Thermoresponsive magnetic ionic liquids: synthesis and temperature switchable magnetic separation

Qipeng Zhaoa, Tun Seng Herngb, Chun Xian Guoa, Dieling Zhaoa, Jun Ding*b and Xianmao Lu*a
aDepartment of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117585. E-mail: chelxm@nus.edu.sg
bDepartment of Materials Science and Engineering, National University of Singapore, Singapore 117574. E-mail: msedingj@nus.edu.sg

Received 15th January 2016 , Accepted 29th January 2016

First published on 1st February 2016


Abstract

Thermoresponsive magnetic ionic liquids (MILs) with lower critical solution temperature (LCST) below 60 °C are synthesized. Magnetic separation of MILs from aqueous solution at temperatures above LCST is demonstrated.


Ionic liquids (ILs), which are molten salts with melting points below 100 °C, have been intensively studied recently.1–7 Due to their interesting properties including low vapour pressure, high thermal and chemical stability, and high conductivity,3 ILs have found applications in fields including catalysis,4,6 electrochemistry,5,7 and separation.8,9 Recently, magnetic ionic liquids (MILs) exhibiting susceptibility to an external magnetic field have attracted increasing attention.10–12 MILs are generally obtained by incorporating high-spin transition metal ions into the structure of conventional ILs. Since the discovery of the first MIL, 1-butyl-3-methylimidazolium tetrachloroferrate, by Hayashi and Hamaguchi in 2004,10 a few MILs based on transition metals such as Co(II),13–15 Mn(II),16 and Gd(III)15,17 have been reported. These MILs have been investigated in various applications,18–20 especially as recyclable catalysts for chemical reactions such as Grignard cross-coupling of alkyl halides21 and fuel desulfurization,22 where the recyclability of catalysts is of significant importance for cost and environmental considerations. However, most MILs synthesized so far are hydrophobic, so their application in catalytic reactions involving aqueous solution is rather limited.23 Although a few hydrophilic MILs have been reported, due to their low magnetic susceptibility and strong interaction with water molecules, they are difficult to separate from aqueous solution magnetically.24,25 Here we show that by incorporating thermoresponsive behaviour to hydrophilic MILs, temperature switchable magnetic separation of hydrophilic MILs from aqueous solution can be achieved with the assistance of mild heating.

It is known that ILs based on phosphonium or ammonium salts can exhibit lower critical solution temperature (LCST) behaviour.26,27 Upon heating to temperatures above their LCSTs, IL solutions undergo phase separation. It is expected that thermoresponsive MILs can also be synthesized by choosing appropriate component ions so that reversible mixing of MILs with water can be attained with thermal stimulus. In this work, by carefully selecting the anions and cations, we prepared four thermoresponsive MILs, namely octyltrimethylammonium bromotrichloroferrate (N1,1,1,8[FeBrCl3]), dodecyltrimethylammonium tetrachloroferrate (N1,1,1,12[FeCl4]), tetradedecyltrimethylammonium bromotrichloroferrate (N1,1,1,14[FeBrCl3]), and 8-butyl-1,8-diazabicyclo[5.4.0]undec-7-ene bromotrichloroferrate (DBU-Bu[FeBrCl3]). All four MILs respond to temperature change (Fig. 1), while two of them exhibit LCST below 60 °C. To the best of our knowledge, this is the first report of thermoresponsive MILs. Detailed synthesis of thermoresponsive MILs can be found in ESI. Briefly, ammonium salts or amidine salts were dissolved in dichloromethane, followed by adding equal mole of FeCl3·6H2O with stirring for 24 h.28,29 The MILs were then purified for further characterisations. The yields were 89.9%, 90.4%, 72.7% and 93.0% for N1,1,1,8[FeBrCl3], N1,1,1,12[FeCl4], N1,1,1,14[FeBrCl3] and DBU-Bu[FeBrCl3], respectively.


image file: c6ra01235c-f1.tif
Fig. 1 Preparation of thermoresponsive MILs with different cations: octyltrimethylammonium bromotrichloroferrate (N1,1,1,8[FeBrCl3]), dodecyltrimethylammonium tetrachloroferrate (N1,1,1,12[FeCl4]), tetradedecyltrimethylammonium bromotrichloroferrate (N1,1,1,14[FeBrCl3]), and 8-butyl-1,8-diazabicyclo[5.4.0]undec-7-ene bromotrichloroferrate (DBU-Bu[FeBrCl3]).

The successful synthesis of MILs was verified by FT-IR and Raman spectroscopy. From the FTIR spectrum of each resulting MIL (Fig. 2a), two peaks were found at 2854 and 2920 cm−1, attributing to –CH2 and –CH3 stretching in the organic moieties of the cations. For N1,1,1,8[FeBrCl3], N1,1,1,12[FeCl4], and N1,1,1,14[FeBrCl3], the peak at 1487 cm−1 corresponds to N–CH2 bending in the ammonium moieties, while for DBU-Bu[FeBrCl3], the peaks at 1327 and 1620 cm−1 are due to the stretching of C–N–C and C[double bond, length as m-dash]N in the amidine group. Raman analysis was conducted to obtain the structure of the anion of each MIL (Fig. 2b). The Raman spectra of N1,1,1,8[FeBrCl3], N1,1,1,14[FeBrCl3] and DBU-Bu[FeBrCl3] are similar – five peaks at 340, 326, 260, 237, and 216 cm−1 corresponding to [FeBrCl3] were found for each spectrum,30 while N1,1,1,12[FeCl4] shows the characteristic peak of Fe–Cl at 329 cm−1.30,31 The presence of either [FeCl4] or [FeBrCl3] in the MILs was further confirmed with UV-vis spectroscopy (Fig. S2 in ESI). Three typical absorption bands of [FeCl4] at 533, 615 and 688 nm can be observed for N1,1,1,12[FeCl4]. Five characteristic bands at 628, 650, 671, 713 and 733 nm attributing to [FeBrCl3] are shown in all three spectra of N1,1,1,8[FeBrCl3], N1,1,1,14[FeBrCl3], and DBU-Bu[FeBrCl3].31 We further conducted elemental analysis of the MILs using X-ray photoelectron spectroscopy (XPS). The atomic concentrations of Fe, N, Cl, and Br in each MIL were listed in Table S1 (ESI). For N1,1,1,8[FeBrCl3], N1,1,1,12[FeCl4], N1,1,1,14[FeBrCl3], and DBU-Bu[FeBrCl3], the atomic ratios of Fe[thin space (1/6-em)]:[thin space (1/6-em)]N[thin space (1/6-em)]:[thin space (1/6-em)]Cl[thin space (1/6-em)]:[thin space (1/6-em)]Br were 1[thin space (1/6-em)]:[thin space (1/6-em)]0.8[thin space (1/6-em)]:[thin space (1/6-em)]2.8[thin space (1/6-em)]:[thin space (1/6-em)]0.7, 1[thin space (1/6-em)]:[thin space (1/6-em)]1.2[thin space (1/6-em)]:[thin space (1/6-em)]4.4[thin space (1/6-em)]:[thin space (1/6-em)]0, 1[thin space (1/6-em)]:[thin space (1/6-em)]1.2[thin space (1/6-em)]:[thin space (1/6-em)]3.4[thin space (1/6-em)]:[thin space (1/6-em)]1.2, and 1[thin space (1/6-em)]:[thin space (1/6-em)]2.5[thin space (1/6-em)]:[thin space (1/6-em)]3.9[thin space (1/6-em)]:[thin space (1/6-em)]1.2, respectively. These ratios are consistent with their nominal ratios in the molecular formula.


image file: c6ra01235c-f2.tif
Fig. 2 (a) FT-IR and (b) Raman spectra of thermoresponsive MILs.

Magnetic property of MILs is important for their application in magnetic extraction. Here, we measured the magnetic properties of our MILs using SQUID method with a magnetic field range of −20[thin space (1/6-em)]000 to 20[thin space (1/6-em)]000 Oe at 300 K (Fig. 3). All four MILs exhibited paramagnetic behaviour. From the slope of the magnetic field dependence, the magnetic susceptibilities of the MILs were determined to be 23.2 × 10−6, 16.7 × 10−6, 27.7 × 10−6 and 21.9 × 10−6 emu g−1 for N1,1,1,8[FeBrCl3], N1,1,1,12[FeCl4], N1,1,1,14[FeBrCl3], and DBU-Bu[FeBrCl3], respectively. The magnetic susceptibilities are comparable to typical MILs such as Bmim[FeCl4] which has a magnetic susceptibility of 40.6 × 10−6 emu g−1.10


image file: c6ra01235c-f3.tif
Fig. 3 Magnetization of (a) N1,1,1,8[FeBrCl3], (b) N1,1,1,12[FeCl4], (c) N1,1,1,14[FeBrCl3], and (d) DBU-Bu[FeBrCl3] as a function of applied magnetic field.

To investigate the thermoresponsive behaviour of MILs, the change of UV-vis transmittance with temperature at 600 nm was measured for each MIL with a concentration of 25 wt%. From Fig. 4a, the LCSTs of N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3] are 60 °C and 50 °C, respectively, at which the transmittances of the MILs dropped abruptly. Although the transmittances for N1,1,1,8[FeBrCl3] and N1,1,1,14[FeBrCl3] also decreased with temperature, no sudden drop in transmittance could be observed, indicating these two MILs do not exhibit LCST behaviour. We further studied the effect of concentration on the LCSTs of N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3] (Fig. 4b). The LCSTs of N1,1,1,12[FeCl4] solutions at concentrations of 25, 30, and 33 wt% were 60, 48, and 38 °C, respectively. For DBU-Bu[FeBrCl3], the LCSTs were 63 and 50 °C at concentrations of 20 wt% and 25 wt%, respectively. When the concentrations were below 25 and 20 wt% for N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3], respectively, no LCST behaviour was observed below 70 °C. Thus, it can be concluded that lower concentrations of thermoresponsive MILs lead to higher LCSTs.


image file: c6ra01235c-f4.tif
Fig. 4 (a) Change in UV-vis transmittance with temperature for of 25 wt% thermoresponsive MILs. (b) Change in UV-vis transmittance with temperature for N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3] at different concentrations. (c–f) Heating DBU-Bu[FeBrCl3] solution from 24 °C to 50 °C and magnetic separation of DBU-Bu[FeBrCl3] from water at 50 °C.

Due to the phase separation of MILs from water at temperatures above LCST, the magnetic separation of MILs from aqueous solutions should be improved. The temperature switchable magnetic separation of thermoresponsive MILs was demonstrated with 25 wt% DBU-Bu[FeBrCl3] aqueous solution as an example (Fig. 4c–f). The solution is translucent at room temperature (24 °C), and no response to external magnetic field can be observed (Fig. 4c). Upon heating up to its LCST (50 °C) for 20 min, the solution became turbid. After 30 min, dark oily droplets were formed at the bottom of the glass vial. By applying an external magnetic field, the oily droplets were immediately attracted to the side of the vial. After cooling down with ice bath, the oily droplets could be dissolved again to form a homogeneous solution as shown in Fig. 4c. This process was also observed for the aqueous solution of N1,1,1,12[FeCl4] at 25 wt% (Fig. S3). By measuring the concentrations of the aqueous phase before and after phase separation, the concentrations of DBU-Bu[FeBrCl3] and N1,1,1,12[FeCl4] solutions dropped to 16.5 wt% and 20.2 wt%, respectively. The water content in the MIL phase was obtained with thermogravimetric analysis (TGA). As shown in Fig. S4 (ESI), 10.6 and 20.5 wt% of water were found for DBU-Bu[FeBrCl3] and N1,1,1,12[FeCl4], corresponding to extraction efficiencies of 43 and 19%, respectively.

Similar to typical thermoresponsive ILs, the LCST behaviour of MILs can be attributed to the hydrophilic/hydrophobic transition at different temperatures. Based on previous studies, it has been found that the hydrophilicity of the component ions and the length of the hydrophobic chain of thermoresponsive ILs could significantly influence their phase behaviour – longer side chains or more halogen atoms could have higher water-shedding effect which gives rise to lower dissociation or lower solubility in water.26,27 In our work, different combinations of cations and anions were tested for water solubility and thermoresponsiveness. The results are summarized in Table 1. Phosphonium cations with long side chains such as tetrabutylphosphonium ([P4,4,4,4]+) and octyltributylphosphonium ([P4,4,4,8]+) were first tried with [FeEDTA] anion. The resulting MILs are soluble in water, but they are too hydrophilic to show LCST. After replacing [FeEDTA] with a more hydrophobic anion [FeBrCl3], these resulting MILs become too hydrophobic to be dissolved in water. Therefore, more hydrophilic ammonium/amidine cations were then employed to form MILs with [FeBrCl3] (Fig. 1). The four MILs all are thermoresponsive. However, as shown in Fig. 4a, only N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3] exhibit LCST, indicating these two MILs are more hydrophobic than N1,1,1,8[FeBrCl3] and N1,1,1,14[FeBrCl3]. From the solubility and thermoresponsiveness tests, the hydrophilicities of the component ions should be in the order of [P4,4,4,8]+ < [P4,4,4,4]+ < [DBU-Bu]+ < [N1,1,1,14]+ < [N1,1,1,12]+ < [N1,1,1,8]+, and [FeCl4] < [FeBrCl3] < [FeEDTA] for cations and anions, respectively. Since the hydrophilicity of cations and anions can remarkably affect the phase behaviour of IL and water mixture, only MILs with cations and anions of moderate hydrophilicity could undergo LCST transition. Due to the importance of hydrophilicity on the LCST behaviour of MILs, we examined their hydrophilicity indices (HIs).27 By measuring the water contents of the separated MILs after phase separation (Fig. S4, ESI), the HIs were calculated to be 3.0 and 6.1 for DBU-Bu[FeBrCl3] and N1,1,1,12[FeCl4], respectively, indicating that DBU-Bu[FeBrCl3] is less hydrophilic than N1,1,1,12[FeCl4]. This explains the lower LCST of DBU-Bu[FeBrCl3] than N1,1,1,12[FeCl4] at the same concentration.

Table 1 Solubility and thermoresponsiveness of MILs with different combinations of cations and anions
Cation Anion Soluble in H2O Thermoresponsive
[P4,4,4,4]+ [FeEDTA] Y N
[P4,4,4,8]+ [FeEDTA] Y N
[P4,4,4,4]+ [FeBrCl3] N
[P4,4,4,8]+ [FeBrCl3] N
[N1,1,1,8]+ [FeBrCl3] Y Y
[N1,1,1,12]+ [FeCl4] Y Y
[N1,1,1,14]+ [FeBrCl3] Y Y
[DBU-Bu]+ [FeBrCl3] Y Y


In conclusion, we synthesized four new MILs – N1,1,1,8[FeBrCl3], N1,1,1,12[FeCl4], N1,1,1,14[FeBrCl3] and DBU-Bu[FeBrCl3]. Among them, N1,1,1,12[FeCl4] and DBU-Bu[FeBrCl3] showed LCSTs of 60 and 50 °C, respectively. Temperature switchable magnetic separation of MILs was demonstrated. These results suggest that appropriate combination of cations and anions can result in LCST-type MILs that may find applications such as recyclable catalysts for reactions in aqueous solutions.

Acknowledgements

We are grateful for the financial support of the National Research Foundation CRP program (Grant# R279-000-337-281).

Notes and references

  1. J. Dupont, R. F. de Souza and P. A. Suarez, Chem. Rev., 2002, 102, 3667–3692 CrossRef CAS PubMed.
  2. J. S. Wilkes and M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 1992, 13, 965–967 RSC.
  3. 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 CrossRef CAS PubMed.
  4. J. P. Hallett and T. Welton, Chem. Rev., 2011, 111, 3508–3576 CrossRef CAS PubMed.
  5. D. R. Macfarlane, M. Forsyth, P. C. Howlett, J. M. Pringle, J. Sun, G. Annat, W. Neil and E. I. Izgorodina, Acc. Chem. Res., 2007, 40, 1165–1173 CrossRef CAS PubMed.
  6. V. I. Parvulescu and C. Hardacre, Chem. Rev., 2007, 107, 2615–2665 CrossRef CAS PubMed.
  7. P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000, 39, 3772–3789 CrossRef CAS.
  8. E. Santos, J. Albo and A. Irabien, J. Membr. Sci., 2014, 452, 277–283 CrossRef CAS.
  9. S. Potdar, R. Anantharaj and T. Baneriee, J. Chem. Eng. Data, 2012, 57, 1026–1035 CrossRef CAS.
  10. S. Hayashi and H. O. Hamaguchi, Chem. Lett., 2004, 33, 1590–1591 CrossRef CAS.
  11. H. Wang, R. Yan, Z. Li, X. Zhang and S. Zhang, Catal. Commun., 2010, 11, 763–767 CrossRef CAS.
  12. M. Valkenberg and W. Hölderich, Appl. Catal., A, 2001, 215, 185–190 CrossRef CAS.
  13. A. Branco, L. C. Branco and F. Pina, Chem. Commun., 2011, 47, 2300–2302 RSC.
  14. T. Peppel, M. Kockerling, M. Geppert-Rybczynska, R. V. Ralys, J. K. Lehmann, S. P. Verevkin and A. Heintz, Angew. Chem., Int. Ed., 2010, 49, 7116–7119 CrossRef CAS PubMed.
  15. R. E. Del Sesto, T. M. McCleskey, A. K. Burrell, G. A. Baker, J. D. Thompson, B. L. Scott, J. S. Wilkes and P. Williams, Chem. Commun., 2008, 44, 447–449 RSC.
  16. S. Pitula and A. V. Mudring, Chem.–Eur. J., 2010, 16, 3355–3365 CrossRef CAS PubMed.
  17. B. Mallick, B. Balke, C. Felser and A. V. Mudring, Angew. Chem., Int. Ed., 2008, 47, 7635–7638 CrossRef CAS PubMed.
  18. N. Deng, M. Li, L. J. Zhao, C. F. Lu, S. L. de Rooy and I. M. Warner, J. Hazard. Mater., 2011, 192, 1350–1357 CrossRef CAS PubMed.
  19. E. Santos, J. Albo and A. Irabien, RSC Adv., 2014, 4, 40008–40018 RSC.
  20. S. K. Panja and S. Saha, RSC Adv., 2013, 3, 14495–14500 RSC.
  21. K. Bica and P. Gaertner, Org. Lett., 2006, 8, 733–735 CrossRef CAS PubMed.
  22. W. S. Zhu, P. W. Wu, L. Yang, Y. H. Chang, Y. H. Chao, H. M. Li, Y. Q. Jiang, W. Jiang and S. H. Xun, Chem. Eng. J., 2013, 229, 250–256 CrossRef CAS.
  23. K. Bica and P. Gaertner, Eur. J. Org. Chem., 2008, 20, 3453–3456 CrossRef.
  24. M. Wang, B. Li, C. Zhao, X. Qian, Y. Xu and G. Chen, Korean J. Chem. Eng., 2010, 27, 1275–1277 CrossRef CAS.
  25. S. H. Lee, S. H. Ha, C. Y. You and Y.-M. Koo, Korean J. Chem. Eng., 2007, 24, 436–437 CrossRef CAS.
  26. Y. Kohno and H. Ohno, Chem. Commun., 2012, 48, 7119–7130 RSC.
  27. Y. Kohno and H. Ohno, Phys. Chem. Chem. Phys., 2012, 14, 5063–5070 RSC.
  28. J. Albo, E. Santos, L. A. Neves, S. P. Simeonov, C. A. M. Afonso, J. G. Crespo and A. Irabien, Sep. Purif. Technol., 2012, 97, 26–33 CrossRef CAS.
  29. O. Nacham, K. D. Clark, H. L. Yu and J. L. Anderson, Chem. Mater., 2015, 27, 923–931 CrossRef CAS.
  30. M. Dobbelin, V. Jovanovski, I. Llarena, L. J. C. Marfil, G. Cabanero, J. Rodriguez and D. Mecerreyes, Polym. Chem., 2011, 2, 1275–1278 RSC.
  31. Y. M. Tang, X. L. Hu, P. Guan, X. P. Lin and X. Q. Li, J. Phys. Org. Chem., 2014, 27, 498–503 CrossRef CAS.

Footnote

Electronic supplementary information (ESI) available: Details of Experimental section, FT-IR spectra of ammonium/amidine salts, UV-vis spectra of MILs, photographs of LCST behaviour of N1,1,1,12[FeCl4]. See DOI: 10.1039/c6ra01235c

This journal is © The Royal Society of Chemistry 2016