**5.2. Ionic-liquid based ferrofluids**

454 Smart Actuation and Sensing Systems – Recent Advances and Future Challenges

conventional MRFs, and high enough for practical purposes.

the fundamental point of view too.

suspensions with [EtMeIm][EtOSO3] as liquid carrier.

[BuMeIm][PF6]. As a matter of fact, this MRF only exhibited a sedimentation ratio of 0.95 after two months of preparation. Having prepared a highly stable IL-based MRF was important, but even more important was the fact that colloidal stability was achieved without the addition of any stabilizing agents. The enhancement of stability was attributed to the physical adsorption of the IL ions on magnetite surface (for which they had strong affinity) that gave rise to steric repulsion between the dispersed particles. In the same work, the rheological behaviour of the samples upon magnetic field application was also studied concluding that the MR effect of the IL-based MRFs was comparable to those obtained for

In a different work, Guerrero-Sánchez et al. (Guerrero-Sánchez et al., 2009) also used ILbased MRFs to study the influence of temperature on their MR behaviour. Here, ILs allowed extending such study to a broader range of temperature than in the case of conventional MRFs, due to their thermal stability and low vapour pressure. Something similar was pointed out by Dodbiba et al. (Dodbiba et al., 2007) who also took advantage of the good performance of ILs at high temperature to study the rheological behaviour upon magnetic field application of a mixture of two types of carbonyl iron powders dispersed in an IL. Therefore, these works make evident that the preparation of IL-based MFs is important from

Our group has also studied IL-based MRFs, especially for what concerns their stability. With this aim, Gómez-Ramírez et al. (Gómez-Ramírez et al., 2011) prepared IL-based MRFs differing both in the disperse phase material (pure iron and silica-coated iron particles) and in the liquid carrier ([EtMeIm][EtOSO3] and ([EtMeIm][Et2OPO3]). Additionally, conventional MRFs with mineral oil as liquid carrier were also synthesized. The particle volume fraction of all the samples was kept as 10 vol %. The most stable MRF was obtained by the combination of silicacoated iron particles and [EtMeIm][Et2OPO3]; for example, the viscosity of such MRF for shear rates over 100 s-1 almost coincided with the theoretical prediction of Batchelor equation, what means that only two-body hydrodynamic interactions existed between the particles, and, therefore, almost no magnetic aggregates were present. In fact, microscopic observations of the suspensions upon magnetic field application showed that field-induced particle chains had a more regular structure in the case of the most stable MRF (see figure 7). The worst results were obtained for mineral oil-based MRFs in which the particles were strongly aggregated and the resulting structures became more irregular. An intermediate behaviour was found for the

**Figure 7.** Field-induced particle structures in MRFs consisting of silica-coated iron microparticles

dispersed in [EtMeIm][Et2OPO3] (a) and mineral oil (b). Bar length: 200 μm.

As it has been pointed out above, one of the advantages of using ILs as carriers is the stability that the physical adsorption of their constituent ions provides with. In the case of MRFs it has been seen that this improvement may be good enough for practical purposes. But what happens in the case of FFs that, by definition, need to be ultrastable? What are the mechanisms involved in the stabilization of IL-based FFs?

The first reference to IL-based FFs appeared in 2009, when Oliveira et al. prepared suspensions of bare maghemite (γ-Fe2O3) and cobalt ferrite (CoFe2O4) nanoparticles dispersed in [BuMeIm][BF4] (Oliveira et al., 2009). Such suspensions were said to be stable in the presence of an external magnetic field, even without the addition of stabilizing agents. The authors attributed such stability to the formation of a semi-organized protective layer by the IL ions around the dispersed particles. Nevertheless, when water was added to the suspensions, they became completely unstable, apparently due to the destruction of such a protective layer. The authors also tried to prepare suspensions using hydrophobic ILs (namely [BuMeIm][PF6] and [BuMeIm][Tf2N]), but these suspensions were not stable at all.

Jain et al. (Jain et al., 2011) also attempted to synthesize IL-based FFs using maghemite nanoparticles, but with particle concentrations much higher than in the case of Oliveira et al. They succeed when using [EtMeIm][Ac] and [EtMeIm][SCN] as carriers, without needing any stabilizers, but they did not in the case of the protic IL [EtN][NO3] and [BuMeIm][BF4]. In the latter case, they failed even at lower concentrations in contrast with the results of Oliveira et al. As a matter of fact, there is still some controversy related to the stabilization of bare magnetic nanoparticles in IL media, although the stabilization provided by the IL ions has been pointed out to work quite well for metal nanoparticles in general (Pârvulescu et al., 2007). This is the case of positively charged Pt nanoparticles dispersed in [BuMeIm][BF4] and

[BuMeIm][PF6], for example (Scheeren et al., 2006). However, in some other cases it is not high enough, and additional stabilization is needed (Pârvulescu et al., 2007; Torimoto et al., 2010).

In most of the suspensions mentioned above, the material for the dispersed phase is maghemite. At this point it is important to mention that both Guerrero-Sánchez et al. (Guerrero-Sánchez et al., 2007) and Oliveira et al. (Oliveira et al., 2009) failed to synthesize ultrastable IL-based suspensions when using magnetite (Fe3O4) nanoparticles. [BuMeIm][BF4] that, according to Oliveira et al., had given rise to good results for maghemite and cobalt ferrite, did not work at all in the case of magnetite. This could be due to the fact that magnetic interaction between magnetite nanoparticles is higher than in the case of maghemite ones, since the saturation magnetization of the first (90 Am2kg-1) is higher than the one of the latter (80 Am2kg-1). Given that both iron oxide surfaces must be composed of the same Fe-OH groups, the formation of the protective layer around magnetite nanoparticles could have taken place, but probably was not protective enough to overcome the magnetic attraction between the particles. An additional stabilization mechanism seemed to be needed.

In 2011 our group reported a true IL-based magnetite FF (Rodríguez-Arco et al., 2011a). In this work, emphasis was made in the fact that only strong steric repulsion was useful to obtain an ultrastable suspension of magnetite in [EtMeIm][EtOSO3]. This was achieved by adsorbing surfactants with long tails on the surface of magnetite nanoparticles that gave rise to such strong steric repulsion (more details about this IL-based FF are given below). In fact, we also tried to prepare suspensions of bare magnetite nanoparticles in the same IL and failed. Later, something similar was pointed out by Jain et al., who observed that bare maghemite nanoparticles were unstable in the protic IL [EtN][NO3], but they became highly stable (even at highly particle concentration) when coated by a layer of short acrylic-acid-*b*acrylamide copolymer (Jain et al., 2012). As a result, steric repulsion seems to be the best stabilization mechanism for the preparation of IL-based FFs.

In view of the results presented above a question may arise: how can we explain that in some cases stabilization is ensured without the addition of stabilizing agents (i.e. surfactants) while in others it is strictly necessary? Here, it is particularly interesting to mention the work by Ueno et al. in which they prepared suspensions of bare and PMMAgrafted silica particles dispersed in imidazolium-based ILs (Ueno et al., 2008). Given that ILs are non-volatile, the *in situ* behaviour of these suspensions was studied by means of transmission electron microscopy (i.e. no escapes of the liquid to the vacuum system took place). They observed that strong aggregation, which appeared for bare silica particles, was almost absent in the case of PMMA-grafted ones (figure 8). They concluded that electrostatic repulsion, that was the only possible stabilization mechanism in the case of bare silica, did not work. They attributed this to the screening effect of the high ionic atmosphere surrounding the particles, since ILs are actually room-temperature molten salts. They also corroborated such affirmation by estimating the interaction potential energy between the dispersed particles and checking that the electrostatic repulsion component was negligible when compared with the steric repulsion one (Ueno et al., 2008).

stabilization mechanism for the preparation of IL-based FFs.

when compared with the steric repulsion one (Ueno et al., 2008).

2010).

mechanism seemed to be needed.

[BuMeIm][PF6], for example (Scheeren et al., 2006). However, in some other cases it is not high enough, and additional stabilization is needed (Pârvulescu et al., 2007; Torimoto et al.,

In most of the suspensions mentioned above, the material for the dispersed phase is maghemite. At this point it is important to mention that both Guerrero-Sánchez et al. (Guerrero-Sánchez et al., 2007) and Oliveira et al. (Oliveira et al., 2009) failed to synthesize ultrastable IL-based suspensions when using magnetite (Fe3O4) nanoparticles. [BuMeIm][BF4] that, according to Oliveira et al., had given rise to good results for maghemite and cobalt ferrite, did not work at all in the case of magnetite. This could be due to the fact that magnetic interaction between magnetite nanoparticles is higher than in the case of maghemite ones, since the saturation magnetization of the first (90 Am2kg-1) is higher than the one of the latter (80 Am2kg-1). Given that both iron oxide surfaces must be composed of the same Fe-OH groups, the formation of the protective layer around magnetite nanoparticles could have taken place, but probably was not protective enough to overcome the magnetic attraction between the particles. An additional stabilization

In 2011 our group reported a true IL-based magnetite FF (Rodríguez-Arco et al., 2011a). In this work, emphasis was made in the fact that only strong steric repulsion was useful to obtain an ultrastable suspension of magnetite in [EtMeIm][EtOSO3]. This was achieved by adsorbing surfactants with long tails on the surface of magnetite nanoparticles that gave rise to such strong steric repulsion (more details about this IL-based FF are given below). In fact, we also tried to prepare suspensions of bare magnetite nanoparticles in the same IL and failed. Later, something similar was pointed out by Jain et al., who observed that bare maghemite nanoparticles were unstable in the protic IL [EtN][NO3], but they became highly stable (even at highly particle concentration) when coated by a layer of short acrylic-acid-*b*acrylamide copolymer (Jain et al., 2012). As a result, steric repulsion seems to be the best

In view of the results presented above a question may arise: how can we explain that in some cases stabilization is ensured without the addition of stabilizing agents (i.e. surfactants) while in others it is strictly necessary? Here, it is particularly interesting to mention the work by Ueno et al. in which they prepared suspensions of bare and PMMAgrafted silica particles dispersed in imidazolium-based ILs (Ueno et al., 2008). Given that ILs are non-volatile, the *in situ* behaviour of these suspensions was studied by means of transmission electron microscopy (i.e. no escapes of the liquid to the vacuum system took place). They observed that strong aggregation, which appeared for bare silica particles, was almost absent in the case of PMMA-grafted ones (figure 8). They concluded that electrostatic repulsion, that was the only possible stabilization mechanism in the case of bare silica, did not work. They attributed this to the screening effect of the high ionic atmosphere surrounding the particles, since ILs are actually room-temperature molten salts. They also corroborated such affirmation by estimating the interaction potential energy between the dispersed particles and checking that the electrostatic repulsion component was negligible

**Figure 8.** In situ TEM picture of dilute suspensions of bare (a) and PMMA-grafted (b) silica particles in [EtMeIm][Tf2N]. Reprinted (adapted) with permission from (Ueno et al., 2008). Copyright (2008) American Chemical Society.

If the results of Ueno et al. (Ueno et al., 2008, 2009) are taken into account, there are two points that seem clear: (i) electrostatic repulsion does not work in IL media because of the highly ionic environment; (ii) only steric repulsion gives rise to long-term stability. On its part, the protective layers formed by the adsorption of the IL ions (i.e. physically or by coordination compounds) could provide with steric repulsion. However, in some cases (as for magnetite suspensions), it is not strong enough to ensure long-term stability and other additives (i.e. surfactants) have to be used.

As it has been pointed out above, our group succeeded in the preparation of true magnetite IL-based FFs. This was only achieved when the surface of the particles was coated with some stabilizing agents, but not all additives worked (Rodríguez-Arco et al., 2011b). For example, citric acid, that had been previously used to stabilize water-based FFs, did not work properly in [EtMeIm][EtOSO3], since the samples consisting of citric acid-coated magnetite slightly settled when subjected to a magnetic field gradient of 10 mTmm-1, and gave rise to strong phase separation after 5 min of centrifugation at 8000*g.* This was due to the fact that citric acid molecule is very short and, therefore, the resulting steric repulsion was quite weak. As a consequence, it seems that only long enough molecules should be used.

In addition, compatibility between the tails of the surfactant and every particular IL must exist. For example, if magnetite particles were just coated with a single layer of oleic acid, the resulting suspension was as unstable as those consisting of bare magnetite. Nevertheless, if a second layer of oleic acid molecules was formed when an excess of oleic acid was added, a true IL-based FF was obtained. The ultrastability of such FF was accurately proved, since it did not settle either upon magnetic field gradient application or after 30 min of centrifugation at 8000*g*. The differences between both particle coatings can be easily observed in figure 9.

It can be seen that the tails facing the IL in figure 9a are non-polar, while the situation is just the opposite when the oleic acid-double layer is formed (figure 9b). In the first case, there is not compatibility between the surfactant tails and the IL (highly polar), whereas in the second one there is. As it has been previously pointed out, Jain et al. (Jain et al., 2012) were

able to stabilize maghemite nanoparticles in [EtN][NO3] by coating them by a layer of short acrylic-acid-*b*-acrylamide copolymer; however, they were not in the case of [BuMeIm][BF4] and [EtMeIm][SCN] just because the polyacrylamide block of the acrylic-acid-*b*-acrylamide copolymer is not soluble in these ILs, and therefore, no compatibility exists. Therefore, as it has been said above, the best way to ensure long-term stability is the use of surfactants with long enough, carrier liquid-compatible tails adsorbed on the surface of magnetite nanoparticles.

**Figure 9.** Monolayer of oleic acid molecules adsorbed on the surface of magnetite nanoparticles (a). Oleic acid-double layer formed when adding an excess of oleic acid (b).

Finally, some results about the rheological behaviour upon magnetic field of the samples mentioned above are shown (Rodríguez-Arco et al., 2011b). In particular, figure 10 shows the yield stress obtained for all of them.

**Figure 10.** Yield stress for IL-based suspensions consisting of bare (triangles), citric acid-coated (circles) and oleic acid-double layer-coated (squares) magnetite. Full and open squares correspond to this latter sample before and after centrifugation respectively (see text).

With regards to the results presented in figure 10 it is important to remind that, theoretically, a true FF should not display considerable yield stresses, since its response to magnetic field is too weak for this to happen. However, in the case of the suspension of bare magnetite, the yield stress is quite high, likely due to the strong particle aggregation, and therefore, to the formation of field-induced structures by the aggregates. When the particles are coated by citric acid, the aggregation degree decreases, and so does the yield stress. In the case of the oleic acid-double layer two different situations appear. The first one corresponds to the original sample, which displays similar values (a little bit lower) than the sample of citric acid-coated magnetite. However, when this sample is subjected to centrifugation, although the supernatant itself is still magnetic, the yield stress of the supernatant becomes negligible. This can be explained because particle aggregates that could be present in the original sample were removed by centrifugation. Similar results were found for the MV effect (Rodríguez-Arco et al., 2011b). In fact, the MV response of these new IL-based MFs was theoretically fitted by a model that was originally proposed by Zubarev et al. (Zubarev et al., 2005) for oil-based FFs, showing that the MV effect increases with both the volume of the largest particles and their concentration and when the distance between the magnetic cores of the particles decreases.

In conclusion, it could be said that much work is being done in the field of IL-based FFs. Thanks to such studies our understanding of the phenomena involving them (i.e. stability) is improving, and it is very likely that better IL-based FFs will be prepared in the future. However, more applied studies are needed, which, for example, analyze their real performance in technological devices like those of section 3.
