Studies on the effects of layered silicate content on the AC electrical treeing and breakdown behaviors of epoxy/ layered silicate nanocomposites were carried out in needle-plate electrode geometry. Wide-angle X-ray diffraction (WAXD) analysis and transmission electron microscopy (TEM) observation showed that 1 wt% of the multilayered silicate was fully exfoliated into nano-sized monolayers in the epoxy matrix however, over 3 wt% of the silicate was in an intercalated state. When 1 wt% layered silicates were incorporated, an electrical tree was initiated in 439 min and propagated at a speed of 2.3 μm/min after applying 781.4 kV/mm, representing a decreased in starting initiation time by a factor of 11.0 and increase in propagation speed by a factor 8.2 in comparison with neat epoxy resin. These values were in great decline after the layered silicate content was increased to 3wt% which implies that the exfoliated silicate blocked the tree initiation and propagation processes effectively. However the effect was largely decreased in the intercalated state. Nanocomposites have special characteristics that may be useful in a number of advanced functional applications. The recent surge of interest in high voltage insulating materials based on layered silicates and epoxy matrices has been inspired by the realization of these nanocomposites which exhibit unusual ac insulation breakdown strength and partial discharge resistance as compared to their bulk counterparts [1 - 3] . Therefore, in the last two decades many researchers have investigated application of these nanocomposites in various heavy electric equipment such as mold-type transformers, current transformers (CT), potential transformers (PT), metering out-fit (MOF) and gas switching gears [4 , 5] . Much work has also been done to estimate the long-term insulation performance of the polymer/layered silicate nanocomposites by studying the electrical treeing phenomena in non-uniform electric fields, having been considered the most important mechanism in the deterioration of polymeric insulators (e.g. high voltage polymeric cables) [6 - 8] . The tree growth mechanism was divided into three processes: (1) the incubation process, (2) the initiation process, and (3)the propagation process [9] . Once an electrical tree was initiated, it would be propagated rapidly and until breakdown occurred. Ideally, the initiation time should be delayed and the propagation rate retarded in order for good insulation. These processes were caused by complicated mechanisms including charge injection-extraction, collision ionization, oxidation decomposition, partial discharge, and partial high temperature. In this study, insulation material was prepared using a layered silicate of 1~5 wt% that was incorporated into an epoxy matrix using a 3-roll mill, after which AC electrical treeing and breakdown tests were carried out using needle-plate electrodes in order to estimate the long-term insulation performance. Cloisite ^® 10 A (Southern Clay Products, Inc.) was used as a layered silicate which was modified with dimethyl-benzyl-hydrogenated tallow quaternary ammonium, where the hydrogenated tallow consisted of C18 (~65%), C16 (~30%), and C14 (~5%), and the anion was chloride. Diglycidyl ether of bisphenol A (DGEBA) type epoxy resin whose trade name was YD 128 (Kukdo Chem. Co. Korea) was used. The epoxy equivalent weight (EEW) was 184~190 g/eq. The Curing agent was 3- or 4-methyl-1,2,3,6- tetrahydrophthalic anhydride (Me-THPA) whose trade name was HN-2200 (Hitachi Chem. Co., Japan). A tertiary amine type accelerator benzyl-dimethyl amine (BDMA, Kukdo Chem. Co. Korea) was also used. A needle-type steel electrode was purchased from Ogura Jewelry Co., Japan. Its diameter and length were 1 mm and 60 mm, respectively with a tip angle of 30° and a curvature radius of 5 μm. The bottom-side of the specimen was coated with a conductive silver paste to form a plate electrode and a plate-type copper electrode with 45×45 mm ² was used as the grounded electrode. The epoxy base resin and layered silicate were shear mixed with a 3-roll mill (EXAKT 80E, EXAKT Advanced Technologies GmbH), and then 80parts-per-hundred (phr) Me-THPA and BDMA (0.9 phr) were mixed with a mechanical stirrer for 10 minutes. The layered silicate content in the epoxy nanocomposite was 1~5 wt%. The mixture was poured into a 15×15×30 mm ³ mould after a needle electrode was placed at a distance of 2.7 mm from the plate electrode. It was then cured at 120℃ for 2 hours and postcured at 150℃ for 2 hours then cooled slowly at a rate of -0.5℃/min to room temperature to avoid internal stress. Finally,the opposite side of the needle electrode in the epoxy specimen was coated with the conductive silver paste. The specimen shape is as shown in Fig. 1 [10] . To measure the tree initiation and propagation, and the breakdown rate, a constant alternating current (AC) of 15 kV (60 Hz) was applied to the specimen in the needle-plate electrode arrangement in a 30℃ insulating oil bath. The specimens were tested. The process was monitored by a video microscope system (ICS-305B, SOMETECH Inc.), as shown in Fig. 1 . High voltage (HV) was applied with an AC Endurance Voltage Tester (Haefely, Germany) and the voltage was increased at a rate of 1 kV/s to a maximum of 15 kV, where it was held until electrical breakdown took place. Tree images were collected every 1 minutie. The change of interlayer distance was measured by wide-angle X-ray diffraction (WAXS, XRD30, Rigaku). The X-ray beam was nickel-filtered Cu K1 (λ = 0.154 nm) radiation operated at a tube voltage of 40 kV and tube current of 30 mA. TEM observation was carried out using a high-resolution transmission electron microscope (Hitachi S-4100). Figure 2 shows TEM images for the layered silicates dispersed in an epoxy matrix and magnified views. The contents of the layered silicate were (in a and b) 1 wt% and (for c and d) 3 wt%. The only difference between (a) and (b) or (c) and (d) was magnification. Figures 2 (a) and 2 (c) showed that evenly Dispsersed layered silicate in the epoxy matrix could be achieved using a 3-roll mill regardless of the silicate content. As 1 wt% of the layered silicate was mixed, the dark lines corresponding to the silicate monolayers were arranged in a disorderly fashion greater lengths in Fig. 2 (b). However they were displayed with ordered arrangement when 3 wt% of the layered silicate was mixed, as shown in Fig. 2 (d). This indicates that the interlayer distance of the layered silicate became broader via the penetration of epoxy resins into intergalleries during the shear mixing and curing, and finally forming an exfoliated state in the 1 wt% system. However, in the 3 wt% system, there was too much layered silicate to be fully exfoliated by the penetration of epoxy resin. These results were also confirmed by WAXD analysis. Figure 3 shows WAXD patterns for the layered silicate powder and cured epoxy nanocomposites with layered silicate (1 and 3 wt%), with 120℃ 2hour cure and 150℃ 2hour postcure. The characteristic peak for the interlayer distance (d-space) of the layered silicate powder was shown to be 2 θ=4.70°, which could be converted to 1.88 nm by Bragg’s formula [11] . As 1 wt% of the layered silicate was mixed, the peak d-space almost disappeared with a faint peak at 2 θ=2.50o (d-space=3.53 nm). This indicate that the interlayer distance of the layered silicate became broader and exfoliated during shear mixing and curing. However, for layered silicate content at 3 wt%, a new characteristic peak at 2 θ=2.50° (3.53 nm) appeared. This implies that the layered silicate would have orderly silicate layers, in other words, it existed in the intercalated state. Results agreed well TEM observation shown in Fig. 2 . Figure 4 shows the tree growth rate curve in the epoxy/layered silicate (1 wt%) in a constant electrical field of 781.4 kV/mm (60 Hz) at 30℃. The electrical tree morphology corresponding to photos (a) and (b) was collected after applying HV for (a) 5,270 minutes and (b) 15,680 minutes. When 15 kV was applied to the specimen in the needle-plate electrode arrangement, the electrical field at the needle tip, E _tip =781.4 kV/mm was calculated using Masons formula, where r is the needle tip radius, V is the applied voltage, and x is the needle-plate distance. An electrical tree began to form after 439 minutes and propagated rapidly at the speed of 2.3 μm/min, and breakdown took place after 1,592 min when 781.4 kV/mm was applied. Typical behaviors of branch type electrical trees were obtained from the morphology observation in Figs. 4 (a) and 4 (b). For charges injected into and extracted from the epoxy insulation from the needle tip, and small electrical tree branches were initiated as shown in Fig. 4 (a). In the next propagation process, for charges injected into and extracted from the epoxy insulation from the carbonized conductive tree tip, several new branches that were darker and more pronounced appeared and grew rapidly ( Fig. 4 (b)). E _tip = 2 V/(r·ln(1 + 4x/r) The inset in Fig. 4 shows the tree growth rate curve in the neat epoxy resin under the same conditions. An electrical tree was began to form after 40 minutes and propagated very rapidly at speed of 19.0 μm/min, and eventually broke down in 182 minutes when HV was applied. As expected, the initiation time decreased by a factor of 11.0 and average propagation rate increased by 8.2 times, in comparison to the epoxy/layered silicate preparation (1.0 wt%). However branch type trees, with relatively few branches, appeared and are shown in Figs. 4 (c) and 4 (d). To study the effect of the layered silicate content on treeing behaviors, tree initiation time and breakdown time were measured under a constant electric field of 781.4 kV/mm (60 Hz) at 30℃ the results of which are displayed in Fig. 5 , These were obtained from the scale parameter of the tree data analyzed by Weibull statistical methods. The scale parameter represents the treeing behaviors with 63.2% of the cumulative probability. When 1 wt% nano-sized silicate layers were incorporated, the tree initiation time and the breakdown time was sharply increased. However, these values declined significantly after 3 wt% of the layered silicate which implies that the exfoliated nano silicate layers (1 wt%) effectively blocked tree formation and propagation. But the effect was largely decreased with the increasing silicate content, and for layered silicate over 3 wt% in the intercalated state, This was confirmed by WAXD analysis in shown in Fig. 3 . Consequently, it was found that the well-dispersed exfoliated silicate could sufficiently improve the electrical stability of the neat epoxy. Effects of layered silicate content on AC electrical treeing and breakdown behaviors in epoxy/layered silicate nanocomposites were carried out in needle-plate electrode geometry. In the epoxy/layered silicate (1 wt%) nanocomposite, the electrical tree began to form after 439 minutes and propagated at the speed of 2.3 μm/min, and finally breakdown took place after 1,592 min when applying 781.4 kV/mm. However, those values in the neat epoxy resin were 40 minutes, 19.0 μm/min and 182 minutes, respectively. The initiation time decreased by a factor of 11.0 and the average propagation rate increased by 8.2 times, as the layered silicate (1.0 wt%) was incorporated. These values were in great decline for layered silicate contents over 3 wt% which implied that the exfoliated silicate blocked tree initiation and propagation. The effect decreased significantly in the intercalated state. All observed electrical trees showed branch type morphology, but that of the epoxy/layered silicate (1 wt%) nanocomposite was much more complicated with more branches than trees in neat epoxy resin, which meant that the injected electrons couldn’t pass through the exfoliated layered silicate, and had to form new paths to avoid the clusters of layered silicates.