Fig. 1: ACDAN hyperspectral imaging enables spatial mapping of dielectric permittivity with pixel resolution.

a Molecular structure of the ACDAN (6-acetyl-2-dimethylaminonaphthalene) fluorescent probe, a water-soluble molecule that partitions into biomolecular condensates. b Perrin–Jablonsky diagram illustrating the dependence of ACDAN fluorescence on solvent permittivity. The sketch depicts changes in ACDAN dipole moment due to photon absorption or emission and the solvent relaxation process. \(h{\nu }_{a}\) and \(h{\nu }_{e}\) represent the absorption and emission energies, respectively. c Schematic representation of confocal hyperspectral imaging and analysis. A full emission spectrum is measured at each pixel from a stack of images acquired at different emission wavelengths (\(\lambda\)). The pixel permittivity (for either condensed or depleted phases) can be determined using two approaches: Gaussian fitting (bottom left) and phasor representation (bottom right), both detailed in the text. d Fluorescence emission profiles of ACDAN from hyperspectral microscopy imaging of calibration solutions of mineral oil or PEG400-H₂O solutions (from left to right: 100%, 90%, 80%, 70%, 60%, 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, 8%, 6%, 4%, 2%, 1%, 0.5%, 0% PEG content). e, f Calibration data using the maximum emission wavelength of ACDAN obtained from a Gaussian fit of the spectrum (e) or the phase in the phasor plot (f) as a function of previously reported or theoretically calculated solvent dielectric constant (\(\varepsilon\)): Orange circles represent different PEG-400 solutions in water, with permittivity values from literature (referenced in Table S1). Blue circles represent ethanol-water mixtures, with dielectric constants calculated using the Maxwell–Garnett law of mixtures (see SI, Table S3 and Fig. S4b). The error bars indicate the spectral bandwidth used for detection, and datapoints represent the maximum emission wavelength averaged over all 512 \(\times\)512 pixels in the image stack (n = 262144). The relation \(\frac{\phi \left({\lambda }_{f}-{\lambda }_{0}\right)}{2\pi }+{\lambda }_{0}={\lambda }_{\max }\) with \({\lambda }_{0}=416{{{\rm{nm}}}}\) and \({\lambda }_{f}=728{{{\rm{nm}}}}\) was used to establish the link between the emission wavelength \({\lambda }_{\max }\), and the phase \(\phi\) for which the Lippert–Mataga equation (Eq. (1), solid curve; fitting parameters given in Table S2) accurately describes the data.