Dynamics of Hydrogen Bonded Systems
Population Relaxation in Methanol Solutions
Figure 1: Pump-probe signal for a/b peak in 5 mol % methanol-d dissolved in CCl4. (A) The initial relaxation results from vibrational excited state decay with a ~2 ps time constant. (B) The breaking of hydrogen bonds increases the population of a/b methanol-d and thus the absorption of the laser, which produces a negative signal. This negative signal decays towards zero signal with a ~20 ps time constant. This represents the rate for hydrogen bond reformation. Ultrafast infrared vibrational transient absorption experiments have been conducted on the a/b band and on the d band of methanol-d in CCl4. The experiments allow us to monitor different methanol populations as a function of time depending on which molecules we probe. In the a/b experiments, we have demonstrated that hydrogen bond breaking occurs (~2-3ps) even when exciting those molecules at the end of an oligomer that do not donate a hydrogen bond (b). This represents a sequential hydrogen bond breaking mechanism since the breaking occurs as a secondary relaxation event. This type of behavior was also demonstrated to occur in ethanol-d and propanol-d solutions. One of the only differences in the dynamics of the three different solutions was that the percentage of hydrogen bonds that went to break quickly decreased with increasing molecular size. This was attributed to the fact that propanol-d has many more modes to relax into relative to methanol-d and thus provides a more effective energy sink for the excited OD stretch to relax into.
Concentration dependent experiments have been conducted on the d band in order to distinguish between intermolecular energy transfer and intramolecular energy relaxation events as they both contribute to the signal at higher concentrations. In an isotopically pure sample, hydrogen bond breaking was discovered to occur on two timescales (~200fs, ~2ps).
The faster component was attributed to breaking of the hydrogen bond directly associated with the OD that was excited, while the slower component reflects energy relaxation within the oligomer leading to dissociation elsewhere. The hydrogen bonds reformed on multiple time scales depending on whether the hydrogen bond that broke was local or not. In an isotopically mixed sample, the dynamics that were observed were simpler because we narrowed the scope of what we were able to probe by decreasing the concentration of OD absorbers while maintaining the oligomer size. Only the fast breaking and hydrogen bond reformation rates were observed to occur since we cannot see OH hydrogen bond dynamics with our probe. The isotopically mixed MeOD/MeOH sample allowed us to demonstrate that the fast initial decay of our signal was due to the lifetime (~500fs) since excitation transfer does not occur.
Figure 2: Pump-probe signal for d peak in 5 mol % methanol-d dissolved in CCl4. (A) The initial relaxation results from vibrational excited state decay with a ~500 fs time constant. (B)/(B’) The breaking of hydrogen bonds decreases the population of d methanol-d and thus the absorption of the laser, which produces a positive signal. This positive signal rises with two characteristic time constants of ~250 fs and ~2.5 ps.
