🔬 optics

Wavelength and Polarization Multiplexed Nonlocal Metasurface for Quantitative Phase Microscopy

This paper demonstrates a compact, single-shot quantitative phase microscopy technique using a wavelength and polarization multiplexed nonlocal metasurface to generate low-noise differential phase contrast images and recover quantitative phase information from unstained biological and material samples.

Original authors: Haiwei Wang, Shikun Ma, Shaban B. Sulejman, Niken Priscilla, Wendy S. L. Lee, Peter Francis Matthew Elango, Lukas Wesemann, Elizabeth Hinde, Ann Roberts

Published 2026-03-04

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Haiwei Wang, Shikun Ma, Shaban B. Sulejman, Niken Priscilla, Wendy S. L. Lee, Peter Francis Matthew Elango, Lukas Wesemann, Elizabeth Hinde, Ann Roberts

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Problem: Seeing the Invisible

Imagine you are trying to take a photo of a clear glass marble sitting on a white table. Because the marble is transparent, the camera sees nothing but white. You can't tell where the marble is, how thick it is, or what's inside it.

In the world of biology, this is a huge problem. Many living cells (like the HeLa cells mentioned in the paper) are transparent. To see them under a normal microscope, scientists usually have to "stain" them with toxic dyes or fluorescent labels. But this kills the cells or changes how they behave.

Phase Microscopy is the solution. It doesn't look at color; it looks at how much the light slows down as it passes through the cell. This "slowing down" creates a phase shift. However, human eyes and standard cameras can't see phase shifts directly. They only see brightness.

The Old Way: The Heavy, Slow Machine

Traditionally, to see these invisible phase shifts, scientists use complex, bulky machines.

The Analogy: Imagine trying to measure the shape of a hill by walking around it with a tape measure, taking hundreds of photos from different angles, and then using a supercomputer to stitch them together.
The Downside: These machines are big, expensive, slow, and often require taking multiple pictures to build one final image. If you want to watch a cell move in real-time, the machine is often too slow to keep up.

The New Solution: The "Magic Filter" (Metasurface)

This paper introduces a tiny, flat chip called a nonlocal metasurface. Think of this chip as a "magic filter" that you can slide right in front of your camera lens.

Instead of taking 100 photos and doing math later, this chip does the heavy lifting instantly as the light passes through it. It turns the invisible phase shifts into visible brightness changes in a single snapshot.

How It Works: The Traffic Director

The researchers built this chip using tiny, microscopic pillars (nanorods) arranged in a specific pattern. Here is how they made it work using two clever tricks: Polarization and Wavelength.

1. The Polarization Trick (The Left-Handed vs. Right-Handed Gloves)

Light can vibrate in different directions. The chip is designed to treat "Left-Handed" light and "Right-Handed" light differently.

The Analogy: Imagine a bouncer at a club. If you enter wearing a "Left" glove, he points you to the left exit. If you wear a "Right" glove, he points you to the right exit.
In the Paper: When light passes through the sample and hits the chip, the chip splits the image into two versions simultaneously: one that highlights edges going up/down and another that highlights edges going down/up. By subtracting these two images, the computer gets a clear, high-contrast picture of the cell's edges without any background noise.

2. The Wavelength Trick (The Color Switch)

The chip is also sensitive to color (wavelength).

The Analogy: Imagine a traffic light that changes the direction of traffic based on the color of the car. Red cars go North; Blue cars go East.
In the Paper: The researchers use two specific colors of light (613 nm and 656 nm).
- At Color A, the chip highlights vertical edges.
- At Color B, the chip highlights horizontal edges.
Why this matters: To get a perfect 3D map of the cell (Quantitative Phase Imaging), you need to know the slope in both directions. By quickly switching colors, the chip captures both the vertical and horizontal slopes in one go.

The Result: A Single-Shot, Crystal Clear Map

By combining these tricks, the researchers created a system that:

Takes one photo (Single-shot).
Uses no toxic dyes (Unstained cells).
Produces a digital map showing exactly how thick the cell is and how much "mass" it has.

They tested this on:

Computer-generated patterns: To prove the math works perfectly.
Real HeLa cells: They successfully imaged living cancer cells, showing details like the nucleus and organelles, with measurements that matched what is known in scientific literature.

Why This Matters for the Future

Currently, high-end microscopes are like mainframe computers: huge, expensive, and require a dedicated room.

This new metasurface is like a smartphone camera app.

It is tiny (microscopic size).
It is cheap to manufacture.
It can be added to existing microscopes to make them super-powerful.

The Bottom Line:
This research paves the way for portable, pocket-sized microscopes that doctors could use in remote villages or emergency rooms to instantly analyze blood or tissue samples without needing a lab full of equipment. It turns a complex, slow scientific process into a simple, instant snapshot.

1. Problem Statement

Imaging transparent, unstained biological samples (e.g., live cells) and materials remains a significant challenge in microscopy. Traditional phase contrast techniques like Zernike phase contrast and Differential Interference Contrast (DIC) provide qualitative images but suffer from limitations such as alignment sensitivity, artifacts, and the inability to directly measure phase values. While Quantitative Phase Microscopy (QPM) offers numerical phase maps useful for calculating dry mass and cell growth, existing QPM methods typically rely on:

Complex and bulky interferometric systems.
Non-interferometric methods requiring multiple image captures, which limits temporal resolution and increases computational load.
Lack of compactness, hindering deployment in point-of-care or remote settings.

The authors aim to develop a single-shot, compact, and low-noise QPM system using metasurfaces to overcome these trade-offs between system complexity, cost, and information content.

2. Methodology

The core innovation is a single-layer nonlocal metasurface placed in the object plane that performs direct spatial frequency filtering. The system utilizes a multiplexing strategy involving two wavelengths and two circular polarizations to extract phase gradients along orthogonal directions.

Metasurface Design

Structure: A periodic array of plasmonic nanorods (Silver/Chromium) on a Silicon Nitride waveguiding layer, sandwiched between a PMMA superstrate and a glass substrate.
Mechanism: The design employs chiral guided-mode optical routing. Unlike symmetric resonant gratings that produce parabolic dispersion, this metasurface exhibits linear, asymmetric spatial dispersion about normal incidence.
Multiplexing Strategy:
- Polarization: The metasurface routes incident light into guided modes based on handedness. By using linearly polarized input and analyzing Left (LCP) and Right (RCP) circular polarizations simultaneously, the system generates two images with opposite gradient contrast (one positive, one negative) for a specific direction.
- Wavelength: The unit cell dimensions are engineered such that guided modes propagate along the $x$ -axis at one wavelength and the $y$ -axis at another. This allows the system to switch the direction of gradient sensitivity (horizontal vs. vertical) by simply changing the illumination wavelength.

Imaging and Reconstruction Process

Acquisition: A sample is illuminated with horizontally polarized light at two specific wavelengths (613 nm and 656 nm). A polarization-sensitive camera captures simultaneous LCP and RCP images.
Differential Phase Contrast (DPC) Calculation: For each wavelength, the DPC image is calculated by subtracting the LCP and RCP intensities and normalizing by their sum:
$I_{DPC} = \frac{I_+ - I_-}{I_+ + I_-}$
This process cancels out common background noise (e.g., speckle) and removes intensity variations, isolating the phase gradient signal.
Phase Retrieval: The DPC images from both wavelengths (providing $x$ and $y$ gradient information) are used as inputs for a non-iterative deconvolution algorithm with Tikhonov regularization to reconstruct the quantitative phase map $\phi(\rho)$ .

3. Key Contributions

Single-Shot Multi-Axis DPC: The system achieves simultaneous acquisition of gradient-reversed images via polarization multiplexing, eliminating the need for mechanical scanning or multiple exposures for a single axis.
Wavelength-Dependent Axis Switching: By leveraging the dispersive nature of the guided-mode resonances, the system captures orthogonal gradient information ( $x$ and $y$ ) by switching wavelengths, enabling robust isotropic phase recovery without complex optical reconfiguration.
Noise Reduction: The differential subtraction of LCP and RCP channels effectively removes background intensity noise and speckle, resulting in high signal-to-noise ratio (SNR) phase images.
Compact Architecture: The metasurface replaces bulky beam splitters and interferometric components, allowing for a significantly smaller footprint suitable for integration into portable devices.

4. Results

The authors validated the system through simulations, optical characterization, and imaging experiments:

Characterization: Angle-dependent transmission spectra confirmed the asymmetric dispersion and polarization routing behavior. Experimental results showed good agreement with simulations (with a minor 40 nm blue-shift attributed to fabrication tolerances).
SLM Validation: The system was tested using a Spatial Light Modulator (SLM) to generate known phase targets (sinusoidal star patterns).
- The retrieved phase maps showed excellent agreement with the target phase values.
- The system demonstrated isotropic resolution, recovering phase features equally well in all directions.
- Signal-to-noise ratio improved with increasing phase excursion.
Biological Imaging (HeLa Cells): The system successfully imaged unstained HeLa (cervical cancer) cells.
- Quantitative Data: The recovered phase excursion values were consistent with literature values for similar wavelengths.
- Comparison: The retrieved phase was used to synthesize a DIC image, which was compared against a commercial DIC microscope. While the commercial DIC offered higher spatial resolution (due to higher NA), the metasurface system provided superior noise immunity and quantitative data in a single shot.
- Resolution: The effective Numerical Aperture (NA) was limited to 0.05, resulting in a spatial resolution of approximately 6.6 μm at 656 nm.

5. Significance and Future Outlook

This work demonstrates a viable pathway toward ultracompact, next-generation QPM systems. By integrating nonlocal metasurfaces with multiplexed optical filtering, the authors have created a platform capable of extracting rich phase information from transparent samples without staining or complex alignment.

Applications: The technology is highly relevant for medical diagnostics, biotechnology (e.g., cell growth monitoring), and materials science.
Future Work: The authors aim to extend the system to operate at a single wavelength (using spectral filters for single-shot acquisition) and increase the effective NA to improve spatial resolution. They also anticipate integrating these metasurfaces with other meta-optical components to create fully integrated, high-performance imaging devices for point-of-care use.