An article from February 2015 bemoans the fact that “...while countless breakthroughs have been announced over the last decade, time and again these advances have failed to translate into commercial batteries with anything like the promised improvements in cost and energy storage.” This is true not just of batteries, but of all devices for the storage of electric charge by chemical means. This includes electrochemical capacitors.
The biggest obstacles to their commercial development are power restrictions and safety issues. This is made worse by the trend of miniature device development, because of the requirement for more techniques which will permit the analysis of electrochemical cell dynamic processes in situ.
This is why new materials for both electrodes and electrolytes are so important in achieving better device performance. With respect to the electrolyte, ionic liquids, comprising of pure salts that melt below 100 oC, have attracted a great deal of attention as potential electrochemical storage media. They have many appealing characteristics, such as high electrochemical windows, good stability to heat, excellent conductivities and low vapor pressure.
One salient field of research is at the place where the ionic liquid electrolytes interface with the electrode, because it is here that electrochemical capacitors (ECs) and similar devices undergo dynamic interactions which affect their performance. In particular, the electrical double layer or EDU is in view here. This is made up of powerful interactions between the ions near the surface of the electrode and the electrode surface itself. They are not the same as those occurring in the solution. Thus the EDL determines how any electrochemical storage device will work. It has been well analyzed in conventional electrolytes, but this is not the same of the ionic liquid electrolytes.
Ionic liquids are unique in that they comprise of alternate anion/cation layers at the interface of solid and liquid. This key difference results in a distinct EDL structural arrangement just below the surface, because ion ordering lessens as the layers increase in depth. The use of molecular dynamic (MD) simulations helps visualize how the ions will undergo arrangement in relation to changes in the electrode material and the surface charge, and thus to understand the meaning of the experimental results.
Atomic force microscopy (AFM) is often selected to examine the structure of ionic liquids as it offers increased spatial resolutions compared to other techniques, and is also appropriate for a more adaptable experimental setting. This is important as ionic liquids are well-suited for novel energy storage devices.
Ball and stick molecular models of two commonly used room-temperature ionic liquids. (Top) propylammonium nitrate (abbreviated PAN), and (Bottom) 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (abbreviated EMIm TFSI here, also known as [EMIm+][Tf2N-]). In all models, carbon atoms are gray, hydrogen atoms are white, nitrogen atoms are blue, sulfur atoms are yellow, oxygen atoms are red, and fluorine atoms are green.
High-Resolution Imaging of EDL in Ionic Liquids with AFM
In many areas of research the solid-liquid interface needs to be analyzed, and this is even truer of charge storage devices. In these systems, the performance depends upon the electrolyte-electrode interface. This layer is also called the EDL or electrical double layer and is very interesting because here, the ions and the electrode surface which is also charged, interact powerfully. With each charge and discharge, the EDL ions shift around to allow for the surface charge. The issue is the lack of understanding of the EDL of ionic liquid electrolytes, unlike that of the traditional solvents.
Conventionally, ionic liquid layers have been analyzed for their lateral structure using scanning tunneling microscopy, or STM. This is harder in liquids because acquiring images with good resolution is difficult with the occurrence of interference with the STM signal by the strong nanoscale structure at the interface.
To surmount this problem, freezing of the ionic liquid has been performed so that it can be imaged under ultrahigh vacuum (UHV). This creates new issues, however, because the real ionic liquid structure under real-time operating conditions is not simulated by the nanostructure of the interface under UHV or frozen conditions.
One way to examine these ionic liquids in their operating environment under conditions which are close to real time is AFM. Lately, amplitude modulation AFM (AM-AFM) or tapping mode AFM has been employed to analyze the interface structure formed between bulk propylammonium nitrates (PAN) and extremely ordered pyrolytic graphite (HOPG) at ambient temperature. Here, the cantilever of the AFM undergoes oscillation near the resonant frequency with a subnanometric free amplitude.
The amplitude set-point ratio for the imaging is set at over 0.7 so that the tip-sample force is minimal, avoiding disruption of the EDL structure as far as possible. Under these conditions the AFM probes the interfacial ions that have been adsorbed on the electrode surface at a resolution below 1 nm, as Figure 1 shows, but without touching the electrode. To verify these conditions the set-point ratio can be reduced, while pushing up the tip-sample force, to the point where the tip probes beyond the ionic layer to contact the surface of the electrode.
This experiment was the first of its kind, in that it determined the structures within a viscous liquid at molecular level with AM-AFM. Again, it was unique in being able to image the interfacial (graphite) electrode-ionic liquid layer at molecular resolution. This study used AFM to achieve a comparable resolution to that obtained using STM to examine frozen ionic liquid monolayers.
Another study done at 25 oC used imaging of the interface ion layer at high resolution. The interface being between 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMIm TFSI) and a HOPG surface. The bias was applied to the electrode at varying levels, and 0.1 wt/wt% Li TFSI or EMIm Cl was added to the ionic liquid as well. Using AFM the effects of these changes on the arrangement of ions at the interface of the electrolyte and electrode could be visualized.
Images at molecular resolution showed alternating layers of anions and cations in a unit cell arranged as sharp rows on the surface, with an open-circuit potential, and with similar ions in the proximity, as Figure 2 shows. A vital point is that the ion layering seen at nanoscale changed with the potential applied to the surface, or when there were low concentrations of chloride or lithium ions. The study provided proof that the topmost ion layer which was in contact with the electrode surface was made up of an additional layer of cations, which due to being the outermost may mask the influence produced by the surface potential on the EDL structure.
Thus in situ, AFM showed the way to differentiating cations and anions on the electrode surface for the first time, and also allowed the determination of their precise locations inside the layer of ionic liquids. Expectedly, the ion layer at ambient temperature had quite a different structure from that produced by frozen monolayers as shown by UHV STM. This leads to the conclusion that EDL definition for ionic liquids may be critically dependent on the use of AFM, which could quite possibly pave the way for their successful use for electrochemical devices of the future.
Figure 1. Height image of a PAN–HOPG interface obtained using tapping mode with a scan amplitude of 1 nm. Twodimensional Fourier analysis confirmed that PAN adsorbs on the HOPG surface with a rhomboidal structure with a lattice spacing of 0.48±0.02 nm, approximately twice that of the underlying HOPG. Adapted with permission from Ref. 10.
Examining Ion Layers with Force Curves
Electrochemical capacitors (ECs) are better than batteries when it comes to storing energy, and are thus hot areas of research. For one, their power densities are higher, at 10 kW/kg, than that of batteries, and they also boast a shorter charge-discharge time. An issue is their lower energy density, which is one important focus of current study. One way to improve this is to boost the operating voltage of the EC with ionic liquid electrolytes so that the stored energy goes up.
This is a better approach than aqueous electrolyte use, which could undergo electrolysis at excessive voltages, leading to the formation of hydrogen gas which is hazardous. As is known, ionic liquid electrolytes do not break down so easily. Thus higher voltages can be achieved and thus greater energy densities.
Another factor determining EC capacitance is the electrical double layer, and this is vital in developing future energy storage devices based on ECs. Thus EC analysis is warranted in bringing out explanations of how electrochemical systems work and the potential uses of ionic liquids for electrochemical storage. For instance, ion layer defects may be uncovered better, seeing that it is already known that the ion layer is not quite even. This needs to be studied to find out whether such defects enhance or degrade EC performance.
Figure 2. Tapping mode phase images of the IL Stern layer adsorbed to a graphite (HOPG) substrate for both, (Left) the pure IL (EMIm TFSI) and (Right) for an IL solution with added chloride anions (EMIm TSFI + 0.1 wt/wt % EMIm Cl). These images were taken at open circuit potential (OCP) (i.e. with no applied bias), corresponding to 0.26 V vs. Pt for the pure IL and 0.42 V vs. Pt for the solution with added chloride anions. For the pure IL (Top), the phase image reveals alternating pattern of two rows of stiffer regions (lighter areas- example rows highlighted in red) followed by two rows of more compliant regions (darker areas- example rows highlighted in gray). Addition of chloride ions (Bottom) changes this structure, resulting in a pattern of two stiffer rows, one compliant row, one stiff row, then one compliant row. Additional work at applied potentials above and below the OCP (not shown here) ascertained that the stiffer rows must correspond to the IL cations and the more compliant rows correspond to the IL anions. Adapted with permission from Ref. 12.
Force-distance (F-D) curves are useful in examining the EDL structure of an ionic liquid in relation to the distance to the surface of the electrode. These were produced using HOPG as a model platform for the use of EC based on carbon. The evolution of bias in the layers of ions was analyzed using F-D curves obtained with AFM, as Figure 3 shows. The reproducibility of the F-D curves was an area of particular care, since this is often ignored in similar experiments.
Using a mix of molecular dynamic (MD) simulations and AFM experiments, the ion layers were modeled by the application of a potential. The results show very good correlation between the two sets of study methods.
In particular AFM was used to do F-D spectroscopies on a sample droplet of EMIm TFSI ionic liquid on HOPG. The measurements of the F-D curves was carried out between 0.1 and 0.3 Hz. Uncoated cantilevers of silicon nitride were used which had a spring constant of 0.38 N/m. Comparison curves were set in alignment to the y-axis so that they could be at zero force at a considerable distance from the sample.
Using the F-D curves, the positions of the ion layer inside the EDL of EMIm TFSI were found. The first layer was at 0.37 nm starting from the surface of the HOPG, while succeeding layers were separated by 0.7 nm. 2D histograms charted with 50 F-D curves confirmed high reproducibility. By comparing these figures with MD simulations, a close familiarity with the ionic layer EDL was gained on both biased and unbiased graphite surfaces. This step had to be done to enable the F-D curves to be properly understood and to make an identification of the ion layers.
Combining MD simulation with AFM F-D spectroscopy experiments provided an understanding of whether each of the ion layers was anion or cation. The MD simulations showed indications that those layers of ions which were nearest to the surface of the electrode were composed of anions and cations oriented in two different ways.
The ions in the topmost layer were found to have a potential difference of +1V and -1V for cations and anions, compared to the zero charge potential. This was thought to be because a Coulombic force is exerted between the electrode and the ion. In succeeding layers, the ions show an anion-cation pairing with characteristic 0.7 nm spacing.
What is important about the study conclusions is the suggestion that it may make future ion layer analysis in electrochemical-based systems easier for ECs. It indicates that AFM is useful in studying every side of the EDL, which provides a big step forward in getting to know how charge storage operates in energy storage devices.
Figure 3. Details of single force−distance curves. (a) Single approach and retract curve showing multiple ion layers. (b) Single force separation plot of curve shown in (a), and 2D histogram of 50 consecutively measured force curves (c). Adapted with permission from Ref. 17.
How to Enhance AFM Cantilever Effects by Photothermal Excitation
The examples show that AFM is an extremely useful tool to achieve imaging of the EDL structure of the ionic liquid at high resolution. One of the issues with instrumentation in these studies is the increased viscosity of these liquids in comparison to water, which is more commonly used in AFM. The problem with viscous liquids, is the increased hydrodynamic drag which means more force must be applied to the cantilever to achieve oscillation in the tapping mode.
In most cases the oscillation is driven by a piezoelectric actuator situated near the cantilever, and this is termed piezoacoustic excitation. In the case of ionic liquids, the requirement for high drive amplitude leads to excitation in other mechanical parts in the AFM, as well as in the cantilever. The result is a contaminated cantilever frequency response because of external resonances, which makes both the operation and the interpretation of the experiment more difficult.
Thus these outside resonance peaks must be eliminated from the cantilever response, which must also be preserved over time. One way to do this is by adopting photothermal excitation which provides direct cantilever drive. This is based on a secondary laser focused on a point near the cantilever base. The laser power is modulated to the cantilever’s resonance frequency, and thus produces oscillation due to the resulting small amount of thermal expansion and contraction. This leaves other mechanical resonances intact, leaving the cantilever response pristine and stable. Results show that photothermal excitation yields better signals when AM-AFM is used to image in ionic liquids.
This potential use of photothermal excitation to examine ion layering with AFM was shown experimentally using EMIm TFSI on mica and HOPG surfaces. To accomplish this, dynamic F-D and small amplitude F-D (SAFD) spectroscopies were both used with AFM in the above setup, so that the effect of the surface or substrate on the components of resistance to squeeze.
With F-D spectroscopy, the layering of EMIm TFSI on mica and HOPG were measured and the results compared with each other, as Figure 4 shows. Mica produced nanostructure layers of greater strength and number, probably due to the higher charge on its surface. The F-D data sets also showed good reproducibility when photo-excitation was used from the 50 data sets used in the analysis.
Figure 4. (from top to bottom) Amplitude, phase, kts, and bts for the cantilever response in EMIm TFSI on mica (left) and HOPG (right). The charge-mediated layers on mica are stronger than the templated layers on HOPG. This leads to an increase in resistance to local flow on mica as the preferentially oriented molecules interlock with each other. On HOPG, the opposite effect is observed, wherseby the ionic liquid has a slight reduction in flow resistance in the presence of the wall. Adapted with permission from Ref. 20.
The study conclusions showed two significant divergences from a previous experiment by the same authors, based on piezoacoustic excitation. One was the lack of resolution of the ionic liquid structure at an interface with HOPG because of the presence of contamination of cantilever resonance by extraneous resonances. A less obvious difference is that the cantilever base must be moved through a large amplitude to cause the tip to move with even a small amplitude, when piezoacoustic excitation is used in viscous ionic liquids. In this situation, the tip motion can be interpreted only by subtracting the movement of the cantilever base, a quantity which is already difficult to measure with precision. However, using photothermal excitation avoids cantilever base movement, leading to simplification of the analysis with increased accuracy.
Ionic liquids promise better electrochemical storage device performance, but before this can be realized, the EDL and the electrode-ionic liquid electrolyte interactions must be fully characterized. AFM is much more useful in this field than other techniques due to its unique capabilities, along with higher resolution and other data which helps clarify the structure of the EDL and the related interfacial dynamics taking place within ionic liquids.
The Cypher AFM from Asylum Research can be used to yield unparalleled imaging resolution as well as stable results when examining images and spectroscopic results in ionic liquids. It can be used in combination with blueDrive photothermal excitation to produce cantilever responses of greater accuracy, making it the best choice for analyzing EDL structure in ionic liquids.
- K. Bullis, “Why We Don’t Have Battery Breakthroughs”, MIT Technology Review, (2015).
- J.M. Tarascon, M. Armand, Nature 414, 359-367 (2001).
- M. Armand, F. Endres, D.R. MacFarlane, H. Ohno, B. Scrosati Nature Mater.8, 621-629 (2009).
- D. Larcher, J. Tarascon, Nature Chem. 7, 19-29 (2015).
- M.P. Mousavi, A.J. Dittmer, B.E. Wilson, J. Hu, A. Stein, P.J. Bühlmann, Electrochem. Soc.162, A2250-A2258 (2015).
- F. Endres, O. Höfft, N. Borisenko, L.H. Gasparotto, A. Prowald, R.Al-Salman, T. Carstens, R. Atkin, A. Bund, S.Z. El Abedin, Phys. Chem.Chem. Phys.12, 1724-1732 (2010).
- T. Welton, Chem. Rev., 99, 2071-2084 (1999).
- M.V. Fedorov, A.A. Kornyshev, Chem. Rev.114, 2978-3036 (2014).
- M. J. Earle, K. R. Seddon, Pure Appl. Chem., 72, 1391-1398 (2000).
- A. J. Page, A. Elbourne, R. Stefanovic, M.A. Addicoat, G.G. Warr, K. Voïtchovsky, R. Atkin, Nanoscale 6, 8100-8106 (2014).
- B. Uhl, T. Cremer, M. Roos, F. Maier, H.-P. Steinrück, R.J. Behm, Phys. Chem. Chem. Phys. 15, 17295-17302 (2013).
- A. Elbourne, S. McDonald, K. Voïchovsky, F. Endres, G.G. Warr, R. Atkin, ACS Nano 9, 7608-7620 (2015).
- F. Buchner, K. Forster-Tonigold, B. Uhl, D. Alwast, N. Wagner, H. Farkhondeh, A. Groß, R.J.R. Behm, ACS Nano 7, 7773-7784 (2013).
- P. Simon, Y. Gogotsi, Nature Mater.7, 845-854 (2008).
- G. Pognon, T. Brousse, D. Bélanger, Carbon 49, 1340-1348 (2011).
- G. Lota, B. Grzyb, H. Machnikowska, J. Machnikowski, E. Frackowiak, Chem. Phys. Lett. 404, 53-58 (2005).
- J. M. Black, D. Walters, A. Labuda, G. Feng, P.C. Hillesheim, S. Dai, P. T. Cummings, S. V. Kalinin, R. Proksch, N. Balke, Nano Lett.13, 5954-5960 (2013).
- A. Labuda, P. Grütter, Langmuir28, 5319-5322 (2012).
- A. Labuda, K. Kobayashi, Y. Miyahara, P. Grütter, Rev. Sci. Instrum.83, 053703 (2012).
- M. Jitvisate, J. R. J. Seddon, Phys. Chem.C120, 4860-4865 (2016).
- J. R. J. Seddon, Phys. Chem.C118, 22197-22201 (2014).
This information has been sourced, reviewed and adapted from materials provided by Asylum Research - An Oxford Instruments Company.
For more information on this source, please visit Asylum Research - An Oxford Instruments Company.