Method of monolithic integration of hyperspectral image sensor

This application claims priority to Chinese Patent Application No. 201610214392.0, titled “METHOD OF MONOLITHIC INTEGRATION OF HYPERSPECTRAL IMAGE SENSOR,” filed with the State Intellectual Property Office of the People's Republic of China on Apr. 7, 2016, which is hereby incorporated by reference in its entirety.

The present disclosure relates to the technical field of hyperspectral imaging, and in particular to a method for monolithic integration of a hyperspectral image sensor.

With help of hyperspectral imaging technology, a continuous spectrum can be divided into multiple spectral bands containing certain range of wavelengths, therefore, desired spectral bands can be further analyzed individually more in detail. Hence, hyperspectral imaging technology is widely applied to technical fields such as remote sensing, food science and bio-medical research. A previous hyperspectral imaging system is based on laboratory setups, therefore it slows in analysis speed, high in cost and not compact at all. In order to solve the above problems, a hyperspectral imager was implemented by monolithically integrating a wedge filter on top of a CMOS image sensor by the researchers of Interuniversity Microelectronics Center (IMEC).

As presented by IMEC, the method for fabricating a hyperspectral image sensor is by monolithically integrating a wedge filter on top of the photosensitive region of a CMOS image sensor. Such kind of hyperspectral image sensors, compared to the laboratory style ones, have much smaller volume, faster analysis speed and lower cost. A filter integrated on the top of the CMOS image sensor is generally a set of Fabry-Pérot interferometers (Fabry-Pérot interferometers), as shown in FIG. 1 . The Fabry-Péinterferometers include two highly reflective layers (mirrors), a bottom reflecting layer 11 and a top reflecting layer 12 ; and a transparent cavity layer is placed in between the two reflecting layers 11 and 12 . The transparent cavity layer 13 is a wedge structure, and each step height corresponds to one spectral band.

The transparent cavity layer defines the cavity length where light can be reflected in between the two mirrors.

In conventional technology as IMEC implemented, the transparent cavity layer with step structures is fabricated by alternating photolithography-etching steps. For example, a transparent cavity layer with eight step heights is fabricated with three sets of photolithography-etching processes, as shown in FIG. 2 .

A height of the cavity layer determines a wavelength of a central spectrum of a corresponding filter, hence in a multi-step Fabry-Pérot interferometer (for example, one with more than 100 step structures), the heights of the cavity layer must be fabricated very precisely. Practically, it is reported that etching depths have within wafer non-uniformity of at least about 2.7% during each etching process, which means that the more etching steps are performed, the greater the non-uniformity. As a transparent cavity layer composed of multiple step structures requires multiple etching steps, thus, accumulation of the non-uniformity increases as the etching steps increases. Therefore, the device may be failed due to being out of the specifications.

In addition, it is difficult to apply plasma etching or wet etching to etch some specific materials, for example, hafnium(IV) oxide (HfO 2 ). Since the boiling points of the halides of the lanthanide transition metals (including Hf) are high (higher than 300 degrees Celsius), it is difficult to etch those halides up through plasma etching. A reagent based on containing rich hydrogen fluoride (HF) molecules may be used for wet etching of HfO 2 , however the reagent containing HF normally will destroy the interface layer of photoresist and HfO 2 . Therefore, it may result of lifting off the photoresist. To overcome the issue, a hard mask has to be added, thereby it increases process steps, cost and uncertainties. Thus, if the transparent cavity layer is fabricated by photolithography-etching processes, the selection of materials for the transparent cavity layer is restricted due to etching process.

In order to solve the problem of bad height uniformity of step structures and device failure caused by etching processes and extend the selection range of materials for the transparent cavity layer, a method for monolithic integration of a hyperspectral image sensor is provided according to the present disclosure.

In order to solve the above technical problems, the following technical solution is provided according to the present disclosure.

A method for monolithic integration of a hyperspectral image sensor is provided, which includes:

preprocessing the surface of the photosensitive region of a CMOS image sensor wafer, to flatten the surface of the photosensitive region;

forming a bottom reflecting layer on the preprocessed surface of the photosensitive region of the CMOS image sensor wafer;

forming a transparent cavity layer on the bottom reflecting layer by using area selective atomic layer deposition (ASALD) processes, where the transparent cavity layer is composed of N step structures, where N=2 m , m≥1 and m is a positive integer; and

forming a top reflecting layer on the transparent cavity layer.

Compared with conventional technology, the invention has the following beneficial effects.

As the atomic layer deposition process has a feature of excellent deposition uniformity, hence in the present disclosure, the N step structures formed on the bottom reflecting layer by using the area selective atomic layer deposition (ASALD) process have the same height within wafer. Thereby, with the method for monolithic integration of a hyperspectral image sensor according to the present disclosure, the problem of height non-uniformity of step structures within wafer is greatly minimized. Heights of the formed step structures are determined by conditions of the area selective atomic layer deposition process, for example, the number of deposition cycles. And, film deposition non-uniformity (<1.5%) of the ALD process is much smaller than that of the photolithography-etching process (2.7%) in practical (for example, ALD-Al 2 O 3 films). Hence, with the method for monolithic integration of a hyperspectral image sensor according to the present disclosure, the problem of non-uniformity accumulation due to the etching process in conventional technology is greatly minimized. Thereby, it is beneficial to reduce device failure rate so to increase the yield.

In addition, the ALD process can be used to generate various types of film materials which are not necessarily to be etched easily. Hence, compared with the photolithography-etching process, the method for monolithic integration of a hyperspectral image sensor according to the present disclosure can be applied more widely in the aspect of material selection of a cavity layer.

In order for a clearer understanding of the embodiments according to the present disclosure, drawings to be used in the description of the embodiments according to the present disclosure are described briefly hereinafter. Apparently, these drawings only describe some embodiments of the invention, and other drawings may be obtained by those skilled in the art without any creative work.

FIG. 1 is a cross section structure of a Fabry-Pérot interferometer according to a related technical filed;

FIG. 2 is a schematic diagram showing cross section of the structures corresponding to fabrication processes of a transparent cavity layer with eight step structures fabricated by using photolithography-etching processes according to conventional technology;

FIG. 3 is a flowchart of an area selective atomic layer deposition process according to the present disclosure;

FIGS. 4A to 4C are cross section of the structures according to a series of fabrication processes of the area selective atomic layer deposition process according to an embodiment of the present disclosure;

FIG. 5 is a flowchart of a specific implementation of step S 301 in a case that a reagent used for the surface hydrophobic treatment is the surface passivator, according to an embodiment of the present disclosure;

FIGS. 6A to 6D are cross section of the structures according to a series of fabrication processes of a specific implementation of step S 301 in a case that a reagent used for the surface hydrophobic treatment is the surface passivator, according to an embodiment of the present disclosure;

FIG. 7 is a flowchart showing a series of fabrication processes of a specific implementation of step S 301 in a case that the reagent used for the surface hydrophobic treatment is a kind of polymer, according to an embodiment of the present disclosure;

FIGS. 8A and 8B are cross section of the structures according to a series of fabrication processes of a specific implementation of step S 301 in a case that the reagent used for the surface hydrophobic treatment is a kind of polymer, according to an embodiment of the present disclosure;

FIG. 9 is a flowchart of a method for fabrication of a hyperspectral imager according to an embodiment of the present disclosure;

FIGS. 10A to 10D are cross section of the structures according to a series of fabrication processes of a method for monolithic integration of a hyperspectral image sensor according to an embodiment of the present disclosure;

FIG. 11 is a flowchart of a first specific implementation of step S 903 according to an embodiment of the present disclosure;

FIGS. 12A to 12J are cross section of the structures according to a series of fabrication processes in the first specific implementation of step S 903 according to an embodiment of the present disclosure;

FIG. 13 is a flowchart of a second specific implementation of step S 903 according to an embodiment of the present disclosure.

FIGS. 14A to 14K are cross section of the structures according to a series of fabrication processes in the second specific implementation of step S 903 according to an embodiment of the present disclosure;

FIG. 15 shows a flowchart of a method for forming a snapshot mosaic transparent cavity layer having four 4×4 snapshot mosaic areas; and

FIG. 16A to FIG. 16G are three dimensional of the structures according to a series of fabrication processes for forming a snapshot mosaic transparent cavity layer having four 4×4 snapshot mosaic areas.

The embodiments according to the present disclosure are described in detail in conjunction with the drawings hereinafter, so that the object, technical solutions and technical effects of the present disclosure can be understood clearly.

Before the embodiments according to the present disclosure are described, an area selective atomic layer deposition (Area Selective Atomic Layer Deposition, abbreviated as ASALD) process is introduced first.

The area selective atomic layer deposition process is a kind of atomic layer deposition process on selected surface where water is one of the precursors.

A material grown with water as one of the precursors generally grows well only on hydrophilic surfaces. In addition, the material grown with water as one of the precursors grows differently on different hydrophobic surfaces. The greater the hydrophobicity of a surface is, the higher the density of a hydrophobic material is, it is more difficult to grow the material with water as one of the precursors. The hydrophobicity of a surface is generally measured by contact angle.

Besides, for a Fabry-Pérot interferometer, in order to be compatible with a conventional CMOS process line with aspect of issues such as contamination, a bottom reflecting layer of the Fabry-Pérot interferometer is generally a Bragg stack which is a composite film layer of silicon oxide and silicon nitride. Hence, a surface of the Bragg stack in contacting with a transparent cavity layer may be processed with SCl (a mixture of ammonium hydroxide solution, hydrogen peroxide solution and water), thereby formation of a relatively high hydrophilicity of the hydrophilic layer because of Si—OH (silanol) groups. Since Si—OH will be replaced with non-hydrophilic Si—O groups gradually in a relatively dry environment, in a case that the Bragg stack surface processed with SCl before is re-processed with a reagent used for the surface hydrophobic treatment, the Si—O groups on the hydrophilic layer formed by Si—OH can be replaced with the hydrophobic groups, thereby forming a high density hydrophobic surface on the surface of the Bragg stack. Since it is difficult for the material grown with water as one of the precursors to grow on a hydrophobic surface, the material layer can grow only on the unprocessed hydrophilic surface not on the hydrophobic surface in the subsequent atomic layer deposition process. In the present disclosure, the stair-like transparent cavity layer is grown based on the principle. Hence, in the embodiments according to the present disclosure, a material composing the transparent cavity layer is grown by the atomic layer deposition process with water as one of the precursors.

For example, the material grown by the atomic layer deposition process with water as one of the precursors may include at least one of: B 2 O 3 , MgO, Al 2 O 3 , SiO 2 , CaO, Sc 2 O 3 , TiO 2 , VO x , CrO x , FeO x , CoO x , NiO, CuO, ZnO, Ga 2 O 3 , GeO 2 , SrO/SrCO 3 , Y 2 O 3 , ZrO 2 , Nb 2 O 5 , In 2 O 3 , Sb 2 O 5 , La 2 O 3 , PrO x , Lu 2 O 3 , HfO 2 , Ta 2 O 5 , WO 3 and BiO x . As an example, in a case that the material grown by using the atomic layer deposition process with water as one of the precursors is a compound, the material may be Hf a Si b O c etc.

Generally, in order to be compatible with the conventional CMOS BEOL process line with aspect of contamination, the material grown by using the atomic layer deposition process with water as one of the precursors may include Al 2 O 3 , SiO 2 , TiO 2 , HfO 2 and Ta 2 O 5 , typically.

For example, FIG. 3 shows an area selective atomic layer deposition process according to the present disclosure, which includes steps S 301 to S 303 in the following.

In step S 301 , a surface of a substrate is processed with a reagent used for the surface hydrophobic treatment, to process a surface region of the substrate, on which a material with water as one of the precursors is not to be grown, into a hydrophobic surface region.

FIG. 4A shows the cross section of the structure corresponding to step S 301 . Specifically, step S 301 includes: performing surface processing on the surface of the substrate 401 with a reagent used for the surface hydrophobic treatment by covering part of the surface of the substrate, on which the material with water as one of the precursors is not to be grown, with a reagent used for the surface hydrophobic treatment 402 . In FIG. 4A , a region covered with the reagent used for the surface hydrophobic treatment 402 is the surface region of the substrate on which the material with water as one of the precursors is not to be grown.

In a case that the substrate is a Bragg stack, before the surface of the substrate is processed with the reagent used for the surface hydrophobic treatment, a surface of the Bragg stack may be preprocessed in advance. The preprocessing may include: preprocessing the surface of the Bragg stack with SCl (a mixture of ammonium hydroxide solution, hydrogen peroxide solution and water) to obtain a hydrophilic layer with relatively high hydrophilicity because of formation of Si—OH (silanol) layer. After the preprocessing, the preprocessed surface of the Bragg stack is processed with the reagent used for the surface hydrophobic treatment. During the hydrophobic treating process, Si—OH groups of the hydrophilic layer formed by Si—OH are being replaced with hydrophobic groups, thereby forming a high density hydrophobic surface on the surface of the Bragg stack.

In step 302 , an area selective atomic layer deposition is performed on the processed surface of the substrate, to form an atomic layer deposition material layer on a non-hydrophobic surface region.

FIG. 4B shows the cross section of the structure corresponding to step S 302 . In performing the area selective atomic layer deposition on the processed surface of the substrate, a deposited material can be grown only on part of the surface of the substrate covered without the reagent used for the surface hydrophobic treatment 402 , and no material can be grown on the surface covered with the reagent used for the surface hydrophobic treatment 402 . In this way, after step S 302 is performed, an atomic layer deposition material layer 403 is formed on the non-hydrophobic surface region of the substrate 401 .

In step S 303 , the reagent used for the surface hydrophobic treatment on the hydrophobic surface region is removed.

The reagent used for the surface hydrophobic treatment covered on the surface of the substrate is removed with an appropriate method based on the characteristic of the reagent used for the surface hydrophobic treatment, i.e., removing the reagent used for the surface hydrophobic treatment on the hydrophobic surface region. FIG. 4C shows the corresponding cross sectional diagram after step S 303 is completed.

In the embodiment of the present disclosure, the reagent used for the surface hydrophobic treatment may be a surface passivator or a patternable polymer. The patternable polymer may be polymethyl methacrylate (PMMA). In some specific embodiments of the present disclosure, step S 301 may be performed through the following two implementations depending on different reagents used for the surface hydrophobic treatment.

First Implementation:

FIG. 5 is a flowchart of a specific implementation of step S 301 in a case that a reagent used for the surface hydrophobic treatment is the surface passivator. As shown in FIG. 5 , the specific implementation includes steps S 301 A 1 to S 301 A 4 in the following.

In step S 301 A 1 , the surface of the substrate is coated with a surface passivator to form a self-assembled monolayer (Self-Assembled Monolayer, abbreviated as SAM) on the whole surface of the substrate.

For example, the surface of the substrate may be coated with the surface passivator by spraying. In another embodiment of the present disclosure, in order to form the self-assembled monolayer on the whole surface of the substrate, the substrate may be soaked in the surface passivator, which is generally applied in laboratories.

FIG. 6A shows a cross section of the structure corresponding to step S 301 A 1 . In FIG. 6A, 601 represents the substrate, and 602 represents the self-assembled monolayer formed with the surface passivator.

In step S 301 A 2 , the self-assembled monolayer is coated with photoresist and the photoresist is baked.

For example, the self-assembled monolayer may be coated with the photoresist by spin-coating in the present disclosure. Then, the photoresist on the self-assembled monolayer is baked, to evaporate the organic solvent in the photoresist.

FIG. 6B shows a cross section of the structure corresponding to step S 301 A 2 . In FIG. 6B, 603 represents the photoresist on the self-assembled monolayer 602 .

In step S 301 A 3 , exposing, developing and weak plasma stripping are performed on the photoresist, to divide the surface of the substrate into a region covered with the photoresist and a region covered without the photoresist, and then the self-assembled monolayer on the region covered without the photoresist is removed.

Specifically, the photoresist 603 is exposed by using a mask with a preset pattern, to transfer the pattern of the mask to the photoresist and then developed, and a soluble region of the photoresist, which is formed by the exposing, is dissolved with a chemical developer, to remove the region of the photoresist which can be dissolved by the developer, thereby dividing the surface of the substrate 601 into the region covered with the photoresist and the region covered without the photoresist. The region covered without the photoresist is coated with the self-assembled monolayer.

After the developing is performed, the self-assembled monolayer on the region covered without the photoresist needs to be removed. FIG. 6C shows the corresponding cross section of the structures after step S 301 A 3 is completed.

For example, a self-assembled monolayer composed of some surface passivation layers may be removed with developers, for example, a self-assembled monolayer composed of hexamethyldisilazane (HMDS). In a case of other passivation layers, the self-assembled monolayer may be oxidized and removed by using a weak plasma stripping process with most of the photoresist remained.

In step S 301 A 4 , the remained photoresist is stripped, to expose the self-assembled monolayer where was covered by the remained photoresist, thereby forming the self-assembled monolayer on the region where was covered with the photoresist, where a surface region of the substrate, on which the self-assembled monolayer is formed, is the hydrophobic surface region.

Since photoresists can be dissolved by organic solvents, the remained photoresist may be stripped by dissolving with organic solvents, such as acetone.

As the organic solvent does not destroy the self-assembled monolayer, the self-assembled monolayer covered by the photoresist is exposed after the photoresist is stripped, thereby forming the self-assembled monolayer on the region where was covered with the photoresist. The surface region of the substrate, on which the self-assembled monolayer is formed, is hydrophobic. FIG. 6D shows the corresponding cross section of the structure after step S 301 A 4 is completed.

It should be noted, in case that a reagent used for the surface hydrophobic treatment is the surface passivator, as shown in step S 303 , the reagent used for the surface hydrophobic treatment on the hydrophobic surface region, i.e., the surface passivator, may be removed by plasma ashing.

In the embodiment of the present disclosure, the surface passivator may include any of the following compounds:

ODTS, octadecyltrichlorosilane;

HMDS, hexamethyl disilazane;

ODTM, octadecyltrimethoxysilane;

ODTE, octadecyltriethoxysilane; and

FOTS, tridecafluoro-1,1,2,2-Tetrahydrooctyltrichlorosilane.

Second Implementation:

FIG. 7 shows a flowchart of the specific implementation of step S 301 in case that the reagent used for the surface hydrophobic treatment is a patternable polymer. As shown in FIG. 7 , the specific implementation includes from steps S 301 B 1 to S 301 B 4 .

In step S 301 B 1 , the substrate is coated with a polymer film.

For example, the substrate 701 may be coated with a polymer film 702 by spin-coating. FIG. 8A shows the corresponding cross section of the structure after step S 301 B 1 is completed.

In a specific example, the polymer film may be made of a polymer dissolved in toluene. A mass fraction of the polymer in the toluene solution may be 1-5 wt. %.

In step 301 B 2 , the polymer film on the surface of a preset region of the substrate is exposed, where the preset region is the hydrophobic surface region to be formed.

Step S 301 B 2 may include exposing the polymer film 702 on the surface of the preset region of the substrate by using a mask with a preset pattern, where the preset region is the hydrophobic surface region to be formed.

For example, the polymer film 702 may be exposed in a deep ultra violet environment (the wavelength is 248 nm).

In step S 301 B 3 , the polymer film which is not exposed is removed.

It should be noted that the exposed polymer film can not be removed. FIG. 8B shows the corresponding cross section of the structure after step S 301 B 3 is completed.

In a specific embodiment, the unexposed polymer film may be removed with a mixed solution of isopropyl alcohol and methyl isobutyl ketone, where a volume ratio of the isopropyl alcohol and the methyl isobutyl ketone may be 1:1.

In step S 301 B 4 , soft baking is performed on the polymer film, afterwards vacuum baking is performed to remove moisture in the polymer film, to keep the polymer film dry.

A specific process condition of step S 301 B 4 may perform soft baking on the polymer film for 5 minutes at 120 degrees Celsius, a temperature of the vacuum baking may be 100 degrees Celsius and a duration of the vacuum baking may be two hours.

In the embodiment of the present disclosure, the patternable polymer may be polymethyl methacrylate which is abbreviated as PMMA.

It should be noted, in case that the reagent used for the surface hydrophobic treatment is a kind of polymer, in step S 303 , the reagent used for the surface hydrophobic treatment on the hydrophobic surface region, i.e., the polymer, may be removed by plasma ashing or dissolving with organic solvents, for example acetone.

The above-described is specific implementations of the area selective atomic layer deposition process used in the embodiment of the present disclosure.

For a method for monolithic integration of a hyperspectral image sensor according to an embodiment of the present disclosure, the substrate described above is a CMOS image sensor wafer with a size of 12 inches or below.

Some specific implementations of the method for monolithic integration of a hyperspectral image sensor according to the present disclosure are described hereinafter in detail in conjunction with the drawings and the area selective atomic layer deposition process described above.

FIG. 9 is the flowchart of a method for monolithic integration of a hyperspectral image sensor according to an embodiment of the present disclosure. As shown in FIG. 9 , the method includes from steps S 901 to S 904 .

In step S 901 , the surface of the photosensitive region of a CMOS image sensor wafer is preprocessed, to flatten the surface of the photosensitive region.

FIG. 10A shows the cross section of the structure corresponding to step S 901 . Step S 901 includes: flattening a whole CMOS image sensor wafer 101 (including the surface of the photosensitive region 102 ) by using a chemical mechanical polishing (CMP) process, to make the surface of the photosensitive region flat.

In step S 902 , a bottom reflecting layer is formed on the preprocessed surface of the photosensitive region of the CMOS image sensor wafer.

FIG. 10B shows the cross section of the structure corresponding to step S 902 . As shown in FIG. 10B , a bottom reflecting layer 103 is formed on the surface of the photosensitive region 102 of the CMOS image sensor wafer. The bottom reflecting layer 103 is a bottom reflecting layer of a Fabry-Pérot interferometer to be formed. The bottom reflecting layer 103 has high reflectivity.

In step S 903 , a transparent cavity layer is formed on the bottom reflecting layer by using the area selective atomic layer deposition process.

FIG. 10C shows a cross section of the structure corresponding to step S 903 . For example, a transparent cavity layer 104 formed is composed of eight step heights structures. Practically, in the embodiment of the present disclosure, the transparent cavity layer 104 formed is not limited to be composed of eight step heights structures, and may be composed of N step heights structures where, N=2 m , m≥1, and m is a positive integer.

In the embodiment of the present disclosure, the transparent cavity layer 104 is a space in which light can be reflected between two reflective layers with high reflectivity. In the transparent cavity layer, each step height structure corresponds to a spectral band of one wavelength range, hence one spectrum containing multiple wavelengths can be divided into spectral bands of the multiple wavelengths by each step height structure, thereby to realize accurate data analyzing for respective spectral bands.

In the embodiment of the present disclosure, N step heights structures can be formed on the bottom reflecting layer by using the area selective atomic layer deposition process for log 2 N+1 =m+1 times, where the N step structures form the transparent cavity layer. The specific implementation is described in detail hereinafter in conjunction with FIGS. 11 to 14J .

In the embodiment of the present disclosure, the N step heights structures may be formed on the bottom reflecting layer by using the area selective atomic layers deposition process for log 2 N+1 =m+1 times through two specific implementations.

First Specific Implementation

For example, it is set N=8. That is, the transparent cavity layer is composed of eight step heights structures. In a case that N=8, the area selective atomic layer deposition process needs to be performed for log 2 8 +1=3+1=4 times. In this case, m=3.

FIG. 11 shows the flowchart of the first specific implementation, which includes steps from S 111 to S 1110 .

In step S 111 , an initial material layer is deposited on the whole surface of the bottom reflecting layer, where the initial material layer has a thickness equal to the height of the bottom step of a transparent cavity layer to be formed.

As shown in FIG. 12A , step S 111 includes: depositing an initial material layer 1041 on the whole surface of the bottom reflecting layer 103 by using the ASALD process, where the initial material layer 1041 has a thickness equal to the height of the bottom step of the transparent cavity layer to be formed.

In step S 111 , the first area selective atomic layer deposition process for forming the transparent cavity layer is performed.

In step S 112 , the surface of the bottom reflecting layer is divided equally into 2 (2 1 ) first subregions 1 A and 1 B, and surface processing is performed on the surface of the first subregion 1 B, to coat the surface of the first subregion 1 B with a reagent used for the surface hydrophobic treatment 121 , thereby forming a hydrophobic surface region on the surface of the first subregion 1 B.

It should be noted that a reference may be made to related descriptions in step S 301 for specific implementation to perform surface processing on the surface of the first subregion 1 B, to coat the surface of the first subregion 1 B with the reagent used for the surface hydrophobic treatment 121 , thereby forming the hydrophobic surface region on the surface of the first subregion 1 B. FIG. 12B shows a cross section of the structure corresponding to step S 112 .

In step S 113 , an area selective atomic layer deposition is performed on the first subregion 1 A, to form a material layer structure 1042 corresponding to a (1+1) th area selective atomic layer deposition process.

The material layer structure 1042 grown through the (1+1) th area selective atomic layer deposition process has a thickness equal to a height of 4 (2 (3-1) ) common steps. FIG. 12C shows a cross section of the structure corresponding to step S 113 . A height of a common step is a height of a single step in the transparent cavity layer to be formed except the bottom step.

In step S 114 , the reagent used for the surface hydrophobic treatment 121 on the first subregion 1 B is removed.

FIG. 12D shows a cross section of the structure corresponding to step S 114 .

From step S 112 to step S 114 , the second area selective atomic layer deposition process is performed. Hereinafter a third area selective atomic layer deposition process is performed.

In step S 115 , the surface of the bottom reflecting layer is divided equally into 4 (2 2 ) second subregions 2 A to 2 D, and surface processing is performed on surfaces of the second subregions 2 B and 2 D, to coat the surface of the second subregions 2 B and 2 D with the hydrophobic treatment agent 121 , thereby forming hydrophobic surface regions on the surfaces of the second subregions 2 B and 2 D.

Practically, step S 115 may be understood as follows: dividing each first subregion equally into two second subregions, and forming a hydrophobic surface region on a surface of a second subregion every another second subregion. FIG. 12E shows a cross section of the structure corresponding to step S 115 .

In step S 116 , an area selective atomic layer deposition is performed on the second subregions 2 A and 2 C, to form a material layer structure 1043 corresponding to a (1+2) th area selective atomic layer deposition process.

The material layer structure 1043 grown through the (1+2) th area selective atomic layer deposition process has a thickness equal to a height of 2 (2 (3-2) ) common steps. FIG. 12F shows a cross section of the structure corresponding to step S 116 .

In step S 117 , the reagent used for the surface hydrophobic treatment 121 on the second subregions 2 B and 2 D is removed.

FIG. 12G shows a cross section of the structure corresponding to step S 117 .

From step S 115 to step S 117 , the third area selective atomic layer deposition process is performed. Hereinafter a fourth area selective atomic layer deposition process is performed.

In step S 118 , the surface of the bottom reflecting layer is divided equally into 8(2 3 ) third subregions 3 A to 3 H, and surface processing is performed on surfaces of the third subregions 3 B, 3 D, 3 F and 3 H, to coat the surfaces of the third subregions 3 B, 3 D, 3 F and 3 H with the reagent used for the surface hydrophobic treatment 121 , thereby forming hydrophobic surface regions on the surfaces of the third subregions 3 B, 3 D, 3 F and 3 H.

Practically, step S 118 may be understood as follows: dividing each second subregion equally into two third subregions, and forming a hydrophobic surface region on a surface of a third subregion every another third subregion. FIG. 12H shows a cross section of the structure corresponding to step S 118 .

In step S 119 , an area selective atomic layer deposition is performed on the third subregions 3 A, 3 C, 3 E and 3 G to form a material layer structure corresponding to the (1+3) th area selective atomic layer deposition process.

The material layer structure 1044 grown through the (1+3) th area selective atomic layer deposition process has a thickness equal to a height of 1(2 (3-3) ) common step. FIG. 12I shows a cross section of the structure corresponding to step S 119 .

In step S 1110 , the reagent used for the surface hydrophobic treatment 121 on the third subregions 3 B, 3 D, 3 F and 3 H is removed.

FIG. 12J shows a cross section of the structure corresponding to step S 1110 .

From step S 118 to step S 1110 , the fourth area selective atomic layer deposition process is performed.

The above-described is an example of the first specific implementation of step S 903 in a case that N=8. Practically, in the embodiment of the present disclosure, the number N of step structures in the transparent cavity layer may be 2 m , where m is a positive integer.

In a case that the number of step structures in the transparent cavity layer is N=2 m , the first specific implementation may be summarized as follows based on the example in which N=8.

The forming N step structures on the bottom reflecting layer by using the area selective atomic layer deposition process for (m+1) times includes: performing a first area selective atomic layer deposition process and performing a (1+i) th area selective atomic layer deposition process, where i∈{1, m}, and i is an integer.

The first area selective atomic layer deposition process includes:

depositing an initial material layer on a whole surface of the bottom reflecting layer, where the initial material layer has a thickness equal to the height of a bottom step of a transparent cavity layer to be formed.

The (1+i) th area selective atomic layer deposition process includes:

dividing the surface of the bottom reflecting layer equally into 2 i i th subregions; and

performing an area selective atomic layer deposition on an i th subregion every the other i th subregions, to form a material layer structure corresponding to the i th area selective atomic layer deposition process.

The material layer structure grown through the i th area selective atomic layer deposition process has a thickness equal to a height of 2 (m-i) common steps, and a height of a common step is a height of a single step in the transparent cavity layer to be formed except the bottom step.

Second Specific Implementation

For example, it is set N=8. That is, the transparent cavity layer is composed of eight step structures. In a case that N=8, the area selective atomic layer deposition process needs to be performed for log 2 8 +1=3+1=4 times. In this case, m=3.

FIG. 13 shows a flowchart of the second specific implementation, which includes from steps S 131 to S 1311 .

In step S 131 , a surface of the bottom reflecting layer is equally divided into 8 (2 3 ) subregions a to h.

FIG. 14A shows a corresponding cross section of the structure after step S 131 is completed.

In step S 132 , an initial material layer is deposited on each subregion, where the initial material layer has a thickness equal to the height of a bottom step in a transparent cavity layer to be formed.

As shown in FIG. 14B , step S 132 includes: depositing an initial material layer 1041 ′ on a whole surface of the bottom reflecting layer 103 by using the ASALD process, where the initial material layer 1041 ′ has a thickness equal to the height of the bottom step in the transparent cavity layer to be formed.

In step S 132 , the first area selective atomic layer deposition process is performed to form the transparent cavity layer.

In step S 133 , surfaces of subregions b, d, f and h are coated with a reagent used for the surface hydrophobic treatment 121 ′, to form hydrophobic surfaces on the surfaces of the subregions b, d, f and h.

Step S 113 may be understood as follows: coating a surface of 1(2 (1-1) ) subregion with the reagent used for the surface hydrophobic treatment every other 1(2 (1-1) ) subregion, to form a hydrophobic surface every other subregion. FIG. 14C shows a corresponding cross section of the structure after step S 113 is completed.

In step S 134 , an area selective atomic layer deposition is performed on surfaces of subregions a, c, e and g, to form a material layer structure 1042 ′ corresponding to a (1+1) th area selective atomic layer deposition process.

Step S 134 may be understood as follows: performing the area selective atomic layer deposition performed on 1(2 (1-1) ) subregion every other 1(2 (1-1) ) subregion, to form the material layer structure 1042 ′ corresponding to the (1+1=2) th area selective atomic layer deposition process. The material layer structure grown through the second area selective atomic layer deposition process has a thickness equal to a height of 1(2 (1-1) ) common step. FIG. 14D shows the corresponding cross section of the structure after step S 134 is completed.

In step S 135 , the reagent used for the surface hydrophobic treatment on the surfaces of the subregions b, d, f and h is removed.

FIG. 14E shows the corresponding cross section of the structure after step S 135 is completed.

From step S 133 to step S 135 , the second area selective atomic layer deposition process of the second specific implementation is performed. Hereinafter a third area selective atomic layer deposition process is performed.

In step S 136 , subregions c, d, g and h are coated with the reagent used for the surface hydrophobic treatment 121 ′, to form hydrophobic surfaces on the surfaces of the subregions c, d, g and h.

Step S 136 may be understood as follows: coating surfaces of 2 (2 (2-1) ) subregions with the reagent used for the surface hydrophobic treatment every other 2 (2 (2-1) ) subregions, to form hydrophobic surfaces every other 2 subregions.

FIG. 14F shows the corresponding cross section of the structure after step S 136 is completed.

In step S 137 , an area selective atomic layer deposition is performed on the surfaces of subregions a, b, e and f, to form a material layer structure 1043 ′ corresponding to a (1+2) th area selective atomic layer deposition process.

Step S 137 may be understood as follows: perform the area selective atomic layer deposition is performed on 2(2 (2-1) ) subregions every other 2(2 (2-1) ) subregions, to form the material layer structure corresponding to the third area selective atomic layer deposition process. The material layer structure grown through the third area selective atomic layer deposition process has a thickness equal to a height of 2(2 (2-1) ) common steps.

FIG. 14G shows the corresponding cross section of the structure after step S 137 is completed.

In step S 138 , the reagent used for the surface hydrophobic treatment 121 ′ on the surfaces of the subregions c, d, g and h is removed.

FIG. 14H shows the corresponding cross section of the structure after step S 138 is completed.

From step S 136 to step S 138 , the third area selective atomic layer deposition process of the second specific implementation is performed. Hereinafter a fourth area selective atomic layer deposition process is performed.

In step S 139 , subregions e, f, g and h are coated with the reagent used for the surface hydrophobic treatment 121 ′, to form hydrophobic surfaces on the surfaces of the subregions e, f, g and h.

Step S 139 may be understood as follows: coating surfaces of 4 (2 (3-1) ) subregions with the reagent used for the surface hydrophobic treatment every other 4 (2 (3-1) ) subregions, to form hydrophobic surfaces every the other four subregions.

FIG. 14I shows the corresponding cross section of the structure after step S 139 is completed.

In step S 1310 , an area selective atomic layer deposition is performed on the surfaces of subregions a, b, c and d, to form a material layer structure 1044 ′ corresponding to a (1+ 3 ) th area selective atomic layer deposition process.

Step S 1310 may be understood as follows: perform the area selective atomic layer deposition on surfaces of 4 (2 (3-1) ) subregions every other 4 (2 (3-1) ) subregions, to form the material layer structure 1044 ′ corresponding to the fourth area selective atomic layer deposition process. The material layer structure grown through the fourth area selective atomic layer deposition process has a thickness equal to a height of 4 (2 (3-1) ) steps.

FIG. 14J shows the corresponding cross section of the structure after step S 1310 is completed.

In step S 1311 , the reagent used for the surface hydrophobic treatment 121 ′ on the surfaces of subregions e, f, g and h is removed.

FIG. 14K shows the corresponding cross section of the structure after S 1311 is completed.

From step S 139 to step S 1311 , the fourth area selective atomic layer deposition process of the second specific implementation is performed.

The above-described is an example of the second specific implementation of step S 903 in a case that N=8. Practically, in the embodiment of the present disclosure, the number of step structures in the transparent cavity layer may be 2 m , where m is a positive integer.

In a case that the number of step structures in the transparent cavity layer is N=2 m , the second specific implementation above may be summarized as follows based on the example in which N=8.

A surface of the bottom reflecting layer is divided equally into N subregions.

A first area selective atomic layer deposition process includes:

depositing an initial material layer on each subregion, where the initial material layer has a thickness equal to the height of a bottom step in the transparent cavity layer to be formed.

A (1+i) th area selective atomic layer deposition process includes:

performing an area selective atomic layer deposition on 2 (i-1) subregions every the other 2 (i-1) subregions, to form a material layer structure corresponding to the i th area selective atomic layer deposition process.

The material layer structure grown through the i th area selective atomic layer deposition process has a thickness equal to a height of 2 (i-1) common steps. A height of a common step is the height of a single step in the transparent cavity layer to be formed except a bottom step; and

i∈{1, m} and i is an integer.

Through the first and second specific implementations of step S 903 , N step structures can be formed on the bottom reflecting layer through less area selective atomic layer deposition processes. It should be noted that in the embodiment of the present disclosure, step S 903 is not limited to the two specific implementations above. For example, in the present disclosure, one step structure may be formed through each area selective atomic layer deposition process. In this way, more area selective atomic layer deposition processes are used and more corresponding masks are fabricated, thereby resulting increasing in cost, tolerances while decreasing in efficiency.

It should be noted that the first specific implementation and the second specific implementation of step S 903 are examples of specific implementations for fabricating a transparent cavity layer having one-dimensional step height structures in one axis.

Based on the transparent cavity layer having the one-dimensional step structures in one axis, one-dimensional step structures in another axis can be fabricated with a similar method. In this way, a transparent cavity layer having two-dimensional step heights structures can be formed.

A snapshot mosaic transparent cavity layer can be formed by splicing multiple transparent cavity layers with the two-dimensional step structures together, where each transparent cavity layer having the two-dimensional step structures may be regarded as an area of the snapshot mosaic transparent cavity layer.

It should be noted that in fabricating the snapshot mosaic transparent cavity layer, one-dimensional step structures in one axis (for example, an X axis) area are fabricated first, and then step structures in the other axis (for example, a Y axis) are fabricated afterwards. In fabricating the snapshot mosaic transparent cavity layer, a method therefore is the same as a method for forming step structures in one axis area, only by designing the masks with divided areas.

For example, FIG. 15 shows a flowchart of a method for forming a snapshot mosaic transparent cavity layer with four 4×4 snapshot mosaic areas. The method includes from steps S 151 to S 155 in the following.

In step S 151 , a first material layer is grown on a whole surface of the photosensitive region surface of a CMOS image sensor wafer by using ASALD process.

As shown in FIG. 16A , the photosensitive region is divided into multiple structures by a 4×4 matrix. The matrix includes rows from 1 to 4 and columns from A to D. Step S 151 includes growing a material layer on each row from 1 to 4 and each column from A to D. FIG. 16A shows a schematic structural stereogram corresponding to step S 151 . It should be noted that the first material layer has a thickness equal to the height of the bottom step in a transparent cavity layer.

In step S 152 , a second material layer is deposited on the surface of a region where the third row and the fourth row are located by using the ASALD process.

FIG. 16B shows a schematic structural stereogram corresponding to step S 152 . The second material layer has a thickness equal to the height of two common steps.

In step S 153 , a third material layer is deposited on a surface of a region where the second row and the fourth row are located by using the ASALD process.

FIG. 16C shows a schematic structural stereogram corresponding to step S 153 . The third material layer has a thickness equal to the height of one common step.

Single axis one-dimensional structures in a row direction are formed by performing steps S 151 to S 153 . A transparent cavity layer with the single axis one-dimensional structures can be used to form a hyperspectral image sensor of a line scan (line scan) structure.

In step S 154 , a fourth material layer is deposited on columns C and D by using the ASALD process.

FIG. 16D shows a schematic structural stereogram corresponding to step S 154 . The fourth material layer has a thickness equal to the height of two common steps.

In step S 155 , a fifth material layer is deposited on columns B and D by using the ASALD process.

FIG. 16E shows a schematic structural stereogram corresponding to step S 155 . The fifth material layer has a thickness equal to the height of one common step.

Single axis one-dimensional step structures in a column direction are formed by performing steps S 154 to S 155 .

In this way, a transparent cavity layer with two-dimensional step structures can be formed by performing steps S 151 to S 155 . The formed transparent cavity layer with the two-dimensional step structures may function as one area of the snapshot mosaic transparent cavity layer with four 4×4 snapshot mosaic areas.

In step S 156 , four single area structures are combined together to form the snapshot mosaic transparent cavity layer with four 4×4 snapshot mosaic areas.

FIG. 16F and FIG. 16G are schematic structural stereograms of the snapshot mosaic transparent cavity layer with four 4×4 snapshot mosaic areas.

In step S 904 , a top reflecting layer is formed on the transparent cavity layer.

FIG. 10D shows a cross section of the structure corresponding to step S 904 , and a top reflecting layer 105 is formed on the transparent cavity layer 104 .

It should be noted that steps S 902 to S 904 compose the process of fabricating a Fabry-Pérot interferometer. According to the steps, the Fabry-Pérot interferometer includes the bottom reflecting layer 103 , the transparent cavity layer 104 formed on the bottom reflecting layer 103 and the top reflecting layer 105 formed on the transparent cavity layer 104 .

A hyperspectral image sensor can be monolithically integrated through the above steps.

It should be noted that steps S 902 to S 904 compose the process of fabricating the Fabry-Pérot interferometer. With the above fabrication method, a Fabry-Pérot interferometer can be monolithically integrated with a CMOS image sensor, thereby forming a hyperspectral image sensor. The hyperspectral image sensor formed in such a way has features of small in volume, fast in analysis speed and low in cost.

In addition, in the embodiment of the present disclosure, the transparent cavity layer is formed by using the area selective atomic layer deposition process. As the atomic layer deposition process has a feature of excellent deposition uniformity, hence in the present disclosure, the N step structures formed on the bottom reflecting layer by using the area selective atomic layer deposition process have the same height. Thereby, the issue of height non-uniformity of step structures is greatly minimized with the method of monolithic integration of a hyperspectral image sensor according to the present disclosure. Heights of the formed step structures are determined by conditions of the area selective atomic layer deposition process, for example, the number of deposition cycles, and film deposition non-uniformity (<1.5%) of the ALD process is much smaller than that of the photolithography-etching process (2.7%) according to actual experience of growing the films (for example, ALD-Al 2 O 3 films). In other words, film uniformity of the ALD process is better than that of the photolithography-etching process. Hence, with the method for monolithic integration of a hyperspectral image sensor according to the present disclosure, the issue of non-uniformity accumulation due to the etching process in conventional technology is greatly minimized, thereby being beneficial to reduce failure rate so to increase the yield of the devices.

In addition, the ALD process can be used to grow many types of film materials, and does not require that the materials can be etched easily. Hence, compared with the photolithography-etching process, the method for monolithic integration of a hyperspectral image sensor according to the present disclosure can be applied more widely in the aspect of material selection of a cavity layer.

The embodiments above are some preferred embodiments of the invention, and the invention is disclosed through the preferred embodiments which, however, are not intended to limit the invention. Those skilled in the art may make numerous possible alternations and modifications to the technical solutions according to the present disclosure or change the technical solutions into equivalent embodiments based on the disclosed methods and technical content, without deviating from the scope of technical solutions according to the present disclosure. Hence, any simple alternations, equivalents and modifications made to the embodiments based on the technical essence of the present disclosure without deviating from the content of the technical solutions according to the present disclosure shall fall within the scope of the technical solutions according to the present disclosure.