Single-Molecule Vibrational Thermometry

Abstract

Molecular probes of temperature (termed "molecular thermometers") have become broadly used for in situ temperature measurements. Here, we describe Boltzmann-edge vibrational thermometry (BET) detected by anti-Stokes fluorescence, where the relative population of vibrationally excited molecules acts as a calibration-free reporter of local temperature based on the Boltzmann distribution. We demonstrate that BET microscopy is readily compatible with biological samples and achieves single-molecule sensitivity. We then show that local environments can be characterized through the modulation of vibrational temperature by mid-infrared absorption, allowing for BET fingerprinting. This work provides a foundation for sensitive vibrational thermometry in biological imaging.

Figure 1.

Working principle of Boltzmann-edge thermometry. (a) Normalized electronic absorption and emission spectra of Rh800 (structure inset) in DMSO at room temperature. (b) Log-scale visualization of the Boltzmann edge (low-frequency tail) of the UV-vis absorption spectrum of Rh800. The slope of the dashed line corresponds to a temperature of 291 K. Additional dotted lines are drawn to illustrate shifts in the Boltzmann edge at higher and lower temperatures.

Figure 2.

Anti-Stokes fluorescence BET microscopy. (a) Anti-Stokes fluorescence-based wide-field imaging configuration (cell image obtained from the NIH BioArt Source, BIOART-000198). (b) Representative anti-Stokes fluorescence image of Rh800-stained HeLa cells. (c) Mapping of the Boltzmann edge in anti-Stokes fluorescent intensity. The stained cells have a mean temperature of 285 ± 9 K (95% confidence), close to room temperature.

Figure 3.

Single-molecule vibrational thermometry. (a) Representative anti-Stokes fluorescence image of Rh800 molecules in a polyvinyl alcohol thin film with four spots indicated. (b) Measurement of the Boltzmann edge in anti-Stokes fluorescent intensity from individual single molecules at the indicated spots (errors are 95% confidence intervals). The temperatures from the individual spots agree with the mean temperature across the ensemble (289 ± 19 K, 95% confidence; Figure S3).

Figure 4.

BET fingerprinting. (a) Instrument configuration for measuring anti-Stokes fluorescence under mid-IR excitation (left) and illustration of local heating (right). (b) Modulation of anti-Stokes fluorescence using mid-IR light with 1 kHz modulation. The solution heats exponentially when the mid-IR is on, leading to an increase in anti-Stokes fluorescence, and cools exponentially back to room temperature when the mid-IR is off. (c) BET at varying mid-IR frequencies for 100 μM Rh800 in DMSO-d₆ with 10 kHz modulation. (d) Vibrational temperature as a function of mid-IR frequency for 100 μM Rh800 in DMSO-d₆. The vibrational temperatures align well with the FTIR spectrum of water (converted to IR extinction coefficient; neat, 55.4 M), indicating that hygroscopically absorbed water is the dominant source of vibrational heating in solution.