The application relates to scalable nanoimprint manufacturing of functional multi-layer metasurface devices.
Optical metasurfaces, consisting of subwavelength-scale meta-atom arrays, hold great promise of overcoming fundamental limitations of conventional optics. Due to their structural complexity, metasurfaces usually require high-resolution yet slow and expensive fabrication processes such as electron beam lithography. Their scalable nanomanufacturing remains a challenge that limits technology transfer and commercialization. Nanoimprint lithography (NIL) has been utilized in single-layer, standalone metasurface fabrication, but demonstrations of double- and multi-layered metasurfaces with precise layer-to-layer overlay remains elusive.
Metasurfaces are capable of manipulating fundamental electromagnetic responses, i.e., phase, amplitude, spectral response, and polarization, at subwavelength scale. They have shown great potential in addressing fundamental limitations of conventional bulky optical systems and realizing ultracompact optical devices and systems for many applications, such as holography, imaging, spectroscopy, beam shaping and steering, etc. Despite significant progress in metasurface design and proof-of-concept laboratory demonstrations, scalable and cost-effective nanomanufacturing with good uniformity and reproducibility remains one of the major challenges that slows down commercialization of metasurface devices. Conventional prototyping nanofabrication methods, such as electron-beam lithography (EBL) or focused-ion beam (FIB), reply on pixel-by-pixel writing for precise nanopatterning but are not suitable for scalable manufacturing due to long writing time, high cost and reproducibility problems over large scale. High-resolution semiconductor optical lithography technologies (such as deep-UV or extreme-UV lithography) are ideal for high-throughput production, but very expensive and complex to operate for prototyping demonstrations. In comparison, nanoimprint lithography (NIL) is suitable for both prototyping demonstration and large-scale production of nanostructures as small as to sub-ten nanometer scale given its unique optical diffraction-free, parallel patterning capabilities. Previously, NIL has been employed successfully in a wide range of optical applications, such as polarizers, anti-reflection coatings, solar absorbers, image or color displays, etc., and demonstrated successful for geometrically simple, stand-alone, single-layer metasurface structures as well as stacked double-layer structures without stringent demand of precise alignment. Yet the organization of multi-layer dielectric, metallic or hybrid metasurface structures into functional devices and systems requires not only high-throughput nanopatterning, but also precise alignment and high quality control in vertical stacking, which therefore requires completely different integration strategy.
According to examples of the present disclosure, an on-chip metasurface polarimetric imaging device is disclosed where the photonic structures and the NIL processes are designed and created using two separate NIL molds over a patterning area of >20 mm2 with designed Moiré alignment markers by electron-beam writing, and further subsequently fabricate silicon and aluminum metasurface structures over two levels. The silicon and aluminum metasurfaces are fabricated both by using NIL as a nanolithography and 3D pattern-transfer technology, respectively, achieving nanometer-scale linewidth uniformity, sub-200 nm overlay accuracy, and high optical polarization selection while minimizing fabrication complexity and surface roughness. This NIL-based, multi-layered nanomanufacturing approach is applicable to scalable production of large-area structures for ultra-compact optic, electronic, and fluidic devices.
According to examples of the present disclosure, a polarimetric metastructure is disclosed that comprises a formation of metasurface structure, a formation of vertically coupled double-layer metallic gratings (VCDG), and a dielectric spacer layer between the metasurface and the double-layer gratings. The geometry of the polarimetric metastructure is anisotropic subwavelength gratings with material of an optical index of larger than 1 in the wavelength of interest. The gratings can have a period in the range of 50 nm to 400 nm. The material can be a semiconductor. For example, the material can be silicon (Si), a dielectric, or a metal. The VCDG can comprise a layer of subwavelength gratings made of transparent dielectric material, a first thin metallic layer on top of the gratings, and a second thin metallic layer within the trenches next to the gratings. The subwavelength gratings can have a period (p) in the range of 50 nm to 400 nm. The subwavelength gratings can have a height (h) in the range of 50 nm to 300 nm. In the first and second metallic layers, the metal thickness (t) is the same for both layers. In the metal thickness, the thickness (t) is in the range of 10 nm to 200 nm. The first and second layer metallic layers can be disconnected and separated by a small gap (g). The gap (g) can be in the range of 5 nm to 80 nm. The dielectric can be silicon dioxide (SiO2). For example, in the grating material, the dielectric can be silicon nitride or a polymer. The polymer can be crosslinked with light illumination or heat radiation. The polymer can be silicon rich and can have an optical refractive index in the range of 1.4 to 1.7. For example, in the metallic material, the metal can be aluminum (Al), silver (Ag), or gold (Au). For example, in the dielectric spacer, the material can be silicon oxide, silicon nitride, or a polymer. The polymer can be crosslinked with light illumination or heat radiation. The polymer can be silicon rich and can have an optical refractive index in the range of 1.4 to 1.7. For example, in the dielectric spacer, its thickness (d) can be in the range of 100 nm to 1000 nm. For example, the polarimetric metastructure can be designed for red-green-blue (RGB) polarimetric imaging having an operation wavelength range in the visible wavelength. The operation wavelength range of the metastructure can be from about 400 nm to about 700 nm.
In accordance with examples of the present disclosure, a polarimetric metastructure imaging sensor is disclosed that is formed by integrating the polarimetric metastructures onto a visible imaging sensor. The imaging sensor provides full-Stokes polarization state analysis of light and polarimetric imaging of objectives. The integration of the metastructures to the visible imaging sensor can be achieved by bonding via a layer of glue. The glue can be a polymer that is crosslinked when exposed to light or heat radiation. The polarimetric imaging sensor provides single-shot full-Stokes imaging, broad-band operation, and full angle of view up to 40°.
In accordance with examples of the present disclosure, a synergistical design and manufacturing method of the disclosed polarimetric metastructures is disclosed that comprises a first nanoimprint lithography (NIL) step and relevant fabrication steps to produce metasurface structures, theoretical and experimental determination of the structural parameters of the VCDGs, including the grating period (p), grating height (h), metal thickness (t) and gap (g) for the targeted polarimetric performance, theoretical and experimental determination of the dielectric spacer thickness (d), performing a second NIL to create a scaffold of VCDG gratings on top of fabricated metasurface structures, depositing metals on the VCDG resist polymer with designed thickness (t). In the manufacturing method, a resist layer is deposited onto a top layer of a substrate, a mold with grating morphology is applied on at least a portion of the resist layer, a pressure is applied on the mold, heat or light radiation is applied while the mold is held onto the resist layer. In the manufacturing method, thin film deposition, metal etching mask deposition, and etching are involved to produce the nanograting structures of designed dimensions. The mold can be made of a material that is nanostructured into gratings of the designed grating period (p) and grating height (h). The mold material can be made in transparent material and UV light can be used as the radiation source during processing for the resist crosslinking. The mold material can be fused silica. The resist layer thickness can be designed such that the final resist after NIL has a thickness of d and the resist layer functions as the disclosed dielectric spacer. The resist polymer after NIL can have the desired grating period (p) and grating height (h). The resist surface can be planarized during NIL processing so that the grating scaffold in the NIL resist has a smooth morphology in both lateral and vertical directions. The metal can be deposited on top of the grating and the trenches next to the gratings to produce two layers of metal gratings with a gap (g) in between. The metal can be deposited in a vacuum system, wherein evaporation can be used for vacuum deposition. The metal deposition rate can be controlled from 0.1 nm per minute to 100 nm per minute. Marks can be designed in both NIL steps for precise alignment with an alignment error in the range of 5 nm to 500 nm. In the alignment process, two sets of gratings of slightly different periods (P1 and P2) can be designed in the metasurface and VCDG structures, and during the second NIL process the substrate and mold can be brought into a close vicinity between them (less than 10 μm), and dark and bright periodic images are formed under white light illumination due to interference of two sets of gratings, which called Moiré fringe. The period of Moiré fringe (Pfringe) is calculated as Pfringe=P1·P2/(P1−P2) and the alignment error (A) is calculated by the Moiré fringe alignment offset(s) as Δ=s·(P2/Pfringe), thus resulting in nanometer-scaled alignment accuracy given Pfringe/P2 is much bigger than 1. The area of the metasurface and the VCDGs are from millimeter scale to less than 1 meter in both of the lateral dimensions.
According to examples of the present disclosure, a scalable NIL integration approach towards manufacturing of metasurfaces over multiple layers is disclosed, creating subwavelength-thick, functional, metasurface polarization filter arrays (MPFAs) as a proof-of-concept demonstration. Here NIL was carried out in two steps to produce two sets of nanostructures in dielectric (silicon) and metallic (aluminum) materials with precise dimensional control (<4 nm in linewidth standard deviation) separated by a dielectric (SiO2) spacer layer. Uniquely, NIL was used as a lithographical technique to faithfully pattern nanostructures in silicon metasurface, and then explored as a transfer-printing method to replicate the three-dimensional geometries of the NIL mold into the imprint resist, which directly functioned as a scaffold for the aluminum metasurface. For precise alignment of the silicon and aluminum metasurfaces, Moiré alignment markers were placed onto both of the NIL molds along with the designed functional nanostructures, which in this work were organized into micrometer-sized arrays and contained dense nanogratings, to achieve a high alignment accuracy (inter-layer translational alignment error ˜200 nm, rotational error <0.017 degrees) between the silicon and aluminum metasurfaces. These vertically integrated micro-arrays, i.e. MPFAs, functioned as polarimetric imaging pixels, comprising broadband linear polarization (LP) filters, i.e. grating polarizers in the aluminum metasurface, and selectively patterned circular polarization (CP) filters across visible wavelength bands (500±20 nm, 550±20 nm, and 600±20 nm) from overlaid multi-layer nanostructures. Finally, the metasurface chips were bonded onto complementary metal-oxide semiconductor (CMOS) imaging sensors, thus enabling broadband polarimetric imaging and full-Stokes parameter analysis across the different colors at a high accuracy (<5%). Using MPFAs as an example, we showed that the NIL-fabricated devices presented a much smoother surface (4 nm roughness) and gratings compared to EBL-MPFAs (˜16 nm), attributed to the in-situ planarization capability of NIL, and the NIL-MPFAs produced much enhanced extinction ratio (ER, ˜10 times and ˜4 times better at blue and red color wavelengths) than EBL ones. This significant improvement in optical performance was attributed to our co-designed and greatly simplified manufacturing strategy that minimized fabrication complexity, reduced structure damages, and suppressed optical scattering loss. This successful multilayer NIL-metasurface integration approach can be adapted to the fabrication of many other metasurface structures towards high-throughput scalable manufacturing of various metasurface devices, which have broad applications in both efficient prototyping and large-scale production of ultra-compact chip-integrated optic, optoelectronic, electronic, and fluidic devices and systems.
A synergistic approach to co-design the multi-layered optical metasurfaces and their scalable NIL manufacturing process is disclosed. Our exemplary polarimetric imaging system was a multilayered metasurface polarization filter arrays (MPFAs) integrated onto a CMOS imaging sensor (
Previously, an EBL-based process (
In this work, because the metasurface design requires complex patterning of dense and small structures in silicon and aluminum (period as small as 180 nm and linewidth as small as 70 nm, see Table 1) with different geometries, orientations and dimensions to act as QWPs and polarizers, respectively in the vertically stacked MPFAs, EBL writing was chosen to create the NIL molds for its design flexibility and high resolution. A low EBL writing current (and therefore a long writing time) was needed for accurate control of structural geometry but in practice limited the demonstrated patterning to ˜20 mm2 even after a 16-hour writing. This area size was designed to fit a CMOS imager sensor for on-chip polarimetric imaging demonstration. Noticeably, NIL is many times faster than EBL once a mold is available (Table 5), and its high throughput advantage can be further manifested when scaling to even bigger areas if the molds can be made from more advanced lithography tool, such as deep-UV or extreme-UV lithography (Table 6). Nevertheless, the use of the MPFAs as an example allows us to establish this new manufacturing approach and quantitatively evaluate the device performance by comparing NIL with EBL based integration processes.
To optimize the manufacturing throughput and device performance, several design parameters were identified that are closely relevant to NIL processes and performance of Si QWP and the VCDGs (
The silicon metasurface-based, microscale QWP array was fabricated by thermal-NIL for its simplicity, using a NIL mold fabricated on a thermal SiO2-coated silicon wafer (
The VCDG linear polarizer arrays were fabricated using UV-NIL with a NIL mold fabricated on a transparent fused silica wafer (
NIL-based multi-layer metasurface fabrication usually faces two major challenges: 1) precise alignment between the different layers, and 2) the adverse impacts of existing surface topography on the subsequent NIL fabrication.
To achieve submicron optical alignment for vertical stacking of the VCDG and Si metasurface micro-filter arrays, we designed and fabricated interference-based Moiré patterns (
On the other hand, the surface topography resulting from the selectively fabricated Si metasurface (
Optical spectroscopic measurements were also performed of the integrated multi-layer metasurfaces and compared the performance of EBL- and NIL-fabricated MPFA samples (
The integrated multi-layer MPFAs were diced (7.2 mm×5.6 mm), optically aligned to the edges of a commercial CMOS sensor (IMX477) on mask aligner, and bonded with UV-curable polymer (
The chip-integrated full stokes polarimetric imaging sensors have a broad range of applications. As proof-of-concept demonstrations, here we show the imaging results of several objects, including a plastic fork, a pair of 3D glasses, and a beetle (
In summary, we propose and demonstrate a scalable manufacturing strategy of multilayer metasurfaces by synergistically co-designing multi-level NIL processes and metasurface structures. Uniquely, for the first time to our best knowledge, the nanometer-scale lithographic resolution and high-fidelity 3D pattern transfer capabilities of NIL were both explored to design and fabricate multi-layer dielectric and metallic hybrid nanostructures in a single device. In particular, NIL was used as a lithography tool to create a first-layer Si metasurface, and then utilized as a transfer printing technique that preserves both the lateral and vertical geometries of the mold in the NIL resist as a 3D scaffold, which was converted into second-layer VCDG metasurface through a simple Al evaporation without defect-prone etching or liftoff steps. We demonstrated successful fabrication of dense nanostructures (period ˜200 nm and critical dimension <100 nm) in micro-sized arrays (4.65×4.65 μm2 for VCDG and 7.75×7.75 μm2 for Si metasurface) over ˜0.2 cm2 area with uniform and accurate dimension control (linewidth standard deviation less than 4 nm) and high interlayer alignment accuracy (translational error ˜200 nm, rotation error <0.017 degrees). Compared to EBL, the NIL processes were not only much simpler and faster, but also greatly reduced the surface roughness (from ˜16 nm to ˜1 nm) and improved the CP extinction ratios by ˜10 and ˜4 times at blue and red color wavelength ranges. The NIL-fabricated MPFAs were bonded to a CMOS imager to create a metasurface polarimetric imaging sensor, i.e. Meta-PolarIm, for compact, single-shot, broadband polarimetric imaging in visible wavelengths, demonstrating a high polarization state measurement accuracy (<5%). These demonstrations proved this new, NIL-based, and simple fabrication method can outperform EBL in manufacturing throughput, cost and device performance. This multilayer NIL-metasurface co-design approach can be adapted to the fabrication of many other metasurface structures, enabling on-chip integration of metasurface devices and their optic, electronic, optoelectronic, or other systems. By speeding up the prototyping process and enabling low-cost, large-scale production of such devices and systems, our design and manufacturing strategy can support future commercialization and broad deployment of metasurface devices in profound applications that are key to next-generation commercial electronics, national security and sustainability.
Poly(benzyl methacrylate) (≥99.0%), Propylene glycol monomethyl ether acetate (≥99.5%), Pentaerythritol tetra acrylate, Isobutyl methacrylate (≥97.0%), Anisole (≥99.7%), and trichloro (1H,1H,2H,2H-perfluorooctyl) silane, Octadecyl acrylate (≥97.0%), and 1H,1H,2H,2H-perfluoro-1-decanol (97%) were purchased from Sigma-Aldrich. BYK-310 and BYK-3570 were purchased from BYK Additives and Instruments. Omnirad 1173 and Omnirad TPO were purchased from IGM Resins. (Acryloxypropyl) methyl siloxane homopolymer was purchased from Gelest. PMMA (950K A2 and 495K A3) was purchased from MicroChem. AMOPRIME was purchased from AMO GmbH. CN-292, SR-9003-B, and CN-975 were purchased from Satomer. AZ-1505 positive photoresist was purchased from MicroChemicals. Gel-box AD-22AS-00 was purchased from Gel-Pak. All chemicals used as received without further purification.
The thermal NIL resist was prepared by diluting thermoplastic polymer (poly-benzyl methacrylate, or PBMA) in Propylene glycol monomethyl ether acetate (PMA) as solvent, with a small amount of surface additive (BYK-310) added for lowering surface tension. The UV-NIL resist was prepared by mixing (Acryloxypropyl) methyl siloxane homopolymer with cross-linker (Pentaerythritol tetraacylated) photo initiators (Omnirad 1173 and Omnirad TPO) and surface additive (BYK-3570) in Isobutyl methacrylate (IBMA). For the CMOS bonding process, another UV-curable polymer was prepared by mixing fast-reacting, low-viscosity, acrylate oligomers (e.g. SR-9003-B and CN-292), a surface additive (1H,1H,2H,2H-Perfluoro-1-decanol BYK-3570), and photo initiators (Omnirad 1173 and Omnirad TPO) into IBMA solvent. All the solutions were stirred overnight at room temperature and filtered before use.
The Si metasurface mold was fabricated by EBL (
To fabricate NIL mold for the VCDGs, a thick fused silica wafer (6 mm) was chosen as the substrate (Supplementary Figure S7) to minimize mold bending during NIL. Then fused silica dicing, sample cleaning, EBL writing, development, Cr evaporation, and lift off were carried following the same process mentioned above to produce the nanostructured Cr hard masks. The EBL exposure doses were adjusted for designed structural dimensions. The Cr mask was used to etch 150 nm deep into fused silica by RIE using the same recipe as aforementioned. Differently, a mesa structure (roughly 1.5 cm2, height=2 μm) was intentionally fabricated in an additional RIE process to better accumulate pressure in the nanopatterned region. The mesa structure provided more uniformly imprinted structures using our imprinter. Both Si and fused silica molds were solvent and RCA-1 cleaned, and they were treated using trichloro(1H,1H,2H,2H-perfluorooctyl) silane in the vacuum oven for 30 min at 100° C. to form the self-assembled monolayers (SAMs) on the surface, which acted as an anti-sticking layer during the NIL process.
First, 130 nm α-Si was deposited on the pre-cleaned fused silica sample using plasma-enhanced CVD (PECVD) (Oxford Plasmalab 100, SiH4=480 sccm, 1200 mTorr, 15 W, 350° C.), followed by 60 nm SiO2 deposition using the same tool (SiH4=170 sccm, N2O=710 sccm, 1000 mTorr, 20 W, 350° C.) without breaking chamber vacuum. After the substrate preparation, a tri-layer structure was employed for the thermal NIL process. Namely, a PMMA layer (950 k A2, thickness of 90 nm) was spin-coated and post-baked 5 min at 200° C., followed by evaporation of ˜15 nm SiO2 mid-layer (Kurt J. Lesker) at a deposition rate of 0.5 Å s−1, and then spin-coating of thermal NIL resist and post-baking (5 min at 180° C.). The thermal NIL was carried out using a nanoimprinter (THU400, Zhenjiang Lehua Electronic Technology Co. Ltd.) at a nominal temperature reading of 55° C. and pressure of 750 KPa for 15 min in vacuum. Then the residual layer was RIE etched by oxygen plasma (O2=10 sccm, 10 mTorr, 100 W), where SiO2 mid-layer acted as the etch-stop layer to enable sufficient over etching time for uniform removal of the residual layer. The nanopatterns in the resist were transferred to SiO2 mid-layer by another RIE etching (CHF3=25 sccm, O2=1 sccm, 10 mTorr, 100 W), and the PMMA bottom layer was RIE etched by oxygen plasma (O2=10 sccm, 10 mTorr, 30 W). The high etching selectivity between SiO2 and PMMA is beneficial for reliable patterning in a relatively thick PMMA layer, and helps form a mushroom-like structure in the SiO2/PMMA stack to minimize accumulation of metal on the sidewall of PMMA, which facilitated high-yield lift-off process and minimized feature distortion. The fabricated sample was immersed in remover PG solution for 15 minutes at 80° C. for lift-off, and later rinsed with IPA and DI water, followed by 10 nm Cr layer deposition by thermal evaporation. The 60 nm SiO2 hard mask layer was etched by RIE (CHF3=40 sccm, O2=3 sccm, 40 mTorr, 250 W) using Cr as a hard mask, and Cr was stripped by chromium etchant. Finally, the 130 nm α-Si film was etched using inductively coupled plasma (ICP) RIE (PlasmaTherm Apex ICP, Cl2=100 sccm, Ar=5 sccm, 10 mTorr, 250 W) using SiO2 as a hard mask to complete Si metasurface fabrication. The SiO2 hard mask layer was left without intentional removal, but its thickness was taken into account of the whole spacer layer thickness calculation.
The VCDGs were fabricated in a significantly simpler fabrication process. First, adhesion promoter (AMOPRIME) was spin-coated on pre-cleaned fused silica chips (some with Si metasurface for device integration and some others without Si metasurface used as process monitors) and post-baked 10 min at 115° C. on a hot plate. The prepared UV-NIL resist was spin-coated on the substrate, followed by UV-NIL using the fabricated fused silica VCDG mold on mask aligner (MJB4, Suss MicroTec). Three different fringes were visualized on the mask aligner TV monitor for the case of alignment. Once alignment was verified, 1.5 s UV exposure was used to cross-link the resist, which turned to be a polymer similar to SiOx in optical index after curing (UV-NIR spectroscopic ellipsometry, J. A. Woollam, M-2000) (Supplementary Figure S8). The UV resist had a low viscosity to easily fill, desired for high-fidelity pattern transfer at relatively low pressure. After UV-NIL, the printed resist scaffold was treated using a mild oxygen plasma process (O2=10 sccm, 10 mTorr, 30 W) to activate the hydroxyl groups on the surface. A layer of Cr (2 nm) was evaporated followed by Al deposition at 2.5 Å s−1 to form the VCDG gratings. A high vacuum level (1 to 3×10−7 Torr) was useful to obtaining smoother surface morphology of VCDG (supplementary Figure S9) by decreasing residual gases in the chamber and reducing contaminants. Finally, a 200 nm SiO2 layer was deposited as an encapsulation layer to avoid further oxidation of the Al surface by using a radio-frequency (RF) sputtering system (Kurt J. Lesker) at a deposition rate of 0.5 Å s−1.
Here two sets of gratings with slightly different periods (e.g. P1 of 4 μm on the substrate and from Si metasurface mold, and P2 of 4.2 μm on the VCDG mold) acted as the Moiré marks. The two gratings would produce periodic stripes under illumination, with the period Pfringe calculated as Pfringe=P1·P2/(P2−P1)=84 μm, when the substrate and mold were brought close to each other, e.g. with a small gap less than 10 μm. To minimize the alignment error, four groups of alignment markers (AM1, AM2, AM3, and AM4, respectively) were placed next to the NIL-patterned area, in another word separated by 7.2 mm horizontally and 5.6 mm vertically from each other. Noticeably, our process differ from previously studies that required metal deposition, because the large optical index difference from α-Si metasurface (n=3.58 at 632 nm) to the substrate SiO2 (n=1.49 at 632 nm) provided distinguishable contrast and eliminated the needs of metallic coating.
The integrated multilayer metasurface chip was diced and bonded onto the customized CMOS sensor as follows. Here AZ-1505 photoresist (PR) was spin-coated on both sides of the fabricated sample and post-baked 1 min at 90° C. as a protection layer during chip dicing, then the sample was diced into 7.2 mm×5.6 mm rectangular shape using a dicing saw (DAD320, DISCO Corporation). Afterwards, the sample was immersed in acetone to remove PR, rinsed in IPA, and dried with nitrogen blow. A thin PDMS film of ˜1 mm was detached from a commercially available Gel-box and attached to a 4-inch borosilicate wafer as an intermediate host layer for the diced chip. A customized CMOS sensor was brought together with a printed circuit board (PCB), and mounted onto a customized support by Kapton tape, formed by stacked glass slides taped on a 4-inch Si wafer, to maintain the surface evenness considering that the backside of PCB had protruding electrical components. Then the UV-curable polymer was spin-coated on the CMOS imager, and the CMOS PCB was loaded into the mask aligner (MJB4, Suss MicroTec). After precise alignment the CMOS PCB was moved up in the z-direction and made contact with the metasurface chip, initiating polymer flow. Then the polymer was cross-linked under UV exposure (365 nm, 350 W) for 10 min to ensure appropriate bonding strength.
The linewidth dimension and surface morphology of the α-Si metasurface and the Al VCDGs were inspected by scanning electron microscopy (SEM, Hitachi S-4700 FESEM) with an acceleration voltage of 15 keV and current of 10 μA. A thin layer of Au/Pd was sputtered (Cressington sputter coater 108) on the α-Si metasurface sample to enhance imaging resolution prior to SEM measurements. Optical properties (refractive index n, extinction coefficient k) of deposited α-Si and SiO2 and cured UV resist films were measured by UV-NIR spectroscopic ellipsometry (J. A. Woollam, M-2000). Olympus BX53 fluorescent microscope coupled Horiba iHR320 imaging spectrometer was utilized to record all the optical images of fabricated samples for the calculation of alignment accuracy. It is noted that the electron microscopy would not be able to effectively detect the α-Si metasurface buried deep under the thick spacer layer (˜500 nm) after the UV NIL effectively planarized the surface topography. To standardize alignment measurement, the optical images were converted to 8-bit black and white images and processed by setting a color threshold. The transmittance spectra were measured by the same tool, then LPER and CPER were calculated.
The finite-difference time-domain (FDTD) simulations were carried out to calculate transmission efficiency and LPER and CPER of the metal-dielectric hybrid chiral metasurface. All the simulations were conducted with empirically measured optical indexes of each material. Periodic boundary conditions and perfectly matched layers were used within a unit cell along the in-plain direction. The plane wave was applied along the grating width and length direction to calculate LPER and efficiency and super-positioned two orthogonally linearly polarized plane waves were used to represent RCP/LCP light input. In initial simulations, ideal VCDG gratings without edge rounding were considered (
Stokes parameters of 16 reference polarization states input were theoretically calculated based on the linear retardance, transmission efficiency, bandwidth of the color filters, the angle of linear polarizer, and super achromatic quarter-wave plate. Firstly, transmission efficiency and linear retardance dispersion data of SAQWP05M-700 (Thorlabs) was obtained from Thorlabs website. Stokes parameter of light transmitted through the linear polarizer with angle of θ1 and quarter waveplate with fast axis along angle θ2 can be modelled using the Mueller matrix of a linear diatenuator and a linear retarder:
Here, θ1 represents the transmission axis of the linear polarizer, q and r represents the maximum and minimum transmission efficiency of linear polarizer, as extracted from data provided by Thorlabs website. LPER can be expressed as LPER=q/r.
Here, θ2 represents the angle fast axis of the retarder, δ represent retardance, as extracted from data provided by Thorlabs website.
It is noteworthy that both δ and fast axis angle θ2 are wavelength dependent. Final Stokes parameters were averaged after we obtained Sin at each wavelength accounting for wavelength dependency of δ and θ2 using the equation:
Device instrument matrix calibration process at red, green, blue colors
For an arbitrary input polarization state Sλ with input wavelength of λ (nm), the captured intensity of a super-pixel as a vector I can be written as the equation below:
where matrix Aλ is wavelength dependent instrument matrix of the metasurface filter array. Sλ can be inversely calculated by solving the equation 5:
To obtain A, at red, green, and blue color ranges, we used three color filters (FBH450-40, FBH550-40, FBH600-40) with bandwidths of 40 nm to select targeted wavelength range respectively. For each color, 10 pre-known polarization states Sλ,4×10 was measured by the device respectively to form an intensity matrix Icam,6×10, the instrument matrix Aλ can then be obtained using the equation:
Here, the rank of Sλ,4×10T should be 4 to make sure Sλ,4×10T is invertible.
A moving window spatial scanning discussed (
i=1, 2, 3, j=1, 2 . . . 8, n=671, p=509, where Si
The design tolerance can be evaluated by assuming small structural dimensional variations from fabrication, e.g., 10 nm linewidth increase or decrease. For example, as shown in
The hours are estimated based on the tools available to the authors (EBL system at University of Arizona, NIL tools and thermal evaporator in Wang and Yao labs, and film deposition and etchers at the NanoFab facility at ASU). For simplicity, sample cleaning, metrology, sample transfer, tool conditioning, etc. were not taken into account. Because EBL writing time scales linearly with the metasurface fabrication area, it becomes too slow and expensive, and therefore impractical, for many applications requiring fabrication over areas larger than centimeter squares.
As shown in
The embodiments set forth below represent the necessary information to enable those skilled in the art to practice the embodiments and illustrate the best mode of practicing the embodiments. Upon reading the following description in light of the accompanying drawing figures, those skilled in the art will understand the concepts of the disclosure and will recognize applications of these concepts not particularly addressed herein. It should be understood that these concepts and applications fall within the scope of the disclosure and the accompanying claims.
It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure. As used herein, the term “and/or” includes any and all combinations of one or more of the associated listed items.
It will be understood that when an element such as a layer, region, or substrate is referred to as being “on” or extending “onto” another element, it can be directly on or extend directly onto the other element or intervening elements may also be present. In contrast, when an element is referred to as being “directly on” or extending “directly onto” another element, there are no intervening elements present. Likewise, it will be understood that when an element such as a layer, region, or substrate is referred to as being “over” or extending “over” another element, it can be directly over or extend directly over the other element or intervening elements may also be present. In contrast, when an element is referred to as being “directly over” or extending “directly over” another element, there are no intervening elements present. It will also be understood that when an element is referred to as being “connected” or “coupled” to another element, it can be directly connected or coupled to the other element or intervening elements may be present. In contrast, when an element is referred to as being “directly connected” or “directly coupled” to another element, there are no intervening elements present.
Relative terms such as “below” or “above” or “upper” or “lower” or “horizontal” or “vertical” may be used herein to describe a relationship of one element, layer, or region to another element, layer, or region as illustrated in the Figures. It will be understood that these terms and those discussed above are intended to encompass different orientations of the device in addition to the orientation depicted in the Figures.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. As used herein, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes,” and/or “including” when used herein specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs. It will be further understood that terms used herein should be interpreted as having a meaning that is consistent with their meaning in the context of this specification and the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
This application claims benefit to U.S. Provisional Patent Application No. 63/500,458 filed on May 5, 2023, the contents of which are hereby incorporated by reference in its entirety.
This invention was made with government support under grant no. DE-EE0008999 awarded by the U.S. Department of Energy, grant nos. 1947753, 1809997, 1847324, and 1838443 awarded by the National Science Foundation. The government has certain rights in the invention.
Number | Date | Country | |
---|---|---|---|
63500458 | May 2023 | US |