Ultrafast Molecular Dynamics

Time Resolved Photoelectron Spectroscopy (TRPES)

Figure 1

The advent of commercially available, broadly tunable femtosecond laser systems over the last few years has enabled the study of molecular dynamics in real time as an initially prepared state, or set of states, evolves towards a final set of products. This is depicted schematically in figure 1. The typical timescale of such dynamical evolution in molecules occurs on the order of tens to hundreds of femtoseconds (1 femtosecond = 10-15 seconds). Using so-called ‘ultrafast’ laser pulses that are similar in temporal duration to these processes therefore allows us to follow dynamical processes in molecules as they happen.

Using photoionization as a probe to follow this dynamical evolution is an ideal tool since in principle all participating molecular states may be observed directly. This is in contrast to techniques such photoabsorption or photoemission spectroscopy, where certain ‘dark states’ may be unobservable, and the overall details of the dynamics are therefore partially obscured.

In contrast to many groups utilizing photoionization as a dynamical probe, we choose to look at the distribution of photoelectron kinetic energies rather than simply the overall photoion yield. This often provides far more subtle insight into the dynamics, as illustrated in figure 2 for the case of the molecule all trans decatetraene. The photoelectron spectrum is clearly seen to evolve as function of the delay time between the pump laser pulse (which prepares the initial state) and the probe laser pulse (which probes the evolution towards the final states at each timestep,t). This is a consequence of Koopmann’s correlations: The initially prepared S2 state ionizes to the ground electronic state of the cation (termed D0), whereas the S1 state into which it evolves ionizes to the first excited electronic cation state (D1). The distribution of photoelectron kinetic energies resulting from each state are therefore markedly different, as can be seen in the figure. An important point to note is that integration over all photoelectron energies at each time step (equivalent to monitoring the photoion yield) obscures this dynamical information. TRPES is therefore a powerful tool for studying molecular dynamics.

Figure 2

Photostability of DNA Bases

One particular area to which we have applied the TRPES method in recent years is in studies of ultraviolet (UV) photostability in DNA. This is an interesting problem which has attracted a great deal of interest from both experimentalists and theoreticians around the world in the last five years or so. The key questions in this area centre around ascertaining the mechanism or mechanisms by which electronic energy absorbed by DNA (in the form of ultraviolet light) may be rapidly converted into vibrational energy and then subsequently dispersed into the surroundings. This rapid redistribution or ‘quenching’ of the electronic energy into vibration (a so-called non-adiabatic process) is important since the only alternative pathway to dissipate the excess energy is the breaking of a chemical bond within a strand of the DNA molecule. The consequences of such strand breaks and the genetic mutations that can subsequently arise are well documented.

As illustrated in figure 3, DNA consists of two sugar-phosphate strands forming a ‘backbone’ that is held together in the famous double helix structure by a series of base pair interactions. The four individual bases, adenine (A), thymine (T), guanine (G) and cytosine (C) always pair up as either A-T or G-C, and the sequencing of these base pairs forms a binary encoding system that is unique to each individual. Upon ultraviolet irradiation, it is the bases within the DNA molecule which absorb the light (these sites are called chromophores).

Figure 3.

As a first step to understanding the subsequent energy redistribution at a high level of detail, we study the isolated bases in the gas phase in a molecular beam under well-defined collision free conditions using pump-probe, time-resolved photoelectron spectroscopy (TRPES). Figure 4, below, shows a 3D contour plot of the decay dynamics in adenine and 9-methyladenine, recorded using the magnetic bottle photoelectron spectrometer in our lab. Placing different chemical substituients at the 9-position is particularly interesting as this is where the sugar phosphate backbone attaches in DNA. More specifically, the * electronic state – recently implicated in playing a significant role in the relaxation dynamics – is localized at this position and methylation was expected to effectively ‘switch off’ this pathway.

Figure 4.

By globally fitting a bi-exponential decay model to the data in figure 4 and comparing the results obtained for adenine and 9-methlyadenine, we were able to experimentally confirm the involvement of the * state in the relaxation dynamics of adenine for the first time. This is a rapid process, taking place in less than 100 fs and may provide a highly efficient mechanism for 'electronically cooling' DNA back to the group electronic state. Work is now underway in our group to better model the effects of the sugar group in real DNA by placing electron withdrawing functional groups at the 9-position.

Further Reading