Figure 5

Experimental intensity correlations for nano-aperture 2fFCS. Intensity correlations for (a). A lipid bilayer directly on glass, (b). A polarization insensitive square hole array in gold, and (c). The crossed slits. Blue curves with circular symbols correspond to autocorrelations, i.e., correlating detectors in the same polarization channel, while red curves with square symbols correspond to cross-correlation of polarization channels. For the reference system (a) and the polarization insensitive antennas (b), the auto- and crosscorrelates are identical, but a strong difference appears for the crossed nanoslit sample (c). Shaded areas around the curve indicate the standard error in the mean, given that curves are averages over 3 runs of 120 seconds each, where we furthermore average over all auto- resp. crosscorrelating detector combinations. Dotted (dash-dotted) black lines indicate the fitted ACF (CCF) according to a simple Gaussian model (see text). Note that for panels b,c the difference is the polarization cross-talk required to fit the data. In (b) all apertures contribute equally independent of polarization leading to an identical auto- and crosscorrelation. In (c) the x(y)-oriented holes contribute 4 times more strongly to MDF _y(x) than the y(x)-oriented holes. In panel c the purple curve shows the difference in correlations, corresponding to the temporal correlation of instantaneous polarization differences. The black curve superimposed on the purple curve is the difference of the fits to the CCF and ACF, which has a rolloff time of 0.93 ms. Diffusion at the aperture substrate is slowed down compared to the reference case (D = 0.33 versus 4.5 μm²/s).