Picking our RTIL Anion- The first major goal before research could begin was choosing which ionic liquid and respective doping salt we were going to use. In order to simplify the system we had to make sure that both the anion of the IL and that of the salt were the same, given the nature of the CAF we wanted to use an anion that had the minimum possible ionic interaction to avoid ionic pairing in solution at least for the Ionic Liquid. Given this the triflate (trifluoromethanesulfonate) anion known commonly as triflate or TFSI (bis(trifluoromethane)sulfonamide) anion known commonly as bistriflimide were considered the most ideal. For reasons I don’t fully understand Prof Fleshman has said that the triflate, given that it is far less of a complex molecule than bistriflimide, would be better for this study than bistriflimide. Personally, outside of the context of the CAF, TFSI would seem far more ideal for any study given the extensive research already being done on doping RTIL TFSIs with lithium, and the complex structures TFSI forms with alkali metals (see “Li+ Solvation and Transport Properties in Ionic Liquid/Lithium Salts Mixtures: A Molecular Dynamics Simulation Study” by Li et al 2012 for more on the coordination structure of Lithium and TFSI in RTILs). It must be noted that although triflate and bistriflimide were the primary candidates for this study many RTIL anions such as PF6– , BF4–, and AlCl4– exist. For a review of different types of ionic liquids see “Ionic liquids as electrolytes” by Galinski et al. 2006.
Picking our RTIL Cation– If you are reading this you are probably at least somewhat familiar with Dr. Fleshman’s Research, and if you are not than this entire document will probably be of very little use to you. Given that we need to alter the dielectric constant to validate the central claim of the CAF, i.e. that the temperature dependence of the exponential prefactor is entirely dependent on the temperature dependence of the dielectric constant, we need a RTIL cation that we can consistently add more and more onto an alkyl chain. Given that almost all RTIL cations are bulky organic molecules this is actually not that hard, making the ‘family’ relatively simple to exploit. Our main debate was between the pyrrolidinium based RTIL cation and a imidazolium based cation.
Now, honestly, there is far more literature out there about doping Lithium salts, especially LiTFSI, into pyrrolidinium based RTILs than there is for imidazolium based RTILs, and by far this would be my first choice and maybe an avenue of study for any reading this. The reason we choose imidazolium was because this continued Allison’s work with Gus Lowrey using imidazolium triflate RTILs for the CAF and was already the direction she was taking with several other colleagues outside of Lawrence. I found surprisingly little in the literature about doping lithium into imidazolium based RTILs, the one notable exception being “Transport Coefficients, Raman Spectroscopy, and Computer Simulation of Lithium Salt Solutions in an Ionic Liquid” by Monterio et al. 2008 which dopes BMMI (1-butyl-2,3-dimethylimidazolium) TFSI with LiTFSI which, considering I found this paper after my 3rd day of literature searching, was quite the gold mine for information.
Doping Concentrations- As we mentioned earlier we will not be using TFSI but triflate, and in terms of resources for doping LiTf salt into the Cnmim series or frankly any imidazolium based RTIL is almost non-existent. Previously someone in the lab had attempted to dope Alkali-Tf salts into some Cmim to find that it was insoluble. Given the very expensive nature of these RTILs there is not a lot of room for mistake, so we want to make sure we have everything under wraps before continuing. On one hand Monterio et al. 2008 (mentioned above) claims that they doped up to a 0.38 LiTFSI mole fraction, which gives some hope, but sadly TFSI is not Triflate. An online resource I found from Berkley called “Ionic Liquids for Rechargeable Lithium Batteries” from 2005 (link: http://www.osti.gov/scitech/servlets/purl/929086) claims that someone has made a 1 Molal solution of LiTf in Bmim meaning that they would have made a solution with 0.22 mole fraction LiTf which is hopeful, but again this is for Bmim only, we might imagine that the solubility changes with a change in alkyl chain length. My hope is that we can do two concentrations of LiTf, one at 0.1 mole frac and the other at 0.2 mole frac, but Allison has yet to hear back from her colleagues. The expansive work done on LiTFSI and pyrrolidinium RTILs is primarily why my first instinct was for that combination. Another thing to note is a lot of studies of doping RTILs use heat or add ACN into the solution to fully dissolve the salt, given the nature of this study this is not an option.