Research Article www.acsami.org
Silica Nanoparticles Treated by Cold Atmospheric-Pressure Plasmas Improve the Dielectric Performance of Organic−Inorganic Nanocomposites Wei Yan,†,‡ Zhao Jun Han,‡ B. Toan Phung,† and Kostya (Ken) Ostrikov*,‡,§ †
School of Electrical Engineering and Telecommunications, The University of New South Wales, Sydney, NSW 2052, Australia Plasma Nanoscience Centre Australia (PNCA), CSIRO Materials Science and Engineering, Lindfield, NSW 2070, Australia § Plasma Nanoscience, Complex Systems, School of Physics, The University of Sydney, Sydney, NSW 2006, Australia ‡
S Supporting Information *
ABSTRACT: We report on the application of cold atmospheric-pressure plasmas to modify silica nanoparticles to enhance their compatibility with polymer matrices. Thermally nonequilibrium atmospheric-pressure plasma is generated by a high-voltage radio frequency power source operated in the capacitively coupled mode with helium as the working gas. Compared to the pure polymer and the polymer nanocomposites with untreated SiO2, the plasma-treated SiO2−polymer nanocomposites show higher dielectric breakdown strength and extended endurance under a constant electrical stress. These improvements are attributed to the stronger interactions between the SiO2 nanoparticles and the surrounding polymer matrix after the plasma treatment. Our method is generic and can be used in the production of high-performance organic−inorganic functional nanocomposites. KEYWORDS: organic−inorganic nanocomposites, atmospheric pressure plasma, insulation, partial discharge, dielectric breakdown
■
INTRODUCTION The incorporation of inorganic nanoparticles into polymer matrices is very promising to improve the properties of organic−inorganic nanocomposites, which are considered very attractive materials for the development of next-generation, high-performance devices for numerous advanced applications.1−7 One of these applications is high-voltage insulation, where polymer-based organic−inorganic nanocomposites have shown major advantages (e.g., flexibility, processability, good performance, low weight, low cost, etc.) over the conventional insulation materials.7 SiO2, TiO2, and Al2O3 are among the most-widely used inorganic nanoparticles to be filled in the dielectric polymers. However, a common problem associated with these nanoparticles is that they often tend to agglomerate because of surface incompatibility between the inorganic fillers and the polymer matrices. As such, the insulation performance of these nanocomposites is often substantially degraded. Using a coupling agent to modify the surface of nanoparticles prior to mixing them with polymers has been attempted in both laboratories and industries.8 Grafting of organic groups (e.g., amine or silane) on the nanoparticle surfaces has been demonstrated to improve the dispersion uniformity in the polymer matrices.9,10 However, the low surface reactivity of these nanoparticles often inhibits strong interactions between the nanoparticles and the surrounding polymers, resulting in © 2012 American Chemical Society
unsatisfactory improvements of the properties of the nanocomposites. To resolve these issues, here we use thermally nonequilibrium, cold atmospheric-pressure plasmas to modify the surface of the nanoparticle fillers. We show that the plasmasurface reactions can effectively increase the surface energy and the reactivity of these nanoparticles. This in turn results in a better dispersity and stronger interactions with the polymer matrices. Compared to their untreated counterparts, the plasma-treated nanocomposites therefore attain higher dielectric breakdown strength and also longer endurance under the constant electrical stress. These results demonstrate a simple yet effective technique to improve the physical and chemical properties of the functional organic−inorganic nanocomposites, thereby advancing their applications in many fields.
■
EXPERIMENTAL SECTION
Materials. Hydrophobic fumed silica nanoparticles grafted with dimethyldichlorosilane (DDS) were provided by Evonik Industries, Germany. The nanoparticles had an average particle size of 16 nm and a surface area of 110 m2/g. The epoxy resin (RX771C/NC) containing bisphenol A diglycidyl ether (BADGE) was obtained from ROBNOR Received: February 21, 2012 Accepted: April 10, 2012 Published: April 10, 2012 2637
dx.doi.org/10.1021/am300300f | ACS Appl. Mater. Interfaces 2012, 4, 2637−2642
ACS Applied Materials & Interfaces
Research Article
RESINS Ltd. The curing agent (ARADUR HY 1300 GB) consisted of triethylenetetramine (TETA) was supplied by Huntsman Advanced Materials. This thermoset epoxy resin has been widely used as insulation materials in high-voltage applications.11 Atmospheric-Pressure Plasma Treatment. The surface modification of the as-received silica nanoparticles was carried out in a
at room temperature for another 48 h. The fabricated nanocomposites with untreated and plasma-treated SiO2 were denoted as UTNCs and PTNCs, respectively. Pure epoxy resin as a control sample was also fabricated and denoted as PER. Insulation Performance Tests. The partial discharge (PD) characteristic is one of the most important parameters for evaluating the performance of an insulation material as the development of PD causes the degradation to the material and eventually leads to the ultimate breakdown.7 The PD measurements were performed using OMICRON MPD600 PD detection unit with the measurement circuit shown in Figure S3 in the Supporting Information. This circuit included a 1000 pF blocking capacitor (Cb) and a detection impedance (Z). The discharge-free detection system was powered by a highvoltage power supply. Each sample was placed in a test cell filled with insulation oil to avoid any unwanted surface discharge (see Figure S4 in the Supporting Information). The maximum voltage that could be achieved in this system was 50 kV at 50 Hz. In our measurements, a voltage was ramped from 0 V up until the occurrence of an ultimate breakdown to obtain the inception voltage and the dielectric breakdown strength of each sample. The PD inception threshold was set as 10 pC. Additionally, an endurance test under a sustained AC voltage was performed using the same circuit as in the PD measurements, except that the frequency was adjusted to 300 Hz to speed up the aging effect as compared to the 50 Hz power frequency. The applied root-mean-square (rms) voltage for the endurance test Vendu was 7.5 kV. Material Characterizations. X-ray photoelectron spectroscopy (XPS) spectra were obtained in both survey and narrow-scan modes using the SAGE150 Compact ESCA System (SPECS) with a monochromatic Mg Kα radiation at 1253.6 eV. The detected spot size of the samples was 1 × 2 mm2. A Gaussian−Lorentzian (30/70) product function was used for XPS peak deconvolutions. For sample preparation, nanoparticles were first transferred onto a molybdenum sheet, followed by dropping one drop of ethanol. Upon the drying of ethanol, uniform and flat layers of both untreated and plasma-treated nanoparticles were obtained with a good surface coverage. In this way, the contribution from the silica nanoparticles in the XPS spectra was minimized. Field emission scanning electron microscopy (FESEM) images was obtained using a Zeiss Ultra Plus microscope with an electron beam energy of 15 keV. For the SEM imaging, the nanoparticle samples were prepared by directly placing them onto a carbon adhesive tape. In contrast, the nanocomposite samples were viewed by scraping the thin cross-section sheets from the bulk material. Both samples were decorated with gold to prevent surface charging in SEM observations. Electric Field Simulation. To study the effect of nanoparticle agglomeration on the uniformity of electric field in their vicinity, a simulation was carried out using the Maxwell-2D software.17 Specifically, a sphere model was used as the agglomerated nanoparticles in the simulation. Spheres with different diameters were placed in a homogeneous electric field to show the correlation between the agglomeration size and the intensity of the distorted electric field.
Figure 1. Setup of the atmospheric-pressure plasma reactor. The bottom wall of the bottle serves as a dielectric barrier. During the treatment, the nanoparticles were placed in the uniform plasma zone and exposed to the plasma. Inset (top) shows the schematic of a coil electrode placed 5 mm above the bottom of the glass bottle. Inset (bottom) shows a real-time photograph of the plasma discharge. custom-designed atmospheric-pressure plasma reactor (Figure 1). Cold atmospheric-pressure plasma was generated by a 350 kHz radio frequency (RF) power supply with a maximum 5 kV peak-to-peak output. A tin-coated copper coil electrode was placed 2 mm above the top of the nanoparticle layer. The ground electrode was placed under the reactor to form a dielectric barrier discharge (DBD) structure.12 Helium was used as the working gas for the discharge. The plasma was generated by a 4 kV peak-to-peak RF voltage (the voltage and current waveforms are shown in Figure S1 in the Supporting Information). The power density in the plasma region was estimated to be 60 mW/ cm3.13−16 In each experiment, we have put 50 mg of the nanoparticles on the bottom of the reactor. These particles formed an evenly distributed, 3 mm thick soft layer. As the outer diameter of the coil was close to the diameter of the cylindrical reactor, the plasma zone generated by the coil electrode can fully cover the whole surface of the nanoparticle layer (also see the inset of Figure 1). We also observed that the plasma can partially penetrate the thin nanoparticle layer and reach the bottom layer of the nanoparticles during the experiments. Moreover, the nanoparticles remained powder-like and no clear changes in the morphology of the nanoparticle layer were found after the plasma treatment. Occasionally, however, it was noted that the treated particles tended to attach to the wall of the reactor. To obtain a fairly uniform exposure, the nanoparticles were first treated by the atmospheric-pressure plasma for 5 min; the plasma was then stopped and the nanoparticles were stirred using a magnetic bar for 10 s. This treatment-stirring process was repeated 6 times and the total treatment time of the nanoparticles was 30 min. Synthesis of Polymer Nanocomposites. First, 0.71 g of curing agent was added into 1.79 g of epoxy resin to obtain a 4:10 weight ratio. High-speed mechanical stirring was continuously performed to obtain a uniform fluid. Due to the high viscosity of the fluid, 0.05 g of SiO2 nanoparticles were dispensed into the resin in three batches, with an interval of 10 min. After being stirred for 40 min, the mixture was degassed for 1 h under vacuum (
