JP5369362B2 - Physical information acquisition apparatus and manufacturing method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem of crystallinity in an imaging apparatus utilizing a band gap. <P>SOLUTION: The band gap is controlled by changing the composition ratio of a mixed crystal system of a compound semiconductor. For example, the absolute value of lattice mismatching &Delta;a is made small by using an AlGaInP-based mixed crystal, an SiGeC-based mixed crystal, a ZnCdSe-based mixed crystal, and an AlGaInN-based mixed crystal. For example, Ge having a lattice constant larger than that of Si is mixed with SiC, thus reducing the absolute value of the lattice mismatching, and increasing crystallinity. At least one superlattice layer having a thickness of approximately 10 nm or smaller is formed at the interface of Si and SiC, or Si and SiGeC-based layer, thus increasing crystallinity further. <P>COPYRIGHT: (C)2006,JPO&amp;NCIPI

Description

  The present invention relates to a physical information acquisition apparatus. More specifically, for example, physical quantity distribution detection that can be read as an electric signal from a physical quantity distribution converted into an electric signal by a detection region using a band gap that is sensitive to electromagnetic waves input from outside such as light and radiation. The present invention relates to a signal acquisition technique suitable for application to a solid-state imaging device using such a semiconductor device.

  There are various physical quantity distribution detection semiconductor devices in which a plurality of unit components (for example, pixels) that are sensitive to changes in physical quantity such as light and radiation input from the outside such as electromagnetic waves are arranged in a line or matrix form. Used in the field.

  For example, in the field of video equipment, a CCD (Charge Coupled Device) type, a MOS (Metal Oxide Semiconductor) type, or a CMOS (Complementary Metal-oxide Semiconductor) type solid that detects changes in light (an example of an electromagnetic wave) that is an example of a physical quantity. An imaging device is used. These read out, as an electrical signal, a physical quantity distribution converted into an electrical signal by a unit component (a pixel in a solid-state imaging device).

  For example, in a solid-state imaging device, a photodiode that is a photoelectric conversion element (light receiving element; photosensor) provided in an imaging unit (pixel unit) of a device unit receives electromagnetic waves input from outside such as light and radiation. The signal charges are detected and generated and accumulated, and the accumulated signal charges (photoelectrons) are read out as image information.

  Here, a conventional solid-state imaging device for single-plate color image imaging is for performing color (identification) selection on the light-receiving surface side of the imaging unit in order to obtain a signal having color information with one imaging device. A filter provided with a color filter (color separation filter) has become the mainstream.

  In an image sensor that identifies a color by using a subtractive color filter, by identifying a color by using a subtractive color-specific color filter, and providing a semiconductor layer for a photoelectric conversion element that detects light under each filter, The light transmitted through the filter is detected separately. In this case, it is necessary to assign individual color components to one photoelectric conversion element forming one pixel, and one unit of color separation is equivalent to the repetition period of the color separation filter.

  For example, as a combination of color filters, primary color systems using three types of red (R), green (G), and blue (B), yellow (Y), cyan (C), magenta (M), and green There is a complementary color system using four types of (G). The primary color system has better color reproducibility than the complementary color system, and the complementary color system is advantageous in terms of sensitivity because the light transmittance of the color filter is high. At the time of video reproduction, signal processing is performed on a color signal (for example, R, G, B primary color signals) obtained using a primary color or complementary color filter, and a luminance signal and a color difference signal are synthesized. .

  By the way, any combination of primary color or complementary color filters performs color selection by transmitting only a specific wavelength region component and guiding it to the photoelectric conversion element while blocking (cutting) other wavelength region components. It is a subtractive color filter.

  For example, by using three subtractive color filters of red, green, and blue, which are the three primary colors, color is identified, and a semiconductor layer that detects light under each filter is provided as a photoelectric conversion element. The three primary color lights transmitted through the subtractive color filter are individually detected.

  However, in the case of the subtractive color filter system, the amount of light to be cut is large, so the light use efficiency is poor. In particular, when the respective colors are identified using the three primary color filters of red, green, and blue, the amount of light falls to 1/3 or less by itself.

  In addition, since a photoelectric conversion element is required for each color, at least three photoelectric conversion elements are required for one unit of color separation, which is an obstacle to realizing a sensor with a high density of pixels. In addition, the cost is increased due to the need for a color separation filter.

  Recently, a sensor for identifying a color by utilizing the fact that the absorption coefficient of a semiconductor differs depending on the wavelength of light has been proposed (see, for example, Patent Document 1).

US Pat. No. 5,965,875

  20A and 20B are diagrams for explaining the mechanism of the sensor described in Patent Document 1, in which FIG. 20A shows a light absorption spectrum characteristic of a semiconductor layer, and FIG. 20B shows a cross-sectional structure of the device. It is a schematic diagram.

  In this mechanism, the absorption coefficient of the three primary colors of the Si (silicon) semiconductor decreases in the order of blue, green, and red as shown in FIG. 20A, that is, blue light, green light included in the incident light L1, As shown in FIG. 20 (B), the red, blue, and red light is emitted in the depth direction from the surface of the Si semiconductor by utilizing the location dependence due to the wavelength in the depth direction. Are sequentially provided.

  However, the mechanism described in Patent Document 1 using the difference in absorption coefficient depending on the wavelength does not decrease the amount of light that can be detected theoretically, but the layer that detects blue light absorbs to some extent when red light or green light passes through. In order to receive, those lights will be detected as blue light. For this reason, even if a blue signal is not originally present, if a green or red signal is input, a blue signal is also generated and a false signal is generated, so that sufficient color reproducibility cannot be obtained.

  In order to avoid this, it is necessary to perform correction by signal processing by calculation for all three primary colors, and a circuit necessary for the calculation is separately required. Therefore, the circuit configuration is complicated and large in size, and the cost is increased. Become expensive. Further, for example, when one of the three primary colors is saturated, the original value of the saturated light is not known, resulting in a calculation error. As a result, the signal is processed differently from the original color. .

  As shown in FIG. 20A, most semiconductors have absorption sensitivity to infrared light. Therefore, for example, in a solid-state imaging device (image sensor) using a Si semiconductor, for example, it is usually necessary to put an infrared cut filter as an example of a color reduction filter in front of the sensor.

  In response to the problem of the mechanism using the difference in absorption coefficient depending on the wavelength, the band gap is used without using the subtractive filter, so that the light quantity conversion is efficient and the color separation is good. There has been proposed a sensor that can detect the light of each of the three primary colors with one sensor (see, for example, Patent Documents 2 to 4). What is disclosed in these is an image sensor having a structure in which the band gap is changed in the depth direction.

JP-A-1-151262 JP-A-3-289523 JP-A-6-209107

  However, in the mechanism described in Patent Document 2, although materials with different band gaps Eg are sequentially laminated in the depth direction of the semiconductor layer on the glass substrate, for example, blue (B), green In the color separation of (G) and red (R), it is merely described that the layers are laminated so that Eg (B)> Eg (G)> Eg (R), and the description of specific materials is as follows. Absent.

  In contrast, Patent Document 3 describes color separation using a SiC material, and Patent Document 4 describes AlGaInAs and AlGaAs materials. However, Patent Documents 3 and 4 do not describe crystallinity at heterojunctions of different materials.

  When materials having different crystal structures are joined, misfit dislocations occur due to the difference in lattice constants, and crystallinity deteriorates. As a result, electrons trapped in the defect level formed in the band gap are discharged, thereby causing dark current.

  The present invention has been made in view of the above circumstances, and provides a mechanism that can solve the problem of crystallinity when a sensor is configured without using a subtractive color filter by using a band gap. Objective.

  In the physical information acquisition apparatus (typically, an imaging apparatus) according to the present invention, the detection region is configured by a lattice matching system or configured to meet the lattice matching conditions by adding a predetermined element. It was decided.

  Further, the invention described in the dependent claims defines a further advantageous specific example of the physical information acquisition apparatus according to the present invention.

According to the present invention, a plurality of detection regions stacked in a depth direction from a light incident surface facing the substrate are provided on a predetermined substrate, and electromagnetic waves are separated by wavelength in each detection region and detected. A physical information acquisition device for detecting a physical quantity distribution that outputs a unit signal corresponding to each wavelength of the electromagnetic wave, wherein the plurality of detection regions output a plurality of lights in a depth direction from the light incident surface. Each of the plurality of detection regions is formed as a plurality of detection regions having different band gaps depending on the wavelength of the light so as to be selectively absorbable, and the band gaps of the plurality of detection regions are separated from the incident surface of the light incident surface. is narrower, the respective detection region, by controlling the bandgap of the respective detection regions by changing the composition ratio of the ternary or higher mixed crystal containing Si, adapted to the substrate and the lattice matching condition I will do it It is configured to have a quantum well structure, characterized in that, the physical information obtaining device is provided.
Moreover, according to this invention, the method of manufacturing the said physical information acquisition apparatus is provided.

  As a mechanism for manufacturing the physical information acquisition device as described above, for example, a plurality of detection regions for detecting electromagnetic waves by wavelength separation using a band gap energy are provided on a predetermined substrate, and the detection regions are lattice-matched. It is preferable to configure the system so as to meet the lattice matching condition by adding a predetermined element.

  Here, the plurality of detection regions may be formed by repeating ion implantation, may be formed by repeating thermal diffusion, or may be formed by performing crystal growth by a plurality of CVD methods. Alternatively, it may be formed by performing a plurality of MBE methods or MOCVD methods.

  According to the present invention, since the detection region is configured by a lattice matching system or by adding a predetermined element so as to meet the lattice matching condition, the band gap is changed in the depth direction. In the case of providing the above structure, the crystallinity problem between the detection regions can be solved.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

<Basic configuration>
FIG. 1 is an example of a physical information acquisition apparatus according to the present invention, and is a diagram for explaining the concept of an imaging apparatus (image sensor) that performs color separation by controlling a band gap.

  As shown in the drawing, materials having different band gaps Eg are sequentially laminated on a glass substrate to separate colors. In this example, blue light (B), green light (G), and red light (R) are stacked so that Eg (B)> Eg (G)> Eg (R) so as to be detected separately. That is, the respective band gaps of the detection layers are sequentially reduced. That is, the band gap Eg is such that the blue light detection layer can absorb only blue light and transmit green light and red light. This point is the same as the mechanism described in Patent Document 2.

  In the blue light detection layer, the band gap Eg (B) of the layer is, for example, Eg (B) = 2.38 to 2.53 eV, and preferably Eg (B) = 2.45 to 2.49 eV. Like that. By doing so, the blue light detection layer detects only blue light and does not detect green light or red light.

  In order to prevent red light from being detected in the green light detection layer, the band gap Eg (G) of the layer is, for example, Eg (G) = 1.98-2.20 eV, and preferably Eg (G) = 2. 0.07 to 2.12 eV. In this way, blue light and green light can be detected.

  Note that if the blue light detection layer is thick enough to absorb blue light, the blue light will not reach the green light layer, so the green light detection layer will only detect green light. It will be possible. For example, when the absorption coefficient of blue light is αB (cm ^ -1; "^" indicates power; the same applies hereinafter), the intensity of blue light is 1 / e or less at the interface with the green light detection layer. The blue light detection layer is made thicker than 1 / αB × 10 ^ −4 (μm). More desirably, the thickness is increased to 2 / αB × 10 ^ −4 (μm) or more so that the strength is 1 / e2 or less.

  Finally, in the red detection layer, the band gap Eg (R) of the layer is set to, for example, Eg (R) = 2.20 eV or less so that red light can be detected. Desirably, Eg (R) = 1.90 eV or less. In this way, red light can be detected.

  Here, when the absorption coefficient of the green light detection layer is αG (cm ^ −1) so that the green light does not enter the red light detection layer, the green light has a green light intensity of 1 / e or less. The thickness of the detection layer is increased to 1 / αG × 10 ^ −4 (μm) or more. Desirably, the thickness is increased to 2 / αB × 10 ^ −4 (μm) or more so that the strength becomes 1 / e2 or less.

<First Method for Obtaining Band Gap; First Example>
FIG. 2 is a diagram for explaining an example of a first method for obtaining a band gap (referred to as a first example of the first method).

  As a method for obtaining the band gap as described above, first, as a first method, a first method for controlling the band gap by changing the composition ratio of a mixed crystal system which is a compound semiconductor can be employed. For example, an AlGaInP mixed crystal, a SiGeC mixed crystal, a ZnCdSe mixed crystal, or an AlGaInN mixed crystal can be used.

  In particular, the first example of the first method shown in FIG. 2 is an aspect in which the same group VI element Ge is put into the SiC system, that is, Ge is put into a SiGeC mixed crystal. As shown in FIG. 2, the SiC group is the same group VI element and has good compatibility with a Si substrate that is generally used. Therefore, the band gap can be efficiently controlled by the composition ratio of Si and C.

  Assuming that the Vegard law is established, if the composition of SiXC1-X is X = 0.68 in the blue light detection layer, for example, Eg (B) = 2.48 eV, and in the green light detection layer, X = If 0.76, Eg (G) = 2.1 eV, which satisfies the above range. Here, the Si detection substrate may be used as it is for the red detection layer.

  Thus, color separation can be performed by controlling the band gap. Since no subtractive filter is used, the conversion efficiency of the optical-electrical signal is increased, and high sensitivity characteristics can be obtained. In addition, since no sensor is required for each color, only one sensor is required for each pixel. Therefore, an image sensor having a high density of pixels can be realized.

<First Method for Obtaining Bandgap; Second Example>
FIG. 3 is a diagram showing a quantum well structure for explaining another example of the first method for changing the composition ratio of the mixed crystal system (referred to as a second example of the first method). The second example of the first method is characterized in that the absolute value of the lattice mismatch Δa is reduced by introducing Ge into a SiGeC mixed crystal.

  That is, when the Si substrate is used, when the band gap is controlled only by the SiC system, the lattice constant of C is smaller than that of Si, as shown in FIG. For this reason, misfit dislocations may be introduced at the Si / SiC interface and the crystallinity may deteriorate. Therefore, by mixing Ge, which is larger than the lattice constant of Si, with SiC, the absolute value of lattice irregularity is reduced and the crystallinity is increased.

The lattice irregularity described here is defined by the following equation (1). Here, a _Si is the lattice constant of Si, and a _SiC is the lattice constant of SiC mixed crystal.

<First Method for Obtaining Band Gap; Other Modifications>
Each of the first methods described above is a method of controlling the band gap by changing the composition ratio of a mixed crystal system which is a compound semiconductor. In particular, the crystallinity is obtained by adding Ge to form a SiGeC based mixed crystal. To increase.

  However, the first example in which the band gap is controlled by changing the composition ratio of the mixed crystal system is not limited to this. For example, a predetermined thickness (for example, a thickness of about 10 nm) is formed at the interface between Si and SiC or SiGeC system layer. One or more superlattice layers of the following thin films may be added.

  In this case, the superlattice thin film only needs to have a lattice constant different from that of Si. That is, SiGeC-based layers having different composition ratios may be stacked in multiple layers.

  By doing so, the strain in the added superlattice layer is relieved, and dislocations are released in the lateral direction, which can increase the crystallinity.

  As described above, in the first method for obtaining the band gap, the generation of crystal defects in the superlattice layer can be suppressed. Thereby, since generation | occurrence | production of the electron trapped by a superlattice layer can be suppressed, dark current can be suppressed. At the same time, light scattering due to crystal defects can be suppressed, and as a result, a high S / N ratio can be obtained.

<Second Method of Obtaining Bandgap; First Example>
4 and 5 are diagrams for explaining an example of a second method for obtaining a band gap (referred to as a first example of the second method). Here, FIG. 4 is a figure which shows the calculation result at the time of producing the quantum well of Si in SiGeC. FIG. 5 is a diagram showing an example of a quantum well stacked structure.

  This second method is characterized in that the band gap is controlled by using the quantum size effect. In order to use the quantum size effect in the second method, it is preferable to employ a method of forming a quantum well structure in which a mixed crystal of three or more elements and a quantum well layer are combined as one method.

  For example, the first example of the second method is characterized in that, as shown in FIG. 3, a Si / SiGeC / Si substrate structure is formed by forming a Si quantum well in SiGeC. Yes. Here, the composition ratio of Si, Ge, and C of SiGeC is 1: 0.2: 1.

In FIG. 4, the thickness L of the Si layer, which is a well, is plotted on the horizontal axis, and as shown in FIG. 3, the energy increment ΔE _Cn from the bottom of the conduction band to the quantum level (where the ground level is n = 1) Similarly, the total energy increment ΔE ₁ _total obtained by adding the energy increment ΔE _Vn (where the ground level n = 1) from the upper end of the valence band to the quantum level is plotted on the vertical axis. .

In this case, the calculation follows the following equation (2-1). In addition, m ^* in Formula (2-1) represents the effective mass of carriers, and the effective masses m _e and m _hh of electrons and heavy holes are expressed by Formula (2-2) and Formula (2-3), respectively. It was. M _{0 in} Formula (2-2) and Formula (2-3) represents the mass of free electrons.

  From this result, it can be seen that the energy increment can be increased by setting the thickness L of the Si layer to 4 nm or less. In particular, when the Si layer has a superlattice structure with a thickness of 0.95 nm, Eg = 2.4 to 2.5 eV, which corresponds to blue detection. Furthermore, by making the thickness of the Si layer a superlattice structure of 1.10 nm, Eg = 2.0 to 2.1 eV, which corresponds to green color detection.

  Note that the superlattice layer may be multilayered to form a quantum well stacked structure. For example, as shown in FIG. 5, a Si substrate 11 with Eg = 1.1 eV is used as a red detection region 12, and a SiGeC layer (the hatched portion in the figure) and a Si layer (hatched line in the figure) are formed on the Si substrate 11. The hatched portions are alternately and repeatedly stacked to form a 1.10 nm thick well layer (Si layer) in the SiGeC layer. A region of Eg = 2.0 to 2.1 eV having a well layer (Si layer) with a thickness of 1.10 nm is defined as a green detection region 14.

  Furthermore, a SiGeC layer (a hatched portion in the figure) and a Si layer (a hatched portion in the figure) are alternately and repeatedly laminated on the green detection region 14 to form a 0.95 nm thick well layer (Si Layer) in the SiGeC layer. A region of Eg = 2.4 to 2.5 eV having a well layer (Si layer) having a thickness of 0.95 nm is defined as a blue detection region 16.

  By adopting such a structure, the band gap Eg can be changed in the depth direction, and regions with different band gaps Eg can be created, thereby enabling efficient color separation.

  Of course, also in the first example of the second method for obtaining the band gap, the occurrence of crystal defects in the superlattice layer can be suppressed. Accordingly, generation of electrons trapped in the superlattice layer can be suppressed, so that dark current can be suppressed. At the same time, light scattering due to crystal defects can also be suppressed. As a result, high S / N ratio can be obtained.

<Signal retrieval method>
6 and 7 are diagrams for explaining a signal extraction method. In FIG. 6 and FIG. 7, it has shown by the example applied to a COMS structure. 6 and 7 show one photodiode group.

  As in the first example illustrated in FIG. 6, the solid-state imaging device 20 having a COMS structure includes light absorption regions (signal charge generation units; detection layers 22, 24, and 26; each of R, G, and B primary colors). The PN junction is arranged corresponding to the photodiode group), and the PN junction is repeatedly present in the vertical direction with respect to the semiconductor layer.

  More specifically, in the P-type Si substrate 21, for each photodiode group, each corresponding position in the depth direction (Z direction in the figure) is doped with n-type impurities to be independent. A red detection layer (red detection region) 22 for detecting red light is provided by forming the n-type Si region 22N.

  Further, in the n-type Si region 22N corresponding to each photodiode group, a p-type impurity is first doped with a p-type impurity at a detection position in the depth direction (Z direction in the figure) corresponding to green light. A green detection is performed in which a region 24P is formed and an n-type Si region 24N is formed at a detection position in the depth direction (Z direction in the drawing) corresponding to green light in the p-type Si region 24P to detect green light. A layer (green color detection region) 24 is provided.

  Further, in the n-type Si region 24N corresponding to each photodiode group, a p-type impurity is first doped with a p-type impurity at a detection position in the depth direction (Z direction in the drawing) corresponding to blue light. 26P is formed, and n-type Si region 26N is formed by doping an n-type impurity at a detection position in the depth direction (Z direction in the figure) corresponding to blue light in p-type Si region 26P. A blue detection layer (blue detection region) 26 for detecting blue light is provided.

  If necessary, the p-type Si region 26P + is formed by doping the surface of the n-type Si region 26N with a high-concentration p-type impurity in order to reduce noise by reducing dark current. That is, a so-called HAD (Hole Accumulated Diode) structure in which a hole accumulation layer made of a P + type impurity region is further stacked on a charge accumulation layer on the surface side of the NP diode made of an N + type impurity region (for example, (See JP-A-5-335548 and JP-A-2003-78125).

  As a result, as shown in the figure, the n-type Si region 22N, the p-type Si region 24P, the n-type Si region 24N, the p-type Si region 26P, and the n-type Si region 26N are sequentially laminated in a curved shape to form a Baumkuchen-like structure. Like to have.

  In the solid-state imaging device 20 having such a structure, each photodiode group (the red detection layer 22, the green detection layer 24, and the blue detection layer 26) includes a P-type layer (p-type Si region) of each pixel and the photodiode group. The light of a predetermined color is detected by the electric charge which generate | occur | produces between the layers of the photoelectric conversion element formed with a corresponding N type layer (n type Si area | region).

  By doing so, in the solid-state imaging device 20, the red detection layer 22, the green detection layer 24, and the blue detection layer 26 that form photodiodes (photoelectric conversion elements) are provided independently for each pixel. Since the detection layers 22, 24, and 26 are arranged in the depth direction of the semiconductor layer, it is practically unnecessary to provide pixels for each color, and one pixel can correspond to detection of three colors. By doing so, the pixel size can be effectively tripled compared to the conventional structure in which pixels of R, G, and B colors are arranged in the plane direction. The amount of light incident per unit area can be increased, the photoelectric-electrical signal conversion efficiency can be increased, and high sensitivity characteristics can be obtained.

  In the solid-state imaging device 20 having such a structure, signal charges are accumulated in N-type layers (n-type Si regions 22N, 24N, and 26N; hatched portions in the drawing) corresponding to blue, green, and red, respectively. Therefore, in order to convert this signal charge into an electric signal, the charge-voltage signal conversion unit 105 converts the signal charge into an electric signal and amplifies and reads it. The configuration of the charge-voltage signal conversion unit 105 may be the same as that of an intra-pixel amplifier using floating diffusion in a known CMOS sensor.

  For example, as shown in FIG. 6, the charge-voltage signal conversion unit 105 has a charge generation unit that has both a photoelectric conversion function of receiving light and converting it into a charge, and a charge storage function of storing the charge ( An amplifying transistor having a source follower configuration that is an example of a detection element that detects a potential change of the reset transistor 136 that is an example of a reset gate unit, a readout line selection transistor 140, and a floating diffusion 138 with respect to the photodiode group 132) 142 for each of R, G, and B. In other words, there is no transfer gate, and the photoelectric conversion elements 132B, 132G, and 132R corresponding to the color separation detection layers 22, 24, and 26 are directly and electrically connected to the charge-voltage signal conversion unit 105. Is adopted.

  The charge-voltage signal conversion unit 105 includes a pixel signal generation unit having an FDA (Floating Diffusion Amp) configuration having a floating diffusion 138 which is an example of a charge injection unit having a function of a charge storage unit. The floating diffusion 138 is a diffusion layer having parasitic capacitance.

  The reset transistor 136 is driven by a reset driving buffer (not shown) via a reset wiring (RST) 156. The read line selection transistor 140 is driven by a selection drive buffer (not shown) via a read selection line (SEL) 152. Each drive buffer can be driven independently.

  In the reset transistor 136 in the charge-voltage signal converter 105, the source is connected to the gate of the floating diffusion 138 and the amplifying transistor 142, the drain is connected to the power supply VN, and the reset pulse RST is reset to the gate (reset gate RG). Input from buffer. The reset transistor 136 has a function of resetting the potential of the floating diffusion 138.

  For example, in the read line selection transistor 140, the drain is connected to the source of the amplifying transistor 142, the source is connected to the pixel line 151, and the gate (particularly, the read selection gate SELV) is connected to the read selection line 152. . The read line selection transistor 140 is not limited to such a connection configuration, and the drain is connected to the power supply VCC, the source is connected to the drain of the amplification transistor 142, and the gate is connected to the read selection line 152. Good.

  A read selection signal SEL (Row) is applied to the read selection line 152. The amplification transistor 142 has a gate connected to the floating diffusion 138, a drain connected to the power supply VCC, a source connected to the pixel line 151 via the drain of the read line selection transistor 140, and further connected to the read signal line 153. It is like that.

  In the charge-voltage signal converter 105 having such a configuration, the floating diffusion 138 is connected to the gate of the amplifying transistor 142, so that the amplifying transistor 142 corresponds to the potential of the floating diffusion 138 (hereinafter referred to as FD potential). The signal is output to the readout signal line 153 through the pixel line 151 in the voltage mode.

  The reset transistor 136 resets the floating diffusion 138. When a read selection transistor (transfer transistor) is provided, the signal charge generated by the charge generation unit 132 is transferred to the floating diffusion 138.

  Many pixels are connected to the readout signal line 153. To select a pixel, the readout line selection transistor 140 is turned on only in the selected pixel. Then, only the selected pixel is connected to the readout signal line 153, and the signal of the selected pixel is output to the readout signal line 153.

  Here, in the circuit configuration shown in FIG. 6, the charge-voltage signal conversion unit 105 is individually provided for each of R, G, and B. For this reason, although the circuit scale is increased, the R, G, and B color signals at the same pixel position can be acquired independently and simultaneously.

  On the other hand, in the circuit configuration of the second example shown in FIG. 7A, one charge-voltage signal converter 105 is provided for a group of R, G, and B. For this reason, it is necessary to acquire R, G, and B color signals by time division.

  Therefore, in comparison with the circuit configuration of the first example shown in FIG. 6, the connection between the photoelectric conversion elements 132B, 132G, and 132R for each color forming the charge generation unit (photodiode group) 132 and the floating diffusion 138. The configuration is different.

  Specifically, a read selection transistor (transfer transistor) which is an example of a charge read unit (transfer gate unit / read gate unit) is provided between the charge generation unit 132 and the gate of the amplification transistor 142. In this case, the read selection transistor is driven by the transfer drive buffer via the transfer wiring (read selection line TX). In other words, each of the photoelectric conversion elements 132B, 132G, and 132R has a structure in which transfer gates 134B, 134G, and 134R are provided for each color. Each transfer gate 134B, 134G, 134R is controlled by a control unit (not shown). In FIG. 7A, the read line selection transistor 140 is omitted.

  In the structure of such a configuration example, the photoelectric conversion elements 132B, 132G, and 132R corresponding to the color separation detection layers 22, 24, and 26 are for amplification through the corresponding transfer gates 134B, 134G, and 134R. The pixel signal is connected to the vertical signal line 153 in accordance with each timing shown in FIG. 7B, which is connected to the charge-voltage signal conversion unit 105 including the transistor 140, the reset transistor 136, and the like, and shows the reset state and the signal read state. Output. By resetting the floating diffusion 138 before reading each signal charge, the pixel signals R, G, and B can be read independently in the order of R → G → B (the order may be changed). Although it is impossible to simultaneously acquire R, G, and B color signals at the same pixel position, the circuit scale can be reduced.

  In addition, the circuit example which takes out a signal from each photoelectric conversion element 132B, 132G, 132R is not restricted to the system shown here, but various signal which reads the signal charge obtained by each photoelectric conversion element 132B, 132G, 132R by which color separation was carried out The known circuit method can be applied as appropriate.

<Second Method of Obtaining Bandgap; Second Example>
8 to 11 are diagrams illustrating an example of a second method for obtaining a band gap (referred to as a second example of the second method). 8 corresponds to FIG. 3, and FIG. 9 corresponds to FIG. 4. FIG. 9 corresponds to FIG. 4, and CsSe quantum well in ZnSSe (composition ratio of S and Se is x: 1-x). It is a figure which shows the calculation result at the time of producing. FIG. 10 corresponds to FIG. 5 and is a diagram showing an example of a quantum well stacked structure, and FIG. 11 is a diagram showing respective lattice constants.

  The second example of the second method has a feature in that the band gap is controlled by using the quantum size effect to completely lattice match with the substrate. In order to use the quantum size effect, as one of the methods, a method using a quantum well structure may be employed.

  For example, the second example of the second method is characterized in that a structure of a ZnCdSe / ZnSSe / GaAs substrate is formed by forming a CdSe quantum well in ZnSSe as shown in FIG. . In this example, the S composition x of the ZnSSe layer is 0.06, and the Se composition “1-x” is 0.94.

In FIG. 9, the horizontal axis is the thickness L of the CdSe layer, which is a well, and the energy increment ΔE _Cn from the lower end of the conduction band to the quantum level as shown in FIG. The total energy increment ΔE ₁ _total with the energy increment ΔE _Vn (where the ground level n = 1) from the top of the valence band to the quantum level is plotted on the vertical axis. ing. Here, the band gap of the CdSe layer is 1.15 eV.

In this case, the calculation follows the following equation (3-1). Note that m ^* in the formula (3-1) represents the effective mass of the carriers, and the effective masses m _e and m _hh of the electrons and heavy holes are the formulas (3-2) and ( 3-3). M _{0 in} formula (3-2) and formula (3-3) represents the mass of free electrons.

  From this result, it can be seen that the energy increment can be increased by setting the thickness L of the CdSe layer to 8 nm or less. In particular, when the CdSe layer has a superlattice structure with a thickness in the range of ± 0.1 nm centering on a thickness of 2.3 nm, Eg = 2.4 to 2.5 eV, which corresponds to blue detection. Furthermore, by setting the thickness of the CdSe layer to a superlattice structure with a thickness in the range of ± 0.3 nm centered on 3.5 nm, Eg = 2.0 to 2.1 eV, which corresponds to green color detection.

  Further, by setting the thickness of the CdSe layer to a superlattice structure of 4.0 nm or more, Eg ≦ 2.0 eV, which corresponds to red detection.

  Also in the second example of the second method, similarly to the first example of the second method, a multilayer structure of quantum wells may be formed by forming a superlattice layer in multiple layers. For example, as shown in FIG. 10, a ZnSSe layer (the hatched portion in the figure) and a CdSe layer (the hatched portion in the figure) are alternately stacked on the GaAs substrate 31 (the lattice hatched part in the figure). Thus, a well layer (CdSe layer) having a thickness of 6.0 nm is formed in a multilayer in the ZnSSe layer. A region of Eg ≦ 2.0 eV having a well layer (CdSe layer) having a thickness of 6.0 nm is defined as a red detection region 32.

  Further, a ZnSSe layer (a hatched portion in the drawing) and a CdSe layer (a hatched portion in the drawing) are alternately and repeatedly stacked on the red detection region 32, thereby forming a well layer (CdSe having a thickness of 3.5 nm). Layer) in the ZnSSe layer. A region of Eg = 2.0 to 2.1 eV having a well layer (CdSe layer) having a thickness of 3.5 nm is defined as a green color detection region 34.

  Further, a ZnSSe layer (a hatched portion in the drawing) and a CdSe layer (a hatched portion in the drawing) are alternately and repeatedly laminated on the green detection region 34, thereby forming a 2.3 nm thick well layer (CdSe). Layer) in the ZnSSe layer. A region of Eg = 2.4 to 2.5 eV having a well layer (CdSe layer) having a thickness of 2.3 nm is defined as a blue detection region 36.

  By adopting such a structure, the band gap Eg can be changed in the depth direction, and regions with different band gaps Eg can be created, thereby enabling efficient color separation.

  By the way, the respective lattice constants are as shown in FIG. 11. From FIG. 11, the S composition x of the ZnSSe layer is set to 0.06, the Se composition “1-x” is set to 0.94, and the like. It can be seen from the Begard rule that the Se composition “1-x” is much larger than the S composition x of the ZnSSe layer, and lattice matching with the GaAs substrate 31 occurs when the S composition x is very small. . In this way, the occurrence of defects such as misfit dislocations can be suppressed by lattice matching with the GaAs substrate 31.

  Further, the lattice constant (a = 6.07A) of the CdSe layer is greatly different from the lattice constant (a = 5.654A) of the GaAs substrate 31, but in this example, the CdSe layer is changed to an extremely thin ultrathin layer (thin superlattice layer). By doing so, the occurrence of defects such as misfit dislocations can be suppressed.

  As described above, also in the second example of the second method for obtaining the band gap, the generation of crystal defects in the superlattice layer can be suppressed. Thereby, since generation | occurrence | production of the electron trapped by a superlattice layer can be suppressed, dark current can be suppressed. At the same time, light scattering due to crystal defects can be suppressed, and as a result, a high S / N ratio can be obtained.

  In the above description, the CdSe well layer has been described. However, a ZnCdSe well layer may be used. In this case, since the bulk band gap not considering the quantum effect changes depending on the composition ratio of the Zn composition x and the Cd composition “1-x”, the optimum well layer thickness considering the quantum effect changes. To do.

  The bulk band gap Eg that does not take the quantum effect into consideration can be expressed as the following formula (4), assuming that the Vegard law holds.

  Accordingly, the composition of the ZnCdSe well layer “Zn: Cd = 1: 1−x” may be constant as the three primary color spectroscopy, and the thickness of the well layer may be changed, or the thickness of the well layer may be changed. Only a Zn composition x may be changed to obtain a predetermined band gap, or both may be changed.

<About manufacturing method>
By the way, the crystal growth of the compound material as described above is performed, for example, by CVD (Chemical Vapor Deposition) or MOCVD (Metal Organic Chemical Vapor Deposition) for forming a thin film on a wafer. ), Ion implantation, MBE (Molecular Beam Epitaxy), laser ablation, sputtering, and other various methods. Hereinafter, a method for manufacturing an image sensor having a structure in which the band gap is changed in the depth direction will be described.

<Ion implantation method>
FIG. 12 is a diagram for explaining an application example of the ion implantation method. Here, an example of application to the case of manufacturing the solid-state imaging device 20 shown in FIG. When ion implantation is used, the structure shown in the solid-state imaging device 20 can be manufactured by repeatedly performing the ion implantation.

  For example, after growing a crystal corresponding to each color on a p-substrate (S10), ions of n-type dopant species are implanted with high acceleration energy (S12). Further, ions of p-type dopant species are implanted at a lower acceleration energy (S14), and ions of n-type dopant species are implanted at a lower acceleration energy (S16). Further, processing similar to S14 and S16 is performed. In this way, by changing the acceleration energy so as to decrease sequentially, the depth of ion implantation can be changed, and as a result, a structure like the solid-state imaging device 20 shown in FIG. 6 can be obtained.

<Thermal diffusion method>
FIG. 13 is a diagram for explaining an application example of the thermal diffusion method. Here, an example of application to a manufacturing process (particularly an impurity diffusion process) for manufacturing the solid-state imaging device 20 shown in FIG. 6 is shown. Even when thermal diffusion is used, the structure shown in the solid-state imaging device 20 can be manufactured by repeatedly performing thermal diffusion, as in the ion implantation method.

  When an n-type layer or a p-type layer is obtained by thermally diffusing impurities, the diffusion region may be manufactured through a process step of sequentially diffusing p-type and n-type impurities using the same mask. With such a process step, the process is simple and the cost can be reduced.

  For example, as shown in FIG. 13, a plurality (multilayer) of photodiode regions can be formed by sequentially thermally diffusing a p-type dopant and an n-type dopant using the same oxide film mask. Therefore, the process is simple and the cost can be reduced.

  In particular, in the case of a Baumkuchen-like layer structure like the solid-state imaging device 20 shown in FIG. 6, all the detection layers 22, 24 and 26 for the three primary color photoelectric conversion elements (photodiodes) are formed. Therefore, a single mask can be used in common.

<CVD method>
14-16 is a figure explaining the application example of CVD method. Here, FIG. 14 is a flowchart showing the processing procedure of the CVD method, FIG. 15 is a schematic diagram showing the layer structure of the solid-state image sensor to be manufactured, and FIG. 16 is the final layer structure of the solid-state image sensor. It is a schematic diagram which shows a form. Here, an application example to the solid-state imaging device 20 as shown in FIG. 6 having a Si / SiGeC quantum well is shown.

As shown in FIG. 14, a SiGeC layer (composition ratio 1: 0.2: 1) crystal is grown on an n-type Si substrate by CVD (S20). For example, first, after cleaning the surface by immersing the Si substrate in a mixed solution of NH ₄ OH, H ₂ O ₂ , H ₂ O (1: 1: 5) for 10 minutes (S21), HF (HF: H ₂ O = 1:50) The natural oxide film is removed by performing the process for 10 seconds (S22). Through such a process, the surface can be cleaned and the crystallinity of subsequent crystal growth can be improved.

Next, this Si substrate is prepared and placed on the substrate holder (S23). Further, monosilane SiH ₄ , C ₃ H _8, and GeH ₄ were mixed at 36 μmol / min, 59 μmol / min, and 5 μmol / min, respectively, under the conditions of a pressure of 1 × 10 ⁴ Pa, a substrate temperature of 1150 ° C., and H ₂ gas of 1 liter / min. The SiGeC crystal is grown on the Si substrate with a thickness of 10 nm by simultaneously supplying under the conditions (S24).

Thereafter, by supplying only SiH ₄ , the Si quantum well layer is grown by about 1.1 nm (S26).

  This process is repeated so that the total thickness becomes 3 μm, thereby producing a green color detection quantum well region (S28).

  Further, by the same procedure, the blue detection quantum is changed so that the total thickness becomes 2 μm by repeating the SiGeC layer (composition ratio 1: 0.2: 1) so that the thickness is 10 nm and the Si quantum well layer is 0.95 nm. A well region is formed (S34 to S38).

  As a result, a multilayer structure of quantum wells as shown in FIG. 15 can be manufactured. FIG. 15 is a simplified illustration of the quantum well stack structure shown in FIG.

  Thereafter, for example, by applying a thermal diffusion method, boron P and phosphorus P, which are p-type and n-type dopants, are alternately diffused at a predetermined depth (S40). Further, if necessary, it is preferable to form a p-type Si region 26P + by doping boron B on the last outermost surface so as to reduce the dark current (S42).

  As a result, as shown in FIG. 16, the green detection layer 24 and the blue detection layer 26 have n-type Si regions 24N, 26N and p-type Si regions 24P, in the Si / SiGeC multilayer quantum well structure. 26P can be formed. In contrast to that shown in FIG. 6, since the Si substrate 21 is changed from p-type to n-type, a p-type Si region 22 </ b> P is formed in the red detection layer 22.

  Although the CVD method is used here, for example, in the laser ablation method, it is also possible to grow a crystal using SiGeC as a target material.

<MBE method>
17 to 19 are diagrams illustrating application examples of the MBE method. Here, FIG. 17 is a flowchart showing the processing procedure of the MBE method, FIG. 18 is a schematic diagram showing the layer structure of the solid-state imaging device to be manufactured, and FIG. 19 shows the final layer structure of the solid-state imaging device. It is a schematic diagram which shows a form. Here, an example of application to the solid-state imaging device 20 as shown in FIG. 6 having a ZnCdSe / ZnSSe quantum well is shown.

As shown in FIG. 17, an MBE method is used on an n-type GaAs substrate 31 to form a Zn _0.94 S _0.06 Se layer and a Zn _0.1 Cd _0.9 Se layer with a thickness of 10 nm and 6.0 nm, respectively. Are alternately grown (S50R). The band gap Eg at this time corresponds to 1.99 eV. By repeating this, the red detection quantum well region (red detection region 32) is fabricated by setting the total thickness of the quantum well structure to 3 μm (S52R).

Further, a green detection quantum well region (green detection region 34) and a blue detection quantum well region (blue detection region 36) are sequentially manufactured by the same procedure. For example, using the MBE method, a Zn _0.94 S _0.06 Se layer and a Zn _0.1 Cd _0.9 Se layer are alternately grown using a shutter so as to have thicknesses of 10 nm and 3.2 nm, respectively (S50G). The band gap Eg at this time corresponds to 2.21 eV. This process is repeated so that the total thickness of the quantum well structure is 2 μm, thereby producing a green color detection quantum well region (S52G).

Similarly, using the MBE method, a Zn _0.94 S _0.06 Se layer and a Zn _0.1 Cd _0.9 Se layer are alternately grown using a shutter so as to have thicknesses of 10 nm and 2.4 nm, respectively (S50B). The band gap Eg at this time corresponds to 2.49 eV. The blue detection quantum well region is fabricated by repeating this so that the thickness of the entire quantum well structure becomes 1.5 μm (S52B).

  As a result, a stacked structure of quantum wells as shown in FIG. 18 can be manufactured. FIG. 18 is a simplified illustration of the quantum well stack structure shown in FIG.

  Thereafter, for example, by applying a thermal diffusion method, phosphorus P and chlorine CL, which are p-type and n-type dopants, are alternately diffused at a predetermined depth (S54). Further, if necessary, it is preferable to form a p-type Si region 36P + by doping phosphorus P on the last outermost surface so as to reduce the dark current (S56).

  By doing so, as shown in FIG. 19, the red detection layer 22, the green detection layer 24, and the blue detection layer 26 have n-type Si regions 22N, 24N, and 26N in the ZnCdSe / ZnSSe multilayer quantum well structure. P-type Si regions 22P, 24P, and 26P can be formed. In contrast to the structure shown in FIG. 16, the Si substrate 21 shown in FIG. 16 is changed to an n-type GaAs substrate 21. As a result, the GaAs substrate 21 is located below the red detection layer 22. It has a provided structure.

  Note that although the MBE method is used here, a similar structure can be produced even by using, for example, the MOCVD method.

  In any of the manufacturing methods, finally, as in the readout circuit shown in FIG. 6 and the like, a CMOS structure is manufactured by a normal process to form a solid-state imaging device.

  These solid-state imaging devices are image sensors having a vertical structure having R, G, and B color separation layers in the depth direction with respect to the semiconductor layer, and can separate the three primary colors without color filters.

  As described above, according to the present embodiment, the layer for color separation is formed by controlling the band gap. In particular, the element structure is defined in consideration of crystallinity. As a result, the occurrence of crystal defects in the superlattice layer can be suppressed, and various problems caused by the crystal defects, such as dark current and light scattering, can be suppressed. Be able to get.

  Of course, the effect of having an image sensor provided with a layer for color separation by controlling the band gap can be enjoyed. For example, wavelength separation (color separation) can be realized without using a subtractive color filter. As a result, the detection layer corresponding to each wavelength can be arranged in the vertical direction (depth direction) of the semiconductor layer, which is practically compared to the conventional structure in which pixels by wavelength (by color) are arranged in the plane direction. The pixel size can be increased, the amount of incident light per unit area can be increased, the conversion efficiency of the optical-electrical signal can be increased, and high sensitivity characteristics can be obtained.

  In addition, since the detection layer can be arranged in the vertical direction (depth direction) of the semiconductor layer, it is virtually unnecessary to provide a pixel for each wavelength, and a single pixel can be detected by separating a plurality of wavelengths. become able to. As a result, it becomes easy to realize a sensor having a high density of pixels. Of course, since no color filter is required, the cost is reduced.

  In addition, in a sensor that uses the difference in absorption coefficient depending on the wavelength, for example, in a layer that detects blue light, when red light or green light passes, it is absorbed to some extent, resulting in color mixing and a decrease in color discrimination ability. In the case of the structure using the band gap, this color mixing problem can be solved and sufficient color reproducibility can be obtained. As a result, it is not necessary to correct the color mixture by signal processing, the corresponding circuit can be omitted, and the cost is reduced.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. Various changes or improvements can be added to the above-described embodiment without departing from the gist of the invention, and embodiments to which such changes or improvements are added are also included in the technical scope of the present invention.

  Further, the above embodiments do not limit the invention according to the claims (claims), and all combinations of features described in the embodiments are not necessarily essential to the solution means of the invention. Absent. The embodiments described above include inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. Even if some constituent requirements are deleted from all the constituent requirements shown in the embodiment, as long as an effect is obtained, a configuration from which these some constituent requirements are deleted can be extracted as an invention.

<< Spectra of four or more colors >>
For example, in the above-described embodiment, the visible light band is divided into the wavelength components of the three primary colors B, G, and R. Actually, however, in the photodiode group formed in the vertical direction of the semiconductor layer in one pixel. By further separating the detection layers (virtual photoelectric conversion elements) for each color to be provided, it becomes possible to perform finer spectroscopic analysis, so that it is possible to detect four or more colors. In this case, it is possible to detect accurate color information that could not be obtained by conventional three-primary-color imaging. That is, it is also possible to detect separately in four or more colors, and more accurate information can be obtained as the spectral color components other than red, green, and blue are increased.

  In an imaging method using only the three primary colors, it is impossible to detect a portion outside the triangle formed by the origins of the three primary colors on the chromaticity diagram. On the other hand, by disposing a detection layer for the fourth color (or higher) that detects the color component outside the triangle and finely dispersing the color within the photodiode group, it is possible to accurately detect the color component outside the triangle. Color can be detected. Therefore, when the signal is reproduced on the display side, color reproduction is performed more accurately. Furthermore, if a multi-color display system of four or more colors is used on the display side, a more accurate image with a wide color reproduction range on the chromaticity diagram can be obtained.

  For example, in addition to each detection layer to detect three colors of red (R), green (G), and blue (B), a detection layer is provided to detect emerald (E) as a fourth color. It can also be configured to perform color spectroscopy. The emerald is located between blue and green in terms of wavelength and is split between blue light and green light. Therefore, in the photodiode group, a detection layer that detects blue light and green light is detected. What is necessary is just to arrange | position the detection layer which detects emerald light between a detection layer.

  In this case, the fourth color pixel E is added to the color pixel in order to improve the color reproducibility, and the overall operation can be made exactly the same as described above, and the same effect as described above can be obtained. You can enjoy it. In the case of these four colors, more accurate information can be obtained. When an image image detected using this sensor is converted into a signal and the signal is reproduced on the display side, color reproduction is also performed accurately. . Furthermore, if a multi-color display system of four or more colors is used on the display side, an accurate video with a wide color reproduction range can be obtained.

  Although a detailed description of the color signal processing is omitted, a matrix for generating three colors of RGB close to the human eye from pixel signals of each color photographed in four colors corresponding to the detection layers for four colors. An image processor for performing the calculation is provided. If emerald (E) is detected in addition to red (R), green (G), and blue (B), the difference in color reproduction is reduced as compared with detection with three colors of R, G, and B. For example, the reproducibility of blue-green or red can be improved.

<< Spectra other than visible light band >>
Furthermore, the above-described technique is not necessarily limited to the spectroscopy to a plurality of wavelength components in the visible light band. If the structure is optimized, infrared light and ultraviolet light can be separated and detected without using a subtractive color filter. In this case, for example, infrared light and ultraviolet light can be detected and imaged together with visible light. For visible light detected at the same time, it is not limited to detecting a monochrome image without splitting, but a color image can also be detected by splitting the visible light band into, for example, three primary color components as described above. it can. Accordingly, image information of infrared light and ultraviolet light that cannot be seen by the eyes can be simultaneously acquired in correspondence with an image image (monochrome image or color image) of visible light that can be seen by the eyes. This expands the application as a key device for new information systems such as night vision cameras and photosynthetic surveillance cameras.

It is a figure explaining the concept of the imaging device (image sensor) which performs color separation by controlling a band gap. It is a figure explaining the 1st example of the 1st method of obtaining a band gap. It is a figure which shows the quantum well type | mold structure explaining the 2nd example of the 1st method of changing the composition ratio of a mixed crystal system. It is a figure which shows the calculation result at the time of producing the Si quantum well in SiGeC explaining the 1st example of the 2nd method of obtaining a band gap. It is a figure which shows the example of a laminated structure of a quantum well explaining the 1st example of the 2nd method of obtaining a band gap. It is a figure explaining the extraction method (1st example) of a signal. It is a figure explaining the extraction method (2nd example) of a signal. It is a figure which shows the quantum well type | mold structure explaining the 2nd example of the 2nd method of obtaining a band gap. It is a figure which shows the calculation result at the time of producing the quantum well of CdSe in ZnSSe explaining the 2nd example of the 2nd method of obtaining a band gap. It is a figure which shows the laminated structure example of the quantum well explaining the 2nd example of the 2nd method of obtaining a band gap. It is a figure which shows the lattice constant explaining the 2nd example of the 2nd method of obtaining a band gap. It is a figure explaining the application example of an ion implantation method. It is a figure explaining the application example of a thermal diffusion method. It is a flowchart which shows the process sequence explaining the application example of CVD method. It is a schematic diagram which shows the layer structure of the solid-state image sensor manufactured explaining the application example of CVD method. It is a schematic diagram which shows the final form of the layer structure of a solid-state image sensor explaining the application example of CVD method. It is a flowchart which shows the process sequence explaining the application example of MBE method. It is a schematic diagram which shows the layer structure of the solid-state image sensor manufactured explaining the application example of MBE method. It is a schematic diagram which shows the final form of the layer structure of a solid-state image sensor explaining the application example of MBE method. It is a figure explaining the mechanism of the sensor given in patent documents 1.

Explanation of symbols

  11, 21 ... Si substrate, 12, 32 ... red detection area, 14, 34 ... green detection area, 16, 36 ... blue detection area, 20 ... solid-state image sensor, 22 ... red detection layer, 24 ... green detection layer, 6 DESCRIPTION OF SYMBOLS ... Blue detection layer, 31 ... GaAs substrate, 105 ... Charge-voltage signal conversion part, 132 ... Charge generation part, 136 ... Reset transistor, 138 ... Floating diffusion, 140 ... Read-out line selection transistor, 142 ... Amplification transistor

Claims (6)

Each of the wavelengths of the electromagnetic wave includes a plurality of detection regions stacked in a depth direction from a light incident surface facing the substrate on a predetermined substrate, and the electromagnetic waves are separated and detected in each detection region. A physical information acquisition device for detecting a physical quantity distribution that outputs a unit signal corresponding to
The plurality of detection regions are formed as a plurality of detection regions having different band gaps depending on the wavelength of the light so that the plurality of light can be selectively absorbed in the depth direction from the light incident surface. ,
The band gaps of the plurality of detection regions are narrowed away from the incident surface of the light incident surface,
Each detection area is
By controlling the bandgap of the respective detection regions by changing the composition ratio of the ternary or higher mixed crystal containing Si, it is configured to fit the substrate lattice matching condition,
Having a quantum well structure,
It is characterized by
Physical information acquisition device. The detection region has a superlattice structure,
The physical information acquisition apparatus according to claim 1. The detection region is configured to meet the lattice matching condition by a mixed crystal of SiC mixed with Ge, and has a quantum well structure.
The physical information acquisition apparatus according to claim 1. The detection region has a thin film of a superlattice layer having a lattice constant different from that of Si at an interface between Si and SiC or a SiGeC-based layer.
The physical information acquisition apparatus according to claim 3. Each of the wavelengths of the electromagnetic wave includes a plurality of detection regions stacked in a depth direction from a light incident surface facing the substrate on a predetermined substrate, and the electromagnetic waves are separated and detected in each detection region. A method for manufacturing a physical information acquisition device for detecting a physical quantity distribution that outputs a unit signal corresponding to
The plurality of detection regions are formed as a plurality of detection regions having different band gaps depending on the wavelength of the light so that the plurality of light can be selectively absorbed in the depth direction from the light incident surface, respectively.
The band gaps of the plurality of detection regions are narrowed away from the incident surface of the light incident surface,
Each detection area is
By controlling the bandgap of the respective detection regions by changing the composition ratio of the ternary or higher mixed crystal containing Si, it is configured to fit the substrate lattice matching condition,
Formed to have a quantum well structure,
A method for manufacturing a physical information acquisition apparatus, characterized in that: The plurality of detection regions are formed by any of the following methods:
Formed by repeating ion implantation,
Formed by repeating thermal diffusion,
It is formed by performing crystal growth by a plurality of CVD methods.
It is formed by performing MBE method or MOCVD method multiple times.
The manufacturing method of the physical information acquisition apparatus of Claim 5.

Images