Fluorescence correlation spectroscopy

Fluorescence correlation spectroscopy (FCS) works by statistically analyzing how the intensity of fluorescent light fluctuates over time, using a technique called time correlation. The idea is rooted in L. Onsager's regression hypothesis, and it reveals kinetic details about the physical processes causing those fluctuations. A key use is studying how the concentration of fluorescent molecules in a solution varies. Here, a very small volume—containing just a few molecules—is observed. Because these molecules move randomly due to Brownian motion, the number inside that tiny space constantly changes around an average, causing the fluorescence signal to flicker. By analyzing these flickers, FCS can calculate the average number of molecules and how long they typically take to diffuse through the observed space, which in turn gives the concentration and size of the particles. These measurements are valuable in biochemistry, biophysics, and chemistry.

FCS is highly sensitive because it works with very few molecules (at nanomolar to picomolar concentrations) inside a volume of about one cubic micrometer. Unlike methods like HPLC, it doesn't physically separate substances; instead, its optics provide the spatial resolution. FCS can also track fluorescently tagged molecules inside living cells, enabling what's called "in situ or in vivo biochemistry"—studying biochemical pathways directly in intact cells and organs.

The technique is often used with optical microscopy, especially confocal or two-photon excitation microscopy. In these setups, light is focused on a sample, and fluctuations in fluorescence intensity—caused by diffusion, chemical reactions, aggregation, or other processes—are analyzed using a temporal autocorrelation function. The method works best when individual molecules are entering or leaving the observation volume (or turning on and off within it). If too many molecules are present, the fluctuations are too small relative to the total signal to detect; if too few, events are too rare and measurements take too long. FCS is essentially the fluorescent version of dynamic light scattering, which uses coherent light scattering instead of incoherent fluorescence.

When a suitable model is available, FCS can provide quantitative data like diffusion coefficients, hydrodynamic radii, average concentrations, rates of chemical reactions, and singlet-triplet dynamics. Because fluorescent markers come in many colors and can be attached to specific molecules (such as proteins, polymers, or metal complexes), it's possible to study individual molecules in complex mixtures. The development of sensitive detectors, like avalanche photodiodes, made it practical to detect fluorescence from single molecules in very dilute samples, expanding FCS applications from materials science to biology. The creation of cells with genetically tagged proteins, such as green fluorescent protein, has made FCS a common tool for studying molecular dynamics in living cells.

The first experimental application of signal-correlation techniques to fluorescence was in 1972 by Magde, Elson, and Webb, who are credited as the inventors of FCS. They and others soon published papers establishing the theoretical foundations and applications. Around 1990, as it became possible to detect very few fluorescent particles, two issues emerged: a non-Gaussian distribution of fluorescence intensity and the three-dimensional confocal measurement volume in laser microscopy. This led to analyzing distributions and moments of the signals, eventually forming a set of methods called Brightness Analyses (see Thompson, 1991, for a review). Starting in 1993, improvements like confocal and two-photon microscopy better defined the measurement volume and reduced background, greatly improving signal-to-noise and enabling single-molecule sensitivity. This renewed interest in FCS; by August 2007, over 3,000 papers using the technique had been published (see Krichevsky and Bonnet for a review). FCS has since been extended to laser scanning and spinning-disk confocal microscopy, cross-correlation between two fluorescent channels (FCCS), and the use of Förster Resonance Energy Transfer (FRET) instead of direct fluorescence.

A typical FCS setup uses a laser (continuous wave from 405 to 633 nm, or pulsed from 690 to 1100 nm) reflected into a microscope objective by a dichroic mirror. The beam is focused into a sample with fluorescent particles diluted so that only a few (usually 1 to 100) are in the focal spot (about one femtoliter). As particles cross into this volume, they fluoresce. The same objective collects the emitted light.