JP2011192838A - Light-receiving element, light-receiving element array, hybrid detecting device, photo sensor device, and manufacturing method for light-receiving element array - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a light-receiving element and a light-receiving element array that have high light-receiving sensitivity in a near-infrared range, a manufacturing method for the light-receiving element array, a hybrid detecting device, and an optical sensor device. <P>SOLUTION: A light-receiving layer 3 having a band-gap wavelength of 1.65 to 3.0 μm is composed of a type 2 MQW formed of (GaAsSb/InGaAs) or (GaAsSb/InGaAsN (P, Sb)). Light-receiving elements are separated from adjacent light-receiving elements via grooves or impurity concentration distributions. An electrode 11 for each light-receiving element is in ohmic contact with an n-type region 6 formed on a cap layer for each light-receiving element. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a light receiving element, a light receiving element array having a light receiving sensitivity from the near infrared to the infrared region, a hybrid detection device using the light receiving element array, an optical sensor device, and a method for manufacturing the light receiving element array. is there.

  Since the near-infrared wavelength region corresponds to an absorption spectrum related to living organisms such as animals and plants and the environment, the development of near-infrared detectors using III-V compound semiconductors in the light-receiving layer has been actively conducted. Yes. For example, by using Extended-InGaAs for the light-receiving layer, a CMOS (Complementary Metal-Oxide Semiconductor) circuit, which is a read-out IC (ROIC), is connected to a light-receiving element array that has sensitivity up to a wavelength of 2.6 μm. An example of a detector that converts a photocurrent into an output signal has been published (Non-Patent Document 1). In the light receiving element array, electrons are collectively collected in a common n-side electrode among electron / hole pairs generated by light incident on the pn junction, and holes are read out from the p-side electrode, which is a pixel electrode, to the CMOS. .

In addition, it has been reported that a light receiving element using an InGaAs / GaAsSb type 2 multi-quantum well structure (MQW: Multi-Quantum Wells) in the light receiving layer has sensitivity up to a wavelength of 2.5 μm (Non-patent Document 2). .
In addition, an example in which a CMOS readout circuit is used for a light receiving element array in a shorter wavelength range (visible range) instead of the near infrared range is disclosed (Patent Documents 1 and 2). In this case, the signal charges of electrons are read out from the pixels in the n-type region using silicon as the semiconductor.

Japanese Patent Laid-Open No. 11-307753 JP 2002-217397 A

Hideo Takahashi et al. “InGaAs photodetector for near infrared”, OPTRONICS (1997), No.3, pp.107-113 R.Sidhu, N.Duan, J.C.Campbell, and A.L.Holmes, Jr., "A 2.3μmcutoff wavelength photodiode on InP using lattice-matched GaInAs-GaAsSb type IIquantum wells” 2005 International Conference on Indium Phosphide and RelatedMaterials

In the near-infrared InGaAs photodetector (image sensor) of Non-Patent Document 1 described above, InGaAs having a composition that does not lattice match with the InP substrate is used as the light-receiving layer, the dark current increases and the sensor noise increases. The buffer layer is devised to gradually alleviate distortion, but there is a limit. In order to reduce this noise and improve the S / N ratio to a practical level, a cooling device is required, resulting in a large-scale device. InP or the like cannot be used for the window layer epitaxially grown on the InGaAs light receiving layer. For example, InAsP (As / P is approximately 0.6 / 0.4) lattice-matched to the InGaAs light receiving layer is used as the window layer. When used, since the InAsP layer has an absorption band at a wavelength of 1 to 1.5 μm, the sensitivity to light in this wavelength range becomes small regardless of whether it is incident on the front surface or the back surface.
In addition, the light receiving element of Non-Patent Document 2 has never been arrayed to form a detection device, and there is no example in which the sensitivity of the arrayed detection device is measured.
In addition, the light receiving elements in Patent Documents 1 and 2 using silicon as a semiconductor do not have sensitivity in the near infrared region.

  The present invention provides a light-receiving element, a light-receiving element array, a hybrid detection device, an optical sensor device, and a method for manufacturing the light-receiving element array using the light-receiving element array, which have high light-receiving sensitivity in the near infrared region. Objective.

In the light receiving element array of the present invention, light receiving elements are arranged. The light receiving element array includes an InP substrate, a light receiving layer having a band gap wavelength of 1.65 μm to 3.0 μm and positioned on the InP substrate, and a cap layer positioned on the light receiving layer. The light-receiving layer is made of any one of type 2 type MQW of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb) that is lattice-matched to the InP substrate. Each of the light receiving elements includes an electrode positioned on the cap layer, and the electrode is in ohmic contact with the cap layer in an n-type region provided for each of the light receiving elements, and the adjacent light receiving element and the groove or n-type It is characterized by being separated by a region that is not formed. Here, the light receiving elements in the light receiving element array may be a one-dimensional array or a two-dimensional array, and may be a regular array or an irregular array in either case.

According to the above configuration, the light receiving layer is formed of type 2 MQW, and the pixel electrode formed of the light receiving element is in ohmic contact with the n-type region. For this reason, the charge due to electrons in the electron-hole pair formed by receiving light with MQW is read out as pixel information.
In general, among carriers (electrons and holes) in a semiconductor band, holes tended to have lower mobility than electrons. Even when a type 2 MQW is used as a light-receiving layer and a pixel readout portion is used as a p-type region, a quantum efficiency of 0.3 to 0. 0 over the near-infrared region when light is incident from the cap layer side opposite to the InP substrate. It had a sensitivity of about 9. However, in the case where light is incident from the back side of the InP substrate, the light receiving sensitivity in the above-mentioned wavelength region is greatly reduced when the type 2 type MQW is used as the light receiving layer and the pixel reading unit is used as the p type region. This phenomenon is a phenomenon that the present inventors have recently known. The mechanism by which this phenomenon occurs is assumed as follows.
“When holes are trapped in one quantum well of MQW, a certain amount of energy is required to exceed the barrier potential of the quantum well. Light incident from the cap layer side opposite to the InP substrate is incident. It is easy to perform photoelectric conversion with MQW in the range close to the cap layer immediately, and there are only a few quantum wells from the quantum well where the hole with low mobility is located in the generated electron-hole pair to the pixel electrode. Can reach the pixel electrode relatively easily and obtain sufficient sensitivity, however, in the case of incidence from the back surface of the InP substrate, the position where light is incident and received immediately is a quantum well in a range close to the InP substrate. For this reason, when the InP substrate is incident on the back surface, holes need to exceed a large number of quantum wells in order to reach the pixel electrode. Therefore, when the type 2 type MQW is used as a light receiving layer and the pixel is used as a p-type region, the cap layer side incident shows high sensitivity. When the back side of the InP substrate is incident, the sensitivity is greatly reduced, and the degree of this sensitivity reduction is affected by the light absorption rate of the light receiving layer that varies depending on the wavelength. "
Until now, data was collected for the incident on the cap layer side for one light receiving element having a light receiving layer of type 2 type MQW. However, when a plurality of light receiving elements are arranged in a two-dimensional array, the backside incidence of the InP substrate is unavoidable so that the pixel wiring does not become an obstacle to incident light.

  The reason why the pixel of the InP III-V compound semiconductor is set to the p-type region is due to the ease of handling of the impurity element and the technology accumulated so far. Zinc (Zn) is widely used because it is very easy to handle as an impurity of InP, and has a lot of accumulated data, and has been regarded as a standard of InP impurities. For this reason, Zn has been selectively diffused into InP, and a technique of using a selective diffusion region (p-type region) as a pixel has been accumulated. Each pixel is isolated by a region that is not selectively diffused. InP has continued to be used without other impurities other than Zn. Zn is such an easy-to-handle impurity for InP. However, since Zn is a p-type impurity of InP, other disadvantages have been dealt with at a high cost by making the pixel a p-type region. For example, taking a multiplexer as an example, in an imaging device using a light receiving element array made of silicon whose pixels are n-type regions, the readout electrode of the CMOS or CCD has a circuit configuration for reading out signal charges by electrons. For this reason, in order to read out and process the signal charge due to holes, it is necessary to cope with changing the polarity.

As described above, when the light-receiving element array is a type 2 MQW and the light-receiving element array having the selective diffusion region (p-type region) as a pixel is incident on the back surface from the InP substrate side, Of these, a significant proportion of holes cannot reach the pixel electrode through the MQW, resulting in a decrease in light receiving sensitivity. Therefore, the basic idea of the present invention is to use the n-type impurity selective diffusion region as a pixel without depending on the ease of use of Zn in InP. For this reason, it is set as a nip type light receiving element array. The present invention is based on the phenomenon that has been confirmed that the light receiving element array by the type 2 MQW is configured in the pin type to significantly reduce the sensitivity. In the present invention, the following effects can be obtained by making the light-receiving layer an MQW of type 2 type and making the light-receiving element array having the selective diffusion region (n-type region) a pixel as the back side incidence from the InP substrate side. .
(1) Electrons can reach the pixel electrode with high efficiency (yield) by exceeding the number of quantum wells relatively easily because of the higher mobility than holes. For this reason, the fluctuation | variation of the sensitivity by a wavelength can be made small.
(2) Moreover, since the effect of the above (1) can be obtained, the pn junction can be arranged in the MQW in the range close to the InP substrate when the InP substrate is incident on the back surface. With such an arrangement, when the InP substrate is incident on the back surface of the InP substrate, the depletion layer from the pn junction can be expanded in a region where the amount of light immediately after entering is large. In such a region with a large amount of light, the degree of light absorption is large. As a result, it is possible to improve the light receiving sensitivity and obtain a light receiving element array in which the variation in sensitivity due to the wavelength is small.
Note that the fact that a given epitaxial layer is lattice-matched with the InP substrate means that the difference from the lattice constant of InP is 0.002 or less, that is, the lattice constant of the InP substrate is ao and the lattice constant of the epitaxial layer is a. Is satisfied | a−ao | /ao≦0.002.

  The n-type region for each light receiving element includes (1) a part of the thickness from the surface of the cap layer, (2) all the thickness of the cap layer, (3) all the thickness of the cap layer, and the cap layer side of the light receiving layer And (4) all thicknesses of the cap layer, and all thicknesses of the light-receiving layer, so as to have a thickness direction distribution of any one of the forms (1) to (4) it can. As described above, by having many options of the depth for introducing the n-type impurity, the position of the pn junction can be arranged in the range of the cap layer to the bottom surface of the MQW. For example, when the position of the pn junction is close to the InP substrate, the depletion layer can be brought close to the incident surface on the back surface, and photoelectric conversion can be performed immediately upon entering. Since electrons have a higher mobility than holes, they can reach a pixel electrode with high efficiency (yield) over a large number of quantum wells relatively easily. For this reason, the fluctuation | variation of the sensitivity by a wavelength can be made small. Further, since it is considered that the crystallinity of MQW deteriorates, it is preferable to select the depth for introducing the n-type impurity in consideration of the crystallinity of MQW.

The cap layer can be an InP layer or a composite layer of an InGaAs layer and an InP layer. Since InP and InGaAs have a small absorption at a wavelength of 1 μm to 1.5 μm, the light receiving sensitivity of the entire range from the wavelength of 1 μm to the near infrared region can be increased.
Since InP and InGaAs are lattice-matched with the InP substrate, the absorption in the above wavelength range is small by using it not only for the cap layer but also for the buffer layer. Absorption can be suppressed, and sensitivity can be increased.
Here, when the cap layer is a composite layer of an InP layer and an InGaAs layer and has a laminated structure of InP layer / InGaAs layer / light receiving layer, the InGaAs layer is referred to as an impurity concentration distribution adjusting layer. The term “cap layer” refers to the case where only the InP layer is referred to, and the case where it refers to the composite layer of (InP layer / InGaAs impurity concentration distribution adjustment layer). However, it should be construed broadly within the spirit of the present invention.
The impurity concentration distribution adjusting layer has a pn junction at various depth positions and a depletion layer as a light receiving layer (low concentration carrier under a reverse bias voltage) in order to make the nip type photodiode in the pixel into various forms. ) Can be arranged to overhang. The impurity concentration distribution adjusting layer may be disposed in order to form a nip type photodiode of an appropriate form according to various manufacturing methods.

  The light receiving element has an n-type region formed by selective diffusion or selective ion implantation, and is different depending on the region not selectively diffused and not n-type or not selectively ion-implanted and not n-type. It can be separated from the light receiving element. As a result, the light receiving element array can be formed only by the impurity concentration distribution without performing mesa etching. For this reason, crystal damage accompanying mesa etching can be avoided.

  Further, the light receiving element may have a cap layer doped with n-type impurities, or a cap layer and light receiving layer doped with n-type impurities, and may be separated from an adjacent light receiving element by a groove. According to this, regardless of selective diffusion, an n-type impurity is doped at the time of epitaxial film formation, and adjacent light receiving elements are isolated by mesa etching. For this reason, when selective diffusion is difficult depending on the type of n-type impurity, an array can be formed while keeping the light receiving elements independent.

  The light receiving element has a cap layer doped with an n-type impurity, or a cap layer doped with an n-type impurity and a light receiving layer, and a p-type region selectively diffused with a p-type impurity, or a p-type impurity with a selective ion The implanted p-type region can be separated from the adjacent light receiving element. As a result, the p-type impurity is selectively diffused or ion-implanted into the region between the light receiving elements without mesa etching so that the n-type region formed by doping during film formation becomes a p-type region. Isolate from next door.

(1) an n ⁻ type layer, (2) a p ⁻ type layer, and (3) an impurity composite layer in which the cap layer side is an n ⁻ type layer and the InP substrate side is a p ⁻ type layer. Any one of 1) to (3), and the carrier concentration can be 1e16 / cm ³ or less. In either case, the carrier concentration in any part of the light receiving layer is 1e16 / cm ³ or less. Thereby, MQW with good crystallinity can be obtained. In addition, the depletion layer can be easily spread widely in the light receiving layer (MQW) from the pn junction, and the sensitivity can be increased.

  The hybrid detection device of the present invention includes any one of the light receiving element arrays described above and a silicon readout circuit, and an electrode for each light receiving element of the light receiving element array and a readout electrode of the silicon readout circuit intervene a bonding bump. The light is incident from the back surface of the InP substrate in the light receiving element array. As a result, a hybrid detection device having high light receiving sensitivity in the near infrared region can be obtained.

  The reverse bias voltage applied to the light receiving elements in the light receiving element array can be set to 0 to −1V. As a result, near infrared light can be received using a small power source. For this reason, it is possible to obtain a hybrid detection device that can be reduced in size and weight and is convenient to carry.

An optical sensor device according to the present invention uses any one of the light receiving element arrays described above or any hybrid detection device. Accordingly, since light can be received with high sensitivity up to a long wavelength region in the near infrared region, for example, inspection can be performed using a plurality of wavelengths that can be used for specifying a predetermined component. For this reason, the accuracy of the inspection can be increased, or an inspection that is not possible before the present invention can be performed.
Here, the optical sensor device is combined with an optical element, for example, an optical system such as a spectroscope or a lens, and performs wavelength distribution measurement or can be used as an imaging device to obtain many useful practical products. Can do. Of course, it often includes a control device such as a microcomputer. Specific examples of the optical sensor device include (1) an imaging device for visual field support or monitoring, (2) a biological component inspection device for inspecting components in the living body, and (3) for inspecting moisture concentration. And (4) a food quality inspection device for inspecting the quality of food, and (5) an environmental monitoring device for monitoring components in gas such as an incinerator.

  The light receiving element of the present invention has an InP substrate, a band gap wavelength of 1.65 μm to 3.0 μm, a light receiving layer located on the InP substrate, and a cap layer located on the light receiving layer. The layer is lattice-matched to the InP substrate and is a type 2 multiple quantum well structure of any one of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb) Comprising an electrode located on the cap layer, the electrode being in ohmic contact with the cap layer in the n-type region, wherein the n-type region is separated from the surrounding by a region that is not n-typed It is characterized by. Accordingly, it is possible to obtain a light receiving element having high light receiving sensitivity up to a long wavelength range in the near infrared region regardless of whether the substrate is incident or the cap layer is incident. This light receiving element may have a low reverse bias voltage, and can contribute to reduction in size and weight of the power source. Furthermore, it is not necessary to make the readout electrode such as CMOS have a special specification corresponding to holes, and an inexpensive optical sensor device can be obtained.

The light receiving element array manufacturing method of the present invention manufactures a light receiving element array in which light receiving elements are arranged. The manufacturing method includes any one of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb) lattice-matched on the InP substrate or the buffer layer. A step of forming a type 2 MQW light receiving layer, a step of forming a cap layer on the light receiving layer, and an isolation process for forming a light receiving element in the cap layer and the light receiving layer separately from the adjacent light receiving element Forming an n-type region in at least a surface layer of the cap layer for each of the light receiving elements in the isolation process, or the isolation process and the cap layer / light-receiving layer forming process. An electrode is formed for each light receiving element in the n-type region.
As a result, a light receiving element array having high light receiving sensitivity in the near infrared region can be manufactured relatively easily.

In formation of an n-type region for each light receiving element in the isolation processing step or the isolation processing step and the cap layer / light receiving layer forming step, (1) selective diffusion or selective ion implantation from the surface of the cap layer. And forming the n-type region and separating it from other light-receiving elements by a region not selectively diffused or a region not selectively ion-implanted. (2) When forming a cap layer or a cap layer and a light-receiving layer, the n-type region is formed. Doping with impurities and providing a groove by etching between the adjacent light receiving element and separating from the adjacent light receiving element by the groove, or (3) When forming the cap layer or the cap layer and the light receiving layer, An n-type impurity is doped, a p-type impurity is introduced between the adjacent light receiving element by selective diffusion or ion implantation, and a region containing the p-type impurity is adjacent to the adjacent light receiving element. Separating the optical element, it is possible.
This makes it possible to obtain an array of light receiving elements isolated from adjacent ones, which is suitable for the characteristics of the impurity element from among the options.

  According to the present invention, a light receiving element array or the like having high light receiving sensitivity in the near infrared region can be obtained.

It is sectional drawing which shows the hybrid type | mold detection apparatus of the near infrared region in Embodiment 1 of this invention. It is the top view which looked at the light receiving element array of FIG. 1 from the light incident side. It is a figure which shows the thickness direction distribution of the carrier density | concentration in one light receiving element in the light receiving element array of FIG. It is sectional drawing which shows the hybrid detection apparatus of the near infrared region in Embodiment 2 of this invention. FIG. 5 is a diagram showing a thickness direction distribution of carrier concentration in one light receiving element in the light receiving element array of FIG. 4. It is sectional drawing which shows the hybrid detection apparatus of the near infrared region in Embodiment 3 of this invention. It is a figure which shows the thickness direction distribution of the carrier density | concentration in one light receiving element in the light receiving element array of FIG. It is sectional drawing which shows the hybrid detection apparatus of the near infrared region in Embodiment 4 of this invention. It is a figure which shows the imaging device or visual field assistance apparatus which is an optical sensor apparatus in Embodiment 5 of this invention. It is a figure which shows the visual field assistance apparatus of the back at night. It is a figure which shows the biological component detection apparatus which is an optical sensor apparatus in Embodiment 6 of this invention. It is a figure which shows the water | moisture content detection apparatus (water distribution image formation apparatus of an eye) in the living body which is an optical sensor apparatus in Embodiment 7 of this invention. It is a figure which shows the apparatus which test | inspects the taste which is a food quality inspection apparatus which is an optical sensor apparatus in Embodiment 8 of this invention. It is a figure which shows the absorption spectrum of the range of 1200 nm-2500 nm of the rice of three different brands. It is a temperature distribution measuring apparatus for obtaining the temperature distribution of the garbage in the refuse combustion furnace, which is the optical sensor apparatus according to the ninth embodiment of the present invention. It is a figure which shows the temperature distribution imaging device 20a in FIG. It is a figure which shows the near-infrared spectrum in a refuse combustion furnace. It is a figure which shows the absorption spectrum of water. It is sectional drawing which shows the hybrid detection apparatus of the near infrared region before this invention. It is a figure which shows the thickness direction distribution of the carrier density | concentration in one light receiving element in the light receiving element array of FIG. It is a figure which shows the level in the valence band of the hole in MQW of the light receiving element array before this invention.

<Problems with Hybrid Type Detection Device Prior to the Present Invention>
The problem with the hybrid type detection device prior to the present invention using the MQW for the near infrared region as the light receiving layer described here is not yet known.
FIG. 19 is a diagram showing a hybrid detection device 110 including a light receiving element array 150 prior to the present invention and a CMOS 170 constituting a readout circuit (ROIC). The light receiving element array 150 has the following laminated structure.
n-type InP substrate 101 / n-type InP (or InGaAs) buffer layer 102 / n ^- type 2 MQW / n type light-receiving layer 103 (InGaAs / GaAsSb) ^- type InGaAs selective diffusion concentration distribution control layer 104 / n ^- -type InP Cap layer 105
The light receiving element or the photodiode includes a pn junction 115 located at the tip of the p-type region 106 introduced by selective diffusion from the surface of the n ⁻ -type InP cap layer 105. Each light receiving element is separated by a region that is not selectively diffused. The p-type impurity to be selectively diffused is Zn. As described above, Zn is very easy to handle as an impurity of InP, and there is abundant technical data accumulated so far. The selective diffusion mask pattern 136 used for selective diffusion is left as it is, and a protective film 137 is covered on the selective diffusion mask pattern 136.
The electrode of the light receiving element or the pixel electrode 111 is disposed so as to be in ohmic contact with the p-type region 106, and the ground electrode 112 is disposed so as to be in ohmic contact with the n-type InP substrate 101 in common with each light receiving element. The pad 171 forming the readout electrode of the CMOS 170 is conductively connected to the pixel electrode 111 with the bonding bump 131 interposed. The ground electrode 172 of the CMOS 170 and the ground electrode 112 of the light receiving element array 150 are grounded to the outside.
FIG. 20 is a diagram showing the carrier concentration distribution along the thickness direction of the light receiving element array 150. In FIG. 20, p-type carriers corresponding to the concentration distribution of selective diffusion of Zn pass through the InGaAs selective diffusion concentration distribution adjusting layer 104 from the high-concentration cap layer 105 in the selective diffusion introduction layer, and p-type in the light receiving layer 103. The carrier concentration is low. The flat n-type carrier concentration in type 2 MQW is due to doping during film formation. A pn junction 115 is formed at the intersection (intersection or interface) between the p-type carrier concentration which is inclined and decreases on the light receiving layer top side (cap layer side) and the flat n-type carrier concentration.
When receiving light, a reverse bias voltage is applied to the pn junction 115, that is, a voltage is applied between the pixel electrode 111 and the ground electrode 112 so that the voltage of the ground electrode 112 is higher than that of the pixel electrode 111. The depletion layer extends to the light receiving layer 103 of the type 2 MQW, and electron-hole pairs are formed by the light reaching here. Since the pixel electrode 111 has a voltage lower than that of the ground, it collects holes, and the charge of the holes forms pixel information. By reading out the charges of the pixels at a predetermined time pitch, it is possible to form an intensity distribution of an image or a measurement signal.
In the type 2 MQW of the light receiving layer, holes are formed in the valence band of GaAsSb as shown in FIG. Since the band diagram shown in FIG. 21 is a band diagram for electrons, holes are read upside down. The holes trapped at the lowest level drift to the cap layer 105 side over the high barrier in the valence band. Light incident from the back surface of the InP substrate 101 is immediately received in the depletion layer to generate holes. At this time, although the holes are driven by a reverse bias electric field, they must reach the p-type region 106 through the MQW light-receiving layer 103 through many high barriers. For this reason, the ratio of the number of holes reaching the p-type region 106 or the pixel electrode 111 is considerably reduced. As a result, the light receiving sensitivity is lowered. Originally, it was known that holes have a larger effective mass than electrons and a low mobility. However, the above-described decrease in light receiving sensitivity cannot be explained by such a general mobility. The mechanism of the phenomenon in which the light receiving sensitivity is reduced is being investigated. However, anyway, the type 2 MQW light receiving layer 103 is provided, and the pixel electrode 111 is disposed in the p-type region 106 so that holes are signal charges. The following experimental fact is confirmed about the light receiving element array 150 or the hybrid type detection device 110.
1. When light is incident from the cap layer 105 side with respect to a light-receiving element using Type 2 MQW as a light-receiving layer, the quantum efficiency of near-infrared light is 0.3 to 0.9. This quantum efficiency is good.
2. However, if the same light receiving element array is incident on the back surface of the InP substrate, the quantum efficiency in the near infrared region is reduced to a very low value of 0.05 to 0.5. In the case of using the light receiving element array, since wiring is provided for each pixel, in order to avoid interference with light due to the wiring, it must be incident from the back surface of the InP substrate.

<Points of the present invention>
The point of the present invention is as follows. The hybrid detection device includes a light receiving element array in which a light receiving layer is a type 2 MQW (for example, (InGaAs / GaAsSb)), and a readout circuit (ROIC). In particular, it is important that type 2 MQW is used for the light receiving layer. In the light receiving element array, an electrode arrangement portion of each light receiving element is used as an n-type region, and a ground electrode is arranged in a p-type region common to each light receiving element. Near-infrared light having a wavelength of 1.65 μm to 3.0 μm is incident from the back surface of the InP substrate of the light receiving element array, and the CMOS constituting the readout circuit is an electron hole pair generated by photoelectric conversion in a type 2 type MQW. Reads out the charge from the electrons.
In the present invention, holes are not used as signal charges, and electrons are used as signal charges. Electrons have higher mobility than holes. By using electrons emitted to the conduction band as signal charges, the light receiving sensitivity or quantum efficiency can be increased. The formation of the n-type region in which the pixel electrode is arranged is not as technically accumulated as Zn of the p-type impurity, but selective diffusion of Sn and other n-type impurities, or other n-type region forming methods not heretofore Alternatively, a pn junction formation method is used.

(Embodiment 1)
FIG. 1 is a cross-sectional view showing a hybrid detection device 10 according to Embodiment 1 of the present invention. The light receiving element array 50 includes a p-type InP substrate 1 / p-type InP buffer layer 2 / p ⁻ -type light receiving layer (light absorption layer) 3 / p ⁻ -type InGaAs impurity concentration distribution adjusting layer 4 / p ⁻ -type InP cap layer 5, It is formed in the laminated body. In each light receiving element, n-type impurity tin (Sn) is selectively diffused to form an n-type region 6, and a pn junction 15 is formed at the tip of the n-type region 6. The n-type region 6 in the light receiving element constituting the pixel P is separated from the adjacent n-type region by a region that is not selectively diffused. Therefore, the light receiving element array 50 with a low dark current can be obtained with a simple structure without forming a mesa structure or the like. A ground electrode 12 is disposed on the InP substrate 1 which is a p-type region common to each light receiving element.
The readout circuit of the hybrid detection device 10 is configured by a CMOS 70 having readout electrodes or pads 71. A ground electrode 72 is also provided in the CMOS 70 and is grounded together with the ground electrode 12 in the light receiving element array 50.

  The electrode 11 of the light receiving element or pixel P of the light receiving element array 50 and the pad 71 serving as the readout electrode of the CMOS 70 are conductively connected by the bonding bump 31. An indium alloy is preferably used as the metal material for forming the bonding bump 31. An antireflection film 35 made of a SiON film is disposed on the back surface of the InP substrate 1 serving as an incident surface on which light is incident. Further, the SiN selective diffusion mask pattern 36 used for selective diffusion for forming the n-type region 6 is left as it is. Furthermore, a protective film 37 such as a polyimide resin covering the opening of the selective diffusion mask pattern 36 or the surface of the InP cap layer 5 and the selective diffusion mask pattern 36 is provided.

  As described above, type 2 MQW is used for the light-receiving layer 3. In order to adjust the impurity concentration when selectively diffusing tin (Sn) or the like, which is an n-type impurity, or the n-type impurity concentration by doping described in the second and subsequent embodiments, the cap layer is formed of an InP cap layer 5 and an impurity. A mixed layer with the concentration distribution adjusting layer 4 is used. That is, the impurity concentration distribution adjusting layer 4 is inserted between the InP cap layer 5 and the light receiving layer 3. The specific structure of the type 2 MQW will be described in detail later.

  In the present embodiment, the bonding bumps 31 are provided on the pixel electrodes 11 that are in ohmic contact with the n-type region 6 of the light receiving element array 50 before being assembled as the hybrid detection device 10. At the time of connection or pressure bonding, the bonding bump 31 is pressure-bonded to the pad 71 of the CMOS 70 and conductively connected.

  FIG. 2 is a plan view of the light receiving element array 50 as viewed from the light incident side. For example, the overall size is 10 mm wide × 9 mm long, the number of pixels P is 320 × 256 (about 82,000 pixels), and the pitch is 25 μm.

FIG. 3 is a diagram showing a carrier concentration distribution in the thickness direction of one light receiving element in the light receiving element array 50. As can be seen from comparison with the carrier concentration distribution of the light receiving element before the present invention shown in FIG. 20, the holes (p-type carrier) and the electrons (n-type carrier), which are carriers of the carrier, are interchanged. That is, the InP substrate 1 / InP buffer layer 2 / type 2 type MQW light receiving layer 3 / InGaAs impurity concentration adjusting layer 4 / InP cap layer 5 are p-type layers. In particular, the p-type carrier concentration of the InP substrate 1 with which the ground electrode 12 is in ohmic contact is set high. In the n-type region 6 formed by introducing an n-type impurity from the surface of the InP cap layer 5 by selective diffusion for forming a light receiving element or pixel, the n-type carrier concentration is high as shown in FIG. Become. In the n-type region 6, the n-type carrier has a high concentration (about 1e18 / cm ³ ) on the cap layer 5 side in the impurity concentration distribution adjustment layer 4, but a low concentration (1e16 in the impurity concentration distribution adjustment layer 4. / Cm ³ or less) and then a gentle gradient, and the light receiving layer 3 is inclined with the gentle gradient.
The flat p-type carrier concentration (about 2e15 / cm ³ ) in the light-receiving layer 3 is formed by doping p-type impurities during the MQW film formation. A boundary surface formed by the intersection of the flat p-type carrier concentration and the gently inclined n-type carrier concentration from the impurity concentration distribution adjusting layer 4 forms the pn junction 15. That is, the pn junction 15 is an inclined junction. In the case of the present embodiment, the pn junction 15 is located near the InGaAs impurity concentration distribution adjusting layer 4 in the light receiving layer 3 and almost at the interface (light receiving layer 3 / impurity concentration distribution adjusting layer 4). In the light receiving layer 3, the p-type carrier concentration is low and flat, and the n-type carrier concentration is inclined from the high concentration and enters the light receiving layer 3. Therefore, the depletion layer generated by applying the reverse bias voltage to the pn junction is The light receiving layer 3 spreads greatly.

During standby for light reception, a reverse bias voltage (preferably a voltage of 0 to −1 V) is applied to the pn junction 15 to spread the depletion layer into the light receiving layer 3 toward the InP substrate 1. As will be described later, in MQW, several tens to several hundreds of InGaAs / GaAsSb are stacked. As shown in FIG. 1, light incident from the back surface of the InP substrate 1 is received in a depletion layer near the InP substrate 1. Electron hole pairs are formed by light reception. In the light receiving element array prior to the present invention, holes are moved to the pixel electrode, but in the present invention, electrons that are n-type carriers are collected in the pixel electrode 11. The features of the present embodiment are as follows.
(1) (i) The electrons are not bound to the bottom level of the valence band of the well in the type 2 type MQW like the holes shown in FIG. Electrons may receive a binding force in the conduction band due to the conduction band barrier, but have a higher mobility than holes as shown in (2) below. For this reason, electrons in the conduction band are not bound as strongly as holes in the valence band. That is, each well acts as a large migration obstacle only for holes.
(Ii) Incident light is not attenuated toward the side closer to the InP substrate 1, and the amount of received light (photoelectric amount) is large. Then, the carrier generated on the side closer to the InP substrate 1 must reach the pixel electrode 11 through the impurity concentration adjusting layer 4 / cap layer 5 over many wells. By changing the carrier from a hole to an electron, even if the depletion layer is extended to a position close to the InP substrate, trapping or annihilation of the signal carrier is unlikely to occur, so that the quantum efficiency can be further increased.
(2) Since the curvature change of the energy wave number (Ek) curved surface of the conduction band in the band structure is larger than that of the curved surface of the valence band, the effective mass of electrons is smaller than the effective mass of holes, Mobility is great. Although the movement obstacle due to MQW has a large effect on holes, even if there is no well structure, electrons are more likely to move than holes due to the conduction band of the band structure and the shape of the Ek curved surface of the valence band. To the pixel electrode 11 easily.
For the reasons (1), (i), (ii), and (2) above, in the hybrid detection device 10 according to the present embodiment, electrons generated by light reception in the light receiving element array 50 efficiently move to the pixel electrode 11. Can contribute to the signal. As a result, it is possible to obtain the hybrid detection device 10 having high quantum efficiency in the near infrared region and excellent light receiving sensitivity.

Next, an example of a manufacturing method will be described (see FIG. 1). First, a p-type InP buffer layer 2 of about 0.1 μm is formed on the InP substrate 1. Next, an MQW light-receiving layer 3 of (InGaAs / GaAsSb) or (GaInNAs / GaAsSb) is formed. The composition of InGaAs is In _0.53 Ga _0.47 As and the composition of GaAsSb is GaAs _0.52 Sb _0.48 so as to lattice match with InP. As a result, the degree of lattice matching (| Δa / a |: where a is the lattice constant of the InP substrate, Δa = ai−a (ai is the lattice constant of each layer) is the difference in lattice constant from the InP substrate). 002 or less.
The InGaAs layer constituting the unit quantum well structure has a thickness of 5 nm, the GaAsSb layer has a thickness of 5 nm, and the number of pairs (the number of repetitions of the unit quantum well) is 250. Next, an In _0.53 Ga _0.47 As layer having a thickness of about 0.3 μm is epitaxially grown on the light receiving layer 3 as the impurity concentration distribution adjusting layer 4, and finally, an InP cap layer having a thickness of about 0.8 μm is finally added. 5 is epitaxially grown. Both the light receiving layer 3 and the impurity concentration distribution adjusting layer 4 are preferably epitaxially grown by MBE (Molecular Beam Epitaxy) method. The InP cap layer 5 may be epitaxially grown by the MBE method, or after the impurity concentration adjusting layer 4 is grown, the InP cap layer 5 may be removed from the MBE apparatus and epitaxially grown by the MOVPE (Metal Organic Vapor Phase Epitaxy) method. .
The n-type region 6 located so as to reach from the InP cap layer 5 to the MQW light-receiving layer 3 is formed by selectively diffusing n-type impurities such as Sn from the opening of the selective diffusion mask pattern 36 of the SiN film. Is done. Introducing selective diffusion within the periphery of each pixel in a limited plane is realized by diffusing using the selective diffusion mask pattern 36 of the SiN film. Since the periphery is limited, the n-type region 6 is separated by a region that is not selectively diffused.
In the n-type region 6, an AuGeNi / Ti / Au n-side electrode 11 is provided in ohmic contact. As the indium alloy for forming the bonding bump 31, for example, an In—Zn alloy of 85% by weight to 97.5% by weight and 2.5% by weight to 15% by weight of Zn may be used. An alloy of In96% -Zn4% is preferable. In addition, In97% -Ag3%, In100%, or the like can be used as a metal that makes ohmic contact with InP. As the metal other than indium, Sn96.5% -Ag3.5% can be used. The ground electrode 12 that is in ohmic contact with the p-type InP substrate 1 is preferably formed of AuZn.

Since the InP substrate 1 is in ohmic contact with the p-side electrode 12 on the back surface, it is preferable to use a substrate containing a p-type impurity such as Zn at a predetermined level or more. For example, a material containing about 1 × 10 ¹⁷ / cm ³ or more of a p-type dopant such as Zn is preferable.
The InGaAs / GaAsSb type 2 type MQW light receiving layer 3, the InGaAs impurity concentration distribution adjusting layer 4, and the InP cap layer 5 may be non-doped or may contain a very small amount of p-type dopant such as Zn (for example, 2 × 10 ¹⁵ / cm 2). ^{About 3} ) Doping may be performed.
The pn junction 15 is an impurity region in the light receiving layer 3 in which the p-type impurity concentration of the region on the side opposite to the side where the n-type impurity element is introduced by selective diffusion is so low that it can be regarded as an intrinsic semiconductor. (Referred to as i region). It also includes a junction (pn junction) 15 formed between the n-type region 6 formed by selective diffusion and the i region. That is, the pn junction may be a ni junction or the like, and includes a case where the p concentration in the np junction is very low.

As described above, using the SiN selective diffusion mask pattern 36 formed on the surface of the InP cap layer 5, n-type impurities such as Sn are selectively diffused from the opening to enter the MQW light receiving layer 3 of InGaAs / GaAsSb. An n-type region 6 is formed so as to reach. The front tip of the n-type region 6 forms a pn junction 15. At this time, a high concentration region having an n-type impurity concentration of about 1 × 10 ¹⁸ / cm ³ or more is preferably limited to the InGaAs impurity concentration distribution adjusting layer 4. That is, the high concentration impurity distribution continues in the depth direction from the surface of the InP cap layer 5 to the InGaAs impurity concentration distribution adjustment layer 4 and greatly decreases on the top side in the impurity concentration distribution adjustment layer 4. It is preferable to make the concentration gradient from about 1 × 10 ¹⁶ / cm ³ to a gentle concentration gradient. The light receiving layer 3 enters with the gentle gradient. As a result, the n-type impurity concentration distribution in the vicinity of the pn junction 15 in the light receiving layer 3 is a distribution indicating an inclined junction.

(Embodiment 2)
FIG. 4 is a cross-sectional view showing hybrid detection apparatus 10 according to Embodiment 2 of the present invention. FIG. 5 is a diagram showing the thickness direction distribution of carrier concentration in one light receiving element in the light receiving element array 50. 4 is different from the light receiving element array 50 of the first embodiment in that the n-type region 6 passes through most of the light receiving layer 3 and reaches a position close to the boundary with the InP buffer layer 2. That is, the n-type region 6 formed by selective diffusion of n-type impurities reaches the position near the boundary between the light receiving layer 3 and the buffer layer 2 through the cap layer 5 / impurity concentration distribution adjusting layer 4. The difference is that the depth of the n-type region 6 is deep, and the other points are the same as those of the first embodiment including the signs of the respective parts.
Numerically, as shown in FIG. 5, the n-type carrier concentration is about 1e18 / cm ^{3 in} the light receiving layer 3 in the vicinity of the boundary with the impurity concentration distribution adjusting layer 4, but in the thickness range of 0.2 μm. It drops sharply to about 1e16 / cm ³ and becomes a gentle gradient. From the above-mentioned boundary to the bottom from about 0.25 μm position, it further decreases with a gentle gradient. The pn junction 15 is formed near the boundary with the InP buffer layer 2. Therefore, even in the case of the inclined junction, the n-type carrier concentration gradient is smaller than that in the first embodiment, and the n-type carrier concentration level is lower. Conversely, the n-type carrier concentration on the n-type carrier side of the pn junction 15 increases stepwise.

  FIG. 4 shows a power supply that applies a reverse bias voltage to the pn junction 15. Not only in the present embodiment but when the n-type impurity selective diffusion region is a pixel, the reverse bias voltage applied to the light receiving element in the light receiving element array can be set to about 0 to −1V. As a result, the power source can be reduced, and miniaturization and weight reduction can be achieved. As a result, it is possible to obtain a hybrid detection device that is convenient to carry. Although the power source is illustrated only for the light receiving element array of FIG. 4, the above reverse bias voltage range can be applied to all the light receiving element arrays of the present invention. Therefore, in all the embodiments, reduction in size and weight can be achieved.

The depletion layer stops on the p-type region side until it contacts the InP buffer layer 2 having a stepped high concentration, and spreads more greatly on the n-type region side toward the light receiving layer 3 having a low n-type carrier concentration. By making most of the thickness of the light receiving layer 3 of the type 2 type MQW n-type, it is ensured that the electrons generated by light reception pass through the n-type region 6 toward the pixel electrode 11 without escaping. 11 can be moved. For this reason, electrons generated by light reception can be captured more reliably, and quantum efficiency can be improved.
Regarding the manufacturing method, parts other than n-type region 6 are based on the method in the first embodiment. In addition, the n-type region 6 in the present embodiment can be realized by making the temperature selectively higher than the temperature of the selective diffusion of the n-type impurity in the first embodiment for a long time and performing the selective diffusion.

(Embodiment 3)
FIG. 6 is a cross-sectional view showing hybrid detection apparatus 10 according to Embodiment 3 of the present invention. FIG. 7 is a diagram showing the thickness direction distribution of carrier concentration in one light receiving element in the light receiving element array 50. In the first and second embodiments, the pixel is separated from the adjacent pixel by selective diffusion of n-type impurities. The present embodiment is different in that it is separated from an adjacent pixel by providing a groove 19 between the light receiving elements or the pixels. Therefore, in this embodiment, the selective diffusion mask pattern 36 used for selective diffusion is not necessary. The groove 19 is provided by mesa etching the epitaxial layer, but the exposed end face of the epitaxial layer and the like must be protected, and is covered with a protective film 37 as shown in FIG.
In the present embodiment, the pn junction 15 is located at the interface between the type 2 MQW light-receiving layer 3 and the InP buffer layer 2. All epitaxial layer impurities are introduced by doping during film formation. For this reason, as shown in FIG. 7, the carrier concentration in each epitaxial layer has a flat distribution in the thickness direction.

In the present embodiment, the pn junction is located at the boundary between the p ⁻ type light receiving layer 3 and the n ⁺ type buffer layer 2, and the depletion layer mainly spreads in the light receiving layer 3 by applying a reverse bias voltage. For this reason, light incident from the back surface of the InP substrate 1 is received by the depletion layer to which an electric field is applied, and effectively generates electrons. The electrons can reach the n-side electrode 11 with high yield via the MQW conduction band in the n-type region 6. That is, substantially the same effect as in the second embodiment can be obtained.
However, in the present embodiment, the pn junction 15 is formed with a wide flat surface over the entire pixel surrounded by the bottom surface of the groove 19. For this reason, the depletion layer looks like a flat rectangular parallelepiped having a uniform thickness over the pixels. As a result, each pixel receives light in the same depletion layer shape regardless of the incident angle of light. As a result, obliquely angled light can be received at the same position as the front incident light. In the first and second embodiments, the n-type region is a tapered convex curved surface, and the depletion layer is formed along the convex curved surface. Depending on the angle of the incident light, light is received at the slope of the convex curved surface or received at the top of the convex curved surface. For this reason, in the first and second embodiments, the incident angle dependency of the light receiving sensitivity is increased, but in the present embodiment, the incident angle dependency is decreased.

The manufacturing method is different from the first and second embodiments in the following points. Epitaxial layers are sequentially grown on the p ⁺ InP substrate as follows. Impurities are doped during film formation.
p ⁺ InP substrate 1 / p ⁺ InP buffer layer 2 / n ⁻ type 2 type MQW light receiving layer 3 / n ⁺ InGaAs impurity concentration distribution adjustment layer 4 / n ⁺ InP cap layer After the above epitaxial layer is formed, the pixel region A groove 19 is provided between them by selective etching. The selective etching can be performed using an existing selective etching method used for InP-based III-V semiconductors. After the groove 19 is formed, it is covered with a protective film 37 such as polyimide as described above. The formation of the electrodes 11 and 12 can be performed in the same manner as in the first and second embodiments.

(Embodiment 4)
FIG. 8 is a cross-sectional view showing hybrid detection apparatus 10 according to Embodiment 4 of the present invention. In the present embodiment, as in the third embodiment, the impurity is introduced by doping when the epitaxial layer is formed. For this reason, the thickness direction distribution of the carrier concentration in the light receiving element is as shown in FIG. The position of the pn junction 15 is also the boundary between the light receiving layer 3 and the buffer layer 2. However, in this embodiment, the means for separating the light receiving elements is performed by selectively diffusing p-type impurities between the light receiving elements. That is, the n-type region 6d by doping is surrounded by the p-type region 6p by selective diffusion. In the third embodiment, the grooves are separated.

  By separating the light receiving element by selective diffusion of p-type impurities without using a groove, it is possible to eliminate the deterioration of crystallinity associated with mesa etching and the accompanying increase in dark current. However, the tip shape of the p-type region 6p formed by selective diffusion of the p-type impurity becomes important, and if it does not reach the buffer layer 2 reliably, the pn junction 15 continues to the adjacent light receiving element. Because it is formed, consideration is necessary.

(Embodiment 5: Optical sensor device (1)-imaging device or visual field support device-)
FIG. 9 is a diagram illustrating an imaging device or a visual field support device that is the optical sensor device 20 according to Embodiment 3 of the present invention. This visual field support device is mounted on a vehicle in order to support a driver's forward visual field when driving a car at night. The vehicle includes a hybrid detection device 10 that includes the light receiving element array 50 described in the first to fourth embodiments, an optical element such as a CMOS or a lens (not shown), and a display monitor 61 that displays a captured image. A control device 60 that drives and controls them is mounted. FIG. 10 is a diagram showing a nighttime rear view assistance device mounted on a vehicle in order to support the driver's rear view during night driving of an automobile. An image captured by the hybrid detection device 10 including the light receiving element array 50 according to the first to fourth embodiments, the CMOS, the lens, and other optical elements attached to the rear part of the automobile is displayed on the display device in front of the driver. 61. The hybrid detection device 10 or the light receiving element array 50 and the display device 61 are driven and controlled by the control device 60.

  In the vehicular field of view assisting device prior to the present invention, the reflected or emitted light in the infrared region from the object is received and used as an image, so that there are the following problems. When using reflected light, a light source is required, a mounting space is required, and the cost is increased. In addition, when using the radiant heat of an object, it is difficult to recognize a pedestrian or the like wearing a non-heating element other than a person or a cold protection device, so it is necessary to use it together with a recognition means other than an infrared camera. In addition, when using a light source, it is necessary to take measures against the human body, that is, eye-safe measures depending on the wavelength range to be used.

In the visual field support device according to the present embodiment, the extra light source and the eye-safe measures as described above are unnecessary. Moreover, it does not matter whether the imaging target is heated or not. Furthermore, a clear image of the object can be obtained even in an environment containing moisture such as in fog. For this reason, the visual field assistance apparatus for vehicles excellent at night can be provided. This uses a reflected light of SWIR (Short Wavelength Infra-Red) cosmic light from an object, uses a light receiving element having a sufficiently small dark current and an excellent dynamic range (S / N). Because.
The above is a vehicle vision support device, but can also be used for night vision devices, navigation support devices, intruder monitoring devices, indoor monitoring devices, city fire monitoring devices arranged at high positions, and the like.

(Embodiment 6: Optical sensor device (2)-biological component detection device-)
FIG. 11 is a diagram showing a biological component detection device that is the optical sensor device 20 according to Embodiment 6 of the present invention. In FIG. 11, the above-described hybrid detection device 10 is used for the light receiving unit, and concentration measurement is performed using an absorption band located in a long wavelength region of glucose in the near infrared region. In the present embodiment, near-infrared light transmitted through the living body is measured to determine the glucose concentration. You may use the reflected light of a human body. The light takes the following path.
Light source 63 → irradiation fiber 64 → detection site (ubiquitous) → information-equipped optical fiber 65 → diffraction grating (spectrometer) 91 → hybrid detector 10 → control unit 85
The spectroscope may be provided between the light source and the irradiation fiber.
By obtaining the absorption spectrum of the blood component at the detection site, the control unit 85 can detect the absolute value of the blood sugar level, or its relative value or magnitude. The example shown in FIG. 11 receives light transmitted through a human finger, but can obtain information on many biological tissues such as skin, muscle, and blood.

The reference signal is measured by the transmitted light of the reference plate driven by the actuator 67 so that the reference signal is retracted when the living body (finger) is inserted and is inserted when the living body is retracted. Although the thickness of the reference plate depends on the material of the reference plate, it is preferable to make it thin so that the amount of transmitted light is sufficient. The reference plate is moved by the actuator 67 so that variations in position and orientation (angle) do not occur.
The above is an example in which the hybrid type detection device 10 is used for measurement of blood glucose level by transmitted light from the human body. In addition, blood glucose level by body reflected light, body fat, collagen of eye cornea, collagen distribution image of face, etc. It can be used for measurement.

(Embodiment 7: Optical sensor device (3)-moisture detector-)
FIG. 12 is a diagram showing a water detection device (water distribution image forming device for eyes) in a living body which is the optical sensor device 20 according to Embodiment 7 of the present invention. There are many symptoms related to water, such as dry eyes and nasty eyes. When such a symptom appears, as shown in FIG. 12, the symptom can be evaluated by taking a water distribution image of not only the cornea C but also the entire front surface of the eye E. For example, it is possible to detect an abnormally high water concentration at a location corresponding to the lacrimal gland. The concave mirror 68 is preferably one having a high reflectivity for near-infrared light, for example, one made of gold (Au). The concave mirror 68 is positioned not on the front of the eye but on the side so as to reflect light from each part of the eye emitted from the light source 63 and form an image of each part of the eye on the imaging device by the hybrid detection device 10. To do. The filter 69 preferably transmits light in the vicinity of 1.4 μm belonging to the water absorption band or light in the vicinity of 1.9 μm. The microcomputer 85b of the control unit 85 forms a water distribution image in the eye E based on the output signal of the pixel of the imaging device (hybrid type detection device) 10 and displays it on the display device 85c. Since the imaging device 10 according to the present invention has a low dark current and a high sensitivity up to the long wavelength side, a clear water distribution image with a high S / N ratio can be obtained. For this reason, it is useful for understanding the action of water in the eye and the movement of water.

Since the eye reacts very sensitively to light, it is preferable not to use the light source 63 if possible. For example, the peak I of the spectrum of this SWIR cosmic light can be used as the light source. The wavelength of the peak I is in the vicinity of 1.4 μm and belongs to the water absorption band. For this reason, the SWIR space light can be substituted with the exception of the light source 63. Alternatively, even if the artificial light source 63 is used, the light may be limited to the near infrared region, and the peak value may be set to, for example, twice the peak intensity of the SWIR space light. Eye safe is reliably realized by using the SWIR space light as a light source. As described above, the reason why SWIR cosmic light or a light source with a low intensity level can be used is that the dark current of the hybrid detection device 10 constituting the imaging device according to the present embodiment can be reduced. . That is, a clear image can be formed even with a weak signal.
The above is an example of an eye moisture detection device that is a part of a living body, but in addition to this, moisture measurement of natural products (moisture measurement of melon (quality test), measurement of sputum contamination rate by moisture, other fruits, laver, Seafood, dairy products, etc.), corneal moisture measurement in corneal correction surgery, moisture measurement of living body such as facial skin, moisture measurement of paper products, moisture measurement in oil in automatic oil drainer, moisture measurement of sludge dehydrated cake, It can be used for moisture measurement of coal, moisture measurement of clothes in a clothes dryer, and the like.

(Embodiment 8: Optical sensor device (4) -food quality inspection device-)
FIG. 13: is a figure which shows the apparatus which test | inspects the taste which is the food quality inspection apparatus which is the optical sensor apparatus 20 in Embodiment 8 of this invention. In FIG. 13, the rice S to be inspected may be a rice grain, or may be crushed and powdered. The near-infrared light emitted from the light source 63 passes through the rice S, is diffracted by the diffraction grating 91 and dispersed, and is received by the hybrid detection device 10 including the light receiving element array 50, and the electrical signal of the absorption spectrum Become. In the control unit 85 including the microcomputer 85b, the display unit 85c, the input keyboard 85d, the printer 85p, etc., the absorption spectrum is analyzed and displayed. As described above, a signal with a high S / N ratio up to a wavelength of 3000 nm can be obtained by using the hybrid detection device 10 including any one of the light receiving element arrays 50 of the first to fourth embodiments.

In general, quality that depends on the human senses, such as taste, is based on the contribution of multiple components instead of a single component. For this reason, the absorption spectrum of the food in the near infrared region is collected. Separately, the food is subjected to a taste sensory test by a paneler to obtain a sensory evaluation value of the food. A correlation function is set between the sensory evaluation value and the absorbance at a specific wavelength of the absorption spectrum of the food, the first derivative value, or the second derivative value. Based on this correlation function, a sensory evaluation value is obtained from the absorption spectrum of near-infrared light to be inspected, and the taste is inspected. Which wavelength of the absorption spectrum is taken into the correlation function depends on the following factors.
(1) Type of food (2) Receivable range of the taste inspection device, particularly in the range where the S / N ratio is high The above (1) and (2) employ any wavelength of the absorption spectrum depending on the food However, not only that, it also depends on the light receiving wavelength range of the taste inspection apparatus. The following approaches are possible for taste inspection.

FIG. 14 shows absorption spectra in the range of 1200 nm to 2500 nm of rice of _three different brands S ₁ , S ₂ , S ₃ . At wavelengths of 1900 nm and above, the absorbance of each component of rice such as amylose, protein, and moisture is large, and the absorbance of each component can be specified. When the sensory test for taste and the concentration of each component were correlated, it was found that the rice taste (sensory test evaluation) was mainly composed of protein, amylose and water in the rice. The protein, amylose, and water content are determined from the absorbance at wavelengths of 2100 nm, 2130 nm, 2270 nm, and 2370 nm. Therefore, protein, amylose, and moisture can be obtained from the absorbance at these wavelengths, and the taste value can be obtained based on the above correlation function. That is, the taste index = f ₁ (x ₁ ) · f ₂ (x ₂ ) · f ₃ (x ₃ ). Here, _{x 1} is amylose concentration, _{x 2} is the protein concentration, _{x 3} is the water _{concentration, f 1, f 2, f} 3 represents the function. In this approach, the taste of rice is combined with each component of the rice, and the taste is examined from near-infrared spectroscopic data.

According to the above approach, rice taste inspection has become possible with sufficient accuracy from near-infrared spectroscopy. The wavelength range to be adopted can be appropriately selected according to the photoelectric conversion device of the device. For rice, the future challenge is to make it possible to identify branded rice by using as wide a range of wavelengths as possible and using statistical methods to obtain a lot of information. By making it possible to identify the brand rice, it is possible to make it meaningless, for example, acquisition of illegal purposes of trademark rights. In order to realize this, it is possible to use the light receiving element array 50 or the hybrid detection device that can obtain a signal with a high S / N ratio.
The above shows an example of a food quality inspection device for inspecting the taste, but in addition to this, component inspection of growing and shipping fruits, vegetables, etc., detection of beef carcass fat content, detection of abnormal meat, fat It can be used for the detection of PCBs in the inside, the inspection of food heating history, and the like.

(Embodiment 9: Optical sensor device (5) -environment monitor device-)
FIG. 15 is a temperature distribution measuring device for obtaining the temperature distribution of the garbage in the refuse combustion furnace, which is the optical sensor device 20 according to the ninth embodiment of the present invention. It is a specific example of the environment monitor apparatus for detecting the component density | concentration in gas. FIG. 16 is a diagram showing the temperature distribution imaging device 20a. In a refuse combustion furnace, carbon or hydrocarbons are in bulk and do not exist in a form suitable for fuel, so there is little soot and there is a large amount of moisture. FIG. 17 shows the near-infrared spectrum in the refuse combustion furnace, and the emission spectrum wavelengths λ ₂ and λ _{3 of} water are remarkable. In the present embodiment, the concentration and temperature of water are monitored together with the absorption spectrum of water shown in FIG. 18 by utilizing the fact that the emission spectrum of water changes with temperature. In FIG. 18, (K1) and (K2) are measured using 10 mm and 1 mm cuvette cells, respectively. Since the intensity of the emission spectrum is also proportional to the concentration of water, it is difficult to measure with high accuracy using only two emission peak wavelengths, so an absorption spectrum is also used.
In the temperature distribution imaging device 20a, the interference filter 10a is important. The interference filter 10a is a filter having a transmission wavelength at each of the light emission peak wavelengths λ ₂ and λ ₃ and the plurality of absorption peak wavelengths. For example, the absorption peak wavelength has two sharp peaks M2 and M3 in the near infrared region as shown in FIG. 18, but the interference filter 10a allows light of these wavelengths to pass. Therefore, in the interference filter 10a, a total of four types of filters or four transmission wavelength filters are arranged together with the above two emission peak wavelengths. It is desirable to provide an automatic selection mechanism that automatically selects these four types of interference filters by an external operation. It is preferable to provide an autofocus mechanism for automatically focusing the optical system 10c such as a lens. For example, corresponding to the above-described four types of interference filters, imaging of dust or a little above it is performed for light of four wavelengths. As a result, images of four wavelengths can be obtained.
For air with the water vapor temperature and water vapor concentration changed in advance, the light intensity at the above wavelength can be obtained, and the temperature regression equation can be obtained. This temperature regression equation is stored in the microcomputer 85b of the control unit. With the above imaging, the intensity for each wavelength can be obtained at each position. If the above temperature regression equation is used, the temperature can be obtained at each position. In this way, by monitoring both the temperature and concentration of water, it is possible to accurately detect the combustion state of waste.
Conventionally, many temperature sensors have been arranged in the waste incinerator, but the number of temperature sensors can be reduced by arranging the apparatus of the present embodiment at the upper part or top of the incinerator. .

  Although the embodiments and examples of the present invention have been described above, the embodiments and examples of the present invention disclosed above are merely examples, and the scope of the present invention is the implementation of these inventions. It is not limited to the form. The scope of the present invention is indicated by the description of the scope of claims, and further includes meanings equivalent to the description of the scope of claims and all modifications within the scope.

  According to the light receiving element array, the hybrid type detection device, and the like of the present invention, using the type 2 type MQW, the selective diffusion region of the n type impurity is used as a pixel, and high light receiving sensitivity is obtained in the near infrared region by the backside incidence of the InP substrate. It is possible to obtain a light receiving element array, a hybrid detection device, and the like.

1 InP substrate, 2 InP buffer layer, 3 type 2 type MQW light receiving layer, 4 impurity concentration distribution adjusting layer, 5 InP cap layer, 6 n-type region by selective diffusion, 6d n-type region by doping, 6p p by selective diffusion Type region, 10 Hybrid type detection device, 10a, 10c Optical component added to the detection device, 11 n side electrode (pixel electrode), 12 p side electrode (ground electrode), 15 pn junction, 19 groove, 20 Optical sensor device 20a Temperature distribution imaging device, 31 Bond bump, 35 Antireflection film, 36 Selective diffusion mask pattern, 37 Protective film, 50 Light receiving element array, 60 Control device, 61 Display device, 63 Light source, 64 Optical fiber for irradiation, 65 Information Mounted optical fiber, 67 actuator, 69 filter, 70 CMOS (readout circuit), 71 pad (readout power) ), 72 ground electrode, 76 a concave mirror, 85 control unit, 85b microcomputer, 85c display unit, 85d keyboard (input unit), 85p printer, 91 a diffraction grating (spectroscope).

Claims (13)

A light receiving element array in which light receiving elements are arranged,
An InP substrate;
A light-receiving layer having a band gap wavelength of 1.65 μm to 3.0 μm and positioned on the InP substrate;
A cap layer located on the light receiving layer,
The light receiving layer is lattice-matched to the InP substrate, and is a type 2 type multiplex of any one of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb). Consisting of quantum well structure,
Each of the light receiving elements includes an electrode positioned on the cap layer, and the electrode is in ohmic contact with the cap layer in an n-type region provided for each of the light receiving elements. A light receiving element array characterized by being separated by a region which is not n-type.   The n-type region for each of the light receiving elements includes (1) a part of the thickness from the surface of the cap layer, (2) all the thickness of the cap layer, (3) all the thickness of the cap layer, and the light receiving (1) to (4) of a part of the thickness part of the cap layer side of the layer, and (4) all thicknesses of the cap layer and all thicknesses of the light receiving layer The light receiving element array according to claim 1, wherein the light receiving element array has a thickness direction distribution.   The light receiving element array according to claim 1, wherein the cap layer is an InP layer or a composite layer of an InGaAs layer and an InP layer.   The light receiving element has an n-type region formed by selective diffusion or selective ion implantation, and the selective diffusion and non-n-type region or selective ion implantation and a non-n-type region, The light receiving element array according to any one of claims 1 to 3, wherein the light receiving element array is separated from other light receiving elements.   The light receiving element has the cap layer doped with n-type impurities, or the cap layer and light receiving layer doped with n-type impurities, and is separated from an adjacent light receiving element by the groove. The light receiving element array according to any one of claims 1 to 3.   The light-receiving element has the cap layer doped with n-type impurities, or the cap layer and light-receiving layer doped with n-type impurities, and a p-type region selectively diffused with p-type impurities, or p-type impurities 4. The light receiving element array according to claim 1, wherein the light receiving element array is separated from an adjacent light receiving element by a p-type region into which selective ions are implanted. 5. The light-receiving layer includes: (1) an n ⁻ type layer, (2) a p ⁻ type layer, and (3) an impurity composite layer in which the cap layer side is an n ⁻ type layer and the InP substrate side is a p ⁻ type layer. 7. The method according to claim 1, wherein the carrier concentration is any one of (1) to (3), and in any case, the carrier concentration is 1e16 / cm ³ or less. Light receiving element array.   A light receiving element array according to any one of claims 1 to 7 and a silicon readout circuit, wherein an electrode for each light receiving element of the light receiving element array and a readout electrode of the silicon readout circuit are bonded bumps. A hybrid type detection device, wherein the hybrid type detection device is interposed and conductively connected, and light enters from the back surface of the InP substrate in the light receiving element array.   The hybrid detection device according to claim 8, wherein a reverse bias voltage applied to the light receiving elements in the light receiving element array is 0 to -1V.   An optical sensor device using the light receiving element array according to any one of claims 1 to 7, or the hybrid detection device according to claim 8 or 9. An InP substrate;
A light-receiving layer having a band gap wavelength of 1.65 μm to 3.0 μm and positioned on the InP substrate;
A cap layer located on the light receiving layer,
The light receiving layer is lattice-matched to the InP substrate, and is a type 2 type multiplex of any one of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb). Consisting of quantum well structure,
An electrode positioned on the cap layer, the electrode being in ohmic contact with the cap layer in the n-type region;
The light receiving element, wherein the n-type region is separated from a surrounding region by a region that is not n-type. A method of manufacturing a light receiving element array in which light receiving elements are arranged,
Type 2 type multiplexing of any one of (GaAsSb / InGaAs), (GaAsSb / InGaAsN), (GaAsSb / InGaAsNP), and (GaAsSb / InGaAsNSb) lattice-matched on the InP substrate or buffer layer Forming a light-receiving layer having a quantum well structure;
Forming a cap layer on the light receiving layer;
A step of separating the cap layer and the light receiving layer to form the light receiving element separately from an adjacent light receiving element;
In the isolation process, or in the isolation process and the cap layer / light-receiving layer forming step, an n-type region is formed on at least the surface layer of the cap layer for each light-receiving element,
An electrode is formed for each of the light receiving elements in the n-type region, and a method for manufacturing a light receiving element array.   In forming the n-type region for each of the light receiving elements in the isolation process or the isolation process and the cap layer / light receiving layer forming process, (1) selective diffusion or selection from the surface of the cap layer The n-type region is formed by ion implantation, and is separated from other light receiving elements by a region that is not selectively diffused or a region that is not selectively ion implanted. (2) The cap layer or the cap layer and the light receiving layer At the time of formation, an n-type impurity is doped, and a groove is provided by etching between the adjacent light receiving element and separated from the adjacent light receiving element by the groove, or (3) the cap layer or the cap layer When forming the light receiving layer, the n-type impurity is doped, and the p-type impurity is introduced between the adjacent light receiving element by selective diffusion or ion implantation. Wherein the separating the adjacent light receiving elements by the region containing the impurity, the manufacturing method of the light-receiving element array according to claim 12.

Images