Current Knowledge on Friction, Lubrication, and Wear of Ethanol-Fuelled Engines—A Review

Abstract

The urgent need for drastic reduction in emissions due to global warming demands a radical energy transition in transportation. The role of biofuels is fundamental to bridging the current situation towards a clean and sustainable future. In passenger cars, the use of ethanol fuel reduces gas emissions (CO₂ and other harmful gases), but can bring tribological challenges to the engine. This review addresses the current state-of-the-art on the effects of ethanol fuel on friction, lubrication, and wear in car engines, and identifies knowledge gaps and trends in lubricants for ethanol-fuelled engines. This review shows that ethanol affects friction and wear in many ways, for example, by reducing lubricant viscosity, which on the one hand can reduce shear losses under full film lubrication, but on the other can increase asperity contact under mixed lubrication. Therefore, ethanol can either reduce or increase engine friction depending on the driving conditions, engine temperature, amount of diluted ethanol in the lubricant, lubricant type, etc. Ethanol increases corrosion and affects tribocorrosion, with significant effects on engine wear. Moreover, ethanol strongly interacts with the lubricant’s additives, affecting friction and wear under boundary lubrication conditions. Regarding the anti-wear additive ZDDP, ethanol leads to thinner tribofilms with modified chemical structure, in particular shorter phosphates and increased amount of iron sulphides and oxides, thereby reducing their anti-wear protection. Tribofilms formed from Mo-DTC friction modifier are affected as well, compromising the formation of low-friction MoS₂ tribofilms; however, ethanol is beneficial for the tribological behaviour of organic friction modifiers. Although the oil industry has implemented small changes in oil formulation to ensure the proper operation of ethanol-fuelled engines, there is a lack of research aiming to optimize lubricant formulation to maximize ethanol-fuelled engine performance. The findings of this review should shed light towards improved oil formulation as well as on the selection of materials and surface engineering techniques to mitigate the most pressing problems.

1. Introduction

Humanity is experiencing a critical moment that requires a complete paradigm shift through a radical energy transition. According to the targets established in the COP26 report, in order to rescue the climate and guarantee sustainable development in the future two key strategic themes are the mobility sector and the generation of clean and renewable energy [1]. In transport, electric vehicles should gradually replace internal combustion engines; however, this transition will occur at different rates in different regions. Heavy vehicles used in sea and air transport are difficult to electrify. Thus, the use of low-carbon renewable fuels will be needed for many years to come in order eliminate fossil fuel use and its very high carbon footprint. Biofuels represent a fundamental bridge in ensuring the transition to a new era dominated by renewable, sustainable, and clean energy technologies, as it is estimated that in 2030 the global fleet of internal combustion engines will be around 2 billion. According to the International Energy Agency, although liquid biofuels only accounted for 4% of global transport energy demand in 2020 it is estimated that this will rise to 14% in 2050 [2]. The main countries with extensive public policies that support growth of biofuel use are Brazil, the United States, Canada, Indonesia, and India [3].

There are serious concerns regarding the use of ethanol fuel due to the large amount of arable land required for the crops used to produce bioethanol and how it can impact food supply [3], although this has been mitigated by second-generation and third-generation ethanol fuel, which is mostly focused on industrial waste and algae raw materials.

The octane number (ON) of ethanol is greater than that of gasoline, although its calorific value is lower, requiring a larger amount of fuel to produce the same engine power [4]. The role of bioethanol in the energy transition is unquestionable, not only because it is renewable but mostly because CO₂ emissions resulting from the use of ethanol as fuel are lower than those from gasoline even when a full life-cycle assessment is considered [5]. Mixtures of gasoline with ethanol produce lower concentrations of unregulated gaseous species such as sulfur dioxide (SO₂) and isocyanic acid (HNCO) as well as fewer harmful polycyclic aromatic hydrocarbons (PAHs) when compared with gasoline [6]. Moreover, the amount of particulate matter (PM) in the exhaust system is lower for fuels containing ethanol as compared with gasoline [6,7]. When the effects of the PM emitted from engines running on mixtures of gasoline and ethanol were evaluated in biological tests, they were shown to be less cytotoxic and to have lower levels of reactive oxygen species (ROS) and lower mutagenicity than for engines running on gasoline alone [6].

However, the use of ethanol fuel poses serious tribological challenges to engines. As ethanol has a high latent heat of vaporization and a moderate boiling point, it can easily reach the tribological contacts, either via the cylinder walls or by dissolving the lubricant in the crank case lubrication reservoir [8]. The challenges that ethanol fuel brings to moving engine components, such as degradation of the materials involved, reduction of lubricant viscosity, interaction with additives in lubricants, tribocorrosion, etc., need to be carefully identified and mitigated. In this review, we focus on the tribological aspects regarding the use of ethanol fuel. Before we delve into the effects of ethanol on friction and wear, we first provide an account of how it affects the main lubricant properties, as these can substantially influence friction and wear. After describing friction and wear when ethanol is present, we detail how several of the described behaviors result from the effects of ethanol on the formation and durability of tribofilms. Finally, we present a discussion of current knowledge gaps, future trends, and lubricant guidelines.

2. Effects of Ethanol on Lubricant Properties

Compared to other fuels such as gasoline and diesel, bioethanol (anhydrous and hydrated) has a higher heat of vaporization [9] and higher boiling point [10]. These higher values enhance the possibility of unburned bioethanol condensation on the cylinder wall being dragged by the piston rings to the oil sump [11,12]. As a consequence, large amounts of bioethanol may accumulate in the crankcase. In fact, amounts from 6% to 25% of diluted bioethanol and its combustion products have been found in the crankcase for both field [13,14] and engine tests [15,16,17]. This amount is higher for operating conditions that prevent proper heating of the lubricant, such as cold running conditions [18,19], short driving/testing cycles [13], start-stop technology [13], as well as higher concentrations of ethanol blended in the fuel. Higher concentrations of ethanol fuel in the oil sump impact the properties and performance of the lubricating oil in a different manner than gasoline, an issue that requires further investigation [9]. Furthermore, ethanol is more hydrophilic (dissolves easily in water) and hygroscopic (water-absorbing) compared to other fuels. Thus, the effect of water in the lubricant must be taken into account. The following section reviews the effects of ethanol on lubricant properties that are directly related to oil service life and performance.

2.1. Effect of Ethanol Fuel on Lubricant’s Physicochemical Properties

The viscosity, total base number (TBN), and total acid number (TAN) play essential roles in the lubricating system. These parameters are often used to evaluate oil quality and define oil change intervals [20]. Although there are no universally accepted warning limits, several researchers have monitored changes in lubricant properties resulting from engine use [21,22,23].

During usage, it is common to observe oil acidification (increase in TAN) due to thermal oxidation and contamination with fuel and combustion byproducts; this is accompanied by a reduction in alkaline reserves (reduction in TBN), which are the compounds that neutralize acid products, as well as a slight reduction in viscosity due to thermal aging and fuel contamination [9,20]. This trend of viscosity and TBN reduction followed by an increase in TAN is accentuated by ethanol dilution [24,25]. At elevated temperatures, oxidation of fuel–oil mixture may generate peroxide in the presence of oxygen; this peroxide undergoes further reaction to form carboxylic acid, ketones, aldehydes, and alcohols, potentially resulting in a further increase in the acidity of the lubricant [26].

In real engine applications the lubricant ages due to thermal and fuel contamination. The acidification of the lubricant caused by thermal aging is accelerated by ethanol dilution due to its higher oxygen content. The role of temperature on lubricant properties has been scrutinized by a number of researchers. Most of the work has focused on comparing lubricant properties and the performance of fresh and thermally aged lubricants [27]. Essentially, these experiments consist of heating oil samples at different temperatures and times and then evaluating their friction and wear response through tribotests such as pin-on-disc and ball-on-disc. The overall observations indicate that aged lubricants present higher friction and wear values than corresponding fresh lubricant under mixed or boundary lubrication conditions. This is attributed to oil oxidation (increased TAN), which alters the dynamics of tribofilm formation. Tribofilms formed from thermally aged lubricants present a different molecular structure with a higher oxygen content [28,29]. A recent experimental methodology involved first contaminating the lubricant with different amounts of ethanol, then performing tribotests [30]. Analogously to the observation for thermally aged lubricants, the fuel-contaminated lubricants presented higher friction and wear in comparison to their fresh versions in steel-to-steel contact. The cause is similar in both cases, being due to tribofilm molecular structural changes; these are described in detail in Section 4. However, such changes occurred even without oil oxidation, and the tribofilms formed in the presence of ethanol did not present the same structure as tribofilms formed from thermally aged lubricants [29]. Thus, tribofilm formation and lubricant performance are affected by tribofilm oxidation and lubricant oxidation, leading to higher friction losses and wear [27,31]. Such acidification of the lubricant increases oxidative and corrosive wear, as observed by the presence of black oxides (magnetite), pits, and mass loss [9].

Viscosity reduction due to fuel dilution has been observed by several researchers in both field [24,29] and lab tests [32,33,34], which is mostly due to the much lower viscosity of ethanol in comparison to usual engine lubricants. On the one hand, engine tests comparing gasoline and gasoline with different concentrations of ethanol showed that viscosity reduction over time depends much more on the lubricant type than the fuel type, as both gasoline and ethanol have much lower viscosity than engine lubricants [24]. On the other hand, Costa and Spikes [32] found larger viscosity reductions for anhydrous than hydrated ethanol diluted in the lubricant. The authors hypothesized that the true solubility of the ethanol in the lubricant is lower for hydrated ethanol, resulting in some of the ethanol forming a microemulsion with water, which has little influence on lubricant viscosity. Modern engines lubricated with synthetic oils presented lower viscosity reduction than simpler mineral oils. After field tests of 3000 km performed with four motorcycles fuelled with gasoline and a mixture of gasoline and ethanol, the viscosity decreased by about 20% and 45% for synthetic and mineral-based oils, respectively, regardless of the fuel used [29]. Oil oxidation did not take place after 3000 km regardless of the tested lubricant and fuel. However, it should be pointed out that tribofilm formation is compromised in the presence of ethanol (see Section 4) even in the absence of oxidation.

It is important to point out that viscosity reduction might bring benefits in terms of friction reduction [35] for components operating under full film lubricating conditions, as lower viscosity reduces shear losses in the lubricant [32]. However, as further detailed in Section 3, the reduced viscosity makes the contact conditions more severe for mixed and boundary lubricating conditions, which may lead to higher friction values and wear [19,27,29,32,34].

2.2. Deposit Formation Due to Ethanol Contamination

Due to environmental concerns related to the emission of particulates, which are related to respiratory diseases, soot generated from the combustion process now enters the crankcase with combustion gas blow-by and is “collected” by the lubricant. The soot, which is mostly formed by carbons, is measured by quantifying the total carbon content of lubricating oil. The bonding of soot, water, and acids forms sludgy masses, which is a widespread and commonly known problem in internal combustion engines. In fact, it is one of the main reasons for failure in lubricated systems during operation [19].

On the one hand, soot emissions from alcohol-fuelled engines are lower relative to other fuels; hence, reduced soot loading enhances the lubricating oil life [36]. On the other hand, ethanol is known to cause higher lubricant oxidation than gasoline. Ethanol is a polar solvent that can break down the additives in engine oil, leading to increased oxidation and degradation of the lubricant. In comparison, gasoline has lower solvency and is less likely to cause lubricant oxidation and additive degradation under the same levels of contamination [37].

In the case of ethanol–lubricant mixtures giving rise to acetic acid formation during operation [38], there is likely to be intensified oxidation and significant resulting sludge formation. This was observed by Besser, Schneidhofer, Dörr, Novotny-Farkas and Allmaier [37], who verified that the addition of acetic acid accelerates the ageing process by increasing oxidation, neutralization number, and sludge formation as well as slightly reducing TBN in comparison to neat lubricant or lubricant mixed with ethanol. However, in the absence of acetic acid ethanol is less likely to form sludge than gasoline.

Figure 1 attempts to summarize the effects of ethanol on lubricant properties and how it affects deposition formation throughout the lubricant lifetime. The use of lubricant decreases TBN and increase TAN, with both effects becoming more severe in the presence of ethanol contaminating the lubricant. Viscosity does not change much during lubricant use (only a slight decreasing tendency) up to a certain point that defines the lubricant lifetime. Within this period, ethanol reduces the lubricant viscosity. Afterwards, intense deposit formation should lead to a significant increase in lubricant viscosity. The onset of intense deposit formation occurs earlier for lubricants contaminated with ethanol, meaning that ethanol reduces the operation limit, thereby shortening the lubricant lifetime. Afterwards, the more intense deposit formation in the presence of ethanol should increase lubricant viscosity compared with neat lubricant.

3. General Trend of the Impact of Ethanol on Friction and Wear

The literature addresses different aspects of the impact of ethanol on engine performance. A number of these publications are directed to the analysis of tribological outputs, i.e., friction and wear. Several tribological systems have been used in these studies, ranging from measurements in actual engines to laboratory tests; thus, care should be taken when exploring the results in light of the varying characteristics of each system. In addition, the materials in contact can vary in different publications, as can the percentage of ethanol in the fuel, from zero ethanol (E0) to 100% ethanol (E100). In fact, engines that use pure ethanol commonly use hydrated ethanol (with around 5% water) instead of anhydrous ethanol, as it does need to dissolve with the gasoline. Therefore, most publications that refer to E100 should instead refer to E95.

Tests conducted directly on spark ignition engines provide a comprehensive comparison of the tribological effect of different fuels. dos Santos Filho, et al. [39] ran a fired engine dynamometer with either Brazilian gasoline (E25) or with hydrated ethanol (E95), then characterized the engine components for comparison. By looking at the topography of the cylinder bores after testing, the authors concluded that localized wear at the bottom dead centre (BDC) was significantly more pronounced for E95 than E25, as shown in Figure 2.

Another study used a Yamaha E150 maritime engine to investigate the tribological impact of ethanol [25]. However, instead of looking at the engine components the authors evaluated the extent of fuel dilution when running the engine with either E0 or with E10 fuels (a blend of gasoline and 16% isobutanol (i-B16) was tested as well). Oil samples were extracted at given intervals of engine service and later used in different laboratory tests (unidirectional sliding, reciprocating sliding, four-ball and block-on-ring scuffing). Despite detecting a reduction in oil viscosity due to fuel dilution, the unidirectional and reciprocating sliding tests did not result in significant differences in terms of friction coefficient for all conditions tested, independently of the fuel type and time of engine service. The differences in terms of wear during the reciprocating and four-ball tests were not significant either. On the other hand, E10 resulted in a 30% reduction in scuffing load, which represents the system load-carrying capacity, in comparison with E0. Used engine oils were explored by Cousseau, Ruiz Acero and Sinatora [29] in laboratory tribological tests. For this, 100 h dynamometer tests were conducted with identical engines running with either E22 or E95 fuels. Oils extracted at the end of these tests were used in reciprocating sliding tests. Although the viscosity for the oil from the E22 test was slightly higher than for the E95 test, no significant differences were found in the friction coefficients measured in the tribological tests.

Laboratory tests to investigate the effects of fuels on tribological performance such as those conducted by Ajayi, Lorenzo-Martin, Fenske, Corlett, Murphy and Przesmitzki [25] are usually relatively simple. These tests involve tribological systems that may represent engine conditions but are inherently different. According to the authors, the reciprocating sliding tests were associated with boundary lubrication conditions, which were responsible for the non-observation of changes in friction coefficient induced by changes in oil viscosity. Similarities in friction behaviour were observed when mixing 6% in volume of a range of fuels (from E0 to E85) with an SO SAE 5W30 lubricant in four-ball tests [9]. In that case, a slight trend of increase in wear scar diameter was observed for the mixtures with higher fractions of ethanol. A different trend regarding friction coefficient was observed by Crepaldi, et al. [40], who conducted lubricated reciprocating sliding tests of actual piston rings against cylinder bores. Four lubricants with different friction modifier additives were considered in this case, each one analysed in fresh condition as well as contaminated with 10% of either anhydrous (AE) or hydrated (HE) ethanol. Noticeable reductions in friction coefficient were observed when the base oil was contaminated with AE or HE, while reductions in friction coefficient were less pronounced for the different types of friction modifiers.

More sophisticated laboratory test configurations may help to expand the analysis of the effect of ethanol on friction to other lubrication regimes. Stribeck curves associated with a given lubricant can be obtained during sliding/rolling contact of a steel ball and a steel disk in a piece of equipment called a Mini-Traction Machine (MTM). Costa and Spikes [32] used MTM tests to analyse the impact of adding 5% of either AE or HE on friction under different lubrication regimes. Under boundary lubrication conditions, ethanol in a fully formulated oil increased friction, probably by interaction with the lubricant additives (see Section 4), while it reduced friction in a base oil. Under mixed lubrication, ethanol increased friction for both the base oil and formulated lubricant, which was because the reduction in viscosity caused by ethanol led to thinner elastohydrodynamic (EHL) films, thereby increasing asperity contact. Under full film lubrication, ethanol reduced friction for both the base oil and formulated lubricant due to the viscosity reduction caused by the fuel dilution enabling the lubricant molecules to slide more easily in relation to each other.

Regarding the materials in tribological contact, different trends may be observed for the impact of ethanol on the lubrication of different materials. For example, Banerji, et al. [41] compared the friction and wear behaviour of an SAE 5W30 synthetic oil to its mixture with E85 (50% vol.) in unidirectional ball-on-disk tests. Ethanol dilution resulted in a significant decrease in friction coefficient for the contact of cast iron discs with DLC-coated AISI 52100 balls. However, the behaviour of diluted and undiluted oil was not significantly different when the cast iron disk was in contact with a cast iron counter-body. An opposite trend was observed in terms of wear, with ethanol dilution being responsible for significant differences in volumetric wear for the cast iron–cast iron contact and insignificant wear differences for the cast iron–DLC contact. Bandeira, et al. [42] studied the effect of ethanol contamination on coated materials in unidirectional sliding contacts. The tests were conducted with CrN-coated AISI 4140 disks against silicon nitride balls lubricated with either fresh SAE 5W40 lubricant or a mixture with HE (5% in vol.). No significant differences were observed in terms of friction coefficients or wear rates when contaminated and non-contaminated oils were compared.

Attempts to run laboratory tests with a mixture of ethanol and lubricant, such as those described in previous paragraphs, usually require attention to the test temperature. The selection of temperatures above the evaporation point of ethanol (around 69 °C) may result in quick ethanol evaporation which, while it may not invalidate the test, needs to be considered during analysis of the results [32,33,34,40]. As an alternative, in a study of the tribological impact of ethanol Lenauer, et al. [43] conducted an artificial engine oil alteration process with addition of ethanol combustion products to the oil, which was later used in reciprocating sliding laboratory tests. Two fully formulated oils were considered (SAE 15W40 and 5W30). The ethanol combustion products were acetaldehyde and acetic acid. The results indicated that the contaminants resulted in significant changes in the friction behaviour of the 15W40 oil, but did not for 5W30. Both contaminants provided similar results in terms of wear for the two oils, increasing wear rates during running-in and lowering wear rates during the steady-state regime.

The tribological impact of ethanol is not restricted to conditions where ethanol contaminates the lubricant oil. High-frequency reciprocating rig (HFRR) tests were conducted to analyse the impact of water contamination on ethanol lubricity [44]. The results showed that water did not result in significant changes in coefficient of friction, while the aspect of the wear scar was modified, showing a trend of decreasing wear scar diameter with the increase in water fraction. HFRR tests have been selected to study the impact of ethanol on the lubricity of various formulations of gasoline as well. The results indicate that ethanol may enhance the lubricity of gasoline [45], and may form protective carbon-rich tribofilms on certain metallic surfaces [46].

4. Effects of Ethanol on the Formation and Durability of Tribofilms

The differences in wear when ethanol is used as a fuel have been associated with different phenomena: i. because the combustion pressure for ethanol needs to be higher than for gasoline, the tribological conditions are more severe [47]; ii. lubricant dilution due to the lubricant becoming contaminated with fuel reduces its viscosity, in turn reducing the thickness of the hydrodynamic lubricant film [32]; iii. the fuel may interact with lubricant boundary additives, affecting the protective tribofilms formed under boundary lubrication conditions [34]; iv. ethanol may lead to corrosion and tribocorrosion of the moving surfaces [48]; and v. ethanol accelerates lubricant degradation [29]. This section reviews the main works in the literature that have aimed to unravel how ethanol fuels affect the formation of protective tribofilms, with special emphasis on tribofilms formed from lubricant boundary additives.

4.1. Ethanol and ZDDP Tribofilms

Zinc dialkyldithiophosphate (ZDDP) is a very important antiwear additive in engine lubricants. In fact, the performance of automotive lubricants strongly relies on the presence of tribofilms formed from ZDDP to protect engine parts against wear [49]. This tribofilm has a very complex structure which may be influenced by the presence of ethanol in the lubricant. Experimental evidence has indicated that the formation of ZDDP tribofilms is induced by shear stresses produced by contact [50], involving an initial reaction of sulfur and phosphorus with metallic surfaces that forms sulphides and phosphates. The zinc present in ZDDP forms zinc sulphide at the interface with the metallic surface (which contains iron sulphide), whereas the continuous decomposition of ZDDP forms additional phosphates that react with zinc phosphate, leading to long-chain zinc polyphosphates. Near the tribofilm surface, long-chain zinc polyphosphates prevail [51]. According to Martin [52], the sulphides near the metallic interface can react with iron that is continuously exposed when the surface is rubbed, preventing its being attacked by oxygen and thereby protecting the surface against severe wear by iron oxide debris.

However, when the lubricant is contaminated with ethanol this tribofilm may be substantially altered. When adding small amounts of ethanol fuel to a formulated lubricant in ball-on-disk tests using a mini-traction machine (MTM), Costa and Spikes [32] found that although ethanol slightly reduced friction under conditions where an elastohydrodynamic film was formed, due to a small reduction in lubricant viscosity it interfered with the tribofilm formed in the boundary lubrication region, increasing the coefficient of friction. These results prompted an investigation of the impacts of ethanol contamination (both anhydrous and hydrated) of lubricants containing ZDDP on the growth and stability of anti-wear tribofilms using the MTM-SLIM technique [34]. The results showed that ethanol significantly retarded the growth of ZDDP tribofilm. Moreover, the final tribofilm was particularly thin and irregular. For a sequence of tests in which the tribofilm was initially allowed to form in a test with uncontaminated lubricant and then ethanol was added, destruction of the preformed tribofilm occurred and it did not regenerate with subsequent sliding in the presence of ethanol.

Different possible mechanisms could be responsible for the changes in thickness and morphology of the ZDDP tribofilm in case of ethanol fuel contamination. For the case of hydrated ethanol, which is used when ethanol is not added to gasoline but is rather used as the main fuel, the effect of the water needs to be considered as well. MTM-SLIM tests have been used to investigate how water affects ZDDP tribofilm formation using a humidity control system, showing thinner tribofilms for higher humidity [53]. The authors used X-Ray photoelectron spectroscopy (XPS) to analyze the tribofilms, showing that the ratio between bonded oxygen and non-bonded oxygen was substantially lower for higher humidity levels. These findings suggested that the presence of water led to shorter phosphates in the tribofilm, which was traced back to possible depolymerization of the longer polyphosphate chains.

In addition to the effects of water, which may either be present in the lubricant fuel for hydrated ethanol or absorbed by ethanol due to its hygroscopic nature, ethanol itself can affect the ZDDP tribofilm. Many hypotheses have been raised in the literature to account for this interaction. Ethanolysis could affect the ZDDP decomposition, similar to what has been observed for methanol [54]. Ethanol could inhibit polymerization of long phosphate chains, and consequently affect the growth of the tribofilms involved in the reduction of film thickness when ethanol and/or water are present [55,56]. Alternatively, ZDDP tribofilms could be solubilized by ethanol while it is formed in a tribological contact [34].

Chemical characterization of ZDDP tribofilms using advanced surface techniques have enabled greater understanding of this interaction. Angle-resolved X-Ray Absorption Near-Edge Structure (XANES) and high-resolution XPS have been used in a synchrotron radiation ring to analyze tribofilms formed on steel during reciprocating tests involving lubricants contaminated with ethanol (AE and HE) [28]. The XANES spectra showed that in addition to zinc sulphide, iron sulphide was present for the oils contaminated with ethanol. In ZDDP tribofilms, the formation of zinc sulphide results from a reaction with the products of the digestion of debris by the phosphates in the tribofilm. According to Martin [52], iron sulphide is only found in ZDDP tribofilms under severe wear conditions. Regarding the effects of ethanol on the phosphates of the tribofilm, XANES showed that metal transition phosphates were more pronounced for the contaminated lubricant, while XPS measurements showed that ethanol decreased the amount of PO₃⁻³ species and increased the amount of short iron phosphates near the substrate [28]. Therefore, ethanol apparently reduces the length of the phosphate chains, perhaps via ethanolysis (for anhydrous ethanol) and/or hydrolysis (for hydrated ethanol), and favours the formation of short iron phosphates. Tribofilms formed with shorter phosphate chains should be thinner and have a reduced “cushioning” effect. Furthermore, if the reservoir of long-chain metaphosphates available to react with iron oxide is substantially reduced, the mechanism of debris digestion is compromised, resulting in increased wear. This hypothesis has been confirmed by the fact that the sulphides at the interface with the substrate are richer in iron when ethanol is present.

Although most of the evidence in the literature points towards thinner ZDDP tribofilms and higher wear for lubricants contaminated with ethanol, other works have suggested the opposite effect. For example, Banerji, Edrisy, Francis and Alpas [8] conducted pin-on-disk tests of an Al-Si alloy using a synthetic lubricant containing ZDDP (neat condition and contaminated with E85). The authors claimed that wear was higher for the neat lubricant without ethanol, as EDS of the wear tracks showed a higher concentration of oxygen for the tests with the contaminated lubricant along with a rougher morphology. They suggested that the hydroxyl (OH) in ethanol could donate an electron pair to help initiate ZDDP decomposition, thereby favouring tribofilm formation. However, they only showed XPS results for the contaminated lubricant, not for the neat lubricant, making it difficult to evaluate the soundness of their arguments. In addition, the amount of fuel dilution in the lubricant was very large (1:1), much larger than the values observed in field tests for lubricant dilution in engines [19]. Later, the same authors found similar results for tests with grey cast iron, although again no chemical analysis of the tribofilms formed from the neat oils was shown, meaning that it was not possible to verify their claim that ethanol helps ZDDP tribofilm formation [41].

Recent works [28,30] have evaluated tribofilm from fully formulated and tailor-made oils using EDS, Raman, XANES, and XPS analysis. EDS analysis showed higher Zn, P, Mo, S, Ca, and O content on the surfaces tested with lubricant containing ethanol compared to neat lubricant. At first, this could have indicated improved tribofilm formation due to the presence of ethanol, as hypothesized by Banerji, Edrisy, Francis and Alpas [8]. However, tribological tests only presented friction [30] and wear [28] reduction during the tests performed with the neat lubricant, which occurred due to the formation of MoS₂ and ZDDP tribofilms, respectively. The Zn, P, Ca, and O probably originated from detergent and ZDDP antiwear additives present in the lubricant. Their higher contents in the presence of ethanol are most likely due to the increased contact severity caused by film thickness reduction and lubricant degradation. In [28], the same elements were shown to result from tribofilms with shorter phosphate chains and higher oxygen content, which do not have the same wear resistance capability as ZDDP tribofilms formed in the absence of ethanol. In a work by Ruiz-Acero, Kessler, Costa and Cousseau [30], the chemical elements observed with EDS formed a compound which had reduced ability to prevent friction compared to MoS₂, as is described further in Section 4.3.

Another relevant point to be addressed is that one of the raw materials used to produce ZDDP is alcohol, meaning that some alcohol is unavoidably retained in commercial ZDDP during industrial production. Hu, et al. [57] explored the effects of alcohol on ZDDP additives in engine oil, reporting that alcohol increased its load-carrying capacity. They attributed this to the adhesion of polar hydroxyl groups on the metal surface and formation a protective film, although no evidence for this was provided.

Figure 3 summarizes how the use of ethanol can contaminate the lubricant in the engine and affect the antiwear ZDDP tribofilm. The presence of ethanol in the lubricant makes the ZDDP tribofilms substantially thinner. This has been confirmed by optical interferometric images of the borders of the wear tracks using the MTM-SLIM technique, which show very thin and scratched tribofilms in the presence of ethanol (Figure 3a). The thickness of the tribofilm increases with time and stabilizes after around 60 min; however, ethanol delays tribofilm formation and reduces the final film thickness (Figure 3b). In seeking to address why this occurs, the chemical characterization of tribofilms using surface advanced techniques such as XANES and XPS has shown that the sulphide layer near the interface with the substrate is richer in iron (Figure 3c). Ethanol reduces the amount of long phosphates (Figure 3d), allowing short phosphates rich in Fe to prevail (Figure 3e). Structures for ZDDP tribofilms are proposed in Figure 3f for a neat lubricant and in Figure 3g for a lubricant contaminated with ethanol.

4.2. Ethanol and Tribo-Oxidation

The presence of hydroxyl groups in ethanol makes both the metallic surfaces and the lubricant prone to oxidation. Kumar and Agarwal [36] have pointed out that partial oxidation of alcohols may generate formic acid, which is very reactive with metals, leading to corrosion and tribocorrosion of engine components. Aiming to investigate this in a ring/cylinder liner system, tribocorrosion tests were conducted on bare and DLC coated stainless steel in ethanol/water solutions [48]. Tests with higher ethanol concentrations showed that the passive layer was more easily removed during rubbing by monitoring the open circuit potential (OCP) during the tests.

dos Santos and Alves [58] compared lubricants contaminated with ethanol and gasoline (5% wt.) using a high-frequency reciprocating rig (HFRR). However, the commercial gasoline they used (Brazilian gasoline) contained between 18 to 27.5% wt. of anhydrous ethanol, meaning that contamination of the lubricant with gasoline already resulted in a considerable amount of ethanol in the lubricant. Their tests with ethanol under boundary lubrication conditions led to more grooving of the balls, which showed dark oxidized patches. The authors suggested that ethanol in the lubricant formed oxide particles due to tribocorrosion during sliding, which then came off of the surface, leading to abrasive wear. For gasoline contamination, grooving of the ball and oxide particles were observed as well, though to a smaller extent, probably because the amount of ethanol in the lubricant contaminated with gasoline is lower than for ethanol fuel.

Another point is that the combustion of ethanol can result in products such as acetaldehyde and acetic acid, which can themselves contaminate the engine lubricant. Lenauer, Tomastik, Wopelka and Jech [43] tested lubricants containing acetaldehyde and acetic acid and concluded that acetaldehyde mainly affected the ZDDP additive, while acetic acid mainly affected the detergents. They used XPS measurements of the wear track in the cylinder liner to estimate a so-called “tribofilm thickness” using concentration profiles of the elements present in ZDDP, which they attempted to correlate with the steady-state wear rates. The reason for this approach was that the wear scars were too small to be measured by either gravimetric or optical methods. They claimed that the contamination with acetic acid led to lower wear rates; however, the method used to estimate the wear rates was unorthodox, and the fact that the concentration of elements from the additive increased in the wear track does not necessarily indicate a lower wear rate. In another study, Costa and Spikes [32] measured thicker tribofilms for lubricants contaminated with acetic acid, although this was associated with a possible surface oxidation.

Figure 4 summarizes the main effects of ethanol on tribo-oxidation. Ethanol increases the boundary film thickness when boundary additives are not present in the lubricant, which is probably due to surface oxidation. Tests using lubricants contaminated with ethanal and acetic acid, which are decomposition products of ethanol at high temperatures, have shown even thicker tribofilms (Figure 4a). Stribeck curves show that friction in the boundary region is lower when such tribo-oxidation occurs, probably because it reduces the shear strength at the sliding interface (see the regions delimited by red ellipses in Figure 4b) [32]. Severe grooving and oxidation have been observed on the worn surfaces when using lubricants contaminated with ethanol, as the increased amount of oxide debris can lead to abrasion (Figure 4c) [58]. Tribo-corrosion tests show that the protective passive films are more easily removed when higher amounts of ethanol are present (Figure 4d) [48].

4.3. Interaction of Ethanol and Friction Modifiers

The interaction of ethanol fuel and friction modifiers (FM) has not yet received considerable attention in the literature; only very recently have the effects of ethanol contamination on the nature and morphology of low-friction tribofilms formed from friction modifier additives begun to be properly identified and understood.

The friction modifiers most widely used in engine lubricants belong to the family of organic friction modifiers (OFMs), in particular amides, amines, and esters, although more complex organo-acid-based compounds have also been used because they improve corrosion resistance. OFMs are amphiphiles having a predominantly linear alkyl chain with one polar head at one extreme; they adsorb at the metal surface, resulting in easy slipping between opposing end groups as well as a “cushioning effect” [59]. For certain OFMs, water can lead to thicker boundary films [59]. Due to their polar nature, both ethanol and water are expected to strongly interact with the adsorbed films caused by OFMs.

When using reciprocating tests of lubricants containing OFMs, relevant effects of ethanol contamination on friction were not found when compared with fresh lubricants [40]. However, in a more fundamental study Costa and Spikes [33] used ultrathin film interferometry to measure the boundary films formed from lubricants containing different OFMs in order to compare their thickness when the lubricant was contaminated with ethanol (HE or AE). Their tests used three different types of commercial OFMs added to base oil in order to disentangle their behaviour from interactions of ethanol with other additives in the lubricant. For the test conditions that led to full film lubrication, ethanol reduced film thickness for all three OFMs due to the reduction of the lubricant viscosity. However, all three OFMs increased the boundary film thickness (Figure 5a,b). In parallel, Stribeck curves were obtained using MTM tests for the same lubricants, showing that the presence of both HE and AE further decreased boundary friction compared with neat base oil + OFM (Figure 5c,d). Because ethanol increased boundary film thickness and reduced boundary friction, particularly at higher test temperatures, the authors suggested that ethanol interacted with the OFMs to form very thick viscous boundary films, possibly involving chemical reactions that are favored by higher temperatures (Figure 5e). Water probably has a significant effect on boundary film formation, as the boundary films were thicker for hydrated than for anhydrous ethanol.

Another important family of friction modifiers consists of soluble organo-molybdenum compounds such as molybdenum dialkyldithiocarbamates (MoDTCs), which have been substitutingOFMs in engine lubricants, particularly in low-viscosity lubricants. Lubricants containing Mo-DTC show a typical friction behaviour characterized by initial high friction followed by a sudden drop. This friction drop only occurs under severe contact conditions involving high shear stresses at the contact. The rationale is that a shear-induced tribochemical reaction of Mo-DTC forms MoS₂ during rubbing, leading to reduced protection against friction and wear [59]. Raman spectroscopy, especially when combined with other surface techniques, is a very powerful way to detect whether MoS₂ tribofilms are formed on the wear scar [60].

The literature suggests that the decomposition of the Mo-DTC molecule and the sequence of tribochemical reactions that lead to the formation of MoS₂ on the surface is strongly influenced by the ZDDP additive [61,62]. Thus, a first important point to raise is that the effects of ethanol on tribochemical reactions involving ZDDP could indirectly affect the formation of low-friction tribofilms.

Cousseau, Ruiz Acero and Sinatora [29] compared the tribological performance of a fresh commercial fully-formulated lubricant (5W30) containing Mo-DTC with two aged lubricants using an Optimol SRV4 reciprocating tribometer. The aged lubricants were drained after dynamometer tests run with different fuels (E22 and E95). The fresh lubricant showed the typical friction drop after a period of rubbing time, while the formation of an MoS₂ tribofilm was confirmed by Raman analysis of the worn tracks, showing the typical peaks at 407 and 370 cm⁻¹. On the other hand, the aged lubricants did not show any friction drop, indicating that no low-friction tribofilm was formed (Figure 6a). Raman analysis did not show peaks for MoS₂ (Figure 6b); instead, it showed peaks for hematite (α-Fe2O3) and magnetite (Fe3O4). Other works have shown that the typical friction drop that characterizes MoS₂ formation for MoDTC-containing lubricants did not occur when the lubricant was intentionally diluted with ethanol during tribological testing [33,40].

A recent work used a combination of advanced surface techniques (angle-resolved XANES, angle-resolved high-resolution XPS, and Raman) to help understand why and how contamination of the lubricant with ethanol hinders the formation of low-friction MoS₂ tribofilms (Ruiz-Acero, Kessler et al. 2023). For this, reciprocating tests using fully formulated MoDTC-containing lubricants as well as Mo-DTC dissolved in a PAO base oil were run with both fresh lubricant and lubricant contaminated with increasing amounts of ethanol. The authors found a critical amount of ethanol dilution that hindered the typical friction drop for Mo-DTC after rubbing. Surface analyses of the tribofilms showed that ethanol favors the formation of sulfates (Figure 6d) and MoO₃ (Figure 6c) at the expense of MoS₂ and MoS_xO_y. According to the literature, MoS₂ and MoS_xO_y are the main compounds responsible for the friction reduction observed in lubricants containing MoDTC [60], thereby justifying why lubricants contaminated with ethanol show significantly higher friction values than fresh lubricants. The authors proposed that ethanol induces an oxygen-rich interfacial medium, favoring the formation of sulfates and molybdenum oxides and reducing the amount of MoS₂ and MoS_xO_y while increasing the amount of iron oxides in the valleys of the topography, as shown schematically in Figure 6e. Because the sulfates and molybdenum oxides do not have good friction-reducing capacity, friction reduction does not occur for oils contaminated with ethanol.

5. Knowledge Gaps, Future Trends, and Lubricant Guidelines

While ethanol fuel has been successfully used worldwide since the 1970s, its use has increased drastically since the 2000s due to the need for cleaner and renewable fuels to replace gasoline during the energy transition period. Its impacts on the tribology of engines are mainly related to the reduction in lubricant viscosity caused by the oxidant nature of ethanol, which both degrades the lubricant and leads to oxidation and triboxidation of metallic parts, as well as the way in which ethanol interacts with the boundary additives present in the lubricants.

The effects of ethanol on friction when mixed into the lubricant have been fairly well established, with ethanol reducing friction under full film lubrication and increasing friction under mixed lubrication. However, under boundary lubrication the outcome is highly dependent on the type of friction modifier present in the lubricant; thus, ethanol can either increase or reduce friction. This issue needs to be better investigated and understood. Regarding wear, ethanol generally increases wear, although some laboratory studies have stated otherwise. The laboratory tests finding higher wear when ethanol is present in the lubricant support findings from the field, which often show more severe wear in engines fuelled with larger amounts of ethanol as compared to gasoline. This seems to be largely related to the effects of ethanol on the structure of ZDDP antiwear tribofilms. However, the interactions of ethanol and boundary additives (friction modifiers and antiwear additives) have not yet been carefully investigated for other surfaces, such as the DLC and plasma spray coatings currently used in cylinder liners.

The premature lubricant degradation due to ethanol can be mitigated by using an adequate lubricant formulation. Lubricants used in ethanol-fuelled engines should be formulated to be more resistant to oxidation and corrosion compared to lubricants used in gasoline engines. In addition, they should contain higher levels of rust and corrosion inhibitors, as well as additional antioxidants. They might contain higher levels of detergents and dispersants as well to help control deposits, maintain engine cleanliness, and improve viscosity control to ensure adequate lubrication under a wide range of temperatures and operating conditions. Similarly, lubricants with higher base numbers can be more effective at neutralizing the acid products generated during engine operation with ethanol and increasing the required lubricant change interval in ethanol-fuelled engines. These small changes in oil formulation are intended to ensure ethanol-fuelled engines’ proper operation. However, there is a lack of research aiming to optimize lubricant formulation to maximize ethanol-fuelled engine performance. In terms of efficiency, Costa and Spikes [33] showed that for certain specific OFMs the presence of ethanol can lead to thicker tribofilms and lower coefficients of friction. Similar studies aiming to develop extreme pressure (EP), antiwear (AW), and FM additives that are empowered by fuel contamination and work well in lubricants formulated with higher levels of detergents and dispersants are of the utmost importance to increase engine reliability and efficiency. For example, because products of the antiwear additive (AW) ZDDP compromise the efficiency of the catalytic converter in the exhaust system, environmental pressures have led to a rush for alternative AW additives that produce lower levels of sulphur and phosphorous emissions [63]. The interaction of alternative AW additives and ethanol should be urgently investigated in order to determine whether they can better protect surfaces in ethanol-fuelled engines while reducing damage to the exhaust system. Such studies will require close interaction between academics, car manufacturers, and the additive and lubricant industry.