Photonic Force Microscopy

• Photonic Force Microscopy is an advanced scanning probe technique that uses optical tweezers to trap nano- or micro-scale probes for high-resolution force mapping.
• It employs a laser-based, non-contact optical trap to detect sub-piconewton forces and nanometer displacements, offering quantitative nanometrology.
• PFM overcomes AFM limitations by minimizing sample artefacts and enabling sensitive imaging of soft interfaces such as living cell membranes.

Photonic Force Microscopy (PFM) is a class of scanning probe techniques in which optical tweezers (OT) are used to trap and manipulate nano- or micro-scale probes, enabling quantitative, high-resolution mapping of local intermolecular and interfacial forces. PFM unifies principles from optical trapping, interferometric displacement sensing, and scanning probe microscopy, offering unique capabilities for nanometrology, mechanobiology, and photonics. By leveraging a laser-based, non-contact trapping "cantilever" with force sensitivity in the sub-piconewton regime, PFM achieves nanometer spatial resolution while minimizing artefacts on delicate or soft samples—such as living cell membranes—where conventional Atomic Force Microscopy (AFM) faces fundamental constraints (2002.01533, Pal et al., 2011, Huang et al., 2016).

1. Core Operating Principles

PFM replaces the physical cantilever of AFM with an optically trapped probe particle, typically a dielectric microsphere or a nano-engineered tip. The basic trapping setup involves a tightly focused infrared laser (commonly 1064 nm, 20–60 mW at the objective) through a high-NA (≥1.2) water- or oil-immersion objective to create a three-dimensional intensity gradient. The optical gradient force confines the particle, and its stochastic Brownian motion is restricted by the trap’s restoring force, which is characterized by a trap stiffness $k_\mathrm{OT}$ (2002.01533, Pal et al., 2011).

The sample, mounted on a piezoelectric positioner, is raster-scanned beneath the stationary optical trap. Features on the sample surface deflect the probe from the trap center, generating nanometer-scale lateral ($\Delta x$, $\Delta y$) and axial ($\Delta z$) displacements that are detected via forward- or back-scattered light, using either quadrant photodetectors (QPDs) or interferometric detection systems (Pal et al., 2011). These displacements are then mapped to forces using Hooke’s law:

$F_i = -k_i x_i, \qquad i = x, y, z$

Trap stiffness values in PFM are typically $k_\mathrm{OT} \approx 0.001$–$0.1$ pN/nm, i.e., 2–3 orders of magnitude softer than standard AFM cantilevers ($k_\mathrm{AFM} \sim 10$ pN/nm), enabling highly sensitive force detection down to $\sim 0.1$ pN in optimized setups (2002.01533).

2. Probe Design and Nano-Fabrication

Resolution in PFM is fundamentally limited by the dimensions of the trapped probe. Historically, micron-sized dielectric beads (e.g., polystyrene, silica, or quartz; $\sim$1 μm diameter) have been used, which restrict achievable lateral resolution. Recent advances have demonstrated the massive parallel nano-fabrication of optically trappable quartz particles with sharp AFM-like tips (apex radius as small as 35 nm) via the following protocol (2002.01533):

• Start with x-cut single-crystal quartz wafer coated with 0.8 μm PECVD SiO₂ and 60 nm Cr.
• Pattern 2D arrays of 1 μm-diameter discs at 1.8 μm pitch via two-step laser interference lithography (LIL).
• Etch SiO₂/quartz pillars ($\Delta x$02.6 μm tall, 80° sidewall) using ICP-RIE.
• Wet etch in 5% HF to selectively remove SiO₂, yielding a 35 nm-radius tip atop each pillar.
• Remove Cr mask; mechanically harvest particles into suspension. Yields are $\Delta x$1 particles per batch, with $\Delta x$230 % retaining intact tips.

This approach yields probes whose sharpness and geometry rival those of AFM tips, enabling imaging with lateral resolution approaching $\Delta x$3nm (2002.01533). This suggests that continued improvements in tip engineering and material selection (e.g., titanium dioxide for higher refractive index) could further enhance the spatial resolution and trapping stiffness.

3. Detection, Calibration, and Noise Performance

Precise force measurements in PFM depend on high-sensitivity displacement detection and meticulous calibration. Displacement signals are typically acquired via QPDs or photodiodes, with voltage signals $\Delta x$4 normalized and converted to true particle displacement through experimentally determined sensitivities (e.g., $\Delta x$5 at 532 nm for 1.1 μm beads) (Pal et al., 2011).

The stochastic motion of an optically trapped particle is governed by overdamped Langevin dynamics. The typical power spectral density (PSD) of displacement follows a Lorentzian:

$\Delta x$6

where $\Delta x$7 and $\Delta x$8 is the drag coefficient. Fitting the PSD yields trap stiffness $\Delta x$9 and drag, with $\Delta y$0 errors for multiple bead sizes (Pal et al., 2011).

Key experimentally reported performance metrics:

Metric	Value/Range	Reference
Trap stiffness ($\Delta y$1)	$\Delta y$2–$\Delta y$3 pN/nm	(2002.01533)
Absolute displacement noise	$\Delta y$4–$\Delta y$5 nm/10–1 ms	(Pal et al., 2011)
Lateral/axial resolution	$\Delta y$6 nm	(2002.01533)
Minimal measurable force	$\Delta y$7 pN	(2002.01533)
Bandwidth (QPD)	>2.5 MHz	(Pal et al., 2011)
Linear response range	$\Delta y$8 nm	(Pal et al., 2011)
Crosstalk between axes	$\Delta y$9%	(Pal et al., 2011)

The use of miniature QPDs—as developed from CD pickup heads—optimizes the photosensitive area-to-beam-radius ratio for maximal sensitivity and low-noise detection. Thermal drift and electronic noise limit resolution for averaging times $\Delta z$0 ms; active mechanical and laser stabilization can mitigate these effects (Pal et al., 2011).

4. Imaging Protocols and Experimental Workflows

PFM imaging involves raster-scanning the sample with a piezo stage in a grid (e.g., $\Delta z$1 pixels over $\Delta z$2, i.e. 10 nm step size). At each pixel, integration times of $\Delta z$3 ms are typical; signals $\Delta z$4, $\Delta z$5, $\Delta z$6, and piezo stage positions are recorded at sub-microsecond intervals. Mean values over the dwell time encode image pixel values; custom drift-correction functions subtract global tilt. By using both $\Delta z$7 for imaging and scan-path correction, sub-nanometer lateral resolution is achievable (2002.01533).

Power spectral or Allan deviation analysis guides the choice of dwell and integration times for optimal SNR. Drift limits displacement accuracy to $\Delta z$8 nm for $\Delta z$9 s at the surface; integration times of $F_i = -k_i x_i, \qquad i = x, y, z$0 s in bulk fluids are required for 1 nm accuracy (2002.01533).

5. Comparison with Atomic Force Microscopy

PFM achieves comparable spatial resolution to AFM on rigid substrates but with far lower probing forces, circumventing tip-induced deformation artefacts that compromise AFM imaging of soft interfaces. On test samples (e.g., etched glass with $F_i = -k_i x_i, \qquad i = x, y, z$1 nm-deep depressions), AFM with $F_i = -k_i x_i, \qquad i = x, y, z$2 pN/nm resolves $F_i = -k_i x_i, \qquad i = x, y, z$3 nm features; PFM images with $F_i = -k_i x_i, \qquad i = x, y, z$4m-tall truncated cones, $F_i = -k_i x_i, \qquad i = x, y, z$5 nm tips, and $F_i = -k_i x_i, \qquad i = x, y, z$6 fN/nm resolve the same features (lateral step $F_i = -k_i x_i, \qquad i = x, y, z$7 nm, $F_i = -k_i x_i, \qquad i = x, y, z$8-noise $F_i = -k_i x_i, \qquad i = x, y, z$9 nm, minimum force $k_\mathrm{OT} \approx 0.001$0 pN) (2002.01533). No force-induced artefact or sample damage is observed at these ultra-low probe forces.

On living Plasmodium falciparum-infected erythrocytes, PhFM reveals $k_\mathrm{OT} \approx 0.001$1–$k_\mathrm{OT} \approx 0.001$2 nm-high “knob” substructures without artefacts from indentation, resolving features that are often invisible to AFM without chemical fixation (2002.01533).

6. Lateral Optical Force Mapping and Advanced Modalities

Laterally induced optical forces can be quantitatively measured using torsional eigenmodes of AFM or leveraging PFM in hybrid optical-mechanical platforms. In the dipole approximation, the lateral gradient force is expressed as:

$k_\mathrm{OT} \approx 0.001$3

where $k_\mathrm{OT} \approx 0.001$4 is the real part of the tip’s polarizability and $k_\mathrm{OT} \approx 0.001$5 is the electric field (Huang et al., 2016). The lateral force generates a torque $k_\mathrm{OT} \approx 0.001$6 about the cantilever axis, exciting its first torsional eigenmode. Microfabrication techniques (e.g., focused ion beam milling) optimize torsional stiffness and resonance frequency, enabling sub-piconewton lateral force detection. Raster-scanning under intensity-modulated laser illumination and lock-in amplifier readout provides spatially resolved maps of both vertical and lateral field components (Huang et al., 2016).

Multi-frequency modulation and multi-channel lock-in amplification extend PFM to full-tensorial force mapping and correlative PFM-Raman or PIFM-TERS, establishing the method as a platform for probing vectorial fields, anisotropic molecular responses, and fast time-resolved dynamics (Huang et al., 2016).

7. Limitations and Current Directions

PFM’s principal limitations stem from nonspecific probe–sample binding (the sub-piconewton trap cannot free stuck particles; improved chemistries are required), imaging speed (scan times of $k_\mathrm{OT} \approx 0.001$7 min for $k_\mathrm{OT} \approx 0.001$8 pixels at $k_\mathrm{OT} \approx 0.001$9 ms/pixel, constrained by Brownian corner frequencies), and sensitivity to drift for long averaging times (2002.01533, Pal et al., 2011). Increasing laser power to stiffen traps risks photodamage to sensitive samples. A plausible implication is that future improvements will prioritize high-index, small-diameter nano-probes, adaptive beam shaping, and multiplexed or high-bandwidth detection systems.

The combination of non-destructive probing, high force sensitivity, and customizable probe geometries positions PFM as a unique approach for quantitative nanomechanics, soft-matter imaging, and plasmonic force mapping, bridging the gap where AFM and purely optical methods are limited (2002.01533, Huang et al., 2016, Pal et al., 2011).