University of Birmingham Synthesis and tribological studies of nanoparticle additives for pyrolysis bio-oil formulated as a diesel fuel

The tribological behaviour of pyrolysis bio-oil with a synthesized nano-Lanthanum oxide (La 2 O 3 ) additive was evaluated using a point contact four ball tribometer under different frictional conditions. Results were compared against a micro ( m )-La 2 O 3 additive and an un-additised bio-oil as a control. The results show that nano-La 2 O 3 impregnated bio-oil had better tribological properties than the control groups. Under the operating loads, the optimum nanoparticle concentration within the bio-oil was investigated. At these levels, the combined action of adsorbed bio-oil ﬁ lms on the worn surfaces and the bearing effects of the nano-La 2 O 3 minimized friction and wear. The tribo-mechanisms were ascribed to adhesive wear as a result of lubrication starvation under high loads, and abrasive wear at high rotational speeds as a result of combined deformation and aggregation of the nano-La 2 O 3 particles.

Synthesis and tribological studies of nanoparticle additives for pyrolysis bio-oil formulated as a diesel fuel 1

. Introduction
Since the energy crisis in 1970s, many countries have focused efforts on the development of alternative sources of energies. In recent decades, bio-oil derived from biomass has been shown as one of the most promising alternatives to the traditional fossil fuels [1e3]. The reasons for this are numerous but include the general availability of the raw materials for the bio-oil that can be derived from waste, and depending on how this is processed there is a potential for the process to be can be carbon neutral. The main components of such bio-oils are oxygen-containing organics including carboxylic acids, alcohols, aldehydes, ketones, phenols etc. Generally, such fuels are prepared though a fast pyrolysis process with the rapid thermal decomposition of biomass in the absence of oxygen at a moderate temperature of~500 C at an extremely high heating rate about 400e500 C/s. As an alternative fuel in internal combustion engines, bio-oil has been shown to enhance combustion characteristics [4]. Bio-oils have also been shown to possess very good lubricating properties. The lubricity of diesel fuels is very important since it has a direct effect on the durability of the fuel injection equipment [5]. Tribologically, some bio-oils have been shown to possess better antifriction properties than those of the conventional fossil fuel. For others however, the wear properties can be much worse, as a result of corrosion and surface degradation. These effects can lead to severe damage of engine components to the extent that the application of bio-oils has been restricted [6e8]. This forms the motivation for seeking methods to upgrade the tribological performance of the bio-oil by inhibiting these negative properties.
Nano-materials exhibit great potential for improving tribological performance and as such they have been widely studied as lubrication additives. Lanthanum oxide is a functional rare earth material and as such, it has found many uses including as a catalyst, and hydrogen storage material [9,10] it also has potential as a tribological additive. For all these applications, the surface shape, particle size, and the dispersed state are of importance. As an example, Xue et al. [11] reported the sliding wear behaviours of nickel composite coatings containing micro-and nano-La 2 O 3 particles. They found that the nano-La 2 O 3 composite coatings exhibited excellent wear resistance. Nano-La 2 O 3 may therefore improve the tribological properties of bio-oil and negate the potential wear limitations outlined above. Furthermore, the use of energetic nanoparticles such as La 2 O 3 , due to small particle size and the relatively low concentration levels, affords a promising method of altering the reactivity of liquid fuels for enhanced combustion stability, resulting in a reduction in ignition delay of some fuels [12].
In this paper, ball-like nano-La 2 O 3 was synthesized and dispersed in a pyrolysis bio-oil to form a Bio-Oil/La 2 O 3 (BO/La 2 O 3 ) suspension via ultrasonic technology. The effect of La 2 O 3 particle size, percentage concentration, friction load and rotational speed on tribological properties were evaluated using a four ball point contact tribometer and the corresponding tribological mechanisms are discussed.

Materials
All of the chemicals used in the present work, such as the lanthanum carbonate (La 2 (CO 3 ) 3 ), nitric acid (HNO 3 ), citric acid monohydrate (C 6 H 8 O 7 H 2 O), acetone and the lanthanum oxide (micro-La 2 O 3 ), were analytical grade reagents.
The BO used was from rice husk via fast pyrolysis technology and was provided by Anhui Province Key Laboratory of Biomass Clean Energy, University of Science and Technology of China. The components of bio-oil were as follows: carboxylic acids 32.72%; alcohols 8.14%; aldehydes 23.49%; ketone 7.09%; phenols 13.98%, and others 14.58%. The main physicochemical properties are as follows: The flash point is 68 C; high heating value is 16.5 MJ/kg; kinematic viscosity is 13.2 mm 2 s À1 at 40 C. It should be noted that although the low heating value and high viscosity of the bio-oil in its present form make it unsuitable for use a fuel, upgrading via emulsification would rectify this. More details of the pyrolysis process, chemical composition and physicochemical properties can be found elsewhere [13].

Preparation of nano-La 2 O 3
The preparation of the nano-La 2 O 3 comprised the following steps. 25 mmol of La 2 (CO 3 ) 3 (~11.4 g) and 150 mmol of HNO 3 (~9.4 ml) were dissolved into 100 ml of distilled water to form a clear reaction solution. This was then heated at 60 C for 5 min after which 8.3 mmol of C 6 H 8 O 7 $H 2 O (~1.74 g) was added to it, and magnetically stirred to form a uniform dispersion. This was then heated in a muffle furnace to remove excess water, allowing a gel to form. The precursor was then dried and the nano-La 2 O 3 (nano lanthanum oxide) was produced via calcination at 850 C for 6 h [14]. Finally the La 2 O 3 was added to bio-oil at concentration levels of 0.2, 0.6, 1.0, 1.4 and 1.8% (all the % in this paper means wt.%). Each BO/La 2 O 3 suspension was stable for at least 7 days after being ultrasonically dispersed for 10 min before each test.

Tribological tests
The tribological tests were carried out on a Jinan MQ-800 fourball tribometer. The schematic of the tribological tests is shown in Fig. 1. Referring to ASTM D4172 (Standard Test Method for Wear Preventive Characteristics of Lubricating Fluid), the experimental conditions selected were nominally a load of 100 N with a rotational speed of 1250 rpm at 25 ± 2 C for 30 min. ASTM E52100 bearing steel test samples were used with a diameter of 12.7 mm, a surface roughness (Ra) of 0.032 mm and a hardness of 61e63 HRC.
Experiments were compared against a control sample comprising the bio-oil without additives. These experimental conditions ensured that all of the tribological tests were performed under the boundary lubrication.
After each test, the steel balls were washed using acetone. Wear rates were was evaluated from the average WSD (wear scar diameter) of the three lower specimens. After an initial test programme, the effects of load and speed for the bio-oil with nano-La 2 O 3 were also evaluated, at 150 N and 200 N and 1450 and 1650 rpm, respectively. Test conditions were repeated three times and the friction coefficient was collected in real-time during each experiment.

Characterization
The particle size distribution of the La 2 O 3 powder was measured using a Bettersize BT-9300H laser particle size analyser. The crystal structure of the precipitate was analysed with a Rigaku D/Max-cB powder XRD (X-ray diffractometer) with Cu Ka radiation (g ¼ 0.541 nm) under 40 kV, 100 mA and a scanning speed of 5 min À1 . The topography of the particles and elemental composition of the worn surfaces was observed using a FEI Sirion 200 SEM (scanning electron microscope) with EDS (energy dispersive spectroscopy). The morphologies and chemical valences of elements on the worn surfaces were analysed using an Olympus TPF-1 optical microscope and a Thermo Scientific ESCALAB 250 XPS (X-ray photoelectron spectroscope), respectively. The WSD of the lower specimens was measured with an accuracy of 0.01 mm. In order to analyse the tribological mechanisms, the thickness of lubricating oil film was calculated with the following formula [15]: where h mean is the mean oil film thickness; h 0 is the viscosity of the lubricant; R is the radius of curvature; U is the entraining surface velocity; E is Young's modulus (E ¼ 210 GPa); a is the Pressureviscosity coefficient, (a ¼ 0.44 Â 10 À8 m 2 /N pa À1 ); W is the contact load and k is the ellipticity parameter, k ¼ 1.03.  approximately 20e1000 nm. The particle size distributions of the La 2 O 3 are shown in Fig. 3a. It should be noted that the sizes of La 2 O 3 shown in Fig. 3a are smaller and more uniformly dispersed than in Fig. 2. This discrepancy is caused by the different test methods employed. The solid powders used in the SEM analysis were prone to agglomeration [16], but for the laser particle size analysis the powder was dispersed in circulating water.
where, D is the average crystallite diameter, K is a dimensionless shape factor and was taken as 0.89, l is the X-ray wave length which in this case was 0.15418 nm, b is the half maximum line width, and q is the Bragg angle, The average crystallite diameters calculated using Scherer's equation were consistent with the XRD spectra, indicating that the synthesis of the nano-La 2 O 3 was successful.
3.2. Effect of additive particle size on the friction and wear properties of bio-oil    formation of bio-oil films with solid particles combined with the particle geometry. It is likely that the cylindrical shape of the m-La 2 O 3 particles resulted in abrasive wear at the contact interface similarly to the effects reported in Refs. [18,19].
Trends for the nano-La 2 O 3 indicate that it had a better wear resistance than the micro additive and in all cases improved the tribological properties compared to bio-oil without additives. The WSD decreased as the blended ratios increased to 1%. Beyond this concentration, wear damage steadily increased. This behaviour could be explained by considering again the geometry of the additive particles. Up to 0.6%, the ball-like nano-La 2 O 3 may have acted as a bearing between the surfaces, sustaining and distributing contact loads. The measured particle sizes of the nano-La 2 O 3 were of a similar magnitude to the surface roughness of the contact interface (32 nm) which could also have contributed to this [20]. Beyond 1% concentrations, the nano-La 2 O 3 was shown to agglomerate into larger clusters thus weakened the effect described above leading to an increased WSD. The frictional characteristics displayed similar behavioural trends. The results shown in Fig. 4 suggested an optimal nano-La 2 O 3 blend ratio (i.e. when the tribological properties were at their best) occurred at a concentration of 1% in the bio-oil.
The optical micrographs of the worn surfaces taken from the upper specimens are shown in Fig. 5. The worn surface of the control sample (Fig. 5a) displayed evidence of corrosive damage and many wide and deep furrows. The corrosion seen on the contact surface was a result of acidic components within the bio-oil (such as acetic and formic acid) [6]. For the m-La 2 O 3 concentrations of between 0.2 and 1% did not show any evidence of corrosion (Fig. 5bed). The corrosion phenomena are likely to have been avoided by the La 2 O 3 particles adsorbing the acidic components of the biofuel, aided by the high specific surface area of the particles [10]. However, the corrosive damage returned when m-La 2 O 3 content was increased above 1% (Fig. 5eef). These observations are consistent with the tribological behaviour described above. When the concentration of m-La 2 O 3 is no more than 1%, the cylindrical shaped particles can roll in plane preventing contact between surfaces and reducing chemical attack. However, adsorption of the acidic components from the bio-oil by the m-La 2 O 3 (above 1%) led to further agglomeration and then larger particle abrasive damage at the contact interface. The freshly abraded surfaces were then more likely to be susceptible to further corrosion followed by wear etc., a cycle similar to that described in Ref. [21].
The evidence of wear and surface damage was much less for the nano-La 2 O 3 experiments as can be seen in Fig. 5gei. At concentration levels up to 1%, furrows on the worn surfaces were shallow and this may be due to the topography of the nano particles. Surface damage and wear increased (Fig. 5j, k) when content exceeded 1% and was most likely due to the aggregation of the nano-additives. This is an effect that has been described in Ref. [22]. Consequently, the nano-La 2 O 3 displayed optimal antifriction and antiwear properties at a weight content of 1% in the bio-oil.
The kinematic viscosity and average oil film thicknesses are shown in Fig. 6, whereas the concentration of La 2 O 3 is shown to have increased, so too does the kinematic viscosity and oil film thickness. The m-La 2 O 3 was calculated to have formed thicker oil films between the frictional surfaces than the nano-La 2 O 3 . This was likely to have been caused by the larger particle diameters of the m-La 2 O 3 which led to a higher viscosity for the respective bio-oil blends. Despite this, the friction coefficient and wear volume of the steel specimens was considerably higher than the nano-La 2 O 3 . The reason for this is that the peak oil film thickness was calculated to be approximately 200 nm, however the average measured diameter of the m-La 2 O 3 particles was at least several micrometres. It is assume therefore, that the m-La 2 O 3 in the bio-oil acted as a form of abrasive debris leading to high wear. Conversely, the average diameter of the nano-La 2 O 3 particles was well below the calculated film thicknesses. This theory holds up to concentrations of 1%. Beyond this point, aggregation of the nano particles (leading to the formation of micron-sized debris) had a negative effect on the   tribological properties of the bio-oil suspensions. This was despite the film thickness continuing to increase.
The results given in Figs. 2e6 overwhelmingly show that the nano-La 2 O 3 impregnated BO had a better tribological performance than those of m-La 2 O 3 , therefore, the following sections focus solely on the performance of the nano-La 2 O 3 dispersed bio-oil. Fig. 7 shows variation of friction coefficients against time with test specimens subjected to a 100 N load, rotating at 1250 rpm for 30 min. The friction coefficient of control sample is the highest of the observed fuels with no clear indication of the transition between running-in and steady state conditions. This may be due to the corrosive action of the acidic components within the bio-oil [13]. However, there are clear transitions between initial and steady state conditions for the nano-La 2 O 3 bio-oils The transient friction coefficients of the bio-oils containing 0.6% and 1.4% nano-La 2 O 3 particles are similar. Variations between these two oils can be attributed to the role of aggregation of the nano particles (first forming and then dispersing similar to phenomenon described in Ref. [23]). The equilibrium point for the optimized lubricity occurs at a content of 1% nano-La 2 O 3 . Fig. 8 shows the Raman spectra of the worn surfaces of the test specimens. Two typical characteristic peaks at the D and G bands, located at 1340 cm À1 and 1570 cm À1 , respectively can clearly be seen. Both of these two peaks were very weak without the additive, however, in the strength grew as the concentration of nano-La 2 O 3 in the bio-oil increased. This indicated that the transmission (i.e. the rolling and compressive action) of the La 2 O 3 through the contact may have contributed to the adsorption, onto the frictional surfaces, of organics from the bio-oil, leading to a reduction in friction. At higher concentration levels, this effect was countered by the increased aggregation of nano-La 2 O 3 particles (e.g. at 1.4%). These results are in agreement with the trends seen in Fig. 6. 3.3. Effects of load on the friction and wear properties of the bio-oil with nano-additives Fig. 9 shows the variation of WSD and transient friction coefficient with varying loads for 1% nano-La 2 O 3 bio-oil at 1250 rpm for 30 min. Both were shown to increase with the increasing load. However neither relationship was linear as the higher loads prohibited the transition from mild to severe wear by increasing the real area of contact [24]. With the increase of the load, the oil film   formed and adsorbed between the worn surfaces were also easier to break down [25]. Increasing the load would also have affected the beneficial effect of the geometry of the nano-La 2 O 3 by causing deformation of the ball-like load bearing particles. These are likely to be the root causes of the increasing friction coefficient with load. These effects may also be contributing factors to the lack of a transitional point between the transient and steady state conditions at the highest load. This may indicate the load limit for the nano-La 2 O 3 . Fig. 10 shows the surface topography of the lower specimens for the various loads (100 N, 150 N, and 200 N) displays the extent of surface damage and wear. The transmission of higher normal loads was likely to prevent the formation, at the contact interface, of robust lubricating oil films, leading to oil starvation and adhesive wear [26,27]. Fig. 11 shows the EDS spectra from the worn surfaces (shown in Fig. 10.). An increase in the loads carried by the specimens, led to both an increase in the deposition at the surface of elemental Lanthanum and a reduction in carbon. These elemental characteristics suggested that though the nano-La 2 O 3 carried part of the load transmitted through the surfaces, an increased loading, significantly reduced the effectiveness of the protective oil film. The combined effect of which was an increased friction and wear profile.
The XPS spectra for the same surfaces are shown in Fig. 12. The peaks at 852.6 eV and 835.2 eV can be ascribed to La3d3 /2 and La3d 5/2 of La3d, respectively, suggesting that Lanthanum on the worn surfaces emanated from the La 2 O 3 . In addition to these, the peaks at 288.6 eV, 285.4 eV and 284.9 eV were ascribed to the eCOOH, CeOH(CeOeC), and CeC(CeH) groups, respectively. These   functional groups will have been derived from the acids, alcohols, and carbon chains in the bio-oil. Another effect of increasing the load was reduction in the area of all the observed peaks (and hence a drop in the chemical contents on the surface). This provides confirmation that the most active wear mechanism was adhesive wear, caused by oil starvation between the specimens.

Effects of rotational speed on the friction and wear properties of the bio-oil with nano-additives
The following tests were conducted with bio-oil containing 1% nano-La 2 O 3 under 100 N for 30 min. Fig. 13 shows a weak correlation between the WSD and friction coefficient with rotational speed. The friction coefficient increased accordingly with speed from 1250 rpm to 1650 rpm. This was caused by the combined action of the nano-La 2 O 3 and the many organics within the polar groups in the bio-oil [28]. These polar groups are adsorbed on to the contact surfaces leading to good frictional properties. These are further enhanced by the bearing effect derived from the ball-like nano-La 2 O 3 particles. The increased rotational speed hindered the adsorption of these polar groups [29] leading to accelerated deformation of the nano-La 2 O 3 . The aggregated and deformed nano-La 2 O 3 becomes abrasive debris. As a result of this, the friction coefficient increases.
The optical micrographs of the worn surfaces of the lower specimens from the rotational speed tests are shown in Fig. 14. These again display very similar characteristics to those already described above (from Figs. 9a and 10a) which also serves to confirm the experimental reliability. The diameter and deformation of wear scar increased with speed. The dimples observed at the highest rotational speed (Fig. 10c), may have been due to a ploughing effect induced by the deformed La 2 O 3 as abrasive particles [30]. Fig. 15 shows the EDS spectra of the same worn surfaces. The elemental patterns indicate that carbon is reduced on the worn surfaces whilst elemental Lanthanum increases. Increasing the rotational speed contributed to the deposition of nano-La 2 O 3 on to the surfaces but would also have been unfavourable for the deposition of organics. The XPS spectra shown in Fig. 16 provide an explanation of the source of the Lanthanum detected on the worn surface. This indicated that it was derived from the La 2 O 3 additive and that the La deposit increased with speed. However, the residue from the tribo-film created by the bio-oil was significantly reduced, which is in agreement with the results presented above.

Conclusions
This paper has described a study of the tribological properties of bio-oil enhanced with a La 2 O 3 suspension. Rotational speed, normal load, particle size (micro and nano particles) and concentration (between 0 and 2%) of synthesized La 2 O 3 were assessed using a MQ-800 four-ball tribometer. Tribological behaviour was fully characterized using a range of observational, numerical and chemical analysis techniques. The following conclusions can be drawn from this study: a. The nano-La 2 O 3 had better antifriction and antiwear properties than those of m-La 2 O 3 as a lubricating additive of bio-oil. For m-La 2 O 3 , when its content in bio-oil was less than 1%, there was little effect on wear and a minor influence on anti-frictional properties. When the concentration exceeded 1%, the particles tended to aggregate and act as an abrasive between the contact surfaces. For the nano-La 2 O 3 , both the wear and friction coefficient first decreased and then increase with La 2 O 3 concentration, the optimal weight content seemed to be 1%. b. The introduction of nano-La 2 O 3 reduced corrosive wear, sometimes prevalent in of the bio-oil and derived from the acidic components within the fuel. It also provided a stable foundation for transient friction properties with a clear and distinct transition between running-in and steady state conditions. c. For bio-oil containing 1% nano-La 2 O 3 , wear and friction coefficient increased with load and speed. However, the specific wear rate decreased and steady state conditions prevailed under the highest loads and rotational speeds. This was indicative of an  excellent lubrication performance for the bio-oil containing 1% nano-La 2 O 3 for all of the frictional conditions simulated. d. The tribological mechanisms for the nano-La 2 O 3 could be attributed to a combined action between the oil films from biooil being adsorbed on to the contact surfaces and a bearing effect from the ball-like nano-La 2 O 3 particles. Adhesive wear due to lubrication starvation throughout the contact was the predominant wear mechanism under high loads. At high rotational speeds, abrasive wear became the principal mechanism of wear, as a result of agglomeration and deformation of nano-La 2 O 3 particles.