1：Experimental Design and Method Selection:

Single particle spectroscopy was used to eliminate sample heterogeneity, with emission spectra recorded for gold nanorods (AuNRs) of varying aspect ratios excited by continuous wave lasers at 405 nm, 633 nm, and 785 nm wavelengths. The design focused on analyzing Stokes and anti-Stokes emission components to extract temperatures and power dependencies.

2：Sample Selection and Data Sources:

Three chemically synthesized AuNR samples with average aspect ratios of 2.1, 3.5, and 5.3 were used, covering longitudinal surface plasmon resonance wavelengths from 568 to 883 nm. Single particles were characterized using scanning electron microscopy (SEM) for dimensions.

3：1, 5, and 3 were used, covering longitudinal surface plasmon resonance wavelengths from 568 to 883 nm. Single particles were characterized using scanning electron microscopy (SEM) for dimensions. List of Experimental Equipment and Materials:

3. List of Experimental Equipment and Materials: Equipment included continuous wave lasers (405 nm, 633 nm, 785 nm), dark-field scattering setup, notch filters for laser light blocking, and finite difference time domain (FDTD) simulation software for absorption cross-section calculations. Materials were gold nanorods synthesized chemically.

4：Experimental Procedures and Operational Workflow:

Emission spectra were measured for single AuNRs, with excitation polarizations aligned to the longitudinal mode. Dark-field scattering spectra were acquired for each particle. Power-dependent measurements were conducted by varying laser power and recording emission intensities.

5：Data Analysis Methods:

Quantum yields (QY) were calculated using incident laser intensities, simulated absorption cross-sections from FDTD, and collected photon counts. Temperature extraction used Bose-Einstein and Boltzmann statistics on Stokes/anti-Stokes intensity ratios. Power-law exponents were derived from nonlinear fitting of emission intensity vs. power data.