Enhancing the tribological performance of PEEK exposed to water-lubrication by filling goethite (α-FeOOH) nanoparticles

Abstract

The formation of a tribo-film on the counterface plays an important role on the tribological performances of polymer subjected to mixed and boundary lubrication conditions. However, when freshwater is used as a lubricant, the formation of a tribo-film usually is hindered. In order to overcome this disadvantage, PEEK/α-FeOOH nanocomposites were developed and their tribological performances were studied under water lubrication conditions in the present work. It was demonstrated that the inclusion of α-FeOOH nanoparticles (NPs) into the PEEK matrix improves significantly the tribological performance of the matrix. The nanostructures and properties of the tribo-films formed on the steel counterface were comprehensively studied. It was identified that the addition of α-FeOOH NPs promotes the formation of a lubricating tribo-film which covers the entire counterface. Based on the investigations on tribo-films, we deem that the α-FeOOH NPs act as precursors for the dehydration reaction promoting the formation of a tribo-film which consists of α-Fe₂O₃ and transferred PEEK. It is assumed that the enhanced tribological performance is related to the possibly high load-bearing capability and “easy-to-shear” characteristic of the tribo-film.

---

1. Introduction

High-performance polymers and their composites are currently being used as sliding components in numerous applications where low friction and wear are desired. This is mainly associated with their superior physical and chemical performances, such as excellent self-lubricating capabilities, mechanical properties, chemical resistance,^1–4 etc. Owing to the lower cohesive energy of the polymer matrix in comparison to those of coupling materials which are usually ceramics and metals, polymers are apt to transfer onto the friction counterface, leading to the formation of a protecting layer, and thereby significant reduction of friction and wear.⁵ Moreover, with the addition of nanosized functional fillers, complex tribo-chemical reactions can occur during friction and help to form a lubricating tribo-film.^5–8 In these regards, formulated polymer composites can be used as sliding materials at high pv (pressure × velocity) conditions without the presence of external lubricants.^5–7 In addition, in order to solve the problems related to friction and wear under boundary lubrication conditions, it is expected that high-performance polymers will be more and more employed in modern engine and propulsion systems etc.

It should be noted that up to now majority of researches on the tribology of polymer-based materials focus on dry sliding conditions. It was demonstrated by various works that the tribological performance of polymer-based materials showed a close dependence on friction conditions.^9,10 In order to develop high-performance composites for applications under boundary and mixed lubrication conditions, it is of fundamental interest to get insight into the tribological mechanisms of polymer-based materials exposed to liquid mediums.

In light of the increasing ecological demand for protecting water environment, to replace mineral or synthetic oil lubricants in hydropower plants and marine ships by natural water can offer an effective way.^11–14 Nevertheless, water usually shows a low lubricating performance related to the low viscosity and its inherent chemical inertness. Especially when fresh water is used as lubricant, formation of high performing boundary layers is sometimes problematic due to the chemical inertness of the medium.¹¹ Owing to the above mentioned reasons, polymer-based materials attract intensive interests for enhancing the lifespan and reliability of sliding components exposed to water medium. Many studies reported that compared to the sliding under dry sliding conditions, the tribological performances of polymer-based materials can be markedly improved when subjected to water lubrication due to the presence of water on the sliding interface.^11,15,16 However, other researchers report the inverse.^15,17 That is, the presence of water leads to higher wear of polymer. The negative effect of water is mainly ascribed to the hindrance of tribo-film formation as water lowers the interface temperature which is often a prerequisite for tribo-chemical reactions. Moreover, water can remove the wear debris and thus makes the compaction of wear product onto sliding surface difficult.

It was identified by different researchers that the formation of a lubricating tribo-film on the counterface is one of the key factors to governing the friction and wear properties of polymer composites under dry sliding conditions.^{10–12,18,19} The addition of nanoparticles into polymer matrix and fiber reinforced composite promote the formation of a homogeneous and tenacious tribo-film which decrease significantly the friction and wear. Based on comprehensive investigations on the structure and properties of the tribo-films of different friction systems, one gets deep insight into the tribological mechanisms of polymer composites under dry sliding conditions. In mixed and boundary lubrication regimes, solid–solid contacts carry significant and even majority of the load. Although the importance of tribo-film formation for water-lubricated systems was identified,^11,13,20,21 deep understanding on the formation and function mechanisms of the tribo-film is still lacking.

Poly-ether-ether-ketone (PEEK), a high-performance thermoplastic, is regarded as one of the most ideal matrix polymers for formulating high-performance tribo-materials for applications under severe conditions, such as high pv factors, strong corrosion mediums,^7,18,22–25 etc. This is mainly attributed to its high mechanic properties, an inherently high chemical stability and very low moisture and liquid adsorption rates.^7,22–25 In addition, large-scale compounding and manufacturing of its composites filled with multifunction fillers can be carried out based on conventional melt processing techniques, such as extrusion, injection molding,¹⁸ etc.

Goethite (α-FeOOH or α-FeO(OH)) is formed close-packed array of O²⁻ and OH⁻ anions with Fe³⁺ and by far the most widespread iron oxyhydroxide.²⁶ In nature, it is an abundant constituent of terrestrial soils, sediments, and oolitic iron ores and a major weathering product of ferrous silicates.^27–29 To the best of our knowledge, precious studied and applications of α-FeOOH are focusing on its catalytic performance,^30,31 adsorptive properties for heavy metals owing to its strong chemical affinity,^32,33 and as pigments in the building industry³⁴ and inorganic dyes,³⁵ etc. Nevertheless, as we know, the tribological performance of α-FeOOH as bulk materials and functional fillers was rarely reported in literatures. Besides the properties and applications above mentioned, α-FeOOH is a chemically reactive material.³⁶ And it is a suitable precursor for the preparation of iron oxide (Fe₂O₃) by thermal process.^31,37 In these regards, it may be a deal functional filler to promote tribo-film formation on the sliding contact surface exposed to water lubrication, especially under severe working conditions, such as mixed and boundary lubrication conditions.

In this work, α-FeOOH NPs as functional fillers were added into PEEK matrix. Moreover, for comparison PEEK filled with α-Fe₂O₃ NPs were also compounded. The tribological performances of PEEK nanocomposites were investigated under water lubrication conditions. The tribological mechanisms were analyzed based on comprehensive characterizations of tribo-film's structure and properties using Raman spectrum, focused ion beam combined with transmission electron microscope (FIB-TEM) and nano-indentation techniques etc.

2. Experimental

2.1. Materials

The PEEK powder (Victrex® PEEK 450 PF, D50 = 50 μm), α-FeOOH (width: 80–110 nm, length: 800–1200 nm) and α-Fe₂O₃ NPs (width: 8–10 nm, length: 40–80 nm) used in this work were commercially supplied by Victrex LPC (Lancashire, UK), Zibo Liqiang Pigment Co., Ltd. (Shandong, China) and Aladdin Industrial Corporation (Shanghai, China), respectively. It is clearly seen from Fig. 1 that both the α-FeOOH and α-Fe₂O₃ NPs used show nanorod forms.

Fig. 1 SEM and TEM micro-morphologies of α-FeOOH and α-Fe₂O₃ NPs used.

2.2. Specimen preparation

After being rigorously stirred in a high-speed stirrer (FW177, Tianjin Taisite Instrument Co., Ltd) at 6000 rpm for 20 min, the hybrid of PEEK and α-FeOOH NPs was put into a mold with a dimension of 50 × 60 × 60 mm³, which was then compression molded at 380 °C and slowly cooled in the mold. When the temperature reached to 170–180 °C, the PEEK composites (PEEK-Coms) plate was removed from the mold and cooled to room temperature. In this work, four PEEK-Coms, i.e. filling 5, 10, 15 and 20 vol% α-FeOOH NPs, were compounded and designated PEEK/5FeOOH, PEEK/10FeOOH, PEEK/15FeOOH and PEEK/20FeOOH, respectively. In order to compare the tribological performances of PEEK filled with α-FeOOH NPs, neat PEEK and PEEK/10Fe₂O₃ plates were also prepared following the same procedure. The specimens for tribological tests were cut from the compression molded plates with a dimension of 19.05 (±0.1) × 12.32 (±0.05) × 12.32 (±0.05) mm³.

2.3. Tribological tests

Prior to each tribological test, the counterpart ring and all polymer specimens were cleaned in acetone and ethanol ultrasonically for 15 min, respectively. The experiments were performed using a block-on-ring (BOR) apparatus (MRH-3, Jinan Yihua Tribology Testing Technology Co., Ltd). Fig. 2a shows its schematic diagram, which consisted of rotating ring sliding against a stationary specimen at the normal load in the range of 50–400 N with a constant sliding speed, i.e. 0.2 m s⁻¹. The sliding surfaces were submerged in tap water at room temperature. The total sliding duration was 180 min. The counterpart ring, i.e. SUS 316, was a water-resistant austenitic stainless steel with a diameter of 49.22 mm, and its chemical composition were C ≤ 0.08%, Si ≤ 1.00%, Mn ≤ 2.00%, S ≤ 0.03%, P ≤ 0.035%, Cr of 16.0–18.5%, Ni of 10.0–14.0% and Mo of 2.0–3.0%. The surface of the ring was abraded with SiC metallographic abrasive papers and the mean liner roughness R_a was controlled at 0.2–0.3 μm with randomly distributed grooves as exhibited in Fig. 2b.

Fig. 2 Schematic diagram of BOR test apparatus (a) and surface morphology of counterpart ring (b).

The friction coefficient was measured by a linear force transducer and recorded by a statistical software affiliated to the BOR tester. It is noted that the friction coefficients were calculated in two ways as follows: for each friction and wear test, (i) when the friction coefficient curves were steady the average value in the steady stage was taken. (ii) If the friction coefficient curves continuously decreased with the sliding time the value at the end of each test was taken. And the specific wear rate (W_S) defined as the material's volume loss per sliding distance and per load was calculated according to eqn (1).

where L′ is the width of polymer testing specimens (12.32 mm), R is the radius of counterpart ring (24.61 mm), W is the projected width of the wear track (depending on material loss, mm), F is the normal load applied on polymer specimens, and L is the total sliding distance (m).

2.4. Characterization and analysis

Morphologies of the fillers used, i.e. α-FeOOH and α-Fe₂O₃, were observed by field emission scanning electron microscope (FE-SEM, Mira 3 Xmu, Tescan) and high-resolution transmission electron microscope (HR-TEM, Tecnai G² TF20, FEI) with accelerating voltages of 20 and 300 kV, respectively.

The worn polymer specimen surfaces were inspected by FE-SEM after being coated with a thin layer of gold. Comprehensive characterizations of structure and properties of the tribo-films formed on the steel counterfaces were carried out, (i) their morphologies were observed by an optical microscope (Axio Imager A2m, Zeiss) and FE-SEM, respectively; (ii) the compositions were identified using an energy dispersive X-ray spectroscopy (SEM-EDX, Energy 350, Oxford) attached to the FE-SEM, and a laser Raman spectrometer (LabRam HR800, Horiba Jobin Yvon S.A.S.) equipped with in an excitation laser source at 532 nm; (iii) their micro-mechanical properties were investigated by nano-indention tests using a nano-triboindentation device (TI 950, Hysitron Inc.) with a Berkovich indenter. All the tests were carried out at a peak load of 3 mN and a max depth of 200 nm, with a holding time of 5 s at the maximum load to minimize time-dependent creep effects. The loading and unloading rates were both 0.1 nm s⁻¹. Nano-indention tests on each tribo-film were repeated for three times at different locations. Thus, the average values and deviations of E-modulus and hardness were calculated, respectively; (iv) the surface topographies of the counterfaces were evaluated using scanning probe microscope (SPM) imaging scan in the nano-triboindentation device; (v) their mirco/nano-structures were characterized using TEM combined with selected area electron diffraction (SAED) (Tecnai G² TF20 S-TWIN, FEI). The TEM-specimens of the tribo-films on the worn surfaces were prepared by focused ion beam (FIB) machining in a DualBeam SEM/FIB instrument (Quanta 3D FEG, FEI). Before cutting cross-sectional lamellae, the area of interest was coated with a platinum cap layer which was deposited in the FIB by ion beam assisted deposition form a gaseous organic Pt-compound.

3. Results and discussion

3.1. Effect of α-FeOOH on friction and wear performances of PEEK

Fig. 3 shows the friction coefficient tendencies versus time and specific wear rates of neat PEEK and PEEK filled various volume fractions of α-FeOOH NPs at 100 N. It is seen that the friction coefficient of neat PEEK decrease continuously with sliding time. In view of the close relation between friction evolution and tribo-film formation,¹⁰ it seems that a steady tribo-film did not form on the worn surface during the 3 hour sliding, thus leading to a high wear rate (cf. Fig. 3b).

Fig. 3 Friction coefficient tendencies versus time (a) and specific wear rates (b) of neat PEEK and PEEK filled various volume fraction α-FeOOH NPs at 100 N.

The addition of α-FeOOH NPs markedly reduce the friction coefficient and shorten the running-in duration (cf. Fig. 3a). In comparison to neat PEEK, the specific wear rates of the PEEK-Coms are decreased by more than one order of magnitude (cf. Fig. 3b). In the range studied, an enhancement of the nanoparticles fraction leads to a slight decrease of the wear rate. It is assumed that α-FeOOH NPs promote the formation of a lubricating tribo-film. Among the five materials concerned, PEEK/20FeOOH has a shortest running-in period. This might give a hint that the formation speed of a stable tribo-film on the steel counterface is the fastest. This can be the reason why PEEK/20FeOOH gives the smallest friction coefficient and specific wear rate.

Fig. 4a shows SEM morphologies of the worn surface of neat PEEK. Deep grooves and fatigue marks (as indicated by red arrows) are clearly observed. According to a recent work,^11,18 the sliding system lies in mixed lubrication regime under the conditions studied in the present work. In this case, solid–solid contact bears significant load and abrasion exerted by the protruding asperities on the counterface constitute an important wear mechanism. Moreover, the adhesion force between sliding pairs leads to stick-slip motion and stains hardening of the PEEK surface layer involved into friction.^11,18 Thus, fatigue of the surface layer can take place. Whereas, the worn surfaces of the nanocomposites filled with α-FeOOH NPs is obviously smoother than that of neat PEEK with much mitigated ploughing marks (cf. Fig. 4b–d). In addition, no obvious fatigue mark is noticed from the worn surface of the nanocomposites. It is deemed that both the abrasion effect and the adhesion force are much reduced when sliding takes place against the nanocomposites in comparison to the sliding against neat PEEK.

Fig. 4 SEM morphologies of worn surfaces of neat PEEK (a), PEEK/5FeOOH (b), PEEK/10FeOOH (c) and PEEK/20FeOOH (d). Thick arrows indicate the sliding directions.

Fig. 5 shows the optical micrographs of the tribo-films formed on the counterpart ring surfaces after sliding against neat PEEK, PEEK/5FeOOH, PEEK/10FeOOH and PEEK/20FeOOH. As compared to that of neat PEEK (cf. Fig. 5a), the tribo-films obtained after sliding against the nanocomposites cover much larger area on the counterfaces (cf. Fig. 5b–d). Moreover, the coverage rate gradually increases with enhancing the fraction of α-FeOOH NPs in PEEK matrix. In particular, the counterface slid against PEEK/20FeOOH is almost completely covered by the tribo-film. The tribo-film can act as a protective barrier that prevents the direct contact between the sliding surfaces and thus mitigates the severity of abrasion and the adhesion force.^38–40

Fig. 5 Optical micrographs of tribo-films on counterpart ring surfaces after sliding against neat PEEK (a), PEEK/5FeOOH (b), PEEK/10FeOOH (c) and PEEK/20FeOOH (d). Arrows indicate the sliding directions.

Fig. 6 illustrates the SEM micrographs and EDX analyses of the tribo-films formed on the steel counterfaces after sliding against neat PEEK and PEEK/10FeOOH. It is demonstrated that the both tribo-films consist of transferred material and iron oxide. From Fig. 6a, the tribo-film of neat PEEK mainly exists in the roughness grooves on the counterface which are still visible. It is deemed that during the sliding against neat PEEK the steel counterbody was oxidized and the oxidation layer was scraped and finally compacted into the grooves. However, the tribo-film obtained after sliding against the nanocomposite shows a distinctly different morphology. That is, the tribo-film spread not only in the roughness grooves but also on the plateau areas between the grooves (cf. Fig. 6c). Fig. 7 presents the SEM micrograph of the nanocomposite's tribo-film with a lower magnification and distribution maps of elements Fe, C and O. It is evidenced that element O covers almost homogeneously the counterface. It is therefore suspected that the tribo-film of the nanocomposite shows a better load-bearing capability than that of the tribo-film obtained when sliding takes place against neat PEEK.

Fig. 6 SEM micrographs and EDX analyses of tribo-films formed on the steel counterfaces after sliding against neat PEEK (a and b) and PEEK/10FeOOH (c and d). Arrows indicate the sliding directions.

Fig. 7 SEM micrograph of the nanocomposite's (PEEK/10FeOOH) tribo-film with a lower magnification and distribution maps of elements Fe, C and O. Arrow indicates the sliding direction.

In order to further evaluate the contribution of the tribo-film obtained after sliding against PEEK/10FeOOH, a new test was performed with the counterbody ring which is covered with the tribo-film. Fig. 8 compares the friction coefficients and the wear rates of neat PEEK when sliding against original and the used (obtained from sliding with PEEK/10FeOOH) rings. Although the friction coefficient value does not vary significantly, the specific wear rate of neat PEEK is lowered by 48.21% in comparison with the sliding against the original ring. This gives evidence about the positive role of the tribo-film on wear reduction. Nevertheless, it should be taken into account that replenishment and removal of tribo-film take place simultaneously. When the PEEK/10FeOOH specimen is altered by neat PEEK, the tribo-film structure and the tribological characteristics as well are also changed. Moreover, similar to short carbon and glass fibers,³ α-FeOOH nanorods added in the nanocomposite can also play a reinforcing effect and thereby, enhance the load-carrying capability of the polymer and reduce the friction and wear. Therefore, the friction coefficient and wear rate of neat PEEK when sliding against the used ring are still higher than those of PEEK/FeOOH (cf. Fig. 3 and 8).

Fig. 8 Enhancements of tribo-film on friction coefficient (a) and specific wear rate (b) of neat PEEK at 100 N for 180 min.

3.2. Effect of normal load on friction and wear performances of PEEK materials

Load applied on the sliding contact surface is one of significant factors to govern the tribological performances of sliding materials, especially under severe working conditions.^10,41 Fig. 9 compares the friction coefficients and specific wear rates of neat PEEK and PEEK/10FeOOH at various applied loads. An increase of the applied loads from 50 N to 400 N leads to linear enhancements of the friction coefficient and wear rate of neat PEEK. This can be ascribed to the increased solid–solid rubbing owing to the load increment. However, the friction coefficient of PEEK/10FeOOH decreases obviously with increasing the applied loads. It is believed that this phenomenon is connected with the tribo-film on the counterface. Nehme et al.¹⁰ reported that increases interface stressing and temperature can trigger more tribo-chemical reactions. It is noticed from Fig. 10 that the thickness and coverage rate of the tribo-film formed at 400 N are superior to that obtained at 50 N. In this case, the tribo-film can prevent more effectively the sliding surfaces from coming into direct contact, and thus the friction coefficient is decreased. On the other hand, especially for a thick tribo-film, the tribo-film is apt to break down under a high contact load due to enhanced interfacial stressing.¹⁰ The accelerated breakdown and replenishment of the tribo-film can be responsible for the higher wear rate of PEEK/10FeOOH with increasing the applied load (cf. Fig. 9b).

Fig. 9 Friction coefficients (a) and specific wear rates (b) of neat PEEK and PEEK/10FeOOH at various applied loads.

Fig. 10 Counterpart rings and SEM micrographs of the wear tracks generated when sliding against PEEK/10FeOOH at 50 N (a) and 400 N (b). Arrows indicate the sliding directions.

3.3. Discussions on tribological mechanisms

Fig. 11 compares the Raman spectra of α-FeOOH and α-Fe₂O₃ NPs and the worn surface of counterpart ring after sliding against PEEK/10FeOOH at 100 N. It is clear that at the end of sliding abundant α-Fe₂O₃ is present on the ring surface. In addition, a small peak of C is also apparent as consistent with EDX results as described above. It is therefore identified that transferred PEEK and α-Fe₂O₃ constitute the tribo-film. Although oxidation of the steel counterface can result in generation of iron oxide, as mentioned above, the tribo-film of PEEK/FeOOH covers much more area on counterface than that of PEEK. Therefore, we deem that at least large quantity of the iron oxide in the tribo-film of PEEK/FeOOH derives from dehydration of α-FeOOH (eqn (2)).

Fig. 11 Raman spectra of neat α-FeOOH and α-Fe₂O₃ NPs and worn surface of counterpart ring after sliding against PEEK/10FeOOH at 100 N.

On the other hand, FIB-HRTEM was employed to characterize the nano-structure of the tribo-film formed on the steel counterface. Fig. 12a shows the overview TEM micrograph of the cross-section of the tribo-film formed on the counterface after sliding against PEEK/10FeOOH at 100 N. It is identified that a tribo-film covers the entire counterface. HRTEM analyses reveal that the tribo-film mainly consists of crystals and amorphous C-matrix deriving from transferred PEEK. Owing to cyclic mechanical interactions semi-crystalline PEEK becomes finally amorphous during the sliding process (bright contrast in Fig. 12a and b). According to inverse FFT (not given here) and SAED analyses (Fig. 12f), evident interplanar spacing is noticed, i.e. 0.368, 0.252, 0.221, 0.184, 0.169 and 0.149 nm, which correspond to the lattice planes of Fe₂O₃, i.e. [012], [110], [113], [024], [116] and [214] (JCPDS no. 33-0664). Thus, it is confirmed that polycrystalline Fe₂O₃ do form during the rubbing process as a result of tribo-chemical reactions. It is deemed that material transfer and tribo-chemical reaction occur simultaneously and transferred PEEK and the product of tribo-chemcial reaction i.e. α-Fe₂O₃, are mixed during friction process.

Fig. 12 Overview TEM micrograph of cross-section of tribol-film formed on counterface after sliding against PEEK/10FeOOH at 100 N (a); (b)–(e) high magnification images of the marked areas in (a), i.e. zone I (b) and (c), II (d) and III (e), respectively; (f) SAED of (e).

In order to get a deeper insight into the tribological mechanisms of α-FeOOH NPs, the friction and wear performances of PEEK/10Fe₂O₃ were examined using the same BOR tester under the same running conditions. Fig. 13 shows the average friction coefficients and specific wear rates of PEEK/10Fe₂O₃ as a function of applied load. It is surprising that in the range studied the friction coefficients and wear rates are obviously higher than those of PEEK/10FeOOH, even higher than those of neat PEEK (cf. Fig. 9 and 13).

Fig. 13 Average friction coefficients and specific wear rates of PEEK/10Fe₂O₃ as a function of applied load.

From Fig. 14, the worn surfaces of PEEK/10Fe₂O₃ are much rougher due to severe abrasion and fatigue damage. Inspections on the wear track reveal that a tribo-film was also formed on the steel counterface. It is observed from Fig. 14 that the original roughness grooves on the counterface are fully filled and a patchy tribo-film covers local regions on the counterface. On the other hand, when comparing the surface topographies of the tribo-films formed on counterfaces after sliding against PEEK/10Fe₂O₃ and PEEK/10FeOOH (Fig. 15), it is found that the surface of tribo-film from PEEK/10FeOOH is much smoother than that of PEEK/10Fe₂O₃ and their root square roughness (R_q) values are 66.45 and 191.61 nm, respectively. Bahadur⁵ pointed out that the formation of a patchy tribo-film could be indicative that the adhesion between the tribo-film and the steel surface was weak. In this regard, it is suspected that when stable iron oxide, i.e. α-Fe₂O₃, is used as fillers, tribo-chemical reaction may not be the prevailing mechanism leading to tribo-film formation. In this case, the tribo-film is formed owing to physical and mechanical effects such as van der Waals force and mechanical compaction etc. It is manifested that tribo-chemical reactions can lead to the formation of a lubricious and stable tribo-film.^5–8 Friedrich et al.³ concluded that the tribological performance of polymer composites can be enhanced when the products of tribo-chemical reactions increases the bonding between the tribo-film and counterface, whereas some fillers lower the wear resistance if they lead to the formation of a non-uniform tribo-film. It is deemed that, the tribo-chemical reactions owing to the inclusion of α-FeOOH NPs play an important role on the formation of a stable tribo-film and the enhancement of tribological performances of PEEK matrix under boundary and mixed lubrication conditions.

Fig. 14 SEM micrographs with low and high magnification of the worn surfaces of PEEK/10Fe₂O₃ and counterpart ring obtained after sliding at 100 N. Arrows indicate the sliding directions.

Fig. 15 2D- and 3D-topographies of tribo-films formed on counterfaces after sliding against PEEK/10FeOOH and PEEK/10Fe₂O₃.

Nanoindentation is a promising technique to investigate the micro-mechanical performance of tribo-films formed on the counterface.⁴² Fig. 16 shows the representative force–displacement curves of nanoindentation tests with the tribo-films formed after sliding against PEEK/10Fe₂O₃ and PEEK/10FeOOH at 100 N, respectively. It is clearly observed that both the elastic modulus (E_r) and hardness of the tribo-film formed by PEEK/10Fe₂O₃ are much higher than those of the tribo-film formed by PEEK/10FeOOH. From Fig. 12a, the Fe₂O₃ produced owing to tribo-chemical reaction of FeOOH NPs tends to aggregate in the amorphous PEEK matrix. That is, the product of tribo-chemical reactions is not evenly dispersed in the tribo-film even under the role of interfacial mixing. Such architecture can impart the tribo-film a much higher lubricating action than that formed owing to only material transfer and mechanical compaction. We assume that the “soft” tribo-film of PEEK/10FeOOH has an “easy-to-shear” characteristic and therefore it shows a lubricous capability. This is deemed as the main reason why the nanocomposite filled with α-FeOOH shows a superior tribological performance (cf. Fig. 3 and 13).

Fig. 16 Representative force–displacement curves of nano-indentation tests with the tribo-films formed after sliding against PEEK/10Fe₂O₃ and PEEK/10FeOOH at 100 N.

4. Conclusions

Intensive efforts have been made to enhance the tribological performance of PEEK exposed to water-lubrication by filling goethite (α-FeOOH) NPs and to understand the relevant tribological mechanisms. Based on the results of tribological tests and tribo-film characterizations with SEM-EDX, Raman spectra, nanoindentation and FIB-TEM, following conclusions can be drawn:

(a) The inclusion of α-FeOOH NPs into PEEK matrix leads to the formation of a stable and lubricating tribo-film on the steel counterface. The tribo-film can show a superior load-bearing capability and an “easy-to-shear” characteristic, thus enhance the tribological performances of PEEK exposed to water lubrication.

(b) The formation speed of a stable tribo-film and its coverage rate on the steel counterfaces is dependent on the volume fraction of α-FeOOH NPs in the PEEK-Coms. PEEK/20FeOOH has a shortest running-in period and the smallest friction coefficient and specific wear rate.

(c) The distinct effect of applied normal loads on the tribological performances of neat PEEK and the nanocomposites is connected with the tribo-film on the steel counterface.

(d) Tribo-chemical reactions play an important role on the formation of a stable tribo-film on the sliding counterface. Different from that of stable α-Fe₂O₃ NPs, the chemical reactivity of α-FeOOH NPs is of great importance for the occurrence of tribo-chemical reactions and the formation of a stable, lubricating, “soft” tribo-film on the steel counterface. Hence, α-FeOOH NPs are identified to be promising fillers for enhancing the tribological performance of polymer matrix under boundary and mixed lubrication conditions.

Acknowledgements

The authors would like to thank the financial support from National Natural Science Foundation of China (Grant No. 51475446), Chinese “Thousand Youth Talents Plan” Project and Project Funded by Gansu Postdoctoral Science Foundation. Moreover, the authors also acknowledge Dr Lijun Xu (Institute of Modern Physics, Chinese Academy of Sciences, China), Dr Chaoshuai Guan (School of Physical Science and Technology, Lanzhou University, China) and Dr Shilong Lv (Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, China) for their help about Raman spectrum, SEM, TEM analyses and FIB cutting, respectively.

References

1. G. Y. Xie, G. X. Sui and R. Yang, Compos. Sci. Technol., 2011, 71, 828–835.
2. G. Zhang, L. Chang and A. K. Schlarb, Compos. Sci. Technol., 2009, 69, 1029–1035.
3. K. Friedrich, Z. Zhang and A. K. Schlarb, Compos. Sci. Technol., 2005, 65, 2329–2343.
4. W. Österle, A. I. Dmitriev, B. Wetzel, G. Zhang, I. Häusler and B. C. Jim, Mater. Des., 2016, 93, 474–484.
5. S. Bahadur, Wear, 2000, 245, 92–99.
6. S. Bahadur, Q. Fu and D. Gong, Wear, 1994, 178, 123–130.
7. Y. Yamamoto and T. Takashima, Wear, 2002, 253, 820–826.
8. T. Saito and F. Honda, Wear, 2000, 237, 253–260.
9. G. Zhang, C. Zhang, P. Nardin, W. Li, H. Liao and C. Coddet, Tribol. Int., 2008, 41, 79–86.
10. G. Nehme, R. Mourhatch and P. B. Aswath, Wear, 2010, 268, 1129–1147.
11. C. Gao, G. Guo, F. Zhao, T. Wang, B. Jim, B. Wetzel, G. Zhang and Q. Wang, Tribol. Int., 2016, 95, 333–341.
12. A. Golchin, K. Friedrich, A. noll and B. Prakash, Wear, 2015, 328–329, 456–463.
13. A. Golchin, K. Friedrich, A. noll and B. Prakash, Tribol. Int., 2015, 88, 209–217.
14. J. Wang, B. Chen, N. Liu, G. Han and F. Yan, Composites, Part A, 2014, 59, 85–92.
15. B. Chen, J. Yang, J. Wang, N. Liu and F. Yan, Tribology Transactions, 2014, 58, 140–147.
16. B. Chen, J. Wang and F. Yan, Tribol. Int., 2012, 52, 170–177.
17. J. Jia, J. Chen, H. Zhou, L. Hu and L. Chen, Compos. Sci. Technol., 2005, 65, 1139–1147.
18. G. Zhang, B. Wetzel and Q. Wang, Tribol. Int., 2015, 88, 153–161.
19. W. Zhai, X. Shi, Z. Xu and Q. Zhang, Mater. Chem. Phys., 2014, 147, 850–859.
20. R. S. Gates, M. Hsu and E. E. Klaus, Tribology Transactions, 2008, 32, 357–363.
21. N. Liu, J. Wang, B. Chen, G. Han and F. Yan, Mater. Des., 2014, 55, 805–811.
22. Z. Rasheva, G. Zhang and T. Burkhart, Tribol. Int., 2010, 43, 1430–1437.
23. J. P. Davim and R. Cardoso, Wear, 2009, 266, 795–799.
24. Y. Yamamoto and M. Hashimoto, Wear, 2004, 257, 181–189.
25. M. Zalaznik, M. Kalin and S. Novak, Tribol. Int., 2016, 94, 92–97.
26. Y. Wang, Y. Gao, L. Chen and H. Zhang, Catal. Today, 2015, 252, 107–112.
27. P. Zarzycki, S. Kerisit and K. M. Rosso, J. Phys. Chem. C, 2015, 119, 3111–3123.
28. M. Usman, M. Abdelmoula, P. Faure, C. Ruby and K. Hanna, Geoderma, 2013, 197–198, 9–16.
29. Y. Yang, W. Yan and C. Jing, Langmuir, 2012, 28, 14588–14597.
30. D. A. Jadhav, A. N. Ghadge and M. M. Ghangrekar, Bioresour. Technol., 2015, 191, 110–116.
31. R. Sahu, B. J. Song, Y. P. Jeon and C. W. Lee, J. Ind. Eng. Chem., 2015, 35, 115–122.
32. M. Ma, H. Gao, Y. Sun and M. Huang, J. Mol. Liq., 2015, 201, 30–35.
33. A. Jaiswal, S. Banerjee, R. Mani and M. C. Chattopadhyaya, J. Environ. Chem. Eng., 2013, 1, 281–289.
34. M. A. Legodi and D. D. Waal, Dyes Pigm., 2007, 74, 161–168.
35. N. Guskos, G. J. Papadopoulos, V. Likodimos, S. Patapis, D. Yarmis, A. Przepiera, K. Przepiera, J. Majszczyk, J. Typek, M. Wabia, K. Aidinis and Z. Drazek, Mater. Res. Bull., 2002, 37, 1051–1061.
36. J. F. Boily, J. Phys. Chem. C, 2012, 116, 4714–4724.
37. J. D. Betancur, C. A. Barrero, J. M. Greneche and G. F. Goya, J. Alloys Compd., 2004, 369, 247–251.
38. X. Fan, L. Wang, W. Li and S. Wan, ACS Appl. Mater. Interfaces, 2015, 7, 14359–14368.
39. C. Gao, Y. Wang, D. Hu, Z. Pan and L. Xiang, J. Nanopart. Res., 2013, 15, 1502.
40. S. Bec, A. Tonck, J. M. Georges and G. W. Roper, Tribol. Lett., 2004, 17, 797–809.
41. B. Kim, J. C. Jiang and P. B. Aswath, Wear, 2011, 270, 181–194.
42. L. Chang, K. Friedrich and L. Ye, J. Tribol., 2013, 136, 021602.

---