Semiconductor optical sensor for visible and ultraviolet light detection and corresponding manufacturing process

Abstract

A semiconductor optical sensor (1) is provided with: a substrate (2) integrating a plurality of photodetector active areas (4); and a CMOS layer stack (6) arranged on the substrate (2) and including a number of dielectric (6a) and conductive (6b) layers. UV conversion regions (10) are arranged above a number of first photodetector active areas (4) to convert UV light radiation into visible light radiation towards the first photodetector active areas (4), so that the first photodetector active areas (4) are designed to detect UV light radiation. In particular, the first photodetector active areas (4) are alternated to a number of second photodetector active areas (4), designed to detect visible light radiation, in an array (15) of photodetection units (16) of the optical sensor (1), defining a single image detection area (15′), sensitive to both UV and visible light radiation with a same spatial resolution.

Claims

1. A semiconductor optical sensor (1), comprising: a substrate (2) integrating a plurality of photodetector active areas (4); and a CMOS layer stack (6) arranged on the substrate (2) and including a number of dielectric (6a) and conductive (6b) layers; further comprising UV conversion regions (10) arranged above a number of first photodetector active areas (4) and configured to convert UV light radiation into visible light radiation towards said first photodetector active areas (4), said first photodetector active areas (4) being thereby designed to define first photodetection units (16) for UV light detection, wherein the first photodetector active areas (4) are alternated to a number of second photodetector active areas (4), integrated in the substrate (2) and designed to define second photodetection units (16) for visible light detection, the first and second photodetection units forming an array (15) of photodetection units (16) of the optical sensor (1), jointly defining a single image detection area (15′) sensitive to both UV and visible light radiation, characterized by further comprising filter regions (11), arranged above said second photodetector active areas (4) and made of a material transparent to visible light radiation and opaque to UV light radiation, and further characterized in that the single image detection area (15′) is designed to be sensitive to both UV and visible light radiation, at a same time and with a same spatial resolution.

2. The sensor according to claim 1, wherein the UV conversion regions (10) include an organic material in which a fluorescent dye is immersed, the fluorescent dye being designed to perform the UV conversion.

3. The sensor according to claim 1, wherein the organic material is a down-conversion material capable of absorbing UV light and emitting visible light radiation, chosen among: Lumogen, Coronene, AlQ3′ and ZnS:Mn.

4. The sensor according to claim 1, wherein the UV conversion regions (10) are alternated to the filter regions (11) within a capping layer (12) arranged above the first and second photodetector active areas (4).

5. The sensor according to claim 4, wherein the CMOS stack (6) is arranged on a top surface (2a) of said substrate (2), opposite to a back surface (2b) of said substrate (2) defining an outer surface of said optical sensor (1); wherein the capping layer (12) is arranged above a top dielectric layer (6a) of the CMOS stack (6).

6. The sensor according to claim 4, wherein the CMOS stack (6) is arranged on a top surface (2a) of said substrate (2), and a top dielectric layer (6a) of the CMOS stack (6) is arranged at an outer surface of said optical sensor (1); wherein the capping layer (12) is arranged above a back surface (2b) of said substrate (2), opposite to said front surface (2a).

7. The sensor according to claim 1, further comprising light channelling regions (7′) formed within trenches through the CMOS stack (6), from a top dielectric layer (6a) towards a front surface (2a) of the substrate (2), the light channelling regions (7′) including a material designed to define the UV conversion regions (10), integrated within the CMOS stack (6); wherein the light channelling regions (7′) are arranged on respective first photodetector active areas (4) and configured to guide light radiation towards the respective first photodetector active areas.

8. The sensor according to claim 7, comprising transparent regions (11′) alternated to the filter regions (11) within a capping layer (12) arranged above a top dielectric layer (6a) of the CMOS stack (6), the transparent regions (11′) designed to be transparent to both visible light and UV light radiation.

9. The sensor according to claim 7, wherein the light channelling regions (7′) are also arranged on respective second photodetector active areas (4) and configured to guide light radiation towards the respective second photodetector active areas (4).

10. The sensor according to claim 1, wherein the filter regions (11) are transparent to a respective portion of the visible light spectrum, to implement RGB, Red, Green and Blue, filters, designed to filter the visible light radiation.

11. A process for manufacturing a semiconductor optical sensor (1), comprising: providing a substrate (2); forming a plurality of photodetector active areas (4) within the substrate (2); and forming a CMOS layer stack (6) arranged on the substrate (2) and including a number of dielectric (6a) and conductive (6b) layers; further comprising forming UV conversion regions (10) arranged above a number of first photodetector active areas (4) and configured to convert UV light radiation into visible light radiation towards said first photodetector active areas (4), said first photodetector active areas (4) being thereby designed to define first photodetection units (16) for UV light detection, further comprising integrating in the substrate (2) a number of second photodetector active areas (4), alternated to the first photodetector active areas (4) and designed to define second photodetection units (16) for visible light detection, the first and second photodetection units forming an array (15) of photodetection units (16) of the optical sensor (1), jointly defining a single image detection area (15′) sensitive to both UV and visible light radiation, characterized by further comprising forming filter regions (11) arranged above said second photodetector active areas (4), made of a material transparent to visible light radiation and opaque to UV light radiation, and further characterized by forming a single color image displaying visible and UV information at a same time and with a same spatial resolution.

12. The process according to claim 11, wherein forming UV conversion regions (10) comprises depositing an organic material in which a fluorescent dye is immersed, the fluorescent dye being designed to perform the UV conversion.

13. The process according to claim 11, wherein forming the UV conversion regions (10) and the filter regions (11) comprises depositing a photo-patternable organic material layer with a fluorescent dye dispersed therein, to create patches of fluorescent material defining the UV conversion regions (10).

14. The process according to claim 11, further comprising: forming light channelling regions (7′) within trenches through the CMOS stack (6), from a top dielectric layer (6a) towards a front surface (2a) of the substrate (2); and filling the light channelling regions (7′) with a material designed to define the UV conversion regions (10) integrated within the CMOS stack (6); wherein the light channelling regions (7′) are each arranged on a respective first photodetector active area (4) and configured to guide light radiation towards the respective first photodetector active area.

15. The process according to claim 14, comprising forming transparent regions (11′) alternated to the filter regions (11) within a capping layer (12) arranged above a top dielectric layer (6a) of the CMOS stack (6), the transparent regions (11′) designed to be transparent to the visible light and the UV light radiation.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) For a better understanding of the present invention, preferred embodiments thereof are now described purely by way of non-limiting example and with reference to the attached drawings, wherein:

(2) FIG. 1 is a schematic representation of a known single-chip optical sensor;

(3) FIG. 2 is a cross-section of a semiconductor optical sensor, according to a first embodiment of the present solution;

(4) FIG. 3 is a schematic top-down representation of the optical sensor of FIG. 2;

(5) FIGS. 4A-4D are cross-sections of the optical sensor of FIG. 2, in subsequent steps of a corresponding manufacturing process;

(6) FIG. 5 is a cross-section of a semiconductor optical sensor, according to a further aspect of the present solution;

(7) FIG. 6 is a cross-section of a semiconductor optical sensor, according to a second embodiment of the present solution;

(8) FIGS. 7A-7D are cross-sections of the optical sensor of FIG. 6, in subsequent steps of a corresponding manufacturing process; and

(9) FIG. 8 is a cross-section of a semiconductor optical sensor, according to a further aspect of the present solution.

BEST MODE FOR CARRYING OUT THE INVENTION

(10) As will be discussed in details in the following, an aspect of the present solution provides a single-chip optical, or image, sensor, provided with both visible and UV detection ability in a same die of semiconductor material; the optical sensor includes an array of sensing cells, alternating in a single detection area both UV detection cells and visible light detection cells, so that the die implements detection of UV and visible light radiation, not only at the same time, but also with a same spatial resolution.

(11) FIG. 2 shows an optical sensor 1 according to a first embodiment of the present solution, relating, by way of example, to a front-side illuminated (FSI) photodetector, made with CMOS semiconductor techniques.

(12) The optical sensor 1 comprises a substrate 2 of semiconductor material, in particular silicon, having a front surface 2a and a back surface 2b, here defining a first external surface of the optical sensor 1.

(13) A plurality of photodetector active areas 4 are formed in the substrate 2, at the first surface 2a thereof, each defining a photodetection area of a respective photodetection unit (so called “pixel”) of the photodetector. Photodetector active areas 4 are doped regions of a doping type opposite to that of the substrate 2 and may be formed e.g. via implantation or diffusion of doping agents. Photodetector active areas 4 are arranged in an array pattern in the substrate 2.

(14) Conveniently, the photodetector active areas 4 can be designed to define photodiode, phototransistor, or photoresistor active areas. The photons impinging on the photodetector active areas 4 are converted into charge carriers producing an output electric signal proportional to the intensity of the incident light.

(15) For example, photodiode active areas are realized, in the simplest form, as p-n (or n-p) junctions configured so that the n (or p) active regions are depleted from charge carriers (such as electron/hole pairs) and, hence, incident photons generate electron/hole pairs collected by the depletion regions of the photodiodes. As is known, also pnp (or npn) junctions can be conveniently formed.

(16) Gate regions 5 of the photodetection units, made of a suitable metal material, are formed on the front surface 2a of the substrate 2, arranged in a position vertically corresponding to the area of separation between two respective photodetector active areas 4.

(17) The optical sensor 1 further includes a CMOS stack 6 comprised of a number of stacked dielectric layers 6a, e.g. made of a silicon oxide, and conductive layers 6b, e.g. made of a metal material. In the example shown in FIG. 2, two metal layers are shown, but it is clear that a different number of metal layers may be provided. A top dielectric layer 6a defines an external surface of the stack 6, opposite to the front surface 2a of substrate 2.

(18) In a manner not shown, but which will be clear for a skilled person, conductive vias and interconnections are formed in the CMOS stack 6 to define electrical conductive paths towards the photodetector active areas 4 and towards electrical contact pads (also not shown) arranged at the external surface of the stack 6, so that output electric signals generated by the photodetector active areas 4 may be provided to an external electrical system.

(19) In the embodiment shown in FIG. 2, light channelling regions (so called “light pipes”) 7 are formed through the stack 6, from the top dielectric layer 6a towards the front surface 2a of the substrate 2, each arranged on a respective photodetector active area 4 and configured to guide light radiation towards the respective photodetector active area 4. The light pipes 7 are formed of a suitable material transparent to visible light radiation, for example an organic polymer, such as a siloxane polymer.

(20) A passivation layer 8, e.g. of silicon nitride, is formed above the stack 6, on the external surface thereof defined by the top dielectric layer 6a.

(21) According to a particular aspect of the present solution, the optical sensor 1 further includes UV conversion regions 10, here formed on the passivation layer 8, made of a suitable material able to convert UV light radiation into visible light radiation. UV conversion regions 10 are formed of an organic material in which a fluorescent dye is immersed, designed to perform the UV conversion; for example, the organic material may be a siloxane polymer, like Silecs SC-480, and the dye a fluorescent phosphor like the Lumogen. Examples of down conversion materials that can absorb UV light and emit visible light include, for example, the above cited Lumogen, Coronene, AlQ3′, ZnS:Mn, and the like.

(22) UV conversion regions 10 are arranged vertically corresponding to the positions of some of the photodetector active areas 4, which are intended to perform detection of the UV light radiation.

(23) Moreover, filter regions 11 are formed on the passivation layer 8, made of a suitable material transparent to visible light radiation and opaque to the UV light radiation, such as a siloxane organic polymer. Filter regions 11 may be transparent only to a respective portion of the visible spectrum, thus implementing RGB (Red, Green and Blue) filters, designed to filter (in a known manner, here not discussed in details) the light radiation.

(24) Filter regions 11 are alternated to the conversion regions 10 on the passivation layer 8, according to a desired array pattern; in particular, filter regions 11 are arranged vertically corresponding to the positions of the photodetector active areas 4, which are intended to perform detection of the visible light radiation.

(25) The optical sensor 1 thus includes a capping layer 12 on the passivation layer 8, formed by the alternated conversion and filter regions 10, 11.

(26) FIG. 3 shows a schematic top-down view of a possible pattern for the array 15 of photodetection units (or pixels) 16 of the optical sensor 1, each including a respective photodetection area. In this case, the array includes rows and columns wherein pixels for detection of UV light radiation are alternated to pixels for detection of a respective color (Red ‘R’, Green ‘Gr’ or Blue ‘B’) of the visible light radiation.

(27) It is underlined that other patterns of alternating pixels for UV and visible light radiations may be equally envisaged in the array 15, as will be clear for a skilled person; for example, each pixel sensitive to UV light radiation may be alternated to two or more pixels sensitive to visible light radiation, along the rows and/or columns of the array 15.

(28) In any case, the array 15 defines a single image detection area 15′, sensitive to both UV and visible light radiation, at a same time and with a same spatial resolution.

(29) The optical sensor 1, as shown in FIG. 2, may further include micro-lens units 18, arranged on the capping layer 12, vertically corresponding to a respective photodetector active area 4 and pixel of the photodetector, in order to suitably focus (in a known manner, here not discussed in detail) the incoming electromagnetic radiation and obtain a collimated beam towards the photodetector active areas 4.

(30) During operation, light radiation, denoted with 20 in the above FIG. 2, from the external environment, reaches the optical sensor 1 and in particular the capping layer 12, via the micro-lens units 18, if present, and, through the same capping layer 12, reaches the underlying photodetector active areas 4. In particular, all the photodetector active areas 4 receive visible light radiations, either directly through the filter regions 11 (possibly implementing RGB filters), or via the down-conversion performed by the UV conversion regions 10, which convert UV light radiation into visible light radiation.

(31) Photodetector active areas 4 collect the received photons and generate corresponding output electric signals that are routed through the CMOS stack 6 towards the electrical contact pads at the external surface thereof, for communication to an external processing system. In a possible embodiment, comparing the signal collected on the pixels covered with conversion material and the pixels covered with the material absorbing UV radiation and transparent to visible light, it is possible to reconstruct with more accuracy the distribution of UV intensity focused on the sensor surface.

(32) As shown in FIG. 4A, the manufacturing process of the optical sensor 1 first envisages formation of the photodetector active areas 4 within the substrate 2, e.g. via diffusion or implantation of dopant atoms.

(33) The gate regions 5 are then formed on the front surface 2a of the substrate 2, via deposition and photolithographic patterning of a metal layer; subsequently, the CMOS stack 6, including the stacked dielectric layers 6a and conductive layers 6b, is formed, via deposition and patterning of alternated dielectric and metal layers, as shown in FIG. 4B.

(34) The light channelling regions 7 are then formed through the stack 6, as shown in FIG. 4C; in particular, trenches are first formed through the layers of the CMOS stack 6, reaching the front surface 2a of the substrate, and the same trenches are filled with a suitable material, transparent to the visible light, via deposition of a photo-patternable material, which is then removed outside of the same trenches. As in the example shown, top dielectric layer 6a may be formed on the light channelling regions 7.

(35) Afterwards, as shown in FIG. 4D, the passivation layer 8 is formed on the CMOS stack 6, via deposition, and the capping layer 12 is formed thereon, with the filter regions 11 alternated to the UV conversion regions 10.

(36) In particular, a possible embodiment may envisage use of a fluorescent dye dispersed in a photo-patternable organic material layer, to create patches of fluorescent material defining UV conversion regions 10, on top of some of the pixels of the sensor array; the remaining pixels of the sensor array are covered with a transparent material, or with red, green, blue filters. The organic material layer may be removed out of the photoactive area with a photolithographic process.

(37) The micro-lens units 18 may then be formed on the capping layer 12, obtaining the resulting structure shown in FIG. 2.

(38) As shown in FIG. 5, the discussed embodiment may be implemented also for a back-side illuminated (BSI) photodetector, made with CMOS semiconductor techniques.

(39) In this case, the external surface of the CMOS stack 6, again formed on the front surface 2a of the substrate, defines the external surface of the optical sensor 1. The passivation layer 8 and the capping layer 12 are in this case formed on the back surface 2b of the substrate 2, with the micro-lens unit 18 again formed on the capping layer 12, if required.

(40) This embodiment may afford the advantage of guiding the light radiation 20 from the external environment towards the photodetector active areas 4 within the substrate 2, minimizing interaction with metal lines.

(41) A second embodiment of the present solution is now discussed, with reference to FIG. 6, relating again to a front-side illuminated photodetector.

(42) In this second embodiment, the light channelling regions, here denoted with 7′ are filled with a suitable material able to convert UV light radiation into visible light radiation, for example an organic material in which a fluorescent dye is immersed, thereby constituting the UV conversion region 10 of the optical sensor 1, here integrated within the CMOS stack 6; the organic material may be a siloxane polymer, containing the dye, such as a fluorescent phosphor like Lumogen.

(43) The light channelling regions 7′ are here formed through the CMOS stack 6, reaching the front surface 2a of the substrate 2, and moreover through the passivation layer 8, reaching an external surface thereof.

(44) In the shown embodiment, the light channeling regions 7′ are formed vertically corresponding to the positions of some of the photodetector active areas 4, which are intended to perform detection of the UV light radiation; on the contrary, no light channeling regions are here formed on the photodetector active areas 4, which are intended to perform detection of the visible light radiation.

(45) Transparent regions 11′ are in this case formed on the passivation layer 8, alternated to the filter regions 11 according to a desired array pattern.

(46) In particular, in this case, transparent regions 11′ are arranged in positions vertically corresponding the light channeling regions 7′ and are designed to be transparent to the wavelength range of both visible and UV light radiation; while filter regions 11, arranged in positions vertically corresponding the photodetector active areas 4, which are intended to perform detection of the visible light radiation, are made of a suitable material transparent to visible light radiation and opaque to the UV light radiation.

(47) The filter regions 11 may be transparent only to a respective portion of the visible spectrum, thus implementing RGB (Red, Green and Blue) filters, designed to filter (in a known manner, here not discussed in details) the light radiation.

(48) The filter and transparent regions 11, 11′ define in this case the capping layer 12 of the optical sensor 1, arranged on the passivation layer 8.

(49) Also in this second embodiment, the optical sensor 1 therefore includes an array 15 of photodetection units (or pixels) 16, wherein pixels for detection of UV light radiation are alternated to pixels for detection of the visible light radiation, within a single image detection area 15′, sensitive to both UV and visible light radiation, at the same time and with a same spatial resolution.

(50) Moreover, the optical sensor 1 may again include the micro-lens units 18, arranged on the capping layer 12, vertically corresponding to a respective photodetector active area 4 and pixel of the photodetector.

(51) The manufacturing process of the second embodiment again envisages, as shown in FIG. 7A, formation of the photodetector active areas 4 within the substrate 2, e.g. via diffusion or implantation of dopant atoms.

(52) The gate regions 5 are then formed on the front surface 2a of the substrate 2, via deposition and photolithographic patterning of a metal layer; subsequently, the CMOS stack 6, including the stacked dielectric layers 6a and conductive layers 6b, is formed, via deposition and patterning of alternated dielectric and metal layers. Then, as shown in FIG. 7B, the passivation layer 8 is formed on the CMOS stack 6.

(53) The light channelling regions 7′ and corresponding UV conversion regions 10 are formed through the stack 6 and passivation layer 8, as shown in FIG. 7C, vertically corresponding to the positions of the photodetector active areas 4, which are intended to perform detection of the UV light radiation; in particular, trenches are first formed through the CMOS stack 6 and passivation layer 8, starting from its external surface and reaching the front surface 2a of the substrate 2, and then the same trenches are filled with the suitable material, designed for conversion of UV light radiation into visible light radiation, via deposition of a photo-patternable material, which is removed outside of the same trenches.

(54) Afterwards, the capping layer 12 is formed on the passivation layer 8, as shown in FIG. 7D, with the alternated filter and transparent regions 11, 11′.

(55) The micro-lens units 18 may then be formed on the capping layer 12, thereby obtaining the structure shown in FIG. 6.

(56) The advantages that the proposed solution allows to achieve are clear from the foregoing disclosure.

(57) In any case, it is again underlined that the disclosed solution offers a number of improvements compared to prior art solutions, among which: the possibility to reconstruct UV and visible light starting from a same focused image on the chip, containing information coming from both spectral ranges, with a same spatial resolution; an easier and cheaper integration; a smaller area occupation, leading to a more compact system.

(58) The disclosed second embodiment may be advantageous, since it allows to detect visible light across a deeper active area, since the UV conversion regions 10 are distributed within the whole depth of the light channelling regions 7′.

(59) The manufacturing process of the semiconductor optical sensor 1 generally requires additional steps that are performed advantageously at the BEOL, or back-end, steps of a standard CMOS manufacturing process, thus not requiring extensive modifications to common processes.

(60) The features discussed above are particularly advantageous in case the semiconductor optical sensor 1 is embedded inside a mobile device, like a smart phone, a tablet, or in general a mobile device or handset; indeed, the discussed solution allows to achieve the low cost and low size requirements that are generally required to mobile devices.

(61) Finally, it is clear that modifications and variations may be made to what has been described and illustrated herein, without thereby departing from the scope of the present invention, as defined in the annexed claims.

(62) In particular, it is again underlined that different patterns may be envisaged in the array 15 of the optical sensor 1 in the image detection area 15′, with respect to the alternated arrangement of the pixels designed to detect visible light radiation and those to detect UV radiation.

(63) Moreover, in a further possible embodiment, shown in FIG. 8, which is otherwise similar to the second embodiment of FIG. 6, the light channelling regions 7′ may be provided on all the photodetector active areas 4.

(64) It is also underlined that the discussed solution is suitable for both front-side and back-side illuminated sensors, and various type of photodetectors, including for example p-n junctions, p-i-n detectors or SPAD avalanche photodiodes.