Silanization of boron nitride nanosheets (BNNSs) through microfluidization and their use for producing thermally conductive and electrically insulating polymer nanocomposites

Journal of Solid State Chemistry 249 (2017) 98–107

Contents lists available at ScienceDirect

Journal of Solid State Chemistry journal homepage: www.elsevier.com/locate/jssc

Silanization of boron nitride nanosheets (BNNSs) through microfluidization and their use for producing thermally conductive and electrically insulating polymer nanocomposites

MARK

⁎

A.Tuğrul Seyhana,b, , Yapıncak Göncüa, Oya Durukana, Atakan Akaya, Nuran Aya a

Department of Materials Science and Engineering, Anadolu University (AU), Iki Eylul Campus, 26550 Eskisehir, Turkey Composite Materials Manufacturing Science Laboratory (CMMSL), Research and Application Center of Civil Aviation (RACCA), Anadolu University (AU), Iki Eylul Campus, 26550 Eskisehir, Turkey b

A R T I C L E I N F O

A BS T RAC T

Keywords: Microfluidization Boron nitride nanosheets (BNNSs) Silane coupling agent Polypropylene Polymer nanocomposites Thermal conductivity

Chemical exfoliation of boron nitride nanosheets (BNNSs) from large flakes of specially synthesized micro-sized hexagonal boron nitride (h-BN) ceramics was carried out through microfluidization. The surface of BNNSs obtained was then functionalized with vinyl-trimethoxy silane (VTS) coupling agent through microfluidization once again in an effort to make them compatible with organic materials, especially those including polymers. The morphology of BNNSs with and without silane treatment was then systematically characterized by conducting various different analytical techniques, including Thermogravimetric analysis (TGA), X-ray diffraction (XRD), Scanning Electron Microscopy (SEM), Bright field Transmission Electron Microscopy (BFTEM), Contact angle analyzer (CAA), Particle size analyzer (PSA) and Fourier Transmission Infrared (FTIR) spectroscopy attached with attenuated total reflectance (ATR) module. As a result, the silane treatment was determined to be properly and successfully carried out and to give rise to the irregularity of large flakes of the BNNSs by folding back their free edges upon themselves, which in turn assists in inducing further exfoliation of the few-layered nanosheets. To gain more insight into the effectiveness of the surface functionalization, thermal conductivity of polypropylene (PP) nanocomposites containing different amounts (1 wt% and 5 wt%) of BNNSs with and without silane treatment was experimentally investigated. Regardless of the weight content, PP nanocomposites containing silanized BNNSs were found to exhibit high thermal conductivity compared to PP nanocomposites containing BNNSs without silane treatment. It was concluded that microfluidization possesses the robustness to provide a reliable product quality, whether in small or large quantities, in a very time effective manner, when it comes to first exfoliating two-dimensional inorganic materials into few layered sheets, and functionalizing the surface of these sheets afterwards to make it possible to utilize them as promising filler constituent in manufacturing thermally conductive and electrically insulating polymer nanocomposites that could be considered as whole or a part of a heat-releasing device.

1. Introduction A surge of scientific attention has been recently directed to the exfoliation of single or very few layered graphene sheets from graphite to benefit from its excellent quantum transport and mechanical properties that are comparable or superior to one-dimensional carbon nanotubes [1–8]. More lately, this attention has shifted to the exfoliation of boron nitride nanosheets (BNNSs) from hexagonal boron nitride (h-BN) ceramics, a structural analogue of graphite wherein alternating B and N atoms are substituted with C atoms [2–6]. In contrast to graphene, BNNSs are electrically insulating materials, yet exhibit various graphitic-like plane (002) properties, such as high ⁎

thermal conductivity and extraordinary mechanical strength. BNNSs have been therefore believed to have as much scientific potential as graphene sheets to find widespread use as promising filler in polymer nanocomposite-based heat-releasing devices [9–15]. For example, if a special engineering case was of concern where an electrically insulating device with high thermal conductivity is of fundamental importance, polymer nanocomposites with BNNSs would certainly have superior advantages over polymer nanocomposites with graphene sheets. Nevertheless, the lack of a feasible and practical approach to synthesizing few layered BNNSs or graphene sheets in quantities large enough to make an electronic device stands as an obstacle in the way of exploiting their use in polymer nanocomposites. [2–14]. Generally speaking, it is,

Correspondence to: Anadolu University Iki Eylul Campus Engineering & Architecture Faculty Materials Science & Engineering Department, 26480 Eskisehir, Turkey. E-mail address: [email protected] (A.T. Seyhan).

http://dx.doi.org/10.1016/j.jssc.2017.02.020 Received 5 October 2016; Received in revised form 28 January 2017; Accepted 19 February 2017 Available online 24 February 2017 0022-4596/ © 2017 Elsevier Inc. All rights reserved.

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

more scalable and sophisticated methods be utilized to unveil the extraordinary physical properties of very few layered BNNSs for innovative engineering applications. In our former studies [20,21], we revealed that microfluidizer is highly capable of high throughput exfoliation of very few thin layered sheets and of dispersion of carbon nanotubes with a relatively short dwell time in an epoxy resin, at the same energy consumption level as in sonication. As for the second challenge, surface functionalization is expected to induce a significant effect on the degree of dispersion of nano-sized materials in organic materials [23–27]. From this perspective, the covalent functionalization of BNNSs with organic molecules to alter the surface chemistry and to enhance the interfacial interactions would be highly reasonable. In other words, it should be therefore of interest to graft organic functional molecules of choice to exfoliated BNNSs in a controllable manner. Although BN allotropes are known to exhibit excellent chemical stability, whether in strongly acidic or basic conditions, several reports have described the covalent functionalization of boron nitride nanotubes (BNNTs) and BNNSs by using oxygen and nitrogen radical species [22]. Parallel to this, the generation of a highly reactive nitrene species was utilized for covalently functionalization of the BNNSs [22]. Moreover, Lewis bases, such as amine molecules with long lipophilic or hydrophilic chains, were also used to form complexes or adducts with the electron-deficient boron atoms on BNNSs. It was reported [23] that the complexation of a Lewis base with BNNSs facilitated the further exfoliation of the layered structure of the bulk material, resulting in thin planar BNNSs that are readily dispersible or soluble in organic solvents. On the other hand, alternatively, the boron sites in the BN lattice could be theoretically activated by oxidation in concentrated acid solutions, which is expected to result in yielding hydroxyl groups at the BN surfaces. Subsequently, the hydroxylated BN could be further modified by grafting with surface coupling agent for better adhesion and dispersion of BN particles in polymers. From this point of view, silane coupling agents, widely used as the surfactant to modify or functionalize the surfaces of inorganic particles, might prove to offer promise. It was already reported that epoxy composites containing 30 wt% of silane treated h-BNs exhibited 6.14 times higher thermal conductivity than neat epoxy at no expense of mechanical and thermal rigidity. In this study, we present a systematic route to the covalent functionalization of the surface of the exfoliated BNNSs utilizing a silane coupling agent through microfluidization in order to render their surfaces more hydrophobic and compatible with common organics, including polymers. BNNSs obtained from the exfoliation of the microsized h-BNs through microfluidization were then subjected to silanization through microfluidization once again. To our best knowledge, this study is the first study in the literature that utilizes a microfluidizer both to exfoliate and to chemically functionalize the surfaces of 2D inorganic materials in sequence. Whether the surface of BNNSs was properly functionalized or whether the BNNSs were morphologically altered was monitored and verified by systematically employing various analytical techniques, including Thermogravimetric analysis (TGA), Xray diffraction (XRD), Scanning Electron Microscopy (SEM), Bright field Transmission Electron Microscopy (BF-TEM), Contact angle analyzer (CAA), Particle size analyzer (PSA) and Fourier Transmission Infrared (FTIR) spectroscopy attached with attenuated total reflectance (ATR) module. To evaluate the performance of the silanized BNNSs when used for materials engineering applications, they were blended with polypropylene (PP) through melt compounding followed by hot pressing. The findings obtained were discussed in a concise manner with emphasis being placed on the silane-induced morphology alteration of the BNNSs and its effects on the thermal conductivity of the resulting PP nanocomposites.

in fact, not only the case for BNNSs and graphene sheets, but it is also a common challenge encountered during exfoliation of any types of 2D materials. Another point to take into account is that it is required that the surface of most inorganic materials, before blended with polymers, be chemically functionalized for the purpose of enhancing their compatibility during composite processing [15–23]. As is in our case, BNNSs with their highly inert surface have a great difficulty, forming a homogeneous dispersion in polymers. More specifically, BNNSs possess a great propensity to present in the form of coarse agglomerates in polymers along which the high interfacial thermal resistance takes place, which would reduce the thermal transport via phonon scattering, thus hammering the coherence of phonon propagation [24–31]. It is for this reason that the surface treatment of BNNSs with functional organic molecules would sound highly promising, when it comes to benefitting from their high thermal conductivity and mechanical strength as promising filler in polymers [32–38]. However, easy and effective methods of modifying the surface of BNNSs are still missing for large volume applications [22–28]. Having focused on the scientific essence of the aforementioned issues, two important challenges seem to require being resolved before talking about BNNS modified polymer nanocomposites with high thermal conductivity that would be employed for the development of new microelectronic devices. The former is to get some ways of robustly and repeatedly exfoliating very few layered nanosheets in large quantities, while the latter is to come up with an efficient methodology to chemically functionalize their surface to make them compatible with polymers. As for the first challenge, mechanical exfoliation by Scotch tape has emerged to be the first method conducted to prepare freestanding or very few layered graphene sheets [7]. Herein, the pulling force during the mechanical cleavage disrupts the weak van der Waals interactions between graphene layers, thereby leaving the sp2 bonded in-plane structure intact. This method was shown to work on other layered materials, too, such as h-BNs and molybdenum disulfide (MoS2). However, since a large number of high-quality nanosheets are to be harvested for advanced engineering applications, mechanical exfoliation by Scotch tape with extremely low yield turned out to be impractical for production of nanosheets in large quantities. As a result, ball milling has been further deemed an impending prospect for thickness reduction of 2D layered materials [6–9]. However, although powerful, ball milling treatments have been reported to be uncontrollable in some cases during which shear force coupled with compression might be detrimental to the in-plane crystal structure, thus causing a great number of structural flaws to occur on the milled particles, even though the large number of relatively small steel balls was proposed to be used during milling in an attempt to alleviate this drawback. As a consequence, a planetary mill was proposed as a more promising machinery tool to control the rolling balls in such a manner that it applies only shear force on the milled particles [7–9]. The use of lubricants was also suggested in planetary milling to avoid the damage to the structures while also eliminating the welding effect. Apart from those methods mentioned above, chemical methods have also come under spotlights for the same purpose to take advantage of the exfoliation capability provided by the polar solvent, despite being claimed to cause high defects in crystal structures compared to pure mechanical exfoliation. Zhi et al. [10] prepared milligram quantities of BNNSs, conducting a simple two-step process that involves exfoliating h-BN powders into nanosheets in dimethylformamide (DMF) through sonication, followed by separation of the exfoliated BNNSs by centrifugation. They concluded that the BNNSs obtained by centrifugation at 8000 rpm and 5000 rpm were of 3 nm and 7 nm thick, respectively. Many other similar studies can be found in the literature that utilized various different chemical solvents to prepare very thin few layered BNNSs by applying sonication at different periods of time and energy consumption levels [9–14]. However, although beneficial to preparing BNNSs in milligram levels, sonication has not been found good enough for large scale production of BNNSs. For this reason, it is required that 99

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

2. Experimental procedure 2.1. Preparation of exfoliated BNNSs through microfluidization Exfoliation of specially synthesized micro-sized h-BN precursors of very thin large flakes into BNNSs was carried out through microfluidization. Generally speaking, microfluidization is a dynamic highpressure homogenization process based on accelerating liquid to velocities of up to several hundreds of meters per second through a specially designed fixed interaction chamber wherein several orders of magnitude higher shear rates are created than are obtainable by any conventional mixing equipment [19]. For the purpose of proper exfoliation, the combination of N,N dimethylformamide (DMF) and chloroform (6–1 on weight basis) were prepared and conducted as solvent. Hexagonal boron nitride precursors (6 wt%) were then added to the prepared solvent (100 ml). To derive chemically exfoliated BNNSs, a high pressure microfluidization was carried out by employing a commercially available microfluidizer (M-110P, Microfluidics Corp.) at constant intensifier pump pressure of 207 MPa (30,000 Psi). Twenty (20) of circulation passes was applied on the dispersions. It takes about 0.6 min for 100 ml solution of h-BN precursors to complete one circulation pass. More information regarding the working principle of a microfluidizer and the corresponding experimental details can be found elsewhere [19–21]. Following exfoliation of large flakes of hBNs, the exfoliated BNNSs were obtained ready for silanization. 2.2. Silanization of BNNSs through microfluidization The aim of silanization herein is to improve solubility of BNNSs in organics as well as to alter their morphology for the sake of their efficient use in functional polymer nanocomposites with high thermal conductivity. Generally speaking, silane coupling agents are multifunctional molecule which reacts at one end with the surface of the inorganic phase, while at the other end with the surrounding matrix phase of choice, whether it is inorganic or organic. Silane coupling agents have a generic chemical formula R(4−n)-Si-(R′X)n (n=1,2) where R is alkoxy, X represents an organofunctionality, and R′ is an alkyl bridge (or alkyl spacer) connecting the silicon atom and the organofunctionality [23–27]. Various different silane structures were shown to assist in enhancing interfacial bonding between inorganic reinforcements, such as glass fibers, and organic polymers, such as polypropylene or unsaturated polyester resin. The way that organofunctionality of the silane interacts with the polymers is dependent to large extent on the type of reactivity of the functional end group with the polymer. Fig. 1 gives the silanization steps where BNNSs were used as reference for having a better understanding of the aforementioned issues involved. In this study, vinyl trimethoxy silane (VTS) obtained from Merck was used for chemically functionalizing the surface of BNNSs. In principle, VTS is first hydrolyzed to the silanol in the aqueous solution to which the BNNSs are exposed (Hydrolysis) [22– 29]. The silanol molecules then compete with water molecules to form hydrogen bonds with the hydroxyl groups bonded to the surface of BNNSs. When BNNSs are dried, the free water is driven off. The condensation reaction takes place afterwards, both at the junction between silanol and BNNS surface as well as between neighboring silanol molecules (Condensation). The result is what is called polysilaxone layer which is bonded to the constituent surface, thus giving an array of the corresponding reactive groups to the surrounding environment (Bond formation). Note that prior to silanization, the surface hydroxylation of BNNSs was enhanced through strong acid oxidation to yield more hydroxyl groups bonded to the B atoms in the edge planes of the BNNSs. For this purpose, 5 g of BNNSs was added to 200 ml solution of nitric acid. The prepared suspension was then allowed to stay overnight at rest. The resulting solution obtained was then continuously mixed, using a magnetic stirrer for 24 h at 70 °C. Deionized water was then used to wash the filtrate until it had a pH

Fig. 1. VTS-Silanization of BNNSs.

value of approximate 7. Then, the corresponding BNNSs obtained were dried in a vacuum oven at 80 °C for 24 h. To get started with the silanization procedure, a mixture of ethanol and deionized water (4:1) was prepared followed by adjustment of its pH value to approximately 4 by dispensing few drops of sulfuric acid (95%) into the mixture. The weight percentage of the silane coupling agent required for the mixture was predicted according to the equation as follows [13–14.]

X = (A / ω). f

(1)

where X is the silane weight content required for a minimum uniform coverage of the filler constituent (in g.), f is the amount of filler constituent (in g.), A is the surface area of the BNNSs (m2/g), and ω is the wetting surface of the corresponding silane coupling agent [4]. Following simple calculation, 3 wt% of silane coupling agent was predicted to be added to the prepared mixture. To promote the silanization reaction, BNNSs were ultrasonically dispersed within the solution for 1 h. Following addition of 5 g oxidized BNNSs to the mixture, the microfluidizer was used for 5 cycles more to ensure proper silanization and thus to obtain less agglomerated BNNSs at the end of the process. Fig. 2 gives the flowchart that summarizes briefly each experimental step mentioned above. The final mixture obtained after microfluidization was continuously blended with a magnetic stirrer for 1 h and 30 min at 50 °C. The solution was then allowed to stay overnight at room temperature for condensation. The solution was then filtered, followed by drying in a vacuum-oven at 110 °C for 5 h for 100

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

Fig. 2. Flowchart that summarizes briefly each experimental step followed to obtain silanized BNNSs.

the sake of formation of siloxane network. The final product obtained was washed several times with deionized water and ethanol in sequence. Please note that methanol was eventually used to remove unreacted silane residues as well as any other prospective contaminants from the surface of BNNSs.

BNs, pristine BNNSs, and the silanized BNNSs were obtained using a Malvern Mastersizer Nano series–Nano ZS. For this purpose, all types of the powders were first put in a mixture of DMF and chloroform (1:1), followed by sonication for 1 h (20 s off and 30 s on). Note that refractive index of boron nitride and its resulting suspensions are measured to be 1.74 and 1.43, respectively.

2.3. XRD characterization 2.5. FTIR analysis XRD patterns were collected using a Bruker AXS D8 4-circle diffractometer equipped with a Cu sealed tube point source. A Göbel mirror optic generating a 2D-collimated parallel beam (divergence 0.03 and a lateral length of 18 mm) was used to collect the data. Monochromatic Cu–K radiation with a wavelength of 1.5405 A° equipped with an anti-scatter slit was conducted to remove background resulting from contribution of the holder. 2θ ranging from 10° to 60° with the scan speed of 2°/min at 40 kV and 30 mA was employed to gain an insight into the prospective alterations in the morphology of pristine and silane-treated BNNSs by taking into accounts the peaks of the starting material, specially synthesized micro-sized h-BN ceramics. The peak positions of (002) in the XRD curves of the starting material (micro-sized h-BNs), pristine and silanized BNNSs were fitted to Lorentzian distribution. The interlayer spacing along with the full width measured at half maximum (FWHM) of the peak intensities were measured. The correlation length (L), which refers to the number of stacked layers along the (c) direction and the layer diameter along the (a) direction on the lateral extent of BNNS layers, was determined according to the Scherrer formulation as follows [27].

L=

Kλ B2Θ cos Θ

FTIR was conducted to monitor and verify whether or not proper attachment of silane molecules onto the surface of BNNSs was successfully accomplished. For this purpose, FTIR spectra were recorded from 4000 to 500 cm−1 at resolution of 4 cm−1 with a scan time of 64 s by using a Bruker FTIR Tensor 27 equipped with attenuated total reflection (ATR) module. 2.6. TGA measurements As BNNSs possess the excellent thermal stability combined with oxidation resistivity, it could sound highly reasonable to monitor removal of organic silane groups grafted onto the surface of BNNSs by heating the modified sheets above the decomposition temperature of the silane groups in a thermal heating chamber with an inert atmosphere. With this fact in mind, TGA measurements were carried out on the pristine and silanized BNNSs, using a TA instruments thermogravimetric analyzer (TGA-Q500) at heating rate of 10 °C/min under nitrogen atmosphere with a flow rate 50 ml/min from 40 to 750 °C to evaluate to what extent the degree of silanization was achieved for BNNSs.

(2)

where λ is the wavelength of the X rays used ~1.5418 Å, B_2Θ is the integral breadth, the width of a rectangle with the same height and area as the diffraction peak. Note that for accurate size-strain broadening, one should use B_2Θ as a measure of peak width and relate it to FWHM depending on the distribution fit model used. K is a constant which is equaled to 0.94 and 1.84 for the calculation of correlation length along the (c) and (a) directions, respectively [30].

2.7. Contact angle measurements Contact angle measurements were carried out on the BNNSs with and without silane treatment by using a Biolin Scientific Attension instrument so as to evaluate to what extent a hydrophobic surface was eventually accomplished. The tests were done according to sessile drop method, using ultra-pure water. Water surface tension was calculated according to the pendant drop method. The tabulated contact angle data correspond to the average value of at least three different measurements which were taken at room temperature from different points over the surface of pelletized sheets with and without silane

2.4. Particle size measurements The particle size distribution of specially synthesized micro-sized h101

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

of nanocomposites [31]. The heat generated then comes across the sample material through the sensor during which a prompt voltage drop takes place at the heating source. The rate of increase in the sensor voltage was then considered to determine the TC of the nanocomposites. Testing was conducted for all nanocomposites at 25 °C.

treatment. In addition to this, since hydrophobic effect is closely associated with the propensity of non-polar substances to aggregate in solutions, pristine and silanized BNNSs were placed in ethanol and their level of suspension at different time intervals was observed for comparison. 2.8. Microstructural characterization

3. Results and discussion

A field emission gun-scanning electron microscope (Zeiss SEM Supra 50 VP) was employed to characterize h-BNs and BNNSs with and without silane treatment. The nanoscopic-scale characterization of BNNSs with and without silane treatment was also performed using a bright field emission transmission electron microscope (Jeol TEM 1100 F), operating at 200 kV.

Fig. 3a and b depict, respectively, SEM images of h-BN micro-sized precursors of very thin large flakes and the pristine BNNSs obtained through exfoliation of the flakes after they were subjected to 20 passes of circulation. As seen in the image, it is definite that pristine BNNSs have a thickness of somewhat around 100–200 nm following the exfoliation through microfluidization. Here, although not mentioned in detail, in light of our hands-on experiences, the degree of quality of the specially synthesized micro-sized h-BNs precursor is of certainly importance in obtaining the BNNSs with as low number of layers as possible. The larger the sheets, and the lower the number of layers, the better thermal conductivity would be achieved in polymers in which they are embedded. Fig. 4 shows XRD patterns of micro-sized h-BNs, and BNNSs with and without silane treatment together with their corresponding intensity peak values in the table underneath it to ease interpretation of the findings. It was revealed that, irrespective of the silane treatment, BNNSs possess characteristics of hexagonal phases whose diffraction peaks occur at 2(Ɵ) values of 26.78°, 41.96°, 44.12°, 50.48°, and 55.44°, which are identical to the peaks of h-BN crystals [5–7]. It can be seen that, after microfluidization, the intensities of the (002), (004), and (100) peaks increased, while the intensities of the (101) and (102) peaks remained almost intact. On the other hand, when BNNSs were silanized through microfluidization, the (002) and (004) peaks appeared to be further intensified, in contrast to the intensity of the peak (100) that has instantaneously decreased. Meanwhile, the intensities of the (101) and (102) peaks remained almost unchanged. This suggests that microfluidization has most probably no influential impact on the short range order of the BN crystals in the c-direction, and that the most of the in-plane layers remain unaffected even after the specially synthesized h-BNs ceramics were exposed to the high shear when passing through the microchannels during the process. More specifically, provided that the short range order of the crystals remain customarily intact, the change or decrease of those peaks, including (100), (101), and (102) might arise from the preferred orientation of uniformly sheared sheets into the substrates of the (002) basal plane that is parallel to the substrate surface, which makes the BNNSs end up with the reduced thickness

2.9. Preparation of BNNS filled PP nanocomposites Polypropylene (PP) nanocomposites containing different weight contents (1 and 5 wt%) of BNNSs with and without silane treatment were produced by melt compounding using a DSM Xplore twin-screw mini extruder with a volume of 15 ml. Prior to melt compounding, low molecular weight PP beads obtained from PETKİM GmbH in Turkey with a Melt flow rate (MFR) of 25g/10 min were blended with pristine and silanized BNNSs in ethanol, followed by drying in a vacuum oven for 3 h at 90 °C to get rid of moisture induced void formation during extrusion. The temperatures of the feed, metering, compression, and die sections in the extruder barrel were set at 195, 190, 195, and 190 °C, respectively. During the whole process, the screw speed and the feeding speed were kept constant at 500 rpm and 30 rpm, respectively. The mechanical torque value generated in the extrusion barrel was taken into account to check the accuracy of the set dwell time. The resulting composite melt was first allowed to be direct quenched in a water bath and then pelletized by using an electronic cutting machine. The final composite strands obtained were then utilized as raw materials for producing hot-pressed nanocomposite sheets at 195 °C under 50 MPa for a certain period of time. 2.10. Thermal conductivity (TC) measurements Thermal conductivity (TC) of the resulting PP nanocomposites was measured using a thermal analyzer developed by C-Therm TCi, CTherm Technologies Ltd., New Brunswick, Canada. Modified transient plane source (MTPS) method was used for the measurements. In the corresponding test configuration, heat is provided by means of a spiraltype heating source which is placed at the center of the probe to give rise to temperature at the interface between the sensor and the sample

Fig. 3. (a) SEM images of pristine h-BN micro-sized precursors of very thin large flakes. (b) The pristine BNNSs obtained after 20 circulation passage of h-BNs ceramics through microfluidizer.

102

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

Fig. 4. XRD patterns of micro-sized h-BNs and BNNSs with and without silane treatment together with their corresponding intensity peak values.

note that the samples were several times washed with methanol to remove any chemical contaminants or residues left to avoid any noise during the measurements. As seen in the figure, the particle size distribution of h-BN precursors decreased with de-agglomeration, followed by exfoliation. It proves to be that average size of micro-sized h-BN powders (755.7 nm) comes down to 12.48 nm after they were exfoliated into few layered sheets (BNNSs), when passing through the designated number of cycles during microfluidization. When BNNSs were silanized, average particle size of the silane treated BNNSs went up to 439.5 nm, which can be thought of as convincing evidence that silane molecules were successfully attached onto the surface of BNNSs. Fig. 6 gives FTIR spectra of BNNSs with and without silane treatment. As seen in the figure, it is obvious that they both have exactly the same absorption bands at 1375 and 760 cm−1 as h-BN ceramics, which, respectively, refer to B-N stretching and B-N bending [11,12,14]. Moreover, BNNSs with and without silane treatment display two distinct absorption bands that are in fact associated with h-

combined with the large diameter to thickness ratio. In other words, VTS vastly seem to have assisted in promoting the further alterations of BN crystals along the (002) and (004) orientations, thus reducing the number of layers while altering the stacking sequence. That seems to be more significant along the (002) orientation than along the (004) orientation. These findings are relevant to the findings reported in the literature [12–15,28–32]. To further back up this approach, the peak widths and the correlation lengths along the (002) orientation were estimated according to the Scherer formulation outlined in detail in the experimental section. The predictions showed that the general trend associated with the samples is that the lowest peak width, the highest correlation length results in. More specifically, silanized BNNSs, pristine BNNSs, and h-BNs were predicted to have 189 nm or 57 layers, 243 nm or 65 layers, and 446 nm or 330 layers, respectively. On the basis of the experimental results obtained and the predictions made, the silane treatment was concluded to offer promise to induce further exfoliation of the sheets. Fig. 5 gives the particle size distribution of specially synthesized micro-sized h-BN precursors, BNNSs, and the silanized BNNSs. Please

Fig. 5. The particle size distributions of specially synthesized micro-sized h-BN precursors, BNNSs exfoliated by the use of microfluidizer for 20 cycles and the silanized BNNSs obtained by the use of microfluidizer on the exfoliated BNNSs for 5 cycles.

Fig. 6. FTIR spectra of BNNSs with and without silane treatment.

103

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

100

significant weight loss up to 750 °C, as expected. As for the silanized sheets, a weight loss that takes place at 100–150 °C can be ascribed to desorption of water contained in the sheets. The weight loss around 200 °C is most probably resulting from removal of weak hydroxyl groups that were introduced through oxidation in the acid treatment. The weight loss around 500 °C might most probably be because of the pyrolysis of chemically bonded silane molecules [30–34]. It seems to continue to take place until 750 °C, suggesting that the whole thermal degradation process of silane molecules is in two steps. More specifically, what deserves mentioning is that despite the use of theoretically predicted amount of silane coupling agent (3 wt%), as elaborated in the experimental section, approximately around 0.84 wt% of silane coupling agents seemed to have covalently attached onto the surface of BNNSs. This would be indeed expected because chemical stability of the BNNSs is such extremely high that it causes the grafting degree of organic molecules to remain low, which sinks the effectiveness of covalent bonding functionalization. On the other side, this might be good for our case, as excessive attachment of silane coupling agent, although aimed at improving compatibility of the BNNSs with polymers while reducing interfacial thermal resistance, would reduce the TC of the BNNSs and of their resulting polymer nanocomposites, too. This is because silane coupling agents have an ability to form a very thick thermal barrier layer that might be capable to cause phonon scattering. This subject will be emphasized later in the text when it comes to discussing TC of the PP nanocomposites. To back up our approach in terms of surface characteristics, further studies were carried out on the silanized and pristine BNNSs by using a contact angle analyzer. Fig. 8a gives the photos of water-based contact angle measurements of the samples of pristine and silanized BNNSs. It was determined that the angle with silanized BNNSs (84.28°) is higher than the angle with pristine BNNSs (41.13°), suggesting that proper silanization was successfully performed, resulting in the surface of BNNSs to be more hydrophobic, as expected. These results were found to be reproducible, even after more experiments were carried out in a similar manner, utilizing various different samples that were randomly selected from diverse batches. Fig. 8b gives the photos of ethanol solutions containing pristine and silanized BNNSs. The mixture of ethanol and the powders were vigorously stirred for a long time enough

Weight [%]

Untreated BNNSs

99.5

Silanized BNNSs 0.84 % weight loss

99

o

100-200 C Water desorption o

200-500 C Silane decomposition

98.5

0

100

200

300 400

500

600

700

o

Temperature [ C] Fig. 7. Thermal degradation of the BNNSs with and without silane treatment.

BNs as well. Specifically, one at 1450 cm−1 refers to broad OH groups coming from B-OH band, while the other at 3000–3600 cm−1 refers to N-H stretching. After silanization of the BNNSs with vinyltrimethoxysilane (VTS), the two new peaks centered at 2922 cm−1 and 2852 cm−1 were observed to show up due to BNNS-VTS reaction, which can be attributed to the asymmetric and symmetric –CH2– stretching vibrations, respectively. Furthermore, B-O-Si band that shows up at 921 cm−1 can be attributed to the reaction between methoxy group of VTS and OH groups of BNNSs. In addition to this, Si–O–C and Si–O– Si stretching take place at 1000 cm−1 and 1100 cm−1, arising from the condensation of single VTS on the surface of BNNSs. These findings obtained can be reasonably cited ample evidence that the silane treatment of the sheets was successfully accomplished, which is proportional to the findings of XRD and particle size measurements. Fig. 7 gives the TGA thermograms of the sheets with and without silane treatment. As seen in the figure, the raw BNNSs showed no

Fig. 8. a) Photos of water-based contact angle measurements of the pristine and silanized BNNSs samples b) Photos of different ethanol solutions containing BNNSs with and without silane treatment.

104

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

undamaged micro-sized flakes that are stacked one by another are seen in Fig. 9b as a thick dark plate with somewhat inclines at both ends. Moreover, BNNSs without silane treatment seem to be flat without any significant folding back involved, as indicated by the arrows in Fig. 9c, whereas especially the free edges of the silanized BNNSs seemed to be cleaved and folded, which can be even noticeable to a naked eye along the region indicated by the arrows in Fig. 9 d and e, suggesting that their morphology was altered. In fact, these findings can be considered relevant to the XRD findings, such that the silane treatment resulted in morphology change, reducing the number of layers of the sheets. On the other hand, how the sheet cleavage or folding back could take place might most likely depend to large extent on how individually or homogenously the micro-sized h-BN ceramics are clustered in the presence of silane solution and take their positions immediate before they enter the micro-channels under a high mechanical shear. Two ways would seem to be possible. The former is that thin sheets could be peeled off the edge of h-BNs when a few or more sheets might collide edge-to-edge with each other, while at the same time as sliding over one by another, which could be pictured as in image (d), while the latter is that sheets could be exfoliated from the top of other sheets when they are hit over the top one by another, which could be pictured as in the image (e). In other words, it seems that the presence of silane coupling agents in the sample solution are also playing a pivotal role in the way the sheets interact before and after entering the micro-channels, inducing further exfoliation of the sheets. These findings are consistent with the earlier findings that silane treatment somewhat altered the sheet morphology, reducing the number of the layers of the sheets. To put all the aforementioned findings in use, BNNS filled PP nanocomposites were produced and their TC was measured. Generally speaking, observing the superior TC of particle filled composites lies in the fact that either the thermal conductive pathways are exploited through incorporation of very high amounts of fillers, or that the interfacial contact resistance be reduced by enhancing filler-matrix affinity, making chemical surface modifications [38–45]. TC of the composites is vastly dependent on the filler content. At the low filler weight content, filler particles are randomly oriented in the polymer matrix with no contact. As the content proceeds to increase, filler particles could start contacting each other by forming the thermally conductive pathways which end up with higher TC [43–45]. Fig. 10 gives the TC of neat PP and its resulting nanocomposites containing different amounts (1 and 5 wt%) of BNNSs with and without silane treatment. As seen in the figure, regardless of weight content, TC of the neat PP was improved with the pristine BNNSs, although not at the

Fig. 9. SEM images of silanized BNNSs (a) as well as the bright field TEM images of micro-sized h-BN ceramics (b), pristine BNNSs (c) and silanized BNNSs (d and e).

by hand and then allowed to settle down at room temperature at different time intervals (4, 24, 48 h). It was found that, regardless of exposure time, silanized BNNSs remain homogenously suspended in ethanol, indicating that they interact with ethanol molecules at the desired level of compatibility. On the other hand, pristine BNNSs exhibited a limited solubility, thus appearing as coarse clusters in ethanol at the bottom of the tube, as expected, due to their highly inert surface. These findings are proportional to the XRD, FTIR, and TGA findings. All the analytical techniques that were employed so far showed that the silane treatment was properly carried out, but have not explained whether or not any change in the physical appearance and the dimensional stability of the BNNSs existed due to the silane treatment. Correlation of the experimental XRD findings with the theoretical predictions implied that the number of the layers of the BNNSs was reduced when they were silanized. In this manner, SEM and TEM examinations were further carried out to see whether there is a discerned difference in appearance between pristine BNNSs and the silanized BNNSs. Fig. 9 gives the SEM image of silanized BNNSs (a) together with the bright field TEM images of micro-sized h-BN ceramics (b), pristine BNNSs (c) and silanized BNNSs (d and e), respectively. When compared to appearance of the BNNSs without silane treatment in Fig. 1b, BNNSs with silane treatment appears to be translucent in Fig. 9a. In other words, the difference between the two samples is that pristine BNNSs are coarse compared to silanized BNNSs. As for BF-TEM images, h-BN ceramics composed of lots of

Fig. 10. TC of the neat PP and its resulting nanocomposites with different weight content of BNNSs with and without silane treatment.

105

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

in the cases where the need arises for composite materials with extreme ratios of thermal to electrical conductivity, which best fits the electrically insulating nature of BNNSs.

desired statistical significance. Specifically, TC of neat PP and its nanocomposites with the pristine BNNSs scattered more or less around the value of 0.22 W m−1 K−1. Similar findings were also reported in another study [44] where it was implied that TC of the PP increased, as the weight content of h-BNs increased. They also found that TC of large-sized h-BN containing PP composites produced by melt compounding followed by compression molding exhibited higher TC than those containing small-sized h-BNs, which emphasizes the importance of selecting the most suitable precursors from which the layered nanosheets are exfoliated. For example, in the same study, TC of the neat PP sheets, which is around 0.33 W m−1 K−1, was shown to increase to 0.41 W m−1 K−1 when they were blended with 5 vol% of large-sized h-BNs. On the other hand, the silane surface treatment considered in this study brings an increase to the TC of the PP nanocomposites at the same filer loading rates. The results obtained that the TC of the PP nanocomposites with silanized BNNSs was improved with the increasing weight content without any statistical insignificance involved. For instance, PP nanocomposites with 5 wt% of BNNSs exhibited the TC of 0.38 W m−1 K−1 which corresponds to the improvement of % 90 when compared to the neat PP. One conceivable reason for the experimental findings could be that the relatively weak interfacial interactions between the pristine BNNSs and the surrounding PP matrix caused the aggregation combined with void formation among the sheets, causing them to have a poor dispersion level accompanied by low homogeneity across the nanocomposites. More specifically, poor dispersion in conjunction with coarse agglomerates and voids might bring about the manifestation of reciprocal phonon vector, which acts as a heat reservoir, thus helping restrict the heat flow diffusion. As a result, the silanized BNNSs prove to be better than pristine BNNSs at providing the PP matrix with high TC. Similar findings were also obtained in another study [45]. They found that, at the same weight content, TC of the PP composites with pristine h-BNs exhibited 0.256 W m−1 K−1 at room temperature, which was then reported to increase to 0.369 W m-1 K-1, when the h-BNs modified by 3-aminopropyl-3-ethoxy-silane (APTES) were used as filler. The authors ascribed it to the increased surface of additional phase boundaries that emerge between the fillers and the silane molecules, causing phonon scattering of propagating heat flux. They also concluded that the silane layer on the surface of h-BNs helps reduce the polymer blend viscosity during extrusion, enabling the h-BNs to be more homogenously dispersed and to be more individually separated within the PP matrix, which enhanced the TC of the resulting composites. In fact, silane layers are supposed to play an important role in the degree of compatibility and interfacial affinity between fillers and polymers. Silane molecules act as a bridge, one end is connected to the BN particle while the other end to the surrounding PP matrix. The strong interfacial interactions encouraged by silane modification are expected to lower interfacial thermal resistance, resulting in high TC. However, on the other hand, excessive grafting of silane coupling agent is capable to reduce the TC, forming a thermal barrier layer that may interfere with phonon scattering, as elucidated earlier during the discussion of TGA findings. It seems that the degree of the silane organic groups grafted on the surface of BNNSs works good enough for obtaining dispersible and soluble BNNSs to be used for polymer composite based heat releasing devices. Another point could be that the reduced number of the sheets through silane treatment, as approved by the XRD findings, prevents the degree of agglomeration to a certain extend by increasing the surface area of uniformly distributed sheets in the PP matrix, which results in high TC values. When recapping all the results obtained, microfluidization turns out to be very successful approach to functionalizing surface of such graphene-like fillers as BNNSs to extend their use in organic based engineering materials, such as polymers. Nevertheless, there are many rooms at the bottom to further consider when developing polymer/ BNNSs nanocomposites with much improved performance, especially

4. Conclusions In this study, we present a systematic route to the covalent functionalization of the surface of BNNSs with a silane coupling agent in order to render their surface more hydrophobic as well as to tailor their interfacial compatibility with polymers when they are used as fillers therein. For this purpose, the exfoliated BNNSs obtained using a microfluidizer were subjected to silanization once again through microfluidization in sequence. Whether or not the BNNS surfaces were properly functionalized was verified by using various analytical techniques including XRD, TGA, SEM-EDX, BF-TEM, FTIR-ATR, CAA, and PSA. It was determined that the proper application of silane treatment was carried out through microfluidization in a very short period of time. It was also revealed that the silane treatment altered the morphology of the BNNSs, giving rise to the irregularity in their physical appearance while reducing their number of layers. The free edges of BNNSs were found to be cleaved due to the silane treatment, while the pristine BNNSs appear in the form of large and regular flakes without their free edges folded. To put theoretical predictions and experimental observations in practical use, PP nanocomposites were produced and the TC of the resulting nanocomposites was measured. Regardless of the weight content, the highest TC values were obtained from the nanocomposites with silanized BNNSs. The results reported herein suggest that, when functionalized through a proper covalent bonding, exfoliated BNNSs could serve the same purpose for similarly high performance polymer nanocomposites with different promising fillers, such as graphene. For example, for potential industrial applications, polymer/BNNS nanocomposites could have superior advantages over nanocomposites based on graphene sheets, especially when electrically insulating nanocomposites with high TC is of specific interest. The key to the success in potential applications lies in accomplishing more effective and controllable exfoliation for higher quality of BNNSs through many different state-of-the-art chemical surface functionalization routes, whether covalent or non-covalent. However, no matter what route was applied, a significant challenge would always appear to be that interlayer bonding in hexagonal BN is much stronger than that in graphene. Acknowledgements This research was funded by the Scientific and Technological Research Council of Turkey (TUBITAK), Turkey through Grant number 110M158. The authors would like to thank BORTEK Ltd. (EskişehirTurkey) for providing us with the specially synthetized h-BN ceramics used in this study. The authors would cordially like to express their appreciation to Dr. Servet Turan, Dr. Hilmi Yurdakul, and Dr. Osman Nuri Çelik for their help in TEM investigation of the samples. References [1] M. Zheng, Y. Gu, Z. Xu, Y. Liu, Synthesis and characterization of boron nitride nanoropes, Mater. Lett. 61 (2007) 1943–1945. [2] Z. Huang, H. Guan, W. Tan, X. Qiao, S. Kulprathipanja, Pervaporation study of aqueous ethanol solution through zeolite-incorporated multilayer poly (vinyl alcohol) membranes: effect of zeolites, J. Membr. Sci. 276 (2006) 260–271. [3] Haoli Zhouab, Yi Sua, Xiangrong Chena, Shouliang Yia, Yinhua Wana, Modification of silicalite-1 by vinyltrimethoxysilane (VTMS) and preparation of silicalite-1 filled polydimethylsiloxane (PDMS) hybrid pervaporation membranes, Sep. Purif. Technol. 75 (2010) 286–294. [4] Y. Kubota, K. Watanabe, O. Tsuda, T. Taniguchi, Deep ultraviolet light-emitting hexagonal boron nitride synthesized at atmospheric pressure, Science 317 (2007) 932–934. [5] N. Alem, R. Erni, C. Kisielowski, M.D. Rossell, W. Gannett, A. Zettl, Atomically thin hexagonal boron nitride probed by ultrahigh-resolution transmission electron microscopy, Phys. Rev. B 80 (2009) 155425.

106

Journal of Solid State Chemistry 249 (2017) 98–107

A.T. Seyhan et al.

containing an acrylic silane, Coupling Agent Chem. Mater. 11 (1999) 2548–2553. [27] P.H. Jung, Y. Sung, H.T. Sun, L.Y. Kwan, W. Sung-Ho, N. Jae-Do, Interphase control of boron nitride/epoxy composites for high thermal conductivity, Rheol. J. 22 (2010) 259–264. [28] X. Yanjun, A.S. Callum, X. Zefang, M. Holger, M. Carsten, Silane coupling agents used for natural fiber/polymer composites: a review, Compos. Part 41 (2010) 806–819. [29] A.T. Seyhan, Z. Sun, J. Deitzel, M. Tanoglu, D. Heider, Cure kinetics of vapor grown carbon nanofiber (VGCNF) modified epoxy resin suspensions and fracture toughness of their resulting nanocomposites, Mater. Chem. Phys. 118 (2009) 234–242. [30] A. Tuna, Y. Okumus, H. Celebi, A.T. Seyhan, Thermochemical conversion of poultry chicken feather fibers of different colors into microporous fibers, J. Anal. Appl. Pyrol. 115 (2015) 112–124. [31] S.Y. Pak, H.M. Kim, S.Y. Kim, J.R. Youn, Synergistic improvement of thermal conductivity of thermoplastic composites with mixed boron nitride and multiwalled carbon nanotube fillers, Carbon 50 (2012) 4830–4838. [32] J.H. Warner, M.H. Rümmeli, A. Bachmatiuk, B. Büchner, Atomic resolution imaging and topography of hexagonal boron nitride sheets produced by chemical exfoliation, ACS Nano 4 (2010) 1299–1304. [33] Y. Wang, Z. Shi, J. Yin, Boron nitride nanosheets large-scale exfoliation in methanesulfonic acid and their composites with polybenzimidazole, J. Mater. Chem. 21 (2011) 11371–11377. [34] C. Li, Y. Bando, C. Zhi, Y. Huang, D. Golberg, Thickness-dependent bending modulus of hexagonal boron nitride nanosheets, Nanotechnology 20 (2009) 146–152. [35] J. Yu, L. Qin, Y. Hao, S. Kuang, X. Bai, Y.M. Chong, W. Zhang, E. Wang, Vertically aligned boron nitride nanosheets chemical vapor synthesis, ultraviolet light emission, and super hydrophobicity, ACS Nano 4 (2010) 414–422. [36] J. Yu, X. Huang, C. Wu, X. Wu, G. Wang, P. Jiang, Interfacial modification of boron nitride nanoplatelets for epoxy composites with improved thermal properties, Polymer 53 (2012) 471–480. [37] M. Zheng, Y. Gu, Z. Xu, Y. Liu, Synthesis and characterization of boron nitride nanoropes, Mater. Lett. 61 (2007) 1943–1945. [38] T.T. Chau, A review of techniques for measurement of contact angles and their applicability on mineral surfaces, Miner. Eng. 22 (2009) 213–219. [39] M. Wang, S. Deb, W. Bonfield, Chemically coupled hydroxyapatite – polyethylene composite: processing and characterization, Mater. Lett. 44 (2000) 119–124. [40] T.G. Gopakumar, J.A. Lee, M. Kontopoulou, J.S. Parent, Influence of clay exfoliation on the physical properties of montmorillonite/polyethylene composites, Polymer 43 (2002) 5483–5491. [41] J. Hou, L. Guohua, N. Yang, L. Qin, E. Maryam, Q. Zhang, N. Wanga, X. Qu, Preparation and characterization of surface modified boron nitride epoxy composites with enhanced thermal conductivity, RSC Adv. 4 (2014) 44282–44292. [42] X. Cui, P. Ding, N. Zhuang, L. Shi, N. Song, S. Tang, Thermal conductive and mechanical properties of polymeric composites based on solution-exfoliated boron nitride and graphene nanosheets: a morphology-promoted synergistic effect, ACS Appl. Mater. Interfaces 7 (2015) 19068–19075. [43] W.L. Song, P. Wang, L. Cao, A. Anderson, M.J. Meziani, A.J. Farr, Y.P. Sun, Polymer/boron nitride nanocomposite materials for superior thermal transport performance, Angew. Chem. Int. Ed. 51 (2012) 6498–6501. [44] W. Cheewawuttipong, D. Fuoka, S. Tanoue, H. Uematsu, Y. Iemoto, Thermal and mechanical properties of polypropylene/boron nitride, Compos. Energy Procedia 34 (2013) 808–817. [45] D.S. Muratov, D.V. Kuznetsov, I.A. Iinykh, I.N. Burmistrov, I.N. Mazov, Thermal conductivity of polypropylene composites filled with silane-modified hexagonal BN, Compos. Sci. Technol. 111 (2015) 40–43.

[6] C. Damm, J. Körner, W. Peukert, Delamination of hexagonal boron nitride in a stirred media mill, J. Nanopart. Res. 15 (2013) 1561–1572. [7] L.H. Li, Y. Chen, G. Behan, H. Zhang, M. Petravic, A.M. Glushenkov, Large-scale mechanical peeling of boron nitride nanosheets by low-energy ball milling, J. Mater. Chem. 32 (2011) 11862–11866. [8] Y. Lin, T.V. Williams, W. Cao, H.E. Elsayed-Ali, J.W. Connell, Defect Functionalization of Hexagonal Boron Nitride Nanosheets, J. Phys. Chem. C 114 (2010) 17434–17439. [9] D. Golberg, Y. Bando, Y. Huang, T. Terao, M. Mitome, C. Tang, C. Zhi, Boron nitride nanotubes and nanosheets, ACS Nano 4 (2010) 2979–2993. [10] C. Zhi, Y. Bando, C. Tang, H. Kuwahara, D. Goldberg, Large-scale fabrication of boron nitride nanosheets and their utilization in polymeric composites with improved thermal and mechanical properties, Adv. Mater. 21 (2009) 2889–2893. [11] R. Meier, I. Kahraman, A.T. Seyhan, S. Zaremba, K. Drechsler, Evaluating vibration assisted vacuum infusion processing of hexagonal boron nitride sheet modified carbon fabric/epoxy composites in terms of interlaminar shear strength and void content, Compos. Sci. Technol. 128 (2016) 94–103. [12] J.P. Hong, S.W. Yoon, T.S. Hwang, Y.K. Lee, S.H. Won, J.D. Nam, Interphase control of boron nitride/epoxy composites for high thermal conductivity, Korea Aust. Rheol. J. 22 (2010) 259–264. [13] Q. Liu, J. Ding, D.E. Chambers, S. Debnath, S.L. Wunder, G.R. Baran, Fillercoupling agent-matrix interactions in silica/polymethylmethacrylate composites, J. Biomed. Mater. Res. 57 (2001) 384–393. [14] X. Ma, N.H. Lee, H.J. Oh, S.C. Jung, W.J. Lee, S.J. Kim, Morphology control of hexagonal boron nitride by a silane coupling agent, J. Cryst. Growth 316 (2011) 185–190. [15] R.R. Madathingal, Stephanie L. Wunder, Thermal degradation of poly(methyl methacrylate) on SiO2 nanoparticles as a function of SiO2 size and silanol density, Thermochim. Acta 526 (2011) 83–89. [16] L. Gao, J. Li, J. Am, Preparation of nanostructured hexagonal boron nitride powder, Ceram. Soc. 86 (2003) 1982–1984. [17] X. Wei, M.S. Wang, Y. Bando, D. Golberg, Electron-beam-induced substitutional carbon doping of boron nitride nanosheets, nanoribbons, and nanotubes, ACS Nano 5 (2011) 2916–2922. [18] M.S.R.N. Kiran, K. Raidongia, U. Ramamurty, C.N.R. Rao, Improved mechanical properties of polymer nanocomposites incorporating graphene-like BN: dependence on the number of BN layers, Scr. Mater. 64 (2011) 592–595. [19] T. Panagiotou, S.V. Mesite, R.J. Fisher, Production of norfloxacin nanosuspensions using microfluidics reaction technology through solvent/antisolvent crystallization, Ind. Eng. Chem. Res. 48 (2009) 1761–1771. [20] H. Yurdakul, Y. Goncu, O. Durukan, A. Akay, A.T. Seyhan, N. Ay, S. Turan, Nanoscopic characterization of two-dimensional (2D) boron nitride nanosheets (BNNSs) produced by microfluidization, Ceram. Int. 38 (2012) 2187–2193. [21] O. Durukan, I. Kahraman, P. Parlevliet, M. Geistbeck, A.T. Seyhan, Microfluidization, time-effective and solvent free processing of nanoparticle containing thermosetting matrix resin suspensions for producing composites with enhanced thermal properties, Eur. Polym. J. 85 (2016) 575–587. [22] T. Sainsbury, A. Satti, P. May, A.O. Neill, V. Nicolos, Y.K. Guko, J.N. Coleman, Covalently functionalized hexagonal boron nitride nanosheets by nitrene addition, Chem. Eur. J. 18 (2012) 10808–10812. [23] Y. Lin, T.V. Williams, J.W. Connell, Soluble, exfoliated hexagonal boron nitride nanosheets, J. Phys. Chem. Lett. 1 (2010) 277–283. [24] S.F. Xavier, J.M. Schultz, K. Friedrich, Fracturepropagation in particulate- filled polypropylene composites influnce of mica surface treatment, J. Mater. Sci. 25 (1990) 2428–2432. [25] E.P. Plueddeman, Silane Coupling Agent, Plenum Press, A.B.D, New York, 1992. [26] H. Jo, F.D. Blum, Characterization of the interface in polymer- silica composites

107