PIFLUX: Plasmonic Localization Microscopy

• The paper introduces PIFLUX, a technique that combines deep-subwavelength plasmonic illumination with widefield detection to achieve few-nanometer localization precision.
• Plasmon-enabled single-molecule localization employs a water–metal–dielectric–metal–water multilayer to generate tunable, high-spatial-frequency illumination patterns decoupled from far-field diffraction limits.
• Validation on a synthetic nuclear pore complex shows PIFLUX outperforms MINFLUX and SIMFLUX, achieving up to 93% improvement in precision and over 100-fold area gain.

Plasmon-enabled high-precision single-molecule localization microscopy denotes a class of single-molecule localization microscopy (SMLM) schemes that exploit near-field plasmonic structuring to obtain localization information beyond the spatial-frequency ceiling of conventional far-field interference. In the formulation proposed as PIFLUX (Plasmon-Illumination FLUX), deep-subwavelength plasmonic illumination is combined with widefield detection, so that few-nanometer localization precision is maintained over a micrometer-scale field of view (FOV). The central mechanism is phase-tunable interference between counter-propagating gap plasmons and a normally incident optical field, yielding a laterally translatable illumination pattern whose spatial period is preserved during phase shifts; this design is intended to overcome the limited FOV and throughput of MINFLUX while surpassing the diffraction-limited pitch constraint of SIMFLUX (Shaukat et al., 30 Jun 2026).

1. Conceptual position within localization microscopy

PIFLUX is situated at the intersection of structured-illumination localization and plasmonic near-field engineering. Standard centroid-based SMLM is photon-limited, with localization precision scaling approximately as $\lambda/(\mathrm{NA}\sqrt{N})$, so improvements are usually sought either by collecting more photons or by increasing the spatial information encoded per detected photon. Patterned-illumination methods raise the latter quantity by modulating the excitation field and inferring emitter position from phase-dependent photon counts rather than from the emission point-spread function alone.

Within that landscape, MINFLUX and SIMFLUX define two distinct operating regimes. MINFLUX achieves very high precision through a donut excitation pattern, but its field of view is limited to approximately $100$ nm because the excitation is sequential. SIMFLUX uses widefield far-field sinusoidal patterns and doubles the precision of standard SMLM, but remains limited by diffraction-limited pitch. PIFLUX combines widefield detection with a subwavelength illumination period generated by surface plasmon polaritons (SPPs), and thereby seeks MINFLUX-class precision without sacrificing micrometer-scale coverage (Shaukat et al., 30 Jun 2026).

A common oversimplification is that few-nanometer precision in patterned-illumination localization is inseparable from either a tightly confined sequential excitation geometry or a diffraction-limited sinusoidal pitch. PIFLUX suggests a different division of labor: the near field supplies the high spatial frequency, while the camera preserves widefield parallelism.

2. Plasmonic architecture and illumination physics

The physical platform is a water–metal–dielectric–metal–water (WMDMW) multilayer structure. Its key element is a nanoscale dielectric spacer of thickness $t_d$ sandwiched between thin metal films of thickness $t_m$. This geometry supports gap-plasmon modes with in-plane wavevectors $\beta'$ much larger than those available to far-field light, which is the origin of the reduced pattern pitch.

The illumination profile is generated by interfering counter-propagating gap plasmons across the spacer together with a normally incident optical field. Along the interface, the total intensity is

$$I(x) = 4e^{-2\beta''|x|} + |E_a|^2 + 4E_0e^{-\beta''|x|}\cos(k_0\sin\chi\,x+\theta)\cos(\beta'x-\varphi).$$

In this expression, $\beta'$ is the real part of the plasmon propagation constant and sets the spatial frequency, while $\beta''$ sets attenuation. The effective pattern period is

$\Lambda_{\rm eff} = \frac{2\pi}{\beta'},$

which can be much smaller than the conventional optical limit of approximately $\lambda/2\mathrm{NA}$. The phase parameter $100$0 laterally translates the standing-wave pattern without changing its spatial period, and that invariance is central to localization because it decouples pattern displacement from pitch variation.

The engineering knobs follow directly from the gap-plasmon dispersion. Smaller gap thickness $100$1, thinner metal thickness $100$2, or higher excitation energy increase $100$3, allowing tunable deep-subwavelength patterning. This tunability is not merely a design convenience; it is the mechanism by which localization precision is decoupled from the far-field diffraction scale.

3. Data acquisition and localization inference

For each molecular blink, PIFLUX acquires six frames: three with phase shifts along $100$4 and three along $100$5, using $100$6. The emitter position is then inferred from the photon counts measured under the known SPP modulation pattern. Because the illumination phase is externally controlled and the pattern pitch is subwavelength, the relative count vector across the six measurements contains substantially more positional information per detected photon than a conventional widefield localization measurement (Shaukat et al., 30 Jun 2026).

The inferential framework is maximum-likelihood based. The paper reports that multi-stage maximum-likelihood estimators confirm the theoretical predictions, including on a synthetic nuclear pore complex. This is significant because the gain claimed for PIFLUX does not rest only on a local Fisher-information argument; it is also supported at the estimator level for a nontrivial synthetic structure.

An important practical consequence of the acquisition design is parallelism. Unlike sequential MINFLUX excitation, PIFLUX and SIMFLUX are camera-based widefield schemes and can localize many emitters simultaneously. This suggests that the method addresses throughput not by relaxing precision, but by moving the precision-enhancing pattern into a plasmonic widefield format.

4. Statistical precision and comparison with MINFLUX and SIMFLUX

The performance analysis is based on the Cramér-Rao bound (CRB) derived from the Fisher information under Poissonian photon-counting statistics. The scalar lateral precision metrics are $100$7 and $100$8, and the isotropic lateral metric is

$100$9

This metric is used to compare PIFLUX with MINFLUX and SIMFLUX over both position and signal-to-background regimes.

For a 6 nm gap, PIFLUX achieves mean lateral localization precision of approximately 2.2 nm, described as matching or exceeding MINFLUX and doubling SIMFLUX. For 500 detected photons, the reported median radial errors (MRE) are approximately 2.2 nm for PIFLUX with $t_d$0 nm, 2.6 nm for MINFLUX, and 4.3 nm for SIMFLUX. The minimum CRB for PIFLUX is reported below 2 nm. In best-case conditions, the paper reports that $t_d$1 for PIFLUX is up to 93% better than SIMFLUX and up to 47% better than MINFLUX at high signal-to-background ratio (Shaukat et al., 30 Jun 2026).

Field of view is the second axis of comparison. MINFLUX is described as limited to roughly $t_d$2 nm, whereas PIFLUX maintains high precision over multi-micrometer ranges, including consistently better-than-SIMFLUX performance over approximately 10 $t_d$3m. The stated area gain relative to MINFLUX is greater than 100-fold. Taken together, these metrics define the method’s central claim: plasmonic near-field structuring can push localization precision beyond the far-field interference limit without collapsing the usable FOV.

5. Synthetic nuclear pore complex reconstruction

The demonstration system is a synthetic nuclear pore complex (NPC) comprising eight emitters on a 110 nm ring, with 500 photons per emitter and 1 background photon per pixel. All compared methods reconstruct the ring, but PIFLUX yields the smallest median radial errors (Shaukat et al., 30 Jun 2026).

The significance of this example is not only geometric familiarity. An NPC ring is a structured, multi-emitter target with a scale commensurate with nanometer-level localization claims. The paper further reports that PIFLUX maintains nanometer precision even when the NPC is displaced across the extended FOV, on the order of a micron from the origin. This is presented as evidence for both uniformity and scalability, and therefore for the suitability of the method to parallel, high-throughput imaging rather than only to local, origin-centered operation.

This synthetic validation also clarifies what the method does and does not establish. It confirms estimator behavior and pattern uniformity in simulation, but it does not by itself constitute an experimental biological imaging result. A plausible implication is that the principal uncertainty shifts from statistical localization theory to fabrication, alignment, and experimental realization of the WMDMW plasmonic platform.

6. Significance, interpretation, and outlook

The technical significance of PIFLUX lies in its decoupling of localization precision from the diffraction limit. Because the effective pitch is set by the gap-plasmon wavevector $t_d$4 rather than by far-field interference geometry, the illumination can access spatial frequencies that conventional widefield patterned illumination cannot. The method therefore reassigns the limiting scale from free-space optics to nanostructured plasmonic dispersion (Shaukat et al., 30 Jun 2026).

Two further advantages follow directly from that choice. First, the spatial period is tunable by geometry and excitation energy, which gives the platform flexibility across experimental requirements. Second, because the detection remains widefield, the method preserves the simultaneous multi-emitter localization capability associated with camera-based SMLM.

The main conceptual caution is that PIFLUX is a proposal validated by Cramér-Rao analysis and by maximum-likelihood reconstruction on synthetic data. Its present status, as described, is therefore methodological and predictive rather than already established as a mature experimental instrument. Even so, the reported combination of few-nanometer precision, micrometer-scale FOV, and plasmonically tunable subwavelength patterning suggests a route toward high-throughput, nanometer-resolved imaging of macromolecular assemblies such as nuclear pore complexes, and more generally toward structured-illumination localization microscopy in regimes inaccessible to purely far-field methods.