Superlubricity induced by partially oxidized black phosphorus on engineering steel

Macroscale superlubricity has attracted increasing attention owing to its high significance in engineering and economics. We report the superlubricity of engineering materials by the addition of partially oxidized black phosphorus (oBP) in an oleic acid (OA) oil environment. The phosphorus oxides produced by active oxidation exhibit lower friction and quick deposition performance compared to BP particles. The H-bond (-COOH⋯O-P, or -COOH⋯O=P) formed between P-O bond (or P=O) and OA molecule could benefit the lubricating state and decrease the possibility of direct contact between rough peaks. The analysis of the worn surface indicates that a three-layer tribofilm consisting of amorphous carbon, BP crystal, and phosphorus oxide forms during the friction, which replaces the shear interface from the steel/steel to carbon—oBP/carbon—oBP layer and enables macroscale superlubricity.


Introduction
Friction is a part of our daily lives and consumes large amounts of energy, accounting for approximately one-third of the global disposable energy [1]. Research indicates that 80% of equipment failures are caused by friction [2]. In China, friction-related energy and material consumption could reach 4.5% of the GDP. Against this background, superlubricity, which presents a near-zero friction state, is attracting increasing attention [3,4]. In the past two decades, two-dimensional (2D) materials [5,6], including graphene [7,8], molybdenum disulfide [9,10], black phosphorus [11][12][13][14], and boron nitride [15,16], have presented excellent potential to promote lower friction to superlubricity regimes, owing to their weak shear strength of the interlayer, desirable structural orders, or orientations. Dispersing these 2D materials into specific liquids like water, and lubricating oil widens their scope of application in superlubricity compared to their current strict limitations in the application of solid lubrication only in atmospheric and vacuum conditions [12,17,18]. Among these applications, research toward practical industrial applications will be the final aim of superlubricity research, implying that engineering materials such as stainless steel and 52100 steel, would be the better tribo-pair for superlubricity research, rather than Si 3 N 4 /sapphire, Si 3 N 4 /SiO 2 , or diamond-like carbon (DLC) [19,20].
In this study, superlubricity in a 52100 steel tribo-pair is realized using black phosphorus (BP). As a novel 2D material, BP has attracted increasing attention in the fields of electricity, optics, mechanics, and tribology owing to its high carrier mobility (~1,000 cm 2 ·V −1 ·s −1 ), light response, high hardness in the zigzag direction, and anisotropic lamellar structure [21][22][23]. Notably, BP powders are sensitive to oxygen, water, and light under atmospheric conditions, which results in degradation and providing disadvantageous for its use in these fields. Therefore, various methods www.Springer.com/journal/40544 | Friction for passivation and protection of BP have been proposed, including packaging [24], functionalization [25], liquid phase protection [26], and doping [27]. Subsequent tribological tests have shown that the oxidation of BP can favor its friction performance. Specifically, the abundant P-OH and P=O bonds formed on the oxidized surface can absorb water molecules at the friction interface, which is beneficial for the formation of an easy-shear film, finally achieving superlubricity in water-based lubrication [17,28]. Inspired by this interesting oxidation property, a BP oil-based superlubricity is designed. This is different from the easily sheared water layer formed in the friction contact area in water-based lubrication, and the ability of P-OH and P=O bonds to absorb water will be weakened under oil lubrication. The amphoteric organic molecule oleic acid (OA) is considered a potential lubricating oil since it can not only form hydrogen bonds with P-OH bonds but also form an excellent tribological carbon-based layer under the catalysis of BP [13,29]. Tang et al. [30] reported superlubricity achieved by BP powders on engineering steel. In this study, we investigate the effect of BP oxidation, which is inevitable in the preparation and the friction process. A hydrogen peroxide (H 2 O 2 , HP) solution is added to actively oxidize the BP crystals. The friction performance of BP and partially oxidized BP (oBP) is explored. This paper extends the application range of oBP from water-based to oil-based superlubricity and provides deeper understanding of the superlubricity performance of oxidized BP.

Sample preparation
Pristine BP powders were prepared using a high-energy ball-milling method with red phosphorus as a raw material [13]. Figure 1 shows a schematic diagram of the active oxidation of BP and the oil sample preparation process. The active oxidation of BP was completed in absolute ethanol, and a certain ratio (1/3.75) of the BP/HP solution was set. The transformation of the morphology, structure, and surface bonding behavior of BP after oxidation was analyzed using scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS) mapping spectra, transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific Inc.). After 30 min of oxidation, the BP ethanol solution was mixed with OA (90 wt%, purchased from Adamas) and then vapored ethanol at 90 °C under a vacuum environment to obtain the oBP-OA oil sample.

Friction test and analysis
Before the friction tests, the oBP-OA and BP-OA oils were sonicated (ultrasonic cell pulverized at 600 W for 1 h in an iced bath environment) and centrifuged (1,500 rpm, 10 min) to obtain a well-dispersed sample. The friction test was performed using a ball-on-disk tribometer (UMT5, Bruker) with a rotation speed of 300 rpm (94.2 mm/s) and a load of 1 N (the corresponding maximum Hertz contact stress: 660 MPa). The entire test conducted in an atmospheric environment (room temperature, 40-50 RH%). The standard ball (Ra 0.02 μm) and disk (Sa ~0.006 μm) composed of the common engineering material 52100 steel (AISI) were used for the tribo-pair. The hardness of ball is 62-64 HRC and the disk has undergone heat treatment whose hardness could reach 53.6 HRC. After the test, the morphology, profile, and bonding behavior of the worn surface and debris were analyzed using a 3D white light interferometer, SEM, TEM, and XPS. In addition to the UMT friction test, atomic force microscope (AFM, Asylum Research Cypher) was used to explore the influence of oxidation on its morphology and friction behavior.

Characterization of pristine BP and oBP
The morphology and structure of the BP and oBP  | https://mc03.manuscriptcentral.com/friction particles (samples prepared before the OA addition step) are presented in Figs. 2 and S1 in the Electronic Supplementary Material (ESM). The bulk BP without active oxidation presents a clear outline, as shown in the SEM images (Figs. 2(a) and S1(a) in the ESM). In addition, an obvious accumulation phenomenon of small particles can be observed on the surface of bulk BP. As for the oBP, the biggest difference in morphology is the pores (Fig. 2(b)) that appeared on the surface. The O and P elemental mapping of the oBP sample (Figs. 2(b1) and 2(b2)) confirmed such a porous structure and revealed that the BP surface was covered by a layer of phosphorus oxide. As the solvent could be adsorbed by this porous BP surface, under the action of the high-energy electron beam, this adsorbed solvent could volatilize, which would turn the shape of the oBP surface from a solid block to the spread honeycomb, as shown in Figs. 2(b) and S1(b) in the ESM.
The TEM images and EDS results in Figs. 2(c), 2(c1), www.Springer.com/journal/40544 | Friction and 2(c2) provide more structural details about the oBP samples. Porous structures caused by oxidation and thin 2D films were observed. In addition, large amounts of BP particles confirmed by the crystalline plane (021)) with sizes within 100 nm were doped on the BP film (Figs. 2(c3) and 2(c4)). Considering that small particles also existed in the pristine BP (Figs. 2(d1) and 2(d2)), it is believed that these small particles are produced by the mechanical ball-milling process and the active oxidation process cannot completely remove these small BP particles. Such a small size plays a crucial role in reducing the friction force. As for the BP samples, besides the small particles, thin 2D films still could be observed in the TEM image ( Fig. 2(d)).
In contrast, no porous structure could be found on the BP surface, which on the other hand, reflects the effect of active oxidation on the morphology and structure of BP. While the morphology and structure of BP change with the effect of active oxidation, the surface chemical bonding mode also changes. The XPS results of the pristine BP and oBP particles (Figs. 2(e) and 2(e1)) reveal this transformation process. The P 2p spectrum can be divided into two groups: the BP crystal bond and the phosphorus oxide bond. The former P-P bond consists of the P 2p 1/2 peak and P 2p 3/2 peak located at 129.5 and 130.4 eV, while the phosphorus oxide bond contains the O-P=O peak and dangling P=O peak located at 134.1 and 135.1 eV [17,28]. For pristine BP powders, both P-P bond and phosphorus oxide bonds could be identified, which indicates that the pristine BP powders were partially oxidized, which may be due to the ambient exposure during the XPS sample preparation and storage process. The relative intensity of the P-P peak is higher than that of the phosphorus oxide peak. However, after the addition of HP, the relative intensity of the P-P bonds of the oBP sample decreased and strong phosphorus oxide bonds (O-P=O bond and dangling P=O bond) dominate the oBP surface. The O 1s spectrum contains two peaks, including the dangling P=O bonds at 531.6 eV and the bridging P-O-P bonds at 533.2 eV. The proportion of these two bonds changed with the active oxidation. For pristine BP, the ratio of P-O-P and P=O bonds (calculated by the ratio of peak area) was 1.56. With the addition of HP, the proportion gradually increased to 2.28 (oBP). It has been reported that oxygen is easily bonded within the P-P bond to form a P-O-P bond with a limited oxygen supply [28,31]. Thus, with the addition of a limited HP solution, the proportion of P-O-P bonds shows an increasing trend.
From the above results, it is concluded that the active oxidation by HP solution not only causes the appearance of a pore structure on the BP surface, but also causes the surface to be covered by more phosphorus oxides. The change in the structure and surface bonding state affect its friction performance.

Macroscale friction test
The macroscale tribological performances of the BP and oBP powders suspended in OA are shown in Fig. 3(a). In general, the addition of BP or oBP powders is conducive to a decrease in coefficient of friction (COF), compared to OA lubrication. For example, the COF of the BP-OA sample stabilized at 0.029 after 9,600 s of friction, which was 70% lower than that of OA (COF: 0.0733). The oBP even gives rise to a greater decrease in the COF value. After 8,600 s of friction, the COF reached the range of 0.005-0.008, entering the superlubricity state. Notably, the COF of the oBP-OA samples showed a downward trend in the initial stage of the test, and then increased rapidly, which is completely different from the rapidly increasing trend of OA and BP-OA samples in the initial stage. The 3D topographic images and optical images of the wear scars are shown in Figs. 3(b) and S2 in the ESM. The addition of BP or oBP resulted in excellent lubricating performance, reducing scratches on the ball surface. As for the ball surface of oBP-OA, it is obvious that the wear scar is surrounded by a black tribofilm, which consisits of C, O, and P elements, as confirmed by the EDS results shown in Fig. S3 in the ESM. In sharp contrast, there is no generation of a black tribofilm around the wear scar lubricated by OA or BP-OA, which indirectly reflects the advantages of oBP-OA oil samples in forming tribofilms. It is believed that the downward trend in the initial stage was caused by the rapid deposition of phosphorus oxides on the friction surface, forming a deposited  film to reduce the rough peak shear behavior under high-stress contact conditions at the beginning of the oBP-OA test. The width of the wear track decreased by 34.8% and 40%, due to the addition of BP and oBP, as shown in Figs. 3(c) and S4 in the ESM, respectively. The depth of the wear track also decreased by 38.3% for BP as compared with OA. Interestingly, for the oBP, a bump with a height of ~90 nm appeared in the wear track area, insteading of a hollow.
Time-of-flight secondary ion mass spectrometry (TOF-SIMS) was performed on the wear scars of BP-OA and oBP-OA to explore surface chemical composition and depth profile. Considering the relative sensitivity factor (RSF), the negative mode was chosen to identify the fragments of carbon oxide and phosphorus oxide. The TOF-SIMS mass spectra of anions derived from the wear scar lubricated by BP-OA and oBP-OA oil samples are shown in Fig. 4. It was found that the types of anionic fragments derived from the two samples were the same, with both including hydrocarbon fragments, carbon-oxygen fragments, and phosphorus-oxygen fragments. Because of the active oxidation, the phosphorus oxide remaining on the oBP-OA wear scar is higher than that on the BP-OA wear scar, as confirmed by the higher intensity of phosphorus-oxygen fragments (PO -, PO 2 -, HP 2 O -).
According to the TOF-SIMS elemental maps (Fig. 5), a large number of phosphorus oxides accumulate around the contact area of oBP-OA, while for BP-OA, these phosphorus oxides are more concentrated in the friction contact area. In addition, the distribution of anion fragments related to carbon on the two friction surfaces also showed great differences. As one of the representatives of OA decomposition products, the C 2 H 2 fragment presents a higher intensity in the oBP-OA friction area, however, no obvious aggregation phenomenon could be found on the BP-OA wear scar. In addition to analyze the difference in surface composition, we explored the distribution of anion fragments in the profile, and the results are shown in Fig. 6. As for Cand C 2 H 2 -, the intensities present a similar downward trend, indicating that the carbon film is mainly concentrated in the top zone of the tribofilm. It is worth noting that the content of the C 2 H 2 fragment in the oBP-OA top zone is slightly higher than that in the BP-OA top zone, as proven by the higher intensity. Considering that the decomposition products of oleic acid are more concentrated in the friction area, it is believed that such a high-content and high-concentration distribution state will provide a better friction reduction effect. The elemental distribution, surface bonding behavior, and composition of the BP-OA and oBP-OA wear track areas were characterized, as shown in Fig. 7. The wear track area of BP-OA was covered by a layer of O and P (Fig. 7(a)). In contrast, the O, P, and C contents in the area around the wear track of oBP-OA were significantly higher than those of the wear track ( Fig. 7(b)), similar to the ball surface. The reason for this phenomenon is the formation of the deposition layer, which can be observed in the SEM image ( Fig. 7(b)). XPS characterization of the wear track area of BP-OA and oBP-OA was performed to explore the influence of active oxidation on the wear surface bonding behavior. The C 1s, O 1s, and P 2p spectra shown in Fig. 7(c) are derived from the wear track area of BP-OA, while the spectra of Fig. 7(d) originate from the wear track area of oBP-OA. The C1s peak in Fig. 7(c) can be fitted by two peaks located at 284.7 and 286 eV, corresponding to C-C (C-H) and C-O (C-O-P). In comparison, on the surface of oBP-OA, the new peak of C=O appears at 288.9 eV, although its content is very small (3.9%) [29,32]. In addition, the ratio of the C-O/C-O-P bond on the surface of oBP-OA was doubled compared with that on the surface of BP-OA, reaching 36.5%, which indicates that a higher concentration of oxide layer was formed [12]. The O 1s spectra derived from the BP-OA and oBP-OA surfaces share the same bonding type, both consisting of Fe 2 O 3 , P=O, and P-O-C (P-O-P) bonds located at 530.6, 531.7, and 533.5 eV, respectively. The  | https://mc03.manuscriptcentral.com/friction ratio of the Fe 2 O 3 peak is higher than that on the surface of oBP-OA, which indicates that the metal substrate has a higher degree of oxidation and the friction condition is worse. The P 2p spectra on the surface of BP-OA are composed of two bonds, including the POx/P-O-C bond and P-P bond located at 133.6 and 129.5 eV [29]. No P-P bond could be detected on the surface of oBP-OA due to active oxidation. The existence of P-O-C bonds in both C 1s, O 1s, and P 2p indicate a reaction between the phosphorus oxide and the decomposition products of OA and it is critical for linking the phosphorus and carbon layers.
We performed focused ion beam (FIB) etching and TEM on the cross-sectional area of the wear track area and the non-wear area of the oBP-OA to analyze the influence of friction on the film structure. Three layers with obviously different morphologies and structures could be identified, as shown in Fig. 7(e), with a total thickness of approximately 66 nm. Combined with the cross-sectional morphology, structure, and element distribution, it can be determined that the inner amorphous layer near the metal substrate (thickness ~11.4 nm) is composed of phosphorus oxide, which is attributed to the deposition of oBP. The www.Springer.com/journal/40544 | Friction middle layer (thickness ~30.9 nm) consists of an amorphous carbon layer with black phosphorus quantum dots which is confirmed by diffraction patterns. The closer to the top layer, the higher the concentration of BP quantum dots could be detected, showing a concentration diffusion trend (enlarged image of Area 1). According to the distribution of the elements, the top layer of the tribofilm is composed of phosphorus oxide and amorphous carbon, which is consistent with the XPS results. In addition, the top layer (thickness 22.4 nm) was mixed with obvious BP crystals (enlarged TEM image of Area 2). In contrast to the tribofilm formed on the worn surface, the oBP dispersed in OA can naturally form a uniform phosphorus oxide layer on the surface of the non-wear metal substrate, which has the same structure as the internal phosphorus oxide layer near the metal substrate in the friction area, but with a larger thickness (~85.5 nm), as shown in Fig. S5 in the ESM. Thus, it is logical to speculate that in the initial stage of friction, the friction surface is covered by a phosphorus oxide layer formed by the natural deposition of oBP, and the friction stress would decrease the thickness of this layer and induce the formation of a new tribofilm on this layer, finally forming a three-layer structure.

Microscale friction test
The AFM lateral force mode was used to study the friction performance of phosphorus oxide in terms of microscale sight. Considering the TEM results that the BP-containing top layer formed on the surface of an amorphous carbon layer, the AFM tip covered with an amorphous carbon layer was equipped to slide against the various BP surfaces. The fresh BP surface was produced by micromechanical exfoliation and then placed in the atmosphere for several days. The SEM, AFM height, and corresponding friction images of the BP surface are shown in Fig. 8. All height and friction images were obtained by applying a normal force of 4.85 nN.
As for the fresh exfoliated BP surface, no oxidation micro bumps could be observed in the SEM (Fig. 8(a)) or AFM height image (Fig. 8(a1)). After being placed in the air for 5 h (25 °C, 50-60 RH%), a large number of blobs appeared and spread over the entire BP surface, as shown in the height image shown in Fig. 8(b).  Under the continuous action of oxygen, water, and light, these blobs gradually grow and are closely distributed on the BP surface (Figs. 8(c)-8(f)), which promotes the generation of a dense micro-bump network. The micro-bump was 1 nm higher than the bottom flat, proved by the height profile shown in Figs. 8(e3) and 8(f2). According to our previous studies, these micro bumps consisit of phosphorus oxides owing to degradation and could favor the lubrication between tribo-pairs. The latter friction image of the BP surface after 7 and 8 days shown in Fig. 8(e2) and 8(f1) confirmed such a favorable effect on the friction force between the micro-bump area and the surrounding area. The friction force of the bump area after 8 days was approximately 12.5 nN, which was 25% smaller than that of the surrounding area.

Mechanism
Owing to the weak bonding force between the 2D material layers, the addition of BP crystals in the oil can undoubtedly reduce the friction coefficient by up to 70%, as confirmed by the friction test results. Meanwhile, the active oxidation of BP could favor its lubricating performance and further reduce the COF to superlubricity. The characterization of pristine BP and oBP revealed that part of the P-P bonds could bond with oxygen to form P-O-P and P=O bonds under the effect of the HP solution, which confirms the transformation from the BP structure to phosphorus oxides ( Fig. 9(a)). The AFM results (Fig. 8) also indicate that the BP surface could produce micro bumps composed of phosphorus oxides due to oxidation when exposed to the atmosphere. The phosphorus oxides present several interesting properties, and the first is the low friction itself. Compared to the fresh BP surface, the phosphorus oxide areas exhibit lower friction, with a reduction of up to 25%, as confirmed by the AFM results. The second property is the ability to deposit on the steel surface. The TEM results of the noncontact metal matrix shown in Fig. S5 in the ESM confirmed the existence of a deposition layer with a thickness of more than 80 nm. In comparison, no obvious deposition layer was found on the surface of the BP-OA. Such a deposition layer composed of phosphorus oxides would work synergistically with the BP crystal to provide initial low friction (down to 0.02), as shown in Fig. 9(b). Meanwhile, owing to the H-bond (-COOH···O-P, or -COOH···O=P) formed between P-O bond (or P=O) with the OA molecule [30], the existence of a phosphorus oxides layer could facilitate the adsorption of OA lubricating oil in the contact area, which benefits the lubricating state and decreases the possibility of direct contact between the rough peaks. However, the subsequent increment of COF indicates that such a natural deposition layer is easily damaged by the friction shear force (Fig. 3(a)), even when supported by BP crystals. According to the TEM results of the cross-sectional area of the wear track, the tribofilm after achieving superlubricity consists of three layers: the inner deposited phosphorus oxide layer, the middle carbon layer, and the top mix layer. The P-O-C bond created in the friction process could link the carbon and phosphorus layers, which promotes the formation of a unique three-layer structure. The result of TOF-SIMS on the ball surface also shows that the friction contact area is covered by a layer of carbon and phosphorus oxide. The carbon layer is formed by a stack of hydrocarbon fragments which are the decomposition products of OA during friction, as shown in Fig. 9(a). Combining the surface information, it is believed that the direct contact of the metal rough peaks is replaced by the sliding contact between the tribofilm (Fig. 9(c)), which is the key factor in realizing superlubricity.

Conclusions
Macroscale superlubricity was achieved by oxidized black phosphorus (oBP)-oleic acid (OA) suspensions on engineering steel, which confirms that the active oxidation of BP could favor its friction performance in an oil environment. The phosphorus oxides produced present lower friction and faster deposition ability on the steel surface, which helps to provide initial low friction. Under the effect of friction stress, the amorphous carbon layer constructed from the decomposition of OA could combine with phosphorus oxides and BP crystals to build a unique three-layer tribofilm to replace the shear interface from steel/steel to carbon-oBP/carbon-oBP layers. This study extends the application range of oBP from water-based to www.Springer.com/journal/40544 | Friction oil-based environments and provides a simple method for the superlubricity between the engineering materials.

Declaration of competing interest
The authors have no competing interests to declare that are relevant to the content of this article. The author Guoxin XIE is the Editorial Board Member of this journal and the author Jianbin LUO is the Editor-in-Chief of this journal.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this licence, visit http://creativecommons. Since 2014, he has worked at Tsinghua University as an associate professor. His research interests include intelligent self-lubrication, electric contact lubrication, etc. He has published more than 50 referred papers in international journals. He won several important academic awards, such as Chinese Thousands of Young Talents, the Excellent Doctoral Dissertation Award of China, and Ragnar Holm Plaque from KTH, Sweden.