Chongfan Technology
News
07
2025
-
04
Nature Photonics | On-chip tunable twisted angle moiré sensor
Author:
Introduction
By Harvard University 、 Stanford University 、 University of California, Berkeley jointly developed a new chip-level " tunable twisted-angle moiré photonic crystal sensor ", for the first time simultaneously decoding spectral and polarization information in a miniature optical chip, realizing a true " hyperspectral + hyperpolarization " multi-dimensional perception, opening up new possibilities for future photonic computing, intelligent imaging, quantum information processing, and other applications. This work was recently published in "Nature Photonics" An adaptive moiré sensor for spectro-polarimetric hyperimaging " in Nature Photonics 。
This work was led by Dr. Hao-Ning Tang (Department of Applied Physics, Harvard University) and Dr. Bei-Cheng Lou (Department of Applied Physics, Stanford University) in collaboration with many top scientists in optics and nanophotonics, including Professor Eric Mazur, Harvard University 、 Professor Shan-Hui Fan, Stanford University ， Professor Yuan Cao, University of California, Berkeley 、 Professor Amir Yacoby, Harvard University and Professor Evelyn Hu, Harvard University and from Fan Du, Harvard University ， Guangqi Gao ， Mingjie Zhang 、 Xueqi Ni students.
Research Background: Decoding all the information in "light"
In nature and the information world, light carries extremely rich information—not only color (wavelength), but also polarization states, and even hidden multi-dimensional information such as phase, direction, and topological structure.
This information is widely used in bio-imaging, material analysis, optical communication, lidar, quantum information and other fields.
However, traditional methods often require a complex set of optical systems to simultaneously acquire these optical degrees of freedom, including:
-
Spectrometer (analyzing wavelength)
-
Polarizers and wave plates (analyzing polarization)
-
Interferometer (analyzing phase)
This not only makes the equipment bulky and difficult to integrate, but also suffers from slow measurement speed, limited resolution, and high cost.
Research Challenge: Can a single chip simultaneously sense the "color + polarization" of light?
In recent years, with the development of computational optics and nanophotonics, people have begun to consider:
Can a highly integrated, miniaturized structure be used to achieve the joint perception and intelligent decoding of multi-dimensional optical information?
The key challenges are:
-
How to build a tunable platform that can dynamically respond to different optical information?
-
How to achieve simultaneous identification of spectrum and polarization instead of sequential scanning?
-
How to accomplish these functions at the chip scale to achieve true integrability and scalability?
Solution: MEMS Tunable Moiré Photonic Crystal Sensor (MEMS-TMPhC)
This study proposes and implements a new type of chip-level, dynamically adjustable sensor that can simultaneously sense both spectrum and polarization— MEMS integrated twisted-angle moiré photonic crystal sensor (MEMS-TMPhC) 。
-
Innovative structural design: Generating a moiré superlattice through two twisted photonic crystal layers to achieve structural control response to light.
-
MEMS dynamic drive: Using an electro-mechanical system to adjust the twist angle and interlayer spacing to achieve reconfigurable optical response.
-
Intelligent computational perception: Combined with an adaptive compressed sensing algorithm to simultaneously decode spectral and polarization states.
-
Compatible with CMOS process: Supports chip manufacturing and large-scale deployment.
This scheme not only demonstrates the perfect combination of computational optics and structural photonics, but also lays the foundation for future intelligent photonic sensors, integrated spectrometers, polarization imaging chips, quantum detectors, and other applications.
Design Ideas
The core objective of this research is to develop a miniaturized, dynamically adjustable optical sensor with multiple degrees of freedom that can be used to simultaneously acquire the frequency (spectrum) and polarization (polarization state) information of light and has reconfigurability and chip-level integration capabilities 。
Figure 1: Computational reconstruction of multi-dimensional information. (a) Pixels in an object contain multi-dimensional optical information, including wavelength (λ), polarization ellipticity (φ), and polarization azimuth (χ). This information can be reconstructed in two steps: First, using an integrated MEMS torsional moiré photonic crystal sensor, the light intensity of the pixel is measured by adjusting multiple degrees of freedom (DoF1, DoF2, DoF3); then, the multi-dimensional information of the target is inferred from the constructed measurement matrix. (b) The degrees of freedom of the double-layer torsional moiré photonic crystal include interlayer coupling distance (h), twist angle (α), and detection angle (θ)
Image Source: Nature Photonics
Key Concepts
Formed by twisting two identical photonic crystal (PhC) structures with the same period relative to each other, Moiré superlattice thereby causing a series of tunable optical effects such as bandgap modulation, singularity control, and spectral modulation.
Introducing MEMS actuators in the chip, electrically controlling the twist angle and interlayer spacing, thereby achieving real-time adjustment of the moiré structure giving the sensor multi-dimensional responsiveness.
Each degree of freedom (e.g., twist angle α, spacing h, detection angle θ) changes the moiré crystal's response to incident light, so we can use these degrees of freedom to "scan" different optical information dimensions (such as wavelength, polarization, phase, etc.), and use an algorithm to invert the original signal 。
Figure 2: Microelectromechanical tunable torsional moiré photonic crystal sensor. (a) Right: Schematic diagram of the main components of the MEMS-TMPhC sensor. PhC represents photonic crystal, TSV represents Through Silicon Vias, and BOX represents Buffered Oxide. Left: Cross-sectional diagram of the MEMS-TMPhC sensor (not drawn to scale), showing two photonic crystal layers integrated through a monolithic process. One photonic crystal layer can be vertically moved and rotated by MEMS actuators. (b) Microscopic image of the MEMS-TMPhC sensor. (c) Scanning electron microscope (SEM) image of the central region of the sensor, the inset showing one of the integrated photonic crystal structures. (d) Microscopic images of the moiré patterns observed at different rotation angles. (e) Left: Vertical MEMS actuator structure; Right: Measurement and fitting driving curve of the actuator. The shaded area indicates the operating range where the moiré pattern appears. (f) Left: Rotary MEMS actuator structure; Right: Measurement and fitting driving curve of the actuator. The shaded area indicates the achievable rotation angle range.
Image Source: Nature Photonics
Working Principle
When the interlayer spacing h of the two photonic crystals is less than the wavelength λ and has a certain twist angle α, a strong interlayer coupling effect is produced, and Moiré superlattice leading to Moiré scattering 。
This scattering introduces tunable spectral response and polarization selectivity into the transmitted light, and these responses dynamically change with changes in h, α, and detection angle θ.
Figure 3: Tunable optical response (a) Measured band structure at different vertical spacing h. The white dashed line indicates the band edge of the corresponding calculation results. With the change of interlayer spacing h, the band edge shifts significantly in frequency. The modal distribution diagram when the modal coupling changes is shown above each band structure diagram. (b) Measured band structure at different twist angles α. As the twist angle increases, the moiré vector length increases (as shown above the figure), and the gap Gm between the moiré band and the non-moiré band also increases. (c) Transmittance spectrum measurement results under different parameter combinations α, h and wave vector k. The shaded area indicates the average standard deviation of the transmittance intensity in multiple measurements. The blue dots are the experimentally measured data, and the blue curve is the fitting result. The polarization state corresponding to different quarter-wave plate angles is also indicated in the figure.
Image Source: Nature Photonics
The device design integrates two types of MEMS drive modules:
-
Vertical actuator: Adjusts the interlayer spacing h between the two photonic crystals, achieving micron-level fine adjustment through the capacitive drive principle.
-
Rotary actuator: Adjusts the relative twist angle α between the crystals, using a three-phase microstepping motor structure to achieve high-precision control with an angular resolution of up to 1/6°.
The detection angle θ can be changed through external optical path design to further expand the detection dimension.
The research uses pre-calibration of multiple sets of response functions to record the transmitted light intensity during the measurement phase and use compressed sensing + regularization optimization algorithm to reconstruct the original spectrum and polarization state.
In addition, the system can also run adaptive sensing algorithm to dynamically select the next optimal measurement configuration based on the collected information, improving sensing efficiency and speed.
In summary, this is a" Measuring light using structure A new paradigm—using MEMS-tunable Moiré structures to achieve multi-dimensional information encoding of light, and then inverting to obtain the spectrum and polarization, breaking the traditional limitation of "one device measuring one piece of information."
Figure 4: Optical sensing principle and experimental results. (a) The spectral components of the optical signal S(λ) are unknown. After passing through the sensor, a response function R is generated under a set of known configurations γ=γ 1 ，γ 2 ，γ 3 ...produces the response function R γ (λ). By measuring the transmission intensity M under various configurations, γ ，the original signal Ŝ(λ) can be reconstructed using a computational algorithm. (b) Comparison of the real and reconstructed curves of different input spectra, covering the entire working frequency range, verifying the reconstruction accuracy. (c) Comparison of errors in spectral reconstruction using adaptive and non-adaptive strategies under different numbers of measurement configurations. The data is based on the four spectral signals in Figure (b), showing the mean error and standard deviation (mean ± SD). (d) Simultaneous reconstruction of the frequency and polarization state of the signal. Frequency information is represented by color, and polarization information corresponds to the position on the Poincaré sphere. The right and bottom panels show the comparison results of the input signal (solid points) and the reconstructed signal (cross points), applicable to cases where both frequency and polarization are unknown. (e) Comparison of reconstruction errors of polarization states (ϕ and χ) using adaptive and non-adaptive strategies under different numbers of configurations. The data is based on 81 sets of measurements, showing the mean error and standard deviation (mean ± SD)
Image Source: Nature Photonics
Figure 5: Hyperspectral and hyperpolarization imaging demonstration. (a) Imaging of the target using an ordinary monochrome camera obtains an image of the luminance distribution at a single wavelength. (b) By adding a MEMS-TMPhC sensor to the optical path and reconstructing the color image through measurement data, hyperspectral imaging of the same target can be achieved. The reconstructed wavelength range is 1476 nm to 1599 nm. (c) When imaging a sample with birefringent properties using a monochrome camera, only intensity information can be obtained, and different polarization states cannot be identified. (d) After inserting the MEMS-TMPhC sensor, hyperpolarization imaging of the target can be achieved, thereby accurately reconstructing its spatial polarization distribution. The reconstructed polarization state range is 0° to 180°. The scale length in all images is 2 mm.
Image Source: Nature Photonics
Large-scale manufacturing, low-cost deployment
The MEMS-TMPhC sensor uses a CMOS-compatible process and can be manufactured with high yield (>70%) on 6-inch wafers, with mass production capabilities. Future applications include:
-
Intelligent spectrometers and chip-level optical analyzers
-
Medical imaging and optical diagnostics
-
Satellite remote sensing and high-precision polarization imaging
-
Optical encryption and information security
-
Ultra-compact quantum light sources and communication systems
Academic Contributions and Future Outlook
This research not only demonstrates the integration possibility of "multi-degree-of-freedom tunable photonic crystals" in practical devices, but also reveals a new dimension of optical Moiré structures in structured light, singular polarization control, and two-dimensional photonics. With the addition of more MEMS degrees of freedom (such as translation, stretching, and shearing), combined with new materials (such as 2D semiconductors and topological materials), the future MEMS-TMPhC platform is expected to evolve into a true "optical brain-like" system.
Paper Information
Tang，H.，Lou，B.，Du，F. et al. Adaptive Moiré sensor for spectral polarization super-imaging. Nat. Photon. (2025).
https://www.nature.com/articles/s41566-025-01650-z
LATEST NEWS
Thank you for visiting the official website of Chongfan Technology. If you have cooperation intentions or suggestions, please contact us through the following methods, and we will reply as soon as possible, thank you!
Address: Room 403, Building 6, Phase III of R&D, No. 36 Xiyong Avenue, High tech Zone, Chongqing, China.
Telephone: +86-13658337211
E-mail: Sales@cfkeji.net
Website: www.cfkeji.net
Mobile Version
