light image of correct shade of color simultaneously.

And, increased by single crystalline layer and to have prevented that electronics is in the potential barrier at surface level place, thereby reduced dark current by electron production.For example, can significantly reduce by 12 in dark current, with the remarkable signal S/N ratio that improves incident light.

Therefore, under image-forming condition, even, can not obtain not having the image of remarkable noise for increasing sensitivity when signal gain is arranged as bigger value such as a small amount of incident light in the darkroom etc.

And, even in having the image device of muting sensitivity, do not consider incident light amount and only the amplification by amplifier just can obtain high-quality image.

And, even when having increased the amount of incident light, also can obtain enough S/N ratios by selected element, and thus only the amplification of the amplifier by being used to compensate muting sensitivity just can obtain not having the gratifying image of remarkable noise.

Therefore, can increase number of pixels in the solid imaging element by the miniaturization element, and can reduce the size that each comprises the optics and the solid imaging element of light receiving element.

Description of drawings

Fig. 1 is that explanation is used to use stacked dielectric film with the accompanying drawing of electromagnetic wave chromatic dispersion (dispersing) for the spectral image sensor principle of predetermined wavelength;

Fig. 2 is the principle exemplary plot that the essential structure of the spectral image sensor of using stacked dielectric film is described;

Fig. 3 is the diagrammatic sketch that the essential structure of the spectral image sensor shown in Fig. 2 is applied to an example of multi-wavelength spectrum separation structure;

Fig. 4 is the structure chart of basic principle of the method for designing of the stacked film of explanation;

Fig. 5 is the reflectance spectrum chart (atlas) of basic principle of the method for designing of the stacked film of explanation;

Fig. 6 is the reflectance spectrum chart of basic principle of the method for designing of the stacked film of explanation;

Fig. 7 A and 7B are the exemplary plot (schematic diagram of reflectance spectrum) of the condition of reflection kernel wavelength X;

Fig. 8 is the reflectance spectrum chart of the condition of example reflection kernel wavelength X;

Fig. 9 is the reflectance spectrum chart of the condition of example reflection kernel wavelength X;

Figure 10 is the topology example figure that uses the spectral image sensor that the single wave spectrum of stacked film separates corresponding to first embodiment according to the invention;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to first embodiment according to the invention to Figure 11;

To be example use the reflectance spectrum chart (concrete reflectance spectrum chart) of the spectral image sensor that single wave spectrum of stacked film separates corresponding to first embodiment according to the invention to Figure 12;

Figure 13 is the topology example figure that uses the spectral image sensor that the single wave spectrum of stacked film separates corresponding to first embodiment according to the invention;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to first embodiment according to the invention to Figure 14;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to second embodiment according to the present invention to Figure 15;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to second embodiment according to the present invention to Figure 16;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to second embodiment according to the present invention to Figure 17;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to second embodiment according to the present invention to Figure 18;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 19;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 20;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 21;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 22;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 23;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 3rd embodiment according to the present invention to Figure 24;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 4th embodiment according to the present invention to Figure 25;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 4th embodiment according to the present invention to Figure 26;

To be example use the structure chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 5th embodiment according to the present invention to Figure 27;

To be example use the reflectance spectrum chart of the spectral image sensor that single wave spectrum of stacked film separates corresponding to the 5th embodiment according to the present invention to Figure 28;

Figure 29 A and 29B are applied in circuit (corresponding to R, G, B and infrared light IR) exemplary plot in the IT_CCD imageing sensor with stacked film;

Figure 30 is applied in circuit (corresponding to visible light VL and infrared light IR) exemplary plot in the IT_CCD imageing sensor with stacked film;

Figure 31 A and 31B are applied in circuit (corresponding to R, G, B and infrared light IR) exemplary plot in the cmos image sensor with stacked film;

Figure 32 is applied in circuit (corresponding to visible light VL and infrared light IR) exemplary plot in the cmos image sensor with stacked film;

Figure 33 A～33F is the exemplary plot of an example of spectral image sensor manufacturing process;

Figure 34 is the structure chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 35 is the structure chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 36 is the structure chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 37 is the structure chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 38 is the exemplary plot corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 39 is the reflectance spectrum chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 40 is the reflectance spectrum chart of example corresponding to the spectral image sensor of single wave spectrum separation of using stacked film according to a sixth embodiment of the present;

Figure 41 A, 41B and 41C are the diagrammatic sketch of an example of color separated filter arrangement;

Figure 42 is the exemplary plot (perspective view) with example of the CCD solid imaging element structure that the color separated filter shown in Figure 41 A, B and the C arranges;

Figure 43 be side by side two wavelength components of separate imaging, be the diagrammatic sketch (sectional structure chart) of an example of the CCD solid imaging element structure of infrared light and visible light;

Figure 44 A, 44B and 44C are the diagrammatic sketch of other example of color separated filter arrangement;

Figure 45 is the exemplary plot (perspective view) with example of the CCD solid imaging element structure that the color separated filter shown in Figure 44 A, B and the C arranges;

Figure 46 A and 46B are that painstakingly (conscious) reduces the exemplary plot of first example that the pixel of resolution arranges;

Figure 47 is the diagrammatic sketch of an example of the transmitted spectrum characteristic of black filter;

Figure 48 A and 48B are the exemplary plot that painstakingly reduces by second example that the pixel of resolution arranges;

Figure 49 A, 49B and 49C are the exemplary plot that painstakingly reduces the 3rd example that the pixel of resolution arranges;

Figure 50 A and 50B are the exemplary plot that painstakingly reduces the 4th example that the pixel of resolution arranges;

Figure 51 A and 51B are the exemplary plot that painstakingly reduces the 5th example that the pixel of resolution arranges;

Figure 52 A and 52B are the exemplary plot that painstakingly reduces the 6th example that the pixel of resolution arranges;

Figure 53 A and 53B are day original exemplary plot of examining disclosed sensor construction among the open No.2004-103964 of patent application;

Figure 54 is the schematic diagram (floor map) that illustrates according to the solid imaging element structure of the 7th embodiment of the present invention;

Figure 55 is the cutaway view of the solid imaging element shown in Figure 54;

Figure 56 is the potential energy distribution diagrammatic sketch;

Figure 57 is the schematic diagram (cutaway view) that solid imaging element structure according to another embodiment of the invention is shown;

Figure 58 illustrates the schematic diagram (cutaway view) that comprises by the solid imaging element structure of the single crystalline layer that forms with the method diverse ways that is used for the figure attitude image device shown in Figure 57;

Figure 59 is the schematic diagram (floor map) that solid imaging element structure according to another embodiment of the invention is shown;

Figure 60 A is the diagrammatic sketch of the potential energy distribution of conventional solid imaging element;

Figure 60 B is the diagrammatic sketch that comprises the potential energy distribution of the structure that is formed on lip-deep p+ layer.

Embodiment

Below, describe embodiments of the invention with reference to the accompanying drawings in detail.

" principle of stacked dielectric film imageing sensor "

Fig. 1 uses stacked dielectric film with the principle exemplary plot of electromagnetic wave chromatic dispersion (dispersing) for the spectral image sensor of predetermined wavelength.Here, become the spectral image sensor of predetermined wavelength to do a description as an electromagnetic example by chromatic dispersion to light wherein.

In Fig. 1, the stacked dielectric film of reference number 1 expression, reference number 10 expression spectral filters.

As shown in Figure 1, stacked dielectric film 1 is stacked member, its have stacked adjacent have between the two difference (refringence is Δ n) refractive index n j (wherein j be 2 or bigger positive integer) and have the structure of the multilayer of predetermined thickness separately.The result is, stacked dielectric film 1 have that electromagnetic predetermined wavelength range component as described below is reflected and residual term by the characteristic of transmission.

Counting constitutes the dielectric layer 1_j of stacked dielectric film 1, for example the thick-layer except that both sides on (layer 1_0 and 1_k) from ground floor to k layer side.Stacked dielectric film 1 roughly comprises the layer the thick-layer on both sides (layer 1_0 and 1_k).

When light incides on the stacked dielectric film 1 with said structure, because the interference in the stacked dielectric film 1 causes that reflectivity (or transmission coefficient) presents the correlation certain to wavelength X.Along with increasing this effect, the refractive indices n of light becomes remarkable.

Especially, when stacked dielectric film 1 have periodic structure or such as the certain condition of the incident light L1 of white light etc. under (for example, each layer thickness is d～λ/4n condition of d), effectively increased particular range of wavelengths reflection of light rate, mainly produce reflected light component L2.That is, reduced transmission coefficient.Reduce other wave-length coverage reflection of light rate, mainly produce transmitted light component L3.In other words, can increase transmission coefficient.

Wavelength X is the centre wavelength of certain wave-length coverage, and n is the refractive index of this layer.In one embodiment of the invention, by utilizing reflectivity (or transmission coefficient) in the stacked dielectric film 1 that the correlation of wavelength is realized spectral filter 10.

The basic structure of the spectral image sensor of the stacked dielectric film of＜application 〉

Fig. 2 is the principle exemplary plot of basic structure of using the spectral image sensor of stacked dielectric film.Fig. 2 illustrates incident light is become infrared light IR (infrared) and visible light VL (visible light) by chromatic dispersion a example.Forming stacked dielectric film 1 makes the infrared light of being longer than the infra-red range medium wavelength λ (the dominant wavelength side is longer than 780nm) of visible light VL at wavelength have high reflectance.In this case, filtering infrared light IR.When not forming this stacked dielectric film 1, can transmitted infrared light IR.

Because stacked dielectric film 1 comprises multilayer, at least two types member (material layer) is applied to each dielectric layer 1_j.When using three when more multi-layered, each dielectric layer 1_j can use different material layers or alternately or with stacked two (or more) layers of any desired order.Replacedly, stacked dielectric film 1 can comprise first and second base materials and can part be replaced by the 3rd (or more) material layers.This will specifically describe below.

The multi-wavelength spectrum image sensor architecture of the stacked dielectric film of＜application 〉

Fig. 3 is the diagrammatic sketch that the basic structure that will comprise the spectral image sensor 11 of the spectral image sensor 10 shown in Fig. 2 is applied to an example of multi-wavelength spectrum separated structures.

In Fig. 3, the stacked dielectric film of reference number 1 expression; Reference number 11, spectral image sensor; Reference number 12, the unit picture element matrix.

Describe with reference to Fig. 2,, do not form stacked dielectric film 1 transmitted infrared light IR by forming stacked dielectric film 1 filtering infrared light IR.Use this fact, rule is removed and stacked dielectric film 1 part corresponding to a plurality of exploring blocks (for example photodiode) position alignment of each wavelength, exploring block component unit picture element matrix 12.That is, in each pixel (unit), feasible only visible light VL imaging and only the infrared light IR imaging or the only mixed light imaging of visible light VL imaging and infrared light IR and visible light VL simultaneously simultaneously of filtering or not filtering infrared light.

Can not be subjected to infrared light IR to obtain monochrome image or coloured image by day with influencing, or can carry out infrared light IR imaging at night.As requested, can export two kinds of images simultaneously.In this case, can not be subjected to visible light VL that the image of infrared light IR is only arranged with influencing by day.

In other words, in the spectral image sensor 11 that separates corresponding to multi-wavelength spectrum, on the photodiode of each the pixel major part that constitutes the regularly arranged unit picture element matrix 12 of pixel wherein, be formed for the stacked dielectric film 1 of reflects infrared light IR, on the basis of the picture element signal that obtains from pixel, be not subjected to infrared light IR to obtain the monochrome image of visible light VL separately with influencing.Be different from disclosed structure among the open No.2002-142228 of Japanese Unexamined Patent Application, can between visible light VL and infrared light IR, do not carry out arithmetical operation and obtain roughly not to be subjected to the monochrome image of the visible light VL that infrared light IR influences.

And, as an example that is used for the wave-length coverage component is separated into the optical component of presetted wavelength ranges of components, can form thereon on each photodiode of stacked dielectric film 1 and be arranged on the colored filter 14 that has the predetermined wavelength transmissison characteristic in the visible light VL scope.In this case, can roughly not be subjected to infrared light IR only to obtain the image of particular range of wavelengths in the visible light VL scope with influencing.

Regularly arranged when visible-range has the colored filter 14x of different wave length transmissison characteristic when on a plurality of photodiodes of corresponding each wavelength (each color), aim at ground on the whole with each photodiode positions, photodiode component unit picture element matrix 12, visible light VL scope can be separated into each wavelength (each color).Therefore, by can only obtain roughly not to be subjected to the coloured image (visible coloured image) of the visible light VL that infrared light IR influences based on synthetic processing from the Pixel Information of each colored pixels.Be different from disclosed structure among the open No.2002-142228 of Japanese Unexamined Patent Application, can between visible light VL and infrared light IR, do not carry out arithmetical operation and obtain roughly not to be subjected to the coloured image of the visible light VL that infrared light IR influences.

In same image device (spectral image sensor 11), for example when unit picture element matrix 12 comprises the pixel that does not wherein form stacked dielectric film 11, can obtain monochrome or the coloured image of visible light VL and infrared light IR all the time separately by the matrix manipulation of pixel output.In addition, because part is removed the stacked dielectric film 1 that is formed on the whole on each photodiode, be different from wherein on image device, to arrange and comprise each stacked dielectric film 1 but do not comprise in the dissociated optical member situation of stacked dielectric film 1, not the problem that can occur aiming at.

For example, can obtain only roughly not being subjected to the image of the mixed light of the image (monochrome image or coloured image) of the visible light VL that infrared light IR influences and infrared light IR and visible light VL simultaneously.In addition, can only obtain roughly not being subjected to the image of the infrared light IR that visible light VL influences by the only synthetic processing of the component of the mixed light of visible light VL and infrared light IR and visible light VL (concrete, difference handle).

Consider that term " roughly unaffected " finally depends on people's eyesight, can exist common eye-observation to a certain extent less than the significantly influence of the light of difference.In other words, for infrared light IR side, may obtain the wherein negligible infrared image of transmission peak wavelength scope (visible light VL) (example of physical message).For visible light VL side, may obtain the wherein negligible normal image of reflected wavelength range component (infrared light IR) (example of physical message).

Colored filter 14 can be to be used for blue color component B (for example to be about 1 at wavelength X=400～500nm place transmission coefficient, be roughly zero at other wavelength place transmission coefficient), green color component G (for example is about 1 at wavelength X=500～600nm place transmission coefficient, be roughly zero at other wavelength place transmission coefficient) or red component R (for example be about 1 at wavelength X=600～700nm place transmission coefficient, be roughly zero at other wavelength place transmission coefficient) primary color filters, B component, G and R are visible light VL (the three primary colors components of wavelength X=380～780nm).

Replacedly, colored filter 14 can be to be used for yellow color component Ye (for example to be roughly zero at wavelength X=400～500nm place transmission coefficient, be about 1 at other wavelength place transmission coefficient), magneta colour component Mg (for example is roughly zero at wavelength X=500～600nm place transmission coefficient, be about 1 at other wavelength place transmission coefficient) or the complementary colours filter of cyan component Cy (for example be roughly zero at wavelength X=600～700nm place transmission coefficient, be about 1 at other wavelength place transmission coefficient).The complementary colours filter has for the three primary colors component of visible light and is roughly zero transmission coefficient.

The complementary colours filter has the sensitivity higher than primary color filters, uses wherein the sensitivity that complementary colours filter that in visible-range transmitted light has the complementary colours of one of corresponding primary colors can increase image device thus.On the contrary, the application of primary color filters has advantage: even also can obtain the signal of primary colors without the difference processing, thereby simplified signal processing.

Term " transmission coefficient is about 1 " refers to wherein the perfect condition of the transmission coefficient of transmission coefficient in other wave-length coverage in certain wave-length coverage.Transmission coefficient needn't be " 1 ".Similarly, term " transmission coefficient be roughly zero " refers to the perfect condition of the transmission coefficient of transmission coefficient in other wave-length coverage in wherein certain wave-length coverage.Transmission coefficient needn't be " being roughly zero ".

In primary systems or complementary colours system, can not consider whether to pass its infrared light scope IR as the reflected wavelength range component, promptly do not consider the transmission of infrared light IR, and its wave-length coverage component as the predetermined color (primary colors or complementary colours) within the visible light VL scope of transmission peak wavelength ranges of components of transmission.This be because stacked dielectric film 1 filtering infrared light IR component.

For example, as shown in Figure 3, not only the pixel 12IR in the unit pixel matrix 12 that comprises four pixels (unit) goes up and forms stacked dielectric film 1, and forms stacked dielectric film 1 respectively on other color red (R), green (G) and blue (B) pixel 12R, 12G and 12B.In addition, red (R), green (G) and blue (B) three primary colors filter 14R, 14G and 14B also are set respectively on each stacked dielectric film 1.

As shown in Figure 3,, on the pixel 12IR that does not have stacked dielectric film 1, do not arrange colored filter 14C, thereby not only infrared tube IR but also visible light VL form signal simultaneously in order to increase sensitivity.In this case, the pixel 12IR that can allow to be used for infrared light is roughly as being used for the pixel of infrared light IR and visible light VL, and not only is used for infrared light IR.

Especially, comprise that the unit picture element matrix 12 of four pixels is divided into pixel 12R, 12G, 12B and 12IR, thereby can not need any compartment of terrain to form the overall structure of image device (spectral image sensor 11), thereby be easy to design.

In this case, composograph on the basis of red (R) that obtain respectively from three pixel 12R, 12G, 12B, green (G) and blue (B) color component can roughly not be subjected to infrared light IR to obtain visible light VL coloured image (being conventional coloured image) thus with influencing.Simultaneously, can obtain the image of infrared light IR on the basis of the blending ingredients of infrared light IR that obtains from pixel 12IR and visible light VL.

Term " image of infrared light IR " means the image of the mixed light of the image that roughly is not subjected to the only infrared light IR that visible light VL influences or infrared light IR and visible light VL.In the structure shown in Fig. 3, in order only to obtain roughly not being subjected to the image of the infrared light IR that visible light VL influences, expectation obtains the difference of the mixed light component of infrared light IR and visible light VL and each red (R) of obtaining respectively from three pixel 12R, 12G, 12B, green (G) and blue (B) color component.This be because, as described below, even when green color filter 12G or black filter 14BK are not set, deduct the indigo plant, the intensity red and green glow that obtain respectively from three pixel 12R, 12G, 12B by output and determine infrared light intensity from the pixel 12IR that receives visible light VL and infrared light IR.

Consider wherein to obtain simultaneously roughly not to be subjected in the application of image of the only infrared light IR that visible light VL influences, such as optical communication applications or by following the tracks of the application of infraluminescence point detecting location, can on pixel 12IR, arrange colored filter 14C, colored filter at least transmission its be the infrared light IR of reflected wavelength range component and its predetermined wavelength component of transmission for the visible light VL of transmission peak wavelength ranges of components.

For example, when the green color filter 14G of transmission infrared light IR and green glow G is set as colored filter C, obtain the blending ingredients of infrared light IR and green visible light LG from pixel 12IR.Yet,, may obtain roughly not being subjected to the image of the only infrared light of visible light (the green glow G in this situation) influence by obtaining the difference of the only green color component visible light that obtains from pixel 12G.Though green color filter 14G preferably is set, deduct the indigo plant that obtains respectively from three pixel 12R, 12G and 12B with not verifying (proving) green color filter 14G, redly compare with the situation of green intensity, just simplified processing.

Replacedly, the black filter 14BK that transmitted infrared light can be set and only absorb visible light VL is as colored filter 14C.In this case, black filter 14BK absorbs visible light VL, only obtaining infrared light component I R from pixel 12IR, thereby even at the image of the only infrared light IR that does not carry out obtaining when difference is handled roughly not being subjected to visible light VL to influence.

At present normally used each R, G and B colored filter have high transmission coefficient for R, G or B and other color are had low transmission coefficient (for example, G and the B in R colored filter situation) at visible light belt.Yet the optical transmission coefficient beyond the visible light belt is indefinite and usually above other color (for example G and the B in R colored filter situation).For example, each filter has the sensitivity of infra-red range and the light of transmission infra-red range.Yet, in this embodiment, even when the perspective coefficient beyond the visible light belt is very high, can not have influence yet.

" example that the method for designing of stacked dielectric film, infrared absorption filter are removed "

The method for designing of＜thickness d j 〉

Fig. 4～6th, the exemplary plot of the basic principle of the method for designing of stacked dielectric film 1.Here, describe the example of a design, wherein stacked dielectric film 1 comprises first and second base materials, and selective reflecting infrared light IR.

Shown in the structure chart of Fig. 4, in the stacked dielectric film 1 that present embodiment is used, stacked a plurality of dielectric layer 1_j that each is made of first or second material layer, dielectric layer 1_j both sides (light incident side is called " layer 0 ", and opposite side is called " layer k ") are maintained at thick silicon oxide SiO ₂Layer (hereinafter is called " SiO ₂") between.In the example shown in Fig. 4, common material is as first and second material layers of dielectric layer 1_j.Particularly, silicon nitride Si ₃N ₄Layer (hereinafter being called " SiN ") and silicon oxide sio ₂Be used separately as first material layer and second material layer, and alternately stacked.Also imagine stacked dielectric film 1 superstructure or below arrange enough thick silicon oxide sio ₂Layer (d0=dk=∞).

When this stacked dielectric film 1 satisfies equation (1), can effectively increase reflectivity.

Equation 1:

dj＝λ ₀/4nj...(1)

In this equation, dj (hereinafter j is the number of plies) expression constitutes the thickness of each dielectric layer 1_j of stacked dielectric film 1, and nj represents the refractive index of each dielectric layer 1_j, λ ₀The centre wavelength (hereinafter being called " reflection kernel wavelength ") of expression reflected wavelength range.

Remove the thick silicon oxide SiO of both sides ₂Count the dielectric layer 1_j of the stacked dielectric film 1 of formation beyond the layer from ground floor to the k layer.For example, dielectric layer 1_j comprises and containing according to from first a SiN layer to k level preface, a SiO ₂Layer and three layers of a SiN layer, or contain a SiN layer, a SiO ₂Layer, a SiN layer, a SiO ₂Layer and five layers of a SiN layer.Fig. 4 illustrates a seven-layer structure.

In addition, when it is the reflection kernel wavelength X 0 of infrared light IR of reflected wavelength range during for 900nm, the refractive index n α of odd-level silicon nitride is 2.03, the 0, even number and k layer silicon oxide sio ₂Refractive index n β be 1.46, refractive indices n is 0.57.

According to equation (1), the thickness d α of silicon nitride SiN (=d1, d3 ..., the j=odd number) be 111nm, silicon oxide sio ₂Thickness d β (=d2, d4 ..., the j=even number) be 154nm.

Fig. 5 illustrates effective fresnel coefficient method of result's (reflectance spectrum chart) the reflectivity R that calculates by to(for) structure shown in the Fig. 4 that adopts common material.The figure shows out the correlation of reflectance spectrum for the number of plies.

Result shown in Fig. 5 represents that when the number of plies increases reflectivity R increases with the central point at the reflection kernel wavelength X 0900nm place of infrared light IR.And, find as reflection kernel wavelength X 0 by selecting wavelength 900nm, can substantial separation infrared light IR and visible light VL.Also find when having five or when more multi-layered, reflectivity R is 0.5 or bigger, especially when having seven or when more multi-layered, reflectivity is expected to above 0.7.

Fig. 6 is the exemplary plot of the varied in thickness correlation (changing relatively) for dielectric layer 1_j.The thickness d j that Fig. 6 is illustrated in each dielectric layer 1_j in the example that comprises seven layers changes ± 10% o'clock result of calculation (reflectance spectrum chart).

According to conditional equality (1), obtain calculated value by the fresnel coefficient method.Yet in fact, the condition of equation (1) is simple and changes.For example, though by the fresnel coefficient method calculate to find when thickness d j have ± during 10% error, still can increase reflectivity effectively.

For example, even Fig. 6 expresses when thickness d j changes, still increased reflectivity R effectively.Particularly, obtain 0.5 or bigger enough reflectivity R at the reflection kernel wavelength X 0900nm place of infrared light IR, and very high at whole infra-red range IR (mainly at 780nm or long wavelength side more) reflectivity.Therefore, when reality considers to have the variation of each dielectric layer 1_j of the thickness d j in equation (2) scope, can effectively have been increased the abundant effect of reflectivity.

Equation 2:

09×λ0/4n≤dj≤1.1×λ0/4n...(2)

The method for designing of＜reflection kernel wavelength X 0 〉

Fig. 7～9th, the condition exemplary plot of reflection kernel wavelength X 0.The bandwidth Delta lambda IR of the digital conditional decision of thickness d j in the infrared external reflection scope of spectrum.Shown in the principle of the reflectance spectrum shown in Fig. 7 (A), when the bandwidth Delta lambda IR broad in the infrared external reflection scope,, central wavelength lambda 0 will become remarkable unless being transformed into the reflection of longer wavelength side visible light VL.Shown in the principle of the reflectance spectrum shown in Fig. 7 (B), when the bandwidth Delta lambda IR in the infrared external reflection scope is narrower, unless central wavelength lambda 0 is transformed into the reflection that visible light VL can not take place shorter wavelength side near the infra-red range of visible light VL.

The absorption spectrum curve of silicon Si is expressed when the infrared light IR of 0.78 μ m≤λ≤0.95 mu m range in the infrared scope is reflected, and infrared absorption filter removes effect will become satisfactory.This is because wavelength is longer than the light of 0.95 μ m to be absorbed seldom in silicon Si, does not experience opto-electronic conversion.Therefore, preferably select the foveal reflex wavelength to make wavelength be reflected at the infrared light IR of 0.78 μ m≤λ≤0.95 mu m range.

Because the visible light VL of 649nm～780nm scope has low visibility in red (R) scope, look so image device can not be subjected to the influence of light reflection.Therefore, even work as 640nm～780nm wave-length coverage reflection having taken place, also any problem can not occur.

And along with the refractive index difference Δ n change of stacked dielectric film 1 is big, the bandwidth Delta lambda IR in the infrared external reflection scope broadens, otherwise along with the refractive index difference Δ n of stacked dielectric film 1 diminishes, the bandwidth Delta lambda IR in the infrared external reflection scope narrows down.Therefore, at SiN/SiO ₂In the multilayer film situation, the bandwidth Delta lambda IR in the infrared external reflection scope narrows down, and at Si/SiO ₂In the multilayer film situation, the bandwidth Delta lambda IR of infrared external reflection scope broadens.

As a result of, at SiN/SiO ₂In multilayer film (the refractive index difference Δ n=0.57) situation, demonstrate for the calculating of Fig. 8 reflectance spectrum chart shown reflection kernel wavelength X 0780nm and 950nm and in the scope of 780nm≤λ 0≤950nm, roughly to satisfy above-mentioned condition.Fig. 8 illustrates the result of calculation of reflectivity R, the feasible reflection kernel wavelength X 0 that only obtains 780nm and 950nm by the thickness d j that changes each dielectric layer 1_j in the stacked structure shown in Figure 13, and it will be described below.

Similar, at Si/SiO ₂In multilayer film (the refractive index difference Δ n=2.64) situation, the reflectance spectrum chart of Fig. 9 demonstrates and roughly satisfy above-mentioned condition in the scope of 900nm≤λ 0≤1100nm.

Therefore, at silicon nitride SiN, silicon Si and silicon oxide sio ₂Combination in, the equation (3-1) below preferred reflection kernel wavelength X 0 satisfies more preferably satisfies equation (3-2).This means that reflection kernel wavelength X 0 is in theory near 900nm.

Equation 3:

780nm≤λ0≤1100nm ...(3-1)

850nm≤λ0≤1000nm ...(3-2)

Certainly, above-mentioned material only is an example, can pass through silicon oxide sio ₂Obtain above-mentioned effect with the combination of materials beyond the combination of silicon nitride SiN layer.From calculate estimating to make that by selecting material can obtain same effect refractive index difference is 0.3 or bigger, more preferably 0.5 or bigger.

For example, can change the composition of SiN film a little according to formation condition.The examples of material that is applicable to the dielectric layer 1_j that constitutes stacked dielectric film 1 is removed silicon oxide sio ₂Beyond silicon nitride SiN, comprise oxide, such as aluminium oxide Al ₂O ₃, zirconia ZrO ₂(refractive index 2.05), titanium oxide TiO ₂(refractive index 2.3～2.55), magnesium oxide MgO and zinc oxide ZnO (refractive index 2.1); Resin material is such as Merlon (refractive index 1.58) and acrylic resin PMMA (refractive index 1.49); And semi-conducting material, such as carborundum SiC (refractive index 2.65) and germanium Ge (refractive index 4～5.5).

Use resin material, can form and have the optical filter that is different from conventional glass optics filbtercharacteristic.That is, plastic cement optical filter weight hypopathia has excellent durability (high temperature, high humidity and bump).

Replacedly, in order effectively to reduce dark current, the single crystalline layer of being made by the material with wide bandgap can be bonded to the surface of the semiconductor layer that wherein forms optical-electrical converter (Semiconductor substrate, semiconductor epitaxial layers, Semiconductor substrate and form semiconductor epitaxial layers thereon etc.), thereby form high potential barrier.

For example, when carborundum SiC layer being bonded to the n type Si laminar surface that wherein forms optical-electrical converter, SiC bandwidth 2.2eV, potential barrier becomes 1.5eV shown in the depth direction potential energy distribution of Figure 56.In this case, potential barrier is higher than the situation shown in Figure 60 A and the 60B, thereby has reduced dark current.

According to above-mentioned Fermi-Di Lake distribution function, dark current has reduced about 12 under the room temperature.

When reducing dark current, also to reduce noise to increase the S/N ratio.The result is that along with incident light quantity diminishes, even when being exaggerated signal by amplifier, it is not too obvious that noise also becomes.

The possible example of wide bandgap materials comprises various materials.

For example, by changing ratio of components change band gap such as the mixed crystal system of compound semiconductor.The example of mixed crystal system comprises AlGaInP mixed crystal, SiC mixed crystal, ZnCdSc mixed crystal and AlGaInN mixed crystal.

When silicon layer when wherein forming the semiconductor layer of optical-electrical converter, the SiC system of the same VI of the advantageous applications family element of considering to produce such as simple and easy.

Yet lattice mismatch has high absolute value between silicon and the SiC, and is easy to take place dislocation (misfit dislocation) at the place, junction interface thus.Lattice mismatch is limited by following equation (equation 11):

Equation 11:

$\mathrm{\Δa}=\frac{a_{\mathrm{SiC}}-a_{\mathrm{Si}}}{a_{\mathrm{Si}}}$

A wherein _SiCAnd a _SiIt is respectively the lattice constant of SiC and Si.

In order to prevent the appearance of dislocation, for example the thickness of SiC film approximately can be reduced to critical thickness or littler.For example, find that by experiment the thickness of SiC film preferably is reduced to 30nm or littler.

When finding that also ratio of components as C among the SiC is high, for example SiC: C is 1: 1, and preferred thickness further is reduced to 15nm or littler.

And, Ge can be added to SiC to form the SiGeC mixed crystal, be used to reduce the absolute value of lattice mismatch Δ a.

Table 1 illustrates the lattice constant of Si, Ge and C crystal structure.

Table 1

Table 1 is illustrated in the SiC system, because the lattice constant of C has increased the absolute value of lattice mismatch Δ a less than Si.

Therefore, by in SiC, mixing the Ge of lattice constant, reduced the absolute value of lattice mismatch Δ a to a certain extent greater than Si.

Even when using SiGeC to form single crystalline layer, the thickness of single crystalline layer is preferably 30nm or littler, more preferably 15nm or littler.

Do not have the band gap narrower owing to do not contain the SiGe of C, when SiGe system is used as single crystalline layer, preferably add C than Si.

As mentioned above, can increase degree of crystallinity by adding Ge formation SiGeC mixed crystal.Yet, can increase degree of crystallinity by other method.

That is, can between such as the semiconductor layer of Si layer etc. and single crystalline layer, insert the tight superlattice layer of at least one thickness smaller or equal to 15nm such as SiGeC layer etc.Tight superlattice layer discharges stress and removes dislocation in the in-plane, thereby increases degree of crystallinity.In this case, superlattice film can be the arbitrary film with the lattice constant that is different from Si.In other words, for example have differently when forming quantitative a plurality of SiGeC system layer, can obtain aforesaid same effect when on the Si substrate, forming.

Film for the single crystalline layer that obtains comprising one of above-claimed cpd, can use arbitrarily general growing method, such as CVD (chemical vapor deposition) method, MOCVD (metal-organic CVD) method, plasma CVD method, MBE (molecular beam epitaxy) method, laser wearing and tearing method, sputtering method etc.

Replacedly, can be on silicon face the carbonaceous material of deposit such as carbon etc., and then annealing so that the silicon face carbonization, thereby form the SiC layer from the teeth outwards.

As mentioned above, when the single crystalline layer with wide bandgap is blocked up, between single crystalline layer and the semiconductor layer dislocation appears.Therefore, thickness is preferably tens nm or littler.

On the other hand, when this layer is thin excessively, tunnel effect takes place, this layer is not enough to as potential barrier thus.Therefore, thickness is preferably 2nm or bigger, more preferably 5nm or bigger.

When the wide bandgap layer is amorphous layer or polycrystal layer, and during non-single crystalline layers, at this layer with arrange the level of formation at the interface between thereunder the semiconductor layer, thereby undesirably can not fully reduce dark current.

" use the spectral image sensor of stacked dielectric film: first embodiment "

Figure 10～14th is according to the exemplary plot of the first embodiment of the present invention corresponding to the spectrum sensor 11 of single wave spectrum separation of using stacked film.First embodiment is used to design the basic skills of the spectral image sensor of using stacked dielectric film.Here will do one to a design example of spectrum sensor 11 and describe, wherein the stacked dielectric film 1 of selective reflecting infrared light IR is used for filtering infrared light IR and receives visible light VL.

When forming with reference to the described stacked dielectric film 1 in Fig. 4～6 on having formed such as the semiconductor device layer of detecting elements such as silicon (Si) photo-detector thereon, the distance that semiconductor layer has between refractive index, semiconductor layer and the stacked dielectric film 1 of each dielectric layer 1_j that is higher than stacked dielectric layer 1 promptly comprises silicon oxide sio ₂The thickness d k of the k dielectric layer 1_k of layer is very important.

This means in structure chart as shown in figure 10 total reflection light LR _TotalAlong with from silicon substrate 1_ ω surface, the interference effect of the reverberation L4 on the surface of the semiconductor device layer (photo-detector etc.) that constitutes by for example silicon Si (refractive index 4.1) and changing just.

Figure 11 is example total reflection light LR _TotalFor comprising silicon oxide sio ₂The spectrum chart of the correlation that changes of the thickness d k of dielectric layer 1_k.Figure 11 illustrates the result of calculation along with the thickness d k variation of dielectric layer 1_k in the stacked dielectric film 1 with seven-layer structure shown in Figure 4.In each spectrum of Figure 11, wavelength (μ m) shows makes abscissa, and reflectivity R shows and makes ordinate.

It is 0.154 μ m that the spectral representation of Figure 11 is worked as thickness d k, and promptly when thickness satisfied the conditional equality (1) of infrared light IR reflection kernel wavelength X 0, reflectance spectrum was influenced hardly, and infrared light IR (wavelength X 〉=780nm) by strong reflection.On the contrary, when thickness d k is 0.3～50 μ m, compare other variation of generation with reflectance spectrum with thickness d k=∞.Therefore find to have the wave-length coverage that infrared light reflection wherein reduces with inclination angle (dip) form.

Yet, when thickness d k is 2.5 μ m or when bigger, the half-breadth at each inclination angle is 30nm or littler in the infrared external reflection, especially when thickness be 5.0 μ m or when bigger, half-breadth is 20nm or littler.The result is that the half-breadth that fully reduces common width (broad) natural daylight is to produce average reflectance.And it is very high that thickness d k is that the spectrum of 0.3～1.0 μ m demonstrates the reflectivity of visible light VL.Know optimum thickness dk by inference preferably near 0.154 μ m from these results, the value of equation (1) that promptly satisfies condition.

Figure 12 is that example is for comprising silicon oxide sio ₂The spectrum chart of the correlation that the thickness d k of the dielectric layer 1_k of layer changes.Especially, Figure 12 illustrates along with at thickness d k being the result that the thickness d k in the 0.154 μ m environs changes.In each spectrum of Figure 12, wavelength (μ m) shows makes abscissa, and reflectivity R shows and makes ordinate.

The result represents, in the thickness d k 0.154 μ m with the equation that satisfies condition (1) is the scope of thickness d k0.14～0.16 μ m at center, suppressed the reflection of visible light VL.

Therefore, the optimum structure of spectral image sensor 11 roughly comprises stacked dielectric film 1A, and it has eight layers that comprise k dielectric layer 1_k, shown in the structure chart of Figure 13.Figure 14 is the spectrum chart of result of calculation that the reflectance spectrum of stacked dielectric film 1A is shown.In other words, stacked dielectric film 1A has on the silicon substrate of the being included in 1_ ω and with four cycles silicon oxide sio is set ₂Structure as second material layer.

" use the spectral image sensor of stacked dielectric film: second embodiment "

Figure 15～18th, the exemplary plot of the spectrum sensor that separates corresponding to single wave spectrum of using stacked film 1 according to a second embodiment of the present invention.Second embodiment adopts first of the first embodiment method for designing to improve example.On the basis of the said method of reference Figure 10～14, improve to reduce the reflection of visible-range.

Improve in the example first, insert the 3rd material layer between k dielectric layer 1_k and silicon substrate 1_ ω, the 3rd material layer has the middle refractive index between the refractive index (=4.1) of the refractive index of k dielectric layer 1_k and silicon substrate 1_ ω.

In addition, according to improve the stacked dielectric film 1 of design first in the constant (constants) of layer 7, the reflection kernel wavelength X 0 of infrared light IR changes to lower wavelength from 900nm, 852nm for example, the thickness d α of ammonification silicon SiN (=d1, d3 ...; The j=odd number) is 105nm, silicon oxide sio ₂Thickness d β (=d2, d4 ...; The j=even number) is 146nm.This is because the new reflectivity that inserts thin SiN layer (30nm) with the reduction visible light reduces near the reflectivity the 780nm boundary between visible light and the infrared light.Therefore, all be switched to these reductions of lower wavelength side, and improve the filtering efficient of boundary vicinity infrared light with the compensatory reflex rate.Certainly, the reflection kernel wavelength X 0 of infrared light IR can remain on the 900nm place.

Particularly, improve in the routine structure, at first shown in Figure 15 at silicon oxide sio ₂Insert silicon nitride SiN thin layer 1_ γ between k layer and the silicon substrate 1_ ω with relatively little thickness d γ as the 3rd material layer.In this example, thickness 1_ γ is 0.030 μ m.Figure 16 illustrates the result of calculation of reflectance spectrum.

Improve in the example first, the 3rd material layer of incorporating into is identical with the first material layer silicon nitride SiN.Yet, can use arbitrary other member with the refractive index that is higher than silicon substrate 1_ ω.

Have first spectral image sensor 11 that improves the stacked dielectric film 1 of example and roughly comprise the stacked dielectric film 1B with nine layers of structure, this structure comprises stacked as a whole dielectric film 1, k dielectric layer 1_k (silicon oxide sio ₂Layer) and seven layers of silicon nitride SiN layer 1_ γ.

And, improve in the routine structure at second shown in Figure 17, insert the 4th material layer between the 3rd material layer that in the first improvement example, inserts and the silicon substrate 1_ ω with the refractive index that is lower than the 3rd material layer.Particularly, between as the silicon nitride SiN layer 1_ γ with thickness d γ of the 3rd material layer and silicon substrate 1_ ω, insert silicon oxide sio ₂Layer 1_ δ.The thickness d δ of the 4th material layer is 0.010 μ m.Figure 18 illustrates the result of calculation of reflectance spectrum.

Improve in the example the 4th material layer of incorporating into and the second material layer silicon oxide sio second ₂Identical.Yet, can use arbitrary other member with the refractive index that is lower than the 3rd material layer (in this example, silicon nitride SiN).

Have second spectral image sensor 11 that improves the stacked dielectric film 1 of example and roughly comprise the stacked dielectric film 1C with ten layers of structure, this structure comprises stacked dielectric film 1, k dielectric layer 1_k (silicon oxide sio as a whole ₂Layer), silicon nitride SiN layer 1_ γ and silicon oxide sio ₂Seven layers of layer 1_ δ.In other words, stacked dielectric film 1C has on silicon substrate 1_ ω and with five cycles silicon oxide sio is set ₂Structure as second material layer.

The difference of first and second examples is silicon oxide sio ₂The existence of layer 1_ δ, but Figure 16 and 18 is illustrated in two examples, has fully reduced the reflectivity of visible light VL.In addition, in second example, obtained by increasing silicon oxide sio ₂Layer 1_ δ reduces the effect of dark current.The preferred d δ of relation between the two layers of thickness＜＜d γ, making can be owing to increasing silicon oxide sio ₂Layer 1_ δ causes the effect that reduces increase silicon nitride SiN layer 1_ γ.

By this way, when at silicon oxide sio ₂When adding the silicon nitride SiN thin layer 1_ γ as the intermediate layer between k layer and the silicon substrate 1_ ω, the intermediate layer comprises having the (=nSiO between refractive index n k ₂) and refractive index n ω ((=nSiN) member can suppress the reflection of visible light VL to=middle refractive index n γ between nSi).Be appreciated that this point by following consideration.

At first, the wavelength of supposing visible light VL is λ VL, and middle refractive index is Nm, and the thickness in intermediate layer is dm, obtains equation (4) from the low-reflection film theory identical with equation (1).When satisfying equation (4), present gratifying effect.

Equation 4:

dm＝λVL/(4×Nm) ...(4)

Wherein wavelength X VL represents whole visible light VL, provides wave-length coverage by following equation (5) thus:

Equation 5:

380nm≤λVL≤780nm ...(5)

In first and second examples, as the intermediate layer increase silicon nitride SiN layer 1_ γ and have refractive index n γ (=nSiN=Nm).Therefore, the equation (5) of expression wave-length coverage is changed to the expression intermediate layer and gets thickness d m, i.e. the equation (6) of the thickness d γ of silicon nitride SiN layer 1_ γ.

Equation 6:

47nm≤dm≤96nm ...(6)

47nm≤dγ≤96nm ...(6)

Though the thickness d m in intermediate layer satisfies equation (6) in theory, thickness d m can depart from from it a little to some extent.According to experiment, allow to have littler thickness d m.Figure 16 and 18 has for example confirmed to present effect during for 30nm as thickness d m.Certainly, because at silicon oxide sio ₂Insert intermediate layer (the 3rd material layer) between k layer and the silicon substrate 1_ ω, the lower limit of intermediate layer thickness is greater than 0nm (except the 0nm).That is, when at silicon oxide sio ₂When inserting the intermediate layer between k layer and the silicon substrate 1_ ω, the thickness d m in intermediate layer and d γ preferably satisfy equation (7).

Equation (7):

0nm≤dm≤96nm ...(7)

0nm≤dγ≤96nm ...(7)

" use the spectral image sensor of stacked dielectric film: the 3rd embodiment "

Figure 19～24th, a third embodiment in accordance with the invention is corresponding to the spectral filter 10 of single wave spectrum separation of using stacked film 1 and the exemplary plot of spectral image sensor 11.Figure 19～22nd constitutes the stacked dielectric film 1 of the spectral filter 10 of a third embodiment in accordance with the invention, the exemplary plot of Figure 23 and 24 spectral image sensor that to be a third embodiment in accordance with the invention separate corresponding to single wave spectrum of using stacked film 1.

In the 3rd embodiment, use second of the first embodiment method for designing and improve example, reduce the number of the dielectric layer 1_j of stacked dielectric film 1.In order to reduce the number of dielectric layer, increase member (material layer) with the refractive index that is higher than first and second base materials that constitute stacked dielectric film 1.

In order to increase member with high index, can be with having two base materials that more the 5th material layer replacement of high index of refraction has high index.The second stacked dielectric film 1 that improves example becomes the stacked dielectric film 1D that roughly comprises the 5th material layer 1_ η.In other words, stacked dielectric film 1D has on silicon substrate 1_ ω and with N cycle silicon oxide sio is set ₂Structure as second material layer.

For the thickness d η of the 5th material layer, the refractive index of supposing the 5th material layer is n η, obtains equation (8) from the low-reflection film theory identical with equation (1).When satisfying equation (8), present gratifying effect.

Equation (8):

dη＝λ0/(4nη) ...(8)

For example, in the example shown in structure chart Figure 19, increase thickness d η and be higher than silicon nitride SiN and silicon oxide sio with 61nm ₂The silicon Si layer of refractive index 4.1 replace a silicon nitride SiN layer (the 3rd dielectric layer 1_3) as the 5th material layer.The result of calculation of reflectance spectrum shown in Figure 20.

Figure 20 illustrates when replacing having total number of plies with silicon Si layer and be the silicon nitride SiN of stacked dielectric film 1 center of odd number, the result of calculation that changes along with total number of plies.

In Figure 19, when designing the constant of each layer of stacked dielectric film 1, the reflection kernel wavelength X of infrared light IR 0 changes to 1000nm from 900nm, the thickness d α of silicon nitride SiN (=d1, d3 ...; The j=odd number) is 123nm, silicon oxide sio ₂Thickness d β (=d2, d4 ...; The j=even number) is 171nm.

In the example shown in structure chart Figure 21, when designing the constant of each layer of stacked dielectric film 1, the reflection kernel wavelength X 0 of infrared light IR is 900nm, the thickness d α of silicon nitride SiN (=d1, d3 ...; The j=odd number) is 111nm, silicon oxide sio ₂Thickness d β (=d2, d4 ...; The j=even number) is 154nm.In addition, increase silicon Si layer and replace a silicon nitride SiN layer as the 5th material layer with 55nm thickness d η.The result of calculation of reflectance spectrum as shown in Figure 22.

The 5th material layer that is increased to the second improvement example is identical with the silicon substrate 1_ ω that constitutes semiconductor element layer.Yet any other member with the refractive index that is higher than other dielectric layer 1_j can constitute stacked dielectric film 1.

The result of calculation of the reflectance spectrum shown in Figure 20 and 22 is represented, even have the very dielectric layer of peanut, can obtain enough reflectivity by in stacked dielectric film 1, increasing refractive index materials layer with the dielectric layer 1_j that is higher than except that the 5th material layer.Especially, because the wide bandwidth of visible light VL, five-layer structure is used for the separation the best between visible light VL and the infrared light IR.

With reference to the as above description among first embodiment of Figure 10～12, form stacked dielectric film 1D in order to go up at semiconductor element layer (silicon substrate 1_ ω), the distance between semiconductor element layer and the stacked dielectric film 1D is a silicon oxide sio ₂The thickness d k of k dielectric layer 1_k is very important.

This means in the structure chart as shown in figure 23 total reflection light LR _TotalAlong with from silicon substrate 1_ ω surface, it is the surface of the semiconductor element layer (photo-detector etc.) that is made of for example silicon Si (refractive index 4.1), reverberation LR interference effect and change.

Figure 24 is example total reflection light LR _TotalFor silicon oxide sio among the stacked dielectric film 1D with five-layer structure as shown in figure 21 ₂The reflectance spectrum chart of the correlation that the thickness d k of dielectric layer 1_k changes.In each spectrum of Figure 24, wavelength X (μ m) shows and makes abscissa, and reflectivity R shows and makes ordinate.

The spectral representation of Figure 24 when thickness d k be 0.15 μ m, promptly when thickness d k is approaching when satisfying the 0.154 μ m of conditional equality (1) of infrared light IR reflection kernel wavelength X 0, reflectance spectrum is influenced hardly, infrared light IR (wavelength X 〉=780nm) by strong reflection.On the contrary, when thickness d k is 0.3～50 μ m, compare other variation of generation with reflectance spectrum with thickness d k=∞.Therefore find to have the wave-length coverage that infrared light reflection wherein reduces with inclination angle (dip) form.This is with as above identical with reference to the description among first embodiment of Figure 11 and 12.

" use the spectral image sensor of stacked dielectric film: the 4th embodiment "

The exemplary plot of Figure 25 and 26 spectral image sensor 11 that to be a fourth embodiment in accordance with the invention separate corresponding to single wave spectrum of using stacked film 1.

The 4th embodiment is the improvement example of the 3rd embodiment that has wherein reduced the number of the dielectric layer 1_j that constitutes stacked dielectric film 1.In the 4th embodiment, further reduced the number of plies.Particularly, in order to reduce the number of plies, a plurality of members (material layer) have been increased with the refractive index that is higher than first and second base materials that constitute stacked dielectric film 1.When increase has a plurality of member of high index, can apparatus replace one of two base materials with the 5th material layer of high index of refraction more, it has higher refractive index.The stacked dielectric film 1 of this improvement example is to have the stacked dielectric film 1E that roughly comprises a plurality of the 5th material layer 1_ η structures.

Similar to the 3rd embodiment, as a plurality of the 5th material layer 1_ η, any member that can use the refractive index that is higher than other dielectric layer 1_j that constitutes stacked dielectric film 1 is as base material.A plurality of the 5th material layers can be identical or different.

As for the thickness d η p of the 5th material layer, the refractive index of supposing the 5th material layer is n η p, obtains equation (9) from the low-reflection film theory identical with equation (1).When satisfying equation (9), demonstrate gratifying effect.

Equation (9):

dηp＝λ0/(4nηp) ...(9)

For example, in the example shown in structure chart Figure 25, form stacked dielectric film 1E, be provided with two and respectively have thickness d η 61nm and be higher than silicon nitride SiN and silicon oxide sio with three-decker ₂The silicon Si layer of refractive index 4.1 replace silicon nitride as the 5th material layer.The result of calculation of reflectance spectrum shown in Figure 20.In other words, stacked dielectric film 1E has on silicon substrate 1_ ω and with four cycles silicon oxide sio is set ₂Layer is as the structure of second material layer.

When designing the constant of each layer of stacked dielectric film 1, the reflection kernel wavelength X 0 of infrared light IR is 1000nm, and the thickness d η (=d1 and d3) of the 5th material layer silicon Si layer is 61nm, two silicon oxide sios ₂The layer thickness d β (=d2) and dk be 171nm.

" use the spectral image sensor of stacked dielectric film: the 5th embodiment "

Figure 27 and 28 is exemplary plot of the spectral image sensor 11 that separates corresponding to single wave spectrum of using stacked film 1 according to a fifth embodiment of the invention.

Similar with second embodiment, in the 5th embodiment, the spectral image sensor 11 that has improved the 3rd or the 4th embodiment is to reduce the reflection of visible-range.

In the example shown in structure chart Figure 27, insert the 3rd material layer between k dielectric layer 1_k in the stacked dielectric film 1E of the 4th embodiment shown in Figure 25 and the silicon substrate 1_ ω, the 3rd material layer has the middle refractive index between the refractive index n ω (=4.1) of the refractive index n k of k dielectric layer 1_k and silicon substrate 1_ ω.Be different from second embodiment, in this embodiment, the reflection kernel wavelength X 0 of infrared light IR remains on the 1000nm place.Certainly, similar with second embodiment, the reflection kernel wavelength X 0 of infrared light IR can be changed to more downside, but not 1000nm.

Specifically, be similar among second embodiment first improve example, in structure shown in Figure 27, at silicon oxide sio ₂Deposit has the silicon nitride SiN layer 1_v of relatively little thickness d v as the 3rd material layer between k layer and the silicon substrate 1_ ω.Thickness d v is 0.030 μ m.Figure 28 illustrates the result of calculation of reflectance spectrum.The stacked dielectric film 1 that improves spectral image sensor 11 in the example is a five-layer structure, and it roughly comprises as a whole stacked dielectric film 1, k dielectric layer 1_k (silicon oxide sio ₂Layer) and three layers of silicon nitride SiN layer 1_v.

The 3rd material layer that is increased to this improvement example is identical with the first material layer silicon nitride SiN.Yet, can use any other member with the refractive index that is higher than silicon substrate 1_ ω.

Though not shown in the drawings, to second of second embodiment improve in the side similar, the 4th material layer that insertion has the refractive index that is lower than the 3rd material layer between the 3rd material layer that can insert in silicon substrate 1_ ω and this improvement example.

In any case, be similar to second embodiment, can reduce the reflectivity of visible light VL scope.Especially, increase the reflectivity of blue B component (wavelength is near 420nm) and green G component (wavelength is near 520nm) a little, but fully reduced the reflectivity of red R component (wavelength is near 600nm).Therefore, this example is applicable to from the separation of infrared light IR.

The image device of＜application of spectral imageing sensor: corresponding to CCD 〉

Figure 29 A, 29B and 30 are the circuit diagrams that the spectral image sensor 11 of above-mentioned arbitrary embodiment are applied to the image device that uses interline transfer (interline transfer) type CCD solid imaging element (IT_CCD imageing sensor).Image device 100 is examples of physical information acquiring device according to an embodiment of the invention.

In these figure, reference number 11 expression spectral image sensors; Reference number 12, the unit picture element matrix; Reference number 100, image device; Reference number 101, the CCD solid imaging element; Reference number 122, vertical transitions CCD; Reference number 124 reads door; Reference number 126, horizontal transfer CCD; Reference number 128, output amplifier; Reference number 140, the picture signal processing unit; Reference number 142, image switching (switch) control assembly; With reference number 146, drive control component.

Similar with Fig. 3, Figure 29 A and 29B illustrate and are used to survey infrared light IR and simultaneously visible light belt VL are separated into R, G and B color component, the structure of blue light B, green glow G and ruddiness R, wherein standalone probe visible light VL and infrared light IR.Unit picture element matrix 12 has the structure that roughly comprises pixel (optical-electrical converter) 12B, the 12G that are used for each wavelength and 12R and do not have the pixel 12IR of stacked dielectric film 1.

For example, shown in Figure 29 A, CCD solid imaging element 101 is included in a plurality of vertical transitions CCD 122 that are provided with on the vertical transitions direction except that unit picture element matrix 12.The charge transfer direction of vertical transitions CCD 122, promptly the read direction of picture signal is consistent with longitudinal direction (directions X among Figure 29 A).

And, between each vertical transitions CCD 122 and each unit picture element matrix 12, put into as the MOS transistor that reads door 124 (124B, the 124G, 124R and the 124IR that are used for each wavelength), and raceway groove stopper (not shown) is set at the boundary of each unit cell (unit component).

Find out that from Figure 29 A a unit picture element matrix 12 has the structure that is used for standalone probe blue light B, green glow G, ruddiness R and infrared light IR, this structure roughly comprises pixel 124B, 124G, 124R and the 124IR that is used for each wavelength (color).In imaging region 110, be provided for a plurality of vertical transitions CCD122 of each row, be used for vertical transitions by reading the signal charge that door 124 reads from sensor element 112, each sensor element 112 comprises unit pixel matrix 12.

In the arrangement of colored filter 14, for example in longitudinal direction (directions X) on the optical receiving surface of silicon substrate 1_ ω, at vertical transitions CCD 122, with blue, green, red, IR, indigo plant, green, red, IR ... order, and also with respect to the same direction (Y direction) of a plurality of vertical transitions CCD 122, with blue, green, red, IR, indigo plant, green, red, IR ... order, arrange filter 14.

In each unit picture element matrix 12 ( pixel 12B, 12G, 12R and 12IR) of each sensor element 112, when applying when reading the driving pulse Φ ROG of pulse ROG to reading door 124, the signal charge that gathers is read out to the vertical transitions CCD 122 of same vertical row.By the transfer that drives vertical transitions CCD 122 based on the driving pulse Φ Vx of the perpendicular transfering clock Vx of for example 3-to 9-, make and shift the signal charge that reads that in horizontal blanking cycle (blanking period) part, is used for corresponding to the various piece of a scan line of vertical direction (delegation) successively.The vertical transitions that is used for each row is called " row moves ".

And, in CCD solid imaging element 101, in the transfer end of a plurality of vertical transitions CCD 122, promptly predetermined direction (for example, horizontal) is set in abutting connection with last column place of vertical transitions CCD 122 and goes up the horizontal transfer CCD 126 (horizontal recording parts or horizontal transfer parts) that arranges with row.By the transfer that drives horizontal transfer CCD 126 based on the driving pulse Φ H1 of for example 2-phase horizontal transfer clock H1 and H2 and Φ H2, make in horizontal blanking cycle (blanking period) horizontal scanning period afterwards horizontal transfer successively from a plurality of vertical transitions CCD 122 row shift signal charge.Therefore, setting is corresponding to a plurality of (two) horizontal transfer electrode of two-phase translator.

And, have the output amplifier 128 that comprises floating point amplifier (floatingdiffusion amplifier) electric charge-voltage conversion unit (FDA) in the transfer end setting of horizontal transfer CCD 126.Output amplifier 128 physical messages are obtained an example of parts, wherein are converted to voltage signal in electric charge-voltage conversion unit successively, are amplified to predetermined level, output then by the signal charge of horizontal transfer CCD 126 horizontal transfers.Voltage signal is according to induction produces the output (V of picture signal as CCD from the reflected by objects light quantity _Out).The result has formed interline transfer type CCD solid imaging element 11.

Produce as CCD output (V from output amplifier 128 inductions _Out) picture signal be input to the picture signal processing unit 140 shown in Figure 29 B.In picture signal processing unit 140, from image switching control assembly 142 input picture switch-over control signals as the example of signal transition control assembly.The driving pulse driven CCD solid imaging element 101 of the self-driven control assembly 146 of origin (example of driver part).

Image switching control assembly 142 instruction changes the output of picture signal processing unit 140 into and roughly is not subjected to visible light VL monochrome that infrared light IR influences or coloured image, roughly is not subjected to the vision-mix of infrared light IR image that visible light VL influences, two images or visible light VL and infrared light IR promptly to increase the false monochrome or the false color image of infrared light IR brightness.In other words, image switching control assembly 142 is controlled the imaging output of visible light VL image and infrared light IR simultaneously and is transformed into picture output.

This instruction can be provided by the outside input that is used to operate image device, perhaps image switching control assembly 142 can instruct by using the visible light brightness that does not contain infrared light automatic processing transformation diagram picture and from 140 outputs of picture signal processing unit.

For example, picture signal processing unit 140 is carried out and is used for the view data R of each pixel synchronously, G, B, or the simultaneous operation of IR, be used for revising the tone noise correction operation of the tone noise (stripe noise) that produces by conditions of streaking (smear phenomenon) and high light phenomenon (blooming phenomenon), be used to control WB (white balance) control operation of WB, the gamma (gamma) that is used to control gradient (gradient) is revised operation, be used to promote the dynamic range of using two fluoroscopic Pixel Information dynamic ranges and promote operation with different charge storage times, or the YC signal generation operation that is used to produce brightness data (Y) and color data (C).The result is to have obtained visible light belt VL image (promptly conventional picture) on the basis of red (R), green (G) and blue (B) primary colors imaging data (R, G, B, IR pixel data).

The pixel data that image signal processing section 140 is also used infrared light IR produces infrared light IR image.For example, colored filter 14C is not set in the pixel 12IR that does not form stacked dielectric film 1 therein, makes when infrared light IR and visible light VL constitute signal simultaneously, the pixel data of using from pixel 12IR has obtained highly sensitive image.Replacedly, when green color filter 14G is set as colored filter 14C, obtained the vision-mix of infrared light IR and green visible light LG.Yet the difference of using the green color component that obtains with pixel 12G obtains the only image of infrared light IR.When black filter 14BK was set as colored filter 14C, the pixel data of using from pixel IR obtained the only image of infrared light IR.

Each image of Chan Shenging is sent to the display unit (not shown) and is sent to other functional part as visual picture in the operator shows, directly is stored in such as the memory device of hard disk device etc. or as deal with data as mentioned above.

Figure 30 illustrates the structure that is used for standalone probe visible light VL (blue light, green glow and ruddiness) and infrared light IR.Though do not describe details, its essential structure is with identical shown in Figure 29 A and the 29B, and each unit picture element matrix (photo diode sets) roughly comprises visible light VL pixel 12W and do not have the pixel 12IR of stacked dielectric film 1.Except the arrangement difference of colored filter 14, this structure is with identical shown in Figure 29 A and the 29B.

In the arrangement of colored filter 14, for example in longitudinal direction (directions X) on the optical receiving surface of silicon substrate 1_ ω, at vertical transitions CCD 122, with visible light VL, infrared light IR, visible light VL, infrared light IR ... order, and also with respect to the same direction (Y direction) of a plurality of vertical transitions CCD 122, with visible light VL, infrared light IR, visible light VL, infrared light IR ... order, arrange colored filter 14.

The image device of＜application of spectral imageing sensor: corresponding to CMOS 〉

Figure 31 A, 31B and 32 are the circuit diagrams that the spectral image sensor 11 of the foregoing description are applied to the image device that uses CMOS solid imaging element (cmos image sensor).Image device 100 is physical message acquisition devices according to an embodiment of the invention.

In these accompanying drawings, reference number 100 expression image devices; Reference number 201, the CMOS solid imaging element; Reference number 205, pixel amplifier; Reference number 207, drive control component; Reference number 219, the vertical signal circuit; Reference number 226, the row processing unit.

Similar with Fig. 3, Figure 31 A and 32B illustrate and are used to survey the structure that infrared light IR is separated into visible light belt VL each color component R, G and B simultaneously.This structure is suitable for blue light B, green glow G and the ruddiness R of standalone probe visible light VL, and infrared light IR, and unit picture element matrix 12 roughly comprises pixel (optical-electrical converter) 12B, the 12G that are used for each wavelength and 12R and the pixel 12IR that does not have stacked dielectric film 1.

Figure 32 illustrates the structure that is used for standalone probe visible light VL (blue light, green glow and ruddiness) and infrared light IR, and each unit picture element matrix 12 (photo diode sets) roughly comprises pixel 12W that is used for visible light VL and the pixel 12IR that does not have stacked dielectric film 1.Except the arrangement difference of colored filter 14, this structure and identical shown in Figure 31 A (identical) with Figure 30.

When spectral image sensor is applied to CMOS, for each pixel (optical-electrical converter) 12B, 12G, 12R and 12IR in the unit picture element matrix 12 are provided with Unit Amplifier.Therefore, in this case, the structure shown in application drawing 31A or Figure 32, wherein picture element signal is amplified then through outputs such as noise deletion circuit by each Unit Amplifier.

For example, CMOS solid imaging element 201 comprises the pixel component of wherein arranging a plurality of pixels with row and column (that is, two-dimensional matrix), and each pixel comprises according to the light receiving element of incident light quantity output signal (examples of charge generation parts).The signal of each pixel output is a voltage signal, image device 201 is typical row type, wherein is parallel to column direction CDS (correlated Double Sampling) (relevant secondary sample) processing capacity parts, digital translation parts (ADC: analog-digital converter) etc. are set.

Particularly, shown in Figure 31 A, CMOS solid imaging element 201 comprises drive control component 207, row processing unit 226 and the output circuit 228 of wherein arranging the pixel component (image-forming block) of a plurality of pixels 12 with row and column and being arranged at pixel component 210 outsides.

In addition, if desired, can be in the same semiconductor region of row processing unit 226 row processing unit 226 before or after AGC (automatic gain control) circuit with single enlarging function is set.When before row processing unit 226, carrying out AGC, carry out simulation and amplify; Yet when carrying out AGC after the row processing unit 226, combine digital is amplified.Because the simple amplification of n-bit digital data can make the gradient degradation, more preferably data are converted into numerical data after amplifying through simulation.

Drive control component 207 has the control circuit function that continues read pixel parts 210 signals.For example, drive control component 207 comprise the horizontal scanning circuit 212 (column scan circuit) that is used to control column address and column scan, be used to control row address and line scanning vertical scanning circuit 214 (line-scan circuit) with have as with the interface of outside be used to produce communicating by letter/timing controlled parts 220 of internal clocking.

Horizontal scanning circuit 212 has the function that reads sweep unit as the count value that is used to read row processing unit 226.Use the suitable technology of manufacturing technology with semiconductor integrated circuit, with the assembly and 210 integrated being formed together of the pixel component in the monocrystalline silicon semiconductor region of drive control component 207, thereby formation is as the solid imaging element (image device) of an example of semiconductor system.

For the purpose of simplifying, Figure 31 A only illustrates a part of row and column, but can arrange tens to several thousand pixels 12 at every row and every row.Each pixel 12 comprises as the unit picture element matrix 12 of light receiving element (charge generation parts) and has the pixel amplifier (Unit Amplifier of amplification semiconductor element (for example transistor) separately; The picture element signal production part) 205 (being used for versicolor 205B, 205G and 205R).

Find out that from Figure 31 A a unit picture element matrix 12 has the structure that is used for standalone probe blue light B, green glow G, ruddiness R and infrared light IR, and roughly comprise pixel 12B, 12G, 12R and the 12IR that is used for each wavelength (color).

In the arrangement of colored filter 14, for example on the optical receiving surface of silicon substrate 1_ ω, on directions X with blue, green, red, IR, indigo plant, green, red, IR ... order, and also on perpendicular to the Y direction of directions X with blue, green, red, IR, indigo plant, green, red, IR ... order, arrange filter 14.

As each pixel amplifier 205, use floating point amplifier.An example of pixel amplifier 205 comprises four universal transistors that are used for cmos sensor, it comprise as select transistor with respect to reading of an example of the electric charge fetch unit of charge generation parts (transfer gate parts/read door part), as the reset transistor of an example of replacement door part, vertically select transistor and as being used to survey the source follow-up amplifier transistor of an example of the detecting element that disintegration voltage changes floated.

As patent No.2708455 is disclosed, can use the amplifier of other type, it comprises three transistors, promptly be connected to amplifier transistor that drain line (DRN) is used to amplify the signal voltage of the signal charge that produces corresponding to the charge generation parts, the reset transistor of the pixel amplifier 205 that is used to reset with select transistor (transfer gate parts) by vertical transformation transistor through shifting reading of wiring (TRF) scanning.

Pixel 12 is connected to vertical scanning circuit 214 through row control circuit 215, and is connected to row processing unit 226 through vertical signal circuit 219.Row control circuit 215 comprises the whole wiring that extends to pixel from vertical scanning circuit 214.For example, be parallel to long scattering object 3 and arrange row control circuit 215.

Horizontal scanning circuit 212 and vertical scanning circuit 214 respectively comprise, for example change transistor and decoder, with corresponding to the control signal start address selection operation (scanning) from communication/timing controlled parts 220.Therefore, going control circuit 125 comprises the various pulse signals that are used to drive pixel 12 (the pulse RST that for example resets, shift pulse TRF, DRN control impuls DRN etc.).

Though it is not shown in the drawings, communication/timing controlled parts 220 comprise the clock that is used to supply operation of components necessity and and to pulse signal the functional block of predetermined timing generator TG (reading an example of address control unit spare) regularly is provided and be used for through terminal 220a receive master clock CLKO, be used to instruct mode of operation through terminal 220b receive data DATA and through terminal 220C output comprise the information of CMOS solid imaging element 201 dateout, the functional block as communication interface.

For example, horizontal address signal and vertical address signal be output valve horizontal demoder and perpendicular recording device respectively, thereby each decoder selects corresponding row or column to drive pixel 12 and row processing unit 226 through drive circuit.

In this case, because pixel is arranged with two-dimensional matrix, on analog pixel signal that pixel amplifier (picture element signal production part) 205 produces and the column direction via the output of vertical signal circuit 219 by position (the being parallel to column direction) access of itemizing, and scan by (vertically) and to read.Then, thus be accessed in perpendicular to the picture element signal on the line direction of vertical direction by (level) scanning picture element signal (for example digital pixel data) is read out to outlet side.Therefore, expectation increases the reading speed of picture element signal and pixel data.Certainly, can carry out arbitrary access, thereby the information of necessary pixel 12 is only read in the address of the pixel 12 that wherein direct appointment will be read.

Communication/timing controlled parts 220 are to each parts of device for example horizontal scanning circuit 212, vertical scanning circuit 214, row processing unit 226 etc., the clock CLKI that supply is identical with the master clock CLKO frequency of exporting through terminal 220a and frequency is divided into 2 or the low-speed clock that obtains of more parts.

Vertical scanning circuit 214 is selected row of pixel component 210 and is supplied necessary pulse to these row.For example, vertical scanning circuit 214 comprises and being used to specify the vertical decoder that reads row on (row of selecting pixel component 210) vertical direction and is used for to corresponding to supplying pulses by the capable control circuit that reads the pixel of locating address (column direction) 12 215 of vertical decoder appointment.Vertical decoder selection is used for the row of electronic shutter and is used to read the row of signal.

Then horizontal scanning circuit 212 select with the synchronous row processing unit 226 of low-speed clock CLK2 in the column circuits (not shown) and with signal guidance to horizontal signal circuit (horizontal output circuit) 218.For example, horizontal scanning circuit 212 comprise and being used to specify read the horizontal demoder of row on (each column circuits of selecting row processing unit 226) horizontal direction and be used to use basis by the selector switch that reads the address 227 of vertical decoder appointment with each signal guidance to the level of horizontal signal circuit 218 of row processing unit 226 (derive) circuit of deriving.For example, when row AD circuit operation figure place n (positive integer) be 10 (=n) time, be 10 corresponding to the number of the horizontal signal circuit 218 of figure place.

In having the CMOS figure attitude image device 201 of above-mentioned structure, the picture element signal output that is used for the pixel 12 of each horizontal path is supplied to the column circuits of row processing unit 226 via vertical signal circuit 219.The signal charge that is stored in the unit picture element matrix 12 ( pixel 12B, 12G, 12R and 12IR) is read out via same vertical signal circuit 219.

Each column circuits of row processing unit 226 receives from the signal of a row pixel and handles this signal.For example, each column circuits has that to be used to use low-speed clock CLK2 be ADC (analog-digital converter) circuit of 10-bit digital data with analog signal conversion.

By using suitable circuit structure, cause immediately following the difference between the signal level (noise level) after the pixel replacement and real (corresponding to incident light quantity) signal level Vsig thereby can handle via the picture element signal of vertical signal circuit 219 with the voltage mode input.Therefore, can remove (FPN) and the noise signal component of replacement noise such as intrinsic plant noise (fixed pattern noise).

The analog pixel signal of handling in the column circuits (or digital pixel data) is transferred into horizontal signal circuit 218 via the horizontal selector switch 217 of selecting signal to drive by the level of horizontal scanning circuit 212 outputs, inputs to output circuit 28 then.Figure place is 10 to be examples, and figure place can be less than 10, and (for example 8) maybe can surpass 10 (for example 14).

In above-mentioned structure, from wherein as the unit picture element matrix 12 of charge generation parts with the pixel of each horizontal path the pixel component 210 of the arranged output pixel signal that continues.Therefore, demonstrate as the aggregate of the picture element signal of whole pixel component 210, corresponding to light receiving element wherein with image, i.e. a two field picture of the pixel component 210 of arranged.

Output circuit 228 is corresponding to the output amplifier in the CCD solid imaging element 101 128, and is similar in the CCD solid imaging element 101, in output circuit 228 back picture signal processing unit 140 is set, shown in Figure 31 B.In addition, be similar in the CCD solid imaging element 101, from image switching control assembly 142 input picture switch-over control signals to picture signal processing unit 140.

The result is, at the imaging data (pixel data R, G, B and IR) of red (R), green (G) and blue (B) primary colors or be used for obtaining on the basis of pixel data of visible light VL the image (promptly conventional the picture) of visible light belt VL, the pixel data that also can use infrared light IR obtains infrared light IR image.

Though not shown in the drawings, when when the basic structure shown in Figure 29 or the 31A removes pixel 12IR, separate each color component R, the G and the B that survey among the visible light belt VL.

In the arrangement of colored filter 14, for example in longitudinal direction (directions X) on the optical receiving surface of silicon substrate 1_ ω, at vertical transitions CCD 122, with blue, green, red, green, blue, green, red, green, blue ... order, and also with respect to the same direction (Y direction) of a plurality of vertical transitions CCD 122, with blue, green, red, green, blue, green, red, green, blue ... order, arrange colored filter 14.Replacedly, in 2 * 2 unit picture element matrix 12, arrange (Bayer arrangement) with so-called Baeyer and arrange two green (G) pixels and redness (R) and blueness (B) pixel respectively, or in three look B, G and R, increase the 4th look (for example emerald green) in order to extend the color reproduction scope.

In this case, only obtain the image of visible light band VL, but the infrared absorption filter that can not arrange in transducer the place ahead as an example of subtractive filter removes filter.Remove filter owing to can not arrange expensive infrared absorption filter, significantly reduced cost.In addition, remove filter owing to can not arrange thick and heavy infrared absorption filter, optical system can be made in light weightly and be compact.Certainly, infrared absorption filter can be set insert/pull out mechanical device, can not increase size of devices thus except that filter.

The infrared absorption filter that this advantage on the cost also can be applicable to wherein replace existing glass to make with stacked dielectric film removes in the structure of filter, and separation is formed into image-position sensor and stacked dielectric film (separate and form exploring block and stacked dielectric film).

For example, remove filter with the infrared absorption filter that uses existing glass to make and compare, device has superiority on cost and is made compact, therefore the image device with excellent portability can be set, for example digital camera etc.

Arrange infrared absorption filter therein in transducer the place ahead and remove in the structure of filter, insert glass substrate in the place ahead of CCD or cmos imaging device, thereby in light path, produce air-glass surface.Therefore, the visible light of transmission is by boundary reflection, thus the problem that causes sensitivity to reduce.And, increasing the number at this interface, the refraction angle of oblique incidence light is with wavelength change thus.Thereby cause because the high light (blooming) that optical path change causes.In this case, use stacked dielectric film and have the advantage that prevents according to the high light of wavelength.

" example of manufacturing process "

Figure 33 A～F is the exemplary plot of an example of manufacturing process of spectral image sensor with sensor construction of above-mentioned arbitrary embodiment.That is, Figure 33 A～F is the exemplary plot of an example of manufacturing process that comprises the spectral image sensor of infrared light IR light receiving part and visible light VL light receiving part.

In the formation of this structure, as Figure 29,30 or 31A and 32 shown in, at first form common CCD or CMOS structural circuit.Then, by silicon oxide deposition SiO successively on the Si photodiode such as for example CVD (chemical vapor deposition) ₂Film and silicon nitride SiN film (Figure 33 B).

Then,, thereby in infrared light IR light receiving part, form opening, make opening arrive undermost silicon oxide sio by RIE etchings such as (reactive ion-etchings) one of four pixels only ₂Film (Figure 33 E).

Then, in order to protect stacked dielectric film 1 etc., by deposit one SiO again on stacked dielectric film 1 of CVD for example ₂Film, stacked dielectric film 1 has the opening that is formed in its part.Can this technology of appropriate change.

Can be with the photoresist (photo-resist) that has corresponding to the opening of infrared IR light receiving part in this technology, thus three pixels of not etching visible light VL (R, G, B component) (Figure 33 C and 33D).In this case, deposit SiO on stacked dielectric film 1 ₂Remove photoresist (Figure 33 D～E) before the film.

Though not shown in the drawings, can also be at the SiO of each pixel ₂Form colored filter and lenticule on the film.

And, when infrared light IR reveals the optical-electrical converter (photodiode etc.) that enters visible light VL a little, can arrange full reduction ultraviolet filtering filter.For example, promptly suitably arrange the infrared absorption filter that has smaller or equal to 50% rejection rate remove filter with the infrared light filtering to the level that roughly can not cause influencing visible light VL, infrared light IR is converged to infrared light IR optical-electrical converter (photodiode etc.), thereby causes enough sensitivity.

In this manufacturing process, the part near the Si substrate is carried out etching, promptly in infrared light IR light receiving part, form and arrive undermost silicon oxide sio ₂The opening of film (Figure 33 E).Therefore, the etching meeting causes damage problem.In this case, can be formed directly into SiO on the Si substrate by increase ₂The thickness of film and reduce damage.

When dk is 2.5 μ m or when bigger, as shown in figure 11, the half-breadth at the infrared light IR scope inclination angle of reflectance spectrum narrows down, thus usually the reflectivity of width natural daylight by average, thereby allow infrared reflection of light.Therefore, the thickness d k of k dielectric layer 1_k is preferably 2.5 μ m or bigger, more preferably 5 μ m or bigger.

Directly be formed on the photodiode on the silicon substrate 1_ ω and the metal line of pixel amplifier on the silicon substrate 1_ ω, promptly be used to form read, as from as the unit signal of the pixel amplifier of the unit signal production part in the image-forming block (detecting area), the wiring layer of signal line of picture element signal, compare with the structure that stacked dielectric film 1 wherein is formed directly on the silicon substrate 1_ ω, with the stacked dielectric film 1 of silicon substrate 1_ ω formation spaced apart.In other words, stacked dielectric film 1 is formed on the metal line, has simplified technology thus, thereby brings advantage cheaply.Particularly, when the number of plies that constitutes stacked dielectric film 1 increases, can obtain the effect of making us quite satisfied.Hereinafter, metal line-conscious (conscious) spectral image sensor will be described.

" use the spectral image sensor of laminated medium film; The 6th embodiment "

Figure 34-the 40th illustrates the accompanying drawing according to the spectral image sensor 11 that separates corresponding to single wave spectrum of the stacked deielectric-coating 1 of the use of the 6th embodiment.In the 6th embodiment, based on such as described technology with reference to accompanying drawing 10-14, in a distance away from silicon substrate 1_ ω of metal line, stacked deielectric-coating 1 and exploring block for example photodiode etc. are formed integrally on silicon substrate 1 ω.

For example, in the CMOS structure, as shown in Figure 34, be formed with therein on the semiconductor element layer of exploring block such as photodiode etc., form a wiring layer.When the thickness of wiring layer is about 0.7 μ m, just can be formed with thereon on the silicon substrate 1_ ω of photodiode and integrally forms the wiring layer of about 0.7 μ m thickness, and can after the technology of carrying out first wiring layer, form stacked dielectric layer 1.In the case, in k layer, this wiring layer is set with about 0.7 μ m thickness d k.

And, as shown in Figure 35, when on semiconductor element layer, the three-layer routing layer of the gross thickness with about 3.2 μ m being set, just can be formed with thereon on the silicon substrate 1_ ω of photodiode etc. and be formed integrally as a sandwich construction, and can after carrying out the technology of the 3rd wiring layer, the top form stacked dielectric layer 1.In the case, in the k layer of the thickness d k with about 3.2 μ m, form this wiring layer.

In the present embodiment, as shown in the drawing, the thickness of about 3.2 μ m represents to have removed the SiO that is provided with on silicon substrate 1_ ω ₂The thickness of about 10nm thickness of layer (δ layer) and the layer k outside about 65nm thickness of the SiN layer (γ layer) that is provided with on this layer.

After having formed stacked dielectric layer 1, can form colour filter 14 and lenticule.

As spectral image sensor 11 corresponding to present embodiment, for example, as shown in Figure 36, in the seven-layer structure of second embodiment shown in Figure 17, employing comprises three layers stacked deielectric-coating 1C as substrate, and these three layers comprise the stacked dielectric layer 1_k of k (silicon oxide sio ₂Layer), silicon nitride SiN layer 1_ γ and silicon oxide sio ₂Layer 1_ δ, k layer dielectric layer 1_k has the thickness of 700nm.In stacked dielectric layer 1C, the silicon nitride SiN layer 1_ γ of relative thin has the thickness d γ of 65nm or 100nm, and is deposited as k layer silicon oxide sio ₂Trilaminate material between layer 1_k and the silicon substrate 1_ ω, and silicon oxide sio ₂Layer 1_ δ has the thickness of 10nm and has the refraction coefficient lower than the refraction coefficient of this trilaminate material, and is deposited as the 4th layer material between trilaminate material and silicon substrate 1_ ω.

In Figure 37, basic stacked deielectric-coating 1 has nine layers of structure, and wherein k layer dielectric layer 1_k has the thickness of 700nm or 3.2 μ m.In stacked deielectric-coating 1C, the silicon nitride SiN layer 1_ γ of relative thin has the thickness d γ of 65nm, and is deposited as k layer silicon oxide sio ₂Trilaminate material between layer 1_k and the silicon substrate 1_ ω, and silicon oxide sio ₂Layer 1_ δ has the thickness d δ of 10nm and has the refraction coefficient lower than the refraction coefficient of this trilaminate material, and is deposited as the 4th layer material between trilaminate material and silicon substrate 1_ ω.

The result of calculation of the reflectance spectrum of three kinds of structures has been shown among Figure 38-40.From Figure 37 and 36, as can be seen, on silicon substrate 1_ ω, be formed with the stacked deielectric-coating 1 of about 0.7 μ m or 3.2 μ m, thus just can simplified wiring technology.Obviously, will be simultaneously on silicon substrate 1_ ω order the SiO of conduct the 4th layer material of 10nm and 65nm (or 100nm) thickness appears having respectively ₂Layer and as the SiN layer of trilaminate material, the stacked deielectric-coating 1 on silicon substrate 1_ ω is just greater than 0.7 μ m or 3.2 μ m.

More than describe every layer and all comprised SiN film and SiO ₂Seven stacked deielectric-coating 1 and the nine stacked deielectric-coating 1 of putting put of film.Yet, as shown in Figure 39, when the number of plies when 7 increase to 9, the reflectivity R in infrared ray IR scope just is enough to increase to 0.9 or higher.

As shown in Figure 40, the thickness d k of k layer dielectric layer 1_k is in the seven-layer structure of 3.2 μ m therein, and it is big that the inclination angle in the infrared reflection scope becomes, and will cause the remarkable increase of reflectivity.Yet, have been found that by the number of plies is increased to 9 and improve the inclination angle that the reflectivity in the infrared ray IR scope will become obvious.

Similarly, Figure 38 showed when as the thickness d γ of the 3rd layer SiN layer when big reflectivity in the visible light VL scope will increase.This possibility be since with the described identical fact in a second embodiment, trilaminate material is set to be used for reduce the intermediate layer of the reflectivity of visible-range, and the thickness d γ of the dielectric layer 1_ γ that is provided with as the intermediate layer just satisfies equation (6) ideally.In other words, on the thin layer side, there is bigger possibility, but on the thick-layer side, then has only little possibility.

As mentioned above, when after conventional Wiring technique, forming stacked deielectric-coating 1, just can simplify manufacturing, and need not increase new technology, will help cost thus.In other words, just simplified the technology that being used to shown in Figure 35 make the CMOS structure, thereby demonstrated good effect.When after having formed stacked deielectric-coating 1, carrying out Wiring technique, just be difficult to remove stacked deielectric-coating 1.

Below each embodiment that present invention will be further described.

Figure 54 is the accompanying drawing of displaying according to the schematic structure of the solid imaging element of the 7th embodiment.

In the present embodiment, will be applied to the CCD solid imaging element according to the solid imaging element of an embodiment.

In Figure 54, reference number 1001 expression solid imaging elements; Reference number 1002 expression ccd registers; Reference number 1003 expression horizontal CCD registers; And reference number 104 expression output amplifiers.

In solid imaging element 1001, according to the photodiode PD of arranged, and the vertical ccd register 1002 of vertical extent (longitudinal direction in the accompanying drawing) is set as light-receiving member, be used for each row of light-receiving member (photodiode PD).Similarly, the horizontal CCD register 1003 of horizontal-extending (horizontal direction in the accompanying drawing) is connected to each terminal of vertical ccd register 1002.And, by output amplifier 1004, output block 1005 is connected to a terminal of horizontal CCD register 1003.

Figure 55 is the profile that contains the solid imaging element 1001 of the light-receiving member shown in Figure 54.

In Figure 55, reference number 1001 expression solid imaging elements; The vertical ccd register of reference number 1002 expressions; Reference number 1011 expression silicon substrates; Reference number 1012 expression p-N-type semiconductor N well regions; Reference number 1013 expression (n-type) electric charge storage regions; Channel region is shifted in reference number 1015 expressions; Reference number 1021 expression optical screen films; And reference number 1025 expression single crystalline layers.PD represents photodiode.

As shown in Figure 55, on n-type silicon substrate 1011, form p-N-type semiconductor N well region 1012, and formation has wherein formed the semiconductor region of photodiode PD with vertical ccd register 1002 in p-N-type semiconductor N well region 1012.

Each photodiode PD is used as optical-electrical converter and is included in the n-type electric charge storage region 1013 of the top formation of p-N-type semiconductor N well region 1012, and these districts 1012 and 1013 have just constituted each photodiode.

In each vertical ccd register 1002, the surface of contiguous p-N-type semiconductor N well region 1012 forms the n-type and shifts channel region 1015, and signal charge is transferred to the n-type and shifts channel region 1015, and shift under the channel region 1015 in the n-type, form the 2nd p-N-type semiconductor N well region 1014.

Similarly, in order to prevent that signal charge from shifting channel region 1015 from the n-type that n-type electric charge storage region 1013 flows into the right side, n-type on the n-of each photodiode PD type electric charge storage region 1013 and right side shifts between the channel region 1015, forms p-type raceway groove and stops to distinguish 1016.

And, between the n-type channel region 1015 in the n-of each photodiode PD type electric charge storage region 1013 and left side, form and read gate regions 1017.

And, have on the silicon substrate 1011 that gate insulating film 1018 is set therebetween, form the transfer electrode of forming by polysilicon 1019.Reading gate regions 1017, shifting channel region 1015 and raceway groove stops to distinguish on 1016, form transfer electrode 1019, and transfer electrode 1019 has an opening that forms on the electric charge storage region 1013 of each photodiode PD.

In solid imaging element according to present embodiment, the single crystalline layer 1025 that the material of for example SiC or SiGeC is formed bonds to the top of the silicon substrate 1011 of the photodiode (optical-electrical converter) that is formed with light-receiving member on it, and SiC or SiGeC have the band gap wideer than the band gap of the silicon of silicon substrate 1011.

Compare with device,, just can significantly reduce dark current by single crystalline layer 1025 is set with conventional structure.

As mentioned above, when adopting SiC or SiGeC as single crystalline layer 1025, the thickness of single crystalline layer 1025 is preferably 30nm or littler, for example, and about 10nm.

Can wait by aforesaid arbitrary method such as CVD and form single crystalline layer 1025.

In solid imaging element 1001 according to present embodiment, will by have than the band gap of the silicon of silicon substrate 1011 more the single crystalline layer 1025 formed of the material of wide bandgap bond to the silicon substrate 1011 that is formed with photodiode PD on it.Because single crystalline layer 1025 has the band gap of broad, therefore just can improve and prevent that electronics is positioned at the potential barrier of surface level, thereby minimizing because of electronics cause dark current.For example, dark current can significantly be reduced 12, therefore just can significantly improve the S/N ratio of the signal that is used for incident light.

As a result, when improving signal gain, also can obtain not have the image of obvious noise even improve sensitivity under for a small amount of incident light in image-forming condition such as darkroom etc.

Similarly, even when solid imaging element 1001 has muting sensitivity, also can only not consider that incident light quantity obtains high-quality image by the multiplication factor of amplifier.

And, even when each pixel of making solid imaging element 1001 imperceptibly when reducing incident light quantity, also can guarantee enough S/N ratios, therefore only the multiplication factor by amplifier compensates muting sensitivity, also can obtain not have the image of obvious noise.

Therefore,, just can improve the pixel quantity in the solid imaging element 1001, and just can reduce the size of solid imaging element by making each pixel of solid imaging element 1001 fine.

(example)

Actual form and verify solid imaging element 1001 according to present embodiment with respect to each feature.

At first, form and to contain as single crystalline layer 1025 and the solid imaging element 1001 of the SiC layer that forms.

By for example CVD method, on silicon substrate 1011, the SiC layer of about 10nm thickness of growing, thus form single crystalline layer 1025.In this step, for example, adopt C ₃H ₈With single silane SiH ₄As source material, and underlayer temperature is 1110 ℃ or lower.

Can form the SiC layer by the method except the CVD method.For example, can be by adopting SiC as the laser abrasion method of the target material SiC layer of growing.

Then, by being used to make the conventional method of CCD solid imaging element, make the solid imaging element 1001 shown in Figure 54 and 55.

In the actual imaging of solid imaging element 1001, can significantly reduce dark current, even, also can obtain not have the image of obvious noise by under dark condition, carrying out imaging with manufacturing.

Then, form the solid imaging element 1001 that contains the SiGeC layer that forms as single crystalline layer 1025.

At first, silicon substrate 1011 immerses NH ₄OH, H ₂O ₂And H ₂The mixed solution of O (mixing ratio is 1: 1: 5) ten minutes is with clean surface.

Then, with HF (HF: H ₂O=1: 50) handle 10 seconds of substrate to remove natural oxide film from the surface of silicon substrate 1101.

In these steps, clean silicon substrate 1011 surfaces, be used to improve the degree of crystallinity of follow-up crystal growth.

On substrate support, fix and remove the silicon substrate 1011 of natural oxide film from it by above-mentioned preliminary treatment.

Then, by low pressure chemical vapor deposition on silicon substrate 1011 deposit as the SiGeC layer of single crystalline layer 1025.

At first, supply pressure is 1 * 10 ⁴The propane C of the 450 μ mol/min of Pa ₃H ₈, underlayer temperature is 1150 ℃, and hydrogen flow rate is 11/min, keeps this state 2 minutes with silicon carbide substrates 1011 surfaces.

And, supply SiH simultaneously with 36 μ mol/min, 59 μ mol/min and 10 μ mol/min respectively ₄, C ₃H ₈And GeH ₄30 seconds by the low pressure chemical vapor deposition SiGeC crystal of growing on silicon substrate.As a result, be deposited to the thickness of about 10nm as the SiGeC layer of single crystalline layer 1025.

Though, in this example, use low pressure chemical vapor deposition, can form SiGeC by additive method.For example, can adopt SiGeC as the laser abrasion method of target material or adopt organo metallic material for example organosilan material etc. come grown crystal by gas source MBE method.

Then, by being used to make the conventional method of CCD solid imaging element, make the solid imaging element 1001 shown in Figure 54 and 55.

In the actual imaging of solid imaging element 1001, can significantly reduce dark current, even, also can obtain not have the image of obvious noise by under dark condition, carrying out imaging with manufacturing.

In the above-described embodiments, on the whole zone of silicon substrate 1011, form single crystalline layer 1025.Yet, in order to improve electrical characteristics, can pass through RIE (reactive ion etching) etc., except the part corresponding to photodiode PD, local corrosion is removed single crystalline layer.In the case, can before etching, protect photodiode by photo etched mask.

The relevant embodiment of situation therewith below will be described.

Figure 57 is an accompanying drawing (cutaway view) of showing the schematic structure of solid imaging element according to another embodiment of the invention.

In Figure 57, reference number 1030 expression solid imaging elements; The vertical ccd register of reference number 1002 expressions; Reference number 1011 expression silicon substrates; Reference number 1012 expression p-N-type semiconductor N well regions; Reference number 1013 expression (n-type) electric charge storage regions; Channel region is shifted in reference number 1015 expressions; Reference number 1019 expression transfer electrodes; Reference number 1021 expression optical screen films; Reference number 1026 expression single crystalline layers; And reference number 1051 expression Unit Amplifier.PD represents photodiode.

In solid imaging element 1030, only on the surface of the photodiode PD of silicon substrate 1011, form the single crystalline layer of for example forming 1026 by SiC or SiGeC according to present embodiment.Single crystalline layer 1026 is bonded to silicon substrate 1011.

Each several part in the solid imaging element 1001 of other parts and the foregoing description is identical, and represents by identical reference number, and is not described hereinafter.

Can form single crystalline layer 1026 as the skim of single crystalline layer 1026 by deposit on whole zone, come local this film of removing by etching such as RIE (reactive ion etching) then, thereby keep and the corresponding part of photodiode PD.

In the structure of the solid imaging element 1030 of present embodiment, the single crystalline layer 1026 with band gap wideer than the band gap of the silicon of silicon substrate 1011 is set, so that it is bonded on the photodiode PD on the silicon substrate 1011.Because single crystalline layer 1026 has the band gap of broad, therefore just can improve and prevent that electronics is positioned at the potential barrier of surface level, thus the dark current that minimizing causes because of electronics.For example, dark current can significantly be reduced 12, thereby can significantly improve the S/N ratio that is used for incident optical signal.

Therefore, when improving signal gain, also can obtain not have the image of obvious noise even improve sensitivity under for a small amount of incident light in image-forming condition such as darkroom etc.

Similarly, even when solid imaging element 1030 has muting sensitivity, also can only not consider that incident light quantity obtains high-quality image by the multiplication factor of amplifier.

And, even when each pixel of making solid imaging element 1030 imperceptibly when reducing incident light quantity, also can guarantee enough S/N ratios, therefore only the multiplication factor by amplifier compensates muting sensitivity, also can obtain not have the image of obvious noise.

Therefore,, just can improve the pixel quantity in the solid imaging element 1030, and just can reduce the size of solid imaging element by making each pixel of solid imaging element 1030 fine.

Actual form and verify solid imaging element 1030 according to the present embodiment shown in Figure 57 with respect to each feature.

As mentioned above, carry out the preliminary treatment and the deposit of SiGeC layer, thereby on silicon substrate 1011, form SiGeC layer as single crystalline layer 1026.

Then, by photoetching and the local SiGeC layer of removing of RIE technology, thereby reservation just only forms the single crystalline layer of being made up of SiGeC 1026 thus corresponding to the part of photodiode PD on photodiode PD.

Then, by being used to make the conventional method of CCD solid imaging element, make the solid imaging element 1030 shown in Figure 57.

In the actual imaging of solid imaging element 1030, can significantly reduce dark current, even, also can obtain not have the image of obvious noise by under dark condition, carrying out imaging with manufacturing.

Replacedly, can adopt a mask only on photodiode PD, to form single crystalline layer.

For example,, cover silicon substrate 1011 with mask except the part of photodiode PD, and the surface of silicon carbide substrates 1011.As a result, just only on photodiode PD, form single crystalline layer.

Figure 58 is an accompanying drawing (profile) of showing the schematic structure of the solid imaging element that contains the single crystalline layer that forms by the method that is different from the solid imaging element shown in Figure 57.

In Figure 58, reference number 1040 expression solid imaging elements; The vertical ccd register of reference number 1002 expressions; Reference number 1011 expression silicon substrates; Reference number 1012 expression p-N-type semiconductor N well regions; Reference number 1013 expression (n-type) electric charge storage regions; Channel region is shifted in reference number 1015 expressions; Reference number 1019 expression transfer electrodes; Reference number 1021 expression optical screen films; And reference number 1027 expression single crystalline layers.PD represents photodiode.

As shown in the profile of Figure 58, solid imaging element 1040 is different from the solid imaging element shown in Figure 57, wherein is formed with single crystalline layer 1027 on photodiode PD, thereby enters silicon substrate 1011.

In each the foregoing description, all on silicon substrate 1011, form the single crystalline layer of forming by SiC or SiGeC 1025,1026 or 1027.Yet,, can adopt other material with band gap wideer than the band gap of the silicon of silicon substrate 1011 for single crystalline layer.

Hereinafter, enumerated the example of the material except SiC with band gap wideer than the band gap of silicon with lattice constant.The all material of hereinafter enumerating has the cubic system identical with silicon Si.This is because of the preferred identical cubic system in order to carry out epitaxial growth on silicon.

Material band gap Eg (eV) lattice constant

GaAS 1.43 5.654

AlAs 2.16 5.66

GaN 3.27 4.55

AlN 6.8 4.45

ZnSe 2.67 5.667

ZnS 3.70 5.41

MgSe 3.6 5.62

MgS 4.5 5.89

Here, GaAs, AlAs, GaN and AlN are III-V group element compound semiconductors, and can use and be AlGaAs ternary mixed crystal or AlGaN ternary mixed crystal.Other III-V group element compound semiconductor comprises AlGaInP quaternary mixed crystal etc.

Similarly, ZnSe and ZnS are II-IV group element compound semiconductors and can use and be ZnMgSSe quaternary mixed crystal.Other II-IV group element compound semiconductor comprises ZnMgO ternary mixed crystal etc.

The present invention can be applied to contain the solid imaging element at semiconductor layer rather than the optical-electrical converter that forms in silicon layer.

For example, the present invention can be applied to contain the solid imaging element of the optical-electrical converter (photodiode) that forms in compound semiconductor layer.

In order to survey the infrared light in 0.9 μ m～1.7 mum wavelength scopes, compound such as GaInAs is used as the semiconductor layer that wherein is formed with optical-electrical converter.

In order to survey the infrared light in 3m～5 mum wavelength scopes, compound such as InSb or PtSi are used as the semiconductor layer that wherein is formed with optical-electrical converter.

In order to survey the infrared light in 8 μ m～14 mum wavelength scopes, compound such as HgCdTe is used as the semiconductor layer that wherein is formed with optical-electrical converter repeatedly.

For example, the detection of these wave-length coverages just can apply the present invention to the photodiode and the solid imaging element (being commonly referred to " infrared emanation technology ") in order to obtain temperature information of silex glass fiber optic communication.

When any in these materials is applied to wherein being formed with the semiconductor layer of optical-electrical converter, the single crystalline layer that just will have wide bandgap bonds to surface of silicon substrate, because therefore the wide bandgap of this material has just produced the same effect with the minimizing dark current that bonding wide bandgap materials is the same between silicon substrate and SiC compound.

The single crystalline layer material is not limited to compound semiconductor, and the present invention can be applied to wherein be formed with the situation of optical-electrical converter (photodiode) in by IV family element rather than the silicon semiconductor layer that for example Ge forms.

In each the above embodiments, the present invention is applied to the CCD solid imaging element.Yet the present invention can be applied to the solid imaging element of other type, for example, and the CMOS solid imaging element.

Figure 59 is an accompanying drawing (schematic plan view) of showing the schematic structure of solid imaging element according to a further embodiment of the invention.In the present embodiment, the present invention is applied to the CMOS solid imaging element.

In Figure 59, reference number 1050 expression solid imaging elements; Reference number 1051 expression Unit Amplifier; Reference number 1052 expression vertical signal lines; Reference number 1053 expression horizontal signal lines; Reference number 1054 expression shift registers; Reference number 1055 expression noise canceller circuits; And reference number 1056 expression horizontal shifting registers.PD represents photodiode.

As shown in Figure 59, solid imaging element 1050 comprises the photodiode PD as light-receiving member, and photodiode PD is arranged in rectangular, and each photodiode PD is connected to corresponding signal lines 1052 and 1053 by Unit Amplifier 1051.Holding wire comprises vertical signal line 1052 and the horizontal signal lines 1053 that is connected to vertical transfer register 1054.Adjacent signal line 1052 and each infall of 1053 are provided with the photodiode PD of each pixel.

Horizontal signal lines 1053 is connected to the holding wire that is used for output signal voltage by noise canceller circuit 1055 with in the MOS transistor shown in the lower part of accompanying drawing.

The grid of MOS transistor is connected to horizontal shifting register 1056, so that MOS transistor is carried out conducting by horizontal shifting register 1056 and is ended.

Although it is not shown in the accompanying drawings, but in the present embodiment, at least therein be formed with on the photodiode PD that forms in the semiconductor layer in the source region of photodiode PD and MOS transistor and drain region, single crystalline layer is set, this single crystalline layer has the band gap wideer than the band gap of semiconductor layer.

The result, each the foregoing description that is applied to the CCD solid imaging element with the present invention wherein is the same, just can reduce dark current significantly, even and each pixel of solid imaging element 1050 made when reducing incident light quantity imperceptibly, also can guarantee enough S/N ratios.Therefore, the multiplication factor of the amplifier by being used to compensate muting sensitivity only just can obtain not have the image of the satisfaction of obvious noise.

Therefore, by making each pixel in the solid imaging element 1050 fine, just can reduce the pixel quantity in the solid imaging element 1050, and can reduce the size of solid imaging element.

And, the present invention not only can be applied to wherein be arranged with by matrix-style the structure of optical-electrical converter such as photodiode, but also can be applied to wherein to arrange the structure of the pixel that comprises optical-electrical converter and wherein according to the structure of delegation or multirow mode (line sensor etc.) arrangement pixel according to interlace mode (verification figure).

The present invention not only can be applied to solid imaging element, and can be applied to contain the light receiving element of the photodiode (optical-electrical converter) that forms in single-crystal semiconductor layer.

For example, the present invention can be applied to photodiode, PIN photodiode or be used as the sbhs sensor with low noise high-performance sensors.

And, can be installed on the common substrate with light receiving element according to an embodiment of the invention with such as the light receiving element of semiconductor laser or light-emitting diode, thereby formation mixed structure, perhaps, Semiconductor substrate can be used as semiconductor layer, thereby form monolithic (monolithic) optics.

In order to make light receiving element or solid imaging element according to an embodiment of the invention, can be used for forming the step and the step that forms polycrystal layer of the formation semiconductor region of optical-electrical converter and other semiconductor region according to any required order at semiconductor layer.

The invention is not restricted to the foregoing description, and can within the cited scope of the present invention, carry out various modifications.

(the setting of color separation filters; First example)

Figure 41 A, 41B and 41C are the accompanying drawings of first example of showing the setting of color separation filters.In first example, be provided for removing visible light and receiving the detecting area of also only surveying the detecting area of infrared light and being used for the visible light colour imaging.

Figure 41 A show each color filter inlay arrangement, be that the Bayer of basic colour filter arranges.In other words, pixel component has a kind of structure that the repetitive of color separation filters wherein comprises 2 * 2 pixels, so that the unit pixel in square lattice is corresponding to three colour filters of red (R), green (G) and indigo plant (B).In order to be provided for removing visible light and to receive and only survey the exploring block (detecting area) of infrared light, replace a colour filter in two green color filters with black colour filter BK.In other words, for the visible light colour imaging, regularly arranged four types colour filter with colour filter characteristic separately, four types colour filter comprise the colour filter of the wave-length coverage (colored parts) that is used for primary colors R, G and B and are used to be different from the black colour filter BK of the infrared light of primary color filter R, G and B component.

For example, be used to survey the first color (redness; R) first colour element is arranged in even number line and odd column; Be used to survey the second color (green; G) second colour element is arranged in odd-numbered line and odd column; Be used to survey the 3rd color (blueness; B) the 3rd colour element is arranged in odd-numbered line and even column; And the 4th colour element (black correction pixel) that is used to survey infrared light IR is arranged in even number line and even column.Therefore, G/B pixel or R/BK pixel just are arranged in the verification figure.In the Bayer of this basic colour filter arranges, repeat G/B or two kinds of colors of R/BK according to line direction and column direction dual mode.

Therefore, by utilizing detection through the corresponding detecting area of primary colors R, G and B, just obtained the visible light coloured image, and the same with visible images, by utilizing detection, just be independent of visible images and obtained infrared image simultaneously through the corresponding detecting area of black colour filter BK.The exploring block (detecting element) that primary color filter R, G and B wherein are set respectively will be separated into each wavelength as the visible region of transmission peak wavelength scope and survey parts separately.

Although adopt primary color filter 14R, 14G and 14B as the color filter 14 that is used for the visible light colour imaging in this example, can adopt complementary colours colour filter Cy, Mg and Ye.The exploring block (detecting element) that complementary colours colour filter Cy, Mg and Ye wherein are set respectively will be separated into each wavelength and survey parts separately as the visible region of the wave-length coverage of changing.In the case, for example, as shown in Figure 41 B, primary color filter 14R, 14G and 14B can be replaced by yellow color filter Ye, magneta colour colour filter Mg and cyan filter Cy respectively.In addition, can replace by the black colour filter BK of correction pixels at one of two magneta colours of diagonal.

In the case, on pixel 12Cy, 12Mg except a pixel that wherein is provided with the black colour filter and among the 12Ye each, form stacked deielectric-coating 1, and on each stacked deielectric-coating 1, complementary colours colour filter 14Cy, 14Mg and 14Ye are set further.In other words, receive cyan Cy, magneta colour Mg among the visible light VL and the light component of yellow Ye by corresponding complementary colours colour filter 14Cy, 14Mg and 14Ye respectively.Owing to forming stacked deielectric-coating 1 on the exploring block of each pixel that is provided with the complementary colours colour filter therein, so just have the function of effectively cutting apart infrared light.

Replace the combination of complementary colours colour filter Cy, Mg and Ye, can adopt combination as the green color filter G and the white colour filter W of one of primary color filter.In the case, its pixel as the black colour filter BK of correction pixels can also be set.For example, as shown in Figure 41 C, made up therein in two kinds of complementary colours colour filter Cy and Mg and the field storing frequencies interleaving systems (fields torage frequency interleave system), can replace among two primary color filter G in four pixels by the black colour filter BK of correction pixels as the green color filter G of primary color filter.

(first example in the sensor construction; Corresponding to CCD)

Figure 42 and 43 is accompanying drawings of the CCD solid imaging element arranged of the color separation filters shown in key diagram 41A, 41B and the 41C, so that absorbs the image in two kinds of wavelength components of infrared light IR and visible light VL separately simultaneously.Figure 42 is a sketch (perspective view) of showing an a kind of example of structure, and Figure 43 is an accompanying drawing of showing the cross-section structure on adjacent substrate surface.Two accompanying drawings have been showed an example of the CCD solid imaging element 101 that is applied to adopt stacked deielectric-coating 1.

In each accompanying drawing, the stacked deielectric-coating of reference number 1 expression; Reference number 11 expression spectral image sensors; And reference number 12 expression unit pixel matrixes.

In the structure of the CCD solid imaging element shown in Figure 42, only demonstrate the unit pixel matrix 12 that contains four pixels.Yet in fact, unit pixel matrix 12 repeats in horizontal and vertical mode.

In four pixels, in pixel 12IR, do not form stacked deielectric-coating 1, and black colour filter 14BK is set in pixel 12Ir, so that only receive infrared light IR by black colour filter 14BK with the periodic arrangement of unit pixel matrix 12.In other words, adopt black colour filter 14BK, remove visible light VL thus, and only receive infrared light IR as color filter 14 at the pixel IR that is used for infrared light IR.The pixel 12IR that wherein is provided with black colour filter 14BK also is called " black picture element 12BK ".

On the other hand, on each pixel of other pixel 12B, 12G and 12R, stacked deielectric-coating 1 is set, and primary color filter 14R, 14G and 14B further are set thereon, so that receive the primary colors component of the blue B among the visible light VL, green G and red R respectively by corresponding primary color filter 14R, 14G and 14B.In other words, be respectively arranged with therein on the exploring block of each pixel of primary color filter and form stacked deielectric-coating 1, just have the function of effective removal infrared light thus.The circuit structure that is adopted has been shown among Figure 29.

The Figure 43 that shows the cross-section structure on adjacent substrate surface shows the pixel that it receives only visible light VL.Its pixel 12IR that receives infrared light IR has a kind of structure that does not comprise stacked deielectric-coating 1 and black colour filter 14BK.In other words, the same with the manufacturing process shown in Figure 33 A～33F, by the CVD method, successful deposit SiN film and SiO ₂Film, thus stacked deielectric-coating formed with structure shown in Figure 13.Then, by photoetching and RIE method, only receive and remove stacked deielectric-coating the infrared light from pixel.Then, deposit SiO once more ₂Film is so that planarized surface.

Have the image device of said structure by employing, just can obtain visible light coloured image and infrared image simultaneously based on former color component.In other words, when its black colour filter 14BK that absorbs visible light VL is set to color filter 14C, just absorb visible light VL, just can obtain infrared IR image thus according to view data from the pixel 12IR that is used for infrared light IR by black colour filter 14BK.

(the arrangement of color separation filters; Second example)

Figure 44 A, 44B and 44C are the accompanying drawings of showing second example of color separation filters arrangement.In second example, be provided for receiving and surveying the detecting area and the detecting area that is used for the visible light colour imaging of all wavelengths component of visible light with infrared light.

The Bayer that Figure 44 A shows basic color filter arranges.In other words, pixel component has a kind of structure that the repetitive of color separation filters wherein comprises 2 * 2 pixels, so that the unit pixel in square lattice is corresponding to three colour filters of red (R), green (G) and indigo plant (B).In order to be provided for receiving and surveying the exploring block (detecting area) of all wavelengths component of visible light, replace a colour filter in two green color filters with white colour filter W with infrared light.In other words, for the visible light colour imaging, regularly arranged four types color filter with colour filter characteristic separately, this colour filter comprise the colour filter of the wave-length coverage (rainbow member) that is used for primary colors R, G and B and are used to be different from the white filter W of the infrared light of primary color filter R, G and B component.

The all wavelengths component (especially near infrared light) of the white pixel transmission that wherein is provided with white colour filter W in from visible light to the infrared light scope.According to this viewpoint, in fact just color filter can be set.

For example, be used to survey the first color (redness; R) first colour element is arranged in even number line and odd column; Be used to survey the second color (green; G) second colour element is arranged in odd-numbered line and odd column; Be used to survey the 3rd color (blueness; B) the 3rd colour element is arranged in odd-numbered line and even column; And the 4th colour element (white pixel) that is used to survey infrared light IR is arranged in even number line and even column.Therefore, G/B pixel or R/W pixel just are arranged in alternating graph.In the Bayer of this basic colour filter arranges, repeat G/B or two kinds of colors of R/W according to line direction and column direction dual mode.

Therefore, by utilizing detection through the corresponding detecting area of primary colors R, G and B, just obtained the visible light coloured image, and, just be independent of visible images and obtained the vision-mix of infrared image or infrared light and visible light simultaneously with visible images by utilizing detection through the corresponding detecting area of white colour filter W.For example, come from the pixel data of the pixel 12IR of its blending ingredients that receives visible light VL and infrared light IR, just obtained the image of the blending ingredients of visible light VL and infrared light IR by employing.Similarly, the image and the visible VL image of the blending ingredients of visible light VL and infrared light IR have just been obtained.Yet, adopt the poor of two kinds of images, just can only obtain infrared IR image.

Although adopt primary color filter 14R, 14G and 14B as the color filter 14 that is used for the visible light colour imaging in this example, can adopt complementary colours colour filter Cy, Mg and Ye.In the case, for example, as shown in Figure 44 B, primary color filter 14R, 14G and 14B can be replaced by yellow color filter Ye, magneta colour colour filter Mg and cyan filter Cy respectively.In addition, can replace by the white colour filter W that is used for infrared imaging at one of two magneta colours of diagonal.

In the case, on pixel 12Cy, 12Mg except a pixel that wherein is provided with white colour filter and among the 12Ye each, form stacked deielectric-coating, and on each stacked deielectric-coating 1, complementary colours colour filter 14Cy, 14Mg and 14Ye are set further.In other words, receive cyan Cy, magneta colour Mg among the visible light VL and the light component of yellow Ye by corresponding complementary colours colour filter 14Cy, 14Mg and 14Ye respectively.Owing to forming stacked deielectric-coating 1 on the exploring block of each pixel that is provided with the complementary colours colour filter therein, so just have the function of effectively cutting apart infrared light.

Replace the combination of complementary colours colour filter Cy, Mg and Ye, can adopt combination as the green color filter G and the complementary filters W of one of primary color filter.In the case, its pixel as the white colour filter W of correction pixels can also be set.For example, as shown in Figure 44 C, made up therein in two kinds of complementary colours colour filter Cy and Mg and the field storing frequencies interleaving systems (field storage frequency interleave system), can replace among two primary color filter G in four pixels by the white colour filter W of correction pixels as the green color filter G of primary color filter.

Because white correction pixels 12W has the sensitivity in the wide wave-length coverage scope that comes from VL and infrared light IR, so compare with the pixel that is used for the visible light colour imaging (wherein being provided with the pixel of primary color filter), signal just is easy to saturated, and saturated phenomenon will become a problem, and imaging will become a problem under bright light environments especially.Particularly, the infrared image that under bright light environments, just can not obtain to be fit to.

In order to solve saturation problem, can control the detection time of the exploring block that wherein is provided with white colour filter W by drive control component 146.For example, under bright light environments, just can control and carry out high speed imaging by the exposure of adopting shutter function (comprising mechanical shutter and electronic shutter).For example, can carry out the exposure of image device with the short period, thus from the image device (especially, exploring block) the read pixel signal, and picture element signal is transmitted through picture signal processing unit 140.

In the case, the speed that for example surpassed for 60 frame/seconds is at a relatively high speed exposed and signal reads, thereby produce the increase effect on saturated.Selectively, can be simply carrying out signal within less than 0.01667 second time (memory time) reads.In the case, can adopt overcurrent that charge signal is discharged into substrate side, so that within the short time, effectively read charge stored.

More preferably, expose under the high speed of 240 frame/seconds and signal reads surpassing, so that further improve saturation effect.Selectively, can carry out signal simply within the time (memory time) less than 4.16 microseconds reads.

Being not limited to proofread and correct white pixel 12W from its reading electric charges so that prevent saturated pixel within the short time (memory time), perhaps can be all pixels that contain other pixel (wherein being respectively arranged with the former color pixel of primary color filter) that is useful on the visible light colour imaging.

Replacedly, the signal that can be integrated in for twice within the short exposure time reads, thereby the weak signal in the black position is converted to strong signal, increases the S/N ratio thus.In the case, for example, by under dark surrounds and bright light environments, carrying out imaging, just sensitivity that can obtain to be fit to and high S/N ratio.In other words,, just can prevent saturated in white correction pixels 12W, and can realize wide dynamic range by the set of signals achievement by high speed imaging.

(first example in the sensor construction; Corresponding to CCD)

Figure 45 illustrates that it has the accompanying drawing of the CCD solid imaging element shown in Figure 44 A, 44B and the 44C, so that obtains the image of two kinds of wavelength components of visible light VL and infrared light IR separately simultaneously.Figure 45 is a sketch (perspective view) of showing an example of the CCD solid imaging element 101 be applied to adopt stacked deielectric-coating.The cross-section structure on adjacent substrate surface is identical with the cross-section structure shown in Figure 43.

In the structure of the CCD solid imaging element 101 shown in Figure 45, only show the unit pixel matrix 12 that contains four pixels.Yet in fact, unit pixel matrix 12 repeats in horizontal and vertical mode.

Among four pixels with the periodic arrangement of unit pixel matrix 12, in pixel 12IR, do not form stacked deielectric-coating 1, color filter 14 is not set in pixel 12Ir, so that need not passes color filter 14 and receive infrared light IR.In the case, pixel 12IR just receives the blending ingredients of infrared light IR and visible light VL.Be called " white pixel 12W " or " gamut transmissive pixel " at its pixel 12IR that is not provided with in each pixel of color filter 14.

Do not form therein among each pixel 12IR of stacked deielectric-coating 1, color filter 14 is not set in white pixel 12W, so that infrared light IR and visible light VL act on signal simultaneously.In the case, the pixel 12IR that is used for infrared light IR is just not only as the pixel of infrared light IR but also as the pixel of infrared light IR and visible light VL.

On the other hand, on each pixel of other pixel 12B, 12G and 12R, stacked deielectric-coating 1 is set, and primary color filter 14R, 14G and 14B further are set thereon, so that receive the primary colors component of the blue B among the visible light VL, green G and red R respectively by corresponding primary color filter 14R, 14G and 14B.In other words, be respectively arranged with therein on the exploring block of each pixel of primary color filter and form stacked deielectric-coating 1, just have the function of effective removal infrared light thus.The same with primary color filter 14R, the 14G and the 14B that are adopted among second embodiment, can adopt first example shown in Figure 46 A.The circuit structure that is adopted has been shown among Figure 29.

Have the image device of said structure by employing, just can obtain vision-mix simultaneously based on visible light coloured image and infrared IR image or the infrared light IR and the visible light VL of former color component.For example, adopt pixel data, just can obtain the image of the blending ingredients of infrared light IR and visible light VL, improve sensitivity thus from the pixel 12IR that is used for its blending ingredients that receives infrared light IR and visible light VL.Similarly, not only can obtain the image and the visible VL image of the blending ingredients of infrared light IR and visible light VL, and just can obtain infrared IR image by the difference that adopts two images.

Although not shown in the accompanying drawings, can replace white colour filter 14W by the green color filter of the G rainbow member in its visible light transmissive scope and infrared component and other component of filtering (B component in the visible-range and R component).In other words, can be provided for receiving and surveying the detecting area of the specific wavelength component in infrared light and the visible light.

In the case, employing is from the pixel data of the pixel 12IR of the blending ingredients that is used for its specific wavelength component that receives infrared light IR and visible light VL, just can obtain the specific wavelength component among infrared light IR and the visible light VL blending ingredients image, improve sensitivity thus.Similarly, not only can obtain the image and the visible VL image of blending ingredients, and just can obtain infrared IR image by the difference that adopts the specific wavelength component among infrared light IR and the visible light VL.

(other example that color separation filters is arranged)

Figure 46～52nd illustrates the accompanying drawing that known pixels that resolution descends is arranged.Arrange about pixel, when the arrangement adopted shown in Figure 41 or 44, the pixel that will be used to survey infrared light (or blending ingredients of infrared light and visible light) is added into the pixel that is used for visible light that wherein is provided with conventional primary color filter of RGB or CyMgYe complementary colours colour filter (or primary color filter G).

For example, replace being used for the green pixel G or the magneta colour pixel Mg of visible light colour imaging by black correction pixel, white pixel, green correction pixel or magneta colour correction pixels, just might reduce the resolution of any visible light coloured image and infrared image thus.For example, when a G pixel in conventional Bayer arranges was replaced by red pixel, resolution will reduce.Yet, when suitable spread correction pixel and its significantly help the pixel (for example, green pixel G) of the wavelength component of resolution, just can solve the problem of resolution.

In the case, important part is, similar with conventional structure, arrange in the color separation filters arrangement of color filter in the embedded figure mode therein, just can be according to the embedded figure mode, utilize predetermined lattice spacing arrange the pixel that is used for infrared light (blending ingredients that is used for infrared light and visible light), and just can be according to the embedded figure mode, utilize the lattice spacing of being scheduled to arrange a kind of color of primary colors RGB of visible light or the pixel of complementary color Cy, Mg and Ye.

Embedded figure means that a kind of color pixel is arranged as predetermined lattice spacing according to the lattice mode, and a kind of color pixel can be not adjacent one another are.In the representative instance that a kind of therein color pixel pixel adjacent one another are is arranged, alternately arrange the square pixels of infrared light and the square pixels of other color according to grid graph (alternating graph) mode.Selectively, selectively arrange a kind of square pixels of color of a kind of color of main color RGB of visible light or complementary color CyMgYe and the square pixels of other color according to grid graph mode (alternating graph).

(for the application example of primary color filter)

For example, for the resolution that adopts the RGB primary color filter to suppress the visible light coloured image descends, just can keep the arranging density of the G pixel in the visible-range, and the pixel of the R of visible-range or B can be corrected black picture element, white pixel or green pixel replacement.For example, as shown in Figure 46 A and B, in 2 * 2 unit pixel matrixes 12, comprise: the colour element G that is used for surveying the green color component of visible-range is arranged in odd-numbered line and odd column and even number line and even column, and proofreaies and correct black picture element (Figure 46 A), white pixel (Figure 46 B) or green pixel (not shown) and be arranged in even number line and odd column.

In addition, in the unit pixel matrix 12 of odd column mode, odd-numbered line place in the odd location picture element matrix 12 on column direction and even column place are provided for surveying the colour element B of the blue color component in the visible-range, and the odd-numbered line place in the even location picture element matrix 12 on column direction and even column place are provided for surveying the colour element R of the red component of visible-range.In the unit pixel matrix 12 in even column, according to colour element B and colour element R being set with above-mentioned opposite arrangement mode.In a word, the repetition period of color filter 14 is finished by 2 * 2 unit pixel matrixes 12.

In the arrangement shown in Figure 46 A and the 46B, alternately arrange a kind of color pixel and other color pixel among the main color RGB of visible light according to the alternating graph mode.In this arrangement, the density of colour element G that significantly is of value to the resolution of visible light coloured image is set to arrange identically with Bayer, just can prevent the decline of the resolution of visible light coloured image thus.

Yet the arranging density of colour element R and colour element B is 1/2 of a Bayer arranging density, has therefore reduced the resolution of color.Yet people are lower than visuality for blue G for the visuality of red R and blue B, so the decline of color resolution just can not become big problem.On the other hand, in adopting the infrared image of correction pixels, the arranging density of correction pixels be used for surveying visible-range green color component colour element G arranging density 1/2, so resolution just is lower than the resolution of visible light coloured image.

For example, by the manufacturing process shown in Figure 33 A～33F, the CMOS solid imaging element (image element circuit structure shown in Figure 31) that test manufacturing has the layer structure shown in Figure 31 (the same with the structure shown in Figure 35, corresponding to the cross-section structure that is used to receive visible light).In the CMOS solid imaging element, adopt black colour filter 14BK, and arrange the black correction pixel according to the graphics mode shown in Figure 46 A with the transmitted spectrum characteristic shown in Figure 47.As a result, will find, obtain the high-resolution coloured image and relative high-resolution infrared image of main color simultaneously with the resolution that is lower than coloured image.

In addition, by the manufacturing process shown in Figure 33 A～33F, the CMOS solid imaging element (image element circuit structure shown in Figure 31) that test manufacturing has the layer structure shown in Figure 37 (the same with the structure shown in Figure 35, corresponding to the cross-section structure that is used to receive visible light).In the CMOS solid imaging element, arrange white pixel according to the graphics mode shown in Figure 46 B.As a result, will find obtained the image of the blending ingredients of the high-resolution coloured image of main color and infrared light and visible light simultaneously, and the image of blending ingredients to have the relative high-resolution of the resolution that is lower than coloured image.Also can find,,, just obtain infrared image simultaneously by reducing blueness, redness and the green density of surveying by former color pixel R, G and B respectively for visible light.

Saturated in order to prevent, all pixels of can exposing within the short time so that reading electric charges information, and can read the signal within the short period and are integrated into twice, are converted to big signal thus.Therefore, in addition under dark surrounds or under bright light environments, all sensitivity that can obtain to be fit to, and expanded dynamic range.

And the structure that has wherein made up the structure of this black correction pixel shown in Figure 46 A and multilayer film or wherein made up this white pixel shown in Figure 46 B and multilayer film has demonstrated the same effect on CCD structure shown in Figure 43 and the CMOS solid imaging element.

For the resolution that suppresses infrared image descends, as shown in Figure 48 A and 48B, can be with proofreading and correct the colour element G that black picture element (Figure 48 A), white pixel (Figure 48 B) or green pixel (not shown) exchange the green color component that is used to survey the visible-range shown in Figure 46 A.In the case, just alternately arrange the infrared image element the same and other color pixel with correction pixels according to the alternating graph mode.In this arrangement, density that can correction pixels is set to identical with density during Bayer arranges, has just prevented that thus the resolution of infrared image from descending.Yet, the arranging density of colour element G that significantly is of value to the resolution of visible light coloured image be correction pixels arranging density 1/2, so the resolution of visible light coloured image will be lower than the resolution of infrared image.This color resolution identical with shown in Figure 46 A and the 46B.

For example, CMOS solid imaging element (image element circuit structure shown in Figure 29, and the same with shown in Figure 43 are corresponding to the cross-section structure of the pixel that is used to receive visible light) is made in test.In the CMOS solid imaging element, adopt the black colour filter 14BK that demonstrates the transmitted spectrum characteristic shown in Figure 47, and arrange the black correction pixel according to the graphics mode shown in Figure 48 A.As a result, will find, obtain high-resolution infrared image and relative high-resolution visible light coloured image simultaneously with the resolution that is lower than infrared image.

In addition, CMOS solid imaging element (image element circuit structure shown in Figure 29, and the same with shown in Figure 43 are corresponding to the cross-section structure of the pixel that is used to receive visible light) is made in test.In the CMOS solid imaging element, arrange white correction pixels according to the graphics mode shown in Figure 48 B.As a result, will find, obtain the high-definition picture of the blending ingredients of infrared light and visible light.Also can find, for visible light, by reducing blueness, redness and the green intensity of surveying by former color pixel R, G and B respectively, just obtained infrared image simultaneously, obtained to have the high-resolution relatively visible light coloured image of the resolution that is lower than infrared image.

Further be sure of, in any image device, need not infraredly cut apart colour filter even in infrared environment, also can carry out the imaging of high color reproducibility.Further find, in the structure that has adopted white pixel, the visible light component that employing obtains from white pixel is proofreaied and correct the luminance signal that obtains based on the visible images of main color, is independent of color rendition thus and the sensitivity that further improved the visible light coloured image.

Saturated in order to prevent, adopt overcurrent, within the short time, only read the electric charge of white pixel, and the signal within the short period can be read and be integrated into twice, just be converted to big signal thus.Therefore, in addition under dark surrounds or under bright light environments, the sensitivity that can both obtain to be fit to, and expanded dynamic range.

And the structure that has wherein made up the structure of this black correction pixel shown in Figure 48 A and multilayer film or wherein made up this white pixel shown in Figure 48 B and multilayer film has demonstrated the same effect on CCD structure shown in Figure 43 and the CMOS solid imaging element.

Figure 49 A, 49B and 49C be illustrate wherein with the visible light coloured image irrelevant and be provided with the accompanying drawing of other example that the pixel of the pixel that is used to obtain infrared image arranges.In these examples,, make up a plurality of color filters for the pixel that is used to obtain infrared image.For example, be combined in the example shown in each Figure 49 A, 49B and the 49C, first example and second example, and in unit pixel matrix 12,, alternately arrange black colour filter 14BK and white colour filter 14W for the pixel that is used to obtain infrared image.Figure 49 A shows the combination of Figure 41 and 44, and Figure 49 B shows the combination of Figure 46 A and 46B, and Figure 49 C shows the combination of Figure 48 A and 48B.

In comprising these arrangements of making up each combination, for example,, mainly adopt white pixel 12W for improving sensitivity; For keeping normal brightness and high brightness, then adopt black picture element 12BK.By making up the output of two types of pixels, just can realize the wide region of the reproducibility scope from the low brightness level to the levels of brightness, and can expand dynamic range.

(for the application example of complementary filters)

In order to adopt CyMgYe complementary colours colour filter to suppress the decline of visible light color image resolution, just can keep the arranging density of the Mg pixel in the visible-range, and the pixel R of visible-range or B can be replaced by the correction black picture element that is used to obtain infrared image, white pixel or green pixel.For example, as shown in Figure 50 A and B, in 2 * 2 unit pixel matrixes 12, the colour element Mg that is used for surveying the magneta colour component of visible-range is arranged in odd-numbered line and even column, and the black picture element (Figure 50 A), white pixel (Figure 50 B) or the magneta colour pixel (not shown) that are used to obtain infrared image are arranged in even number line and odd column.In addition, a kind of magneta colour pixel Mg can be replaced by green pixel G.

In the case, alternately arrange pixel Mg and other color pixel of a kind of complementary color Cy, Mg and the Ye of visible light according to interlace mode.Under this arrangement mode, the density of colour element Mg that significantly is of value to the resolution of visible light coloured image is set to identical with the Bayer arrangement mode, prevents the decline of visible images resolution thus.

Yet, the arranging density of colour element Cy and Ye be colour element Mg arranging density 1/2, therefore reduced color resolution.Yet people are visual low for color, so the decline of color resolution just can not become big problem.On the other hand, in the infrared image that adopts correction pixels, the arranging density of correction pixels (infrared image element) be used for surveying visible-range the magneta colour component colour element Mg arranging density 1/2, so resolution will be lower than the resolution of visible light coloured image.

In order to suppress the decline of infrared image resolution, as shown in Figure 51 A and 51B, can exchange the colour element Mg of the magneta colour component that is used to survey visible-range with the magneta colour colour element (not shown) of proofreading and correct black picture element (Figure 51 A), white pixel (Figure 51 B) or being used to obtain infrared image.In the case, alternately arrange the infrared image element the same and other color pixel according to the alternating graph mode with correction pixels.Under this arrangement mode, density that just can correction pixels is set to the same with the Bayer arrangement mode, and just anti-thus resolution that has made infrared image descends.Yet, the arranging density of colour element Mg that significantly is of value to the resolution of visible light coloured image be correction pixels arranging density 1/2, so the resolution of visible light coloured image just is lower than the resolution of infrared image.This color resolution identical with shown in Figure 50 A and the 50B.

Though be used for suppressing the above-mentioned arrangement that resolution descends, according to embedded figure mode (in representative instance, alternating graph), arrange green G or magneta colour Mg pixel, (for example arrange other color pixel according to the alternating graph mode with high as far as possible density, R and B, or Cy and Ye).In the case, can obtain identical in fact effect.Certainly, in order to improve resolution and color identifying ability, preferably according to the embedded figure mode, arrange colour filter with high visual rainbow member with high as far as possible density.

(for the application example of oblique arrangement)

Although in above-mentioned example, arrange color filter, can arrange color filter according to inclination lattice mode according to square lattice mode.For example, in the arrangement shown in Figure 52 A, the rotation of the arrangement shown in Figure 46 B 45 is spent according to clockwise direction.In the arrangement shown in Figure 53 B, the rotation of the arrangement shown in Figure 48 B 45 is spent according to clockwise direction.In this manner, in oblique arrangement, will be in vertical direction and increase picture element density in the horizontal direction, further improved the resolution on the both direction thus.

Though described the present invention with reference to each embodiment, technical field of the present invention is not limited to the foregoing description, and within the cited scope of the present invention, can carry out variations and modifications.These variations and modification are contained in technical field of the present invention.

Similarly, the invention is not restricted to each embodiment, and all combinations described in each embodiment can be used to solve each problem.Each embodiment in the various embodiments described above comprises various modifications, can adopt described a plurality of features in any required combination.Some features described in each embodiment can be removed, and can obtain effect of the present invention.

Said structure not merely is various embodiments of the present invention, as mentioned above, to have a kind of wherein multilayer and have between adjacent layer different refraction coefficients, every layer and have the stacked member (stacked deielectric-coating) of the structure of predetermined thickness and allow wave spectrum to separate in order to use, can adopt another kind of similarly structure, this stacked member also has the predetermined wavelength component of reflection electromagnetic wave and the characteristic of transmission remainder (remainder).

And above-mentioned technology is not limited to a kind of technology that chromatic dispersion is visible light and infrared light that is used for.For example, light can be visible light and ultraviolet light by chromatic dispersion and can be detected, and ultraviolet light can be detected with visible light, thereby forms an image.And the image of the visible light of Tan Ceing is not limited to the monochrome image without chromatic dispersion simultaneously, and can aforesaidly be used for versicolor color filter by adopting, the visible frequency band chromatic dispersion is surveyed coloured image for the three primary colors component.

Therefore, can be with its visual picture that can see with eyes (monochrome image or coloured image), obtain the information of the ultraviolet light that can see with eyes simultaneously.As a result, the present invention can also be applied to the Primary Component of fresh information system such as synchronous monitor camera of optics etc.

For example, by employing be used for the visible light VL the same with the reflected wavelength range component and the same with the transmission peak wavelength ranges of components less than visible light VL the wavelength side (for example, ultraviolet light) stacked deielectric-coating 1 just can carry out chromatic dispersion and be visible light VL and less than the wavelength side of visible light VL and their detection.

Though it is not shown in the accompanying drawings, in the arrangement shown in Figure 41 A～41C, can be on the pixel 12IR of four pixels under the periodic arrangement mode of unit pixel matrix 12, be formed under greater than the wavelength of visible light VL wavelength, reflecting the stacked deielectric-coating 1 of a component, thereby receive less than the light (ultraviolet light) on the wavelength side of visible light VL.In addition, on each pixel of other three pixel 12B, 12G and 12R, do not form stacked deielectric-coating 1, but color filter 14 (14R, 14G and 14B) is set thereon, so that receives blue B, green G in the visible light VL of lower wavelength side (ultraviolet light) and the three primary colors component of red R.

In order to obtain the signal of the visible light VL the same, preferably between the ultraviolet light component the same with the reflected wavelength range component, carry out arithmetical operation with the transmission peak wavelength ranges of components with its reflected wavelength range component that not influenced by the transmission peak wavelength ranges of components.As color filter 14 (14R, 14G and 14B), can adopt the color filter of 0 transmissivity that has the ultraviolet light the same basically with the transmission peak wavelength ranges of components.In the case, by color filter (14R, 14G and 14B), just will just can cancel arithmetical operation thus less than the component on the visible light VL wavelength side (ultraviolet light).

It will be appreciated by those skilled in the art that according to design needs and other factor and just can carry out various modifications, combination, inferior combination and replacement that they will belong to the scope that accessory claim or they are equal in this scope.

Claims (39)

1. one kind is used to adopt a kind of device to obtain the method for physical message based on unit signal, this device is used to survey the physical distribution that is used for predetermined purpose, this device comprises: as unit block, one is used to survey electromagnetic exploring block and one and is used for the unit signal production part that produces corresponding unit signal and export this unit signal according to the electromagnetic wave amount of being surveyed by this exploring block, and according to predefined procedure this unit block is set on same substrate;

Wherein this exploring block comprises having a stacked member that wherein is stacked with the structure of multilayer, this multilayer has different refraction coefficients and every layer and has predetermined thickness between adjacent layer, this stacked member is set on the incidence surface side that electromagnetic wave incides, this stacked member has the characteristic of this electromagnetic predetermined wavelength range component of reflection, and the transmission residual term; And

Survey transmission this transmission peak wavelength ranges of components by this exploring block, and obtain the physical message that is used for predetermined purpose based on the unit signal of this transmission peak wavelength ranges of components that obtains from this unit signal production part by this stacked member.

According to claim 1 be used to adopt a kind of device to obtain the method for physical message based on unit signal, wherein, this stacked member is not set on the incidence surface side that electromagnetic wave incides in being different from other exploring block of described exploring block, this other exploring block is not the exploring block that is used for the transmission peak wavelength ranges of components, so that survey the reflected wavelength range component by this other exploring block, obtain to be used for the physical message of the second predetermined purpose thus based on the unit signal of the reflected wavelength range component that obtains from this unit signal production part.

According to claim 2 be used to adopt a kind of device to obtain the method for physical message based on unit signal, wherein, select and output based on first physical message of the unit signal of this transmission peak wavelength ranges of components with based on one of second physical message of the unit signal of this reflected wavelength range component, perhaps export two physical messages simultaneously.

According to claim 1 be used to adopt a kind of device to obtain the method for physical message based on unit signal, further be included in and be used to survey the optical component that is provided with on a plurality of exploring blocks incidence surface side separately of transmission peak wavelength ranges of components, be used for this transmission peak wavelength ranges of components is separated into different wave-length coverage components;

Wherein survey separately transmission peak wavelength ranges of components by these a plurality of exploring blocks respectively, and the unit signal that makes up this transmission peak wavelength ranges of components separately obtains other physical message about this transmission peak wavelength ranges of components, and this unit signal obtains from this unit signal production part.

According to claim 4 be used to adopt a kind of device to obtain the method for physical message based on unit signal, wherein,, adopt transmitted light in the visible-range wherein to comprise the primary color filter of trichromatic wavelength component as optical component.

According to claim 4 be used to adopt a kind of device to obtain the method for physical message based on unit signal, wherein,, adopt transmitted light in the visible-range wherein to comprise the complementary colours colour filter of each trichromatic complementary colours as optical component.

According to claim 2 be used to adopt a kind of device to obtain the method for physical message based on unit signal, wherein, the difference processing of the unit signal by the reflected wavelength range component and the unit signal of transmission peak wavelength ranges of components, the negligible physical message of influence that obtains transmission peak wavelength ranges of components wherein is as second physical message.

According to claim 7 be used to adopt a kind of device to obtain the method for physical message based on unit signal, comprising:

On the incidence surface side that the exploring block electromagnetic wave that is used for surveying the transmission peak wavelength ranges of components incides, optical component is set, is used for the predetermined wavelength component of this transmission peak wavelength ranges of components of transmission;

On the incidence surface side that other exploring block electromagnetic wave that is used for surveying the reflected wavelength range component incides, optical component is set, is used for the predetermined wavelength component of this reflected wavelength range component of transmission and this transmission peak wavelength ranges of components; And

Survey the synthetic component of predetermined wavelength component in this transmission peak wavelength ranges of components and this reflected wavelength range component and the predetermined wavelength component of this transmission peak wavelength ranges of components by each exploring block, thereby obtain the negligible physical message of influence of transmission peak wavelength ranges of components wherein as second physical message based on unit signal from the different wavelength range component of this unit signal production part acquisition.

According to claim 2 be used to adopt a kind of device to obtain the method for physical message based on unit signal, comprising:

On the incidence surface side that other exploring block electromagnetic wave that is used for surveying the reflected wavelength range component incides, optical component is set, is used for this reflected wavelength range component of transmission and this transmission peak wavelength ranges of components of filtering; And

Survey the reflected wavelength range component of having got rid of the transmission peak wavelength ranges of components by this exploring block, thereby the physical message of influence that obtains wherein can to ignore this transmission peak wavelength ranges of components based on the unit signal of the reflected wavelength range component that obtains from this unit signal production part is as second physical message.

10. one kind is adopted the next device that obtains physical message based on unit signal of a kind of device, this device is used to survey the physical distribution that is used for predetermined purpose, this device comprises: as unit block, living corresponding unit signal of electromagnetic volume production that is used to survey electromagnetic exploring block and is used for surveying and the unit signal production part of exporting this unit signal according to exploring block, and this unit block is arranged on the same substrate according to predefined procedure, and this device comprises:

A stacked member that on the light incident side of this exploring block, is provided with, electromagnetic wave incides this light incident side, this stacked member has a structure and also has the predetermined wavelength range component of reflection electromagnetic wave and the characteristic of transmission residual term, has a plurality of layers that different refractivity and each layer all have predetermined thickness in this structure between the stacked adjacent layer; And

Signal processing unit is used for according to surveying by sensing element and transmission obtains to be used for the physical message of predetermined purpose by the unit signal of the transmission peak wavelength ranges of components of this stacked member, and this unit signal obtains from the unit signal production part.

11. come to obtain based on unit signal the device of physical message according to a kind of device of the employing of claim 10, wherein, this stacked member and this exploring block integrate.

12. come to obtain the device of physical message according to a kind of device of the employing of claim 11 based on unit signal, wherein, this Signal Processing Element obtains the physical message that is used for second purpose based on the unit signal of reflected wavelength range component, this unit signal be from this unit signal production part according to by sensing element rather than be used to survey the reflected wavelength range component that the sensing element of transmission peak wavelength ranges of components surveys and obtain, this reflected wavelength range component can transmission by this stacked member.

13. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, further comprise: signal switching controls parts, be used to control this Signal Processing Element, thereby select and output is output simultaneously based on first physical message of the unit signal of transmission peak wavelength ranges of components with based on one of second physical message of the unit signal of reflected wavelength range component or both.

14. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, further comprise: survey the optical component that is provided with on the exploring block incidence surface side separately of transmission peak wavelength ranges of components in a plurality of being used to, be used for the transmission peak wavelength ranges of components is separated into separately wave-length coverage component;

Wherein according to each transmission peak wavelength ranges of components of surveying by a plurality of exploring blocks respectively, this Signal Processing Element obtains other physical message about the transmission peak wavelength ranges of components by the unit signal that makes up each transmission peak wavelength ranges of components, and this unit signal obtains from this unit signal production part.

15. come to obtain based on unit signal the device of physical message according to a kind of device of the employing of claim 14, wherein, this optical component is a primary color filter, the light in this primary color filter in the visible-range of transmission comprises trichromatic wavelength component.

16. come to obtain based on unit signal the device of physical message according to a kind of device of the employing of claim 14, wherein, this optical component is the complementary colours colour filter, the light in this complementary colours colour filter in the visible-range of transmission comprises each trichromatic complementary colours.

17. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, wherein, the difference processing of the unit signal by the reflected wavelength range component and the unit signal of transmission peak wavelength ranges of components, this Signal Processing Element are obtained the negligible physical message of influence of transmission peak wavelength ranges of components wherein as second physical message.

18. come to obtain the device of physical message according to a kind of device of the employing of claim 17, further comprise based on unit signal:

Optical component on the incidence surface side that the exploring block electromagnetic wave that is used for surveying the transmission peak wavelength ranges of components incides is used for the predetermined wavelength component of this transmission peak wavelength ranges of components of transmission; And

Optical component on the incidence surface side that another exploring block electromagnetic wave that is used for surveying the reflected wavelength range component incides is used for the predetermined wavelength component of this reflected wavelength range component of transmission and this transmission peak wavelength ranges of components; And

Wherein, by according to the predetermined wavelength component of transmission peak wavelength ranges of components and the unit signal of the predetermined wavelength component of the transmission peak wavelength ranges of components that from the unit signal production part, obtains and according to the predetermined wavelength component of reflected wavelength range component and transmission peak wavelength ranges of components and the difference processing between the unit signal of the synthetic component that obtains from the unit signal production part, this Signal Processing Element obtains the physical message of the influence that wherein can ignore the transmission peak wavelength ranges of components as second physical message.

19. come to obtain the device of physical message according to a kind of device of the employing of claim 12, further comprise based on unit signal:

Optical component on the incidence surface side that the exploring block electromagnetic wave that is used for surveying the reflected wavelength range component incides is used for this reflected wavelength range component of transmission and filtering transmission peak wavelength ranges of components;

Wherein, by use the unit signal of the reflected wavelength range component that obtains according to the reflected wavelength range component of having got rid of the transmission peak wavelength ranges of components from the unit signal production part, this Signal Processing Element obtains the physical message of the influence that wherein can ignore the transmission peak wavelength ranges of components as second physical message.

20. come to obtain the device of physical message according to a kind of device of the employing of claim 10, the j layer material that constitutes this stacked member expression formula A that meets the following conditions based on unit signal:

0.9×λ ₀/4n≤dj≤1.1×λ ₀/4n …A

Wherein dj is a thickness, λ ₀Be the centre wavelength of reflected wavelength range component.

21. come to obtain the device of physical message according to a kind of device of the employing of claim 10 based on unit signal, wherein, the central wavelength lambda of this reflected wavelength range component ₀The expression formula that meets the following conditions B:

780nm≤λ ₀≤1100nm …B

22. come to obtain the device of physical message according to a kind of device of the employing of claim 21 based on unit signal, wherein, the central wavelength lambda of this reflected wavelength range component ₀Be 900nm.

23. come to obtain the device of physical message according to a kind of device of the employing of claim 10 based on unit signal, wherein, when this transmission peak wavelength range lambda ₁When meeting the following conditions expression formula C1, the expression formula C2 that will meet the following conditions of the layer γ material in the stacked member of exploring block side setting:

380nm≤λ ₁≤780nm …C1

0nm＜dγ≤96nm …C2

24. come to obtain the device of physical message according to a kind of device of the employing of claim 23 based on unit signal, wherein, the layer γ material expression formula C3 that meets the following conditions:

47nm＜dγ≤96nm …C3

25. come to obtain the device of physical message according to a kind of device of the employing of claim 10 based on unit signal, wherein, this stacked member comprises the two layers of material at least that is selected from the following material: the oxide such as silicon nitride, silica, aluminium oxide, zirconia, titanium oxide, magnesium oxide and zinc oxide; Polymeric material such as polycarbonate and acrylic resin; And the semi-conducting material such as carborundum and germanium Ge.

26. come to obtain based on unit signal the device of physical message according to a kind of device of the employing of claim 25, wherein, this stacked member comprises as the silicon nitride of ground floor material with as the silica of second layer material.

27. come to obtain the device of physical message according to a kind of device of the employing of claim 11 based on unit signal, wherein, this stacked member comprises: the wiring layer that is provided with on this exploring block side, adopt this wiring layer to be used to form the single line that reads unit signal from this unit signal production part; And stack membrane, this stack membrane has a kind of structure and also has the characteristic of reflection this electromagnetic predetermined wavelength component and this residual term of transmission, and stacked in this structure have a plurality of layers that different refraction coefficients and each layer all have predetermined thickness between adjacent layer.

28. come to obtain the device of physical message according to a kind of device of the employing of claim 27 based on unit signal, wherein, this stack membrane has and comprises as the silicon nitride of ground floor material with as the structure of the silica of second layer material, on two outsides, the ground floor material is set, and two layers of material replaces and stackedly is 9 layers or more multi-layered altogether, and the distance between this stack membrane and this exploring block is 700nm.

29. come to obtain the device of physical message according to a kind of device of the employing of claim 27 based on unit signal, wherein, this stack membrane has a kind of comprising as the silicon nitride of ground floor material with as the structure of the silica of second layer material, on two outsides, the ground floor material is set, and two layers of material replaces and stackedly is 9 layers or more multi-layered altogether, and the distance between this stack membrane and this exploring block is 3.2nm.

30. come to obtain the device of physical message according to a kind of device of the employing of claim 12, further comprise the driver part of this detection time that is used to control another exploring block based on unit signal.

31. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, wherein, the unit signal of for several times integrated reflected wavelength range component of surveying by another exploring block of this Signal Processing Element, thus this integrated unit signal of transmission peak wavelength ranges of components used to obtain to be used for second physical message of being scheduled to purpose.

32. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, wherein, come periodically to be provided for surveying the exploring block and the exploring block that is used to survey the reflected wavelength range component of transmission peak wavelength ranges of components according to the mode of constant digital ratio.

33. come to obtain the device of physical message according to a kind of device of the employing of claim 32, wherein, be provided for surveying an exploring block of reflected wavelength range component with respect to a plurality of exploring blocks that are used to survey the transmission peak wavelength ranges of components based on unit signal.

34. come to obtain the device of physical message according to a kind of device of the employing of claim 12, wherein, be provided for surveying the exploring block and the exploring block that is used to survey the reflected wavelength range component of transmission peak wavelength ranges of components according to 1: 1 mode based on unit signal.

35. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, wherein, the exploring block that is used to survey the transmission peak wavelength ranges of components also comprises: be used for that the transmission peak wavelength ranges of components is separated into the wavelength component and survey the combination of a plurality of detecting elements of this component, arrange the exploring block that is used to survey the reflected wavelength range component and be used to survey a plurality of detecting elements of the exploring block of transmission peak wavelength ranges of components according to the two-dimensional crystal lattice mode, so that arrange the detecting element that is used to survey the predetermined wavelength component in a plurality of detecting elements according to the alternating graph mode.

36. come to obtain the device of physical message according to a kind of device of the employing of claim 12 based on unit signal, wherein, the exploring block that is used to survey the transmission peak wavelength ranges of components also comprises: be used for that the transmission peak wavelength ranges of components is separated into the wavelength component and survey the combination of a plurality of detecting elements of this component, arrange the exploring block that is used to survey the reflected wavelength range component and be used to survey a plurality of detecting elements of the exploring block of transmission peak wavelength ranges of components according to the two-dimensional crystal lattice mode, so that arrange the exploring block that is used to survey the reflected wavelength range component according to the alternating graph mode.

37. a manufacture method that is used to survey the semiconductor device that physical quantity distributes, this semiconductor device comprises: as unit block, be used to survey electromagnetic exploring block; And be used for producing the also unit signal production part of output unit signal according to the electromagnetic wave amount of surveying by this exploring block, this unit block is set on the identical Semiconductor substrate according to predefined procedure, and this method may further comprise the steps:

On this Semiconductor substrate, form semiconductor element layer with exploring block and unit signal production part;

On this semiconductor element layer, be formed for the wiring layer of holding wire, be used for reading unit signal from this unit signal production part; And

On this wiring layer, form stacked film, this stacked film has a kind of structure and has the predetermined wavelength component of reflection electromagnetic wave and the characteristic of transmission residual term, and stacked in this structure have different refraction coefficients and every layer of a plurality of layer with predetermined thickness between adjacent layer.

38. the manufacture method that is used to survey the semiconductor device that physical quantity distributes according to claim 37 further comprises step: according to removing the part of this stacked film with mode corresponding to the exploring block position alignment of each wavelength.

39. the manufacture method that is used to survey the semiconductor device that physical quantity distributes according to claim 37, further comprise step: according to mode corresponding to the exploring block position alignment of each wavelength, form optical component on this stacked film, this optical component is used for the predetermined wavelength of transmission transmission peak wavelength ranges of components.

Images