🔬 optics

Programmable photonic nanojets via phase-only time-reversal: a numerical study

This paper presents a numerical study demonstrating a phase-only time-reversal framework that enables robust, mechanically static steering and subwavelength confinement of programmable photonic nanojets, showing high tolerance to fabrication and alignment errors.

Original authors: Tobias Abilock Mikkelsen, Cristian Placinta, Jesper Glückstad, Mirza Karamehmedović

Published 2026-04-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Tobias Abilock Mikkelsen, Cristian Placinta, Jesper Glückstad, Mirza Karamehmedović

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 Idea: Steering Light Without Moving Parts

Imagine you have a tiny, super-bright flashlight beam (called a Photonic Nanojet) that can see things smaller than a hair or manipulate tiny particles. Usually, to move this beam to a different spot, you have to physically move the lens or the object, like adjusting a microscope. This is slow, clunky, and prone to breaking.

This paper proposes a "magic trick" to move that beam instantly using only software and a special screen, with no moving parts and no dimming of the light.

The Problem: The "Foggy Window"

Think of a glass marble (the microelement) sitting on a table. If you shine a flashlight at it, the light bends and creates a tiny, super-intense spot of light on the other side. That's the Photonic Nanojet.

The problem is: If you want to move that intense spot to a different location, you usually have to:

Physically move the marble.
Or, use a complex system to change the shape of the flashlight beam.

Both methods are slow or require expensive, complicated hardware.

The Solution: The "Time-Reversal" Trick

The authors use a concept called Time-Reversal. Here is the analogy:

Imagine you are in a room with a broken window. You throw a ball, it hits the window, bounces off the walls, and lands in a specific corner.

Normal Physics: You throw the ball, and it goes where it goes.
Time-Reversal: Imagine you could film the ball landing in that corner, and then play the movie backwards. The ball would fly out of the corner, bounce off the walls in reverse, and land perfectly back in your hand.

The researchers do this with light:

The "Fake" Source: They pretend a light source exists exactly where they want the nanojet to be. They calculate how light would scatter out from there, hit the glass marble, and bounce back to a control screen.
The "Reverse" Signal: They take that scattered light, flip it backwards in time (mathematically), and look at the pattern of the light hitting the control screen.
The Magic Screen: They take that pattern and display it on a Spatial Light Modulator (SLM). Think of the SLM as a high-tech "digital window" that can twist the light waves.
The Result: When they shine a real laser through this "twisted" window, the light travels forward, hits the glass marble, and—because of the time-reversal math—it naturally focuses perfectly into a nanojet at the exact spot they wanted.

The best part? They only twist the timing (phase) of the light, not its brightness. This means the system is 100% efficient; no light is wasted.

What They Discovered

1. Shape Doesn't Matter Much
They tested thousands of weird, random shapes for the glass marble (using a mathematical formula called the "superformula" that can make shapes look like stars, blobs, or diamonds).

The Finding: It turns out, you don't need a perfectly engineered, complex shape. Even a simple square or a circle works almost as well as the most complex shapes.
The Analogy: It's like trying to tune a radio. You might think you need a super-complex antenna to get a clear signal, but a simple wire works just fine if you know how to tune the frequency. The "software" (the time-reversal math) does the heavy lifting, not the shape of the object.

2. It's Tough to Break (Robustness)
In the real world, things aren't perfect. The glass marble might be slightly the wrong size, or the material might be slightly impure, or it might be placed a tiny bit off-center.

The Finding: The system is very forgiving. Even if the marble is slightly "wrong," the light still focuses correctly.
The Analogy: It's like a GPS navigation system. If you take a slightly wrong turn or the road is a bit bumpy, the GPS recalculates and still gets you to the destination. The system is "smart" enough to handle small errors without crashing.

3. One Screen, Many Beams
They showed that you can program the screen to create multiple nanojets at the same time.

The Analogy: Instead of one spotlight, you can make a "light show" with several spotlights dancing around the stage, all controlled by one single computer screen.

Why Does This Matter?

This technology could revolutionize how we build microscopes, laser cutters, and medical tools.

No Moving Parts: No motors to wear out or break.
Super Fast: You can switch the beam's position in microseconds (faster than a blink).
Cheap & Simple: You don't need a custom-made, perfect glass marble for every job. A simple, cheap square piece of glass works just fine because the computer does the hard work.

In a nutshell: They found a way to use a "time-travel" math trick to steer a super-tiny laser beam with a digital screen, proving that you don't need fancy, expensive hardware to get perfect results—just a smart algorithm and a simple piece of glass.

1. Problem Statement

Photonic Nanojets (PNJs) are highly concentrated, high-intensity optical structures formed on the shadow side of illuminated mesoscale dielectrics. They are critical for applications such as super-resolution imaging, nanopatterning, optical manipulation, and label-free metrology.

However, a significant barrier to their widespread deployment is the lack of agile, calibration-friendly control over the PNJ's position and morphology. Existing methods rely on:

Mechanical motion: Physically moving the microelement or the beam, which is slow and complex.
Microelement shaping: Fabricating specific geometries or arrays, which lacks flexibility.
Beam shaping: Controlling focus and waist, which often requires complex hardware and multi-parameter calibrations.

The authors aim to develop a method for software-defined PNJ control that allows for wide-range steering (lateral and axial) without mechanical movement or amplitude modulation, using only phase control.

2. Methodology

The paper proposes a phase-only time-reversal framework implemented via numerical simulation. The core concept utilizes the reciprocity of wave propagation to determine the necessary incident field phase to create a PNJ at a specific target location.

A. Time-Reversal Algorithm

The method operates in two main steps:

Backward Simulation (Synthetic Source): A synthetic source $s(\mathbf{r})$ (modeled as a Gaussian) is placed at the desired target PNJ location ( $\mathbf{r}_{PNJ}$ ). The wave propagation is simulated backward through the microelement to a control line ( $\Gamma_{ctrl}$ ) located above the element. This yields a scattered field $H_{back}$ .
Phase Extraction & Forward Illumination: The field on the control line is decomposed into amplitude $A(x)$ and phase $\phi(x)$ . To achieve a phase-only implementation compatible with a Spatial Light Modulator (SLM), the amplitude is flattened to a constant, and only the conjugated phase is retained:
$H_{inc}(x) = e^{-i\phi(x)}$
This phase profile is then used as the incident field in a forward scattering simulation. Due to time-reversal symmetry, this field reconstructs the energy at the target location, forming the PNJ.

B. Numerical Implementation

Solver: Full-wave Finite-Difference Frequency-Domain (FDFD) simulations using the Python library ceviche.
Physics: 2D Transverse Electric (TE) time-harmonic scattering.
Parameters: Free-space wavelength $\lambda_0 = 532$ nm; Microelement refractive index $n_M = 1.49$ ; Ambient index $n_0 = 1$ .
Grid: High-resolution grid (30 points/ $\mu$ m) with Perfectly Matched Layers (PML) to simulate an open domain.
Metrics: PNJ quality is quantified using Waist Width ( $w_{PNJ}$ ) and Axial Length ( $\ell_{PNJ}$ ) defined by the Full Width at Half Maximum (FWHM) of the intensity profile along the local Poynting vector direction.

3. Key Contributions

Phase-Only Steering Framework: Demonstrated a lossless setup using a single SLM and a fixed microelement to steer PNJs across a broad lateral and axial range without mechanical parts or amplitude modulation.
Geometric Insensitivity: Conducted a parametric study using the "superformula" to generate 1,000 random microelement shapes. Found that PNJ formation is largely insensitive to moderate boundary variations, suggesting that simple, high-symmetry shapes are sufficient.
Multi-Target Capability: Showed that the framework can generate multiple simultaneous nanojets by superposing backward sources from multiple target locations, with weighting factors used to balance intensity based on axial depth.
Robustness Analysis: Performed uncertainty quantification (UQ) to assess the system's tolerance to fabrication errors (radius, refractive index) and alignment errors (lateral position).

4. Key Results

A. Steering Performance

The method successfully demonstrated robust lateral and axial steering. The phase profiles derived from the time-reversal process successfully focused energy to target coordinates (e.g., varying from $y=-0.5\mu m$ to $y=-6\mu m$ ) with subwavelength confinement and low sidelobes.

B. Geometric Optimization vs. Simplicity

Superformula Search: A search through 1,000 random shapes identified that shapes with four dominant corners performed well.
The "Square" Surprise: When comparing the best-performing complex superformula shapes against a simple square microelement, the square outperformed the complex shapes by approximately 40% in a composite score (intensity/confinement/length).
Conclusion: Simple, highly symmetric shapes (like squares) provide a stable, predictable mapping that is more efficient than complex, anisotropic shapes for time-reversal steering.

C. Uncertainty Quantification (Robustness)

The system was tested against perturbations in:

Refractive Index ( $\delta n \in [-0.05, 0.05]$ )
Radius ( $\delta R \in [-0.03, 0.03]$ )
Alignment ( $\delta x, \delta y \in [-0.10, 0.10] \mu m$ )

Findings:

Waist Width: Showed moderate sensitivity (Coefficient of Variation, CV $\approx$ 0.15–0.22).
Axial Length: Showed higher sensitivity (CV $\approx$ 0.35–0.57), indicating that the length of the jet is more affected by errors than the width.
Peak Intensity: Generally robust, with CVs ranging from 0.08 to 0.32 depending on the error type.
Conclusion: The system is robust enough for practical applications, particularly regarding alignment and refractive index variations, though axial length stability requires careful fabrication.

D. Multiple Nanojets

The study successfully generated double-nanojets. It was observed that targets at the same axial depth required equal weighting, while targets at different depths required progressive weighting to compensate for propagation losses and interference imbalances.

5. Significance and Impact

Hardware Simplification: This approach eliminates the need for complex mechanical scanning stages or multi-element arrays, relying instead on a single fixed microelement and a programmable SLM.
Design Flexibility: The finding that simple shapes (squares) outperform complex optimized shapes simplifies the manufacturing process for PNJ devices.
Application Readiness: The demonstrated robustness to fabrication and alignment errors suggests this method is viable for real-world applications in microscopy, nanofabrication, and optical trapping.
Scalability: While the study is 2D, the authors note the framework generalizes directly to 3D, paving the way for volumetric control of optical fields.

In summary, the paper establishes a software-defined, phase-only control paradigm for photonic nanojets that is robust, efficient, and compatible with standard optical hardware, overcoming the limitations of mechanical and amplitude-based control methods.