Line shape analysis of two-dimensional infrared spectra.

Ultrafast two-dimensional infrared (2D IR) spectroscopy probes femtosecond to picosecond time scale dynamics ranging from solvation to protein motions. The frequency-frequency correlation function (FFCF) is the quantitative measure of the spectral diffusion that reports those dynamics and, within certain approximations, can be extracted directly from 2D IR line shapes. A variety of methods have been developed to extract the FFCF from 2D IR spectra, which, in principle, should give the same FFCF parameters, but the complexity of real experimental systems will affect the results of these analyses differently. Here, we compare five common analysis methods using both simulated and experimental 2D IR spectra to understand the effects of apodization, anharmonicity, phasing errors, and finite signal-to-noise ratios on the results of each of these analyses. Our results show that although all of the methods can, in principle, yield the FFCF under idealized circumstances, under more realistic experimental conditions they behave quite differently, and we find that the centerline slope analysis yields the best compromise between the effects we test and is most robust to the distortions that they cause.