COMMUNICATION
www.rsc.org/analyst | The Analyst
Using XPS to determine solute solubility in room temperature ionic liquids{ Debbie S. Silvester,a Tessa L. Broder,a Leigh Aldous,b Christopher Hardacre,b Alison Crossleyc and Richard G. Compton*a Received 8th January 2007, Accepted 26th January 2007 First published as an Advance Article on the web 1st February 2007 DOI: 10.1039/b700212b
X-Ray Photoelectron Spectroscopy (XPS) was used to quantify the amount of bromide ions present in two samples of [C4mpyrr]Br dissolved in the room temperature ionic liquid (RTIL) [C4mpyrr][N(Tf)2]. One sample was of a known concentration (0.436 Br atom%); the other was a saturated solution. The results obtained from quantitative XPS analysis indicated that the saturated sample had a concentration, or solubility, of 0.90 Br atom% (746 mM) at 298 K, which was then independently confirmed by potential-step chronoamperometry of the same solution.
Introduction X-Ray Photoelectron Spectroscopy (XPS) is a useful quantitative spectroscopic technique employed in a variety of industrial applications. It is used to determine the quantity of elements present (within 10 nm of the surface), the empirical formula of the material, the binding energy (BE) of one or more electronic states, the density of the electronic states, and any contamination species.1 To date, XPS has mainly been used to study solid species, and occasionally gases. Recently, various researchers2–7 have used this technique to study room temperature ionic liquids (RTILs). Up until now, XPS studies on liquids have been impossible due to evaporation of the liquid under ultra-high vacuum conditions, but the unique non-volatility characteristics of RTILs mean that this is no longer a problem in this medium. Ionic liquids are defined as salts with melting points below 100 uC, and are composed entirely of ions, generally a bulky, unsymmetrical organic cation and an inorganic anion. Room temperature ionic liquids (RTILs), as their name suggests, are ionic liquids that are liquid at around room temperature. RTILs have been employed in a variety of fields including ‘green’ organic synthesis,8 catalysis with transition metals,9 and in many electrochemical applications.10 They are particularly useful as solvents in electrochemistry due to their wide electrochemical windows and intrinsic conductivity, and their high viscosity means that diffusion a
Physical and Theoretical Chemistry Laboratory, University of Oxford, South Parks Road, Oxford, UK OX1 3QZ. E-mail:
[email protected]; Fax: +44 (0) 1865 275 410; Tel: +44 (0) 1865 275 413 b School of Chemistry and Chemical Engineering/QUILL, Queen’s University Belfast, Belfast, Northern Ireland, UK BT9 5AG c Department of Materials, University of Oxford Begbroke Business and Science Park, Sandy Lane, Yarnton, Oxford, UK OX5 1PF { Electronic supplementary information (ESI) available: experimental details. See DOI: 10.1039/b700212b
196 | Analyst, 2007, 132, 196–198
coefficients of electroactive species are lowered by 1–2 orders of magnitude compared with traditional solvents such as acetonitrile. Their non-volatility and high thermal stability also makes them advantageous for gas sensing.11 XPS studies on RTILs are relatively rare, but a few recent reports show the technique to be useful for various purposes, and have focussed exclusively on pure RTILs. For example, Crespo and co-workers3 first showed the use of RTILs in XPS experiments by testing the stability of RTILs in a supported liquid membrane. Caporali and co-workers2 used XPS to show that the composition of the surface of the RTIL 1-butyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide [C4mim][N(Tf)2] is the same as the bulk, despite low energy ion scattering indicating that the outer layer is composed of F atoms. Licence and co-workers reported an XPS study of the chemical structure of the RTIL 1-ethyl-3methylimidazolium ethylsulfate [C2mim][EtSO4],7 and followed this with a more detailed study of six representative imidazoliumbased RTILs,6 showing that both the presence of alkyl substituents and the anion type does not affect the electron distribution of the imidazolium ring. Then, Kempter and co-workers5 studied the electronic structure of the RTIL 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide [C2mim][N(Tf)2] using XPS among other techniques. Most recently, Wasserscheid and coworkers4 used XPS to study the surface composition of [C2mim][EtSO4] and detect silicon impurities left over from the synthesis of the RTIL. These reports have generally focussed on the structure of the pure liquids, but to the best of the authors’ knowledge, there is only one specifically focussed on species knowingly dissolved in RTILs. That concerns using XPS to monitor the decay of a Pd complex in [C2mim][N(Tf)2].7 We now demonstrate that this technique can be used to determine the solubility of dissolved salts (in the range of 0.5–1.0 atom% of bromide) in these media, and we confirm these results using cyclic voltammetry and potential-step chronoamperometry to verify the concentrations found using XPS.
Results and discussion The RTIL used in this study was 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide [C4mpyrr][N(Tf)2] since the bromide salt of this cation was readily available, and the electrochemical window of this RTIL (structure shown in Fig. 1) is wide (+5 V) and showed no obvious features under vacuum. XPS experiments were performed on two samples, one of a known concentration and one of a saturated solution of [C4mpyrr]Br in the RTIL [C4mpyrr][N(Tf)2], using Mg X-ray excitation (see ESI{ for full experimental details). The spectra obtained on several This journal is ß The Royal Society of Chemistry 2007
Fig. 1 Molecular structures of the cation and anion employed as the solvent in this study.
Fig. 3 XPS spectra of the bromine (3d) photoemission, observed at characteristic binding energy for both the saturated and the 0.436 Br atom% solutions.
Fig. 2 Survey scan XPS spectrum of a 0.436 Br atom% solution of [C4mpyrr]Br in [C4mpyrr][N(Tf)2].
attempts showed well-defined characteristic emissions, and no obvious degradation of the ionic liquid was observed during or after exposure to the X-ray source. Fig. 2 shows a typical XPS spectrum obtained from a 0.436 Br atom% solution of [C4mpyrr]Br in the RTIL [C4mpyrr][N(Tf)2]. The figure is clearly labelled with the characteristic photoemission lines corresponding to C, N, O, F and S atoms present in the RTIL structure. Table 1 shows the quantification (atom%) obtained from the analysis of Fig. 2, and also the results obtained for the saturated solution. In both cases, the elements that make up the RTIL are approximately present in the correct ratio as predicted from its structure (see Fig. 1). There is also a small peak at a binding energy of 69.0 eV, which corresponds to the Br 3d photoemission. Fig. 3 shows a close-up view of this photoemission for both the 0.436 Br atom% (solid line) and the saturated Br solutions (dashed line). The saturated solution peak is clearly larger than that of the known concentration, and the areas under the peaks are quantified to 2 dp12 and are given in Table 1. The concentration obtained from XPS for the more dilute solution was 0.45 Br atom%, which agrees well with the actual value for the concentration (0.436 Br atom%). This therefore leads us to believe Table 1 Peak quantification (expressed as concentration, atom%) for each element presenta in both the 0.436 Br atom% and saturated solutions of [C4mpyrr]Br in the RTIL [C4mpyrr][N(Tf)2] at 298 K Solution
%C
%O
%N
%S
%F
% Br
0.436 Br atom% Saturated Br
42.71 43.57
15.77 15.33
7.74 7.69
8.80 8.42
24.53 24.09
0.45 0.90
a Note: the % concentrations are quoted without the hydrogen atoms, which are also present in the RTIL structure.
This journal is ß The Royal Society of Chemistry 2007
that the saturated solution has a concentration of ca. 0.90 Br atom%, giving the solubility of bromide in [C4mpyrr][N(Tf)2] as ca. 746 mM. In order to further verify the concentrations obtained, cyclic voltammetry and potential-step microdisk chronoamperometry was performed on the two solutions. Both solutions were diluted up approximately to 42 times, giving final concentrations (assuming that the XPS results for the saturated solution was correct) of 8.6 and 17.8 mM respectively, since the initial concentrations (361 mM and 746 mM) were too large for typical voltammetric analysis. Fig. 4 shows cyclic voltammograms for the oxidation of (a) 8.6 mM and (b) 17.8 mM [C4mpyrr]Br in [C4mpyrr][N(Tf)2] on a platinum microdisk electrode (diameter 10 mm) at a scan rate of 100 mV s21. Two oxidative waves are observed at approximately +1.1 V and +1.4 V vs. Ag (with some
Fig. 4 Cyclic voltammograms at 298 K for the oxidation of (a) the saturated solution, and (b) the 0.436 Br atom% solution of [C4mpyrr]Br in [C4mpyrr][N(Tf)2] at a Pt microelectrode (diameter 10 mm). Scan rate 100 mV s21. Note: the solutions were diluted by 42 times to give final concentrations of 8.6 and 17.8 mM respectively, in order for an accurate analysis. Inset shows the experimental (—) and fitted theoretical (#) chronoamperometric transients for the oxidation of [C4mpyrr]Br in [C4mpyrr][N(Tf)2] from (a) 0 to 1.17 V, and (b) 0 to 1.22 V.
Analyst, 2007, 132, 196–198 | 197
difference in the potential noted due to the shift in the silver wire pseudo-reference electrode). The oxidation of Br2 has been previously investigated by Allen et al. in the RTIL [C4mim][N(Tf)2],13 and are thought to occur via the following processes: For the first oxidation: 2Br2 2 2e2 A Br2
(1)
Br2 + Br2 = Br32
(2)
and for the second oxidation: Br32
that the assumption of the concentration of the saturated solution was correct, and that XPS is a valuable technique for quantifying the solubility of bromide in ionic liquids, giving the solubility of Br in [C4mpyrr][N(Tf)2] as ca. 0.90 Br atom%, or 746 mM.
Acknowledgements D. S. S. thanks Schlumberger Cambridge Research and L. A. thanks the Department of Education and Learning in Northern Ireland and Merck GmBH for financial support.
Notes and references 2
= Br + Br2
(3)
Br2 2 e2 A KBr2
(4)
Br32 2 e2 A Br2
(5)
Overall:
Since the two oxidative waves observed in the present study are similar in shape and position to the previous study,13 we can conclude that the same mechanism is taking place. In order to calculate accurate concentrations, chronoamperometry was then performed on the two solutions. The inset to Fig. 4 shows typical chronoamperometric transients carried out on both the (a) 8.6 mM and (b) 17.8 mM solutions. The potential was stepped from 0 V (corresponding to no faradaic current) to a potential after the first oxidative wave, and the experimental data were fit following the Shoup and Szabo expression.14 The results from fitting gave a diffusion coefficient of 4.0 ¡ 0.4 6 10212 m2 s21, which compares relatively well to the value of 1.4 ¡ 0.2 6 10211 m2 s21 obtained in [C4mim][N(Tf)2],13 given the much higher viscosity of the liquid used in the present study {268 cP for [C4mpyrr][N(Tf)2] compared with 44 cP for [C4mim][N(Tf)2]} and concentrations of 9.0 ¡ 0.4 and 17.6 ¡ 0.5 mM for the two solutions respectively. The concentrations are accurate to within experimental error for both the 8.6 and 17.8 mM solutions, so therefore we can conclude
198 | Analyst, 2007, 132, 196–198
1 Practical Surface Analysis, Vol 1: Auger and X-Ray Photoelectron Spectroscopy, ed. D. Briggs and M. P. Seah, John Wiley and Sons, Chichester, 1994. 2 S. Caporali, U. Bardi and A. Lavacchi, J. Electron Spectrosc. Relat. Phenom., 2006, 151, 4–8. 3 R. Fortunato, C. A. M. Afonso, J. Benavente, E. Rodriguez-Castello´n and J. G. Crespo, J. Membr. Sci., 2005, 256, 216–223. 4 J. M. Gottfried, F. Maier, J. Rossa, D. Gerhard, P. S. Schulz, P. Wasserscheid and H.-P. Steinru¨ch, Z. Phys. Chem., 2006, 220, 1439–1453. 5 O. Ho¨fft, S. Bahr, M. Himmerlich, S. Krischok, J. A. Schaefer and V. Kempter, Langmuir, 2006, 22, 7120–7123. 6 E. F. Smith, F. J. M. Rutten, I. J. Villar Garcia, D. Briggs and P. Licence, Langmuir, 2006, 22, 9386–9392. 7 E. F. Smith, I. J. Villar Garcia, D. Briggs and P. Licence, Chem. Commun., 2005, 5633–5635. 8 M. J. Earle and K. R. Seddon, Pure Appl. Chem., 2000, 72, 1391–1398. 9 P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000, 39, 3772–3789. 10 M. C. Buzzeo, R. G. Evans and R. G. Compton, ChemPhysChem, 2004, 5, 1106–1120; F. Endres and S. Zeil El Abedin, Phys. Chem. Chem. Phys., 2006, 8, 1–16; D. S. Silvester and R. G. Compton, Z. Phys. Chem., 2006, 220, 1247–1274. 11 M. C. Buzzeo, C. Hardacre and R. G. Compton, Anal. Chem., 2004, 76, 4583–4588. 12 Note: although normally it is not usual to quote figures to two decimal places, here it seems necessary given the small expected values for the concentration of Br. However, these figures merely represent the limit of detection rather than the accuracy, since a 10% error is possible for values quoted. 13 G. D. Allen, M. C. Buzzeo, I. G. Davies, C. Villagra´n, C. Hardacre and R. G. Compton, J. Phys. Chem. B, 2004, 108, 16 322–16 327. 14 D. Shoup and A. Szabo, J. Electroanal. Chem., 1982, 140, 237.
This journal is ß The Royal Society of Chemistry 2007
