Time-Dependent Stokes Shift

The measurement of a time-dependent Stokes shift is a technique that has been employed extensively to measure the bulk dynamics of a variety of solvents. In this method, a molecular probe is selected such that, upon excitation, a significant change in the dipole moment in the excited state induces the surrounding solvent to reorganize to compensate for the change in dipole. This process is subject to Onsager’s regression hypothesis (linear response theory) stating that the solvent relaxation relaxation occurs through equilibrium solvent fluctuations. In the experiments, the liquid response is observed by measuring the time-dependent shift of the fluorescence emission maximum (Stokes shift) until the solvent has completely relaxed to accommodate the dipole moment change when the chromophore is excited. The process is illustrated in Figure 1. Immediately after excitation, the fluorescence emission is centered at ν(t = 0). As the solvent reorganizes, the emission center of the fluorescence moves toward longer wavelengths until it ceases to shift at ν(t = ∞) because the solvent has reached its new equilibrium configuration. In ionic liquids, the Stokes shift is relatively slow, extending well well into the nanosecond regime making this an ideal experiment to be measure by time correlated single photon counting.

Figure 1. A schematic illustration of the time-dependent Stokes shift (exaggerated for clarity). As time progresses, the fluorescence emission wavelength shifts to the red (lower frequency).

Membrane sciences have generated a considerable amount of interest in the field of separations, particularly supported liquid membranes (SLMs). Their ability to provide a high degree of selectivity while consuming considerably less solvent than alternative separation methods has made them desirable for applications in industry as they limit the amount of chemical pollutants released into the atmosphere. Room temperature ionic liquids (RTILs) are also very useful solvents in the field of separations, demonstrating a high ability for selectivity in capturing CO2 molecules. Ionic liquids in porous membranes are referred to as supported ionic liquid membranes (SILMs), and have emerged as an active area of study. The low volatility of the RTIL results in an even lower solvent loss than conventional solvents, and the RTIL stabilizes the membrane more effectively. Studies in which RTILs confined within polyethersulfone (PES) membranes indicate that the influence of an interface on the RTIL structure persists for a significant distance into the liquid, ~350nm.

Here time-dependent fluorescence Stokes shift experiments were performed to measure solvation dynamics using a fluorescent probe molecule in a series of 11-alkyl-33-methylimidazolium bis(trifluoromethylsulfonyl)imide (CnmimNTf2, n = 2, 4, 6, 10, ethyl --Emim, butyl -- Bmim, hexyl -- Hmim, decyl -- Dmim) RTILs as bulk ionic liquids and confined within a PES SILM with 350 nm diameter pores. In these experiments, the fluorescence probe is coumarin 153 (C153), seen in figure 2, is used to study the solvation dynamics of the RTIL samples. Because of the long lifetime of the fluorescent probe, 6 ns, it was possible to study the full range of dynamics in the liquids.

Figure 2. Chemical structures of Coumarin 153

Because the fluorescence emission band shape did not change substantially, each spectrum was fit with a Gaussian around the emission peak maximum (±5 nm) to determine the peak center position. The Gaussian fit was used only to determine the peak maximum and provided a good fit of the data over a limited range. The peak centers as a function of time were used to obtain spectral shift correlation functions (SSCFs) seen below, C(t), which are given by

C(t) = [ν(t) − ν(∞)] / [ν(0) − ν(∞)]

where ν(t), ν(0), and ν(∞) represent the center of the emission peak at a given time t, t = 0, and t = ∞, respectively. The SSCFs of C153 in the bulk NTf2-ionic liquids are shown in figure 3A and their PES200 confined counterparts can be found in figure 3B. All SSCF curves can be fit will as multiexponentials.

Figure 3. C153 SSCF decays of the A:bulk RTIL samples B: PES 200 confined samples. The colored curves are the SSCF data and the dashed black curves are multiexponential fits.

As observed in previous experiments, there is a distinct slowing in the solvation time of Emim NTf2 with the slowest solvation time constant increasing by a factor of 4.4. This slowing factor decreases with increasing chain length (Bmim slowing by a factor of 2, Hmim by a factor of 1.15) to the point where the Dmim NTf2 PES 200 SSCF is effectively identical the bulk Dmim NTf2 SSCF. We identify the slowest component of solvation as the total liquid randomization time by rigorous comparison to Optical Kerr Effect measurements meaning the confined RTILs may randomize much more slowly than their bulk counterparts. We believe that membrane-RTIL interface establishes a local structure very different than the typical bulk structure. Once established, this non-bulk structure propagates out from the interface for a long distance producing the observed large change in the solvation dynamics. Long chained RTILs such as Dmim NTf2 possess the capability to segregate into polar/apolar regimes. These regimes act as a buffer, impeding the propagation of the interfacial structure of the RTIL. Emim NTf2 is not capable of forming such networks and as such, the entire liquid is reorganized. The future studies we hope to better characterize the surface structures of RTILs and the effects these interfaces may have on the liquid as a whole.