Principles of Fluorescence Spectroscopy

PRINCIPLES OF FLUORESCENCE SPECTROSCOPY 829
Figure 24.43. Association of prion proteins (PrP) by dual wavelength
cross-correlation FCS. PrP were labeled with either Cy5 or Oregon
Green (OG). Revised from [139].
(MW) 1/3 . The use of FCS to measure the hydrodynamics
and internal dynamics of macromolecules is promising
because the timescale is not limited by fluorescence lifetimes.
However, there have been relatively few publications
on rotational diffusion 148–152 because there are a number of
physical limitations and technical challenges. To measure
rotational motions using FCS it is necessary to account for
photon antibunching and triplet formation, which can occur
on the same timescale as rotational diffusion.
In a typical FCS experiment the timescale of interest is
the diffusion time τ D , which is usually much longer than the
fluorescence lifetime τ F . Hence there is time for the fluorophore
to return to the ground state to be excited again
while still in the observed volume. However, rotational correlation
times θ are usually comparable to the lifetimes and
much shorter than the diffusion times. In order to observe a
correlation it is necessary to detect more than a single photon
from the molecule before its orientation is randomized
by rotational diffusion. While the fluorophore is in the
excited state it cannot be excited again. As a result there is
always some time delay, comparable to the fluorescence
lifetime, between detection of two photons from the same
fluorophore. This delay is called photon antibunching, to
Figure 24.44. Coincidence analysis to detect λ phage DNA. Coincidence
events are shown by the vertical dashed lines. Revised and
reprinted with permission from [143]. Copyright © 1997, American
Chemical Society.
indicate that detection of a second photon is statistically
less probably as the time delay becomes smaller. 153–155
An optical configuration for anisotropy FCS (AFCS) is
shown in Figure 24.45. The sample is excited with polarized
light. The emission is observed using two detectors.
The emission is split by a beamsplitter that randomly transmits
or reflects the photons. The detection electronics is
similar to that used for TCSPC. The difference in arrival
times of the emitted photons is measured with the timedomain
electronics. Because the photons are randomly distributed
the first photon can arrive in either channel, so the
correlation function will appear to be symmetrical around τ
= 0. Figure 24.45 shows polarizers in the emission light
paths, but they are not necessary because photoselection
occurs upon excitation.
The theory for AFCS can be complex, 148–149 so we will
present the simplest case. Assume that the lifetime τ F is
much shorter than the rotational correlation time θ, and that