Metasurface-based terahertz micro-spectrometer
By using a metasurface-based terahertz micro-spectrometer, which employs a substrate, depletion layer, two-dimensional material absorption layer, and metal metasurface filter for spectral encoding and combines it with deep learning decoding, the problems of large size and high cost of spectrometers have been solved, and high-resolution spectral data collection and reconstruction have been achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV CITY COLLEGE
- Filing Date
- 2023-08-03
- Publication Date
- 2026-05-29
AI Technical Summary
Existing spectroscopic instruments are large and heavy, limiting their application scenarios and costing a lot, especially at terahertz frequencies where expensive light sources and sensors are required. Furthermore, existing spectrometers based on metasurfaces and deep learning are also large and complex to use.
A terahertz micro spectrometer based on metasurfaces is employed, comprising a substrate, a depletion layer, a two-dimensional material absorption layer, and a metal metasurface filter. Spectral encoding is performed using the metal metasurface filter, and deep learning decoding is combined to achieve the collection and reconstruction of spectral data.
It achieves miniaturization of the spectrometer, with a spectral resolution of 0.00078 terahertz, enabling efficient collection and reconstruction of spectral data of target objects and simplifying the usage process.
Smart Images
Figure CN116973333B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of spectral detection and relates to a terahertz micro spectrometer based on metasurfaces. Background Technology
[0002] Existing spectrometers measure the spectral information of objects by spectral dispersion, measuring data for one wavelength at a time, thus increasing the data collection time. Furthermore, spectrometers equipped with spectroscopes and detectors are large and heavy, limiting their application scenarios. Spectrometers operating at terahertz frequencies require not only a terahertz light source but also corresponding sensors, making them very expensive. Current spectrometers using metasurface encoding and deep learning decoding require collecting data from multiple detectors across multiple spectral channels, resulting in larger size and more complex operation. Summary of the Invention
[0003] The purpose of this invention is to address the shortcomings of existing technologies by providing a terahertz micro spectrometer based on metasurfaces.
[0004] The objective of this invention is achieved through the following technical solution: a terahertz micro spectrometer based on metasurface, with a working frequency of 1-5 terahertz, consisting of a substrate, a depletion layer, a two-dimensional material absorption layer, and a metal metasurface filter from bottom to top.
[0005] The two-dimensional material absorption layer has a band gap of 0, enabling it to absorb the spectrum in the 1-5 terahertz band, and its absorption rate can be tuned by the Fermi level.
[0006] The metal metasurface filter is used for spectral encoding to randomly adjust the spectral absorbance of the spectrometer.
[0007] The substrate serves as the gate of the spectrometer, and the source and drain electrodes are set at both ends of the metal metasurface filter to form a terahertz micro spectrometer based on the metasurface.
[0008] Spectral encoding is performed using a metal metasurface filter. Different voltages are applied to the source, drain, and gate to modulate the Fermi level of the two-dimensional material absorption layer. The encoded spectral data is collected through the two-dimensional material absorption layer. The spectral data of the target object is then decoded using deep learning.
[0009] Furthermore, the substrate is made of a conductive material with a transmittance of approximately 0 at 1-5 terahertz, such as semiconductor materials like germanium and gallium nitride, and metal materials like gold, silver, and aluminum.
[0010] Furthermore, the depletion layer serves as an insulating channel between the source, drain, and gate, blocking carrier migration. The material can be insulating materials such as silicon dioxide, hafnium oxide, aluminum oxide, boron nitride, and bismuth selenide. The thickness of silicon dioxide can be 100-300 nanometers, the thickness of hafnium oxide and aluminum oxide can be 10-30 nanometers, and the thickness of boron nitride and bismuth selenide can be 5-20 nanometers.
[0011] Furthermore, the two-dimensional material absorption layer can effectively absorb a wide spectrum with a frequency of 1-5 terahertz. The material is a two-dimensional material with a band gap of 0, such as graphene, and the thickness can be 0.34-15 nanometers.
[0012] Furthermore, the metal metasurface filter includes a plurality of metal loop units arranged in an array; by changing the distance between adjacent metal loop units, the width of each loop, and the distance between adjacent loops, the spectrum of the metal metasurface filter is encoded, thereby generating a random spectral absorption rate.
[0013] Furthermore, the thickness of the metal loop unit is 50-200 nanometers, and the distance between adjacent metal loop units is 10-60 micrometers; the width of each loop in the metal loop unit is 10-60 micrometers, and the distance between adjacent loops is 10-60 micrometers.
[0014] Furthermore, by applying different voltages to the source, drain, and gate, the Fermi level of the two-dimensional material absorption layer is modulated, including:
[0015] Apply voltage V to the source and gate GS And exceeds the threshold V th To start the spectrometer, i.e., V GS >V th When different voltages are applied to the source and drain electrodes, with voltage values ranging from 0 to 200V, the Fermi level of the absorption layer can be effectively controlled, with Fermi level values ranging from 0 eV to 0.9 eV. Combined with the spectral modulation of the metal metasurface filter, the spectral absorption rate of the spectrometer can be randomly controlled, generating absorption rate curves with various waveforms.
[0016] Furthermore, the deep learning decoding is an end-to-end decoding method implemented through a deep neural network model, which includes an input layer, several hidden layers, and an output layer. The spectral encoding is input into the deep neural network model, and after model training, the spectral data of the target object at frequencies of 1-5 terahertz is reconstructed, with a spectral resolution of up to 0.00078 terahertz.
[0017] Compared with the prior art, the present invention has the following beneficial technical effects:
[0018] 1. The terahertz micro spectrometer based on metasurface of the present invention requires only a metal metasurface filter and a two-dimensional material absorption layer to collect spectral data of the target object. After deep learning decoding, the spectral curve of the object at frequencies of 1-5 terahertz can be reconstructed.
[0019] 2. The terahertz micro spectrometer based on metasurface of the present invention has a spectral resolution of up to 0.00078 terahertz. Attached Figure Description
[0020] Figure 1 Fabrication diagram of a metasurface-based terahertz micro spectrometer provided in an embodiment of the present invention;
[0021] Figure 2 This is a schematic diagram of a metasurface filter provided in an embodiment of the present invention;
[0022] Figure 3 The Fermi level modulation diagram of the spectral curve of the metasurface filter provided in the embodiment of the present invention;
[0023] Figure 4 Absorption spectrum curve of a terahertz micro spectrometer based on metasurface provided in an embodiment of the present invention;
[0024] Figure 5 The image shows the spectral reconstruction of a target object by a metasurface-based terahertz micro-spectrometer, as provided in an embodiment of the present invention. Detailed Implementation
[0025] To better understand the technical solution of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.
[0026] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.
[0027] The terminology used in the embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. The singular forms “a,” “the,” and “the” used in the embodiments of this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.
[0028] This invention provides a terahertz micro spectrometer based on metasurfaces, with an operating frequency of 1-5 terahertz, consisting of a substrate, a depletion layer, a two-dimensional material absorption layer, and a metal metasurface filter from bottom to top.
[0029] The two-dimensional material absorption layer has a band gap of 0, enabling it to absorb the spectrum in the 1-5 terahertz band, and its absorption rate can be tuned by the Fermi level; the metal metasurface filter is used for spectral encoding, which allows for random control of the spectrometer's spectral absorption rate; the substrate serves as the spectrometer's gate, and the source and drain electrodes are set at both ends of the metal metasurface filter, forming a metasurface-based terahertz micro-spectrometer.
[0030] Spectral encoding is performed using a metal metasurface filter. Different voltages are applied to the source, drain, and gate to modulate the Fermi level of the two-dimensional material absorption layer. The encoded spectral data is collected through the two-dimensional material absorption layer. The spectral data of the target object is then decoded using deep learning.
[0031] Specifically, the substrate material is a conductive material with a transmittance of approximately 0 at 1-5 terahertz, such as semiconductor materials like germanium and gallium nitride, and metal materials like gold, silver, and aluminum.
[0032] The depletion layer serves as an insulating channel between the source, drain, and gate, blocking carrier migration. The material can be insulating materials such as silicon dioxide, hafnium oxide, aluminum oxide, boron nitride, and bismuth selenide. The thickness of silicon dioxide can be 100-300 nanometers, the thickness of hafnium oxide and aluminum oxide can be 10-30 nanometers, and the thickness of boron nitride and bismuth selenide can be 5-20 nanometers.
[0033] Two-dimensional material absorption layers can effectively absorb a wide spectrum with frequencies of 1-5 terahertz. The material is a two-dimensional material with a band gap of 0, such as graphene, and the thickness can be 0.34-15 nanometers.
[0034] The metal metasurface filter comprises several arrayed metal loop units, each with a thickness of 50-200 nanometers and a distance of 10-60 micrometers between adjacent units. The width of each loop within the metal loop unit is 10-60 micrometers, and the distance between adjacent loops is also 10-60 micrometers. By changing the distance between adjacent metal loop units, the width of each loop, and the distance between adjacent loops, the spectrum of the metal metasurface filter is encoded, thereby generating a random spectral absorbance.
[0035] Specifically, the Fermi level of the two-dimensional material absorption layer is modulated by applying different voltages to the source, drain, and gate, including: applying voltage V to the source and gate. GS And exceeds the threshold V th To start the spectrometer, i.e., V GS >V thWhen different voltages are applied to the source and drain electrodes, with voltage values ranging from 0 to 200V, the Fermi level of the absorption layer can be effectively controlled, with Fermi level values ranging from 0 eV to 0.9 eV. Combined with the spectral modulation of the metal metasurface filter, the spectral absorption rate of the spectrometer can be randomly controlled, generating absorption rate curves with various waveforms.
[0036] Specifically, deep learning decoding is an end-to-end decoding method implemented through a deep neural network model, which includes an input layer, several hidden layers, and an output layer. The spectral encoding is input into the deep neural network model, and after model training, the spectral data of the target object at frequencies of 1-5 terahertz is reconstructed.
[0037] In one embodiment, the following is an example of the fabrication process for a metasurface-based terahertz micro-spectrometer, including the following steps:
[0038] First, such as Figure 1 As shown in (a), conductive materials with a transmittance of approximately 0 at 1-5 terahertz, such as semiconductor materials like germanium and gallium nitride, and metal materials like gold, silver, and aluminum, are selected as substrates, which also serve as the gate of the spectrometer.
[0039] The second step, as Figure 1 As shown in (b), a 300-nanometer-long silicon dioxide layer is deposited on the substrate using a thin-film evaporation device. The material and thickness can also be 10-20 nanometers of hafnium oxide or aluminum oxide; or 5-10 nanometers of boron nitride or bismuth selenide can be mechanically stripped off.
[0040] Then, as Figure 1 As shown in (c), two-dimensional materials such as 0.34-15 nanometer graphene are transferred as an absorption layer above the depletion layer by mechanical exfoliation.
[0041] like Figure 1 As shown in (d), a metal metasurface filter is then fabricated above the absorption layer, coated with 500 nm positive resist PMMA, and a circular pattern is lithographically formed using an electron beam lithography machine. After development, a 5 nm chromium layer and a 200 nm gold layer are deposited using a thin film evaporation device. The resist is then removed using acetone, isopropanol, and water.
[0042] Finally, as Figure 1 As shown in (e), the source and drain electrodes are fabricated, a layer of positive photoresist PMMA is deposited, and the rectangular source and drain electrodes are photolithographically etched on both sides of the metal metasurface filter using an electron beam lithography machine and overlay technique. After development, a 5-nanometer chromium layer and a 200-nanometer gold layer are deposited using a thin film evaporation machine. Finally, the spectrometer is fabricated by removing the photoresist using acetone, isopropanol and water.
[0043] Figure 2This is a schematic diagram of a metasurface filter provided in an embodiment of the present invention. In this embodiment, the metal metasurface filter is a 5*5 metal loop unit array, and each metal loop unit includes four nested loops.
[0044] Figure 3 The Fermi level modulation diagram of the spectral curve of the metal metasurface filter provided in the embodiment of the present invention; Figure 4 Absorption spectrum curve of a terahertz micro spectrometer based on metasurface provided in an embodiment of the present invention; Figure 5 The image shows a spectral reconstruction of a target object using a metasurface-based terahertz micro-spectrometer provided in an embodiment of the present invention. As can be seen from the image, the spectrometer can reconstruct the spectral curve of the object at frequencies of 1-5 terahertz, with a spectral resolution of up to 0.00078 terahertz.
[0045] The above description is merely a preferred embodiment of one or more embodiments of this specification and is not intended to limit the scope of one or more embodiments of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this specification should be included within the scope of protection of one or more embodiments of this specification.
Claims
1. A terahertz micro-spectrometer based on a metasurface, operating at a frequency of 1-5 terahertz, characterized in that, From bottom to top, the layers are: substrate, depletion layer, two-dimensional material absorption layer, and metal metasurface filter; the substrate is a conductive material with a transmittance of approximately 0 at 1-5 terahertz. The two-dimensional material absorption layer has a band gap of 0, enabling it to absorb the spectrum in the 1-5 terahertz band, and its absorption rate can be tuned by the Fermi level. The metal metasurface filter is used for spectral encoding to randomly adjust the spectral absorbance of the spectrometer. The metal metasurface filter includes several arrayed metal loop units. By changing the distance between adjacent metal loop units, the width of each loop, and the distance between adjacent loops, the spectrum of the metal metasurface filter is encoded, thereby generating a random spectral absorbance. The substrate serves as the gate of the spectrometer, and the source and drain electrodes are set at both ends of the metal metasurface filter to form a terahertz micro spectrometer based on the metasurface. Spectral encoding is performed using a metal metasurface filter. Different voltages are applied to the source, drain, and gate to modulate the Fermi level of the two-dimensional material absorption layer. The encoded spectral data is collected through the two-dimensional material absorption layer. The spectral data of the target object is then decoded using deep learning.
2. The terahertz micro-spectrometer based on metasurfaces according to claim 1, characterized in that, The substrate is made of germanium, gallium nitride, gold, silver, or aluminum.
3. The terahertz micro-spectrometer based on metasurfaces according to claim 1, characterized in that, The depletion layer serves as an insulating channel between the source, drain, and gate, blocking carrier migration. The material is silicon dioxide, hafnium oxide, aluminum oxide, boron nitride, or bismuth selenide. When silicon dioxide is used, the thickness is 100-300 nanometers; when hafnium oxide or aluminum oxide is used, the thickness is 10-30 nanometers; and when boron nitride or bismuth selenide is used, the thickness is 5-20 nanometers.
4. The terahertz micro-spectrometer based on metasurface according to claim 1, characterized in that, The material of the two-dimensional material absorption layer is graphene, and the thickness is 0.34-15 nanometers.
5. The terahertz micro-spectrometer based on metasurfaces according to claim 1, characterized in that, The thickness of the metal loop unit is 50-200 nanometers, and the distance between adjacent metal loop units is 10-60 micrometers; the width of each loop in the metal loop unit is 10-60 micrometers, and the distance between adjacent loops is 10-60 micrometers.
6. The terahertz micro-spectrometer based on metasurfaces according to claim 1, characterized in that, By applying different voltages to the source, drain, and gate, the Fermi level of the two-dimensional material absorption layer can be controlled. This includes applying voltages to the source and gate that exceed a threshold to start the spectrometer, and applying different voltages to the source and drain electrodes, with voltage values ranging from 0 to 200V, effectively controlling the Fermi level of the absorption layer, with Fermi level values ranging from 0 eV to 0.9 eV. Combined with the spectral modulation of the metal metasurface filter, the spectral absorbance of the spectrometer can be randomly controlled, generating absorbance curves with various waveforms.
7. The terahertz micro-spectrometer based on metasurfaces according to claim 1, characterized in that, The deep learning decoding is an end-to-end decoding method implemented through a deep neural network model, which includes an input layer, several hidden layers, and an output layer. The spectral encoding is input into the deep neural network model, and after model training, the spectral data of the target object at frequencies of 1-5 terahertz is reconstructed.