Effect of the addition of coated SiO2 nanoparticles on the tribological behavior of a low-viscosity polyalphaolefin base oil

Wear 530-531 (2023) 205025 Available online 22 June 2023 0043-1648/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Effect of the addition of coated SiO 2 nanoparticles on the tribological behavior of a low-viscosity polyalphaolefin base oil F´ atima Mari˜ no a , Jos´ e M. Li˜ neira del Río a , b , David E.P. Gonçalves b , Jorge H.O. Seabra c , Enriqueta R. L´ opez a , Josefa Fern´ andez a , * a Laboratory of Thermophysical and Tribological Properties, Nafomat Group, Department of Applied Physics, Faculty of Physics and Institute of Materials (iMATUS), Universidade de Santiago de Compostela, 15782, Santiago de Compostela, Spain b Unidade de tribologia, vibraçoes e manutençao industrial, INEGI, Universidade do Porto, Porto, Portugal c FEUP, Facultade de Engenharia da Universidade do Porto, Rua Dr. Roberto Frías s/n, 4200-465, Porto, Portugal ARTICLE INFO Keywords: Low-viscosity oil Nanoadditives Anti-wear additives Surface modification of NPs Electric vehicles Automatic transmission fluids ABSTRACT This work reports tribological properties of PAO6 containing SiO 2 nanoparticles modified with stearic-acid (SiO 2 - SA) as additives at concentrations (0.05, 0.10, 0.20, 0.30) wt% and the same concentration of SA as a dispersant. Tribological experiments were performed at 120 ◦C in pure sliding and rolling-sliding conditions (5% slide-to-roll ratio). All nanolubricants have better anti-friction capabilities than PAO6. The optimum concentration for friction reduction was 0.30 wt% for both tribological conditions. The best anti-wear results for the specimens tested in pure sliding conditions were achieved with PAO6 +0.20 wt% SiO 2 -SA with reductions of 55%, 86% and 92%, in wear track width, wear track depth and wear area, respectively. Tribological mechanisms of the nanoparticles have been analyzed through roughness measurements, concluding that polishing, tribofilm and adsorption of the additives occur. 1. Introduction Electric vehicles (EVs) are gaining relevance in recent years to reduce fuel consumption and air pollution. To further increase this positive outcome, it is important to enhance their performance. One way to improve the EV efficiency is to optimize the tribological behavior of the mechanical parts, such as the power transmission. For some configurations of EVs, transmission fluids must fulfill different requirements to those formulated for combustion vehicles. For the configurations where the electric engine and the transmission are in the same housing, the electric transmission fluids (ETF) should a) have ability to limit corrosion of copper and be compatible with polymers used in of the housing and in the electronic components; b) have low viscosity; c) show adequate electric properties [1–4]. The reason for using low viscosity lubricants is due to the high torque and operational speeds of tribological elements in EVs. By reducing the viscosity of the oil, viscous drag and viscous heating decrease and heat transfer is increased [5–7]. However, reducing the lubricant viscosity leads to a shift from full film to boundary lubrication, which can lead to severe surface contact and wear, which means that improved anti-wear and anti-friction properties are needed. One of the most successful methods to reduce friction and wear is the use of nanoadditives in lubricants [8,9]; which leads to a reduction in energy losses as well as emissions of pollutants, protecting the environment, and improves the life of machine elements [10–12]. An important factor in attaining a good nanolubricant is its stability for a prolonged period of time, the agglomeration of the nanoparticles (NPs) may lead to additive loss, efficiency reduction and machinery damage due to abrasive wear [9]. Several methods can be applied to enhance stability, such as physical treatment, use of surfactants and surface modification [13,14]. The need for more stable nanodispersions is especially relevant for those with lower viscosities due to the lower stability of the NPs in these fluids. In addition to high torque and high speeds, the ETFs for EVs should endure high temperatures. Polyalphaolefins are broadly used in different applications due to the great capacity as lubricants and their oxidative and thermal stability, leading to better results compared to mineral oils [15]. In this work, polyalphaolefin 6 (PAO6) has been chosen owing to meet those characteristics as well as its low viscosity. Concerning PAO nanolubricants, some authors prepared dispersions with PAO6, which only have enough stability time to do the tribological * Corresponding author. E-mail address: [email protected] (J. Fern´ andez). Contents lists available at ScienceDirect Wear journal homepage: www.elsevier.com/locate/wear https://doi.org/10.1016/j.wear.2023.205025 Received 13 April 2023; Received in revised form 12 June 2023; Accepted 14 June 2023 Wear 530-531 (2023) 205025 2 measurements [16–18]. Ali et al. [19] solved this issue using oleic acid as a dispersant in a PAO (54 cSt at 40 ◦C) with Al 2 O 3 , TiO 2 or a combination of both NPs as nanoadditives remaining the nanodispersions stable for at least 35 days. As regards the nanoadditives chosen in this work, silicon oxide NPs (SiO 2 NPs) have unique physical, chemical, and optical properties, so they can be applied in many fields [20]. These NPs have been analyzed as lubricant additives, showing great capability as friction and wear modifiers [21–25]. Cort´ es et al. [21] analyzed the effectiveness of uncoated spherical SiO 2 NPs (diameter 20–30 nm) as additives for a vegetable oil, obtaining reductions up to 77% for friction and 74% for wear volume. Zawawi et al. [26] used both SiO 2 NPs (30 nm) and Al 2 O 3 NPs (13 nm) to prepare nanolubricants based on a PAG finding an optimum composite volume concentration of 0.02% at which the friction and wear rate reductions are 4.8% and 13%, respectively. Several authors [22–25,27] modified the surface of SiO 2 NPs to improve the stability of their nanodispersions, being a common modification their functionalization with oleic acid as was done by Peng et al. [22] achieving a stability time of around 30 days for a paraffin oil (43 cSt at 40 ◦C) with concentrations varying from 0.05 to 1.0 wt% of oleic acid modified SiO 2 NPs (92–110 nm). Peng et al. [23] also studied the influence of the size of SiO 2 NPs on tribological properties concluding that nanodispersions of the smallest oleic acid modified SiO 2 NPs (58 nm) in a paraffin oil showed the best tribological results and a stability of at least 30 days. Singh et al. [27] modified the surface of SiO 2 NPs (35 nm) with isopropyl alcohol and prepared nanodispersions of the modified NPs in epoxidized Madhuca indica oil (45.1 cSt at 40 ◦C) achieving strong friction and wear reductions with the nanodispersions of 0.8 wt% modified SiO 2 NPs. Silanization of SiO 2 NPs is another common surface-modification. Sui et al. [24] prepared hairy silica NPs (HSNs) with different ratios of two tethered silanes one polar and the other nonpolar, the more nonpolar tethered silane leads to the best stability time (2 months) due to the lower polarity of the resulting modified surface, which is more compatible with PAO100. Sui et al. [25] further confirmed the improvement of stability of HSNs with a nonpolar modification by comparing the effect of different tethered functionalities in amino silane modified SiO 2 NPs, the best stability results (2 months) in PAO100 were obtained for CH 3 terminated HSNs. In this article, commercial SiO 2 NPs (8 nm) were chemically modified through stearic acid (SA) to improve its stability in PAO6. Several nanodispersions were tribologically tested using pure sliding conditions and rolling-sliding conditions (5% slide-to-roll ratio) at high temperature (120 ◦C). The worn surface in pure sliding condition tests was examined using a 3D optical profilometer and confocal Raman microscopy to understand the NPs tribological performance. For comparative reasons, the tribological effects of SiO 2 -SA NPs were compared with those of a zinc dithiophosphate (ZDDP) and uncoated SiO 2 NPs. ZDDP is one of the most currently used friction and wear modifier additives in engine, industrial and transmission oils but for environmental restrictions and operational problems this additive should be replaced [28–30]. 2. Experimental section 2.1. Materials and reagents SiO 2 NPs (8 nm, 99%) were acquired from US Research Nanomaterials, Inc. (Houston, TX, USA). The reagents used in the esterification reaction were sodium hydroxide (0.05 M) from Honeywell (Charlotte, NC, USA), stearic acid (95%) from Sigma-Aldrich (Saint Louis, MO, USA) and hydrochloric acid (37%) from Merck (Darmstadt, Germany). The solvent used to transfer the NPs to PAO6 is n-hexane (95%) from Labkem (Barcelona, Spain). All these products were used with no additional purification. ZDDP was provided by REPSOL. Polyalphaolefin base oil, PAO6, was provided by Repsol and has the following thermophysical properties: a density of 0.8114 g cm −3 and a kinematic viscosity of 30.62 cSt, both at 40 ◦C, as well as a viscosity index of 138. PAO6 oil was characterized by means of Fourier Transform Infrared Spectrometry (FTIR) and by Raman spectroscopy (Supplementary information, Figs. S1 and S2, respectively). 2.2. Stearic acid functionalization of SiO 2 An esterification reaction was used to functionalize the SiO 2 NPs with SA (Fig. 1), following a procedure similar to that of Mari˜ no et al. [31]. Thus, commercial SiO 2 NPs (200 mg) were dispersed in distilled water (20 mL) in a round-bottom flask, containing a magnetic stir bar, and heated to 60 ◦C with 350 rpm agitation, using a magnetic hot-plate (Fig. 1a). Meanwhile, SA (415 mg) was dissolved in distilled water (10 mL) with the aid of ultrasonication (Fisherbrand Ultrasonic Bath, 180 W, 37 kHz). Once the SiO 2 aqueous dispersion reached the set temperature (60 ◦C), 5 mL of NaOH (0.05 M) were added (Fig. 1b). A minute later, the SA aqueous solution was added dropwise. The obtained mixture was kept for 30 min at 60 ◦C, and then refluxed at 95 ◦C for 90 min, finally obtaining the SA coated SiO 2 NPs (SiO 2 -SA) dispersed in water (Fig. 1c). The chemical composition of the SiO 2 NPs was analyzed by FTIR with a Nexus Thermo Nicolet spectrometer, using the attenuated total reflectance configuration (4000–400 cm −1 ) and by a confocal Raman microscope (WITec Alpha300R+) with a 532 nm laser processing the data with the WITec Project FIVE software. The SA and SiO 2 -SA NPs were also analyzed using FTIR and Raman to compare their spectra with those of uncoated NPs to confirm that the esterification reaction was carried out successfully. 2.3. Preparation and characterization of the nanodispersions The dispersion method of the NPs in PAO6 is summarized in Fig. 2. After the esterification reaction, the aqueous nanodispersion was neutralized with HCl (37%), followed by a washing step using a MPW MUniversal Centrifuge (4000 rpm, 15 min): firstly, with distilled water, then with ethanol and finally with hexane. The collected NPs were redispersed in n-hexane for storage and subsequent handling. The reason for this washing step is to separate the SiO 2 -SA NPs from the excess of SA and reaction impurities (such as NaCl) remaining in the water. After this purification process, the concentration of SiO 2 -SA NPs was obtained by difference in weight of 1 mL of the nanodispersion before and after the evaporation of n-hexane. Then, a known amount of n-hexane nanodispersion was added to a given amount of PAO6. This new dispersion was homogenized by ultrasonication for 15 min and, after that, heated up to 95 ◦C, remaining at this temperature for 60 min to remove the n-hexane, obtaining a 4.7 wt% SiO 2 -SA +PAO6 nanodispersion. Finally, this dispersion was diluted to obtain nanodispersions with concentrations of 0.3, 0.2, 0.1 and 0.05 wt% in SiO 2 -SA. In order to further enhance the stability time of these last nanodispersions, the same wt% of SA as of SiO 2 -SA was added as surfactant. Furthermore, viscosity of the base oil and the formulated nanolubricants were measured at 0.1 MPa from 5 to 100 ◦C using a rotational viscosimeter Stabinger SVM3000 (Anton Paar, Graz, Austria). Experimental values of viscosity are presented in the supplementary information (Table S1). More details about the Stabinger SVM3000 were previously reported [31,32]. 2.4. Tribological methods Pure sliding tribological analyses with PAO6 and the SiO 2 -SA +SA nanodispersions were performed with an Anton Paar MCR 302 modular rheometer equipped with a tribology cell T-PTD200 unit working in the rotational ball-on-three-pins configuration. Temperature was precisely monitored with a Peltier hood HPTD200. More details on configuration, characteristics, and experimental conditions of these experiments can be found in Table 1 and reference [32]. Furthermore, the same tests conditions were used for PAO6 +0.2 wt% SA, PAO6 +0.2 wt% SiO 2 NPs, F. Mari˜ no et al. Wear 530-531 (2023) 205025 3 PAO6 +0.2 wt% SiO 2 -SA NPs and PAO6 +0.2 wt% ZDDP. During each test, the values of the coefficient of friction, CoF, were recorded against the sliding distances, and their mean CoF value was determined. For each lubricant, the tests were repeated at least three times. Thus, from the mean CoF values of the repetitions, the average CoF value for each lubricant was determined and its standard deviation was calculated. Fig. 1. Schematic drawing of the esterification reaction: a) SiO 2 NPs in water, b) after the addition of NaOH and c) after the SA addition. Fig. 2. Scheme of the dispersion method. Table 1 Pure sliding tests: configuration of the tribometer, specimen details, and test conditions. Tribology Cell T-PTD200 (Anton Paar) Ball-on-three pins configuration Peltier hood HPTD200 Tribo-pair Experimental conditions 100Cr6 steel ball: 12.7 mm diameter and 0.15 μ m Ra 100Cr6 pins: 3 mm radius and a 0.3 μ m Ra Both: Hardness: 58–65 HRC Young modulus: 190–210 GPa Poisson ratio: 0.29 Sample: 0.5 mL (3 replicates) Axial force (F): 20 N Tribological normal force (F N ): 9.43 N Maximum Hertzian contact pressure: 1.1 GPa Sliding distance: 340 m Rotational speed: 213 rpm Sliding speed: 0.10 m/s 120 ◦C Table 2 Profilometer specifications and measured wear parameters. S Neox (Sensofar) Specifications Measured parameters Mode: confocal Magnification objective: 10 × Software: SensoScan and SensoMap Wear Track Width (WTW) Wear Track Depth (WTD) Worn area (Area) Surface roughness (Ra): •ISO 4287 standard •Gaussian filter wavelength cut-off: 0.025 mm F. Mari˜ no et al. Wear 530-531 (2023) 205025 4 Prior to the wear analysis, the worn surface of the pins and the balls was cleaned with hexane. Subsequently, the wear-track on each pin and on the ball was characterized through a non-contact 3D optical profilometer (Sensofar S Neox), whose specifications are summarized in Table 2. The average values of these parameters and their standard deviations were obtained from the worn track profiles of the nine pins tested with each lubricant. To examine the effect of the presence of the NPs in the tribo-contact, the composition of the tribofilm formed in the worn pin surfaces was analyzed by confocal Raman microscopy (WITec Alpha300R+). To further analyze the effect of the additives in the worn pin surfaces, SEM micrographs of the PAO6 base oil and its formulated lubricants (PAO6 +0.2 wt% SA, PAO6 +0.2 wt% SiO 2 NPs PAO6 +0.2 wt% SiO 2 -SA NPs +0.2 wt% SA, PAO6 +0.2 wt% SiO 2 -SA NPs and PAO6 +0.2 wt% ZDDP) were obtained with a Zeiss Ultraplus Field Emission Scanning Electron Microscope, FESEM, at three magnifications (500x, 1000x and 2000x). Moreover, rolling-sliding tribological tests were also carried out to measure the friction coefficient in different lubrication regimes using an EHD2 ball-on-disc test rig. The configuration of this apparatus consists of a rotating steel ball in contact with a rotating steel disc in rollingsliding arrangement. The friction coefficient is measured with a torque cell coupled to a rotating ball shaft. Three steel discs with different roughness were used to cover the different lubrication regimes. The specimens and experimental conditions are presented in Table 3. The friction coefficient is determined from the friction force measured in negative and positive slide-to-roll ratio, SRR, (ball rotating faster than the disc and vice versa). The friction coefficient is obtained as the average of those achieved from two different friction tests: one ramp raising speed and the other one reducing speed. Additional information about this device was presented in an earlier article [33]. 3. Results and discussion 3.1. Nanoparticle characterization A JEOL JEM-1011 Transmission Electron Microscope (TEM) was used to observe the morphology of the uncoated SiO 2 nanopowders. TEM images (Fig. 3) of the nanopowders, dispersed previously in water, show small agglomerates of roughly spherical NPs. The infrared spectra of pure SA, commercial SiO 2 NPs and dried SiO 2 -SA NPs are displayed in Fig. 4. Table 4 shows details of the characteristic peaks of these spectra. The FT-IR spectrum of SiO 2 -SA NPs shows signals from both SiO 2 NPs and SA, which are: two symmetrical stretching bands at 2917 and 2848 cm −1 corresponding to C–H bond in –CH 3 and –CH 2 from the SA, as well as two wagging vibration bands at 1458 and 1450 cm −1 of C–C in –CH 2 [34]; some important weaker peaks appear at 1683 and 1064 cm −1 , the former due to the C – – O stretching of the SA and the later from the Si–O–Si anti-symmetric stretching of the SiO 2 NPs [35]. All these signals present a slight shift compared to those of SA and SiO 2 counterparts, most likely due to the strong interactions between both core and shell. An important evidence for the existence of covalent bonds between them is the absence of FT-IR signals corresponding to in-plane and out-of-plane bending O–H vibrations [36] at 1429 and 939 cm −1 respectively, which appear in the FT-IR spectrum of SA. Furthermore, the drop in the intensity of the C – – O stretching band shows that some of the SA molecules have reacted with both types of oxygen atoms, –OH and C – – O, as it has been represented in Fig. 1. To further study the composition of SiO 2 NPs, SA dispersant, and SiO 2 -SA NPs, a Raman analysis was performed (Supplementary information, Fig. S2), whose Raman spectrum was also used in section 3.3 for the Raman analysis of the worn surface. 3.2. Nanolubricant stability The stability of the nanodispersions was assessed by visual observation over time until sedimentation was detected. The images are shown in Fig. 5. Just after their preparation, the SiO 2 and SiO 2 -SA +SA nanodispersions presented a high transparency (Fig. 5a and d). The nanodispersion with uncoated NPs sedimented after 24 h. Thus, a yellow-white turbidity appeared in the bottom of the flask (Fig. 5b), turning more yellow and compact over time (Fig. 5c). In contrast, the SiO 2 -SA nanodispersion kept the same slightly white and transparent appearance even after 100 days (Fig. 5f). Therefore, the nanodispersion containing the SA-modified SiO 2 NPs showed an important enhancement of the stability compared to that containing unmodified SiO 2 NPs. 3.3. Tribological results 3.3.1. Pure sliding conditions Table 5 and Fig. 6 summarize the average CoF values for PAO6 and for the studied lubricants. All the nanodispersions (PAO6 +SiO 2 -SA NPs +SA) lead to coefficients of friction lower than the ones measured for the PAO6 base oil, the reductions varying from 15 to 56%, depending on the concentration. The mass concentrations of the dispersions leading to the most significant CoF reductions are (0.2 and 0.3) wt% SiO 2 -SA NPs +(0.2 or 0.3) wt% SA in PAO6. Comparing the coefficients of friction of the PAO6 +0.2 wt% SiO 2 -SA NPs +0.2 wt% SA nanodispersions with that of PAO6 +0.2 wt% SA mixture, the former showed a 17% coefficient of CoF reduction compared to the latter. SA and SiO 2 -SA NPs seem to have a synergistic effect: the reduction of CoF of the combined additives (53%) is higher than the reductions of friction obtained when only one of the additives is present, the PAO6 +SA mixture shows a 44% reduction and the PAO6 +SiO 2 -SA NPs nanodispersion a 15% reduction, both compared to the base oil. In addition, the SA coating on the SiO 2 NPs surfaces leads to a higher CoF reduction than uncoated SiO 2 NPs, which did not have any CoF reduction effect on the PAO6 base oil. Besides, the PAO6 +0.2 wt% SiO 2 -SA NPs +0.2 wt% SA nanodispersion has shown higher coefficient of friction reduction than that of PAO6 + 0.2 wt% ZDDP mixture, which had a 33% reduction compared to PAO6. The wear of the worn surfaces of the pins tested in the friction tests was analyzed using a 3D profilometer. Table 6 summarizes the average values of the Wear Track Width (WTW), Wear Track Depth (WTD) and worn area (area) and the expanded uncertainties, U (k=2). The WTD values are plotted in Fig. 6 for the studied lubricants. All the nanodispersions (SiO 2 -SA +SA) lead to lower WTD values than PAO6. In what regards to wear, the optimum nanolubricant leading to the highest reductions for all the wear parameters (55%, WTW; 86%, WTD; and 92% worn area) is the PAO6 +0.20 wt% SiO 2 -SA +0.20 wt% SA nanodispersion. In Fig. 7 the profiles of the worn surfaces obtained with PAO6 and each nanodispersion (PAO6 +SiO 2 -SA NPs +SA) are shown. The largest groove is obtained using the base oil, and the narrowest one using PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA nanodispersion. Comparing the PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA nanodispersion with the PAO6 +0.20 wt% SA mixture, the former showed an 86% reduction in WTD compared to the latter, which only led to a wear reduction of 3% in Table 3 Rolling-sliding tribometer: specimen characteristics and experimental conditions. EHD2 apparatus (PCS Instruments) Disc and Ball Specimens Experimental conditions Material: 100Cr6 steel Temperature: 120 ◦C Young modulus: 210 GPa Load: 50 N Poisson ratio: 0.29 SRR: 5% Disc diameter: 100 mm Speed: Ramp from 0.05 to 2 m s -1 Ball diameter: 19.05 mm Sample: 120 mL Surface roughness smooth disc, Ra =0.02 μ m rough disc 1, Ra =0.10 μ m rough disc 2, Ra =0.34 μ m ball, Ra =0.02 μ m. F. Mari˜ no et al. Wear 530-531 (2023) 205025 5 WTD compared to PAO6. The reduction in WTD obtained with PAO6 + 0.20 wt% SiO 2 -SA NPs was also lower (30%) than using combination of both additives (86%), which indicates a great synergy between both additives. This trend is similar to that obtained for the coefficient of friction results. Hence, the nanodispersion containing both SiO 2 -SA NPs and SA showed an enhanced tribological behavior. Concerning the SA coating effect on the SiO 2 anti-wear results, the nanodispersion containing commercial SiO 2 NPs was the only one that worsen the PAO6 Fig. 3. TEM micrographs of the uncoated SiO 2 NPs. Fig. 4. FTIR spectra of uncoated SiO 2 NPs (blue), SA (orange) and SiO 2 -SA NPs (grey). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Table 4 FTIR spectra: wavenumber and vibration mode assigned to each chemical bond. Substance Peak (cm −1 ) Vibration mode Chemical bond SiO 2 1064 ν as Anti-symmetric stretching Si–O–Si 950 ν Stretching Si–O 782 δ Bending Si–O–Si 451 ρ Rocking Si–O SA 2913 ν s Symmetrical stretching C–H bond in –CH 3 2846 ν s Symmetrical stretching C–H bond in –CH 2 1700 ν Stretching C – – O 1429 δ ip In-plane bending O–H 1400–1180 ω Wagging C–C in –CH 2 939 δ oop Out-of-plane bending O–H SiO 2 -SA 2917 ν s Symmetrical stretching C–H bond in –CH 3 2848 ν s Symmetrical stretching C–H bond in –CH 2 1683 ν Stretching C – – O 1454 ω Wagging C–C in –CH 2 1450 ω Wagging C–C in –CH 2 1064 ν as Anti-symmetric stretching Si–O–Si Fig. 5. Photographs of PAO6 +0.2 wt% SiO 2 and PAO6 +0.2 wt% SiO 2 -SA + 0.2 wt% SA dispersions at 0 h (a,d), 24 h (b,e) and 2400 h (c,f) after their preparation. Table 5 Average coefficients of friction, CoF, at 120 ◦C and the expanded uncertainties, U (k=2), for PAO6 base oil and all studied lubricants. Lubricants CoF U Reduction % compared to PAO6 PAO6 0.1542 0.0026 0.05 wt% SiO 2 -SA +0.05% SA 0.1316 0.0021 15 0.10 wt% SiO 2 -SA +0.10% SA 0.1264 0.0014 18 0.20 wt% SiO 2 -SA +0.20% SA 0.0719 0.0015 53 0.30 wt% SiO 2 -SA +0.30% SA 0.0680 0.0003 56 0.2 wt% SA 0.0864 0.0048 44 0.2 wt% SiO 2 0.1554 0.0039 0 0.2 wt% SiO 2 -SA 0.1303 0.0063 15 0.2 wt% ZDDP 0.1034 0.0022 33 F. Mari˜ no et al. Wear 530-531 (2023) 205025 6 wear results, increasing a 15% the WTD compared with PAO6. Furthermore, the optimal nanodispersion of PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA also has a better wear reduction capability compared to the PAO6 +0.20 wt% ZDDP, which had a 75% WTD reduction compared to PAO6 base oil. The arithmetic average roughness, Ra, was also measured through 3D profilometry, considering a gaussian filter 0.25 mm. This parameter is useful to identify the anti-wear mechanisms of NPs. In Table 7 the average values of Ra are plotted for all the worn surfaces lubricated with PAO6 and PAO6 +SiO 2 -SA NPs +SA nanodispersions, as well as for the unworn surface of a pin. All the nanodispersions led to worn tracks with smaller Ra than those of the unworn surface and of the worn pins lubricated with PAO6. The lowest Ra is that of the worn surface lubricated with PAO6 +0.3 wt% SiO 2 -SA NPs being 41% lower than that of Fig. 6. Average coefficients of friction, CoF, (blue) and wear track depth, WTD, (orange) for PAO6 and PAO6 and the studied lubricants. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Table 6 Average values of the worn pin surface width, WTW, depth, WTD, and area of worn tracks and the expanded uncertainties, U (k=2), for PAO6 and all studied lubricants. Lubricants WTW/ μ m U/ μ m WTD/ μ m U/ μ m Area/ μ m 2 U/ μ m 2 PAO6 435 10 2.60 0.19 806 65 0.05 wt% SiO 2 -SA + 0.05 wt% SA 256 14 0.99 0.12 200 32 0.10 wt% SiO 2 -SA + 0.10 wt% SA 268 19 0.55 0.05 104 18 0.20 wt% SiO 2 -SA + 0.20 wt% SA 195 17 0.35 0.06 40 9 0.30 wt% SiO 2 -SA + 0.30 wt% SA 240 18 0.51 0.07 61 12 0.20 wt% SA 422 19 2.54 0.25 723 76 0.20 wt% SiO 2 471 21 3.01 0.22 967 69 0.20 wt% SiO 2 -SA 342 14 1.82 0.21 353 44 0.20 wt% ZDDP 232 12 0.64 0.11 81.2 13 Fig. 7. Profile images of worn pin tracks lubricated with PAO6 (blue) and PAO6 +wt% SiO 2 -SA +wt% SA nanodispersions (0.05 wt%, orange; 0.1 wt% grey; 0.2 wt %, yellow; 0.3 wt%, green). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Table 7 Average values of the Ra and Rq of untested pin surface, worn tracks and the expanded uncertainties, U (k=2), for PAO6 and PAO6 +wt% SiO 2 -SA NPs +wt % SA nanodispersions. Ra/nm U/nm Rq/nm U/nm PAO6 11.2 1.4 15.3 2.4 0.05 wt% SiO 2 -SA NPs+0.05 wt% SA 8.5 0.6 10.4 0.7 0.10 wt% SiO 2 -SA NPs+0.10 wt% SA 6.1 1.0 8.8 2.5 0.20 wt% SiO 2 -SA NPs+0.20 wt% SA 6.2 1.3 7.3 1.5 0.30 wt% SiO 2 -SA NPs +0.30 wt% SA 5.2 0.7 6.2 0.9 Untested Pin 8.9 0.6 10.4 0.6 F. Mari˜ no et al. Wear 530-531 (2023) 205025 7 the unworn surface, and 53% of the worn surface lubricated with PAO6. However, 0.1 and 0.2 wt% nanodispersions exhibit very similar results with more than a 30% reduction compared to the unworn surface and a 45% compared to the worn surface from PAO6. Another parameter used to characterize the topology of the surfaces is the root mean square roughness (Rq), which is more sensitive to large deviation from the mean line than Ra [37]. The same trend with lubricants and pins for Ra and Rq have been found, as can be seen in Table 7. 3D images of the worn tracks of the steel balls used in the friction tests are shown in Fig. 8. The nanolubricant containing SiO 2 -SA and SA as dispersant led to the lowest wear on the ball, 173 μ m for WTW. Interestingly, the absence of SA dispersant (PAO6 +0.20 wt% SiO 2 NPs and PAO6 +0.20 wt% SiO 2 -SA NPs) resulted in WTW values similar to or higher than that obtained with neat PAO6 base oil. However, the use of only SA as an additive generated a slightly higher WTW value (201 μ m) compared to the optimal nanodispersion, which is coherent with the wear results of the pins, thus a synergistic effect of both additives is observed. The ball lubricated with PAO6 containing ZDDP as additive showed good wear reduction, although the WTW value was higher (199 μ m) than that obtained for the optimal nanodispersion. Furthermore, Raman mappings of the worn tracks of pins show distinctive areas corresponding to different compounds indicated with several colors (Figs. 9 and 10). For the assignation of the components in the worn surface, the Raman signals were separated using the software Project FIVE. Concerning the worn pins tested with the base oil (Fig. 9) the presence of PAO6 (blue), as well as of iron oxides (yellow) and of carbon (green) coming from the lubricant degradation were observed [31]. On the other hand, the Raman mapping from the 0.20 wt% SiO 2 -SA NPs +0.20 wt% SA nanodispersion in PAO6 (Fig. 10) indicates the presence of SiO 2 NPs (red), as well as of PAO6 (blue) and of carbon (green). The presence of SiO 2 NPs is confirmed by the appearance of an intense and sharp peak close to 500 cm −1 [38]. This peak does not appear in the Raman spectrum of iron oxide, which instead presents peaks of similar intensity in the lower region of the spectrum (Fig. 9) [39], i.e., with the addition of SiO 2 -SA NPs, iron oxides were not observed. On the other hand, in the Raman spectra of the mapping of Fig. 10, the lack of SA signal (absence of C–H characteristic peaks around 3000 cm −1 ) reveals that, during the tribo-tests, some tribo-chemical reactions occur promoted by the high both temperatures and pressures due to the friction process. Thus, these conditions cause the breaking of the bonds between SA and the coated SiO 2 NPs, similarly to what was found by Zhang et al. [40] for SA modified TiO 2 NPs through XPS. Assuming the same hypothesis as these authors [40], the uncoated SiO 2 NPs are easily adsorbed on the worn surface, generating a boundary lubricating film. Moreover, the SA could be physically adsorbed on the steel surface during the tribotests [41–43] and be removed with the hexane solvent in the cleaning process of the worn pins before the Confocal Raman Analysis. In Figs. 9 and 10, the presence of the chemical components coincides with the direction of the grooves formed during the tribological experiments. In addition, there is a great presence of PAO6 in the worn pin tested with the base oil. When adding the SiO 2 -SA NPs and SA to the PAO6 lubricant the tribofilm created in the worn track is mainly composed by SiO 2 NPs and carbon, being the tribological mechanisms governed by the nanoparticles rather than the base oil. Considering the roughness and these Raman results it can be assumed that on the worn surface tribofilm formation, and polishing and mending effects occur. Regarding the SEM micrographs of the worn pins (Fig. 11), inspection of these images reveals that abrasive wear in the sliding direction is the main wear mechanism in the case of the PAO6-lubricated worn surface, with an evident ploughing even at the lowest magnification (500x). This type of wear is slightly reduced when SA is used as the only additive. The addition of SiO 2 or SiO 2 -SA NPs reduce the grooves in the worn surface but microcracks appear on the surfaces. Furthermore, when ZDDP is used as PAO6 additive, although the wear track is greatly reduced, the surface shows similar damage. The least damaged surface is the one obtained using PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA as lubricant, which once again highlights the positive synergies between both additives, which increase scuffing resistance and reduce wear. These results agree with those obtained with profilometry for the pins. 3.3.2. Rolling-sliding conditions Rolling-sliding tribological results obtained with PAO6 base oil and the SiO 2 -SA NPs +SA nanodispersions are presented as Stribeck curves (Figs. 12 and 13). For this aim, these curves are built by plotting the coefficient of friction against the specific lubricant film thickness, Λ, that was defined as: Λ =h c / Ra (1) where h c corresponds to the theoretical central film thickness and Ra is the equivalent average surface roughness of the contacting surfaces, given by Ra = (Ra2 disc +Ra2 ball) √. In this case, the lubricant central film thickness, h c , at the operating temperature (120 ◦C) was predicted by means of the Hamrock and Downson’s equation [44], taking into account different parameters related to the geometrical and mechanical properties of ball and discs, lubricant properties (viscosity and pressure−viscosity coefficient) and finally the test parameters like speed, load and SRR (slide-to-roll ratio). For each tested lubricant, the experimental dynamic viscosity values from 5 ◦C to 100 ◦C (Table S1) were correlated through the Fig. 8. 3D images of the worn ball surfaces for (a) PAO6, (b) PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA, (c) PAO6 +0.20 wt% SA, (d) PAO6 +0.20 wt% SiO 2 NPs, (e) PAO6 +0.20 wt% SiO 2 -SA NPs, (f) PAO6 +0.20 wt% ZDDP, and their WTW values. F. Mari˜ no et al. Wear 530-531 (2023) 205025 8 Vogel−Fulcher−Tammann (VFT) equation. The dynamic viscosity of each nanolubricant and base oil at 120 ◦C and 0.1 MPa was obtained by extrapolation of the VFT correlations. This methodology was used because the Stabinger SVM3000 viscometer does not permit viscosity measurements above 100 ◦C. The pressure−viscosity coefficient of PAO6 was taken from Ref. [45]. Due to the low concentration of the NPs, for the three nanolubricants their pressure−viscosity coefficient values were considered equal to those of PAO6. Fig. 12 shows the Stribeck curves for PAO6 neat oil as well as for SiO 2 -SA nanolubricants using three different discs: smooth, rough 1 and rough 2 discs (see Table 3). As expected, ball-on-disc tests using the rough discs generate lower Λ values and consequently show higher coefficients of friction than those performed with the smooth disc. At the lowest specific lubricant film thickness, i.e., lowest entrainment speeds, the coefficient of friction values are much lower with any nanolubricants (SiO 2 -SA +SA) than with the PAO6 neat oil, for all discs. Therefore, the key effect of SiO 2 -SA nanoparticles is to significantly reduce the friction under boundary film when the lubricant film build-up is poor (low speeds and low Λ values). Concerning the effect of mass concentration of nanoadditives in the tribological performance, the CoF decreases as the concentration of nanoadditives rises in the studied range. For this reason, the Stribeck curve of PAO6 neat oil together with that of 0.30 wt% SiO 2 -SA +0.30 wt % SA nanolubricant was plotted in Fig. 13. This result agrees with the coefficient of friction measurements achieved for the pure sliding tests, where the 0.30 wt% SiO 2 -SA NPs +0.30 wt% SA was one of the optimal concentrations as anti-friction nanolubricant. At high entrainment speeds (right section of the Stribeck curves) and therefore large specific lubricant film thickness, the coefficient of friction is quite similar for all nanolubricants and base oil. Generally, under elastohydrodynamic (EHD) lubrication, a typical full Stribeck curve shows three lubrication regimes: boundary film, mixed film, and full film lubrication. In general, boundary film lubrication appears if Λ <Λ 0 , mixed film lubrication if Λ 0 ≤Λ ≤Λ 1 , and EHD lubrication if Λ >Λ 1 [46,47]. The values of Λ 0 and Λ 1 , depend on several parameters, in particular the composite surface roughness, lubricant additives, lubricant temperature (120 ◦C) and the application (e.g., ball-on-disc geometry, rolling bearings, gears). In Fig. 13 the coefficients of friction corresponding to PAO6 and to the nanodispersion (PAO6 +0.30 wt% SiO 2 -SA NPs +0.30 wt% SA) are plotted against Λ defined by equation (1). In the case of the nanodispersion, it can be observed that for Λ <0.08 (approximately) the coefficient of friction is almost constant (CoF ≈0.04) which is typical of the boundary film lubrication regime of lubricants containing additives. Fig. 9. Elemental mapping and Raman spectra of worn pins tested with PAO6 base oil. Fig. 10. Elemental mapping and Raman spectra of worn pins tested with PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA nanolubricant. F. Mari˜ no et al. Wear 530-531 (2023) 205025 9 Fig. 11. SEM micrographs at three magnifications of the worn pin surfaces lubricated by (a) PAO6, (b) PAO6 +0.20 wt% SiO 2 -SA NPs +0.20 wt% SA, (c) PAO6 + 0.20 wt% SA, (d) PAO6 +0.20 wt% SiO 2 NPs, (e) PAO6 +0.20 wt% SiO 2 -SA NPs, (f) PAO6 +0.20 wt% ZDDP. F. Mari˜ no et al.