Evaluating Water Miscible Deep Eutectic Solvents (DESs) and Ionic Liquids as Potential Lubricants

Although mineral oils are commonly used as lubricants their emission particularly in marine environments can cause significant impact. In the current study the properties of water miscible Deep Eutectic Solvents and ionic liquids are compared with a typical mineral base oil to ascertain their efficacy for potential marine lubricants. The environmental compatibility of some of the liquids, particularly choline chloride and glycerol, makes it an interesting potential base lubricant. Surprisingly some DESs showed very low corrosion rates with steel, nickel and aluminium even when the liquids contained water. This is a surprising result given that the chloride ion concentration is approximately 5 mol dm -3 .


Introduction
Lubricants are ubiquitous and perform a variety of functions such as cooling of surfaces, avoidance of corrosion, power transfer, offering a liquid seal at sliding contacts in addition to removal and suspension of wear products. Generally lubricating oils are formulated from a mixture of two or more base stocks (mineral oil, bio-based oil or synthetic oil) together with a variety of additives. 1 Most common additives which are used in lubrication are intended to perform several roles, some of them are used to protect the surface of sliding solids, while other are aimed to enhance the performance of the lubricant itself and others are used to maintain the lubricant's properties. Examples of common additives include dispersants, antioxidants, friction modifiers, anti-wear agents, detergents and viscosity index improvers. 2 While they generally have low toxicity on a small scale emission at sea in particular has significant environmental consequences. An additional issue for marine lubricants is the enhanced corrosion resulting from the incorporation of sea water in mineral oil based systems. 3 Nowadays, the development of new products from renewable bio-based materials has become an important potential in the area of lubricant manufacturing. Triglycerides derivatives from seed oils are valued as important alternatives to conventional petroleum based base stocks as they are sustainable, generally environmentally compatible, nontoxic, biodegradable as well as small changes in viscosity as a function of temperature.
Additionally, they are able to adhere to the metal surface giving them a great boundary lubrication role. Moreover, vegetable oils are readily able to dissolve additives and polar contaminants. However, their drawbacks are associated with their high sensitivity to oxidation, hydrolysis and high pour points. 4,5 There are, however several polyalkylene glycol based water miscible lubricants which are also available although they have tended to be used for specialist applications. 6 While most lubricants are non-polar and have low surface tension there have recently been a significant number of publications using ionic liquids as lubricants. 7 It has been shown that these neoteric liquids have viscosity indexes which are superior to many mineral base oils and comparable to many semi-synthetic base oils. They have also been studied as additives for a range of base fluids particularly where the ionic liquids have surfactant properties. Under extreme wear conditions some ionic liquids can break down to produce tribofilms which protect the surface. Naturally there are some environmental issues with selected ionic liquids and the complex synthesis means that the cost of production limits their use to small volume specialised applications. The numerous combinations of cations and anions allow the physical and mechanical properties to be tailored for the application. Their wide liquid range, lack of flammability, high thermal stability and non-volatility makes them interesting candidates as lubricants. 8 Where i corr is the corrosion current density, B is the Stern-Geary constant and Rp is the polarisation resistance. The value of icorr was obtained automatically from software and converted to corrosion rates by inserting equivalent weights and density values of each element in LSV and adjusted to electrode surface area (cm 2 ). For electrochemical impedance spectroscopy, EIS, values of (R P ) have been derived from semi-circular Nyquist plots and converted to i corr. using following equation Then, values are converted to corrosion rates in mm/year from this equation Where CR is the corrosion rate in mm/year, EW is the equivalent weight (g/mole) and ρ is the density of corroding sample g cm -3 Friction coefficient data were determined using the standard pin on disc technique using a Teer Coatings ST-200 wear tester. In all cases the friction coefficient -time profile remained flat over (2000 cycles) and 1 hour at 0.005 ms -1 under load of 30 N at 293±1 K.

Results and Discussion
Thermophysical Data: To be suitable as lubricants fluids need to retain an appropriate viscosity profile over a wide range of temperature. This is generally expressed through the viscosity index (VI) which is defined as; Where U is the kinematic viscosity at 40 o C of the unknown oil; L is the kinematic viscosity at It has previously been shown that the addition of quaternary ammonium salts lead to a decrease in the friction coefficient of fluids and the minimum friction occurs at the eutectic composition. 12 In the current study four DESs were tested at their eutectic composition using choline chloride as the quaternary ammonium salt together with different hydrogen bond donors. These are compared with four water-miscible, imidazolium-based ionic liquids as shown in Table 1.

Deep Eutectic Solvents
Ionic Liquids   Table 2 shows thermo-physical and corrosion data for four DESs, four ionic liquids and a standard mineral base oil. It should be highlighted that this study is comparing the properties of the ionic fluids to those of the base lubricating fluid. The properties of finished lubricating fluids are usually superior to those in Table 2 due to complex additive packages.
The viscosity index is usually calculated using a U-tube viscometer to measure the kinematic viscosity. This method is only valid for truly Newtonian fluids where the viscous stress is proportional to the local shear rate. It is known that some ionic liquids have some non-Newtonian character. 13 Most liquids in Table 2 showed Newtonian behaviour and good correlation was observed between the viscosity obtained from the Ubbelohde and rotating cylinder techniques. Reline and Oxaline showed non-Newtonian behaviour and so to give comparable results the average viscosity at a series of rotation rates was calculated. It should therefore be stressed that the VI data in   Table 2 shows the corrosion rates for mild steel in 4 DESs and four ionic liquids. The corrosion was studied using both linear sweep voltammetry and electrochemical impedance spectroscopy (EIS) and Table 2 shows that both had similar trends and magnitudes. It is well known that chloride based media can significantly enhance the corrosion rate in aqueous solution as it breaks down passive films. 16 As such it would seem logical that it should be avoided for lubricant applications. Table 2 shows the initial corrosion rates as determined using Tafel slopes from slow-rate linear sweep voltammetry. These are extremely low for Ethaline, Glyceline and Reline but extremely high for Oxaline. The origin of this difference could be due to the presence of an insulating film on the metal surface or the kinetics of either the anodic or cathodic processes. The Tafel slope shows that in Ethaline, Reline and Glyceline the cathodic slope is shallow suggesting that the cathodic process is rate limiting. In these liquids the cathodic process is the reduction of oxygen i.e.
The very high chloride concentration will decrease the activity of water as it is highly hydrogen bonded. In the oxalic acid based eutectic however the cathodic process is extremely fast as it is just the reduction of protons to form hydrogen gas allowing much faster corrosion.
Interestingly in this liquid bright yellow sheets of iron oxalate form parallel to the metal surface until all the metal is dissolved, which effectively filled the sample tube by the end of the experiment.
The very unusual aspect of DESs is their propensity to prevent corrosion even when the liquid is doctored with an electrolyte solution. Figure 1 shows that when 1 wt % water containing 3 wt % NaCl (a mimic of sea water) is added to the mineral oil significant 8 corrosion is observed of the mild steel immersed in the liquid after 2 weeks. Clearly the aqueous phase partitions to the steel surface where corrosion is caused by the relatively high chloride concentration. Figure 1 shows an analogous experiment where choline chloride is added to the aqueous solution in place of sodium chloride and it can be seen that corrosion occurs but to a lesser extent. This surprisingly shows that there is a cation effect to the corrosion mechanism which is not fully understood at present but may be related to the relative solubility of the corrosion products. Figure 1 also shows corrosive effect of Ethaline and Glyceline to which the same aqueous electrolytes were added in the same amounts. No visible signs of corrosion were detected on mild steel even after 6 months whereas in wet base oil corrosion was clearly visible within two days. This unusual observation shows that, counter to perceived wisdom, mild steel does not show signs of corrosion in a high chloride medium (approximately 5 mol dm -3 ). Table 3 shows the initial corrosion rate for iron, nickel and aluminium in four DESs as a function of water content. It can be seen that the glycol based liquids are relatively insensitive to the addition of water, whereas the urea and oxalic acid significantly increase the rate of corrosion for iron compared to an aqueous solution. Surprisingly aluminium only shows very slow corrosion when water is added even with a high chloride concentration presumably due to oxalate being able to passivate the metal surface. The formation of a passivating film can clearly be observed using a.c. impedance spectroscopy.
The low corrosion rate of common metals with glycol-based DESs even with significant water content suggests that they may be useful media as base lubricants particularly for marine applications.
Electrochemical a.c. impedance spectroscopy shows that while slow corrosion occurs initially and a passivating layer appears to form for a range of metals including Ni and Al.
Comparing the corrosion rates for the imidazolium based ionic liquids with those for the DESs and base oil shown in Table 2 it can be seen that the corrosion rate of mild steel in the most hydrophobic anion, ethylsulfate, is considerably higher than either hydrogen sulfate or SCN presumably due to its inability to bind traces of water. The acetate anion showed the lowest rate of corrosion. It appears to hold that ionic fluids with more Lewis basic anions tend to demonstrate enhanced corrosion resistance.    Table 4 shows that the DESs show lower friction coefficients than base oil with stainless steel, but slightly higher values for aluminium couples. This is thought to result from the relative abilities of the liquids to wet the different surfaces. This concept was tested by measuring the contact angle of the metals with different DESs. Clearly mineral oils with their low surface tension and low contact angles should be able to better wet surfaces but it depends to some extent on the nature of the surface. An indicator of the hydrophilicity of the surface activity of the liquid metal and Table 4 shows that aluminium has a low surface energy whereas iron based alloys have a high surface activity. It seems logical therefore that ionic liquids with high surface tensions and high contact angles should show improved friction coefficients over base oils with iron based alloys whereas the reverse is the case for aluminium and bronze alloys.

Conclusion
This study has shown that DESs, particularly those based on glycols together with choline chloride have the correct lubricity, toxicity and corrosivity to be viable options for base lubricants. Their ability to be miscible with water in all proportions and to suppress corrosion even when slightly wet means that they could be of particular use for marine lubricants.
Friction coefficient measurements suggest that they have improved tribological properties with respect to mineral oils for iron-based alloys which are clearly the important materials for lubrication whereas the converse is the case for more hydrophobic surfaces such as aluminium.