JP2013021032A - Sensor element array and manufacturing process of the same, and imaging apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a sensor element array which improves its sensitivity while achieving one bump per one pixel.SOLUTION: A sensor element array comprises: first to fourth pixels 10A to 10D arranged in groups of two in long and width and having a structure where a first active layer 8 for a first wavelength band and a second active layer 9 for a second wavelength band are laminated; a first bump 6A electrically connected to the first active layer of the first pixel 10A and the first active layer of the second pixel 10B; a second bump 6B electrically connected to the second active layer of the second pixel; a third bump 6C electrically connected to the second active layer of the first pixel and the second active layer of the third pixel 10C; and a fourth bump 6D electrically connected to the first active layer of the fourth pixel 10D, in which each of the first to fourth bumps is provided above one of the first to fourth pixels, respectively.

Description

  The present invention relates to a sensor element array, a manufacturing method thereof, and an imaging apparatus.

2. Description of the Related Art Conventionally, there is an image sensor that is provided in an imaging apparatus and has a sensor element array in which a plurality of pixels having a structure in which two active layers for two types of wavelength bands are stacked are two-dimensionally arranged.
For example, an infrared having an infrared focal plane array (IRFPA) in which each pixel is constituted by a quantum well infrared photodetector (two-wavelength QWIP) having a structure having sensitivity in two types of wavelength bands. There is an image sensor.

  For example, as shown in FIG. 21, the two-wavelength QWIP constituting each pixel of the IRFPA 100 is formed by sequentially laminating a lower contact layer 101, a lower active layer 102, an intermediate contact layer 103, an upper active layer 104, and an upper contact layer 105. The structure is Here, the lower active layer 102 is an active layer having sensitivity to one wavelength band (wavelength 2), and the upper active layer 104 is an active layer having sensitivity to another wavelength band (wavelength 1). It is. The lower contact layer 101, the intermediate contact layer 103, and the upper contact layer 105 are n-GaAs layers. The two-wavelength QWIP is also referred to as an infrared detection element.

  Further, three bumps 106 (metal bumps; for example, In bumps) are provided per pixel so as to be connected to each of the three contact layers 101, 103, and 105. The upper and lower active layers 102 and 104 are biased through bumps 106 connected to the intermediate contact layer 103. Further, an output signal from the lower active layer 102 is taken out from the bump 106 connected to the lower contact layer 101, and an output signal from the upper active layer 104 is taken out from the bump 106 connected to the upper contact layer 105. .

  Furthermore, as shown in FIG. 22, the IRFPA 100 having such a two-wavelength QWIP is electrically and mechanically connected to a signal processing circuit chip 107 (for example, Si signal processing circuit chip) by flip chip bonding (FCB). Connected. That is, the two-wavelength QWIP provided as a plurality of pixels in the IRFPA 100 is electrically and mechanically connected to each of the plurality of elements provided in the signal processing circuit chip 107 via the bumps 106 provided for each pixel. Has been.

JP-A-9-107121

By the way, as IRFPA increases in scale (multiple pixels) and the number of pixels increases, it is necessary to make the pixel size smaller, and the bump interval (bump pitch) becomes narrower. Then, as shown in FIG. 23, the bumps 106 are displaced during FCB, and adjacent bumps come into contact with each other, and a defect due to a short circuit is likely to occur.
For this reason, in order to cope with the large scale of IRFPA, it is necessary to reduce the number of bumps per pixel. Although it is conceivable to increase the scale without changing the bump interval (without changing the element size), in this case, since the chip size is increased, various problems as described later arise. So it's not realistic.

  Therefore, as shown in FIG. 24, the lower contact layer 101 is common to all pixels, the intermediate contact layer 103 is divided into upper and lower parts, an insulating layer 108 is provided therebetween, and the intermediate contact layer 103A above the insulating layer 108 is provided. And the lower contact layer 101 are connected by a wiring 109. Then, the lower active layer 102 is biased through the lower contact layer 101 and the upper active layer 104 is biased through the lower contact layer 101, the wiring 109, and the intermediate contact layer 103A. Then, an output signal from the lower active layer 102 is extracted from the bump 106 connected to the intermediate contact layer 103B below the insulating layer 108 via the wiring 110, and the upper active layer is extracted from the bump 106 connected to the upper contact layer 105. The output signal from 104 is taken out. Accordingly, it is only necessary to provide two bumps 106 per pixel without providing a bump for biasing, and the number of bumps per pixel can be reduced. Here, the insulating layer 108 is an i-GaAs layer.

  Further, as shown in FIG. 25, the lower contact layer 101 is common to all the pixels, and the surface side wiring 111 connected to all the pixels is connected to the upper contact layer 105, so that the upper and lower active layers 102, 104 are connected. It may be possible to bias the. In this case, an output signal from the lower active layer 102 and an output signal from the upper active layer 104 are taken out from the bump 106 connected to the intermediate contact layer 103 via the wiring 112. Accordingly, it is only necessary to provide one bump 106 per pixel, and the number of bumps per pixel can be further reduced.

  However, when the number of bumps per pixel is reduced to one in this way, the output signal from the upper active layer 104 and the output signal from the lower active layer 102 are alternately extracted by switching the bias. Therefore, it must be disadvantageous in terms of sensitivity. In other words, only about half of the integration time for taking out the output signals from the upper and lower active layers 102 and 104 can be taken, which is disadvantageous in terms of sensitivity.

  Therefore, it is desirable to improve sensitivity while realizing 1 pixel 1 bump.

  The sensor element array has a structure in which a first active layer for the first wavelength band and a second active layer for the second wavelength band are stacked, and the first to fourth pixels arranged in two vertical and horizontal directions, A first bump electrically connected to a first active layer of one pixel and a first active layer of a second pixel; a second bump electrically connected to a second active layer of a second pixel; A third bump electrically connected to the second active layer of one pixel and the second active layer of the third pixel; and a fourth bump electrically connected to the first active layer of the fourth pixel. The first to fourth bumps are required to be provided one above the first to fourth pixels, respectively.

  The imaging apparatus includes the sensor element array, a signal processing circuit chip connected to the sensor element array via a plurality of bumps including at least the first to fourth bumps, and a control operation connected to the signal processing circuit chip. And the control calculation unit outputs an output signal for two pixels read from the first active layer of the first pixel and the second pixel via the first bump to the second active layer of the second pixel. Distribution based on the ratio of the output signal read from the second active layer of the first pixel and the output signal from the first active layer of the first pixel and the second pixel An output signal from the first active layer is obtained, and an output signal for two pixels read from the second active layer of the first pixel and the third pixel through the third bump is used as the first pixel of the first pixel. The output signal read from the active layer and read from the first active layer of the third pixel The output signal from the second active layer of the first pixel and the output signal from the second active layer of the third pixel are distributed based on the ratio to the output signal generated. And

  In this sensor element array, a plurality of pixels having a structure in which an upper active layer for another wavelength band is stacked above a lower active layer for one wavelength band, and the lower active layer are divided into two pixel sizes, Among the plurality of isolation trenches for dividing the upper active layer into a size for one pixel and the plurality of upper active layers having a size for one pixel, the lower active layer having a size for two pixels extends in the extending direction. A plurality of wirings electrically connecting two upper active layers adjacent to each other in the orthogonal direction and a bump connected to each of a plurality of lower active layers having a size corresponding to two pixels have a size corresponding to two pixels. Above each of the plurality of lower active layers having a thickness, arranged in a direction perpendicular to the direction in which the lower active layer extends and shifted by one pixel from each other and connected to each of the two adjacent upper active layers connected by wiring The bump to be connected A plurality of pixels each provided above each of the plurality of pixels so as to be shifted from each other by one pixel in the extending direction of the lower active layer above each of the two adjacent upper active layers connected by the wiring. It is a requirement to have a bump.

  In the manufacturing method of the sensor element array, an upper active layer for another wavelength band is stacked above a lower active layer for one wavelength band, the lower active layer is divided into two pixel sizes, and the upper active layer is divided. A plurality of separation grooves that are divided into a size for one pixel are formed, and among a plurality of upper active layers having a size for one pixel, the lower active layer having a size for two pixels is orthogonal to the extending direction. A plurality of bumps connected to each of a plurality of lower active layers each having a size corresponding to two pixels are electrically connected to each other by two wirings adjacent to each other in the direction. Bumps that are arranged above each of the lower active layers of the first active layer and are shifted by one pixel in the direction perpendicular to the direction in which the lower active layer extends and connected to each of the two adjacent upper active layers connected by the wiring However, due to wiring It is necessary to form one bump above each pixel above each of two adjacent upper active layers connected to each other so as to be shifted by one pixel from each other in the direction in which the lower active layer extends. And

  Therefore, according to the sensor element array, the manufacturing method thereof, and the imaging device, there is an advantage that the sensitivity can be improved while realizing one pixel and one bump.

(A), (B) is a schematic diagram which shows the structure of the sensor element array of this embodiment. (A), (B) is a schematic diagram which shows the structure of the sensor element array of this embodiment. It is a schematic diagram which shows the structure of the imaging device of this embodiment. (A)-(C) are the figures for demonstrating the process in the control calculating part with which the imaging device of this embodiment is equipped. It is a flowchart which shows the process in the control calculating part with which the imaging device of this embodiment is equipped. It is a figure which shows intensity distribution of the output signal from each pixel in the infrared rays detection element array in which each pixel was comprised by 2 wavelength QWIP. It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in the control calculating part with which the imaging device of this embodiment is provided, when equal light injects into four adjacent pixels. It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in the control calculating part with which the imaging device of this embodiment is provided, when equal light injects into four adjacent pixels. It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in the case of the pixel which does not have incident light in four adjacent pixels in the control calculating part with which the imaging device of this embodiment is equipped. . It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in the case of the pixel which does not have incident light in four adjacent pixels in the control calculating part with which the imaging device of this embodiment is equipped. . It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in case the incident light changes comparatively smoothly in four adjacent pixels in the control calculating part with which the imaging device of this embodiment is provided. It is a figure which shows the example of a calculation of the output signal from each active layer of each pixel in case the incident light changes comparatively smoothly in four adjacent pixels in the control calculating part with which the imaging device of this embodiment is provided. It is a figure for demonstrating the spatial resolution in the outline part of the target in the imaging device of this embodiment. It is a schematic diagram which shows arrangement | positioning of the bump in the case of 1 pixel 3 bump. It is a schematic diagram which shows arrangement | positioning of the bump in the case of 1 pixel 2 bumps. (A)-(C) are the schematic diagrams which show the specific structure of the sensor element array of this embodiment, Comprising: (A) is a top view, (B) follows the XY line of (A). Sectional drawing and (C) are sectional drawings which follow the YZ line of (A). It is a schematic diagram for demonstrating operation | movement of the sensor element array of this embodiment. (A), (B) is a schematic diagram which shows the structure of the modification of the sensor element array of this embodiment. (A), (B) is a schematic diagram which shows the structure of the modification of the sensor element array of this embodiment. (A), (B) is a schematic diagram which shows the structure of the modification of the sensor element array of this embodiment. It is typical sectional drawing which shows the structure of the conventional 1 pixel 3 bump IRFPA. It is a typical sectional view showing flip chip bonding of IRFPA and a signal processing circuit chip. It is typical sectional drawing for demonstrating the subject at the time of flip chip bonding of IRFPA and signal processing circuit chip of the conventional 1 pixel 3 bump. It is a typical sectional view showing composition of IRFPA of 1 pixel 2 bump. It is typical sectional drawing which shows the structure of IRFPA of 1 pixel 1 bump.

Hereinafter, a sensor element array, a manufacturing method thereof, and an imaging apparatus according to an embodiment of the present invention will be described with reference to FIGS.
The imaging device according to the present embodiment is, for example, an infrared imaging device. As shown in FIG. 3, the infrared imaging device 1 includes an infrared image sensor 2, a control calculation unit 3 that performs signal processing and various controls, and a monitor 4 that displays a captured image.

Here, the infrared image sensor 2 includes an infrared detection element array 5 having a plurality of infrared detection elements (pixels), and a signal processing circuit chip 7 connected to the infrared detection element array 5 via bumps 6 and including a signal processing circuit. With.
Among these, the infrared detection element array 5 is a two-dimensional infrared detection element array in which a plurality of pixels 10 constituted by infrared detection elements that generate a photocurrent according to the amount of incident infrared rays are two-dimensionally arranged ( (See FIG. 14).

  The infrared detection element array 5 includes a quantum well infrared detector that uses two multiple quantum well (MQW) layers having sensitivity to different wavelength bands as infrared absorption layers. This is a well-type infrared detection element array. That is, each infrared detection element, that is, each pixel provided in the infrared detection element array 5 is configured by two-wavelength QWIP having sensitivity to two types of wavelength bands. The two-wavelength QWIP has a structure in which a first MQW layer 8 for the first wavelength band and a second MQW layer 9 for the second wavelength band are stacked (see FIG. 14). The first MQW layer 8 is also referred to as a first active layer, a surface side active layer, and an upper active layer. The second MQW layer 9 is also referred to as a second active layer, a substrate side active layer, and a lower active layer. The wavelength included in the first wavelength band is also referred to as wavelength 1, and the wavelength included in the second wavelength band is also referred to as wavelength 2.

  The infrared detection element array 5 includes a sensor element array, a sensor array, a (two-dimensional) detection element array, a (two-dimensional) imaging element array, a (two-dimensional) light-receiving element array, a quantum infrared detection element array, and a two-dimensional element. Also referred to as an array, a (two-dimensional two-wavelength) infrared focal plane array (IRFPA), or a QWIP focal plane array (QWIP-FPA). The infrared detection element is also referred to as a quantum infrared detection element, a light receiving element, a sensor element, a photosensor, a QWIP element, a two-wavelength QWIP element, a two-wavelength element, QWIP, or a two-wavelength QWIP.

  The signal processing circuit chip 7 is provided with a signal processing circuit such as an input circuit that receives an electric signal output from each pixel 10 and amplifies the electric signal, and a reading circuit that reads the output of the input circuit to the outside. Here, the readout circuit is a circuit that sequentially reads out the output voltage corresponding to the amount of current that has flowed through each infrared detection element when infrared rays are incident. This is also called ROIC (readout integrated circuit). Note that the input circuit and the readout circuit may be collectively referred to as a readout circuit. In the present embodiment, the signal processing circuit chip 7 is made of, for example, a Si-based semiconductor material. The signal processing circuit chip 7 is also referred to as a signal processing circuit array or a signal reading circuit chip.

  The infrared detection element array 5 and the signal processing circuit chip 7 are connected via a plurality of bumps 6 (metal bumps; for example, In bumps) provided in each pixel 10. That is, the infrared detection element array 5 and the signal processing circuit chip 7 made of different materials are hybridly connected. Here, each of the plurality of bumps 6 is provided above each pixel 10, and each of the cells in the signal processing circuit chip 7 corresponding to each pixel 10 is electrically connected to each pixel 10 via these bumps 6. It is connected to the. As a result, output signals from the first active layer 8 and the second active layer 9 are extracted to each cell in the signal processing circuit chip 7 via the bumps 6.

  Further, in the present embodiment, the infrared image sensor 2 including the infrared detection element array 5 and the signal processing circuit chip 7 configured as described above is in a vacuum container 11 (detector container) having a window through which infrared rays can enter. Is provided. The infrared image sensor 2 is attached to a cold head 13 connected to a cooling system (refrigerator) 12 to be cooled. That is, the infrared image sensor 2 is a cooling infrared image sensor. A lens 14 is attached to the window of the vacuum vessel 11, and infrared rays from the imaging target (target) are imaged on the infrared detection element array 5 by the lens 14.

  In addition, a control calculation unit 3 is connected to the signal processing circuit chip 7 provided in the infrared image sensor 2 described above. That is, the output wiring 15 is drawn out from the signal processing circuit chip 7 to the outside of the vacuum container 11, and the signal processing circuit chip 7 and the control arithmetic unit 3 (external device) are connected via the output wiring 15. Then, an output signal from the signal processing circuit chip 7 is sent to the control calculation unit 3 and is subjected to signal processing by the control calculation unit 3.

Here, the control calculation unit 3 is configured by a computer or a controller. The control calculation unit 3 includes a drive circuit and a signal processing circuit, outputs power and drive pulses for driving each element included in the infrared detection element array 5, and processes output signals from each element. The image signal is output to the monitor 4.
By the way, in this embodiment, in order to improve sensitivity while realizing 1 pixel 1 bump, the infrared detection element array 5 is configured as follows, and in the control calculation unit 3 according to this configuration, the following is performed. I'm trying to do the right thing.

  First, in this infrared detection element array 5, as shown in FIGS. 1A and 1B, in four adjacent pixels, that is, in the first to fourth pixels 10A to 10D arranged in two vertical and horizontal directions, The first to fourth bumps 6A to 6D are respectively provided above the first to fourth pixels 10A to 10D. That is, the two-wavelength QWIP constituting each pixel 10 of the infrared detection element array 5 is the two-wavelength QWIP of one pixel and one bump.

Here, the four adjacent pixels 10A to 10D are arranged in a square shape. 1A schematically shows a state in which the first active layer 8 for two pixels is electrically connected in the horizontal direction, and FIG. 1B shows two pixels in the vertical direction. The state in which the second active layer 9 is electrically connected is schematically shown, and in reality, these are laminated.
The first bump 6A is electrically connected to the first active layer 8 of the first pixel 10A and the first active layer 8 of the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 of the second pixel 10B and the second active layer 9 of the fourth pixel 10D. The third bump 6C is electrically connected to the second active layer 9 of the first pixel 10A and the second active layer 9 of the third pixel 10C. The fourth bump 6D is electrically connected to the first active layer 8 of the third pixel 10C and the first active layer 8 of the fourth pixel 10D.

  Thus, in the four adjacent first to fourth pixels 10A to 10D, the first active layer 8 is electrically connected to the two pixels of the first pixel 10A and the second pixel 10B in the lateral direction. In addition, two pixels of the third pixel 10C and the fourth pixel 10D are electrically connected in the horizontal direction. The first bump 6A is electrically connected to the first active layer 8 electrically connected to the first pixel 10A and the second pixel 10B, and the third pixel 10C and the fourth pixel 10D are connected to each other. The fourth bump 6D is electrically connected to the first active layer 8 that is electrically connected.

  In the four adjacent first to fourth pixels 10A to 10D, the second active layer 9 is electrically connected to two pixels of the first pixel 10A and the third pixel 10C in the vertical direction, and The two pixels of the second pixel 10B and the fourth pixel 10D are electrically connected in the vertical direction. The third bump 6C is electrically connected to the second active layer 9 electrically connected to the first pixel 10A and the third pixel 10C, and the second pixel 10B and the fourth pixel 10D are connected to each other. The second bump 6B is electrically connected to the second active layer 9 that is electrically connected.

Here, in the adjacent four pixels 10A to 10D, two pixels are electrically connected in the horizontal direction immediately above the two second active layers 9 in which two pixels are electrically connected in the vertical direction. Two first active layers 8 are arranged.
When this is seen as a whole of the infrared detection element array 5, as shown in FIG. 2A, one of the two active layers 8 and 9 for the two wavelength bands is in the horizontal direction. The two pixels are electrically connected to each other in a two-dimensional arrangement. Also, the other second active layer 9 out of the two active layers 8 and 9 for the two types of wavelength bands is electrically connected for two pixels in the vertical direction as shown in FIG. Things are arranged in two dimensions. 2A schematically shows a state in which a plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction are arranged two-dimensionally in the vertical and horizontal directions. ing. FIG. 2B schematically shows a state in which a plurality of second active layers 9 in which two pixels are electrically connected in the vertical direction are arranged two-dimensionally in the vertical and horizontal directions. ing. In practice, they are in a stacked state.

  Then, as shown in FIG. 2A, bumps 6 electrically connected to each of the plurality of first active layers 8 to which two pixels are electrically connected in the horizontal direction are two in the horizontal direction. Above each of the first active layers 8 to which pixels are electrically connected, a plurality of pixels are arranged so as to be shifted from each other by one pixel in the vertical direction, and two pixels are electrically connected in the vertical direction. As shown in FIG. 2B, the bumps 6 electrically connected to each of the second active layers 9 are respectively connected to each of the second active layers 9 in which two pixels are electrically connected in the vertical direction. A plurality of bumps 6 are provided above each pixel 10 so as to be shifted upward by one pixel in the horizontal direction.

Due to such a structure, an output signal read out from each of the plurality of active layers 8 and 9 electrically connected to two pixels through one bump 6 is an output signal for two pixels. It becomes.
That is, in the four adjacent first to fourth pixels 10A to 10D, output signals for two pixels from the first active layer 8 of the first pixel 10A and the second pixel 10B are output via the first bump 6A. Will be. Further, output signals for two pixels from the second active layer 9 of the second pixel 10B and the fourth pixel 10D are output via the second bump 6B. Further, output signals for two pixels from the second active layer 9 of the first pixel 10A and the third pixel 10C are output via the third bump 6C. In addition, output signals for two pixels from the first active layer 8 of the third pixel 10C and the fourth pixel 10D are output via the fourth bump 6D.

For this reason, in accordance with such a configuration, the control arithmetic unit 3 performs the following processing.
That is, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A to the first pixel 10A and the third pixel 10A. Based on the ratio of the output signal read from the second active layer 9 of the pixel 10C and the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D, the first signal is distributed. An output signal from the first active layer 8 of the pixel 10A and an output signal from the first active layer 8 of the second pixel 10B are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D via the second bump 6B to the first pixel 10A and the second pixel 10D. Based on the ratio between the output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D, the second signal is distributed. An output signal from the second active layer 9 of the pixel 10B and an output signal from the second active layer 9 of the fourth pixel 10D are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C via the third bump 6C to the first pixel 10A and the second pixel 10C. Based on the ratio between the output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D, the first signal is distributed. An output signal from the second active layer 9 of the pixel 10A and an output signal from the second active layer 9 of the third pixel 10C are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D via the fourth bump 6D to the first pixel 10A and the third pixel 10D. Based on the ratio between the output signal read from the second active layer 9 of the pixel 10C and the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D, the third signal is distributed. An output signal from the first active layer 8 of the pixel 10C and an output signal from the first active layer 8 of the fourth pixel 10D are obtained.

In this way, the output signal for two pixels from one active layer, that is, the output signal for two pixels in one wavelength band, is supplied to the other two active layers provided at positions corresponding to the one active layer. Output signals from the active layers 8 and 9 of each pixel 10 are obtained by distributing the output signals for two pixels from the pixel, that is, the ratio of the output signals for two pixels in the other two wavelength bands. ing.
That is, an output signal for two pixels read from the first active layer 8 of the two electrically connected pixels 10 through the bump 6 is provided at a position below one side of the first active layer 8. Based on the ratio between the output signal read from the second active layer 9 and the output signal read from the second active layer 9 provided at the lower position on the other side of the first active layer 8 The output signal from the first active layer 8 of one of the two pixels 10 and the output signal from the first active layer 8 of the other pixel 10 are obtained by distribution.

  Similarly, an output signal for two pixels read out from the second active layer 9 of the two electrically connected pixels 10 via the bumps 6 is placed at a position above one side of the second active layer 9. Based on the ratio between the output signal read from the provided first active layer 8 and the output signal read from the first active layer 8 provided at the upper position on the other side of the second active layer. The output signal from the second active layer 9 of one of the two pixels 10 and the output signal from the second active layer 9 of the other pixel 10 are obtained by distribution.

  For example, as shown in FIG. 4A, output signals (outputs; signal strengths) from adjacent first to fourth pixels 10A to 10D are UL, UR, DL, and DR, respectively. That is, the output signals from the first active layer 8 of the adjacent first to fourth pixels 10A to 10D are UL1, UR1, DL1, and DR1, respectively, and the output signals from the second active layer 9 are UL2 respectively. , UR2, DL2, DR2. Further, as shown in FIG. 4B, the output signal from the first active layer 8 of the first pixel 10A and the second pixel 10B that are electrically connected is U, and the electrically connected first pixel 10A and the second pixel 10B are electrically connected. Assume that D is an output signal from the first active layer 8 of the three pixels 10C and the fourth pixel 10D. Also, as shown in FIG. 4C, the output signal from the second active layer 9 of the first pixel 10A and the third pixel 10C that are electrically connected is L, and the electrically connected first pixel 10A and the third pixel 10C are electrically connected. The output signals from the second active layer 9 of the two pixels 10B and the fourth pixel 10D are R.

  In this case, the output signal U from the first active layer 8 of the first pixel 10A and the second pixel 10B, the output signal L from the second active layer 9 of the first pixel 10A and the third pixel 10C, and the second pixel The output signal UL1 from the first active layer 8 of the first pixel 10A and the second pixel 10B are distributed at a ratio (L: R) of 10B and the output signal R from the second active layer 9 of the fourth pixel 10D. The output signal UR1 from the first active layer 8 is obtained.

  Further, the output signal D from the first active layer 8 of the third pixel 10C and the fourth pixel 10D, the output signal L from the second active layer 9 of the first pixel 10A and the third pixel 10C, and the second pixel 10B. And the output signal R1 from the second active layer 9 of the fourth pixel 10D (L: R), and the output signal DL1 from the first active layer 8 of the third pixel 10C and the fourth pixel 10D. An output signal DR1 from the first active layer 8 is obtained.

  Further, the output signal L from the second active layer 9 of the first pixel 10A and the third pixel 10C, the output signal U from the first active layer 8 of the first pixel 10A and the second pixel 10B, and the third pixel 10C. And the output signal D from the first active layer 8 of the fourth pixel 10D (U: D), and the output signal UL2 from the second active layer 9 of the first pixel 10A and the third pixel 10C An output signal DL2 from the second active layer 9 is obtained.

  Further, the output signal R from the second active layer 9 of the second pixel 10B and the fourth pixel 10D, the output signal U from the first active layer 8 of the first pixel 10A and the second pixel 10B, and the third pixel 10C. And the output signal D from the first active layer 8 of the fourth pixel 10D (U: D), and the output signal UR2 from the second active layer 9 of the second pixel 10B and the fourth pixel 10D. An output signal DR2 from the second active layer 9 is obtained.

  In this way, the control calculation unit 3 outputs two pixels that are respectively read out through the first to fourth bumps 6A to 6D provided one above the first to fourth pixels 10A to 10D. By distributing the signals, output signals from the first and second active layers 8 and 9 of the first to fourth pixels 10A to 10D are obtained. In other words, by distributing (dividing / distributing) the output signals for two pixels that are read out via the bumps 6 provided one above each pixel 10, the first and second activations of each pixel 10 are obtained. The output signals from the layers 8 and 9, that is, the output signals of the respective wavelengths of the respective pixels 10 are obtained.

Specifically, the control calculation unit 3 performs signal processing (calculation processing) in the following procedure.
That is, as shown in FIG. 5, first, the control calculation unit 3 starts from the first and second active layers 8 and 9 of the two pixels 10 electrically connected via the bumps 6, respectively. An output signal for two pixels is read (step S10).

  For example, in the four adjacent first to fourth pixels 10A to 10D, the control calculation unit 3 has two pixels from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A. Read the output signal. In addition, the control calculation unit 3 reads output signals for two pixels from the second active layer 9 of the second pixel 10B and the fourth pixel 10D via the second bump 6B. In addition, the control calculation unit 3 reads output signals for two pixels from the second active layer 9 of the first pixel 10A and the third pixel 10C via the third bump 6C. In addition, the control calculation unit 3 reads output signals for two pixels from the first active layer 8 of the third pixel 10C and the fourth pixel 10D via the fourth bump 6D.

Next, the control calculation unit 3 calculates a ratio (division ratio) for distributing the read output signals for two pixels (step S20).
For example, in the four adjacent first to fourth pixels 10A to 10D, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B. As the distribution ratio, the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D and the output signal read from the second active layer 9 of the first pixel 10A and the third pixel 10C. And calculate the ratio.

Further, the control calculation unit 3 distributes the output signals for the two pixels read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D as the ratio of the first pixel 10A and the second pixel 10B. A ratio between the output signal read from the first active layer 8 and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D is calculated.
In addition, the control calculation unit 3 distributes the output signals for the two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C as the ratio of the first pixel 10A and the second pixel 10B. A ratio between the output signal read from the first active layer 8 and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D is calculated.

In addition, the control calculation unit 3 determines the ratio of the output signals for the two pixels read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D as the ratio of the first pixel 10A and the third pixel 10C. The ratio between the output signal read from the second active layer 9 and the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D is calculated.
As described above, the control arithmetic unit 3 determines the ratio of the output signals for the two pixels read from the first active layer 8 of the two electrically connected pixels 10 as the ratio of the first active layer 8. The output signal read from the second active layer 9 provided at the lower position on one side and the second active layer 9 provided at the lower position on the other side of the first active layer 8 are read. Calculate the ratio to the output signal.

Similarly, the control calculation unit 3 uses one of the second active layers 9 as a ratio for distributing output signals for two pixels read from the second active layer 9 of the two pixels 10 that are electrically connected. The output signal read from the first active layer 8 provided at a position above the first active layer 8 and the output signal read from the first active layer 8 provided at a position above the second active layer 9 on the other side Calculate the ratio to the output signal.
Next, the control calculation unit 3 divides the read output signal for two pixels based on the ratio obtained as described above, and the first of the two pixels 10 that are electrically connected to each other. Then, the output signals for one pixel, that is, the output signals from the first and second active layers 8 and 9 of each pixel 10 are obtained by allocating to the second active layers 8 and 9 (step S30).

  For example, in the four adjacent first to fourth pixels 10A to 10D, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B. The ratio of the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D to the output signal read from the second active layer 9 of the first pixel 10A and the third pixel 10C. Based on this, an output signal from the first active layer 8 of the first pixel 10A and an output signal from the first active layer 8 of the second pixel 10B are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D to the first active layer of the first pixel 10A and the second pixel 10B. The second active layer of the second pixel 10B is distributed based on the ratio of the output signal read from the output signal 8 and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D. 9 and the output signal from the second active layer 9 of the fourth pixel 10D.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C to the first active layer of the first pixel 10A and the second pixel 10B. The second active layer of the first pixel 10A is distributed based on the ratio of the output signal read from the output signal 8 and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D. 9 and the output signal from the second active layer 9 of the third pixel 10C.

  In addition, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D to the second active layer of the first pixel 10A and the third pixel 10C. The first active layer of the third pixel 10C is distributed based on the ratio between the output signal read from the output signal 9 and the output signal read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D. 8 and the output signal from the first active layer 8 of the fourth pixel 10D.

  As described above, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the two electrically connected pixels 10 on one side of the first active layer 8. An output signal read from the second active layer 9 provided at a lower position and an output signal read from the second active layer 9 provided at a lower position on the other side of the first active layer 8 And the output signal from the first active layer 8 of one of the two pixels 10 and the output signal from the first active layer 8 of the other pixel 10 are obtained.

  Similarly, output signals for two pixels read from the second active layer 9 of the two pixels 10 that are electrically connected are provided at a position above one side of the second active layer 9. The distribution is based on the ratio between the output signal read from the first active layer 8 and the output signal read from the first active layer 8 provided at the lower position on the other side of the second active layer 9. Thus, an output signal from the second active layer 9 of one of the two pixels 10 and an output signal from the second active layer 9 of the other pixel 10 are obtained.

Next, the control calculation unit 3 outputs a signal to the monitor 4 based on the output signals from the first and second active layers 8 and 9 of each pixel 10 obtained as described above, and the monitor 4 The image is displayed on (Step S40).
Here, in the case of the infrared detection element array 5 in which each pixel 10 is configured by two wavelengths QWIP, when looking at the same target, the intensity distribution of the output signal from each pixel 10 is as shown in FIG. The signal strengths of the output signals from the active layers 8 and 9, that is, the signal strengths of the wavelengths 1 and 2 are proportional to each other.

Therefore, as described above, as the ratio when distributing the output signals for two pixels read from one active layer of the two pixels 10 that are electrically connected to the output signals of each pixel 10, The ratio of output signals read from the other active layer of the pixel 10 is used (see FIG. 4).
In this case, when the difference between the output signals (outputs; signal strength) between the four adjacent pixels 10A to 10D is small, that is, when the change of the output signal is small, distribution can be performed without a large error as follows. It is.

First, a case where uniform light is incident on four adjacent pixels 10A to 10D will be described with reference to FIGS.
When equal light enters four adjacent pixels, as shown in the left half of FIG. 7, output signals UL1, UR1, DL1 from the first active layer 8 (upper layer; wavelength 1) of each pixel 10A to 10D. , DR1 are equal and the output signals UL2, UR2, DL2, DR2 from the second active layer 9 (lower layer; wavelength 2) should be equal. Here, values of “1”, “1”, “1”, and “1” are obtained as UL1, UR1, DL1, and DR1, respectively, and values of “0.5” are obtained as UL2, UR2, DL2, and DR2, respectively. "," 0.5 "," 0.5 ", and" 0.5 "are obtained.

  On the other hand, when configured as described above (see FIG. 4), in the four adjacent pixels 10A to 10D, the first active layer 8 of the first pixel 10A and the second pixel 10B is electrically connected in the lateral direction. Has been. Therefore, the value (UL1 + UR1) obtained by adding the output UL1 from the first active layer 8 of the first pixel 10A and the output UR1 from the first active layer 8 of the second pixel 10B is the first pixel 10A and the second pixel. As an output U from the first active layer 8 of 10B. Here, as shown in the right half of FIG. 7, since UL1 and UR1 both have a value of “1”, a value of “2” is obtained as the output U.

  Similarly, the first active layer 8 of the third pixel 10C and the fourth pixel 10D is electrically connected in the lateral direction. Therefore, the value (DL1 + DR1) obtained by adding the output DL1 from the first active layer 8 of the third pixel 10C and the output DR1 from the first active layer 8 of the fourth pixel 10D is the third pixel 10C and the fourth pixel. As an output D from the first active layer 8 of 10D. Here, as shown in the right half of FIG. 7, DL1 and DR1 both have a value of “1”, so that a value of “2” is obtained as the output D.

  The second active layer 9 of the first pixel 10A and the third pixel 10C is electrically connected in the vertical direction. Therefore, a value (UL2 + DL2) obtained by adding the output UL2 from the second active layer 9 of the first pixel 10A and the output DL2 from the second active layer 9 of the third pixel 10C is the first pixel 10A and the third pixel. As an output L from the second active layer 9 of 10C. Here, as shown in the right half of FIG. 7, UL2 and DL2 both have a value of “0.5”, and thus a value of “1” is obtained as the output L.

  Similarly, the second active layer 9 of the second pixel 10B and the fourth pixel 10D is electrically connected in the vertical direction. Therefore, a value (UR2 + DR2) obtained by adding the output UR2 from the second active layer 9 of the second pixel 10B and the output DR2 from the second active layer 9 of the fourth pixel 10D is the second pixel 10B and the fourth pixel. It is obtained as an output R from the 10D second active layer 9. Here, as shown in the right half of FIG. 7, since both UR2 and DR2 have a value of “0.5”, a value of “1” is obtained as the output R.

Therefore, the value “2” obtained as the output U is distributed in a ratio (1: 1) between the value “1” obtained as the output L and the value “1” obtained as the output R. Thus, a value of “1” is obtained as the output UL1 from the first active layer 8 of the first pixel 10A and the output UR1 from the first active layer 8 of the second pixel 10B.
Similarly, the value “2” obtained as the output D is distributed in a ratio (1: 1) between the value “1” obtained as the output L and the value “1” obtained as the output R. Thus, a value of “1” is obtained as the output DL1 from the first active layer 8 of the third pixel 10C and the output DR1 from the first active layer 8 of the fourth pixel 10D.

Also, the value “1” obtained as the output L is distributed by the ratio (1: 1) between the value “2” obtained as the output U and the value “2” obtained as the output D. A value of “0.5” is obtained as the output UL2 from the second active layer 9 of the first pixel 10A and the output DL2 from the second active layer 9 of the third pixel 10C.
Similarly, the value “1” obtained as the output R is distributed in a ratio (1: 1) between the value “2” obtained as the output U and the value “2” obtained as the output D. Thus, a value of “0.5” is obtained as the output UR2 from the second active layer 9 of the second pixel 10B and the output DR2 from the second active layer 9 of the fourth pixel 10D.

As described above, when uniform light is incident on the four adjacent pixels 10A to 10D, the output of two pixels can be accurately distributed, and an output that should originally be obtained can be obtained.
Here, FIG. 8 is a graph of this result. In FIG. 8, the horizontal axis represents the value of the output signal that should be originally obtained from the four adjacent pixels 10 </ b> A to 10 </ b> D, and the vertical axis represents the value of the output signal obtained by the arithmetic processing as described above. Yes.

In the above-described case, both coincide with each other, and therefore, the slope is on a straight line having a slope of “1”.
Next, a case where there is a pixel without incident light (a pixel whose output signal value is “0”) among the four adjacent pixels 10 </ b> A to 10 </ b> D will be described with reference to FIGS. 9 and 10.
When there is a pixel without incident light among the four adjacent pixels 10A to 10D, output signals (output; signal intensity) from the active layers 8 and 9 of each pixel 10 are as shown in the left half of FIG. Should be. Here, values of “1”, “0”, “0”, and “1” are obtained as UL1, UR1, DL1, and DR1, respectively, and values of “0.5” are obtained as UL2, UR2, DL2, and DR2, respectively. ”,“ 0 ”,“ 0 ”,“ 0.5 ”.

  However, when configured as described above (see FIG. 4), the value of the output signal obtained by the arithmetic processing as described above is adjusted by the four pixels 10A to 10D as shown in the right half of FIG. Therefore, it becomes a uniform value. Here, values of “0.5”, “0.5”, “0.5”, and “0.5” are obtained as UL1, UR1, DL1, and DR1, respectively, and UL2, UR2, DL2 And DR2, values of “0.25”, “0.25”, “0.25”, and “0.25” are obtained, respectively.

For this reason, when this is graphed, as shown in FIG. 10, it does not fall on a straight line with an inclination of “1”.
Next, a case where incident light changes relatively smoothly in the four adjacent pixels 10A to 10D will be described with reference to FIGS.
When incident light changes relatively smoothly in the four adjacent pixels 10A to 10D, the output signals (outputs; signal strength) from the active layers 8 and 9 of each pixel 10 are as shown in the left half of FIG. Should be. Here, values of “1”, “0.8”, “0.5”, and “0.3” are obtained as UL1, UR1, DL1, and DR1, respectively, and as UL2, UR2, DL2, and DR2, The values “0.5”, “0.4”, “0.25”, and “0.15” are obtained, respectively.

  However, when configured as described above (see FIG. 4), the value of the output signal obtained by the arithmetic processing as described above is relatively good as shown in the right half of FIG. Here, values of “1.0385”, “0.7615”, “0.4615”, “0.3385” are obtained as UL1, UR1, DL1, DR1, respectively, and UL2, UR2, DL2, DR2 are obtained. As a result, values of “0.5192”, “0.3808”, “0.2308”, and “0.1692” are obtained, respectively.

For this reason, when this is graphed, as shown in FIG. 12, it almost falls on a straight line having an inclination of “1”.
Therefore, for example, when the target as shown in FIG. 13 is viewed, for example, the contour line may appear to spread over four pixels. However, a good approximate value can be obtained in a region where other signals do not vary greatly.

In this way, by using one bump per pixel, making the pixel size ¼, and increasing the number of pixels to four times the number of pixels, the spatial resolution (resolution) does not reach four times, but the signal is As long as the smoothly changing area is viewed, the area can be increased by about 4 times, and the spatial resolution before the multi-pixel conversion is compensated for in the outline portion.
Moreover, 1 pixel 1 bump is realizable by comprising as mentioned above. Also, by using 1 bump per pixel, with the same bump pitch as in the case of 3 bumps per pixel (see FIG. 14), the pixel size is reduced to 1/4, and the number of pixels is quadrupled without increasing the chip size. Pixelization is possible. Further, even if one pixel has one bump, it is not necessary to alternately read out the signal from the first active layer 8 and the signal from the second active layer 9 by switching the bias, so that the sensitivity can be improved.

  That is, in the case of 1 pixel 1 bump, if the bump pitch (bump center distance) is the same, the pixel size can be reduced to 1/4 compared to the case of 1 pixel 3 bump (see FIG. 14). . Thereby, the number of pixels can be quadrupled with the infrared detecting element array 5 of the same size. In addition, the pixel size can be halved as compared with the case of one pixel and two bumps (see FIG. 15), thereby doubling the number of pixels with the infrared detecting element array 5 of the same size. Can do.

However, even if the number of pixels can be quadrupled with the same size of the infrared detection element array 5, the spatial resolution decreases in a region where the output signal difference is large, such as near the boundary line of the target. The spatial resolution is not 4 times as it is throughout. However, in the region where the difference between the output signals is small, the spatial resolution is almost quadrupled.
Further, when one pixel is set to one bump and the pixel size is reduced to ¼ to increase the number of pixels by four times, the chip size remains small and the bump pitch does not change. Has a great advantage in manufacturing. That is, if the number of pixels is increased by a factor of 4 without changing the pixel size, the chip size increases by a factor of 2 in length and a factor of 4 in area. For this reason, there are disadvantages in the manufacturing process such as a reduction in the number of chips in the wafer and a limitation on the shot size of photolithography. FCB also increases the difficulty, such as an increase in required weight and a decrease in tolerance for tilt. Considering these points, even if some spatial resolution is sacrificed, by reducing the number of bumps to one and reducing the pixel size to 1/4, the chip size is not increased and the bump pitch is not changed. It can be said that it is more advantageous to increase the number of pixels by four times.

  For example, in the infrared detection element array 5 in which 256 pixels having a pixel size of about 40 μm x about 40 μm are arranged vertically and 256 horizontally, the minimum pitch of the bumps is about 20 μm in the case of 3 bumps per pixel. The size of the bump pitch is a number close to the limit in consideration of performing FCB. In addition, in the infrared detecting element array 5 in which 512 pixels are arranged vertically and 512 pixels are arranged without changing the chip size and the number of pixels is quadrupled, the bump pitch is further reduced, which is almost impossible to perform FCB. Possible number. On the other hand, without changing the pixel size, that is, without changing the bump pitch, if 512 pixels are arranged vertically and 512 pixels are arranged in a multi-pixel configuration, the area of the effective pixel portion alone is about 20.5 mm square, When the pixel peripheral part is added, it does not become smaller than about 22 mm square. This size exceeds the limit of one shot depending on the exposure apparatus. Even if exposure is possible, since the chip size is large, the number of chips that can be formed in the wafer is limited. Further, when the chip size is doubled in length, the amount of inclination allowed at the time of FCB alignment is halved, and very strict alignment accuracy is required. For this reason, as described above, the number of bumps per pixel is reduced to one and the pixel size is reduced (1/2 in length and 1/4 in area), so that the chip size is not increased. It is advantageous to increase the number of pixels by four times without changing the bump pitch.

Next, a specific structure (element structure) of the infrared detection element array 5 of the present embodiment will be described with reference to FIG.
In the present embodiment, as shown in FIG. 16, the infrared detection element array 5 includes a plurality of pixels 10 (infrared detection elements) arranged two-dimensionally and a plurality of separation grooves 20 and 21 that separate the plurality of pixels 10. With.

  Here, each pixel 10 includes, on a semiconductor substrate (not shown), a lower contact layer 22, a second MQW layer (lower active layer) 9 that absorbs infrared light in one wavelength band, an intermediate contact layer 23, and other wavelengths. It has a structure in which a first MQW layer (upper active layer) 8 and an upper contact layer 24 that absorb the infrared rays of the band are sequentially laminated. That is, the upper active layer 8 for the other wavelength band is stacked above the lower active layer 9 for the one wavelength band. Further, a passivation film (insulating film) 25 is formed so as to cover the entire surface of the semiconductor multilayer structure.

For example, the semiconductor substrate is a GaAs substrate. The lower contact layer 22, the intermediate contact layer 23, and the upper contact layer 24 are GaAs contact layers. That is, in the present embodiment, the infrared detection element array 5 is formed of, for example, a GaAs semiconductor material. The passivation film 25 is a SiN film or the like.
Further, bumps 6 (metal bumps; bump electrodes; here, In bumps) are formed above each pixel 10 on the surface side of each pixel 10 with a passivation film 25 interposed therebetween.

As the plurality of separation grooves 20, 21, a plurality of separation grooves 21 (first separation grooves) extending in the vertical direction and a plurality of separation grooves 20 (second separation grooves) extending in the lateral direction perpendicular thereto are provided. Yes. The vertical direction is also referred to as a direction in which the lower active layer 9 extends, and the horizontal direction is also referred to as a direction orthogonal to the direction in which the lower active layer 9 extends.
The vertical separation grooves 21 are provided in parallel with an interval of one pixel. The vertical separation groove 21 has a depth that divides up to the lower contact layer 22. That is, the vertical separation groove 21 has a depth that divides both the stacked two active layers 8 and 9.

  Further, the horizontal separation grooves 20 are provided in parallel with an interval of one pixel. As the lateral separation grooves 20, separation grooves 20 </ b> A having a depth that divides up to the lower contact layer 22 and separation grooves 20 </ b> B having a depth reaching the surface of the intermediate contact layer 23 are alternately provided. That is, as the lateral separation groove 20, the separation groove 20 </ b> A having a depth that divides both of the two stacked active layers 8 and 9 and one of the two stacked active layers 8 and 9. Separation grooves 20B having a depth to divide only the film are provided alternately.

  By providing such a vertical separation groove 21 and a horizontal separation groove 20, the first MQW layer as the lower active layer 9 is divided into the size of two pixels, and the second MQW as the upper active layer 8 is divided. The layer is divided into the size of one pixel. That is, a plurality of separation grooves 20 and 21 are provided that divide the lower active layer 9 into two pixels and divide the upper active layer 8 into one pixel.

As a result, a plurality of lower active layers 9 extending in the vertical direction and having a size corresponding to two pixels and a plurality of upper active layers 8 having a size corresponding to one pixel are provided.
In this case, the upper contact layer 24 is divided into a size corresponding to one pixel and becomes a plurality of upper contact layers 24 having a size corresponding to one pixel. Further, the intermediate contact layer 23 is divided into a size corresponding to two pixels, and becomes a plurality of intermediate contact layers 23 having a size corresponding to two pixels.

Here, the lower contact layer 22 is divided into upper and lower portions, and an insulating layer 26 is provided therebetween.
In this case, the lower contact layer (second lower contact layer) 22A on the upper side of the insulating layer 26 is divided into a size corresponding to two pixels to form a plurality of second lower contact layers 22A having a size corresponding to two pixels.

  On the other hand, the lower contact layer (first lower contact layer) 22B on the lower side of the insulating layer 26 is not divided by the isolation grooves 20 and 21, and is connected to all the pixels 10. That is, the first lower contact layer 22B is a contact layer (common electrode layer) common to all the pixels 10. The first lower contact layer 22B extends from each pixel 10 to the device outer periphery, and is connected to a bias supply unit of the signal processing circuit chip 7 provided on the device outer periphery through a bump. The first lower contact layer 22B is electrically connected to the intermediate contact layer 23 of each pixel 10 via a wiring 27 (including a pad 28; a lead wiring). Accordingly, a bias voltage is applied to the upper and lower active layers 8 and 9 provided in each pixel 10 via the first lower contact layer 22B, the wiring 27 (including the pad), and the intermediate contact layer 23. It is like that. The wiring 27 is also referred to as metal wiring or wiring metal. The pad 28 is also referred to as a contact metal or a contact electrode.

  Specifically, in the four first to fourth pixels 10A to 10D arranged adjacent to each other in a square shape, corners facing diagonally, that is, the separation grooves 20 and 21 of the first pixel 10A. Wirings 27 (including pads 28) for applying a bias voltage are provided at corners facing the separation grooves 20 and 21 of the first and fourth pixels 10D. That is, the intermediate contact layer exposed by providing the contact holes 33 at the corners facing the separation grooves 20 and 21 of the first pixel 10A and the corners facing the separation grooves 20 and 21 of the fourth pixel 10D. A pad 28 is provided on the surface 23. A pad 28 is provided on the bottom surface of the portion where the separation grooves 20 and 21 in the vicinity of these corners intersect, that is, on the surface of the first lower contact layer 22B exposed by the opening of the insulating layer 26. . These pads 28 are electrically connected by wiring 27 provided along these corners with the passivation film 25 interposed therebetween.

  However, the present invention is not limited to this, and wiring may be formed on the surface side so that all of the plurality of pixels 10 are connected. A bias voltage is applied to the upper active layer 8 provided on the upper side of each pixel 10 via a metal wiring, while the lower contact layer 22 is applied to the lower active layer 9 provided on the lower side of each pixel 10. A bias voltage may be applied via the. The wiring is also referred to as surface wiring, metal wiring, or wiring metal.

  Further, as described above, since the upper active layer 8 has a size corresponding to one pixel, two upper portions adjacent to each other in the horizontal direction among the plurality of upper active layers 8 having a size corresponding to one pixel. The active layers 8 are electrically connected to each other by wirings 29 (including pads 30). The horizontal direction is also referred to as a direction orthogonal to the direction in which the lower active layer 9 having a size for two pixels extends. The wiring 29 is also referred to as metal wiring, surface wiring, or wiring metal. The pad 30 is also referred to as a contact metal or a contact electrode.

  In this case, in the four adjacent first to fourth pixels 10A to 10D, the two lower active layers 9 extending in the vertical direction and having the size of two pixels are respectively the first pixel 10A and the third pixel 10C. These are an active layer that electrically connects the two pixels, and an active layer that electrically connects the two pixels of the second pixel 10B and the fourth pixel 10D. The bumps are electrically connected to the lower active layer 9 via the lower contact layer 22 and the wiring 31 (including the pad 32), and are provided above the second pixel 10B and the third pixel 10C. 6 are the second bump 6B and the third bump 6C, respectively. The wiring 31 is also referred to as metal wiring or wiring metal. The pad 32 is also referred to as a contact metal or a contact electrode.

  In the four adjacent first to fourth pixels 10A to 10D, two upper active layers 8 that are adjacent in the horizontal direction and connected by the wiring 29 (including the pad 30) are connected to the first pixel 10A and the second pixel 10A. The active layer electrically connects two pixels of the pixel 10B and the active layer electrically connects two pixels of the third pixel 10C and the fourth pixel 10D. The bumps are electrically connected to the upper active layer 8 via the upper contact layer 24 and the wiring 29 (including the pad 30), and are provided above the first pixel 10A and the fourth pixel 10D. 6 are the first bump 6A and the fourth bump 6D, respectively.

  Specifically, in the four first to fourth pixels 10A to 10D arranged adjacent to each other in a square shape, corners facing diagonally, that is, the separation grooves 20 and 21 of the second pixel 10B. And a wiring 31 (pad 32 for drawing a signal to the surface side from the lower active layer 9 having a size of two pixels, respectively, at corners facing the separation grooves 20 and 21 of the first pixel 10C and the third pixel 10C. Including: lead-out wiring). That is, the second lower portion exposed by providing the contact holes 34 at the corners facing the separation grooves 20 and 21 of the second pixel 10B and the corners facing the separation grooves 20 and 21 of the third pixel 10C. A pad 32 is provided on the surface of the contact layer 22A. These pads 32 are respectively connected to the second bumps 6B and the third pixels 10C provided above the second pixels 10B by the wirings 31 provided along the corners with the passivation film 25 interposed therebetween. It is electrically connected to the third bump 6C provided above.

  Further, in the four first to fourth pixels 10A to 10D arranged adjacent to each other in a square shape, the upper active layer 8 and the first active layer 8 of the first pixel 10A adjacent to each other in the lateral direction and separated by the separation groove 21 are arranged. A wiring 29 (including a pad 30) for electrically connecting the upper active layer 8 of the two pixels 10B is provided above the first pixel 10A, the second pixel 10B, and the separation groove 21 between them. Yes. That is, the pads 30 are provided on the surfaces of the upper contact layer 24 of the first pixel 10A and the upper contact layer 24 of the second pixel 10B. These pads 30 are electrically connected by a wiring 29 provided with the passivation film 25 interposed therebetween. The wiring 29 is electrically connected to the first bump 6A provided above the first pixel 10A.

  Similarly, in the four first to fourth pixels 10 </ b> A to 10 </ b> D arranged adjacent to each other in a square shape, the upper active layer 8 of the third pixel 10 </ b> C adjacent in the lateral direction and separated by the separation groove 21 A wiring 29 (including a pad 30) for electrically connecting the upper active layer 8 of the fourth pixel 10D is provided above the third pixel 10C, the fourth pixel 10D, and the separation groove 21 between them. ing. That is, the pads 30 are provided on the surfaces of the upper contact layer 24 of the third pixel 10C and the upper contact layer 24 of the fourth pixel 10D. These pads 30 are electrically connected by a wiring 29 provided with the passivation film 25 interposed therebetween. The wiring 29 is electrically connected to the fourth bump 6D provided above the fourth pixel 10D.

In this way, signals for two pixels are taken out from the two lower active layers 9 having a size for two pixels through one wiring 31 and bumps 6B and 6C, and the two pixels are electrically connected by the wiring. Signals for two pixels can be extracted from the two upper active layers 8 connected to each other through one wiring 29 and bumps 6A and 6D.
That is, in the four adjacent pixels 10 (first to fourth pixels 10A to 10D), one bump 6 (first to fourth bumps 6A to 6D) is provided above each pixel 10, and these bumps are provided. 6 are electrically connected to an upper active layer 8 and a lower active layer 9 in which two pixels are electrically connected. As a result, a signal (photocurrent signal) for two pixels can be extracted from the upper active layer 8 and the lower active layer 9 to which two pixels are electrically connected via the respective bumps 6. ing.

  Instead of dividing the lower contact layer 22 vertically and providing the insulating layer 26 therebetween, the intermediate contact layer 23 is divided vertically and an insulating layer is provided between them. The intermediate contact layer and the lower contact layer may be connected by wiring (see FIG. 24). In this case, the wiring for connecting to the bump for reading the signal from the lower active layer 9 may be connected to the intermediate contact layer below the insulating layer.

  In the infrared detection element array 5 configured as described above, when viewed as a whole, the bumps 6 connected to each of the plurality of lower active layers 9 having a size for two pixels have a size for two pixels. Above each of the plurality of lower active layers 9, each of the two adjacent upper active layers 8 that are shifted by one pixel in the direction orthogonal to the extending direction of the lower active layer 9 and that are connected by the wiring 29. The plurality of bumps 6 are arranged so that the bumps 6 connected to the two upper active layers 8 adjacent to each other connected by the wiring 29 are shifted by one pixel in the extending direction of the lower active layer 9. Is provided above each pixel 10 (see FIG. 2).

By the way, in the infrared detection element array 5 configured as described above, the four first to fourth pixels 10A to 10D adjacent to each other operate as follows.
That is, as shown in FIG. 17, first, the intermediate contact layer 23 and the second pixel 10B connected to the first pixel 10A and the third pixel 10C via the first lower contact layer 22B and the wiring 27 are connected to the fourth pixel. A bias voltage is supplied to each of the intermediate contact layers 23 connected to the pixel 10D.

  A bias voltage is applied to each of the upper active layers 8 of the first to fourth pixels 10 </ b> A to 10 </ b> D having a size corresponding to one pixel through the intermediate contact layer 23. Further, the lower active layer 9 having a size corresponding to two pixels, that is, the lower active layer 9 and the second pixel 10B connected by the first pixel 10A and the third pixel 10C through the intermediate contact layer 23. A bias voltage is applied to each of the lower active layers 9 connected to the fourth pixel 10D.

  Further, a signal for two pixels is taken out from the upper active layer 8 of the first pixel 10A and the second pixel 10B, in which two pixels are electrically connected by the wiring 29, via the first bump 6A. Further, a signal for two pixels is taken out from the upper active layer 8 of the third pixel 10C and the fourth pixel 10D in which two pixels are electrically connected by the wiring 29 via the fourth bump 6D. Further, signals for two pixels are taken out from the lower active layer 9 connected by the first pixel 10A and the third pixel 10C through the third bump 6C. Further, signals for two pixels are taken out from the lower active layer 9 connected by the second pixel 10B and the fourth pixel 10D through the second bump 6B.

Next, a method of manufacturing the infrared detection element array 5 configured as described above and the infrared image sensor 2 including the same will be described with reference to FIG.
First, on a semiconductor substrate (not shown) such as a GaAs substrate, an n-GaAs first lower contact layer 22B, an i-GaAs insulating layer 26, an n-GaAs second lower contact layer 22A, a lower active layer 9, n A GaAs intermediate contact layer 23, an upper active layer 8, and an n-GaAs upper contact layer 24 are grown. Thereby, a semiconductor laminated structure (wafer structure) constituting each pixel (infrared detecting element) 10 is formed. In this manner, the upper active layer 8 for the other wavelength band is stacked above the lower active layer 9 for the one wavelength band.

  Next, separation grooves 20 and 21 for separating the pixels 10 are formed by, for example, dry etching. That is, the separation grooves 20 and 21 are formed so that the plurality of pixels 10 are formed. Here, the isolation grooves 20A and 21 extending from the surface of the n-GaAs upper contact layer 24 to the surface of the i-GaAs insulating layer 26, and the surface of the n-GaAs intermediate contact layer 23 from the surface of the n-GaAs upper contact layer 24 A separation groove 20 </ b> B extending to is formed.

In this way, the lower active layer 9 is divided into two pixels, and a plurality of isolation grooves 20 and 21 are formed to divide the upper active layer 8 into one pixel. Thereby, a plurality of pixels 10 arranged two-dimensionally are formed.
Here, in the four adjacent first to fourth pixels 10A to 10D, the lower active layer 9 is connected in the vertical direction to two pixels of the first pixel 10A and the third pixel 10C, and in the vertical direction. Two pixels of the second pixel 10B and the fourth pixel 10D are connected. That is, in the four adjacent first to fourth pixels 10A to 10D, the upper active layer 8 is electrically connected to two pixels of the first pixel 10A and the second pixel 10B in the horizontal direction, Two pixels of the third pixel 10C and the fourth pixel 10D are electrically connected in the horizontal direction.

  Further, contact holes 33 and 34 are formed at the corners of the four adjacent pixels 10A to 10D, for example, by dry etching. Here, the n-GaAs intermediate contact layer 23 extends from the surface of the n-GaAs upper contact layer 24 to a corner located on one diagonal line of the four pixels 10A to 10D arranged adjacent to each other in a square shape. A contact hole 33 extending to the surface is formed. Further, from the surface of the n-GaAs upper contact layer 24 to the surface of the n-GaAs second lower contact layer 22A at the corner located on the other diagonal line of the four pixels 10A to 10D arranged adjacent to each other in a square shape. A contact hole 34 extending to the upper end is formed.

Next, the passivation film 25 is formed so as to cover the entire surface including the separation grooves 20 and 21 and the contact holes 33 and 34 by, for example, plasma CVD (Chemical Vapor Deposition).
Next, a part of the passivation film 25 formed on the bottom surface of the contact hole 33 positioned on one diagonal line of the adjacent four pixels 10A to 10D, that is, on the surface of the n-GaAs intermediate contact layer 23 is dried, for example. After removal by etching, the contact electrode 28 is formed, for example, by vapor deposition. Further, the bottom surface of the portion where the separation grooves 20 and 21 near the contact hole 33 located on one diagonal line of the four adjacent pixels 10A to 10D intersect, that is, on the surface of the n-GaAs first lower contact layer 22B. After removing a part of the i-GaAs insulating layer 26 and the passivation film 25 formed by, for example, dry etching, the contact electrode 28 is formed by, for example, vapor deposition. Further, after removing a part of the passivation film 25 formed on the surface of each pixel 10 by, for example, dry etching, the contact electrode 30 is formed by, for example, vapor deposition.

  Next, a wiring 29 that electrically connects the two contact electrodes 30 formed on the surfaces of the two adjacent pixels 10 is formed. Here, two upper active layers 8 adjacent to each other in a direction orthogonal to a direction in which the lower active layer 9 having a size of two pixels extends among the plurality of upper active layers 8 having a size of one pixel are respectively illustrated. The wiring 29 is electrically connected. In particular, in the four adjacent first to fourth pixels 10A to 10D, the upper active layer 8 electrically connects two pixels of the first pixel 10A and the second pixel 10B in the lateral direction by the wiring 29, and The two pixels of the third pixel 10C and the fourth pixel 10D are electrically connected by the wiring 29 in the horizontal direction. Here, for example, a wiring metal is sputtered, and unnecessary portions are removed by, for example, ion milling to form the wiring 29.

  Further, the contact electrode 28 formed on the bottom surface of the contact hole 33 located on one diagonal line of the four adjacent pixels 10A to 10D and the bottom surface of the portion where the separation grooves 20 and 21 near the contact hole 33 intersect. A wiring 27 that electrically connects the formed contact electrode 28 is formed. Here, for example, a wiring metal is sputtered, and unnecessary portions are removed by, for example, ion milling to form the wiring 27.

Next, a passivation film 25 is formed again so as to cover the entire surface.
Next, a part of the passivation film 25 formed on the bottom surface of the contact hole 34 located on the other diagonal line of the four adjacent pixels 10A to 10D, that is, on the surface of the n-GaAs second lower contact layer 22A is formed. For example, after removing by dry etching, the contact electrode 32 is formed by, for example, vapor deposition.

Next, a wiring 31 is formed extending from the upper part of the contact electrode 32 formed on the bottom surface of the contact hole 34 located on the other diagonal line of the four adjacent pixels 10A to 10D to the region where the bump 6 on the surface is provided. Here, for example, a wiring metal is sputtered and unnecessary portions are removed by, for example, ion milling to form the wiring 31.
Next, a passivation film 25 is formed again so as to cover the entire surface.

Next, after the passivation film 25 in the region where the bump 6 of each pixel 10 is provided is removed by, for example, dry etching, for example, vapor deposition or plating is performed on each surface of the wirings 29 and 31 exposed in the region where the bump 6 is provided. Thus, the bump 6 is formed. In this way, one bump 6 is formed above each pixel 10.
Here, the bump 6 connected to each of the plurality of lower active layers 9 having a size corresponding to two pixels is disposed above each of the plurality of lower active layers 9 having a size corresponding to two pixels. The bumps 6 are arranged adjacent to each other by one pixel in the direction orthogonal to the extending direction 9 and connected to each of the two adjacent upper active layers 8 connected by the wiring 29. One bump 6 is formed above each pixel 10 above each of the two upper active layers 8 so as to be shifted by one pixel from each other in the direction in which the lower active layer 9 extends (see FIG. 2). ).

  In particular, in the four adjacent pixels, that is, in the first to fourth pixels 10A to 10D arranged in two vertical and horizontal directions, the first to fourth bumps 6A to 6D are disposed above the first to fourth pixels 10A to 10D. Are formed one by one. That is, the first bump 6A electrically connected to the upper active layer 8 of the first pixel 10A and the upper active layer 8 of the second pixel 10B, the lower active layer 9 of the second pixel 10B, and the lower portion of the fourth pixel 10D. Second bump 6B electrically connected to active layer 9, third bump 6C electrically connected to lower active layer 9 of first pixel 10A and lower active layer 9 of third pixel 10C, third bump 6C A fourth bump 6D that is electrically connected to the upper active layer 8 of the pixel 10C and the upper active layer 8 of the fourth pixel 10D is formed.

  As described above, the first bump 6A, the second pixel 10B, and the fourth pixel 10D are electrically connected to the upper active layer 8 that is electrically connected to the first pixel 10A and the second pixel 10B. The second bump 6B electrically connected to the lower active layer 9 connected electrically, and the lower active layer 9 electrically connected by the first pixel 10A and the third pixel 10C are electrically connected. The third bump 6C and the fourth bump 6D electrically connected to the upper active layer 8 electrically connected by the third pixel 10C and the fourth pixel 10D are formed.

Next, after dicing (cutting) into a chip size to form an infrared detection element array 5 in which a plurality of pixels 10 are two-dimensionally arranged, this is attached to the signal processing circuit chip 7 by flip chip bonding (FCB). Match. In this way, each pixel 10 is electrically connected to the signal processing circuit chip 7 via the bump 6.
In this way, the infrared detection element array 5 according to the present embodiment and the infrared image sensor 2 including the same are completed.

  Therefore, according to the sensor element array, the manufacturing method thereof, and the imaging apparatus according to the present embodiment, there is an advantage that the sensitivity can be improved while realizing 1 pixel 1 bump. That is, one pixel and one bump can be realized. Also, by using 1 bump per pixel, with the same bump pitch as in the case of 3 bumps per pixel (see FIG. 14), the pixel size is reduced to 1/4, and the number of pixels is quadrupled without increasing the chip size. Pixelization is possible. Further, even if one pixel has one bump, it is not necessary to alternately read out the signal from the first active layer 8 and the signal from the second active layer 9 by switching the bias, so that the sensitivity can be improved.

Note that the present invention is not limited to the configurations described in the above-described embodiments and modifications, and various modifications can be made without departing from the spirit of the present invention.
For example, in the above-described embodiment, when the infrared detection element array 5 is viewed as a whole, the plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction are aligned in the vertical direction and the horizontal direction. A plurality of second active layers 9 that are electrically connected in the vertical direction are arranged in the vertical direction and the horizontal direction, and the two pixels are electrically connected in the vertical direction. Two first active layers 8 in which two pixels are electrically connected in the horizontal direction are arranged immediately above the two second active layers 9 (see FIG. 2). It is not limited to this, and may be shifted.

  For example, as shown in FIG. 18A, a plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction are arranged in the horizontal direction and shifted by one pixel in the vertical direction. May be arranged. Here, as shown in FIG. 18B, the plurality of second active layers 9 in which two pixels are electrically connected in the vertical direction are arranged in the vertical direction and the horizontal direction. Note that the structure illustrated in FIG. 18A and the structure illustrated in FIG. 18B are actually stacked.

  In this case as well, as in the case of the above-described embodiment, as shown in FIG. 18A, each of the plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction is electrically connected. Bumps 6 connected to each other are arranged so as to be shifted by one pixel in the vertical direction above each of the first active layers 8 in which two pixels are electrically connected in the horizontal direction, and FIG. ), The bump 6 electrically connected to each of the plurality of second active layers 9 electrically connected to two pixels in the vertical direction is electrically connected to two pixels in the vertical direction. A plurality of bumps 6 are provided above each pixel 10 above each connected second active layer 9 so as to be shifted by one pixel in the lateral direction.

  Also in this case, the first to fourth bumps are located above the first to fourth pixels 10A to 10D in the four adjacent pixels, that is, the first to fourth pixels 10A to 10D arranged in two vertical and horizontal directions. Each of 6A to 6D is provided. The first bump 6A is electrically connected to the first active layer 8 of the first pixel 10A and the first active layer 8 of the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 of the second pixel 10B and the second active layer 9 of the fourth pixel 10D. The third bump 6C is electrically connected to the second active layer 9 of the first pixel 10A and the second active layer 9 of the third pixel 10C. The fourth bump 6D is electrically connected to the first active layer 8 of the fourth pixel 10D and the first active layer 8 of pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. Is done.

  Thus, in the four adjacent first to fourth pixels 10A to 10D, the first active layer 8 is electrically connected to the two pixels of the first pixel 10A and the second pixel 10B in the lateral direction. Further, the third pixel 10C in the horizontal direction and two pixels other than the first to fourth pixels 10A to 10D adjacent thereto are electrically connected, and the fourth pixel 10D and the fourth pixel 10D are connected in the horizontal direction. Two pixels other than the adjacent first to fourth pixels 10A to 10D are electrically connected. The first bump 6A is electrically connected to the first active layer 8 that is electrically connected between the first pixel 10A and the second pixel 10B, and the first pixel adjacent to the fourth pixel 10D. The fourth bump 6D is electrically connected to the first active layer 8 that is electrically connected to pixels other than the fourth pixels 10A to 10D.

  The second active layer 9 is electrically connected to two pixels of the first pixel 10A and the third pixel 10C in the vertical direction, and 2nd of the second pixel 10B and the fourth pixel 10D in the vertical direction. Pixels are electrically connected. The third bump 6C is electrically connected to the second active layer 9 electrically connected to the first pixel 10A and the third pixel 10C, and the second pixel 10B and the fourth pixel 10D are connected to each other. The second bump 6B is electrically connected to the second active layer 9 that is electrically connected.

In this case, in the adjacent four pixels 10A to 10D, two pixels are electrically connected in the horizontal direction immediately above the two second active layers 9 in which two pixels are electrically connected in the vertical direction. One pixel of one first active layer 8 and two first active layers 9 in which two pixels are electrically connected in the horizontal direction are arranged.
In the case of such a configuration, the control calculation unit 3 outputs the output signals UL1, UR1, DL1, DR1, UL2, UR2, and the like from the first active layer 8 and the second active layer 9 of each pixel 10 as follows. What is necessary is just to obtain | require each value (calculated value) of DL2 and DR2.

Here, the output signal output through each bump 6 is an alphabet attached to the bump 6 and a number “1” or second indicating that the output signal is from the first active layer 8 (wavelength 1). It is expressed by a combination with a numeral “2” indicating that it is an output signal from the active layer 9 (wavelength 2). For example, the output signal from the first active layer 8 output via the bump A is “A1”.
UL1 = F1 × B2 / (B2 + G2)
UR1 = F1 × G2 / (B2 + G2)
DL1 = A1 × B2 / (B2 + E2)
DR1 = C1 × G2 / (D2 + G2)
UL2 = B2 × F1 / (A1 + F1)
UR2 = G2 × F1 / (C1 + F1)
DL2 = B2 × A1 / (A1 + F1)
DR2 = G2 × C1 / (C1 + F1)
Here, the control calculation unit 3 performs the following processing in accordance with the configuration of the infrared detection element array 5 described above.

  That is, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A to the second pixel 10B and the fourth pixel. Based on the ratio of the output signal read from the second active layer 9 of the pixel 10D and the output signal read from the second active layer 9 of the first pixel 10A and the third pixel 10C, the first signal is distributed. An output signal from the first active layer 8 of the pixel 10A and an output signal from the first active layer 8 of the second pixel 10B are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D via the second bump 6B to the first pixel 10A and the second pixel 10D. The output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto. And the output signal from the second active layer 9 of the second pixel 10B and the output signal from the second active layer 9 of the fourth pixel 10D are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C via the third bump 6C to the first pixel 10A and the second pixel 10C. The output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the third pixel 10C and the pixels other than the first to fourth pixels 10A to 10D adjacent thereto. Output signal from the second active layer 9 of the first pixel 10A and output signal from the second active layer 9 of the third pixel 10C. Is to ask for.

  In addition, the control calculation unit 3 includes two pixels read from the first active layer 8 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto via the fourth bump 6D. Output signals read from the second active layer 9 of the second pixel 10B and the fourth pixel 10D, the pixels other than the first to fourth pixels 10A to 10D adjacent to the second pixel 10B, and the fourth pixel. Distributing based on the ratio of the output signal read from the second active layer 9 of the pixels other than the first to fourth pixels 10A to 10D adjacent to the pixel (output signal from the bump indicated by reference sign D), An output signal from the first active layer 8 of the fourth pixel 10D is obtained.

  In addition, the control calculation unit 3 uses the bumps other than the first to fourth bumps 6A to 6D (shown by the reference symbol A) other than the third pixel 10C and the first to fourth pixels 10A to 10D adjacent thereto. The output signal for two pixels read from the first active layer 8 of the pixel is adjacent to the output signal read from the second active layer 9 of the first pixel 10A and the third pixel 10C and the first pixel 10A. Output signals read from the second active layer 9 of the pixels other than the first to fourth pixels 10A to 10D and the pixels other than the first to fourth pixels 10A to 10D adjacent to the third pixel 10C (indicated by symbol E) The output signal from the first active layer of the third pixel is obtained based on the ratio to the output signal from the bump).

  For example, as shown in FIG. 19B, a plurality of second active layers 9 in which two pixels are electrically connected in the vertical direction are arranged in the vertical direction, and one pixel in the horizontal direction. You may make it arrange | position by shifting. Here, as shown in FIG. 19A, the plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction are arranged in the vertical direction and the horizontal direction. Note that the structure shown in FIG. 19A and the structure shown in FIG. 19B are actually stacked.

  In this case as well, as in the case of the above-described embodiment, as shown in FIG. 19A, each of the plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction is electrically connected. Bumps 6 connected to each other are arranged so as to be shifted by one pixel in the vertical direction above each of the first active layers 8 in which two pixels are electrically connected in the horizontal direction, and FIG. ), The bump 6 electrically connected to each of the plurality of second active layers 9 electrically connected to two pixels in the vertical direction is electrically connected to two pixels in the vertical direction. A plurality of bumps 6 are provided above each pixel 10 above each connected second active layer 9 so as to be shifted by one pixel in the lateral direction.

  Also in this case, the first to fourth bumps are located above the first to fourth pixels 10A to 10D in the four adjacent pixels, that is, the first to fourth pixels 10A to 10D arranged in two vertical and horizontal directions. Each of 6A to 6D is provided. The first bump 6A is electrically connected to the first active layer 8 of the first pixel 10A and the first active layer 8 of the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 of the second pixel 10B and the second active layer 9 of pixels other than the first to fourth pixels 10A to 10D adjacent to the second pixel 10B. Is done. The third bump 6C is electrically connected to the second active layer 9 of the first pixel 10A and the second active layer 9 of the third pixel 10C. The fourth bump 6D is electrically connected to the first active layer 8 of the third pixel 10C and the first active layer 8 of the fourth pixel 10D.

  Thus, in the four adjacent first to fourth pixels 10A to 10D, the first active layer 8 is electrically connected to the two pixels of the first pixel 10A and the second pixel 10B in the lateral direction. In addition, two pixels of the third pixel 10C and the fourth pixel 10D are electrically connected in the horizontal direction. The first bump 6A is electrically connected to the first active layer 8 electrically connected to the first pixel 10A and the second pixel 10B, and the fourth pixel 10D and the third pixel 10C are connected to each other. The fourth bump 6D is electrically connected to the first active layer 8 that is electrically connected.

  The second active layer 9 is electrically connected to two pixels of the first pixel 10A and the third pixel 10C in the vertical direction, and the first pixel adjacent to the second pixel 10B in the vertical direction. Two pixels other than the fourth pixels 10A to 10D are electrically connected, and the fourth pixel 10D in the vertical direction and two pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. Minutes are electrically connected. The third bump 6C is electrically connected to the second active layer 9 that is electrically connected between the first pixel 10A and the third pixel 10C, and the second pixel 10B and the first adjacent to the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 electrically connected to pixels other than the fourth pixels 10A to 10D.

In this case, in the adjacent four pixels 10A to 10D, two pixels are electrically connected in the vertical direction immediately below the two first active layers 8 in which two pixels are electrically connected in the horizontal direction. One pixel of one second active layer 9 and two second active layers 9 in which two pixels are electrically connected in the vertical direction is arranged.
In the case of such a configuration, the control calculation unit 3 outputs the output signals UL1, UR1, DL1, DR1, UL2, UR2, and the like from the first active layer 8 and the second active layer 9 of each pixel 10 as follows. What is necessary is just to obtain | require each value (calculated value) of DL2 and DR2.

Here, the output signal output through each bump 6 is an alphabet attached to the bump 6 and a number “1” or second indicating that the output signal is from the first active layer 8 (wavelength 1). It is expressed by a combination with a numeral “2” indicating that it is an output signal from the active layer 9 (wavelength 2). For example, the output signal from the first active layer 8 output via the bump A is “A1”.
UL1 = F1 × J2 / (E2 + J2)
UR1 = F1 × E2 / (E2 + J2)
DL1 = I1 × J2 / (M2 + J2)
DR1 = I1 × M2 / (M2 + J2)
UL2 = J2 × F1 / (F1 + I1)
UR2 = E2 × F1 / (A1 + F1)
DL2 = J2 × I1 / (F1 + I1)
DR2 = M2 × I1 / (I1 + N1)
Here, the control calculation unit 3 performs the following processing in accordance with the configuration of the infrared detection element array 5 described above.

  That is, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A to the first pixel 10A and the third pixel 10A. The output signal read from the second active layer 9 of the pixel 10C and the output signal read from the second active layer 9 of the pixel other than the second pixel 10B and the first to fourth pixels 10A to 10D adjacent thereto. And the output signal from the first active layer 8 of the first pixel 10A and the output signal from the first active layer 8 of the second pixel 10B are obtained.

  In addition, the control calculation unit 3 reads two pixels read from the second active layer 9 of the pixels other than the second pixel 10B and the first to fourth pixels 10A to 10D adjacent to the second pixel 10B via the second bump 6B. Output signals read from the first active layer 8 of the first pixel 10A and the second pixel 10B, the pixels other than the first to fourth pixels 10A to 10D adjacent to the first pixel 10A, and the second pixel. Distribution based on the ratio of the output signals (output signals from the bumps indicated by symbol A) read from the first active layer 8 of the pixels other than the first to fourth pixels 10A to 10D adjacent to the pixel 10B. The output signal from the second active layer 9 of the second pixel 10B is obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C via the third bump 6C to the first pixel 10A and the second pixel 10C. Based on the ratio between the output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D, the first signal is distributed. An output signal from the second active layer 9 of the pixel 10A and an output signal from the second active layer 9 of the third pixel 10C are obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D via the fourth bump 6D to the first pixel 10A and the third pixel 10D. The output signal read from the second active layer 9 of the pixel 10C and the output signal read from the second active layer 9 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto. Output signals from the first active layer 8 of the third pixel 10C and output signals from the first active layer 8 of the fourth pixel 10D. Is to ask for.

  In addition, the control calculation unit 3 uses the bumps other than the first to fourth bumps 6A to 6D (indicated by the reference symbol M) other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto. The output signals for two pixels read from the second active layer 9 of the pixel are adjacent to the output signals read from the first active layer 8 of the third pixel 10C and the fourth pixel 10D and the third pixel 10C. Output signals read from the first active layer 8 of the pixels other than the first to fourth pixels 10A to 10D and the pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D (indicated by reference numeral N) The output signal from the second active layer 9 of the fourth pixel 10D is obtained based on the ratio to the output signal from the bump).

  For example, as shown in FIG. 20A, a plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction are arranged in the horizontal direction, and one pixel in the vertical direction. As shown in FIG. 20B, a plurality of second active layers 9 in which two pixels are electrically connected in the vertical direction are arranged in the vertical direction, and 1 in the horizontal direction. You may make it arrange | position by shifting by a pixel. Note that the structure shown in FIG. 20A and the structure shown in FIG. 20B are actually stacked.

  In this case as well, as in the case of the above-described embodiment, as shown in FIG. 20A, each of the plurality of first active layers 8 in which two pixels are electrically connected in the horizontal direction is electrically connected. Bumps 6 connected to each other are arranged so as to be shifted by one pixel in the vertical direction above each of the first active layers 8 in which two pixels are electrically connected in the horizontal direction, and FIG. ), The bump 6 electrically connected to each of the plurality of second active layers 9 electrically connected to two pixels in the vertical direction is electrically connected to two pixels in the vertical direction. A plurality of bumps 6 are provided above each pixel 10 above each connected second active layer 9 so as to be shifted by one pixel in the lateral direction.

  Also in this case, the first to fourth bumps are located above the first to fourth pixels 10A to 10D in the four adjacent pixels, that is, the first to fourth pixels 10A to 10D arranged in two vertical and horizontal directions. Each of 6A to 6D is provided. The first bump 6A is electrically connected to the first active layer 8 of the first pixel 10A and the first active layer 8 of the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 of the second pixel 10B and the second active layer 9 of pixels other than the first to fourth pixels 10A to 10D adjacent to the second pixel. The The third bump 6C is electrically connected to the second active layer 9 of the first pixel 10A and the second active layer 9 of the third pixel 10C. The fourth bump 6D is electrically connected to the first active layer 8 of the fourth pixel 10D and the first active layer 8 of pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. Is done.

  Thus, in the four adjacent first to fourth pixels 10A to 10D, the first active layer 8 is electrically connected to the two pixels of the first pixel 10A and the second pixel 10B in the lateral direction. Further, the third pixel 10C in the horizontal direction and two pixels other than the first to fourth pixels 10A to 10D adjacent thereto are electrically connected, and the fourth pixel 10D in the horizontal direction is connected to the fourth pixel 10D. Two pixels other than the first to fourth pixels 10A to 10D adjacent thereto are electrically connected. The first bump 6A is electrically connected to the first active layer 8 that is electrically connected between the first pixel 10A and the second pixel 10B, and the first pixel adjacent to the fourth pixel 10D. The fourth bump 6D is electrically connected to the first active layer 8 that is electrically connected to pixels other than the fourth pixels 10A to 10D.

  The second active layer 9 is electrically connected to two pixels of the first pixel 10A and the third pixel 10C in the vertical direction, and the first pixel adjacent to the second pixel 10B in the vertical direction. Two pixels other than the fourth pixels 10A to 10D are electrically connected, and the fourth pixel 10D in the vertical direction and two pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. Minutes are electrically connected. The third bump 6C is electrically connected to the second active layer 9 that is electrically connected between the first pixel 10A and the third pixel 10C, and the second pixel 10B and the first adjacent to the second pixel 10B. The second bump 6B is electrically connected to the second active layer 9 electrically connected to pixels other than the fourth pixels 10A to 10D.

  In this case, in the adjacent four pixels 10A to 10D, one first active layer 8 in which two pixels are electrically connected in the horizontal direction and two pixels in the horizontal direction are electrically connected 2 Immediately below one pixel of the first active layer 8, one second active layer 9 in which two pixels are electrically connected in the vertical direction and two pixels in the vertical direction are electrically connected. One pixel of the two second active layers 9 is arranged.

In the case of such a configuration, the control calculation unit 3 outputs the output signals UL1, UR1, DL1, DR1, UL2, UR2, and the like from the first active layer 8 and the second active layer 9 of each pixel 10 as follows. What is necessary is just to obtain | require each value (calculated value) of DL2 and DR2.
Here, the output signal output through each bump 6 is an alphabet attached to the bump 6 and a number “1” or second indicating that the output signal is from the first active layer 8 (wavelength 1). It is expressed by a combination with a numeral “2” indicating that it is an output signal from the active layer 9 (wavelength 2). For example, the output signal from the first active layer 8 output via the bump A is “A1”.
UL1 = F1 × J2 / (J2 + G2)
UR1 = F1 × G2 / (J2 + G2)
DL1 = I1 × J2 / (J2 + M2)
DR1 = K1 × O2 / (O2 + L2)
UL2 = J2 × F1 / (F1 + I1)
UR2 = G2 × F1 / (C1 + F1)
DL2 = J2 × I1 / (F1 + I1)
DR2 = O2 × K1 / (K1 + N1)
Here, the control calculation unit 3 performs the following processing in accordance with the configuration of the infrared detection element array 5 described above.

  That is, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A to the first pixel 10A and the third pixel 10A. The output signal read from the second active layer 9 of the pixel 10C and the output signal read from the second active layer 9 of the pixel other than the second pixel 10B and the first to fourth pixels 10A to 10D adjacent thereto. And the output signal from the first active layer 8 of the first pixel 10A and the output signal from the first active layer 8 of the second pixel 10B are obtained.

  In addition, the control calculation unit 3 reads two pixels read from the second active layer 9 of the pixels other than the second pixel 10B and the first to fourth pixels 10A to 10D adjacent to the second pixel 10B via the second bump 6B. Output signals read from the first active layer 8 of the first pixel 10A and the second pixel 10B, pixels other than the first to fourth pixels 10A to 10D adjacent to the second pixel 10B, and Distributing based on the ratio of the output signals read from the first active layer 8 of the pixels other than the first to fourth pixels 10A to 10D adjacent to each other (output signals from the bumps indicated by symbol C), An output signal from the second active layer 9 of the pixel 10B is obtained.

  Further, the control calculation unit 3 outputs the output signals for two pixels read from the second active layer 9 of the first pixel 10A and the third pixel 10C via the third bump 6C to the first pixel 10A and the second pixel 10C. The output signal read from the first active layer 8 of the pixel 10B and the output signal read from the first active layer 8 of the third pixel 10C and the pixels other than the first to fourth pixels 10A to 10D adjacent thereto. Output signal from the second active layer 9 of the first pixel 10A and output signal from the second active layer 9 of the third pixel 10C. Is to ask for.

  In addition, the control calculation unit 3 includes two pixels read from the first active layer 8 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto via the fourth bump 6D. Output signal read from the second active layer 9 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D (output signal from the bump indicated by symbol O). And the second active layer 9 of pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D and pixels other than the first to fourth pixels 10A to 10D adjacent to the second pixel 10B. The output signal from the first active layer 8 of the fourth pixel 10D is obtained based on the ratio with the output signal (the output signal from the bump indicated by the symbol L).

  In addition, the control calculation unit 3 uses the bumps other than the first to fourth bumps 6A to 6D (indicated by reference numeral I) other than the third pixel 10C and the first to fourth pixels 10A to 10D adjacent thereto. The output signals for two pixels read from the first active layer 8 of the pixel are adjacent to the output signals read from the second active layer 9 of the first pixel 10A and the third pixel 10C and the third pixel 10C. Output signals read from the second active layer 9 of the pixels other than the first to fourth pixels 10A to 10D and the pixels other than the first to fourth pixels 10A to 10D adjacent thereto (from the bumps indicated by symbol M) The output signal from the first active layer 8 of the third pixel 10C is obtained based on the ratio to the output signal).

  In addition, the control calculation unit 3 uses the bumps other than the first to fourth bumps 6A to 6D (shown by the symbol O) to the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. The output signals for two pixels read out from the second active layer 9 of the pixel are read out from the first active layer 8 of the pixels other than the fourth pixel 10D and the first to fourth pixels 10A to 10D adjacent thereto. The first active layer 8 of the output signal and the pixels other than the first to fourth pixels 10A to 10D adjacent to the third pixel 10C and the pixels other than the first to fourth pixels 10A to 10D adjacent to the fourth pixel 10D. The output signal from the second active layer 9 of the fourth pixel 10D is obtained based on the ratio with the output signal (output signal from the bump indicated by the symbol N) read out from.

  In short, the infrared detection element array 5 has a structure in which the first active layer 8 for the first wavelength band and the second active layer 9 for the second wavelength band are stacked, and the first to fourth pixels 10A to 10A adjacent to each other. 10D, the first bump 6A electrically connected to the first active layer 8 of the first pixel 10A and the first active layer 8 of the second pixel 10B, and the second active layer 9 of the second pixel 10B electrically Connected second bump 6B, third bump 6C electrically connected to second active layer 9 of first pixel 10A and second active layer 9 of third pixel 10C, and fourth pixel 10D. 4th bump 6D electrically connected to the 1st active layer 8 of this, and 1st-4th bump 6A-6D is provided one above each of the 1st-4th pixels 10A-10D. What is necessary is just to comprise.

  Then, the control calculation unit 3 outputs the output signals for two pixels read from the first active layer 8 of the first pixel 10A and the second pixel 10B via the first bump 6A to the second pixel 10B. Distributing based on the ratio of the output signal read from the active layer 9 and the output signal read from the second active layer 9 of the first pixel 10A, from the first active layer 8 of the first pixel 10A. The output signal and the output signal from the first active layer 8 of the second pixel 10B are obtained, and 2 read out from the second active layer 9 of the first pixel 10A and the third pixel 10C through the third bump 6C. The output signal for pixels is distributed based on the ratio of the output signal read from the first active layer 8 of the first pixel 10A and the output signal read from the first active layer 8 of the third pixel 10C. The output signal from the second active layer 9 of the first pixel 10A and the third pixel 10C It may be configured to determine an output signal from the second active layer 9.

In the above-described embodiment, the case where the quantum well layers are used as the active layers 8 and 9 is described as an example. However, the present invention is not limited to this. For example, a quantum dot layer is used as the active layer. Alternatively, a semiconductor layer made of a bulk material or a superlattice may be used.
In the above-described embodiment, the infrared image sensor 5 is described as an example. However, the present invention is not limited to this. For example, the present invention can be applied to an image sensor used for detection of visible light. it can. Further, the configuration of the imaging apparatus is not limited to that of the above-described embodiment.

DESCRIPTION OF SYMBOLS 1 Infrared imaging device 2 Infrared image sensor 3 Control calculating part 4 Monitor 5 Infrared detection element array 6 Bump 6A 1st bump 6B 2nd bump 6C 3rd bump 6D 4th bump 7 Signal processing circuit chip 8 1st MQW layer (1st activity) Layer; upper active layer)
9 Second MQW layer (second active layer; lower active layer)
10 pixels (infrared detector)
10A 1st pixel 10B 2nd pixel 10C 3rd pixel 10D 4th pixel 11 Vacuum container 12 Cooling system (refrigerator)
13 Cold head 14 Lens 15 Output wiring 20, 20A, 20B Separation groove (lateral separation groove; second separation groove)
21 Separation groove (longitudinal separation groove; first separation groove)
22 Lower contact layer 22A Lower contact layer (second lower contact layer)
22B Lower contact layer (first lower contact layer)
23 intermediate contact layer 24 upper contact layer 25 passivation film (insulating film)
26 Insulating layer 27 Wiring 28 Pad 29 Wiring 30 Pad 31 Wiring 31
32 Pad 33, 34 Contact hole

Claims (8)

A first to fourth pixel having a structure in which a first active layer for a first wavelength band and a second active layer for a second wavelength band are stacked, and arranged two by two vertically and horizontally;
A first bump electrically connected to the first active layer of the first pixel and the first active layer of the second pixel;
A second bump electrically connected to the second active layer of the second pixel;
A third bump electrically connected to the second active layer of the first pixel and the second active layer of the third pixel;
A fourth bump electrically connected to the first active layer of the fourth pixel,
The sensor element array according to claim 1, wherein the first to fourth bumps are provided one above the first to fourth pixels, respectively. The second bump is further electrically connected to the second active layer of the fourth pixel,
The sensor element array according to claim 1, wherein the fourth bump is further electrically connected to the first active layer of the third pixel. The second bump is further electrically connected to the second active layer of the fourth pixel,
2. The sensor according to claim 1, wherein the fourth bump is further electrically connected to a first active layer of a pixel other than the first to fourth pixels adjacent to the fourth pixel. Element array. The second bump is further electrically connected to a second active layer of a pixel other than the first to fourth pixels adjacent to the second pixel,
The sensor element array according to claim 1, wherein the fourth bump is further electrically connected to the first active layer of the third pixel. The second bump is further electrically connected to a second active layer of a pixel other than the first to fourth pixels adjacent to the second pixel,
2. The sensor according to claim 1, wherein the fourth bump is further electrically connected to a first active layer of a pixel other than the first to fourth pixels adjacent to the fourth pixel. Element array. A first to fourth pixel having a structure in which a first active layer for a first wavelength band and a second active layer for a second wavelength band are stacked, and arranged two by two vertically and horizontally;
A first bump electrically connected to the first active layer of the first pixel and the first active layer of the second pixel;
A second bump electrically connected to the second active layer of the second pixel;
A third bump electrically connected to the second active layer of the first pixel and the second active layer of the third pixel;
A fourth bump electrically connected to the first active layer of the fourth pixel,
A sensor element array in which the first to fourth bumps are provided one above the first to fourth pixels, respectively;
A signal processing circuit chip connected to the sensor element array via a plurality of bumps including at least the first to fourth bumps;
A control operation unit connected to the signal processing circuit chip,
The control calculation unit is
An output signal for two pixels read from the first active layer of the first pixel and the second pixel through the first bump is read from the second active layer of the second pixel. Distributing based on the ratio of the output signal and the output signal read from the second active layer of the first pixel, the output signal from the first active layer of the first pixel and the second pixel An output signal from the first active layer is obtained, and an output signal for two pixels read from the second active layer of the first pixel and the third pixel through the third bump is obtained. Distributing based on the ratio of the output signal read from the first active layer of the first pixel and the output signal read from the first active layer of the third pixel, Output signal from the second active layer and output signal from the second active layer of the third pixel Characterized in that it is configured to determine the imaging device. A plurality of pixels having a structure in which an upper active layer for another wavelength band is stacked above a lower active layer for one wavelength band;
A plurality of isolation grooves that divide the lower active layer into two pixels and divide the upper active layer into one pixel;
Of the plurality of upper active layers having a size corresponding to one pixel, two upper active layers adjacent to each other in a direction orthogonal to a direction in which the lower active layer having a size corresponding to two pixels extends are electrically connected to each other. Multiple wiring to
Bumps connected to each of the plurality of lower active layers having a size corresponding to the two pixels extend in a direction in which the lower active layer extends above each of the plurality of lower active layers having a size corresponding to the two pixels. The bumps connected to each of the two adjacent upper active layers connected by the wiring are arranged adjacent to each other in the direction orthogonal to the two upper active layers connected by the wiring. And a plurality of bumps provided one above each of the plurality of pixels so as to be shifted from each other by one pixel in the extending direction of the lower active layer. Element array. Laminating an upper active layer for another wavelength band above a lower active layer for one wavelength band,
Dividing the lower active layer into two pixels and forming a plurality of isolation grooves to divide the upper active layer into one pixel;
Of the plurality of upper active layers having a size corresponding to one pixel, two upper active layers adjacent to each other in a direction orthogonal to a direction in which the lower active layer having a size corresponding to two pixels extends are electrically connected to each other by wiring. Connected to
Bumps connected to each of the plurality of lower active layers having a size corresponding to the two pixels extend in a direction in which the lower active layer extends above each of the plurality of lower active layers having a size corresponding to the two pixels. The bumps connected to each of the two adjacent upper active layers connected by the wiring are arranged adjacent to each other in the direction orthogonal to the two upper active layers connected by the wiring. A bump is formed above each of the pixels so as to be shifted from each other by one pixel in the direction in which the lower active layer extends.