Plasmon-Enabled High-Precision Single Molecule Localization Microscopy over an Extended Field of View

Technical Summary: Plasmon-Enabled High-Precision Single Molecule Localization Microscopy (PIFLUX)

Problem Statement
Single-molecule localization microscopy (SMLM) has revolutionized biophysics by enabling molecular-scale mapping beyond the diffraction limit. However, conventional SMLM localization precision is fundamentally limited by the scaling law $\lambda/(NA\sqrt{N})$, where $N$ is the number of detected photons. Achieving few-nanometer precision typically requires large photon budgets, which are constrained by photobleaching and blinking kinetics. While patterned-illumination approaches like MINFLUX achieve $\sim1$ nm precision, they suffer from an intrinsically small field of view (FOV, $\sim50\text{--}100$ nm) and low throughput due to sequential excitation. Conversely, widefield patterned-illumination methods like SIMFLUX offer large FOVs but are limited by the diffraction-limited pitch of the illumination pattern, restricting precision improvements to roughly a twofold enhancement over conventional SMLM. The central challenge is to reduce the illumination-pattern pitch to approach MINFLUX-level precision while retaining the large FOV and parallelism of widefield detection.

Methodology
The authors propose PIFLUX (Plasmon-Illumination FLUX), a scheme that integrates deep-subwavelength plasmonic illumination with widefield camera detection. The core of the method relies on a Water-Metal-Dielectric-Metal-Water (WMDMW) architecture. This structure supports gap-plasmon modes where counter-propagating surface plasmon polaritons (SPPs) interfere with a normally incident optical field.

Key technical components include:

• Illumination Generation: The interference between counter-propagating gap plasmons and an incident plane wave generates a standing-wave illumination pattern. The spatial period of this pattern ($\Lambda_{eff} = 2\pi/\beta'$) is determined by the in-plane propagation constant $\beta'$ of the gap plasmon, which can be significantly larger than the free-space wavevector, enabling deep-subwavelength pitches.
• Tunability: The spatial frequency ($\beta'$) and confinement of the plasmonic field are tunable via excitation energy, dielectric spacer thickness ($t_d$), and metal film thickness ($t_m$). The position of the interference pattern can be tuned by adjusting the plasmon phase ($\phi$) without altering the spatial period.
• Acquisition Protocol: For each localization event, six images are acquired: three with the excitation pattern phase-shifted along the $x$-axis and three along the $y$-axis (successive phase shifts of $2\pi/3$). This allows the extraction of positional information by correlating photon counts with the known excitation patterns.
• Analysis: The authors employ a Cramér-Rao bound (CRB) analysis based on Fisher information to theoretically evaluate localization precision. Additionally, a multi-stage maximum-likelihood estimator (MLE) is developed and validated using synthetic data representing a nuclear pore complex (NPC).

Key Contributions and Results

1. Theoretical Precision: The CRB analysis demonstrates that PIFLUX achieves few-nanometer localization precision over a micrometer-scale FOV. Specifically, PIFLUX with a 6 nm spacer ($t_d=6$ nm) provides an overall lateral precision ($\sigma_r$) that is approximately 93% better than SIMFLUX and even 47% better than MINFLUX (under high signal-to-background ratio conditions) across a $\sim10$ $\mu$m range.
2. Field of View and Parallelism: Unlike MINFLUX, which addresses emitters sequentially, PIFLUX utilizes widefield detection, maintaining a FOV of several micrometers and enabling parallel localization of multiple emitters.
3. Parameter Optimization: The study identifies that thinner dielectric spacers (e.g., 6 nm) yield higher spatial frequencies and better precision but may be more challenging to fabricate. A 10 nm spacer offers a favorable tradeoff, still providing a 50% improvement over SIMFLUX and a larger region where both $x$ and $y$ precisions exceed those of SIMFLUX.
4. Experimental Validation (Synthetic): Using a synthetic NPC model with 500 photons per emitter, the MLE confirmed the theoretical predictions.
   • In high signal-to-background conditions, PIFLUX ($t_d=6$ nm) achieved a median radial error (MRE) of 2.2 nm, outperforming SIMFLUX (4.3 nm) and matching or slightly exceeding MINFLUX (2.6 nm).
   • In low signal-to-background conditions, PIFLUX maintained superior performance (MRE $\sim3.8$ nm) compared to SIMFLUX (7.1 nm) and remained competitive with MINFLUX (3.7 nm).

Significance
The paper claims that PIFLUX represents a significant advancement by decoupling localization precision from the diffraction-limited illumination pitch. By exploiting the large in-plane wavevectors of evanescent plasmonic near fields, the method achieves sub-diffraction illumination patterns over an extended field of view. The authors posit that PIFLUX offers a route to high-throughput fluorescence imaging with single-digit nanometer resolution, bridging the gap between the high precision of MINFLUX and the large FOV/parallelism of SIMFLUX, without requiring sequential emitter-by-emitter excitation. The work suggests that plasmonic architectures can serve as versatile platforms for nanophotonic integration and high-resolution imaging, extending the capabilities of structured illumination beyond the constraints of conventional far-field optics.