Additions and corrections


The authors wish to publish the following as an addition to their article

Measurement of vapour pressures of ionic liquids and other low vapour pressure solvents

Ortrud Aschenbrenner, Somsak Supasitmongkol, Marie Taylor and Peter Styring

Green Chemistry, 2009, 11, 1217–1221; DOI: 10.1039/b904407h. Amendment published 5th February 2010

Measurement of vapour pressures of ionic liquids and other low vapour pressure solvents: a clarification

Introduction

Our recent paper1 has attracted considerable debate. The question of whether true vapour pressures were observed for the ionic liquids has been at the centre of that debate. The values we reported were three orders of magnitude higher than those reported in the literature2-4, although ours were reported at different temperatures and the literature data extrapolated. Our data were derived using a modification of the Langmuir equation whereby the vapour pressures were measured relative to a standard material that gave similar mass loss characteristics with time. It was assumed that the vapourisation constant (α) for the ionic liquids studied and the reference standard glycerol was constant over all temperatures and compositions. Subsequent studies on other liquids have shown there to be a variation in α with both temperature and molecular structure and this may give rise to the difference in reported vapour pressures and the existing literature. We therefore intend to undertake an extensive study of the variation in α over a wider range of materials. All samples except ethylmethylimidazolium ethylsulfate [C2mim][ES] were commercial samples that were stored in a vacuum desiccator and used as supplied. [C2mim][ES] was synthesised in our laboratories strictly following the literature procedure but adding an extra drying step.5

The evaporation of ionic liquids is attracting continued interest, as is the nature of the species in the vapour phase. Earle et al.6 showed that ionic liquids could be distilled under reduced pressure, however, the nature of the species in the vapour phase were not fully elucidated as NMR was used as the primary analytical technique. However, the paper did show that ionic liquid was recovered in the condensate. Ludwig and Kragl have reviewed the volatility of ionic liquids7 and conclude that it is important to understand the physical properties of any material that is to be considered for use. Recently Seeberger et al. have reported studies of the stability of ionic liquids under ramped elevated temperatures.8 They concluded that different mechanisms are dominant under different temperature programmes with mass loss through thermal evaporation or thermal decomposition. At low heating rates evaporation dominates while at high heating rates decomposition dominates. Therefore, in low temperature isothermal studies, evaporation should be the dominant process. Widegren et al. have also looked at ionic liquid evaporation and have reported relative volatilities9 for a number of materials under reduced pressure using a sublimation technique. Rebelo et al. have used ESI-MS to analyse residue and condensate in a reduced pressure distillation of an aprotic ionic liquid.10 The studies show that there is no decomposition during the distillation process. A number of papers have appeared in recent years where the nature of the species present in the vapour phase have been studied and the heat of vapourisation determined for a wide range of ionic liquids.11-20

Experimental procedures

[emin][ES] was prepared according to the literature procedure5 via ethyl group exchange between diethylsulfate and methylimidazole in toluene, in a water free environment. The reaction solvent was removed under vacuum (13 mbar) using a diaphragm pump at 70 °C until there was no further mass loss (typically four hours). The ionic liquid was then stored in a vacuum desiccator. A small sample was transferred into a short-path distillation bulb and this was further dried in a Kugelrohr at 100 °C / 0.1 mbar for 24 hours. A sample was studied by TGA and the vial was sealed with a screw cap and paraffin film then stored under vacuum until required. Karl Fischer studies (Department of Chemistry, The University of Sheffield) showed a viscous neat sample of [C2mim][ES] contained 0.14% (1380 ppm ) water by mass. A sample was diluted in Karl Fischer grade dichloromethane to aid manipulation and decrease the transfer time and was found to contain 0.07% (657 ppm) water by mass.

Elemental analysis on a dried sample of [C2mim][ES] as used in the Carl Fischer titration was carried out on a Perkin Elmer 2400 Series II analyser. LC-MS studies were carried out on a Waters LCT system. The sample of [C2mim][ES] was dissolved in a mixture of acetonitrile–water and eluted through the LC column. Unfortunately, it was not possible to identify individual species as all components eluted under a single broad peak. Time of flight mass spectrometry was carried out using positive (ES+) and negative (ES-) electrospray techniques. The sample in acetonitrile–water was nebulised at 150 °C onto the probe.

Thermogravimetric analysis-mass spectrometry (TGA-MS) was performed on a NETZSCH Luxx STA 409 PG thermogravimetric analyser with the exhaust gases passing to a NETZSCH Aëolos QMS 403 C mass spectrometer. In order to prevent condensation, the transfer line was maintained isothermally at 200 °C. The sample was placed in a ceramic crucible and placed in the furnace. The furnace was evacuated to remove oxygen and purged with dry nitrogen. 1% of the exhaust gases were split and sent to the mass spectrometer through a transfer line maintained at 200 °C.

Headspace MS was carried out on a VG Autospec mass spectrometer on a sample of the vapour obtained on heating a sample of [C2mim][ES] at 200 °C and collecting the vapour in a glass tube that was then evacuated onto the MS probe and analysed using EI. The positioning of the tube ensured that there was no possible contact with the ionic liquid through splashing and any material collected was from the vapour phase only.

Results and discussion

Elemental analysis was expected to show C 40.67%, H 6.83 %, N 11.86% for C8H16N2O4S. The values found were C 40.90%, H 7.05%, N 12.42%, which is close to the expected value but outside acceptable limits for nitrogen. The TGA-MS data indicates the presence of methylimidazole (mim) in the vapour phase. This could come from either unreacted starting material or as a decomposition product. The expected values were recalculated in Microsoft Excel to account for 0.07% water and different mole fractions of mim. The best fit was obtained (C 40.73%, H 6.82%, N 12.23%) for a 2 mol% contamination of mim in the ionic liquid. In terms of mass, this corresponds to 0.7%. Therefore, the total impurities by mass are <0.8% of the sample analysed.

If we consider a thermogram of [C2mim][ES], then a plot of %weight loss against time should lead to a plateau at around 99.2% once all the water and free mim are removed, if no other weight loss process occurs. The average rate of weight loss over the initial TGA experiment at 120 °C was 2.1 μg min-1 over a 30 minute period, however, this followed a period of high temperature drying in the TGA instrument. The weight loss was linear and finite and was measured relative to glycerol as the reference standard. Using this differential approach means that any baseline drift in the instrument can be eliminated from calculations. The experimental procedure was as follows: the weighed sample was held isothermally for 70 min at 120 °C then allowed to cool to 20 °C at 20 °C/min. The sample was then heated isothermally at the experimental temperature (100 or 120 °C) for at least 30 minutes to give a stable weight loss line of constant gradient.

A typical sample mass used in the experiments was 25 mg, which, according to Karl Fischer data contains up to 34.5 μg of water. This means complete water loss would be achieved in under 17 min at 120 °C, which is during the initial sample conditioning / drying period. Therefore, any water would be lost before the gradient of the weight loss slope was determined. This corresponds to the rapid initial weight loss observed during the conditioning phase, so suggests water is not evolved during the data collection period. However, during the 30 min data collection phase, 63 μg of material is lost from the liquid to the vapour phase.

We repeated the studies at 120 and 200 °C using combined TGA-MS analysis on a NETZSCH instrument (Sheffield Hallam University). The resulting thermograms are shown collectively in Figure 1.

Fig. 1 Isothermal thermograms for [C2mim][ES] at 120 and 220 °C after an initial in situ drying procedure.

The thermal programme was as follows. All samples were heated at a rate of 10 °C/min under a nitrogen flow. Samples PS1 and PS2 were heated isothermally at 120 °C while PS5 was heated to 220 °C. The exhaust gases were split and a 1% volume fraction sent to the mass spectrometer. Sample PS1 was sampled for all ions using a quick scan method while the other samples were scanned selectively for expected ions and fragments.

The blank run in Figure 1 corresponds to an empty ceramic crucible and indicates there is no baseline drift over the course of the isothermal scan. Scans for PS1 and PS2 show a similar trend with a constant mass loss over time at 120 °C. When the temperature was increased to 220 °C (PS5) the thermogram showed a similar profile but with a steeper gradient (dm/dt) as expected. MS analysis did not show a mass ion peak for [C2mim][ES] but did show evidence for mim and ethylsulfate fragmentation (m/z 44, 41,40, 18, 17, 16). The peaks at m/z 16-18 could be interpreted as being from adsorbed water, however, they appear across the complete time scan at mass loss values far greater than shown by the Karl Fischer titration or elemental analysis results. Therefore, it is proposed that these fragments arise from decomposition of the ethylsulfate group.

Time of flight (TOF) electospray mass spectrometry in ES+ and ES- modes showed evidence of the [C2mim][ES] species in the form of clusters. The mass spectra are shown in Figures 2-4. While the molecular ion was not observed at m/z 236, we did observe peaks in ES+ mode for [[C2mim]2[ES]]+ (m/z 347.2), [[C2mim]3[ES]2]+ (m/z 583.3), [[C2mim]4[ES]3]+ (m/z 819.5), [[C2mim]5[ES]4]+ (m/z 1055.6), [[C2mim]6[ES]5]+ (m/z 1291.7); and in ES- mode for [[C2mim][ES]2]- (m/z 361.1), [[C2mim]2[ES]2]- (m/z 397.2).

Fig. 2 Electrospray mass spectrum (ES+) of [C2mim]+[ES]- showing clusters.

Fig. 3 Expanded electrospray mass spectrum (ES+) of [C2mim]+[ES]- showing clusters.

Fig. 4 Electrospray mass spectrum (ES-) of [C2mim]+[ES]- showing clusters.

One possibility for the continued mass loss over prolonged isothermal heating studies is that methylimidazole (mim) and diethyl sulphate (DES) are being formed by a reversible reaction of [C2mim][ES] back to the starting materials at the elevated temperatures. The other is that the ionic liquid is in equilibrium with the vapour phase. While the reversible process has been reported for protic ionic liquids21,22 it has not been the case for alkylated derivatives. The second possibility is that the sample is decomposing isothermally at 120 °C, however, while the onset of catastrophic decomposition was not observed until above 300 °C on a thermal ramping experiment it is quite conceivable that decomposition occurs isothermally at elevated temperatures, particularly in an open top platinum crucible and over prolonged temperatures. It has been shown by Seeberger et al.8 that decomposition occurs at high scan rates, especially at higher temperatures. Differential scanning at high rates tends to give rise to decomposition well below the expected decomposition onset. Therefore, at lower temperatures under isothermal conditions evaporation should be the primary mass loss process. Evaporation experiments under ultra high vacuum coupled with line of sight mass spectrometry have shown evidence for the neutral alkyl imidazole in the vapour phase,11 but no evidence of discrete tight ion pairs. One method used to measure the enthalpy of vapourisation used isothermal TGA to determine weight loss which is then used to determine the enthalpy value although the nature of the species in the vapour phase was not determined.23 In these cases the vapour pressure was considerably lower than those we reported for similar materials.

We therefore sampled the headspace above a sample of [C2mim][ES] which was held at 200 °C using a glass tube held at the same temperature. The tube was then transferred to the mass spectrometer in which the source and probe were maintained at 200 °C at a reduced pressure of 5 x 10-7 bar. The elevated temperature was chosen as it is shown in the TGA experiments to give a faster rate of mass loss while the mode of action does not appear to change. Electron ionisation gave a complex spectrum of fragment ions but no molecular ion at m/z 236. The mass spectrum is shown in Figure 5. However, there was evidence for the molecular ion with loss of two ethyl groups (186-188). All bracketed values represent numerical values of m/z for the fragment ions. A peak was also observed corresponding to [ES-CH3]+ (139) and [ES-CH2CH3]+ (123). Perhaps the most interesting peak was observed at m/z 111 which corresponds to the [C2mim]+ fragment. As this could only have been observed via transfer from the ionic liquid to the vapour phase it is evidence that the ionic liquid loses this fragment through either evaporation or decomposition. There is evidence of fragmentation of both the [C2mim]+ (96) corresponding to [C2mim-CH3]+ and the mim group (81 (molecular ion of mim-H), 67, 54). The peak at m/z 96 could be due to fragmentation of [C2mim]+ in the MS, but equally could be a decomposition product as a result of a hydride shift in the [C2mim]+ cation to give ethylimidazole. The peak at 100% intensity was observed at m/z 108. The only rationale for this is [ES-O]+ which is consistent with the fragmentation pattern observed for [DES]. No molecular ion was observed for [DES] itself, so it can be concluded that the peak at 108 is derived from the [ES]- anion.

Fig. 5 Electron ionisation mass spectrum (EI+) of [C2mim]+[ES]-.

In conclusion, it appears likely that the phenomenon we reported was perhaps not simple vapourisation of the ionic liquid but possibly evaporation plus a slow isothermal decomposition process. Figure 1 clearly shows that mass loss is considerable and increases with temperature. Studies are now being carried out to fully understand the process, looking for temperature dependent decomposition products both in the vapour and liquid phases for [C2mim][ES] and other ionic liquids with diverse structures and to determine the evaporation coefficient for use in the Langmuir equation.

References

1. O. Aschenbrenner, S. Supasitmongkol, M. Taylor and P. Styring, Green Chem., 2009, 11, 1217.

2. Q. Dong, C. D. Muzny, A. Kazakov, V. Diky, J. W. Magee, J. A. Widegren, R. D. Chirico, K. N. Marsh and M. Frenkel, J. Chem. Eng. Data, 2007, 52, 1151.

3. http://ilthermo.boulder.nist.gov/ILThermo/ (accessed 07 May 2009).

4. D. H. Zaitsau, G. J. Kabo, A. A. Strechan, Y. U. Paulechka, A. Tschersich, S. P. Verevkin and A. Heintz, J. Phys. Chem. A, 2006, 110, 7303–7306.

5. D. J. Holbrey, R. W. Matthew, P. R. Swatloski, A. G. Broker, W.R. Pitner, K. R. Seddon and R. D. Rogers, Green Chem., 2002, 4, 407.

6. M. J. Earle, J. Esperanca, M. A. Gilea, J. N. C. Lopes, L. P. N. Rebelo, J. W. Magee, K. R. Seddon and J. A. Widegren, Nature, 2006, 439, 831–834.

7. R. Ludwig and U. Kragl, Angew. Chem., Int. Ed. Engl., 2007, 46, 6582–6584.

8. A. Seeberger, A.-K. Andresen and A. Jess, Phys. Chem. Chem. Phys., 2009, 11, 9375–9381.

9. J. A. Widegren, Y. M. Wang, W. A. Henderson and J. W. Magee, J. Phys. Chem. B, 2007, 111, 8959–8964.

10. L. P. N. Rebelo, J. N. C. Lopes, J. Esperanca, H. J. R. Guedes, Lachwa, V. Najdanovic-Visak and Z. P. Visak, Acc. Chem. Res., 2007, 40, 1114–1121.

11. J. P. Armstrong, C. Hurst, R. G. Jones, P. Licence, K. R. J. Lovelock, C. J. Satterley and I. J. Villar-Garcia, Phys. Chem. Chem. Phys., 2007, 9, 982–990.

12. J. P. Leal, J. Esperanca, M. E. M. da Piedade, J. N. C. Lopes, L. P. N. Rebelo and K. R. Seddon, J. Phys. Chem. A, 2007, 111, 6176–6182.

13. D. Strasser, F. Goulay, M. S. Kelkar, E. J. Maginn and S. R Leone, J. Phys. Chem. A, 2007, 111, 3191–3195.

14. R. G. Jones, P. Licence, K. R. J. Lovelock and I. J. Villar-Garcia, Ind. Eng. Chem. Res., 2007, 46, 6061–6062.

15. M. S. Kelkar and E. J. Maginn, J. Phys. Chem. B, 2007, 111, 9424–9427.

16. V. N. Emel’yanenko, S. P. Verevkin, A. Heintz, J.-A. Corfield, A. Deyko, K. R. J. Lovelock, P. Licence and R. G. Jones, J. Phys. Chem. B, 2008, 112, 11734–11742.

17. K. R. J. Lovelock, A. Deyko, J.-A. Corfield, P. N. Gooden, P. Licence and R. G. Jones, Chem. Phys. Chem., 2009, 10, 337–340.

18. A. Deyko, K. R. J. Lovelock, J. A. Corfield, A. W. Taylor, P. N. Gooden, I. J. Villar-Garcia, P. Licence, R. G. Jones, V. G. Krasovskiy, E. A. Chernikova and L. M. Kustov, Phys. Chem. Chem. Phys., 2009, 11, 8544–8555.

19. N. Akai, D. Parazs, A. Kawai and K. J. Shibuya, J. Phys. Chem. B, 2009, 113, 4756–4762.

20. O. Borodin, J. Phys. Chem. B, 2009, 113, 11463–11478.

21. J. von Zawidski, Z. Phys. Chem., 1900, 35, 129–203.

22. R. G. Treble, K. E. Johnson and E. Tosh, Can. J. Chem., 2006, 84, 915–924.

23. H. Luo, G. A. Baker and S. Dai, J. Phys. Chem. B Letters, 2008, 112, 10077–10081.


Back