As natural analogs of graphene, hexagonal boron nitride nanosheets (BNNSs) have exhibited more superior properties than graphene, such as anti-oxidation ability, thermal-conductivity, electronic application [1, 2]. On the basis of these preferable properties, BNNSs were arousing great attention in many applications [3-6]. Since the superlubricity speciality in two dimensional nanomaterial was discovered [7, 8], their application in lubrication showed a tremendous application prospect [9-12].

Since the mechanical exfoliation method was successfully carried out for producing graphene, few-layered BNNSs derived from its bulk crystal has been intensively studied in recent years [13, 14]. However, boron nitride (BN) layers had stronger ionic characteristics between the interlayers than the weak van der Waals force between the graphene layers [15, 16]. Therefore, this stronger intermolecular force caused the exfoliating process much difficultly.

Up to now, few-layered BNNSs were accessed via either solid-state peeling/exfoliation or liquid-phase exfoliation. In contrast to both of them, the latter was deemed to be more highly efficient than former [17-19]. Nevertheless, liquid-phase exfoliation was often subjected to the use of a large amount of chemical reagents [20-22], serious pollution discharge and residues [23, 24], high cost and poor efficiency [25, 26]. Besides these issues, the chemical impurities were inevitably retained on the as-exfoliated few-layered BNNSs, so the as-received products had to carry out the post-treatments for the sake of purification. Therefore, a preferable strategy for producing few-layered BNNSs should have the characters of facile processing, avoiding the use of toxic and harmful chemical reagents and without or less environment pollution. The appearance of thermal expansion exfoliation method for few-layered nanosheets was considered to be an alternative approach to be accessible to this strategy [27-30]. However, the drawbacks of this method also were obvious in past, such as lower output, poor quality [31, 32]. Recently, there have emerged many fresh solid-phase exfoliation methods for high-quality, few-layered BNNSs, such as microwave and liquid nitrogen [33]. The advantages of these harsh approaches could obviously reduce the exfoliation duration and less solvent-using in preparation. There were many successful cases thus far, such as the microwave-assisted method for graphene [34], MoS₂[35], BN [36], and liquid nitrogen assisted exfoliation for graphene [37], MoS₂[38], BN [30]. The powerful physical means could destroy the stronger Van der Waals force between BN layers, thus leading to the curling of the nanosheets edge of layered materials [29]. Hence, the essence of liquid nitrogen exfoliation relied on the huge temperature difference to trigger the curling of nanosheets edge because the gasification of liquid nitrogen resulted in the tailoring force to destroy the force of interlayer of h-BN. Since the microwave function for heating belonged to the molecule level, the surface oxidation of h-BN nanosheets in high temperature was suitable to adopt this way. Regretfully, the single physical way to exfoliate the layered material was usually difficult to access the high-quality and high-yield nanosheets [39, 40]. Therefore, the strategy for the combination route based on microwave and liquid nitrogen ways to access BNNSs was probably prospective to address the obstacle of production.

In the present work, we proposed a highly efficient combined process with high quality and throughput for few-layer BNNSs. In this process, the rapid thermal expansion of microwave was employed in first step, followed by suffering the fast gasification in the liquid nitrogen. The advantages of this combination process for BNNSs were fast, environmental-friendly and non-residual. We have developed a combination strategy route to exfoliate BNNSs. The possible exfoliation mechanism was proposed in discuss. Finally, the tribological behavior of the as-exfoliated BNNSs was investigated by the sliding frict- ion in reciprocating motion.

In a typical experiment, 2 g h-BN powder (white-color powder, 30-90 nm in average thickness and 0.1-0.5 µm in average size) was suffered into the household microwave oven (700 W, 220 V and ~50 Hz) for 20 min, denoted M-BNNSs. Following, the above sample was swiftly impregnated into 250 mL liquid-nitrogen (L-N₂) until the L-N₂ was gasified thoroughly, marked as M-L BNNSs. After the ending, the sample was dispersed into ethanol and subsequently centrifuged at 1000 rpm for 30 min for the sake of removal of the aggregated sheets with poor quality. Finally, the sample was dried, denoted as-exfoliated BNNSs. For calculating remaining concentration in ethanol along with settling time, the dispersion solution of each day was filtered through an inorganic membranes with 0.5 µm of pore size. After the ending, the membranes were dried and then the mass of the deposited material was weighted. Thus, the residual mass in dispersion was counted by the total mass reducing the disposition mass. During the whole process, the solution volume maintained the uniformity.

The quality of BNNSs was determined by a series of character-izations. Fig. 1 displays the TEM images and AFM characterization of the M-BNNSs and as-exfoliated BNNSs. In contrast to bulk h-BN (Fig. S1 in Supporting information), the SEM images of the M-BNNSs and BNNSs (Fig. S2 in Supporting information) show the similar morphology and maintain intact lateral size, indicating that the exfoliations under the microwave and liquid nitrogen have not damage in the lateral structure. As shown in Fig. 1 and Fig. S1b, the TEM images of the M-BNNSs and BNNSs indicate that their thickness become flimsy after experiencing microwave and liquid nitrogen treatments due to being more transparent. In addition, the size and shape of the as-exfoliated samples are uniform, confirming the SEM result. Convincingly, the thicknesses are confirmed by atomic force microscopy (AFM) (Figs. 1c and d, Fig. S3 in Supporting information). Obviously, the thickness of the M-BNNSs and BNNSs are up to approximately 3.22-3.77 (10-12 layers) and 2.59 nm (8 layers), respectively. These results demon-strate that the exfoliation for few-layer nanosheets is achieved by the microwave and liquid nitrogen.

As can be seen, the as-exfoliated BNNSs consist of a great number of randomly aggregated crumpled nanoflakes (Fig. 2). The corresponding SAED (Fig. 2b) indicates that as-exfoliated BNNSs are of the typical six-fold symmetry characteristic of h-BN. The spacing between adjacent fringes was measured to be 0.33 nm, indicating that the lattice fringes are the (002) crystal planes of h-BN. Figs. 2c and d corroborate that the layer number of nanosheets is about 8-12 layers.

Fig. 3 displays the XRD patterns, Raman patterns and UV-vis spectra of h-BN, the M-BNNSs and as-exfoliation BNNSs. These three samples have the similar characteristic peaks at 26.5°, 41.6°, 55.0° and 76.0°, corresponding to the (002), (100), (004) and (110) planes, respectively. After being born the thermal expansion of microwave power, the intensity of the as-exfoliated BNNSs is weaker than bulk h-BN, indicating much loose stacking at the c-direction in the M-BNNSs [41]. This reveals the change of interlayer structure for as-exfoliated BNNSs. Moreover, the full widths at half-maximum of as-exfoliated BNNSs (FWHM = 0.42) at the (002) diffraction peak are larger than bulk h-BN (FWHM = 0.38), indicating the decrease of layer number [42]. Distinctly, the (002) diffraction peaks of h-BN (26.62°) and the M-BNNSs (26.54°) shift toward the lower angel, corroborating the enlargement of their interplanar distance [11]. However, the (002) diffraction peaks of as-exfoliated BNNSs (26.74°) appear blue shift, which confirms the decreased number of layers (thickness reduction), owing to exfoliation of the M-BNNSs [41]. As shown in Fig. 3c, the characteristic Raman peak of bulk h-BN at 1366.84 cm^-1 is attached to E_2g mode with similar mode in graphene [43]. Nevertheless, the characteristic peaks of the M-BNNSs (1365.42 cm^-1) and BNNSs (1364 cm^-1) move towards the orientation of the low wavenumber, which is consistent to the previous reports [44]. Alternatively, the declining intensity of E_2g can reflect the weaker interaction between layers because of the exfoliation [45]. The FWHM values of the M-BNNSs and BNNSs broaden from 9.93 cm^-1 (h-BN) to 11.35 cm^-1 and 12.77 cm^-1, respectively. This is caused from the decline of the layer-to-layer interaction of the exfoliated product [46], demonstrating the presence of exfoliation effect by microwave and liquid nitrogen processes. As indicated in the FT-IR spectra (Fig. S4 in Supporting information), the typical peak of bulk h-BN at 1373 cm^-1 whereas that of the as-exfoliated BNNSs at 1406 cm^-1 has a blue shift, which is associated to the decreasing of thickness from h-BN to as-exfoliated BNNSs [47, 43]. Additionally, the out of plane bending mode of B-N -B at 802 cm^-1 is not a shift for them. Compared to h-BN, the UV-vis adsorption spectra (Fig. 3d) of the as-exfoliated BNNSs at about 210 nm shows a wider band, corresponding to a band gap of ~5.92 eV. This is closer to the reported value (~6.15 eV) [48].

Fig. S5 (Supporting information) shows the XPS energy spectra of bulk h-BN and as-exfoliated BNNSs. The survey spectra of two samples both own two mainly binding energy peaks of B 1s and N 1s at 190.5 and 398.4 eV, respectively. It is in good accordance with the reports [49]. In terms of calculation, the B/N atom ratio is 1.20, high than the chemical ratio of 1:1. We speculated that the atom ratio related to the oxidation of B atoms [4]. The additional oxygen and carbon peaks is most likely to result from the oxidation due to the exposure in air during the XPS measurement preparation [50, 51]. In particular, the deconvoluted B 1s spectrum of the as-exfoliated BNNSs emerges two peaks at 190.5 and 191.7 eV. The former comes from the binding energy peak B-N and the latter is attributed to the B-O bond. Meantime, the deconvoluted N 1s spectrum of the as-exfoliated BNNSs also arises two peaks at 398.4 and 399.2 eV. The latter is assigned to the N-O bond. Dramatically, the O atomic concentrations of the bulk-h-BN and as-exfoliated BNNSs are up to 2.81% and 5.75%, respectively. The remarkably rising reason of O content in the as-exfoliated BNNSs is that the oxidation has been occurred during the exfoliation of microwave powered stage [4]. Fig. S6 (Supporting information) represents the statistical data of the thickness and the lateral size of as-exfoliated BNNSs (Fig. S7 in Supporting information). The statistical results of the thickness indicate that 95% of as-exfoliated BNNSs nanosheets have 2-6 nm in the thickness and the average value is approximately 4.08 nm, closer to the AFM characterization. The statistical lateral size shows that the average lateral size of the as-exfoliated BNNSs is about 215.8 nm. The dependence of the concentration varying of nanosheets in ethanol to the standing time is shown in Fig. S8 (Supporting information). The result can reflect the overall quality and output of nanosheets with different size distributions. As delaying the standing time, the concentration of the as-exfoliated BNNSs in ethanol is dropped and ultimately up to the relatively stable value at 0.565 mg/mL after standing for 7 days.

Fig. 4a shows the friction coefficient (COF) curves and the average COFs (inset) of grease, 0.08 wt% h-BN- and 0.08 wt% BNNSs-based grease along with friction. In this respect, the BNNSs-based grease exhibits the superior antifriction performance than other two samples. In contrast to grease, the COF of the BNNSs-based grease reduces by 20.10%. Fig. 4b presents the varying of the COFs of grease, the h-BN- and BNNSs-based grease as a function of concentration. Within adding the concentration, the lowest COFs appear at 0.08 wt% concentration, attaining the optimal adding content. As shown in Figs. 4c and d, the wear track depths of cross section and 3D optical surface morphologies indicate that the BNNSs-based grease has the preferable antiwear performance, decreasing by 55.8% and 45.1% relative to grease and h-BN-based grease (Fig. S9 in Supporting information), respectively.

In summary, we proposed a novel approach to massively exfoliate BNNSs via microwave-assisted liquid nitrogen exfoliation. The characterizations demonstrated that the as-exfoliated exhib-ited a good quality both thickness and yield. Encouragingly, the as-exfoliated BNNSs as additive in grease exhibited the excellent anfriction and antiwear performance.

We have no conflicts of interest to declare.

This work was funded by Jiangsu Province Six Talent Peaks Project (No. 2014-XCL-013) and Jiangsu Industrial-academic-research Prospective Joint Project (No. BY2016069-02). The authors also acknowledge the Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions and Top-notch Academic Programs Project of Jiangsu Higher Education Institutions (No. PPZY2015B112). The data of this paper originated from the Test Center of Yangzhou University.

Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.cclet.2020.01.019.