1 Introduction

Since the beginning of the present century [1] the research on the lubrication performance of ionic liquids (ILs) has grown exponentially. During the past decade, most studies were focused on aprotic ionic liquids (APILs) containing mainly imidazolium, quaternary ammonium, pyridinium, or phosphonium cations combined with halogenated, especially fluorinated anions, both as base lubricants, and as additives in mineral or synthetic oils, for a variety of sliding contact conditions and tribopair materials [2,3,4,5,6,7,8,9,10]. New nanolubricants have also been recently developed by the combination of ILs and nanomaterials, particularly carbon nanophases [11,12,13].

In spite of their outstanding tribological performance, one major inconvenience of the use of these APILs is the possible tribochemical processes which can give rise, upon degradation, to the evolution of toxic and corrosive species [14].

This has lead to the study of the tribological performance of new halogen-free APILs [15,16,17,18] and, finally, of fully organic heteroatom-free ionic liquids, including those derived from natural resources [19, 20].

Halogen-free protic ionic liquids (PILs) [21] have been described as more environmentally friendly than APILs, and could even be biocompatible [22, 23].

Protic ammonium carboxylate species are readily available from the corresponding amines and carboxylic acids, by a direct neutralization reaction, which produces the new PILs [24, 25]. These protic ammonium carboxylate salts have been used as neat lubricants and as additives in mineral oils [26, 27].

Those PILs with short alkyl chains are soluble in water and highly polar solvents, and can be used as base lubricants or as additives in aqueous lubrication. We have shown the good tribological performance of protic ammonium carboxylate PILs synthesized from bio-based carboxylic acids in copper–copper lubrication [26, 27], as base fluid for graphene dispersed lubricant [11,12,13, 28], and as additives in new epoxy resin porous self-lubricating polymers [29].

The issue of water content as an impurity, adsorption of water and interactions between water and IL molecules, and their consequences for practical applications of ILs, including the tribological performance of ILs is attracting much attention [30,31,32,33].

Early studies on IL lubricants addressed the use of imidazolium fluorine containing ILs as additives in water for the lubrication of ceramic–ceramic contacts [34, 35], and their effect on the reduction of the running-in high friction period.

Precedents of the use of ammonium carboxylate PILs as additives in water have shown that, under sliding conditions, evaporation of water takes place after a certain sliding distance. When neat water is used, the evaporation produces a transition from lubricated to dry contact, with the corresponding increase in coefficient of friction and wear rate, and the severe oxidation of the wear track on steel.

When di[2-(hydroxyethyl)ammonium] succinate (MSu) is present as additive, evaporation of water leads to a transition to a boundary regime with ultralow coefficient of friction [36], but (Water + PIL) lubricants still present a high friction and wear running-in period before water evaporation takes place at the interface.

Mono- and bis[2-(hydroxyethyl)] ammonium PILs containing carboxylate anions derived from long-chain fatty acids have been very recently used [37] as additives to reduce the high friction coefficient of water, in the form of emulsions, as they are not soluble in water.

In order to eliminate the running-in period in water + short alkylchain protic ammonium carboxylate, a boundary layer of PIL was obtained for tri[bis-(2-hydroxyethyl)ammonium] citrate (DCi), composed of three diprotic ammonium cations and a tricarboxylic anion, under mild static conditions [38]. This method provides a friction reduction and antiwear performance similar or superior to that of the neat PIL. DCi has also shown its ability to disperse graphene in 0.1 wt% concentration, to obtain a new nanolubricant which prevents wear volume loss from steel. However, the DCi + 0.1 wt% graphene dispersion changes the steel surface topography and composition by the deposit of a graphene-containing layer on the sliding track. This mechanism has been also described for APILs + graphene lubricants [11, 39].

In the present study, we describe the results of using a new PIL composed of two diprotic bis(2-hydroxyethyl)ammonium cations, and the dicarboxylic succinate anion as additive in water for fluid film and thin film lubrication, and as base lubricant for graphene dispersions with different graphene concentrations. The main goals were to reduce or eliminate high friction coefficient during running-in, and to protect the steel surface from wear and tribocorrosion processes.

2 Experimental

The protic ionic liquid di[bis(2-hydroxyethyl)ammonium] succinate (DSu) (Fig. 1) was synthesized from 2-hydroxyethylamine and succinic acid [24, 25] by Dr. Iglesias (Federal University of Bahia, Brazil) and used as received.

Fig. 1
figure 1

Di[bis(2-hydroxyethyl)ammonium] succinate (DSu)

Full size image

1–10 layer graphene (> 99% purity; thickness 0.6–3.8 nm; specific surface area 500–1000 m2/g) was purchased from Iolitec (Germany). DSu + 0.1 wt% graphene (DSu + 0.1G) and DSu + 0.05 wt% graphene (DSu + 0.05G) dispersions were obtained following the previously described [38] procedure of manual mechanical milling in an agate mortar for a period of 10 min, followed by ultrasonication for 30 min. After completely covering the AISI 316L surface with Water + DSu, DSu thin layers (with a thickness < 10 µm, as determined from profilometry measurements) were obtained by controlled evaporation of water using previously described conditions [38], in a vacuum oven at 60 °C for 3 h and cooling to room temperature in a desiccator.

Thermogravimetric analysis (TGA) was recorded from room temperature to 600 °C at a heating rate of 10 °C/min under a N2 atmosphere (40 mL/min). Degradation temperatures (for a 50% weight loss) were 297.0 °C for DSu, 298.5 °C for DSu + 0.05G, and 311.8 °C for DSu + 0.1G.

Viscosity values (Table 1) were determined using a rotational rheometer (AR-G2, TA Instruments) (USA) following the described procedure [40].

Table 1 Viscosity variation with temperature for DSu and DSu + graphene lubricants (average values of three tests; standard deviation values in parenthesis)
Full size table

Contact angles of the lubricants on AISI 316L surface (Table 2) were determined using a DSA30B equipment (Krüss, Germany).

Table 2 Contact angles (average values of three tests; standard deviation values in parenthesis)
Full size table

A pin-on-disk tribometer (ISC 200PC, USA) was used for tribological tests. 0.75 mm sphere radius 99.9% Al2O3 balls (Goodfellow, UK) with a hardness of 2750 HV, a Young modulus of 445 GPa and a Poisson’s ratio of 0.27, and AISI 316L stainless steel disks (25 mm diameter; 5 mm thickness; average surface roughness, Ra = 0.05 µm) with 200 HV hardness, a Young modulus of 197 GPa and a Poisson’s ratio of 0.27 were used as sliding materials under a normal load of 0.98 N (1.3 GPa mean contact pressure) and a sliding velocity of 0.10 ms−1, for a sliding distance of 1500 m, at ambient conditions [23(± 2) °C; 50(± 5)% relative humidity]. Lubricants were added in an approximate volume of 0.5 mL to completely cover the steel surface before each test. Samples were cleaned with water and ethanol and dried in air.

Raman spectra were obtained using a Nicolet Almage spectrometer (Thermo Electron) and an Olympus microscope. A 514 nm laser was used. SEM and TEM microscopy, XPS analysis and optical profilometry equipments and conditions have been previously described [38]. Electron microscopy images were obtained with a Hitachi S3500N scanning electron microscope (SEM) and with a JEOL JEM 2100 transmission electron microscope (TEM). Surface analysis (XPS) was performed by means of an ESCA5701 Physical Electronics (PHI), taking the C1s binding energy at 285.0 eV as reference. Surface roughness, surface topography images, and wear volumes were determined with a Talysurf CLI profilometer.

Graphene from DSu + graphene dispersions before and after tribological tests were obtained for TEM observation (on a copper grid) by centrifugation at 4000 rpm for 3 min, and washing with ethanol. This procedure was repeated 5 times, before drying the particles in an oven at 100 °C for 24 h and cooling them at room temperature in a desiccator.

3 Results and Discussion

3.1 Properties of the Lubricants

The dispersion of 0.1 wt% graphene in DSu (DSu + 0.1G) increases viscosity by a 2.4 factor with respect to DSu (Table 1). This strong increase anticipates a possible increase of the friction coefficient with respect to neat DSu in the full-fluid lubrication regime, and a possible decrease of friction in the thin film lubrication regime.

The thermal stability (Fig. 2a, b) of DSu + 0.5G dispersion is similar to that of neat DSu, while that of DSu + 0.1G is slightly higher. Degradation temperatures values for 50% weight loss are 297.0 °C for DSu and 298.5 °C for DSu + 0.05%G. A significant increase to 311.8 °C id found only for DSu + 0.1%G.

Fig. 2
figure 2

a TGA curves for DSu and DSu + graphene dispersions. b Detail of the region around degradation temperatures for 50 wt% loss

Full size image

The wettability of the different fluids on stainless steel has been studied by contact angle measurements (Table 2). In all cases, initial contact angles change with time until they reach constant values after an interval of 5 min.

While neat DSu shows lower wettability than water, both graphene dispersions show higher contact angles after 5 min than DSu. The presence of graphene nanoplatelets could reduce the interactions between DSu molecules and the steel surface, but at the same time, could improve the load-carrying ability of the lubricants.

Table 2 shows the poor wettability of neat DSu on the stainless steel surface, with a very high initial contact angle (89.87°), 48% higher than the contact angle of water on the same surface. After 5 min, once the steady state angle is reached, the interaction between DSu molecules and steel surface produces a 37% reduction in the contact angle of DSu with respect to the initial value. Addition of 1 wt% DSu to water increases the final contact angle after 5 min with respect to neat water but, as expected, the value is closer to that of water than to that of neat DSu, due to the low concentration of PIL.

3.2 DSu as Additive in Water Lubrication

3.2.1 Friction Coefficients and Wear Rates

Table 3 and Fig. 3 compare, respectively, friction coefficient mean values and evolution of friction coefficients with sliding distance for neat DSu, Water + DSu and DSu thin film lubricants. Neat DSu shows a relatively poor performance as neat lubricant with respect to other PILs such as tri[bis(2-hydroxyethyl)ammonium] citrate (DCi) [38], probably due to its lower viscosity and its lower load-carrying ability (Table 1), which could be related to the lower number of ammonium, hydroxyl, and carboxylate polar groups present in DSu with respect to DCi.

Table 3 Coefficients of friction and wear rates (average values of three tests; standard deviation values in parenthesis)
Full size table
Fig. 3
figure 3

Variation of coefficients of friction (COF) with sliding distance for neat DSu, Water + 1 wt% DSu and for the thin film on stainless steel, obtained from Water + 1 wt% DSu after water evaporation

Full size image

Table 3 shows that DSu reduces the coefficient of friction with respect to Water + DSu, but the wear rates are of the same order of magnitude for both lubricants. The best performance is obtained for the thin film, which affords a 45% reduction of the friction coefficient and a reduction in wear rate of two orders of magnitude with respect to the full-fluid Water + DSu lubricant.

A better understanding of the performance of each fluid can be attained from Fig. 3. The high friction coefficients and wear rates obtained for Water + DSu and neat DSu are due to the high friction values during the long running-in periods. Thus, Water + DSu shows a friction coefficient of up to 0.26 (mean friction coefficient 0.236 (± 0.011) from 0 to 800 m), while neat DSu shows an initial friction coefficient of 0.24, which gradually reduces to 0.09 after 330 m.

In contrast, the thin film DSu lubricant shows a low initial friction coefficient of 0.07 which reaches a steady state value of 0.10, from 355 to 1500 m.

The common feature between the different lubricants studied in the present work, is their ability for the formation of surface layers from the beginning of the sliding. In the case of DSu and Water + DSu, friction cannot be reduced until water has evaporated under sliding conditions. The evaporation of water under static conditions to obtain Water + DSu thin film leads to lower friction values from the start of the sliding.

Most important, from the point of view of surface protection is the very mild wear obtained in the presence of the boundary DSu film, confirmed both by profilometry surface topography (Fig. 4a) and by cross section profiles (Fig. 4b), and also by SEM microscopy (Fig. 5).

Fig. 4
figure 4

a Surface topography and b cross section profiles of wear tracks on stainless steel disks after lubrication with DSu, Water + DSu and DSu thin film

Full size image
Fig. 5
figure 5

SEM micrographs of AISI 316L disk after lubrication with: a DSu, b Water + DSu, c DSu thin film

Full size image

The wear plastic deformation and abrasion marks present in the stainless steel wear tracks after lubrication with neat DSu (Fig. 5a) and with Water + DSu (Fig. 5b) are absent or are very mild after thin film lubrication (Fig. 5c).

These results could be related with the presence of water in DSu and Water + DSu during the high friction running-in period. The experimental method followed in the preparation of DSu thin layer, with the formation of a surface film of DSu molecules, allows a more effective lubrication from the start of the sliding, preventing severe wear and reducing friction coefficient.

3.2.2 Surface Analysis

Surface analysis results obtained by XPS for Water + DSu and DSu thin film lubricant are shown in Table 4. A slight increase of the atomic percentages corresponding to –C–O and –C(O) C1s binding energies, at 286.4 and 288.7 eV, is observed for Water + DSu inside the wear track (8% and 4%, respectively), with respect to the values found outside the wear track (6.4% and 3.3%, respectively). DSu thin film shows very similar results inside and outside the wear track. When compared with Water + DSu, the thin film shows a reduction of the atomic percentage corresponding to the 285.0 eV aliphatic adventitious carbon C1s peak, and an increase to 4% of the atomic percentage for the Cr2p binding energy at 576.5 eV, assignable to Cr2O3, from 1% for Water + DSu.

Table 4 XPS binding energies and atomic percentages on AISI 316L after lubrication with Water + DSu and DSu thin film
Full size table

Another difference between both lubricants is the reduction of the percentage of the Fe 2p peak at 712 eV, assignable to Fe(OH)O, together with the increase of the percentages for the rest of the Fe 2p peaks between 707 and 710.5 eV (corresponding to metallic iron and to iron oxides, respectively) for the DSu thin film, with respect to Water + DSu.

These results would be in agreement with mild surface modification by tribochemical processes caused by Water + DSu, which are not observed for the DSu thin film.

The results obtained for DSu thin layer are similar to those recently described for DCi thin layer [38], and could be extended to other carboxylate water soluble ionic liquids.

3.3 Lubrication with Graphene Dispersions in DSu

When water cannot be used, the strategy we have discussed in the previous section would not be applicable.

Previous results for the ammonium carboxylate PIL, tri-[bis(2-hydroxyethyl)ammonium] citrate (DCi), have shown its ability to disperse graphene in a concentration of 0.1 wt.% [38].

Table 3 shows that the DSu + 0.1G dispersion is not able to improve the tribological performance of neat DSu. When the concentration of graphene is reduced to 0.05 wt% to obtain the new dispersion DSu + 0.05G, viscosity is increased only by a 10% (Table 1) and the mean friction coefficient is slightly reduced to 0.11 (Table 3), with respect to DSu.

More significant than the mean friction coefficient values are their evolution with sliding distance. Figure 6 shows the variation of friction coefficients with sliding distance for DSu + 0.1%G and DSu + 0.05%G, compared with that of neat DSu.

Fig. 6
figure 6

Variation of coefficient of friction with sliding distance for neat DSu, DSu + 0.1%G, and DSu + 0.05%G

Full size image

The addition of 0.1 wt% graphene almost doubles the running-in distance, from 330 m for neat DSu, to 644 m for DSu + 0.1%G. During that long running-in period, the value of the friction coefficient is always higher, up to 100%, for DSu + 0.1%G with respect to neat DSu. In contrast, DSu + 0.05%G shows no high friction running-in, with a 75% reduction of the initial friction coefficient from 0.24, for DSu and DSu + 0.1%G, to 0.06 for DSu + 0.05%G. In this way, the addition of a lower graphene concentration achieves the lowest initial friction coefficient, without the need of the previous deposition of thin layers.

Another most remarkable feature of the DSu + 0.05G dispersion is its outstanding antiwear performance. In fact, surface damage cannot be observed by surface topography (Fig. 7a) or cross section profilometry (Fig. 7b).

Fig. 7
figure 7

a Surface topography and b Cross section profiles of wear tracks on AISI 316L after lubrication with neat DSu, DSu + 0.1%G and DSu + 0.05%G

Full size image

SEM microscopy (Fig. 8) shows a plastically deformed wide wear track after lubrication with DSu + 0.1G (Fig. 8a), and the complete absence of wear after lubrication with DSu + 0.05G (Fig. 8b).

Fig. 8
figure 8

SEM micrographs: a Wear track on AISI 316 disk after lubrication with DSu + 0.1G, b region of the sliding path on AISI 316L disk after lubrication with DSu + 0.05G

Full size image

Raman microscopy shows the absence of graphene from DSu + 0.05G on most of the wear track surface. The presence of graphene was detected only on some very limited spots on the steel surface (Fig. 9). In this case, no permanent deposition of a continuous graphene layer [11, 38, 39] on the steel surface has been observed, probably due to the absence of a worn track.

Fig. 9
figure 9

Raman microscopy spectra of graphene, DSu and of a spot on the AISI 316L sliding path surface, after lubrication with DSu + 0.05G

Full size image

Bands at 1601.3 and 1359.0 cm−1 are assigned to graphene G and D bands, respectively [41]. DSu spectrum shows strong bands, assignable to C–H stretching, at 2964.1 cm−1 and 2939.0 cm−1.

TEM microscopy has been used in order to determine the possible mechanisms which have taken place. Graphene nanoplatelets were recovered from DSu + G dispersions before and after the tribological tests. The results of TEM microscopy observation are shown in Figs. 10 and 11.

Fig. 10
figure 10

TEM micrographs of graphene from DSu + 0.1%G: a before and b after the tribological test

Full size image
Fig. 11
figure 11

TEM micrographs of graphene from DSu + 0.05%G: a before and b after the tribological test

Full size image

Figure 10a shows graphene from DSu + 0.1G dispersion before the lubrication test. Figure 10b shows the presence of nanometer-sized wear debris particles on the graphene platelets surface after the tribological test. This is in agreement with previous observations for APIL [EMIM][DCA] + 1 wt% graphene [39], where wear debris nanoparticles are adhered to graphene sheets. In contrast, graphene from DSu + 0.05G dispersion is free form wear debris both before (Fig. 11a) and after (Fig. 11b) the tribological test, in agreement with the absence of wear or surface damage.

Thus, DSu + 0.05G prevents wear without significantly changing the composition, surface topography, or roughness of steel.

4 Conclusions

Two different strategies in the use of protic ammonium carboxylate ionic liquid lubricants for friction reduction and surface protection have been confirmed in the present study, namely, their use in aqueous lubrication and in combination with graphene.

A thin film obtained on the steel surface after evaporation of water from a solution of the ionic liquid in water reduces wear rate and eliminates the high friction period during running-in. Surface analysis confirms the absence of tribocorrosion for the thin film lubricant.

Under full-fluid lubrication, in the absence of added water, the dispersion of a low concentration of graphene nanoplatelets in the neat ionic liquid also eliminates the initial high friction sliding distance. Because volume loss takes place during the high friction running-in sliding distance, the elimination of running-in prevents surface damage and the formation of wear debris.

The objective of eliminating the high friction running-in periods have been reached both in the case of the ionic liquid thin film and in the case of the 0.05% graphene dispersion, although only the latter one is able to combine low friction coefficient with the absence of wear.