Ionic liquid dynamics in nanoporous carbon: A pore-size- and temperature-dependent neutron spectroscopy study on supercapacitor materials

The influence of spatial confinement on the thermally excited stochastic cation dynamics of the room-temperature ionic liquid 1-N-butylpyridinium bis-[(trifluoromethyl)sulfonyl]imide ([BuPy][${\mathrm{Tf}}_2\mathrm{N}$]) inside porous carbide-derived carbons with various pore sizes in the sub- to a few nanometer range is investigated by quasielastic neutron spectroscopy. Using the potential of fixed window scans, i.e., scanning a sample parameter, while observing solely one specific energy transfer value, an overview of the dynamic landscape within a wide temperature range is obtained. It is shown that already these data provide a quite comprehensive understanding of the confinement-induced alteration of the molecular mobility in comparison to the bulk. A complementary, more detailed analysis of full energy transfer spectra at selected temperatures reveals two translational diffusive processes on different time scales. Both are considerably slower than in the bulk liquid and show a decrease of the respective self-diffusion coefficients with decreasing nanopore size. Different thermal activation energies for molecular self-diffusion in nanoporous carbons with similar pore size indicate the importance of pore morphology on the molecular mobility, beyond the pure degree of confinement. In spite of the dynamic slowing down we can show that the temperature range of the liquid state upon nanoconfinement is remarkably extended to much lower temperatures, which is beneficial for potential technical applications of such systems.

Figure 1

Pore size distribution of the carbide-derived nanoporous carbon samples, as obtained using a density functional theory based analysis of the nitrogen sorption isotherms (lines between the points are guides for the eye.).

Figure 2

Structure formula of the cation (left) and the anion (right) of the ionic liquid [BuPy][${\mathrm{Tf}}_2\mathrm{N}$] with estimate of the ions’ size.

Figure 3

Intensity of neutrons with an energy transfer of $\pm 2$ $\mu \mathrm{eV}$, when scattered at (left) bulk [BuPy][${\mathrm{Tf}}_2\mathrm{N}$] and (right) in the nanoconfinement of MoC-15 as a function of wave vector transfer $Q$, while scanning the temperature from 2 K to 355 K with a rate of 1 K/min. Back panel: Intensity averaged over the available $Q$ range.

Figure 4

Fixed window scans in a temperature range from 2 K to 355 K with a heating rate of 1 K/min for the ionic liquid in bulk and in nanoconfinement of different pore sizes (intensity averaged over a $Q$ range from 0.44 Å${}^{-1}$ to 1.90 Å${}^{-1}$): (a) Elastic fixed window scans, i.e., $\mathrm{\Delta }E=0\mu \mathrm{eV}$; (b) inelastic fixed window scans at an energy transfer of $\mathrm{\Delta }E=\pm 2\mu \mathrm{eV}$.

Figure 5

Fit to the elastic fixed window scan data of the low-temperature dynamics of the bulk ionic liquid with the model as delineated in the text.

Figure 6

Inelastic fixed window scan at sample MoC-15$+$IL at four exemplary wave vector transfers $Q$ with the corresponding results of the fit to the data, which is performed for all 16 $Q$ simultaneously with a common parameter set and with the model as described in the text.

Figure 7

(a) Self-diffusion coefficients $D$ and (b) residence times ${\tau }_0$ of the [BuPy] cation, as a function of temperature for different nanopore sizes, determined using the Singwi-Sjölander jump-diffusion model. Two translational diffusive processes on different time scales are observed. Squares denote data from the FOCUS time-of-flight spectrometer and circles such from the IN16B backscattering spectrometer, at which open circles show values derived from the inelastic fixed window scans. Solid lines are fits with an Arrhenius temperature dependence [see, e.g., Eq. (8)]. Dashed/dotted lines are guides for the eye. Bulk data are taken from Embs et al. [37].

Figure 8

Selected QENS spectra for the ionic-liquid-filled nanoporous carbon MoC-15 as obtained from (a) IN16B ($Q=1.18\text{Å}{}^{-1}$) and (b) FOCUS ($Q=1.26\text{Å}{}^{-1}$). Solid lines denote the best fit to the data with the model as described in the text.

Figure 9

Fraction $f$ of elastic scattering in the quasielastic spectra, resulting from immobilized cations of the ionic liquid. Squares denote the fast diffusive process as seen with FOCUS and circles the slower one, observed with IN16B.