Spectral Diffusion at Room Temperature

Another effect commonly observed in far-field low-temperature single molecule spectroscopy is spectral shifting due to effects other than macroscopically applied fields. Some shifts, termed spectral diffusion, are spontaneous and others, such as those responsible for hole burning, are photo-induced. At liquid He temperatures, spectral diffusion due to local physical perturbations has been observed to occiu in narrow frequency intervals ( GHz). As the temperature is raised from a few Kelvin to room temperature, the rate of diffusion and the spectral linewidth increase dramatically, which tends to wash out these effects on time scales accessible to measurement. So it was somewhat of a surprise when spectral shifts in emission spectra were observed at room temperature. 

In the first room-temperature, spectrally-resolved, single-molecule experiments, Trautman, Macklin and coworkers observed spectral shifting [17]. The sample consisted of dil (an indocarbocyanine dye) dispersed on polymethylmethacrylate 

Middle Panel Static Stark shift method, traces labeled by the tip voltage, Vf. Bottom Panel Stark shift with transverse dithering method, traces labeled by PV. The scale is exact for the lowest trace in each panel with the other traces shifted vertically upward. Zero detuning = 592.067 nm. Adapted from Ref. 11. 

To obtain fluorescence lifetimes time correlated single photon counting (TCSPC) was used. In TCSPC, the elapsed time is measured between an excitation pulse from a pulsed laser and a detected photon. A histogram of the elapsed times provides a fluorescence decay curve, from which the fluorescence lifetime, Zf, is extracted. Examples of decay curves for bare silica and single R6G molecules on silica taken with a near-field probe are shown in Fig. 11 [21]. 

For R6G on silica, experiments were performed on populations of molecules to obtain the mean, unperturbed lifetime, Zf, and its distribution. Five unperturbed fluorescence decay curves were measured for tips positioned 1.0 to 1.1 pm above high-coverage surfaces ( 10 to 10 molecules illuminated) (Fig. 11(D)). The decay curves were fitted to a single exponential with Zf = 3.65 + 0.04 ns, which is in good agreement with the 3.5+ 0.1 ns obtained previously [22]. Statistical noise and instrument nonlinearity place an upper limit on a possible Gaussian standard devia- 

---