Single-Particle Photothermal Microscopy Using On-Chip Silicon Nitride Microring Resonators

Yulia Podorova, Cecilia H. Vollbrecht, Sam Evans, Hannah Rarick, Arnab Manna, Arka Majumdar, Randall H. GoldsmithThe Journal of Physical Chemistry A 2025 · DOI

TL;DR

This paper presents a novel approach to single-particle photothermal microscopy using on-chip silicon nitride microring resonators (MRRs). By leveraging the high sensitivity of these resonators to temperature changes, the researchers demonstrate the ability to image and characterize individual, non-emissive carbon nanotubes, paving the way for more sensitive analysis of nanoscale materials.

What Problem Does This Paper Address?

Detecting and characterizing individual, non-emissive nanoparticles and molecules is a significant challenge in physical chemistry. Traditional microscopy techniques often struggle with sensitivity, especially for particles that don't fluoresce. Photothermal microscopy offers a solution by detecting the heat generated when a particle absorbs light, but achieving high sensitivity and spatial resolution, particularly for individual nanoscale objects, requires sophisticated optical setups and highly sensitive detection platforms. Previous work has utilized microtoroids and microbubbles as optical microresonators for this purpose, but these non-planar structures present fabrication challenges and can be sensitive to environmental factors. This paper addresses the need for a more robust, chip-scale, and easily fabricated platform for single-particle photothermal microscopy, exploring the potential of planar silicon nitride microring resonators (MRRs).

Methods

The core of this work lies in the integration of silicon nitride microring resonators (MRRs) with a photothermal microscopy setup.

  • On-Chip MRR Fabrication: The researchers fabricated MRRs on a silicon nitride platform. These MRRs are designed with integrated grating couplers (GCs) for efficient coupling of light from optical fibers. The fabrication process involved standard microfabrication techniques like low-pressure chemical vapor deposition (LPCVD) for silicon nitride deposition, electron beam lithography for patterning, and plasma etching.
  • Photothermal Microscopy Setup: A two-beam geometry was employed. A tunable probe laser, locked to the MRR resonance using the Pound-Drever-Hall (PDH) technique, monitors the resonance shift. A pump laser, modulated at 1.1 kHz, is focused onto the MRR surface to excite the target particle. The resulting temperature increase at the MRR surface causes a shift in the probe laser's resonance, which is detected by a lock-in amplifier.
  • Particle Deposition: Multiwall carbon nanotubes (MWCNTs) were used as model non-photoluminescent nanoabsorbers. They were dissolved in N-methyl-2-pyrrolidone and drop-cast onto the MRR wafers, allowing for chip-scale particle deposition.
  • Finite-Element Simulations: COMSOL Multiphysics was used extensively to model:
  • The optical modes within the MRRs.
  • The thermal properties of the MRR and wafer.
  • Heat transfer from the excited particle to the MRR and surrounding environment.
  • The background photothermal signal from the substrate.
  • The impact of different substrate geometries on signal strength.

Datasets

Not discussed in the paper. The experiments involved imaging individual MWCNTs deposited on the fabricated MRRs.

Key Insights and Ideas

  • Planar MRRs as a Viable Platform: The study demonstrates that planar silicon nitride MRRs can serve as effective platforms for single-particle photothermal microscopy, offering advantages in fabrication and integration compared to non-planar microresonators.
  • Sensitivity to Thermal Perturbations: The high Q-factors of the MRRs (around 6 x 10⁴ in this study) make them highly sensitive to the small temperature changes induced by single-particle absorption.
  • Quantifying Nanoparticle Properties: The technique successfully allowed for the imaging of individual MWCNTs and the quantification of their absorption cross-section per atom, which agreed well with literature values.
  • Understanding Thermal Pathways: Finite-element simulations were crucial in dissecting the thermal pathways, revealing how heat dissipates from the particle through the MRR and into the substrate. This understanding is key to optimizing sensitivity.
  • Trade-offs in Planar Geometry: While planar MRRs offer fabrication benefits, the study highlights a significant drawback: the substrate's thermal conductivity can mute the photothermal signal from the particle and amplify background absorption. This is a critical insight for future design considerations.
  • Polarization Dependence: The photothermal signal exhibited polarization dependence, which qualitatively matched the orientation of the MWCNTs, providing an additional layer of information about the particle.

Key Results

  • Successful Imaging of MWCNTs: Individual MWCNTs were clearly imaged using the MRR-based photothermal microscopy.
  • Absorption Cross-Section Agreement: The determined absorption cross-section per atom for MWCNTs was comparable to previously reported values (1.6 ± 0.5 × 10⁻¹⁸ cm²/C atom).
  • Signal-to-Noise and Signal-to-Background Ratios: The signal-to-noise ratio (SNR) was measured at 282 ± 7, and the signal-to-background ratio (SBR) was 3.5 ± 0.1. The lower SBR compared to microtoroids is attributed to the planar geometry and substrate effects.
  • Simulation-Substantiated Thermal Models: Simulations accurately reproduced experimental observations of background photothermal signals on clean MRRs, validating the understanding of thermal contributions from the silicon wafer.
  • Substrate Impact on Signal: Simulations predicted that removing the thermally conductive silicon substrate could enhance the signal by a factor of 2.2, and completely eliminating the substrate could yield a 53-fold increase, underscoring the importance of thermal isolation.

Limitations and Future Work

  • Lower SBR: The primary limitation identified is the lower signal-to-background ratio (SBR) compared to non-planar microresonators like microtoroids. This is attributed to the thermal conductivity of the silicon substrate, which can mute the signal and amplify background absorption.
  • Lower Q-factor: The MRRs used had Q-factors of 6 × 10⁴, which are lower than the 2 × 10⁷ achieved with some microtoroids. This contributes to a lower SNR.
  • Future Work Suggestions:
    • Improving SBR: Fabricating MRRs on all-transparent platforms like quartz or sapphire is suggested to improve the SBR.
  • Increasing Q-factor: Further fabrication efforts, such as polishing or annealing silicon nitride, could raise the Q-factor and enhance overall sensitivity.
  • Exploring Different Materials: The platform could be extended to image other non-emissive particles and molecules.
  • Advanced Simulations: Continued simulation efforts can further refine the understanding of thermal flows and optimize device design.

Summary updated

Abstract

Optical microresonators confine light in time and space, enabling high sensitivity for the detection of small thermal perturbations. We utilized these properties and investigated on-chip integrated silicon nitride microring resonators (MRRs) for single-particle photothermal microscopy. We calibrate the technique using individual nonphotoluminescent carbon nanotubes and determine values for absorption cross-section per atom that agree with literature values. Finite-element simulations are performed to track all relevant thermal gradients. The combination of experiment and simulation allows for discussion of the benefits and drawbacks of the planar MRR geometry for single-particle microscopy.
BibTeX
@article{podorova2025single,
  author = {Yulia Podorova and Cecilia H. Vollbrecht and Sam Evans and Hannah Rarick and Arnab Manna and Arka Majumdar and Randall H. Goldsmith},
  title = {{Single-Particle Photothermal Microscopy Using On-Chip Silicon Nitride Microring Resonators}},
  journal = {The Journal of Physical Chemistry A},
  year = {2025},
  doi = {10.1021/acs.jpca.5c06077},
}
Figure