Are rare, long waiting times between rearrangement events responsible for the slowdown of the dynamics at the glass transition?

The dramatic slowdown of the structural relaxation at the glass transition is one of the most puzzling features of glass dynamics. Single molecule orientational correlation times show this strong Vogel-Fulcher-Tammann temperature dependence typical for glasses. Through statistical analysis of single molecule trajectories, we can identify individual glass rearrangement events in the vicinity of a probe molecule in the glass former poly(vinyl acetate) from 8 K below to 6 K above the glass transition temperature. We find that changes in the distribution of waiting times between individual glass rearrangement events are much less dramatic with temperature, the main difference being a small, but decisive number of increasingly long waiting times at lower temperatures. We notice similar individual, local relaxation events in molecular dynamics trajectories for a variety of glassy systems further from the glass transition, leading to waiting time distributions with similar features as those observed in the single molecule experiments. We show that these rare long waiting times are responsible for the dramatic increase in correlation time upon cooling.

Identification of intensity ratio break points from photon arrival trajectories in ratiometric single molecule spectroscopy.

We describe a statistical method to analyze dual-channel photon arrival trajectories from single molecule spectroscopy model-free to identify break points in the intensity ratio. Photons are binned with a short bin size to calculate the logarithm of the intensity ratio for each bin. Stochastic photon counting noise leads to a near-normal distribution of this logarithm and the standard student t-test is used to find statistically significant changes in this quantity. In stochastic simulations we determine the significance threshold for the t-test's p-value at a given level of confidence. We test the method's sensitivity and accuracy indicating that the analysis reliably locates break points with significant changes in the intensity ratio with little or no error in realistic trajectories with large numbers of small change points, while still identifying a large fraction of the frequent break points with small intensity changes. Based on these results we present an approach to estimate confidence intervals for the identified break point locations and recommend a bin size to choose for the analysis. The method proves powerful and reliable in the analysis of simulated and actual data of single molecule reorientation in a glassy matrix.

Single molecules reveal the dynamics of heterogeneities in a polymer at the glass transition.

The notion of heterogeneous dynamics in glasses, that is, the spatial and temporal variations of structural relaxation rates, explains many of the puzzling features of glass dynamics. The nature and the dynamics of these heterogeneities, however, have been very controversial. Single rhodamine B molecules in poly(vinyl acetate) at the glass transition reorient through sudden jumps. With a statistical search for the most likely break points in the logarithm of the ratio of the two perpendicular fluorescence polarizations, we determine the times of these angular jumps. We interpret these jumps as an indication for individual glass rearrangements in the vicinity of the probe molecule. Time-series analysis of the resulting sequence of waiting times between jumps shows that dynamic heterogeneities in the matrix exist, but are short lived. From the correlation of the logarithm of the waiting time between subsequent jumps, we determine an upper limit for the lifetime of heterogeneities in the sample. The correlation time of τ(het) = 32 s is three times shorter than the orientational correlation time of the probe molecule, τ(orient) = 90 s, in the sample at this temperature, but 13 times longer than the structural relaxation time, τ(α) = 2.5 s, estimated for this sample from dielectric experiments. We present a model for glass dynamics in which each rearrangement in one region causes a random change in the barrier height for subsequent rearrangements in a neighboring region. This model, which equates the dynamics of the heterogeneities with the dynamics of the glass itself and thus implies a factor of one between heterogeneity lifetime and structural relaxation time, successfully reproduces the statistics of the experimentally observed waiting time sequences.

Heterogeneous dynamics and dynamic heterogeneities at the glass transition probed with single molecule spectroscopy.

We studied the temperature dependence of the structural relaxation in poly(vinyl acetate) near the glass transition temperature with single molecule spectroscopy from Tg-1 K to Tg+12 K. The temperature dependence of the observed relaxation times matches results from bulk experiments; the observed relaxation times are, however, 80-fold slower than those from bulk experiments at the same temperature. We attribute this factor to the size of the probe molecule. The individual relaxation times of the single molecule environments are distributed normally on a logarithmic time scale, confirming that the dynamics in poly(vinyl acetate) is heterogeneous. The width of the distribution of individual relaxation times is essentially independent of temperature. The observed full width at half maximum (FWHM) on a logarithmic time axis is approximately 0.7, corresponding to a factor of about 5-fold, significantly narrower than the dielectric spectrum of the same material with a FWHM of about 2.0 on a logarithmic time axis, corresponding to a factor of about 100-fold. We explain this narrow width as the effect of temporal averaging of single molecule fluorescence signals over numerous environments due to a limited lifetime of the probed heterogeneities, indicating that heterogeneities are dynamic. We determine a loose upper limit for the ratio of the structural relaxation time to the lifetime of the heterogeneities (the rate memory parameter) of Q<80 for the range of investigated temperatures.