Introduction

Ionic liquids (ILs) have become increasingly of interest in engineering applications, including chemical processes,1, 2 CO2 capture,3, 4 lubrication,5,6,7 electrochemical devices8,9,10,11 and corrosion protection.12,13,14,15,16 Their desirable properties can include low volatility, chemical and electrochemical stability and they can be potentially more environmental friendly and biologically more compatible.17 All these make them ideal replacements for more traditional solvents or surface treatments. The interest in ILs from both the research and application perspective has seen enormous growth over the past decade or so with more than 30,000 research articles and over 4000 patents filed over the past 5 years. The increased use of ILs, as solvents and heat transfer fluids for example, has also seen research into their corrosiveness and the susceptibility of common engineering metals to corrosion by ionic liquids under various conditions has been investigated.18, 19 This has demonstrated that for some ionic liquid/metal combinations, more extensive corrosion can occur, whereas other combinations lead to passivation of the metal, as was demonstrated for magnesium alloys and discussed below.

Deep eutectic solvents (DES) have also recently been explored as potentially inexpensive media in the same context. Unlike conventional solvents which typically have volatility and may be chemically/electrochemically reactive, DES materials display many of the favourable properties of ILs. DES are prepared by combining a hydrogen bond donor and a hydrogen bond acceptor, which may in themselves be solid materials or volatile solvents, to produce a room temperature liquid, whereby essentially the same low volatility and chemical stability as in ILs can be achieved, albeit with a lower electrochemical window.20 One of the most common DES materials has been the combination of cholinium chloride with urea (see Fig. 1). In the context of this paper, this material has been used as a solvent for electroplating of Zn for example.21

Fig. 1
figure 1

Typical ionic liquids and DES systems used in passivation of reactive metals

Full size image

Figure 1 presents a summary of the structures and nomenclature of some of the important ionic liquid cations and anions (as well as DES) considered thus far. In this perspective, we will summarise the key literature pertaining to the use of ionic liquids in controlling the electrochemistry of reactive metals, including Li, Na and Mg, for energy storage and corrosion control. We will then consider the future directions of this important field with respect to engineering the interphases of reactive surfaces to optimise performance and discover new applications.

Protecting metal battery anodes via solid–electrolyte interphase (SEI) formation

Air and moisture stable ionic liquids have been utilised in energy storage research since their introduction as promising electrolytes for lithium metal batteries since the early 1990s.22,23,24,25,26 Prior to that, chloroaluminate-based ionic liquids combined with the imidazolium cation were extensively investigated by Wilkes as potential battery electrolytes, however, these presented challenges with respect to stability even in dry environments due to their intrinsic acidity.27,28,29,30 With the discovery of the pyrrolidinium bis(trifluoromethane) sulfonamide (TFSI) family of ionic liquids, a study described by Howlett et al. revealed that ionic liquids could successfully support >99% cycling efficiency of Li+/0 onto platinum and copper working electrodes without dendrite formation.31 Since then, ionic liquids have been intensively explored as candidates for safe battery electrolytes, providing different solvation of ions and large electrochemical stability.32, 33 For several decades, researchers have been seeking to understand the formation and control of a beneficial SEI in batteries, particularly in those studies involving highly reactive metals, such as lithium in contact with an electrolyte, Fig. 2a–d.34, 35 Ionic liquids seem particularly promising in this respect, particularly those containing fluorinated anions.

For example in Fig. 2e, f, g, very recently it was shown that in the IL C3mPyrFSI (i.e., N-methyl-N-propylpyrrolidinium bis(fluorosulfonyl)imide—Fig. 1), the pretreated SEI formed chemically prior to cycling has been characterised and shows long cycle lifetimes are possible in a full cell arrangement vs. LiFePO4.8 The SEI formed upon lithium metal in that cell arrangement is made up of LiF, LiOH, Li2S, LiSO2F and NSO− species. This was also hinted at in the earlier work by us.36 These compounds come about after the initial formation of LiF that produces an intermediate radical ion, SO 2 NSO 2 F −, which persists at the surface for adequate time to undergo further reduction.37 Evidence for ring opening of the cation is also observed within the interphase, which can be explained by a β-Hofmann elimination taking place.38 These products passivate the reactive lithium metal, limiting further corrosion of the anode while permitting the ingress and egress of the charge carrying Li+ during battery operation. The preparation of an SEI in this manner through chemical/electrochemical pathways is facile, and less expensive than recent physical techniques. Interestingly, these reactions are reminiscent of those that are proposed to occur on other reactive engineering alloys, such as magnesium and aluminium, where the reduction of the anion (either TFSI or organophosphate) and subsequent deposition of the metal salt on the surface leads to a passivating layer39, 40 as will be discussed in a later section.

An additional component to the formation of this invaluable interphase is the degree of adventitious moisture within the ionic liquid electrolyte. Recent reports have detailed the large role moisture can play in the formation of a beneficial SEI for both lithium41 and sodium metal anodes.42 Electrochemical characterisation of these interphases shows an increased polarisation during symmetrical cell cycling, which is attributed to a larger impedance of the electrode–electrolyte interphase. In case of passivation against corrosion of Mg or Al, low concentrations (0.1 wt%) of acid have also shown to be beneficial.40

Investigations of highly concentrated ionic liquid electrolytes (approaching 50 mol% salt, Fig. 2h) have revealed interesting results in line with reports for more conventional electrolytes.43 When using such highly concentrated lithium electrolytes composed of a LiFSI salt in a phosphonium-based ionic liquid, P111i4FSI, we have demonstrated that the resultant interphase permits an extensive amount of lithium to be plated and stripped at high current density for extended duration (up to 6 mAh cm−2).9, 44 This is also the case in the P111i4FSI45 and also the analogous P1i4i4i4FSI46 ionic liquid when using up to 50 mol% NaFSI. Initial electrochemical characterisation of the interphase formed in these systems has shown low electrode–electrolyte impedance. Work is underway to more extensively characterise this interphase, as it is seen as critical for controlling Na electrochemistry by halting corrosive side reactions while still allowing Na+ transport during battery operation.

Ionic liquid generated SEI layers are also critically relevant in the cycling of Mg anodes in ionic liquid-based Mg-metal battery electrolytes.47 In this case, it has recently become clear that small amounts of water in the electrolyte can play a more dominating role in the formation and properties of the SEI; the predominately MgO-type surface layer, being quite strongly passivating, rapidly quenches the activity of the anode during cycling.48 The recent work has shown that chemical removal of these residual amounts of water allows extended cycling of Mg in the IL electrolyte, with a very thin, more IL-dominated SEI layer probably allowing transport of Mg to and from the [NTf2]−-based IL electrolyte. On the other hand, when FSI was used as the anion, passivation was observed and X-ray powder diffraction revealed the formation of MgF2.47

The same behaviour has been demonstrated for Mg primary batteries where a phosphonium chloride ionic liquid, with either water or ethylene glycol additives, can sustain controlled discharge when a current is drawn but otherwise passivates the magnesium surface, thus preventing corrosion and self-discharge, through the formation of a complex magnesium chloride gel in which the phosphonium IL cation is also contained.49, 50

The work described above demonstrates that the effect of impurities/additives on the surface electrochemistry and the interphase layer on reactive metals is an ongoing research challenge and possibly a practical opportunity.40

Fig. 2
figure 2

Recent developments in the preparation of stable interphases at reactive metal electrodes. a Schematic diagram of the ingress/egress of lithium through a hollow carbon nanosphere scaffold SEI. Scanning electron micrographs of the scaffold before (b) and after (c) lithium plating and stripping and d cross-section after lithiation. e Schematic diagram of time-dependant SEI formation via facile chemical treatment prior to cell fabrication. f Cycling behaviour of Li|LiFePO4 cells comparing lithium anodes, pretreated SEI surface (red boxes) vs. a pristine foil (blue boxes). g Cross-section micrograph of the ionic liquid pretreated lithium foil after cycling shown in f. h Voltage–time profile for a Li|Li cell when using highly concentrated ionic liquids at high current density exhibiting low polarisation (0.1 V, 3 mAh cm−2) and micrographs of the resultant surface morphology (i) and cross-section (j). Figure reproduced/adapted with permission from: a, b, c, d, ref. 35 Macmillan Publishers Ltd; e, f, g, ref. 8 Macmillan Publishers Ltd; h, i, j, ref. 9, Elsevier Ltd.

Full size image

Ionic liquid-derived protective/passivating surface films on Mg and Al alloys

As a protective surface treatment, ionic liquids have been investigated predominantly for magnesium, but also on aluminium alloys.13, 16, 51 The reader is referred to a review by Huang et al. in 2013, which summarises the state of the art at that point. Briefly, following on from the SEI formation on reactive metals, such as lithium, it was shown that immersion of Mg alloys in an anionic liquid containing TFSI as the anion also led to the formation of a protective surface film. A combination of surface techniques indicated that the films are likely composed of a Mg (or Al) salt containing the breakdown of the IL anion, with IL cation somehow entrained in this film. Various parameters have been investigated, such as ionic liquid structure, concentration, the effect of dilution in a solvent, temperature, bias voltage and surface pickling. Many of these treatments are employed to homogenise the surface and encourage a uniform reaction layer between the IL and the surface to increase the level of protection. While the fluoride salts that formed on a Mg surface immersed in a TFSI IL offered almost a 100-fold protection against corrosion in a chloride containing solution, it was quickly shown that ionic liquids containing the phosphate or phosphonate anion were significantly more protective, with thin compact protective films less than 100 nm forming on the surface under certain controlled conditions (see Fig. 3).13, 14, 16 Interestingly, the uniformity of these films is enhanced by activation of the surface, either by using a potential bias (preferably anodic) or by heating the surface to between 50 and 100 °C.12, 52 In both activated cases, the application of a potential or high temperature renders the alloy surface more reactive and thus metal oxidation occurs more rapidly, releasing for example Mg2+ ions near the surface, which then react with the nearby anions forming a surface film. This appears to be a more uniform process under the activated conditions, rather than metal oxidation and hence subsequent film deposition occurring at localised grain boundaries or next to secondary phases/intermetallic particles.12 The surface properties of this film are significantly modified as can be seen from Fig. 3c, which shows the contact angle decreasing significantly as a more uniform protective interphase forms (Fig. 3).

Fig. 3
figure 3

a Scanning electron microscope (SEM) image of a FIB cross-section specimen of AZ31 treated with P1444dpp b polarisation date for same AZ31 alloy immersed in a simulated body fluid c contact angle for water on bare AZ31 and surface immersed in P1444dpp at 50 and 100 °C. Thicker deposit can be seen near the intermetallic. The polarisation data shows two orders of magnitude reduction in anodic and cathodic current densities. Figure reproduced/adapted with permission from: a, b ref. 16, Elsevier Ltd. c Reprinted with permission from ACS Applied Materials and Interfaces, 2014. 6(21): p. 18989–18997. Copyright 2014 American Chemical Society ref. 52.

Full size image
Fig. 4
figure 4

Performance of ILs as corrosion and microbial inhibitors. a Potentiodynamic polarisation results for mild steel in 0.01 M NaCl at pH 2 after 30 min with and without 4 mM of 2-methylimidazolinium 4-hydroxycinnamate. SEM images of mild steel immersed for 24 h in 0.01 M NaCl at pH 2 b control and c with 4 mM 2-MeHImn 4-OHCin. d Shows the percentage survival of the bacteria Vibrio alginolyticus in the presence of cetrimonium nalidixate. a, b, c Reprinted with permission from ACS Sustainable Chemistry & Engineering, 2016. 4(3): p. 1746–1755. Copyright 2016 American Chemical Society ref. 58. d Figure reproduced/adapted with permission from ref. 60, Royal Society of Chemistry.

Full size image

Ionic liquid corrosion inhibitors and antimicrobials for steel

Ionic liquids have also been used as corrosion inhibitors to form a protective surface film on metal surfaces, predominantly for mild steel in acidic solutions, but also for a range of other metals, such as aluminium and magnesium.15, 53,54,55,56,57 Seen as an environmentally friendlier alternative to traditional corrosion inhibitors, particularly those that included the highly effective but highly toxic chromium VI, ionic liquids have been used in solution and as a surface treatment prior to immersion in an aggressive solution. In solution, certain structural properties have been identified that enhance the surface protection offered by these compounds. The presence of π electrons from double, triple or aromatic bonds and free electrons from elements, such as nitrogen, encourage adsorption at the surface, displacing water to form a protective film.15, 54 Consequently, for steel and aluminium, imidazolium cations have been most widely used, some even with benzyl groups attached to increase adsorption. Extended alkyl chain length has also been proposed to increase the strength of adsorption, either on the imidazolium or through the use of quaternary ammoniums.55

While many of these studies have shown the effect of the cation structure, the effect of the anion structure does not appear to be as clear. Many of the ILs investigated consisted of a simple anion, such as chloride or bromide. When compared to more complex anion structures, such as bis(trifluoromethanesulfonyl)amide, the simpler anions have often shown a comparable or better performance.53, 56, 57 More recently Chong et al. showed a synergistic effect in the performance of an imidazolinium cinnamate, which at a pH of 2 clearly outperformed the imidazolinium chloride and the sodium cinnamate in protecting a mild steel surface (Fig. 4a–c).58 This type of material is being referred to as a dual active IL, in which both the cation and anion contribute to the performance. In fact the imidazolinium cinnamate was designed to combat microbially influenced corrosion (MIC), with the imidazolinium cation proposed to inhibit microbial growth, while the cinnamate inhibits corrosion. Since a number of the compounds likely to inhibit microbial growth are similar in structure to corrosion inhibitors, such as quaternary ammoniums and imidazoles, dual active ILs for MIC inhibition have been proposed. Seter et al. investigated a range of ILs as MIC inhibitors. It was found that cetrimonium nalidixate was the most effective inhibitor of microbial growth (Fig. 4d) and was able to prevent both biofilm formation and the initiation of pitting corrosion.59, 60 Thus, the prospect of designing ionic liquids (or indeed organic salts which are based on similar ion chemistry and structure) to target both corrosion inhibition and MIC as dual actives is an area ripe for further investigation, given the diversity of chemistry available and the literally millions of combinations of anion and cation that are possible.

Future prospects

Thus far, ionic liquids have clearly demonstrated their potential for controlling electrochemical reactions on either engineering alloys or battery anode materials, although understanding the chemistry of the interphases is still in its infancy. That there is a role for the anion chemistry in particular, and the presence of impurities or the deliberate addition of reactive small components such as water, halides or acid, is evident from the research thus far. However, the precise chemical mechanisms are far from well understood in order to enable strategic design of interphases. The use of advanced surface characterisation methods to understand the composition and morphology of these interphases and their relationship to IL composition and processing conditions will help develop this field further. Specifically, the development of methodologies to observe the formation and propagation of these films in situ is a key area of research required to drive IL design for interphase control.

Furthermore, the possibility of designing dual active ILs that can protect against corrosion and antifouling is also attractive. The scope for new chemistry in this case is only just emerging and the DES systems may offer an additional, more economical alternative in this space.

Finally, little work has been done on other reactive surfaces, such as Cu, Zn or ZincAlum (of key importance with respect to galvanised steel for example). Designing task-specific ILs to create a unique protective surface for a given metal or metal alloy or indeed to enable new battery chemistries offer exciting possibilities. It is hoped that the ideas in this perspective will trigger further research in chemical design, mechanistic understanding and novel surface modification to engineer interphases using novel IL and DES materials.