Multispectral chip and spectral camera comprising same

Abstract

The utility model relates to a multispectral chip and a spectral camera comprising the same. The multispectral chip comprises a silicon-based substrate which is provided with an image sensing layer, and the image sensing layer comprises a plurality of pixels; the spectrum modulation material layer is arranged on the image sensing layer, the spectrum modulation material layer comprises a plurality of light filtering units arranged in an array, and each light filtering unit comprises a plurality of light filtering sub-units with different transmittance curves; at least one inclined plane lap joint area inclined relative to the silicon-based substrate exists between each filter sub-unit and the adjacent filter sub-unit, and the orthographic projection of the inclined plane lap joint area on the silicon-based substrate bridges two adjacent pixels.

Description

Technical Field

[0001] This application relates to the field of spectral modulation technology, specifically to a multispectral chip and a spectral camera containing the same. Background Technology

[0002] Spectral imaging has seen rapid development in recent years, enriching traditional imaging methods and providing unprecedented detail of objects. As a high-dimensional perception method, it plays an increasingly important role in fields such as precision agriculture, food safety inspection, environmental monitoring, and medical imaging.

[0003] Imaging multispectral chips are crucial components of spectral imaging and detection systems, enabling both two-dimensional imaging of targets and the detection of their spectral information. Current development of imaging multispectral chips primarily employs different technological approaches, such as metamaterial spectral modulation and thin-film material spectral modulation. The metamaterial spectral modulation approach utilizes periodic porous nanostructure arrays of varying sizes fabricated on a dielectric film to achieve narrowband filtering of the target image across the multispectral bands. However, this approach suffers from practical limitations in applications, including a small full width at half maximum (FWHM) of the transmission spectrum, low energy efficiency, reduced spatial resolution, complex fabrication processes, and high costs, making it unsuitable for everyday applications. In contrast, the thin-film material spectral modulation approach involves forming a periodic structure of a single-layer filter film on a photoelectric conversion substrate. Each periodic unit includes multiple modulation channels made of different spectral modulation materials, thereby achieving spectral dispersion.

[0004] However, with Figure 1 As shown in the example of a 9-channel array made from 9 different spectral modulation materials, all exhibit the following characteristics: Figure 2 The transmittance curves shown exhibit aliasing in at least some channels within certain wavelength ranges. This aliasing becomes more severe with increasing channel number, leading to low spectral resolution and making accurate identification and modulation of spectral signals difficult, thus limiting the application of multispectral chips in complex environments. Utility Model Content

[0005] In view of the problems existing in the prior art, the purpose of this application is to provide a multispectral chip and a spectral camera containing the same.

[0006] This utility model provides a multispectral chip, comprising:

[0007] A silicon-based substrate having an image sensing layer, wherein the image sensing layer includes a plurality of pixels;

[0008] A spectral modulation material layer is disposed on the image sensing layer, wherein the spectral modulation material layer includes a plurality of filter units arranged in an array, wherein each filter unit includes a plurality of filter sub-units with different transmittance curves, and there is at least one inclined overlapping area relative to the silicon substrate between each filter sub-unit and its adjacent filter sub-units, wherein the orthographic projection of the inclined overlapping area on the silicon substrate spans two adjacent pixels.

[0009] In one embodiment, the filter subunit is frustum-shaped.

[0010] In one embodiment, two adjacent filter subunits are a regular frustum and an inverted frustum, respectively.

[0011] In one embodiment, the angle between the beveled overlap area and the silicon substrate is between 60 and 90 degrees.

[0012] In one embodiment, the thickness of the filter subunit ranges from 500 to 1000 nm.

[0013] In one embodiment, the photofilter subunit is a colloidal cured film composed of a mixture of resin material, photoinitiator material, pigment and solvent material.

[0014] In one embodiment, each filter unit includes 3*3 filter subunits C1-C9, wherein the first row along the first direction (from left to right in the figure) consists of C1-C3, the second row along the first direction consists of C4-C6, and the third row along the first direction consists of C7-C9, wherein C2, C4, C6 and C8 are regular square frustums, and C1, C3, C5, C7 and C9 are inverted square frustums.

[0015] In one embodiment, the transmittance curve of light incident on the multispectral chip, modulated by the spectral modulation material layer, satisfies the following constraint:

[0016] TC1: Valley transmittance wavelength is 400-550nm; T(430nm-495nm)≤10%; T(550nm-900nm)≥70%;

[0017] TC2: Valley transmittance wavelength is 400-575nm; T(420nm-550nm)≤15%; T(585nm-900nm)≥65%;

[0018] TC3: Valley transmittance wavelength is 400-580nm; T(455nm-550nm)≤15%; T(600nm-900nm)≥75%;

[0019] TC4: Valley transmittance wavelength is 500-680nm; T(400nm-500nm)≥65%; T(580nm-645nm)≤10%; T(700nm-900nm)≥65%;

[0020] TC5: Valley transmittance wavelength is 500-675nm; T(400nm-500nm)≥50%; T(550nm-650nm)≤10%; T(700nm-900nm)≥60%;

[0021] TC6: Valley transmittance wavelength is 550-700nm; T(400nm-535nm)≥65%; T(600nm-665nm)≤10%; T(700nm-900nm)≥75%;

[0022] TC7: Valley transmittance wavelength is 550-750nm; T(400nm-550nm)≥65%; T(630nm-725nm)≤10%; T(800nm-900nm)≥75%;

[0023] TC8: Valley transmittance wavelength is 550-750nm; T(460nm-545nm)≥55%; T(625nm-740nm)≤10%; T(775nm-900nm)≥55%;

[0024] TC9: Valley transmittance wavelength is 600-700nm; T(465nm-540nm)≥70%; T(680nm-745nm)≤10%; T(785nm-900nm)≥70%.

[0025] In one embodiment, the multispectral chip further includes:

[0026] A base material layer, disposed between the image sensing layer and the spectral modulation material layer, has a first transmittance curve, wherein the shape of the first transmittance curve is designed such that the transmittance curve of light incident on the multispectral chip, modulated by the spectral modulation material layer and the base material layer, is more sensitive in a preset wavelength range than in a range outside the preset wavelength range.

[0027] In one embodiment, the transmittance of the first transmittance curve within the preset wavelength range is higher than the transmittance outside the preset wavelength range.

[0028] In one embodiment,

[0029] If the waveform of the filter subunit is plateau-shaped within the preset wavelength range, the modulated transmittance curve will show a peak within the preset wavelength range; or

[0030] If the waveform of the filter subunit has a peak within the preset wavelength range, and the modulated transmittance curve still shows a peak within the preset wavelength range, the full width at half maximum (FWHM) of the modulated peak is smaller than that of the peak before modulation; or

[0031] If the waveform of the filter subunit is concave within the preset wavelength range, the waveform of the modulated transmittance curve within the preset wavelength range will show a peak.

[0032] In one embodiment, the base material layer is a colloidal cured film composed of a mixture of resin material, photoinitiator material, and solvent material.

[0033] In one embodiment, the base material layer is a monolithic structure.

[0034] In one embodiment, the sum of the thicknesses of each filter subunit and its corresponding base material layer region in the entire multispectral chip is equal; and within the same filter subunit, the thicknesses of the base material layer regions corresponding to different filter subunits are stepped.

[0035] In one embodiment, the base material layer comprises: a mesh structure formed of a dielectric material, wherein the orthographic projection of each mesh of the mesh structure onto the image sensing layer is aligned with the orthographic projection of one or more filter subunits onto the image sensing layer; and a base material filling the mesh, wherein the base material is the same for each mesh.

[0036] In one embodiment, the base material in each mesh of the mesh structure corresponds to a filter subunit, wherein,

[0037] Throughout the entire multispectral chip, the sum of the thickness of the base material in all meshes and the thickness of their corresponding filter sub-units is equal; and

[0038] Within the same filter unit, the thickness of the base material corresponding to different filter sub-units is stepped.

[0039] In one embodiment, the thickness of the base material layer ranges from 50 to 500 nm.

[0040] In one embodiment, each filter unit includes 3*3 filter subunits C1-C9, wherein the first row along the first direction consists of C1-C3, the second row along the first direction consists of C4-C6, and the third row along the first direction consists of C7-C9. C2, C4, C6, and C8 are regular frustums, and C1, C3, C5, C7, and C9 are inverted frustums.

[0041] The preset wavelength range is 500-650nm;

[0042] The transmittance curve of light incident on the multispectral chip, modulated by the spectral modulation material layer and the base material layer, satisfies the following constraint:

[0043] TA1: Peak transmittance wavelength is 520-640nm; T(420nm-495nm)≤10%; T(580nm-625nm)≥55%; T(670nm-900nm)≥50%;

[0044] TA2: Peak transmittance wavelength is 570-645nm; T(425nm-550nm)≤10%; T(600nm-630nm)≥70%; T(670nm-900nm)≥55%;

[0045] TA3: Peak transmittance wavelength is 575-645nm; T(460nm-555nm)≤15%; T(580nm-620nm)≥70%; T(660nm-900nm)≥55%;

[0046] TA4: Valley transmittance wavelength is 550-650nm; T(400nm-500nm)≥45%; T(580nm-645nm)≤10%; T(700nm-900nm)≥50%;

[0047] TA5: Valley transmittance wavelength is 550-675nm; T(425nm-500nm)≥35%; T(550nm-665nm)≤10%; T(700nm-900nm)≥45%;

[0048] TA6: Valley transmittance wavelength is 500-650nm; T(350nm-450nm)≤40%; T(485nm-615nm)≥50%; T(800nm-900nm)≥60%;

[0049] TA7: Valley transmittance wavelength is 575-750nm; T(400nm-570nm)≥45%; T(630nm-730nm)≤10%; T(775nm-900nm)≥50%;

[0050] TA8: Valley transmittance wavelength is 600-750nm; T(460nm-565nm)≥40%; T(620nm-745nm)≤10%; T(800nm-900nm)≥45%;

[0051] TA9: Valley transmittance wavelength is 600-760nm; T(460nm-580nm)≥50%; T(635nm-645nm)≤15%; T(800nm-900nm)≥50%.

[0052] Another aspect of this application provides a method for fabricating a multispectral chip, comprising:

[0053] An image sensing layer is formed on a silicon substrate, wherein the image sensing layer includes a plurality of pixels;

[0054] A spectral modulation material layer is formed on the image sensing layer, wherein the spectral modulation material layer includes a plurality of filter units arranged in an array, wherein each filter unit includes a plurality of filter sub-units with different transmittance curves, and there is at least one inclined overlapping area between each filter sub-unit and its adjacent filter sub-units, the orthographic projection of the inclined overlapping area on the silicon substrate spans two adjacent pixels.

[0055] In one embodiment, a spectral modulation material layer is formed on the image sensing layer, comprising:

[0056] The first filter sub-unit material of each filter unit is formed on the image sensing layer;

[0057] The material of the first filter subunit is patterned, wherein the patterned first filter subunit is in the shape of a regular frustum;

[0058] The j-th filter unit material in each filter unit is formed on the patterned i-th filter unit material and the exposed area;

[0059] The material of the j-th filter sub-unit is patterned, wherein the patterned j-th filter sub-unit is a regular frustum shape, wherein i and j traverse from 1 to n and the i-th filter sub-unit and the j-th filter sub-unit are not adjacent in physical space, and n is the total number of filter sub-units included in each filter unit.

[0060] The areas exposed outside the regions forming all the frustum-shaped filter sub-units are filled with the remaining filter sub-unit material to form each filter sub-unit. Each filter sub-unit includes multiple filter sub-units with different transmittance curves, and there is at least one inclined overlap area between each filter sub-unit and its adjacent filter sub-units. The orthographic projection of the inclined overlap area on the silicon substrate spans two adjacent pixels.

[0061] In one embodiment, each filter unit includes 3*3 filter subunits C1-C9, wherein the first row along the first direction consists of C1-C3, the second row along the first direction consists of C4-C6, and the third row along the first direction consists of C7-C9. C2, C4, C6, and C8 are regular frustums, and C1, C3, C5, C7, and C9 are inverted frustums.

[0062] In the method, C2, C4, C6 and C8 are prepared first, and then C1, C3, C5, C7 and C9 are filled.

[0063] In the scheme of this application, the sloping overlap region helps to separate the mixed transmittance curves, thereby improving the distinction between different spectra and enhancing the effect of spectral modulation. Attached Figure Description

[0064] The disclosure of this application will become more readily understood with reference to the accompanying drawings. It will be readily understood by those skilled in the art that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this application. Furthermore, similar numbers in the drawings are used to denote similar components, wherein:

[0065] Figure 1 This is a schematic diagram of the structure of a multispectral chip based on existing technology;

[0066] Figure 2 It is based on the light transmittance curve of a 9-channel multispectral chip in existing technology, in which there is an aliasing region;

[0067] Figure 3 This is a schematic diagram of the structure of a multispectral chip according to an embodiment of this application;

[0068] Figure 4 It is based on Figure 3 A partial schematic diagram of the structure of the multispectral chip shown;

[0069] Figure 5 This is a light transmittance curve of a 9-channel multispectral chip according to an embodiment of this application, wherein some channels are demixed;

[0070] Figure 6 This is a schematic diagram of a structure having a base material layer according to an embodiment of this application;

[0071] Figure 7 This is a schematic diagram of the first transmittance curve of the base material layer according to an embodiment of this application;

[0072] Figure 8 This is a schematic diagram of the transmittance curve after modulation of a filter subunit having a base material layer and having the shapes of a regular square truncated pyramid and an inverted square truncated pyramid, according to an embodiment of this application.

[0073] Figure 9 This is a schematic flowchart of a method for fabricating a multispectral chip according to an embodiment of this application. Detailed Implementation

[0074] Some embodiments of this application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of this application and are not intended to limit the scope of protection of this application.

[0075] In the description of this application, for ease of description, spatial relative terms such as "below," "under," "below," "above," and "on" may be used to describe the relationship between one element and another. When an element or layer is referred to as "on," "adjacent to," or "connected to" other elements or layers, it may be directly on, adjacent to, or connected to other elements or layers, or there may be intervening elements or layers. Conversely, when an element is referred to as "directly on," "directly adjacent to," or "directly connected to" other elements or layers, there are no intervening elements or layers.

[0076] It should also be understood that, for ease of description, the term "A and / or B" refers to all possible combinations of A and B, such as only A, only B, or A and B. The terms "at least one A or B" or "at least one of A and B" have a similar meaning to "A and / or B" and may include only A, only B, or A and B. The singular forms of the terms "a" or "this" may also include plural forms.

[0077] The specific embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0078] like Figure 1 As shown, the existing multispectral chip 1 includes:

[0079] The silicon-based image sensor 10 is specifically a CMOS image sensor or a CCD image sensor;

[0080] The spectral modulation layer 12 includes multiple spectral modulation units 120 arranged in an array (shown by dashed boxes in the figure), wherein each spectral modulation unit 120 includes multiple sub-modulation units, which are nine in the example in the figure, and are respectively denoted as C1, C2, C3, C4, C5, C6, C7, C8 and C9.

[0081] In practical applications, the light source emits a spectrum towards the object to be imaged, so that the spectrum passes through the object and is then incident on the multispectral chip 1. The intensity of the spectrum modulated by the incident light through the nine sub-modulation units arranged in the array is sensed and detected by the pixels 100 on the silicon-based image sensor 10, thereby determining each pixel point. Finally, all pixels are integrated to form an image.

[0082] Among them, the nine modulation channels are made of different spectral modulation materials, and the nine spectral modulation materials have the following properties: Figure 2 The transmittance curves shown exhibit aliasing in at least some channels within certain wavelength ranges. Experiments revealed that this aliasing worsens with increasing channel count. This results in low spectral resolution, making accurate identification and modulation of spectral signals difficult and limiting the application of multispectral chips in complex environments.

[0083] Typically, when fabricating each filter subunit, the desired ideal morphology is vertical or nearly vertical, as shown in Figure 1. Researchers have explored various methods for fabricating vertical or near-vertical filter subunits. However, the inventors of this application have discovered through experimentation that when the edges of the filter subunits are sufficiently inclined—that is, when adjacent filter subunits have an overlap area (inclined interface)—… Figure 2 The waveform in the aliasing region shown will demix, which gives the inventors a new idea to further optimize the transmittance curve.

[0084] Therefore, in one embodiment of this application, a multispectral chip is provided, comprising:

[0085] A silicon-based substrate having an image sensing layer, wherein the image sensing layer includes a plurality of pixels;

[0086] A spectral modulation material layer is disposed on the image sensing layer, wherein the spectral modulation material layer includes a plurality of filter units arranged in an array, wherein each filter unit includes a plurality of filter sub-units with different transmittance curves, and there is at least one inclined overlapping area relative to the silicon substrate between each filter sub-unit and its adjacent filter sub-units, wherein the orthographic projection of the inclined overlapping area on the silicon substrate spans two adjacent pixels.

[0087] For example, such as Figure 3 As shown, where Figure 3 (a) is a top view of the multispectral chip, and 3(b) is a cross-sectional view along the dashed line in the figure.

[0088] In one embodiment, the image sensing layer may be a CMOS image sensor or a CCD image sensor formed on a silicon substrate, wherein the image sensing layer includes a plurality of pixels.

[0089] For the filter unit, each filter unit includes 3*3 filter sub-units C1-C9, wherein the first row along the first direction consists of C1-C3, the second row along the first direction consists of C4-C6, and the third row along the first direction consists of C7-C9.

[0090] In one specific example, as shown in the figure, the filter sub-unit corresponds one-to-one with the pixel 200 below it, that is, their orthogonal projections on the silicon substrate are aligned with each other. However, this application is not limited to this. In another specific example, one filter sub-unit corresponds to multiple pixels 200 below it. In this case, the multiple pixels 200 corresponding to one filter sub-unit together constitute a "superpixel". In this application, the terms "superpixel" and "pixel" are not distinguished.

[0091] Furthermore, in this application, the filter subunits C1-C9 are not limited to being fabricated using only spectral modulation materials. Some of the filter subunits can be fabricated using RGB filter materials, thus enabling the filter unit to include two types of filter arrays. The first type of filter array is used to acquire high-resolution low-spectral images; for example, the first type of filter array can be an RGB filter material array. The second type of filter array is used to acquire low-resolution high-spectral images; for example, the second type of filter array can be a filter array composed of spectral modulation materials. This type of filter array structure allows for the acquisition of high-resolution multispectral images in a single imaging process. It not only has a simple structure but also reduces the cost of improving the resolution of high-spectral images.

[0092] In a preferred embodiment, C2, C4, C6 and C8 are regular frustums, and C1, C3, C5, C7 and C9 are inverted frustums.

[0093] Please note that "regular" and "inverted" are relative concepts. In this application, a regular truncated square refers to a face on the upper and lower surfaces that is closer to the image sensing layer with a larger area than the face away from the image sensing layer. That is, relative to the image sensing layer, the filter subunit is smaller at the top and larger at the bottom. An inverted truncated square refers to a face on the upper and lower surfaces that is closer to the image sensing layer with a smaller area than the face away from the image sensing layer. That is, relative to the image sensing layer, the filter subunit is larger at the top and smaller at the bottom.

[0094] from Figure 3 As can be seen in (b), C2 is a regular trapezoid (the shape of the cross-section, which corresponds to a regular square frustum in the three-dimensional concept), and the adjacent C1 and C3 are inverted trapezoids (the shape of the cross-section, which corresponds to an inverted square frustum in the three-dimensional concept). The adjacent filter sub-units form a sloping overlap area that is inclined relative to the silicon substrate.

[0095] Such a structure can achieve such a modulation effect, such as Figure 4As shown, taking C1-C3 as an example, the light entering pixel P1 is modulated by C1 and C2 (the R1 area to the left of the dashed line L1); the light entering pixel P3 is modulated by C3 and C2 (the R2 area to the right of the dashed line L2); and the light entering pixel P2 is modulated by C1 (the R3 area to the right of the dashed line L3), C2, and C3 (the R4 area to the left of the dashed line L4).

[0096] Figure 5 As shown, the transmittance curve of light incident on the multispectral chip after being modulated by the spectral modulation material layer satisfies the following constraint:

[0097] TC1: Valley transmittance wavelength is 400-550nm; T(430nm-495nm)≤10%; T(550nm-900nm)≥70%;

[0098] TC2: Valley transmittance wavelength is 400-575nm; T(420nm-550nm)≤15%; T(585nm-900nm)≥65%;

[0099] TC3: Valley transmittance wavelength is 400-580nm; T(455nm-550nm)≤15%; T(600nm-900nm)≥75%;

[0100] TC4: Valley transmittance wavelength is 500-680nm; T(400nm-500nm)≥65%; T(580nm-645nm)≤10%; T(700nm-900nm)≥65%;

[0101] TC5: Valley transmittance wavelength is 500-675nm; T(400nm-500nm)≥50%; T(550nm-650nm)≤10%; T(700nm-900nm)≥60%;

[0102] TC6: Valley transmittance wavelength is 550-700nm; T(400nm-535nm)≥65%; T(600nm-665nm)≤10%; T(700nm-900nm)≥75%;

[0103] TC7: Valley transmittance wavelength is 550-750nm; T(400nm-550nm)≥65%; T(630nm-725nm)≤10%; T(800nm-900nm)≥75%;

[0104] TC8: Valley transmittance wavelength is 550-750nm; T(460nm-545nm)≥55%; T(625nm-740nm)≤10%; T(775nm-900nm)≥55%;

[0105] TC9: Valley transmittance wavelength is 600-700nm; T(465nm-540nm)≥70%; T(680nm-745nm)≤10%; T(785nm-900nm)≥70%.

[0106] Compare Figure 5 and Figure 2 It is evident that the combination of regular and inverted square pyramid shapes helps to separate partially overlapping transmittance curves, thereby improving the distinction between different spectra and enhancing the effect of spectral modulation.

[0107] The above examples use regular and inverted square truncated pyramids as illustrations; however, this application is not limited to these. Any truncated pyramid structure is acceptable. That is, a common modulation effect can be achieved when a filter subunit and its adjacent filter subunits have at least one beveled overlap area. For example, C1 and C2, C4 each have beveled overlap areas. Those skilled in the art, guided by the teachings of this application, can adjust the area and slope of the overlap area to ensure that the light entering each pixel is modulated by at least two different modulation subunits, thereby finely adjusting the linearity of the light transmittance curve entering the pixel and alleviating aliasing at certain wavelengths.

[0108] In a preferred example, the angle between the beveled overlap area and the surface of the spectral modulation material layer is between 60 and 90 degrees.

[0109] In one embodiment, the thickness of the filter subunit ranges from 500 to 1000 nm. Experiments have shown that the thickness of the modulation layer is 500-1000 nm. For the modulation layer, if its thickness is less than 500 nm, it loses its broadband modulation capability and cannot achieve the ideal modulation spectral line shape; while if the modulation layer thickness exceeds 1000 nm, it makes the fabrication process difficult and, while increasing spectral crosstalk, also reduces the transmittance of the modulation spectral line.

[0110] In one embodiment, the photofilter subunit is a colloidal cured film composed of a mixture of resin material, photoinitiator material, pigment and solvent material.

[0111] The resin materials include soluble resins, such as phenolic resins, polyurethane resins, polyvinyl alcohol resins, and maleic anhydride resins; and photocurable resins, such as polyimide resins, polyvinyl alcohol resins, epoxy resins, and styrene resins.

[0112] Photoinitiator materials include benzophenones, alkyl phenyl ketones, benzoin and its derivatives, iodonium salts, and iron aromatics.

[0113] Pigments include aniline pigments, phthalocyanine pigments, azo pigments, and pyrrole pigments.

[0114] Solvent-based materials include ethylene glycol methyl ethers, propylene glycol methyl ether acetates, triethylene glycol methyl ethers, and ethyl 3-ethoxypropionate.

[0115] By selecting and proportioning these materials, the transmittance curves of each of the aforementioned filter sub-units can be obtained.

[0116] Furthermore, from Figure 5 As can be seen, the modulated light transmittance curve does not achieve a better response line in certain key wavelength ranges (also known as key bands, such as 500-650nm when used in cameras and other imaging tools for visible light imaging). As shown in the figure, some modulation channels exhibit plateau regions or even notch lines in this band. When performing color reproduction, such a structure is not sensitive to the response of this key band and will be affected by interference from bands outside this band, especially under low light conditions.

[0117] The conventional approach is to try to further adjust the material selection and ratio of each spectral modulation subunit. However, this method does not easily achieve satisfactory results because it requires consideration of the material selection of each modulation subunit. Even if it is achieved, when the application scenario changes (which means the key band is different), it is necessary to try to select and match the materials of each modulation subunit for the new key band, which is too complicated.

[0118] Therefore, in a preferred embodiment, a multispectral chip, such as Figure 6 As shown, it also includes:

[0119] A base material layer is disposed between the image sensing layer and the spectral modulation material layer, wherein the linear shape of the first transmittance curve is designed such that the transmittance curve of light incident on the multispectral chip, after being modulated by the spectral modulation material layer and the base material layer, has a more sensitive response within a preset wavelength range than its response outside the preset wavelength range.

[0120] Unlike existing technologies, this application adds a base material layer, and the incident light is modulated sequentially by the spectral modulation material layer and the base material layer.

[0121] The inventors introduced a base material layer into the original structure and designed the shape of its transmittance curve (hereinafter referred to as the first transmittance curve). This makes the transmittance curve of the incident light after being modulated by the spectral modulation material layer and the base material layer more sensitive in the preset wavelength range than in the range outside the preset wavelength range. Essentially, this makes the transmittance of the multispectral chip higher for light in the key band (which, from a design perspective, can also be called the preset wavelength range) than for light outside that range.

[0122] In a preferred embodiment, the transmittance of the first transmittance curve within a preset wavelength range is higher than the transmittance outside the preset wavelength range.

[0123] Figure 7 The figure shows the first transmittance curve designed for the visible light band of 500-650nm. Specifically, the transmittance of the first transmittance curve satisfies the following constraints: peak transmittance wavelength is 500-640nm, T(400nm-490nm)≥70%, T(585nm-620nm)≥90%, and T(650nm-900nm)≥70%.

[0124] The process of modulating the transmittance curves of each filter unit using the first transmittance curve is mathematically equivalent to multiplying the transmittances of the two units at the same wavelength as a new transmittance, thus obtaining a wavelength-transmittance function. Alternatively, it can be described as using the transmittance on the first transmittance curve as a modulation factor to modulate the transmittance on the transmittance curves of each filter unit. This design of the first transmittance curve allows for the following: Figure 5 The transmittance curve shown is modulated within a preset wavelength range, which can improve or alleviate situations where some modulation channels plateau or even dip within that wavelength range.

[0125] Specifically, if the waveform of the filter subunit within the preset wavelength range is plateau-shaped, the modulated transmittance curve within the preset wavelength range will exhibit a peak; if the waveform of the filter subunit within the preset wavelength range has a peak, the modulated transmittance curve within the preset wavelength range will still exhibit a peak, and the full width at half maximum (FWHM) of the modulated peak will be smaller than that of the peak before modulation; if the waveform of the filter subunit within the preset wavelength range is concave-shaped, the modulated transmittance curve within the preset wavelength range will exhibit a peak.

[0126] like Figure 8 As shown, the transmittance curves of light incident on the multispectral chip, modulated by the spectral modulation material layer and the base material layer, satisfy the following constraints:

[0127] TA1: Peak transmittance wavelength is 520-640nm; T(420nm-495nm)≤10%; T(580nm-625nm)≥55%; T(670nm-900nm)≥50%;

[0128] TA2: Peak transmittance wavelength is 570-645nm; T(425nm-550nm)≤10%; T(600nm-630nm)≥70%; T(670nm-900nm)≥55%;

[0129] TA3: Peak transmittance wavelength is 575-645nm; T(460nm-555nm)≤15%; T(580nm-620nm)≥70%; T(660nm-900nm)≥55%;

[0130] TA4: Valley transmittance wavelength is 550-650nm; T(400nm-500nm)≥45%; T(580nm-645nm)≤10%; T(700nm-900nm)≥50%;

[0131] TA5: Valley transmittance wavelength is 550-675nm; T(425nm-500nm)≥35%; T(550nm-665nm)≤10%; T(700nm-900nm)≥45%;

[0132] TA6: Valley transmittance wavelength is 500-650nm; T(350nm-450nm)≤40%; T(485nm-615nm)≥50%; T(800nm-900nm)≥60%;

[0133] TA7: Valley transmittance wavelength is 575-750nm; T(400nm-570nm)≥45%; T(630nm-730nm)≤10%; T(775nm-900nm)≥50%;

[0134] TA8: Valley transmittance wavelength is 600-750nm; T(460nm-565nm)≥40%; T(620nm-745nm)≤10%; T(800nm-900nm)≥45%;

[0135] TA9: Valley transmittance wavelength is 600-760nm; T(460nm-580nm)≥50%; T(635nm-645nm)≤15%; T(800nm-900nm)≥50%.

[0136] Compare Figure 8 and Figure 5It is evident that adding a base material layer to the frustum-shaped filter subunit not only helps to demix the aliased transmittance curves, improving the distinction between different spectra and enhancing the spectral modulation effect, but also, by introducing a base material layer with a transmittance curve shape designed for a preset wavelength range, further modulates the transmittance curve of the original light modulation subunit. This optimizes the spectral camera's response to the preset wavelength range, improves image quality, reduces interference from other bands (the presence of the base material layer suppresses light transmittance outside the preset wavelength range), and ensures accurate color reproduction during shooting. Even if the application scenario changes (the key wavelength changes), only the transmittance curve of a single base layer needs to be designed for the new target wavelength, rather than adjusting the transmittance curves of multiple filter subunits in a coordinated manner. Furthermore, by using a combination of regular and inverted frustum shapes, the transmittance of the multispectral chip within the preset wavelength range is also relatively improved.

[0137] The following explains how to achieve the transmittance curve designed for the base material layer. The inventor's idea is to make it compatible with the materials and processes of the filter subunit.

[0138] As mentioned above, in the embodiment where the material of the photofilter unit is a colloidal curable film composed of resin material, photoinitiator material, pigment, and solvent material, the base material layer is selected to be made of a colloidal curable film composed of resin material, photoinitiator material, and solvent material.

[0139] exist Figure 6 In the example shown, the base material layer is a monolithic structure, meaning it is deposited as a single layer on top of the image sensing layer, covering it. However, in this case, when light exits from a filter unit and enters the base material layer, it can scatter into areas outside the corresponding region of the base material layer, causing crosstalk. This is particularly noticeable when the base material layer is thick.

[0140] Therefore, this application provides another preferred structure. Specifically, the base material layer includes:

[0141] A mesh structure formed of a dielectric material, wherein the orthographic projection of each mesh of the mesh structure onto the image sensing layer is aligned with the orthographic projection of one or more filter subunits onto the image sensing layer;

[0142] The base material is filled into the mesh, wherein the base material is the same in each mesh.

[0143] The medium material is preferably a light-absorbing material, such as the black matrix commonly used in the display panel industry.

[0144] As those skilled in the art can anticipate, the addition of a base material layer reduces the energy utilization of light, a situation that is further highlighted in embodiments with a grid structure formed by light-absorbing materials.

[0145] Typically, the solution that comes to mind is to reduce the thickness of the base material layer. However, if the base layer is too thin, it will be impossible to effectively modulate the key bands.

[0146] Therefore, in one embodiment, experiments revealed that the thickness of the base layer ranges from 50 to 500 nm, and the thickness of the modulation layer ranges from 500 to 1000 nm. Within this range, the spectral modulation channels of the base layer and modulation layer are advantageous for fabrication and can achieve better spectral modulation performance. A base layer thickness exceeding 500 nm increases the optical path length of the modulation channel, leading to spectral crosstalk between adjacent pixels in the image sensing layer. For the modulation layer, if its thickness is less than 500 nm, it loses its broadband modulation capability and cannot achieve the ideal modulation spectral line shape; while if the modulation layer thickness exceeds 1000 nm, it makes the fabrication process difficult and, in addition to increasing spectral crosstalk, also reduces the transmittance of the modulation spectral lines.

[0147] However, this overall adjustment method does not take into account the differences in materials of different filter sub-units (e.g., different wavelengths of light scatter the same material to different degrees). A more preferred embodiment is given below.

[0148] In embodiments where the base material layer is a monolithic structure, within the entire multispectral chip, the thickness of the base material layer corresponding to different filter sub-units in the same filter unit exhibits a stepped structure. Taking C1-C9 as examples, the thickness of the base material region corresponding to C1 > the thickness of the base material region corresponding to C2 > the thickness of the base material region corresponding to C3 > the thickness of the base material region corresponding to C4 > the thickness of the base material region corresponding to C5 > the thickness of the base material region corresponding to C6 > the thickness of the base material region corresponding to C7 > the thickness of the base material region corresponding to C8 > the thickness of the base material region corresponding to C9. The overall base material layer exhibits a stepped structure.

[0149] As will be discussed in the process section later, the thickness varies depending on which filter sub-unit is fabricated first; the thickness of the corresponding base material region fabricated first is greater than that of the corresponding base material region fabricated later. Of course, the overall thickness of the base layer can still be controlled within the range of 50-500nm.

[0150] In the specific process, filter sub-units corresponding to wavelength ranges insensitive to scattering by the base layer material can be selected as C1 (made first), and filter sub-units corresponding to wavelength ranges sensitive to scattering by the base layer material can be selected as C9 (made later). This takes into account the crosstalk problem. At the same time, for the subsequent device flatness, the sum of the thickness of each filter sub-unit and the thickness of its corresponding base material layer region in the projection relationship is equal. Of course, the thickness of the modulation layer can still be controlled within 500-1000nm. In this way, with the total thickness being the same, the thickness ratio of each filter sub-unit to the corresponding base layer material can be varied, which is equivalent to the material ratio in the vertical direction being varied in each channel. This increases the means to adjust the transmittance curve shape, thereby enabling a more refined obtaining of the desired curve shape.

[0151] In the above-described mesh structure implementation, the above concept can also be realized. In the entire multispectral chip, the sum of the thickness of the base material in all meshes and the thickness of the corresponding filter sub-units are equal; and in the same filter unit, the thickness of the base material corresponding to different filter sub-units is stepped.

[0152] A second aspect of this application provides a spectroscopic camera, including the multispectral chip described in the first aspect. Of course, the spectroscopic camera may also include a microlens array formed on a spectral modulation layer, etc.

[0153] A third aspect of this application provides a method for fabricating a multispectral chip, such as... Figure 9 As shown, it includes:

[0154] S10. An image sensing layer is formed on a silicon substrate, wherein the image sensing layer includes a plurality of pixels;

[0155] S20. A spectral modulation material layer is formed on the image sensing layer, wherein the spectral modulation material layer includes a plurality of filter units arranged in an array, wherein each filter unit includes a plurality of filter sub-units with different transmittance curves, and there is at least one inclined overlapping area between each filter sub-unit and its adjacent filter sub-units, the orthographic projection of the inclined overlapping area on the silicon substrate spans two adjacent pixels.

[0156] In one embodiment, step 20 includes:

[0157] The first filter sub-unit material of each filter unit is formed on the image sensing layer;

[0158] The material of the first filter subunit is patterned, wherein the patterned first filter subunit is in the shape of a regular frustum;

[0159] The j-th filter unit material in each filter unit is formed on the patterned i-th filter unit material and the exposed area;

[0160] The material of the j-th filter sub-unit is patterned, wherein the patterned j-th filter sub-unit is a regular frustum shape, wherein i and j traverse from 1 to n and the i-th filter sub-unit and the j-th filter sub-unit are not adjacent in physical space, and n is the total number of filter sub-units included in each filter unit.

[0161] The areas exposed outside the regions forming all the frustum-shaped filter sub-units are filled with the remaining filter sub-unit material to form each filter sub-unit. Each filter sub-unit includes multiple filter sub-units with different transmittance curves, and there is at least one inclined overlap area between each filter sub-unit and its adjacent filter sub-units. The orthographic projection of the inclined overlap area on the silicon substrate spans two adjacent pixels.

[0162] by Figure 3 Taking a 9-channel structure as an example (using the combination of a regular square truncated pyramid and an inverted square truncated pyramid as an example), the specific steps include:

[0163] First, fabricate the frustum-shaped filter sub-units, for example, C2, C4, C6, and C8 in sequence, i.e., fabricate the filter sub-units in their order of arrangement. Then, fill in the corresponding positions with C1, C3, C5, C7, and C9.

[0164] Those skilled in the art will understand that although the above example shows the fabrication of a regular square truncated pyramid in the order of C2, C4, C6, and C8, this application is not limited to this. C6 or C8 can be fabricated after C2, as long as the filter subunits fabricated one after the other are not physically adjacent.

[0165] It should be noted that although the steps in the above embodiments are described in a specific order, those skilled in the art will understand that in order to achieve the effect of this application, different steps do not necessarily have to be executed in such an order. They can be executed simultaneously (in parallel) or in other orders, and these variations are all within the scope of protection of this application.

[0166] The technical solutions of this application have been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of this application is obviously not limited to these specific embodiments. Without departing from the principles of this application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of this application.

Claims

1. A multispectral chip, characterized in that, include: A silicon-based substrate having an image sensing layer, wherein the image sensing layer includes a plurality of pixels; A spectral modulation material layer is disposed on the image sensing layer, wherein the spectral modulation material layer includes a plurality of filter units arranged in an array, wherein each filter unit includes a plurality of filter sub-units with different transmittance curves, and there is at least one inclined overlapping area relative to the silicon substrate between each filter sub-unit and its adjacent filter sub-units, wherein the orthographic projection of the inclined overlapping area on the silicon substrate spans two adjacent pixels.

2. The multispectral chip according to claim 1, characterized in that, The filter subunit is frustum-shaped.

3. The multispectral chip according to claim 2, characterized in that, The two adjacent filter sub-units are a regular square frustum and an inverted square frustum, respectively.

4. The multispectral chip according to any one of claims 1-3, characterized in that, The angle between the inclined overlapping area and the silicon substrate is between 60 degrees and 90 degrees.

5. The multispectral chip according to claim 4, characterized in that, The thickness of the filter subunit ranges from 500 to 1000 nm.

6. The multispectral chip according to claim 5, characterized in that, The filter unit is a colloidal cured film composed of resin material, photoinitiator material, pigment and solvent material.

7. The multispectral chip according to claim 3, characterized in that, Each filter unit includes 3*3 filter subunits C1-C9, wherein the first row along the first direction consists of C1-C3, the second row along the first direction consists of C4-C6, and the third row along the first direction consists of C7-C9. C2, C4, C6, and C8 are regular square frustums, and C1, C3, C5, C7, and C9 are inverted square frustums.

8. The multispectral chip according to claim 7, characterized in that, The transmittance curve of light incident on the multispectral chip after being modulated by the spectral modulation material layer satisfies the following constraint: TC1: Valley transmittance wavelength is 400-550nm; T(430nm-495nm)≤10%; T(550nm-900nm)≥70%; TC2: Valley transmittance wavelength is 400-575nm; T(420nm-550nm)≤15%; T585nm-900nm)≥65%; TC3: Valley transmittance wavelength is 400-580nm; T(455nm-550nm)≤15%; T(600nm-900nm)≥75%; TC4: Valley transmittance wavelength is 500-680nm; T(400nm-500nm)≥65%; T(580nm-645nm)≤10%; T(700nm-900nm)≥65%; TC5: Valley transmittance wavelength is 500-675nm; T(400nm-500nm)≥50%; T(550nm-650nm)≤10%; T(700nm-900nm)≥60%; TC6: Valley transmittance wavelength is 550-700nm; T(400nm-535nm)≥65%; T(600nm-665nm)≤10%; T(700nm-900nm)≥75%; TC7: Valley transmittance wavelength is 550-750nm; T(400nm-550nm)≥65%; T(630nm-725nm)≤10%; T(800nm-900nm)≥75%; TC8: Valley transmittance wavelength is 550-750nm; T(460nm-545nm)≥55%; T(625nm-740nm)≤10%; T(775nm-900nm)≥55%; TC9: Valley transmittance wavelength is 600-700nm; T(465nm-540nm)≥70%; T(680nm-745nm)≤10%; T(785nm-900nm)≥70%.

9. A spectroscopic camera, characterized in that, The multispectral chip included in any one of claims 1-8.

10. The spectroscopic camera according to claim 9, characterized in that, Also includes: A microlens array formed on the spectral modulation layer.

Images