Memory and Spectral Diffusion in Single-Molecule Emission

Abstract

The blinking dynamics of single violamine R (VR) molecules embedded in crystals of potassium acid phthalate (KAP) are analyzed using threshold and change-point detection (CPD) methods (Watkins, L. P.; Yang, H. J. Phys. Chem. B 2005, 109, 617). Analysis employing thresholding resulted in power-law distributions of on and off times corresponding to a power-law exponent of ∼2, consistent with a distributed kinetics model for population and depopulation of the nonemissive state. When the same emission time traces were analyzed using the CPD method, a power-law exponent of ∼1.5 is obtained. In addition, multiple emission states are observed using CPD, inconsistent with a simple two-state blinking model, and indicative of spectral diffusion. The role of spectral diffusion in the distributed blinking kinetics of KAP/VR is investigated by spectrally decomposing the emission using a dichroic mirror. Combining the CPD method with this experiment yielded the emission energy, intensity, and temporal duration of blinking events. A wide distribution of emission energies is observed, consistent with molecules experiencing a variety of dielectric environments within the crystal host. Time-dependent fluctuations in the spectral decomposition are observed, corresponding to spectral diffusion. Blinking events exhibited by single VR molecules in KAP are correlated, an effect referred to as memory. To our knowledge, this is the first reported observation of memory for a molecular system. Positive correlations are observed for consecutive on times and consecutive off times. In addition, adjacent on and off times demonstrate anticorrelation. These observations support the observation of spectral diffusion in this crystal environment, with this diffusion contributing to population and depopulation of the nonemissive state.

Single molecule power-law behavior on a crystalline surface

“…19 More recently, this has also been confirmed for on distributions of molecules embedded in a crystalline environment. 28,29 Note that exclusion of the multilevel trajectories does not affect the shape of the distributions. Defining the lowest intensity as off and all others as on following Reid and co-workers, 28 the same power-law coefficients are recovered ͑m off = 1.17, m on = 1.25͒.…”

“…28,29 Note that exclusion of the multilevel trajectories does not affect the shape of the distributions. Defining the lowest intensity as off and all others as on following Reid and co-workers, 28 the same power-law coefficients are recovered ͑m off = 1.17, m on = 1.25͒. This indicates that the same underlying process is likely at work for both types of trajectories.…”

“…We would like to mention that in contrast to off distributions, 19 the existence of power-law behavior in on distributions from a single molecule is still debated, since such distributions could possibly result from a static distribution of molecular states. Wustholz et al 28,29 report, however, essentially identical power-law coefficients from small ensembles and single molecules for violamine doped into potassium acid phthalate. For the data presented here, the close agreement and low values of both on and off power-law coefficients observed for C 4 -PTCDI on Al 2 O 3 ͑0001͒ suggest a relationship between both distributions and might shed light on the charge transfer process.…”

“…8 Even though it is generally thought that spectral diffusion is difficult to observe at room temperature, several reports have documented discrete spectral jumps of 10-40 nm in the emission spectrum of isolated molecules at room temperature, presumably associated with similar jumps in the absorption spectrum and hence the absorption cross section. 28,[30][31][32][33][34][35] The connection between these spectral jumps, fluorescence intensity, and fluorescence lifetimes is, however, generally not clear. Sauer and co-workers 34 found no correlation between these observables for three different classes of dyes, although this study did not include perylene bisimides.…”

“…If the fluorescence intermittency originated in sudden spectral jumps, we would expect to see a substantial fraction of molecules showing transitions between multiple subsequent intensity levels, as observed in fact by several other groups. 28,34,35 Alternatively, the molecules would have to jump repeatedly either between a single bright and a single dark state or between several states with identical emission properties. Aside from seeming unlikely even on a crystalline surface, such behavior would most likely be diffusive and should give rise to power-law coefficients of 1.5, at variance with our observations.…”