Polyethylene Nanocomposites A Solution Blending Approach by Kwan
Polyethylene Nanocomposites – A Solution Blending Approach by Kwan Yiew Lau 1, 2 with Prof. Alun S. Vaughan 1 Dr. George Chen 1 Dr. Ian L. Hosier 1 1 University of Southampton 2 Universiti Teknologi Malaysia
Introduction
Background • Polymeric insulators – widely used as standard materials in power delivery systems. • The current commercial trend is to add micro-sized filler into polymer. – Benefits: Enhanced mechanical & thermal properties. – Trade-off: Worsened electrical properties. 3
Background • Introduce nano-sized filler into polymer. – Polymer nanocomposites / nanodielectrics. • Polymers with nanometre-sized fillers homogeneously dispersed at just a few wt%. 4
Research Trend The number of publications in nanodielectrics (Nelson, 2010) • Improved breakdown strength, mitigated space charge formation, enhanced partial discharge resistance, etc. ! y t ni tu r o p p O 5
Problems • Lots of uncertainties concerning nanocomposite applications in dielectric systems remain unanswered. • The mechanisms that lead to the unique dielectric properties of nanocomposites remain unclear. Lack of understanding! especially of the underlying physics and chemistry… 6
Challenges • Dispersion of nanoparticles in polymers. • Small size = agglomeration ≠ single particles. • Various preparation techniques are proposed to obviate, or at least minimise, unwanted clustering effects. 7
Materials and Preparation
Materials • Polymers: – 80 wt% LDPE grade LD 100 BW (Exxon. Mobil Chemicals) – 20 wt% HDPE grade Rigidex HD 5813 EA (BP Chemicals) • Nanofiller: – Si. O 2 nanopowder (Sigma Aldrich), 10 - 20 nm. – Unfunctionalized. 9
Preparation of Materials • Solution blending method: – Nano-Si. O 2 was added into xylene, sonicated for 1 hour. – PE blend was then added. – The mixture was heated to the boiling point of xylene & stirred simultaneously. – The hot mixture was precipitated in methanol. – Filtering, drying and melt pressing. • Unfilled PE prepared in the same way. 10
Results and Discussion
Thermal Analysis All nanofilled PE exhibited reduced induction time and faster crystallisation. Nanoparticles act as nucleation sites. Non-linear Avrami fitting on unfilled PE Non-linear Avrami fitting on 5 wt% nanofilled PE 12
Crystallisation Rate Constant, K 3 • At any given temperature, nanofilled PE shows higher K 3. • 2 wt% - increased K 3. • 5 wt% - higher K 3 data – Increased interactions? – Increased nucleation sites? • 10 wt% - K 3 values saturated. – Suppression effect caused by the reduced growth rate? – Indicative of the onset of nanosilica aggregation? Plot showing the content of nano 13 Si. O 2 on K 3 parameter of PE
Subsequent Melting Behaviour Unfilled PE • Nanofilled PE The melting behaviour was similar, except at Tc = 111 ºC. – Pronounced double peak (unfilled PE) vs. more singular peak (nanofilled PE). 14
Crystallinity • The addition of nano. Si. O 2 does not affect the final crystallinity. • A hint to similar melting trace? Tc = 111 ºC Tc = 115 ºC Tc = 119 ºC Sample 0 wt% X/% X/% 66. 3 58. 5 53. 4 2 wt% 65. 7 58. 9 48. 4 5 wt% 66. 9 57. 8 49. 8 10 wt% 65. 5 58. 4 49. 2 – The thickness of the lamellae is similar. • Nano-Si. O 2 acts as nucleating agent but does not increase the final crystallinity. – Nucleation effect + topological confinement. 15
Polarised Optical Microscopy • For crystallised unfilled PE, spherulites can be clearly observed. • For nanofilled PE: – The size of the spherulites was smaller. 0 wt% 2 wt% 5 wt% 10 wt% – Nano-inclusion appears dramatically to suppress spherulitic development. Crystallised at 117 ºC 16
Scanning Electron Microscopy • Unfilled PE, crystallised 115 ºC: – Open banded spherulitic structures, space filling. • 2 wt% nanofilled PE: – Banded spherulites can still be observed. 0 wt% – Smaller spherulites size. – Nucleation effect. – Nanofiller well distributed, but agglomeration could not be avoided. 2 wt% 17
Scanning Electron Microscopy • Aggregation becomes more apparent with increasing amount of nanofiller. • At 5 wt%, the effect of spherulite banding becomes less pronounced, and the texture was significantly perturbed. 5 wt% • At 10 wt%, the growth of spherulite is largely suppressed, resulted in highly disordered system. 10 wt% 18
AC Breakdown Test Crystallised at 115 ºC Sample �� / k. V mm-1 0 wt% 152 ± 3 19 ± 6 2 wt% 152 ± 2 33 ± 10 5 wt% 150 ± 2 26 ± 7 10 wt% 121 ± 2 21 ± 7 �� • No difference between 0 wt%, 2 wt% and 5 wt%? • Severe aggregations in 10 wt% nanofilled PE reduced the breakdown strength. 19
AC Breakdown Test Quenched Sample 0 wt% �� / k. V mm-1 148 ± 4 �� 16 ± 5 2 wt% 147 ± 4 16 ± 4 5 wt% 144 ± 3 23 ± 7 10 wt% 115 ± 3 16 ± 5 • Same breakdown trend in quenched systems. • Nanosilica does not alter AC breakdown strength. • At severe aggregations, AC breakdown strength would be 20 reduced.
Conclusions and Future Work
Conclusions • Nano-Si. O 2 enhances the nucleation density. – Evidenced from the shorter crystallisation process and higher value of crystallisation rate constant. • The DSC melting traces of the nanocomposites were similar to unfilled PE. – Nano-Si. O 2 did not exert on appreciable effect on the melting behaviour. • Nano-Si. O 2 did not possess significant effect towards the final crystallinity. 22
Conclusions • From POM & SEM: – Nano-Si. O 2 suppresses spherulitic development and thus perturbed the morphological structure of the isothermally crystallised material. • From SEM, nanosilica is well-distributed in PE through solution blending approach. – Agglomeration is unavoidable. • Nano-Si. O 2 does not alter AC breakdown strength of PE. – But the breakdown strength will reduce if the dispersion is poor. 23
Future Work • Dielectric spectroscopy: – Dielectric response of the nanocomposites. – Water absorption behaviour. • Pulse electro-acoustic: – Space charge behaviour. • Surface treatment of nano-Si. O 2. 24
Thank you!
Appendices
Experimental Techniques • DSC – Perkin Elmer DSC 7 with Pyris software. – Sample ~5 mg in a sealed aluminium pan. – Nitrogen atmosphere. – Avrami analysis was performed by DSC. • Heating rate: 10 ºC min-1 • Cooling rate: 100 ºC min-1 • POM – Linkam THM 600 hot stage. – Melt press sample between two microscope slides 27
Experimental Techniques • SEM – JEOL Model JSM-5910. • Gun voltage = 15 k. V; Working distance = 11 mm. – Standard permanganic etching technique. • Permanganic reagent composed of 1 % w/v solution of potassium permanganate in an acid mixture composed of concentrated sulphuric acid, phosphoric acid & water at ratio 5: 2: 1. • After etching, the reagent was quenched using hydrogen peroxide & dilute sulphuric acid at ratio 4: 1. 28
Experimental Techniques • Dielectric Breakdown Test – Samples of ~85 µm in thickness were prepared by using a Specac press (150 ºC, 3 tonne). – Dielectric breakdown test based upon ASTM Standard D 149 -87. – The test sample was placed between two 6. 3 mm ballbearing electrodes immersed in silicone fluid. – An AC voltage of 50 Hz and a ramp rate of 50 V(RMS) s-1 was applied until failure. 29
Avrami Analysis • The crystallinity fraction at time t: • The obtained experimental values of X and t were fitted to the equation using a non-linear approach to estimate the Kexp, ti and n. • Kexp = experimental rate constant or overall crystallisation rate constant containing contributions from both nucleation and growth • n = Avrami exponent or dimensionality of the growth 30
Crystallisation Rate Constant, K 3 • N = the number of nucleation sites per unit volume • G = the growth rate of the crystallising objects 31
Crystallinity Calculation • The enthalpies of melting was determined as a function of crystallisation temperature for each material and then converted into the percentage of HDPE present in each blend that was involved in each phase transition (Mandelkern, 1992). ∆H = melting enthalpy ∆Ho = the value of enthalpy corresponding to the melting of a 100 % crystalline material (293 J g-1 PE) ω = the weight fraction of the crystallisable material. 32
Weibull Analysis • Two-parameter Weibull distribution: P(E) = cumulative probability of failure at E E = experimental breakdown strength α = scale parameter, represents the breakdown strength at the cumulative failure probability of 63. 2 % β = shape parameter • The cumulative probability of failure, P(E) was approximated using the median rank method: i = progressive order of failed tests n = total number of tests 33
SEM Micrographs “Dielectric properties of XLPE/Si. O 2 nanocomposites based on CIGRE WG D 1. 24 cooperative test results” (Tanaka et al. , 2011) IEEE TDEI, 18(5), 1484 -1517 XLPE containing 5 wt% of unfunctionalized nanosilica XLPE containing 5 wt% of functionalized nanosilica 34
- Slides: 34