POLYMER COMPOSITION AND ELECTRICAL DEVICES

Abstract

The invention relates to a polymer composition comprising a polymer (a) and a nanoparticle filler (b), wherein the polymer composition comprises a volume percentage (vol. %) of the nanoparticle filler (b), which is Dvol vol. %, and has a center-to-center average distance, in nanometer (nm), in two dimensions (2D) and with a free radius, from one nanoparticle to its nearest nanoparticle neighbour, which is R1st nm, and wherein the polymer composition shows a dependency between said center-to-center average distance to nearest neighbour, R1st, and said volume percentage, Dvol vol. %, which is R1st=E/(Dvol+0.3)+F, wherein Dvol.sub.1DvolDvol.sub.2, E.sub.1EE.sub.2, F.sub.1FF.sub.2, and Dvol.sub.1 is 0.010 and Dvol.sub.2 is 4.4, E.sub.1 is 100 and E.sub.2 is 280, and F.sub.1 is 50 and F.sub.2 is 140; an electrical device, e.g. a power cable; and a process for producing an electrical device.

Claims

1. A polymer composition comprising a low-density polyethylene (LDPE) polymer (a) and a nanoparticle filler (b), wherein the nanoparticle filler (b) comprises a plurality of surface modified nanoparticles, the plurality of surface modified nanoparticles comprising a plurality of nanoparticles silanized with octadecyl(trimethoxy)silane (OdTMS-C18), octyl(triethoxy)silane (OTES-C8), methyltrimethoxysilane, or a combination thereof, and the plurality of nanoparticles comprising MgO or ZnO, wherein the polymer composition comprises a volume percentage (vol. %) of the nanoparticle filler (b) of Dvol vol. %, wherein each nanoparticle in the nanoparticle filler within the polymer composition is separated from its nearest nanoparticle neighbor by a center-to-center average distance, in nanometer (nm), in two dimensions (2D), of R1st nm, and wherein the polymer composition shows a dependency between said center-to-center average distance to nearest neighbor, R1st, and said volume percentage, Dvol vol. %, which is
R1st=E/(Dvol+0.3)+F, wherein Dvol.sub.1DvolDvol.sub.2, E.sub.1EE.sub.2, F.sub.1FF.sub.2, and Dvol.sub.1 is 0.010, Dvol.sub.2 is 4.4, E.sub.1 is 100, E.sub.2 is 280, F.sub.1 is 50, and F.sub.2 is 140.

2. (canceled)

3. (canceled)

4. (canceled)

5. The polymer composition according to claim 1, wherein Dvol.sub.1 is 0.20 and Dvol.sub.2 is 2.5.

6. The polymer composition according to claim 1, wherein E.sub.2 is 275.

7. The polymer composition according to claim 1, wherein F.sub.2 is 135.

8. The polymer composition according to claim 1, wherein E.sub.2 is 270 and F.sub.2 is 130.

9. The polymer composition according to claim 1, wherein the polymer composition has a level of charging currents of 110.sup.9 ampere (A) or less after applying a direct current (DC) voltage of 2.6 kV to a ca. 80 m thick sample of the polymer composition at 32 kV/mm and at 60 C. for 10.sup.3 seconds.

10. The polymer composition according to claim 1, wherein the amount of LDPE polymer (a) in the polymer composition is at least 35 wt. % of the total weight of polymer component(s) present in the polymer composition.

11. (canceled)

12. (canceled)

13. (canceled)

14. (canceled)

15. (canceled)

16. (canceled)

17. The polymer composition according to claim 1, wherein the LDPE polymer (a) is an optionally unsaturated LDPE homopolymer or an optionally unsaturated LDPE copolymer of ethylene with one or more comonomer(s).

18. The polymer composition according to claim 1, wherein the LDPE polymer (a) is an unsaturated LDPE copolymer of ethylene with at least one polyunsaturated comonomer and optionally with one or more other comonomer(s).

19. (canceled)

20. (canceled)

21. (canceled)

22. The polymer composition according to claim 1, wherein the polymer composition is extruded.

23. An electrical device comprising the polymer composition of claim 1, wherein the electrical device is a power cable; a capacitor film; or a photovoltaic (PV) module.

24. A direct current (DC) power cable comprising a conductor which is surrounded at least by an inner semiconductive layer, an insulation layer, and an outer semiconductive layer, in that order, wherein at least the insulation layer comprises the polymer composition according to claim 1.

25. (canceled)

26. A process for producing an electrical device, wherein the process comprises a step of dry processing the polymer composition according to claim 1.

27. A method of use of a surface modified nanoparticle filler, the method comprising using the surface modified nanoparticle filler to reduce the conductivity of a polymer composition, wherein the polymer composition comprises a low-density polyethylene (LDPE) polymer (a) and the surface modified nanoparticle filler (b), wherein the surface modified nanoparticle filler (b) comprises a plurality of silanized nanoparticles comprising an inorganic oxide, wherein the polymer composition comprises a volume percentage (vol. %) of the surface modified nanoparticle filler (b) of Dvol vol. %, wherein each nanoparticle in the surface modified nanoparticle filler within the polymer composition is separated from its nearest nanoparticle neighbor by a center-to-center average distance, in nanometer (nm), in two dimensions (2D), of R1st nm, and wherein the polymer composition shows a dependency between said center-to-center average distance to nearest neighbor, R1st, and said volume percentage, Dvol vol. %, which is
R1st=E/(Dvol+0.3)+F, wherein Dvol.sub.1DvolDvol.sub.2, E.sub.1EE.sub.2, F.sub.1FF.sub.2, and Dvol.sub.1 is 0.010, Dvol.sub.2 is 4.4, E.sub.1 is 100, E.sub.2 is 280, F.sub.1 is 50, and F.sub.2 is 140.

28. The polymer composition according to claim 18, wherein the polyunsaturated comonomer comprises a straight carbon chain with at least 8 carbon atoms and at least two non-conjugated carbon-carbon double bonds, wherein the polyunsaturated comonomer comprises at least 4 carbons between the at least two non-conjugated carbon-carbon double bonds, and wherein at least one of the non-conjugated carbon-carbon double bonds is terminal.

29. The polymer composition according to claim 18, wherein the polyunsaturated comonomer comprises a diene with at least eight carbon atoms and two non-conjugated carbon-carbon double bonds, wherein at least one of the non-conjugated carbon-carbon double bonds is terminal.

30. The polymer composition of claim 18, wherein the polyunsaturated comonomer is a C.sub.8- to C.sub.14-non-conjugated diene with at least one terminal double bond.

31. The polymer composition of claim 18, wherein the polyunsaturated comonomer comprises 1,7-octadiene; 1,9-decadiene; 1,11-dodecadiene; 1,13-tetradecadiene; 7-methyl-1,6-octadiene; 9-methyl-1,8-decadiene; or mixtures thereof.

32. The polymer composition of claim 18, wherein the polyunsaturated comonomer comprises 1,7-octadiene; 1,9-decadiene; 1,11-dodecadiene; 1,13-tetradecadiene; or any mixture thereof.

33. The electrical device of claim 23, wherein the electrical device is a high voltage (HV) power cable or an ultra high voltage (UHV) power cable.

Description

LEGENDS OF FIGURES

[0345] FIG. 1. Reconstruction in three dimensions of the quantification of the distance for particle interaction on the charge current at different weight percentages in the LDPE/MgO-systems, where the systems in the upper row comprise unmodified MgO nanoparticles and the systems in the lower row comprise C8-modified MgO nanoparticles, i.e. octyl(triethoxy)silane (OTES-C8) silanized MgO nanoparticles. Starting from the left in the top row in FIG. 1: FIG. 1a, FIG. 1b and FIG. 1c, and from the left in the bottom row in FIG. 1: FIG. 1d, FIG. 1e and FIG. 1f.

[0346] FIG. 2 shows IR spectra of unmodified (bottom curve), octa(triethoxy)-silane-modified MgO (middle curve) and octadecyl(trimethoxy)silane-modified (top curve) MgO by using FT-IR technique. The curves have been shifted along the y-axis for visibility.

[0347] FIG. 3 shows thermogravimetric data of unmodified, octa(triethoxy)silane-modified MgO and octadecyl(trimethoxy)silane-modified MgO heated at 10 C./min under nitrogen flow.

[0348] FIG. 4 shows X-ray photospectroscopy (XPS) data of the C1s signal from unmodified, C8-modified MgO and C18-modified MgO. A reduction in carbonated species at 289.9 eV, formed by adsorption of ambient CO.sub.2, can be seen for the unmodified MgO.

[0349] FIG. 5a shows X-ray photospectroscopy data of the O 1s signal of unmodified MgO.

[0350] FIG. 5b shows X-ray photospectroscopy data of the O 1s signal of C8-MgO.

[0351] FIG. 5c shows X-ray photospectroscopy data of the O 1s signal of C18-MgO.

[0352] FIG. 6. A sketch of an electrometer (Keithley 6517A) measuring the charging current.

[0353] FIGS. 7a, 7b and 7c show the charging current against time at 32 kV/mm at 60 C.:

[0354] FIG. 7a. LDPE/MgO unmodified filler comparative examples.

[0355] FIG. 7b. LDPE/MgO C8-coated nanofiller (the MgO C8-coated nanofiller is herein also called C8-modified MgO nanoparticles), i.e. example of the polymer composition of the present invention.

[0356] FIG. 7c. LDPE/MgO C18-coated nanofiller (the MgO C18-coated nanofiller is herein also called C18-modified MgO nanoparticles), i.e. example of the polymer composition of the present invention.

[0357] FIGS. 8a and 8b show the conductivity with an applied electrical field of 32 kV/mm after 10 min (FIG. 8a) and 11 hours (FIG. 8b) for LDPE/MgO nanocomposites, where circles represent nanocomposites comprising unmodified MgO (i.e. comparative examples), squares represent nanocomposites comprising C8-modified MgO, herein also called LDPE/MgO C8-coated nanofiller, (i.e. inventive examples) and diamonds represent nanocomposites comprising C18-modified MgO, herein also called LDPE/MgO C18-coated nanofiller, (i.e. inventive examples) and the dashed line is the corresponding value for reference polyethylene.

[0358] FIGS. 9a and 9b present the scanning electron microscopy images (i.e. micrographs) of the cryo-fractured LDPE/MgO nanocomposites, which images show the overall dispersion of the MgO phase in the PE matrix for the different nanocomposites. The unmodified MgO nanoparticles displayed poor particle dispersion, see FIG. 9a, as compared to the surface modified MgO nanoparticles, see FIG. 9b.

[0359] FIG. 10 shows an SEM-image of LDPE/MgO nanocomposite converted into a black and white image to facilitate particle distribution analysis.

[0360] FIG. 11a shows the average center-to-center distance, in nm in 2D for the nanoparticles in the LDPE/MgO nanocomposites, to nearest (i.e. the 1.sup.st) neighbour with free radius as a function of volume percentage (vol. %), where circles represent unmodified MgO, squares represent C8-modified MgO and diamonds represent C18-modified MgO. The volume percentages correspond to 1, 3, 6, and 9 wt. % MgO in the LDPE.

[0361] FIG. 11b shows the average center-to-center distance, in nm in 2D for the nanoparticles in the LDPE/MgO nanocomposites, to 51.sup.st neighbour with a free radius as a function of volume percentage (vol. %), where circles represent unmodified MgO, squares represent C8-modified MgO and diamonds represent C18-modified MgO. The volume percentages correspond to 1, 3, 6, and 9 wt. % MgO in the LDPE.

[0362] FIG. 11c shows the fraction of particles tied up in aggregates, i.e. the ratio between MgO-phase larger than 2 particles (>132 nm) and the complete area of MgO-phase is shown, with unmodified MgO consistently displaying a high ratio of aggregated particles, where a ratio equal to 1 means that all particles are aggregated.

[0363] FIG. 12 shows the conductivity after 10 min plotted as a function of the interaction radius of the MgO nanoparticles, provided that the total volume was filled to 95% by interaction spheres (see large light gray spheres in FIG. 1). The horizontal line in the top gives the value of the unfilled LDPE.

[0364] FIG. 13 shows scanning electron micrograph of the unmodified MgO nanoparticles dispersed in the polyethylene and the poor adhesion between the unmodified MgO aggregate and the polyethylene matrix is here visible.

[0365] FIG. 14 shows the infrared spectra of the pristine, i.e. uncoated or unmodified 25 nm ZnO nanoparticles (ZnO-25-U), (bottom curve) and silane-coated nanoparticles, i.e. silane-modified ZnO nanoparticles: C1 is methyltrimethoxysilane (ZnO-25-C1) (second from bottom curve), C8 is octyltriethoxysilane (ZnO-25-C8) (second from top curve) and C18 is octadecyltrimethoxysilane (ZnO-25-C18) (top curve) by using FT-IR technique. The curves have been shifted along the y-axis for visibility.

[0366] FIG. 15 shows the normalized mass plotted as a function of temperature for pristine and silane-coated ZnO nanoparticles (ZnO-25-U) after normalization to the mass loss value at 140 C.

[0367] FIGS. 16a-16d show scanning electron micrographs of ZnO particles with different size and their LDPE nanocomposites based on the 3 wt. % nanoparticles.

[0368] FIG. 16a shows that zinc acetate precursor yielded separate and mono-domain nanoparticles (ZnO-25-U) (the number in the middle of the names indicates the average size of the particles in nm).

[0369] FIG. 16b shows nanocomposite based on the C8-coated (or C8-modified) ZnO nanoparticles, i.e. ZnO-25-C8 (the number in the middle of the names indicates the average size of the particles in nm).

[0370] FIG. 16c shows that zinc nitrate precursor yielded submicron (ZnO-550-U) star-shaped particles with a symmetrical habit.

[0371] FIG. 16d shows LDPE/3 wt. % ZnO-550-U.

[0372] FIG. 17 shows scanning electron micrograph of LDPE nanocomposites based on 3 wt. % unmodified ZnO nanoparticles (ZnO-25-U).

[0373] FIG. 18 shows stress-strain curves of unfiled LDPE and LDPE nanocomposites filled with 3 wt. % of ZnO nanoparticles (ZnO-25-U) with different surface coating. The second yield point positions are indicated by arrows in the graph.

[0374] FIG. 19a-19c show charging current of pristine LDPE and its nanocomposites based on different weight fraction of ZnO nanoparticles (ZnO-25-U) with different surface coating obtained at 2.6 kV (E=32.5 kV mm-1) at 60 C.

[0375] FIG. 19a shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C1, and unfilled LDPE (Reference).

[0376] FIG. 19b shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C8, and unfilled LDPE (Reference).

[0377] FIG. 19c shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C18, and unfilled LDPE (Reference).

[0378] FIG. 20 shows charging current of pristine LDPE and its nanocomposites based on different weight fraction of ZnO submicron particles (ZnO-550-U) obtained at 2.6 kV (E=32.5 kV mm-1) at 60 C.

[0379] FIG. 21 shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C1 with a nanoparticle content of 3 wt. % ZnO.

[0380] FIG. 22 shows the corresponding bitmap of FIG. 21 with ZnO nanoparticles as the black phase.

[0381] FIG. 23a shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C8 with a nanoparticle content of 3 wt. % ZnO.

[0382] FIG. 23b shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C18 with a nanoparticle content of 3 wt. % ZnO.

[0383] FIGS. 24 and 25 show the overall dispersion of the ZnO phase in the PE matrix for the nanocomposites with different weight fractions of nanoparticles, i.e. regarding the aggregated ZnO nanoparticles as discrete phases (in accordance with FIG. 22). Note that ZnO_25_C1, ZnO_25_C8 and ZnO_25_C18 in FIGS. 24 and 25 mean ZnO-25-C1, ZnO-25-C8 and ZnO-25-C18 as defined herein.

[0384] FIG. 24 shows centre-to-centre distance for the 1.sup.st neighbour as a function of volume percentage of 25 nm ZnO nanoparticles surface modified with C1, C8 and C18 alkyl chain on the silane at 1 wt. % and 3 wt. %.

[0385] FIG. 25 shows centre-to-centre distance for the 51.sup.st neighbour as a function of volume percentage of 25 nm ZnO nanoparticles surface modified with C1, C8 and C18 alkyl chain on the silane at 1 wt. % and 3 wt. %.

EXPERIMENTAL SECTION

Experimental

Synthesis of MgO Nanoparticles and Surface Modification (i.e. the Nanoparticle Filler (b) Preparation)

[0386] Aqueous precipitated Mg(OH).sub.2 was synthesized accordingly to Pallon et al, J. Mater. Chem. A, 2015, 3, 7523, by adding 1 L of a 0.75 M magnesium chloride solution (MgCl.sub.2.6H.sub.2O, ACS Reagent, Sigma-Aldrich) to a 1 L 1.5 M sodium hydroxide solution (NaOH, 98%, Sigma-Aldrich) in stoichiometric balance under rapid stirring (400 rpm). The precipitate was washed with milliQ-water in three cycles and Na.sup.+, Cl.sup. and other residual reagents were removed by centrifugation, using a Rotina 420 centrifuge (Hettich) and ultrasonic bath to break clusters (DTH 2510, Branson), see A. M. Pourrahimi, D. Liu, L. K. H. Pallon, R. L. Andersson, A. Martinez Abad, J.-M. Lagarn, M. S. Hedenqvist, V. Strm, U. W. Gedde and R. T. Olsson, RSC Adv., 2014, 4, 35568-35577. The precipitate was dried at 90 C. overnight and ground before calcination into MgO platelets for 1 h at 400 C. in a muffle furnace (ML Furnaces). The properties of the MgO nanoparticles formed during calcination from Mg(OH).sub.2 were carefully characterized by Pallon et al, J. Mater. Chem. A, 2015, 3, 7523, using X-ray powder diffractometry (XRD), BET (Brunauer, Emmett and Teller) (specific surface area 167 m.sup.2/g), and scanning- and transmission microscopy (average size of a MgO nanoparticle was 66 nm) studies were performed to verify the applicability of the synthesised nanoparticles. Upon precipitation the Mg(OH).sub.2 formed rounded hexagonal platelets with a mean diameter of 43 nm and a thickness of 10 to 20 nm. During the calcination and phase transformation, MgO retained the shape of the Mg(OH).sub.2 particles, while the crystal lattice changed from hexagonal Mg(OH).sub.2 to cubic MgO (same crystal structure as NaCl) and several crystallites (ca 10 nm) were formed inside the retained particle shape. Due to the weak polycrystallite structure the MgO nanoparticles did not always maintain their structure during the processing with ultrasound and extrusion, but broke apart into individual crystallites (ca 10 nm).

[0387] Anhydrous silanzation was performed in n-heptane to avoid a phase transformation into Mg(OH).sub.2. The calcined MgO powder was dispersed in n-heptane (>99%, VWR) with an ultrasonic bath (DTH 2510, Branson) and then transferred to a ball reactor. 0.5 g MgO was dispersed in 0.165 L n-heptane, and under rapid stirring (400 rpm) 0.9 mL of octadecyl(trimethoxy)silane (OdTMS-C18) was added. For the octyl(triethoxy)silane (OTES-C8) 4.5 mL was used.

[0388] The smaller amount of OdTMS-C18 was used to compensate for the higher reactivity of the methoxygroups of OdTMS-C18, see E. P. Plueddemann, Silane Coupling Agents, 2.sup.nd ed., 1991, Springer, New York (ch3, pp 56). The reaction proceeded for 24 hours, after which the nanoparticles, i.e. the nanoparticle filler (b), were washed with heptane in three cycles to remove excess silanes (using the Rotina 420 centrifuge (Hettich) and that ultrasonic bath). The MgO nanoparticles will herein also be referred to as UN-MgO for the unmodified MgO, C8-MgO for the octyl(triethoxy)silane modified MgO and C18-MgO for the octadecyl(trimethoxy)silane modified particles.

Synthesis of ZnO Nanoparticles and Surface Modification (i.e. the Nanoparticle Filler (b) Preparation)

[0389] Zinc nitrate hexahydrate (Zn(NO.sub.3).sub.2.6H.sub.2O, 98 wt. %, Sigma Aldrich), zinc acetate dihydrate (Zn(CH.sub.3COO).sub.2.2H.sub.2O, 99%, Sigma Aldrich), sodium hydroxide (98 wt. %, Sigma Aldrich), methyltrimethoxysilane (CAS number 1185-55, referred to as C1, 98%, 178.3 Da, Sigma Aldrich), octyltriethoxysilane (CAS number 2943-75-1, referred to as C8, 98%, 276.5 Da, Sigma Aldrich) and octadecyltrimethoxysilane (CAS number 3069-42-9, C18, 90%, technical grade, 374.7 Da, Sigma Aldrich), ammonia hydroxide (25 wt. %, Sigma Aldrich), 2-propanol (99.5 wt. %, VWR), ethanol (96 wt. %, VWR), n-heptane (99 wt. %, VWR), Irganox 1076 (CAS number 2082-79-3, octadecyl-3-(3,5-di-tert-butyl-4-hydroxyphenyl)-propionate), Ciba Specialty Chemicals, Switzerland) and potassium bromide (KBr, 98 wt. %, FTIR grade, Sigma Aldrich) were used as received. High resistivity Milli-Q water (18.2 M cm at 25 C.) was used in all the aqueous reactions.

[0390] ZnO nanoparticles were prepared by an aqueous precipitation method described by A. M. Pourrahimi et al., RSC Adv., 2014, 4, 35568-35577, and A. M. Pourrahimi et al., J. Mater. Chem. A, 2015, 3, 17190-17200. ZnO nanoparticles with an average size of 25 nm were prepared by adding a 0.5 M NaOH aqueous solution to a 0.2 M zinc acetate aqueous solution under vigorous stirring. ZnO particles with an average size of 550 nm were prepared by precipitation of a 0.5 M NaOH aqueous solution and a 0.2 M Zn(NO.sub.3).sub.2 aqueous solution. ZnO particles ca. 2 m in size were prepared by using half the concentrations of both the Zn(NO.sub.3).sub.2 and NaOH precursors used in the synthesis of the 500 nm particle preparation. The ZnO particles were purified thrice in Milli-Q water under ultrasonication, dried at 80 C. and normal pressure, ground to a fine powder with a pestle and mortar and finally dried at 60 C. and 20 kPa for 2 h.

[0391] 0.6 g ZnO nanoparticles (ZA-8g) nanoparticles were dispersed in a solution of water (40.8 mL) and 2-propanol (188.4 mL), and then ultrasonicated for 15 min to obtain a homogeneous suspension. Ammonia (25 wt. %; volume=V.sub.ammonia) was added to the suspension under vigorous stirring. The suspension was stirred for 15 min, and silane (volume=V.sub.silane) was finally added and allowed to react for 3 h at room temperature. The quantities of the different compounds used for the different surface modifications are listed in Table 1. The coated particles were centrifuged and washed thrice with ethanol and dried overnight at 80 C. under reduced pressure (0.5 kPa).

TABLE-US-00001 TABLE 1 Reaction parameters in the silanization of ZnO nanoparticles (ZnO-25-U) V.sub.ammonia V.sub.silane (mL) (mL) C1-coated 1.28 5.40 C8-coated 5.10 5.40 C18-coated 1.28 0.41

[0392] The metal oxide particles used in this study were divided into three categories, zinc oxide particles (designated ZnO-25-U, ZnO-550-U, ZnO-25-C1, ZnO-25-C8 and ZnO-25-C18). The number in the middle of the names indicates the average size of the particles in nm. The surface functionality of the particles is indicated by the last part of the abbreviation: U=uncoated, C1=coated (or silanized) with methyltrimethoxysilane, C8=coated (or silanized) with octyltriethoxysilane and C18=coated (or silanized) with octadecyltrimethoxysilane.

Preparation of the Low-Density Polyethylene, i.e. the Polymer (a)

[0393] Ethylene with recycled CTA was compressed in a 5-stage precompressor and a 2-stage hypercompressor with intermediate cooling to reach initial reaction pressure of ca 2781 bar. The total compressor throughput was ca 30 tons/h. In the compressor area approximately 5.3 litres/hour of propionaldehyde (PA, CAS number 123-38-6) was added together with approximately 83 kg propylene/hour as chain transfer agents to maintain an MFR of 0.72 g/10 min. The compressed mixture was heated to 171 C. in the preheating section of the front feed three-zone tubular reactor with an inner diameter of ca 40 mm and a total length of 1200 meters. A mixture of commercially available peroxide radical initiators dissolved in isododecane was injected just after the preheater in an amount sufficient for the exothermal polymerisation reaction to reach peak temperatures of ca 283 C. after which it was cooled to approximately 203 C. The subsequent 2.sup.nd and 3.sup.rd peak reaction temperatures were 275 C. and 265 C. respectively with a cooling in between to 223 C. The reaction mixture was depressurised by a kick valve, cooled and the low-density polyethylene, i.e. the polymer (a), was separated from unreacted gas.

Base Resin Properties of the Low-Density Polyethylene, i.e. the Polymer (a):

[0394] MFR 2.16 kg, at 190 C. [g/10 min] 0.75
Density [kg/m.sup.3] 922.5

Vinyl [C=C/1000 C] 0.27

Vinylidene [C=C/1000 C] 0.16

Trans-vinylene [C=C/1000 C] 0.04

Crystallinity [%] 53.9

[0395] Melting point, Tm [ C.] 110
Preparation of LDPE/MgO Nanocomposites (i.e. Preparation of the Polymer Composition of the Present Invention and Comparative Examples)

[0396] C8-MgO, C18-MgO and UN-MgO nanoparticles were separately mixed with ground the low-density polyethylene (Borealis) powder containing 200 ppm antioxidant Irganox 1076 in n-heptane solution, followed by shaking for 60 min using a Vortex Genie 2 shaker (G560E, Scientific Industries). The heptane facilitated the simultaneous dispersion of the antioxidant and the MgO. The mixture was dried overnight at 80 C. at 20 kPa to achieve complete removal of n-heptane.

Preparation of Film Samples

[0397] The dried LDPE/MgO nanocomposites, i.e. the polymer composition of the present invention and comparative examples, were extruded at 150 C. for 6 minutes at 100 rpm using a Micro 5 cc Twin Screw Compounder (DSM Xplore). Compression moulding was performed at ambient atmosphere in a LabPro 400 (Fontijne Grotnes) at 130 C. for 10 min under contact pressure and for 10 min at 200 kN, this pressure being maintained during cooling to 30 C. A 75-rim stainless steel mould was used to form the film samples, i.e. film samples of the polymer composition of the present invention and of comparative examples. Before compression moulding, the nanocomposites were degassed at 100 C. under reduced pressure (20 kPa) overnight in a Fisher Scientific Vacucell (MMM Group). A further reference sample (i.e. a further comparative example) of unfilled LDPE with the same concentration of antioxidant, for conductivity measurement, was extruded at 150 C. and hotpressed under the same conditions as the nanocomposites. Three series of MgO-nanoparticles with different weight percentages (0.1, 1, 3, 6, 9 wt. %) were added to the LDPE (see Table 2).

TABLE-US-00002 TABLE 2 displays the series of MgO-nanoparticles added to the LDPE OTES - C8 - OdTMS - C18- modified modified (i.e. OTES - (i.e. OdTMS - Unmodified C8-silanized) C18-silanized) MgO- MgO- MgO- Sample nanoparticles nanoparticles nanoparticles Wt. % (MgO) 0.1, 1, 3, 6, 9 0.1, 1, 3, 6, 9 0.1, 1, 3, 6, 9
Preparation of LDPE/ZnO Nanocomposites (i.e. Preparation of the Polymer Composition of the Present Invention and Comparative Examples)

[0398] The low-density polyethylene pellets (Borealis) were cryo-ground to particles sized 0.5 mm. ZnO nanoparticles (different weight fractions; 0.1, 1 and 3 wt. % of the final formulation) and Irganox 1076 (0.02 wt. % of the final formulation) were added to n-heptane. The slurry was ultra-sonicated for 15 min at 23 C., after which cryo-ground LDPE powder was added and the slurry was mixed using a Vortex Genie 2 shaker (G560E, Scientific Industries) at 25 C. for 1 h. The mixtures were dried at 80 C. overnight, after which they were mixed by the shaker for 1 h.

Preparation of LDPE/ZnO Nanocomposites Film Samples

[0399] The dried LDPE/ZnO nanocomposites, i.e. the polymer composition of the present invention and comparative examples, were melt compounded in a Micro 5 cc Twin Screw Compounder (DSM Xplore) at 150 C. for 6 min with a screw speed of 100 rpm. The extruded nanocomposite rods were cut into pellets and compression-moulded under a load of 200 kN into 80 m thick films using a TP400 laboratory press (Fontijne Grotnes B.V., the Netherlands) at 130 C. for 10 min. The samples were finally cooled to 25 C. at a rate of 20 C. min while maintaining the compressive load.

LDPE/MO Nanocomposites

Particle (Nanoparticle) Dispersion Analysis

[0400] A field emission scanning electron microscope (FE-SEM), Hitachi S-4800, was used to assess the nanoparticle dispersion and distribution in the low-density polyethylene. The samples were prepared by cracking a frozen notched sample in liquid nitrogen, which was further coated with Pt/Pd before insertion in the FE-SEM. The coating/sputtering time was 30 s, at an 80 mA operating current of the Cressington 208HR sputter.

[0401] A Field emission scanning electron microscopy (FE-SEM) image analysis of the LDPE/MgO nanocomposites, i.e. the polymer composition of the present invention and comparative examples, were performed on the surface of the freeze-cracked samples. The first step in the analysis was to identify the MgO nanoparticles in the LDPE manually using Adobe Photoshop CS4, due to the low contrast between the filler and matrix, the rough surface of the polymer, and the irregular shape of the MgO nanoparticles. Once the particles had been marked (100-560 MgO-particles/clusters per specimen) the image was converted into black and white and exported to Matlab to assess mean particle radius <r> and the average center-to-center distance in two dimensions (2D) to the Nth nearest neighbour <R.sub.N> using a free radius of the highlighted particle/cluster. The images were also assessed by using a fixed radius (33 nm) of the MgO-nanoparticles, as determined by Pallon et al, J. Mater. Chem. A, 2015, 3, 7523, to compensate that clusters and aggregates are highlighted as one structure. Any single structure larger than corresponding two particles, as pre-determined from virgin particle diameters, was divided into X number of particles forming a cluster, in order to give a true picture of the dispersion. The particle dispersion and distribution was further quantified by using the deviation ratio N, where the assessed average center-to-center values <RN> of the system of fixed radius are compared with the corresponding <R0N> value of a completely random system with the volume percentage, see J. W. Leggoe, Scr. Mater., 2005, 53, pp. 1263-1268. A deviation ratio; N<1, indicated a system with less good distribution (areas of higher concentration of particles) than a completely random system with the same weight percentage of particles. A N>1 indicated a system that is better distributed than can be expected from a complete random system. In order to confirm that the average and mean values acquired were a good estimate Bootstrap statistics with 1000 bootstrap samples were used, see B. Efron, Ann. Stat., 1979, 7, pp 1-26. For more information, see M. Wahlander, F. Nilsson, E. Larsson, W.-C. Tsai, H. Hillborg, A. Carlmark, U. W. Gedde, E. Malmstrom, Polymer, 2014, 55, pp 2125-2138. Due to the small number of particles in the 0.1 wt. % group, these systems could not be analysed from the SEM-images.

[0402] To quantify the distance for particle interaction on the charge current at different weight percentages, the MgO-systems were reconstructed in three dimensions, see FIG. 1. The reconstructions were based on the size distribution of the MgO-phase as determined for all 1-9 wt. % MgO-systems. To reconstruct the 0.1 wt. % MgO-system, the data for the 1 wt. % system were used as this was consider to best represent the 0.1 wt. % system. Using this reconstruction, the conductivity was related to the interaction distance of the MgO-phase at different filling contents in the polyethylene matrix.

Evaluation of Surface Modification Protocols

[0403] The attachment of both C8- and C18-functional silicone oxide (silsesquioxane) coatings to the MgO nanoparticles, i.e. the preparation of OTES-C8-modified (i.e. OTES-C8-silanized) and OdTMS-C18-modified (i.e. OdTMS-C18-silanized) MgO-nanoparticles, was confirmed by infrared spectroscopy (using FT-IR), as shown in FIG. 2.

[0404] The triple peak of CH.sub.2 stretching band at 2924 cm.sup.1 and the two CH.sub.3-stretching bands at 2960 and 2850 cm.sup.1 showed the presence of alkyl substituents in the silsesquioxane coatings, and the broad peak between 1110 and 1010 cm.sup.1 confirmed the formation of SiOSi bonds or alternatively of SiOR with R being an alkyl unit, see P. Larkin, In Infrared and Raman Spectroscopy, edited by P. Larkin, Elsevier, Oxford, 2011. The condensed silanes contributed to a lower broad peak at 3750-3200 cm.sup.1, which was attributed to the coordinated surface-OH groups acting as proton donors in surface hydrogen bonds. The more distinct peaks at 3760 (C18-MgO), 3698 (C8-MgO) and the shoulders at 3751 (C8), 3715 cm.sup.1 (UN-MgO) were related to the stretching of isolated (1-coordination) and multiple coordinated surface-OH groups, where the possible coordination has been reported to depend on the exposed crystal facet, see E. Knizinger, K. H. Jacob, S. Singh, P. Hofmann, Surf. Sci., 1993, 290, pp 380-402. The shifts in the peaks for the silanized MgO were indicating that certain facets are more favourable for silane condensation. The C8-MgO-spectrum, i.e. the spectrum of C8-modified (i.e. OTES-C8-silanized) MgO-nanoparticles, was more similar to the unmodified MgO (UN-MgO) than the C18-MgO-spectrum, i.e. the spectrum of C18-modified (i.e. OdTMS-C18-silanized) MgO-nanoparticles, with a residual broad peak at about 1467 cm.sup.1 and a more distinct peak at about 1630 cm.sup.1. On the slope up to the MgO-bulk absorption (800 cm.sup.1), see D. Cornu, H. Guesmi, J. M. Krafft, H. Lauron-Pernot, J. Phys. Chem. C, 2012, 116, pp 6645-6654, the small peaks at 835 (C18-MgO) and 850 cm.sup.1 (C8-MgO) have been attributed to SiO stretching in the SiOH, see P. Larkin, In Infrared and Raman Spectroscopy, edited by P. Larkin, Elsevier, Oxford, 2011, while the peak at 862 cm.sup.1 for unmodified MgO (UN-MgO) was assigned to OH from the adsorbed H.sub.2O, see H. A. Prescott, Z. J. Li, E. Kemnitz, J. Deutsch and H. Lieske, J. Mater. Chem., 2005, 15, 4616-4628. The sharp silane peak at 1467 cm.sup.1, visible in both the C18-MgO and C8-MgO spectra was attributed to CH bending, and is known to be sharper for longer alkyl chains.

[0405] The broad peak centered at 1467 cm.sup.1 (1600-1300 cm.sup.1) was attributed to chemisorbed CO.sub.2 in the form of unidentate carbonate (OCO) with one or two bridging bonds, covering a band of wavelengths (1710-1270 cm.sup.1), see D. Cornu, H. Guesmi, J. M. Krafft, H. Lauron-Pernot, J. Phys. Chem. C, 2012, 116, pp 6645-6654, and H. A. Prescott, Z. J. Li, E. Kemnitz, J. Deutsch and H. Lieske, J. Mater. Chem., 2005, 15, 4616-4628. This sort of CO.sub.2 chemisorbed on MgO was previously also reported with a similar relative intensity for adsorbed carbon dioxide from polluted air, see Y. Y. Li, K. K. Han, W. G. Lin, M. M. Wan, Y. Wang, J. H. Zhu, J. Mater. Chem. A, 2013, 1, pp. 12919-12925. The suppressed CO.sub.2 and H.sub.2O adsorption signals (1460 and 1635 cm.sup.1, respectively) from the C18-modified MgO-nanoparticles indicated a better steric hindrance towards the adsorption of these species than the C8-modified MgO-nanoparticles, and the former also displayed a more intense signal at 2950-2850 cm.sup.1 see again both D. Cornu, H. Guesmi, J. M. Krafft, H. Lauron-Pernot, J. Phys. Chem. C, 2012, 116, pp 6645-6654, and H. A. Prescott, Z. J. Li, E. Kemnitz, J. Deutsch and H. Lieske, J. Mater. Chem., 2005, 15, 4616-4628. This was in contrast to spectrum of the unmodified MgO nanoparticles that showed the presence of surface water (3600-3200 and 1635 cm.sup.1). This surface water was expected to facilitate the hydrolysis of the silanes on the particle surfaces during the surface modification reactions.

[0406] FIG. 3 shows the thermogravimetrical (TG) data of unmodified, C8- and C18-modified MgO nanoparticles heated under nitrogen to a temperature of 600 C.

[0407] Up to 230 C. the mass losses were almost identical and could be attributed to the evaporation of crystal water and carbonates species formed from adsorbed CO.sub.2, which has been reported to desorb at 200 C., see V. K. Diez, C. R. Apesteguia, J. I. Di Cosimo, J. Catal, 2006, 240, 235-244. The total mass loss of the unmodified MgO was 2.2% up to 600 C., which was attributed to surface adsorbed water and possibly residual Mg(OH).sub.2 that remained after the calcination of the Mg(OH).sub.2. The C8-modified MgO-nanoparticles and C18-modified MgO-nanoparticles showed a mass loss of 8.2% and 14.0%, respectively. After normalization with respect to the mass loss of the unmodified MgO, the mass losses of 6.2% for C8-modified MgO-nanoparticles and 12.1% C18-modified MgO-nanoparticles could be related to the loss of the organic material, since the relative ratio corresponded to the ratio of the masses of the alkyl chains on the two silanes. The condensation of the silane molecules to form a condensed SiOSi network was therefore confirmed to have occurred in a similar manner for both silanes.

[0408] The silicone oxide network density (.sub.silane [silane per nm.sup.2]) was calculated as in equation (1):

[00001]$\begin{array}{cc}{}_{\mathrm{silane}}=\frac{\left(w_{\mathrm{unmod}}-w_{\mathrm{mod}}\right).Math.N_A}{M_{\mathrm{volatile}}.Math..Math.w_{\mathrm{unmod}}.Math.\mathrm{SSA}}& \left(1\right)\end{array}$

where W.sub.unmodw.sub.mod where W.sub.unmod and w.sub.mod are respectively the mass losses of the unmodified (UN-MgO) and modified MgO (C8-MgO and C18-MgO), N.sub.A is the Avogrado number, M.sub.w (volatile part) is the molar mass of the volatile part of the silane and SSA is the specific surface area. The calculation resulted in a condensed silane molecule coverage of 1.97 silanes/nm.sup.2 for the C8-coating and 1.72 silanes/nm.sup.2 for the C18-coating on the MgO nanoparticles, assuming that the silane molecules grafted as a monolayer with full access to the 167 m.sup.2/g surface area of the pristine MgO nanoparticles. The values were of the same order as those reported for Fe.sub.3O.sub.4 and Al.sub.2O.sub.3 nanoparticles, see D. Liu, A. M. Pourrahimi, L. K. H. Pallon, R. L. Andersson, M. S. Hedenqvist, U. W. Gedde and R. T. Olsson, RSC Adv., 2015, 5, 48094-48103, and also in agreement with the ca. 1.7-2.0 silanes/nm.sup.2 surface coverage reported by McCarthy et al. for a 70% coverage of the available surface, assuming one silane molecule covers 0.4 nm.sup.2, see S. A. McCarthy, G. L. Davies and Y. K. Gun'ko, Nature Protocols, 2012, 7, 1677-1693.

[0409] X-ray photospectroscopy (XPS) of the three MgO nanoparticle samples, i.e. samples of unmodified, C8-modified and C18-modified MgO-nanoparticles, confirmed the attachment of silanes by displaying a Si 2p peak and an increase in the C is intensity (FIG. 4) for the C8-MgO (i.e. C8-modified MgO-nanoparticles) and C18-MgO (i.e. C18-modified MgO-nanoparticles), compared to the unmodified MgO. The atomic percentage of carbon, identified as C is at 285.0 eV, was 16.98 at. % for C8-MgO and 33.51 at. % for the C18-MgO. The C is carbon ratio (0.507) correlated well with the carbon related TG mass losses for the different silanes, considering surface coverages associated with the different silanes and their respective molecular masses. As seen in FIG. 4, a small amount of carbon species (2 at. %) was also found for the unmodified MgO at 289.9 eV, while only traces of carbonates were seen for the C8-modified MgO and no carbonates at all for the C18-modified MgO, see W. K. Istone, Surface Analysis of Paper, ed. T. E. Conners, S. Banerjee, pp 247, 1995, CRC, New York. The steric protection provided by the alkyl chains against the adsorption of carbon dioxide (CO.sub.2), suggested by the disappearance of the broad FT-IR peak around 1460 cm.sup.1 for the C18-MgO in FIG. 2 was thus supported by the XPS results. When CO.sub.2 is adsorbed onto the surface, different carbon-oxygen species are formed depending on the coordination to the surface, see D. Comu, H. Guesmi, J. M. Krafft, H. Lauron-Pemot, J. Phys. Chem. C, 2012, 116, pp 6645-6654. The XPS data also revealed that all the samples showed a relatively large amount of surface-located oxygen (O 1s-531.4 eV and 532.6 eV) compared to the lattice-embedded oxygen at 529.4 eV (FIGS. 5a, 5b and 5c). In the case of the unmodified particles, these energy bands arise from oxygen that exists in the form of terminating MgO, Mg(OH).sub.2, MgOH, crystal-H.sub.2O and carbonated species adsorbed on the surface, see W. K. Istone, Surface Analysis of Paper, ed. T. E. Conners, S. Banerjee, pp 247, 1995, CRC, New York, J. F. Moulder, Handbook of X-ray Photoelectron Spectroscopy, ed. Chastain, pp. 45, and D. Cornu, H. Guesmi, J. M. Krafft, H. Lauron-Pernot, J. Phys. Chem. C, 2012, 116, pp 6645-6654. For the C8-MgO and C18-MgO samples a reduction in relative intensity of the shifted O 1s at 531.4 eV (surface oxygen) compared to the O 1s peak at 529.4 eV (lattice oxygen) was observed after the silanization, probably due to a reduction in the presence of Mg(OH).sub.2 and MgOH, which is in agreement with the condensation of silanol groups (SiOH) that occurs during the condensation of the silsesquioxane coatings onto the surface of the MgO nanoparticles. An additional observation was that the surface modification by silanization resulted in the removal of 0.5 atomic percentage of chloride ions (Cl.sup.), which were detected on the unmodified MgO. These Cl.sup. ions are known to remain adsorbed on the surface of the nanoparticles and to originate from the salt used to precipitate the nanoparticles, see A. M. Pourrahimi, D. Liu, L. K. H. Pallon, R. L. Andersson, A. Martinez Abad, J.-M. Lagarn, M. S. Hedenqvist, V. Strm, U. W. Gedde and R. T. Olsson, RSC Adv., 2014, 4, 35568-35577.

Conductivity Measurement Method

[0410] The volume electrical conductivity measurements were performed following the Conductivity Measurement Method, i.e. standard procedure according to IEC, in Methods of Test for Volume Resistivity and Surface Resistivity of Solid Electrical Insulating Materials, Standard 60093, 1980, applying a direct current (DC) voltage (Glassman FJ60R2) over the film sample, i.e. the polymer composition of the present invention and comparative examples, and measuring the charging current with an electrometer (Keithley 6517A), see FIG. 6. The current signal was recorded by LabVIEW software incorporated in a personal computer and stored for further analysis. An oven was used to control temperature, whereas an overvoltage protection secured the electrometer from damaging due to possible overshoots and a low-pass filter removed high frequency disturbances. A stainless steel three-electrode system was used, in which the high voltage electrode was a cylinder with a diameter of 45 mm, the current measuring electrode was 30 mm in diameter, and the guard ring eliminated surface currents. Good contact between the high-voltage electrode and the film sample was achieved by placing an Elastosil R570/70 (Wacker) layer between them (see L. K. H. Pallon, R. T. Olsson, D. Liu, A. M. Pourrahimi, M. S. Hedenqvist, A. T. Hoang, S. Gubanski and U. W. Gedde, J. Mater. Chem. A, 2015, 3, 7523-7534). The experiments were conducted on LDPE/MgO nanocomposites and unfilled LDPE as reference sample at 60 C. for 410.sup.4 s (11.1 h). The applied voltage was 2.6 kV corresponding to an electric field of 32.5 kV/mm, giving conditions (40-90 C.) in temperature and electric field resembling the stress conditions in the insulation of a real HVDC cable, see C. C. Reddy and T. S. Ramu, IEEE Trans. Dielectr. Electr. Insul., 2006, 13, 1236-1244. The test was repeated twice for each material to assess the reproducibility.

Effect of Particle Loading on DC-Conductivity

[0411] FIG. 7a to 7c show the charging current as a function of time at 32.5 kV/mm and at 60 C. for the unmodified MgO nanoparticles, see FIG. 7a, and the surface modified MgO nanoparticles (C8-coated, see FIG. 7b, and C18-coated, see FIG. 7c) extruded into polyethylene at particle contents from 0.1 wt. % to 9 wt. %. All measurements were carried out on ca. 80 m thick films, i.e. samples of the polymer composition of the present invention and of the comparative examples, which samples were prepared by hot pressing of the extruded polymer composition. Good repeatability was shown.

[0412] From the obtained data (see FIG. 7a to 7c), it was apparent that a distinct drop in charging current occurred during the first 100 seconds for all nanocomposites with nanoparticle (i.e. MgO nanoparticle) content more or equal to 1 wt. % (see phase 1 in FIGS. 7a, 7b and 7c). The distinct initial drop in charge current was not present for the 0.1 wt. % sample with unmodified nanoparticles, while the 0.1 wt. % modified particles showed an initial drop, but not as pronounced as at 1 wt. % nanocomposites. After 100 s the charging current reached a transition and levelled out with a slower decay, much similar to that of the unfilled LDPE. The initial drop in charge current was not present for the 0.1 wt. % sample with UN-MgO, while the samples with 0.1 wt. % surface-modified particles (C8-MgO and C18-MgO) showed an initial drop, but not as pronounced as with the 1 wt. % nanocomposites. The initial drop has been attributed to a polarization effect, but could also be explained by a formation of charges in the vicinity of the electrodes due to the nanoparticles which leads to a higher charge injection barrier, and thus reduced charging current, see V. Adamec, J. H. Calderwood, J. Phys. D. Appl. Phys., 1981, 14, 1487, and R. C. Smith, C. Liang, M. Landry, J. K. Nelson and L. S. Schadler, IEEE Trans. Dielectr. Electr. Insul., 2008, 15, 187-196). The lowest charging current was found for the 1-3 wt. % nanocomposites, with no essentially difference between different MgO-particles. All 1-3 wt. % nanocomposites displayed ca. 2 orders of magnitude lower charging current than the unfilled LDPE after 10 min, and 10-20 times lower charging current after 11 h compared to the unfilled LDPE (FIG. 7a-c).

[0413] The surface modification of the MgO nanoparticles had a clear effect on the ability of the 6-9 wt. % nanocomposites to maintain the lowest levels of charging current over the whole measurement time. The charging current for 6-9 wt. % C8-MgO and C18-MgO was reduced 30-50 times after the initial drop (610.sup.2 s), and was reduced more than 20 times over the entire measurement that lasted for 11 hours (410.sup.4 s). In comparison, the unmodified MgO nanoparticles showed at the same MgO contents significantly higher charging current, with the 9 wt. % samples performing equivalent with an unfilled LDPE after 11 hours (410.sup.4 s), see FIG. 7a.

[0414] The dashed red lines display the corresponding volume conductivity of the unfilled LDPE. A limited drop in volume conductivity can be seen for 0.1 wt. % for the C8-MgO and C18-MgO, which was most pronounced after 10 min. The minimum conductivity was acquired for nanocomposites with 1-3 wt. %, reaching conductivity values of 2-6*10.sup.15 S/m after 10 min.

[0415] This was followed by an almost constant or increased conductivity for the 6-9 wt. % MgO nanoparticle content after both 10 min and 11 h. The absence of surface modification clearly limited the nanoparticle fraction to a maximum of ca. 3 wt. %, with marked increase in volume conductivity at higher weight fractions. At the same time, a more reliable reduction in conductivity was present for the silsesquioxane modified C8-MgO and C18-MgO nanoparticle systems over the whole measurement time, which consistently showed ca. 1-2 orders of magnitude lower conductivity, independently of nanoparticle content. The difference in impact of the surface modification was most apparent for the highest (9 wt. %) nanoparticle content (see FIGS. 8a and 8b). FIGS. 8a and 8b show the measured conductivity values for the different LDPE/MgO nanocomposites after 10 min (FIG. 8a) and 11 hours (FIG. 8b) with an applied electrical field of 32 kV/mm.

Inter-Particle Distance and Correlation of 3-Dimensional Modelled Phase Distribution with DC Properties of LDPE/MgO Nanocomposites

[0416] The dispersion of the nanoparticles was analysed to identify relations between nanoparticle dispersion state and conductivity values in the LDPE/MgO nanocomposites. FIGS. 9a and 9b show the SEM images of the cryo-fractured 6 wt. % LDPE/MgO nanocomposites, with unmodified MgO nanoparticles (FIG. 9a) as compared to the C8-modified MgO nanoparticles (FIG. 9b).

[0417] The unmodified nanoparticles were mostly present as nanoparticles inside up to ca. 1 m large aggregates, with occasionally intercalated polymer, visible in the aggregates (shown in the upper right corner of the micrograph, i.e. FIG. 9a). A 1 m aggregate contained ca. 8000 nanoparticles. In contrast, the surface-modified nanoparticles were evenly distributed over the entire cross-sectional fracture surface area, with only a few aggregates as large as ca. 200 nm, see the micrograph, i.e. FIG. 9b. The even distribution of the nanoparticles confirmed that the silanization reactions effectively had resulted in surface modification of solitary nanoparticles. Overall, both the C8- and the C18-coatings resulted in very good dispersion of the nanoparticles, in contrast to the unmodified nanoparticles.

[0418] For a more in-depth understanding of the dispersion of the nanoparticles, a dispersion analysis was carried out using Bootstrap statistics. The image was based on multiple micrographs, i.e. microscopy images, (>25 for different concentrations) including approx. 250 MgO discrete phase structures per sample (aggregated or non-aggregated nanoparticles). The results are presented herein as 2-dimensional (2D) analyses of the cryo-fractured surfaces with MgO nanoparticle inter-MgO phase distances (FIGS. 9a and 9b, and FIG. 10), and as 3-dimensional modelling of the discrete MgO phases, as determined from the size distribution of the MgO phases in the micrographs, see FIGS. 1a-1f, FIG. 12 and FIG. 13.

[0419] MgO phase distribution in the fractured surfaces 2-dimensions (2D) FIGS. 11a and 11b show the overall dispersion of the MgO phase in the PE matrix for the nanocomposites with different weight fractions of nanoparticles, i.e. regarding the aggregated MgO nanoparticles as discrete phases (in accordance with FIG. 10). FIG. 11a shows that the centre-to-centre MgO phase distance decreased from 900 to 350 nm when the amount of unmodified nanoparticles increased from 1 to 9 wt. %. The modified nanoparticles show a decrease from ca. 400 to 100 nm over the same range for the 1st discrete MgO phase neighbour. From FIG. 11b it is apparent that this pattern also was representative over longer ranges, considering the 51st neighbour. In FIG. 11c the fraction of particles tied up in aggregates are presented. An aggregate is here defined, in its smallest definition, as a discrete MgO-phase structure with a cross-section that was two times larger than the diameters of two solitary particles (>132 nm). Thus, this is the smallest threshold for classification of an aggregate herein, compare with the first aggregate size and the second aggregate size also as defined herein. Overall, the distance between MgO-structures was always greater for the unmodified MgO nanoparticle phase, which was present as smaller clusters and aggregates from 200 nm up to 10 m. On the contrary, very rarely aggregates could be observed in the surface-modified systems where the particles were mostly present as solitary particles. The improved dispersion of the surface-modified nanoparticles was consistent with that the aggregated phase constituted only 20% of the entire MgO phase in FIG. 11c. On increasing the defined threshold for classification of an aggregate from 2 adjacent particles (i.e. >132 nm) to 3 particles resulted in completely aggregation free samples. In summary, the general trend was that unmodified nanoparticles were severely aggregated with an aggregated MgO fraction of ca. 90%, whereas both the systems of the C8- and C18-modified MgO always showed an aggregated content of ca. 15%. The aggregates were also substantially smaller when the nanoparticles had been coated with the C8- or C18-functional silsesquioxane coatings.

Modelling of Nanocomposites Structure in Relation to Measured Conductive Properties (3-Dimensional)

[0420] The 3-dimensional MgO phase distribution was modelled to determine an approximate value for the necessary radius of interaction of the MgO-phase to show an impact on the conductivity of the nanocomposites. FIG. 1 displays the 0.1 wt. % (a and d), 1 wt. % (b and e) and 9 wt. % (c and f) composite interiors based on the size distribution on the nanoparticles in the image analysis. Only the C8-modified MgO nanoparticle system was modelled due to the large similarity with the C18-modified nanoparticle system in dispersion. The smallest spheres represent individual MgO nanoparticles, whereas larger spheres represent the aggregated MgO nanoparticles. The semi-transparent large light gray spheres show the maximum radius of interaction of the MgO phase; see FIGS. 1a, b and d. The nanocomposite comprising 0.1 wt. % C8-modified MgO nanoparticles was used as a reference point since it represented the nanocomposite with the lowest nanoparticle content showing a distinct difference in the measured conductivity, compared to unfilled LDPE. The interaction radius was determined to 775 nm for the surface-modified MgO phase when the sum of all the spheres had reached an interaction volume equal to 95% of the total volume of the entire 0.1 wt. % nanocomposite (large light gray spheres in FIG. 1d). At this point, the large light gray spheres overlapped extensively and filled out the material with margin. The value for the total overlapping interaction volumes (see darker grey region between the large light gray spheres) in the entire 3-dimensional illustration reached 204% when all the spheres had been inserted in the model. For the same nominal radius of interaction (775 nm), it can be observed from FIG. 1a that the spheres no longer were overlapping, which was consistent with the aggregation of the unmodified 0.1 wt. % MgO nanoparticles. The total interaction volume of the spheres reached 40% of the entire sample. The modelled results were consistent with the micrograph observations that considerable portions of the material present as empty LDPE matrix, which in turn was synonymous with an electrical conductivity that corresponded to the pristine LDPE material (FIG. 8a). In contrast, a nanoparticle content above 1 wt. % always resulted in total interaction volumes reaching 100% (FIGS. 1b, 1c, 1e and 1f) for a radius of interaction equal to 775 nm, and overlapping spheres equal to 392% and >1000% for the 1 wt. % unmodified and C8-modified MgO nanoparticles, respectively.

[0421] FIG. 12 shows how the interaction radius of the spheres changes if the composite samples were to be filled with sufficient MgO phase to always cover the 95 vol. % interaction volume of the entire samples. The interaction radius was here plotted against the conductivity after 10 min, showing that ca. 800 nm was the minimal necessary radius of interaction of the MgO-phase to show a marked effect on the measured conductivity values, i.e. circa one order of magnitude decrease in conductivity. This distance was clearly also surpassed for the 1 wt. % unmodified MgO nanoparticles, which showed an interaction radius of 636 nm with a conductivity of 410.sup.15 S/m, even if the nanoparticles were aggregated (see FIG. 1b). The more severe aggregation within the unmodified nanoparticles system started showing an affect already at 3 wt. %, with progressively increasing values of conductivity, i.e. with a decreasing insulation capacity, for the unmodified nanoparticles up to 9 wt. % (FIG. 1c). From the diamond and square markers (FIG. 12), it is apparent that the surface-modified nanoparticles could retain a preserved insulation capacity below 10.sup.14 S/m for nanoparticle contents up to 9 wt. %. The lowest conductivity values were observed for nanocomposites with the 3 wt. % C8-modified MgO nanoparticles and the 1 wt. % C18-modified MgO nanoparticles, respectively, reaching 210.sup.15 S/m at interaction radius from 200-350 nm.

[0422] On one hand, the data show that surface modifications effectively allow the MgO nanoparticle phase to act as an insulation promoter at high nanoparticle content, and that the nanoparticles phase functioned even down to interaction radius values as small as ca. 100 nm. At the same time it is apparent that the unmodified nanoparticles for equivalent interaction radius can be extrapolated to show no effect, i.e. if sufficient amounts of aggregated MgO nanoparticles would have been added to reach the 100 nm interaction volume radius. These observations allow us to conclude that one major effect of the nanoparticles presence is related to the created particle interface to polymer, which may act as surface for distribution and collection of charges within the multiphase nanocomposites. It was also apparent that the measured conductivity values are lower than what has been reported for pure MgO crystals: 110.sup.13 to 110.sup.12 S/m, see F. Freund, M. M. Freund and F. Batllo, J. geophys. Res., 1993, 98, 22209-22229. At the same time, the difference in the interaction radius between 1 wt. % unmodified MgO and 0.1 wt. % C8-modified MgO or, alternatively, and 0.1 wt. % C18-modified MgO was only 50-100 nm, with an unproportionate conductivity difference of one order of magnitude (FIG. 12). This observation indicates that not only the distribution of nanoparticles was essential for the conductivity reduction. The reduction in conductivity may also be related to the intrinsic characteristics of the nanoparticles, where a greater mass of MgO will provide a larger presence of inorganic lattice defects and uneven surfaces (edges, corners and vacancies), which may give rise to surface states with charge trapping capacity, see T. Knig, G. H. Simon, H. P. Rust, G. Pacchioni, M. Heyde and H. J. Freund, J. Am. Chem. Soc., 2009, 131, 17544-17545. At the same time, the increased conductivity for the nanocomposites with unmodified MgO (6 and 9 wt. %) was clearly related to the presence of aggregates, where mainly the adsorbed H.sub.2O (and CO.sub.2) is suggested to have provided a conduction path through the aggregates, with locally higher conductivity. It was also apparent that the adhesion between the aggregates of unmodified MgO nanoparticles and polyethylene, see FIG. 13, was inadequate, with voids formed in the aggregate interface. Voids were considered to be undesired and have previously been shown to have negative effect on electrical insulating properties, see L. Testa, S. Serra and G. C. Montanari, J. Appl. Phys., 2012, 108, 034110.

[0423] It is demonstrated that the conductivity of low-density polyethylene (LDPE) can be reduced by circa 2 orders of magnitude to 210.sup.15 S/m by inclusion of magnesium oxide (MgO) nanoparticles as space charge collecting nanofiller material at 60 C. (32 kV/mm). These values are not only smaller than that of the traditional pristine LDPE polymer (210.sup.13 S/m), but also smaller than that reported for pure MgO crystals: 110.sup.13 to 110.sup.12 S/m, see F. Freund, M. M. Freund and F. Batllo, J. Geophys. Res., 1993, 98, 22209-22229. The lowest observed conductivity was ca. 710.sup.16 S/m for 3 wt. % surface coated nanoparticles. A cornerstone in the development of the presented materials was to apply a thin and selectively condensed silsequioxane coating to the nanoparticles, which allowed for significantly improved nanoparticle dispersions resulting in repeatedly measured high insulation capacity of the nanocomposites. In absence of the silsesquioxane coatings, the nanoparticles aggregated severely and at fractions above 1-3 wt. % resulted in conductivity values approaching that of the pristine LDPE. It is suggested that the conductivity within these aggregated volumes show a different, and higher conductivity due to the presence of adsorbed humidity, e.g. the conductivity of high resistivity MilliQ water (H.sub.2O) is approx. 10.sup.7-10.sup.5 S/m after exposure to ambient conditions for ca 30 min. However, not only water but also CO.sub.2 existed within the aggregates as clearly demonstrated from the C1s peak visible from the XPS data (290 eV). The octadecyl functional (C18) silsesquioxane coatings completely prevented the CO.sub.2 and H.sub.2O, and consequently showed the lowest conductivities repeatedly measured. It is therefore suggested that the coating techniques for inorganic nanofillers used within high voltage insulation must rely on the preparation of coatings sufficiently saturated with alkyl (CH.sub.2) moieties to prevent adsorption of hygroscopic substances, which normally are associated with inorganic nanoparticles. Modelling allowed to conclude that the radius of interactions related to the functional MgO phase was required to show values below ca. 800 nm to effectively reduce the conductivity of the nanocomposites as compared to pristine LDPE. The most effective compositions showed a radius of interaction of ca. 200 nm. The interaction radius was defined as the distance of an arbitrary selected neighbouring MgO phase. The presented successful nanoparticle dispersion has also been confirmed by thermal characterization which revealed that the most evenly dispersed surface coated nanoparticles also functioned to delay the onset of thermal degradation of the nanocomposites with ca. 100 C.

LDPE/ZnO Nanocomposites

Characterization of Nanoparticles

[0424] Transmission infrared spectra, based on 32 scans per spectrum and with a resolution of 4 cm.sup.1, were taken on IR pellets using a Perkin-Elmer Spectrum IR Spectrometer 2000. The pellets (diameter=13 mm; thickness=0.85 mm) consisted of 3 mg nanoparticles and 300 mg KBr. Transmission electron micrographs were obtained using a Hitachi HT7700 microscope operated at 100 kV in the high contrast mode. An ultrasonicated suspension of the nanoparticles in ethanol was deposited onto a carbon-coated 400 mesh copper grid (Ted Pella, Inc., USA) and dried at 50 C. under reduced pressure (0.5 kPa). A Mettler-Toledo TG/DSC1 was used to determine the mass loss of the pristine and coated nanoparticles. All the samples were dried in an oven at reduced pressure (0.5 kPa) at 50 C. overnight before the thermogravimetry. The samples (mass=4.00.5 mg), placed in 70 L aluminium oxide crucibles and heated from 30 to 800 C. at a rate of 10 C. min.sup.1 while being purged with dry nitrogen (flow rate=50 mL min.sup.1).

Particle (Nanoparticle) Dispersion Analysis

[0425] A field emission scanning electron microscope (FE-SEM), Hitachi S-4800, was used to assess the nanoparticle dispersion and distribution in the low-density polyethylene. The samples were prepared by cracking a frozen notched sample in liquid nitrogen, which was further coated with Pt/Pd before insertion in the FE-SEM. The coating/sputtering time was 30 s, at an 80 mA operating current of the Cressington 208HR sputter.

[0426] A Field emission scanning electron microscopy (FE-SEM) image analysis of the LDPE/ZnO nanocomposites, i.e. the polymer composition of the present invention and comparative examples, were performed on the surface of the freeze-cracked samples. The LDPE/ZnO nanocomposites may further be analysed in line with the corresponding analysis as described for the LDPE/MgO nanocomposites under LDPE/MgO nanocomposites, Particle (nanoparticle) dispersion analysis above.

Electrical Conductivity Measurements

[0427] The electrical conductivity measurements were performed by applying a 2.6 kV DC voltage from a power supply (Glassman FJ60R2) across the 80 m thick film sample and measuring the current using a Keithley 6517A electrometer. The electric field across the film was 32.5 kV mm.sup.1. The detected current signal was recorded by LabVIEW software incorporated in a personal computer and stored for further analyses. An oven was used to control temperature, whereas an overvoltage protection secured the electrometer from damaging due to possible overshoots and a low-pass filter removed high frequency disturbance. A three-stainless steel electrode system was used, in which the high voltage electrode was a cylinder with a diameter of 45 mm; the current measuring electrode was 30 mm in diameter, whereas the guard ring allowed for eliminating surface currents. A good contact of the high voltage electrode and the film sample was obtained by placing an Elastosil R570/70 (Wacker) layer between them. The experiments were conducted at 60 C. for 410.sup.4 s (11.1 h).

Characterisation of Functional Silane-Coated Particles

[0428] Three alkyl-containing silanesmethyltrimethoxysilane (C1 in FIG. 14), octyltriethoxysilane (C8 in FIG. 14) and octadecyltrimethoxysilane (C18 in FIG. 14)were used to tailor the surface properties of ZnO nanoparticles (ZnO-25-U). FIG. 14 shows the infrared spectra of the pristine and silane-coated nanoparticles. The absorbance band at 450-600 cm.sup.1 is assigned to the ZnO stretching and the absorbance band at 880 cm.sup.1 originates from the stretching vibration of ZnOH, see A. M. Pourrahimi, D. Liu, V. Strm, M. S. Hedenqvist, R. T. Olsson and U. W. Gedde, J. Mater. Chem. A, 2015, 3, 17190-17200. The CH stretching bands at 2800-3000 cm.sup.1 were present in the spectra of the silane-coated nanoparticles, but the spectrum of the pristine nanoparticles showed no such bands. The spectrum of the C1-coated nanoparticles showed absorption at 1270 cm.sup.1, which is assigned to the SiCH.sub.3 stretching vibration, see R. T. Olsson, M. S. Hedenqvist, V. Strm, J. Deng, S. J. Savage and U. W. Gedde, Polym. Eng. Sci., 2011, 51, 862-874. The spectrum of the C8 and C18-coated nanoparticles also showed the peak at 1470 cm.sup.1 assigned to the CH.sub.2 unit not covalently bonded to silicon, due to terminal alkyl group of the silane, see A. Grill, Annu. Rev. Mater. Res., 2009, 39, 49-69. An absorption band at 1120 cm.sup.1 assigned to the SiOSi stretching vibration which indicated the formation of a cross-linked silicon oxide structure on the nanoparticle surfaces, see D. Liu, A. M. Pourrahimi, R. T. Olsson, M. S. Hedenqvist and U. W. Gedde, Eur. Polym. J., 2015, 66, 67-77. This peak was not visible in the spectra of the all silane coated nanoparticles (FIG. 14). Hence, most of the silanol groups of the hydrolysed silanes condensed with the hydroxyl groups on the nanoparticle surfaces, rather than reacting with other silanol groups to form a cross-linked coating layer around the particles. The absence of the SiOSi stretching band in the spectra of coated nanoparticles suggested that these coatings were monolayers, see D. Liu, A. M. Pourrahimi, R. T. Olsson, M. S. Hedenqvist and U. W. Gedde, Eur. Polym. J., 2015, 66, 67-77. The silane coatings of ZnO nanoparticles with the thickness ca. 2 nm were confirmed by high resolution TEM presented elsewhere, see D. Liu, A. M. Pourrahimi, L. K. H. Pallon, R. L. Andersson, M. S. Hedenqvist, U. W. Gedde and R. T. Olsson, RSC Adv., 2015, 5, 48094-48103.

[0429] FIG. 15 shows the normalized mass plotted as a function of temperature for pristine and silane-coated ZnO nanoparticles (ZnO-25-U) after normalization to the mass loss value at 140 C. Since all nanoparticles regardless of coating had showed a mass loss at 20-140 C. due to removal of loosely bound water, the curve values was normalized to the mass loss value at 140 C. The mass loss at 140-800 C. was attributed to removal of hydroxyl groups in the case of pristine nanoparticles. Since, the hydroxyl groups were dominantly condensed with the silanes during coating of the nanoparticles; the mass loss at 140-800 C. for coated nanoparticles was due to silane transformation into a silica layer. The silane coverage on the nanoparticle surfaces was calculated by normalizing the amount of silane molecules with respect to the surface area of the ZnO-25-U nanoparticles, 34 m.sup.2 g.sup.1, (Table 3). The C1- and C8-coated nanoparticles respectively had the highest and lowest coverage of silane on the nanoparticle surface. These values were in accordance with the silane coverage of aluminium oxide with the specific surface area close to ZnO nanoparticle, see D. Liu, A. M. Pourrahimi, R. T. Olsson, M. S. Hedenqvist and U. W. Gedde, Eur. Polym. J., 2015, 66, 67-77.

TABLE-US-00003 TABLE 3 Coverage of silanes on ZnO nanoparticles (ZnO-25-U) Mass loss at Silane coverage Silane coverage Sample 800 C. (%) .sup.a (mol m.sup.2) (molecules (nm).sup.2) Pristine 1.87 C1-coated 2.30 5.2 .sup.b (11.1 .sup.c) 8.6 .sup.b (18.4 .sup.c) C8-coated 2.79 1.5 .sup.b (1.6 .sup.c) 2.5 .sup.b (2.6 .sup.c) C18-coated 6.64 3.4 .sup.b (3.5 .sup.c) 5.6 .sup.b (5.8 .sup.c) .sup.a The mass loss normalized to the value at 140 C. in order to remove the effect of loosely bound water. .sup.b The calculation was based on the mass loss between uncoated and silane coated nanoparticles were due to only the degradation of hydrocarbon moieties (R group in RSiO.sub.1.5 coatings) .sup.c The calculation was based on assuming that the inorganic part of the silane SiO.sub.1.5 was oxidized to silica (SiO.sub.2)

Characterisation of the LDPE/ZnO Nanocomposites

[0430] FIGS. 16a-d show scanning electron micrographs of ZnO particles with different size and their LDPE nanocomposites based on the 3 wt. % particles. The zinc acetate precursor yielded separate and mono-domain nanoparticles (ZnO-25-U), neither of which showed any intra-particle porosity (FIG. 16a). The zinc nitrate yielded submicron (ZnO-550-U) star-shaped particles with a symmetrical habit (FIG. 16c), see A. M. Pourrahimi, et al., RSC Adv., 2014, 4, 35568-35577 and A. M. Pourrahimi, et al., J. Mater. Chem. A, 2015, 3, 17190-17200. These star-shaped particles consisted of c-axis oriented primary nanoparticles along each petal (spike) director. ZnO-550-U contained pores placed among the primary nanoparticles with an average size of 3.5 nm. The nanocomposite based on the C8-coated ZnO nanoparticles (ZnO-25-C8) showed very uniform particle dispersion in the crystalline LDPE lamellae; the agglomerates were always smaller than 100 nm (FIG. 16b). The cryo-fractured of all nanocomposites based on coated nanoparticles (C1, C8 and C18) showed no sign of big aggregation; whereas the nanocomposites which contained uncoated ZnO (ZnO-25-U) nanoparticles showed big aggregates sized 1-50 m (FIGS. 16a and 17). The applied shear force during extrusion (150 C.) was not sufficient to break uncoated ZnO nanoparticles hard aggregates into solitary particles, but the compatibility between LDPE and hydrophobic surface of coated ZnO nanoparticles resulted in no major aggregation. The submicron and micron sized star-shaped ZnO particles showed good dispersion without any aggregation, while no coating applied on their hydrophilic surfaces'. Assuming a perfect dispersion, for a face-centred cubic lattice arrangement of particles in the polymer, the theoretical IPD (centre to centre distance) is given by (equations (2) and (3)), see A. M. Pourrahimi, et al., J. Polym. Sci., Part B: Polym. Phys., 2011, 49, 327-332:

[00002]$\begin{array}{cc}\mathrm{IPD}={\left(\frac{\sqrt{2}.Math.}{6.Math.V_f}\right)}^{1.Math.\text{/}.Math.3}d& \left(2\right)\\ V_f=\frac{w_f.Math.{}_m}{w_f.Math.{}_m+\left(1-w_f\right).Math.{}_f}& \left(3\right)\end{array}$

where V.sub.f is the volume fraction of nanoparticles in the nanocomposite, d is the diameter of the nanoparticles, w.sub.f is the mass fraction of nanoparticles in the nanocomposite (3% in all nanocomposites), pr is the density of the nanoparticles (5610 kg m.sup.3) and .sub.m is the density of the LDPE (920 kg m.sup.3). The theoretical IPDs for the nanocomposites based on 3 wt. % ZnO were ca. 90, 2000 and 7500 nm respectively, for particles sizes of 25, 550 and 2000 nm. The much lower IPD for the nanoparticle nanocomposite systems indicated their high interfacial surface area with LDPE matrix.

[0431] In order to gain more information about the interfacial adhesion between nanoparticles and LDPE matrix, tensile testing was performed (FIG. 18). The strain-at-break (%) of the nanocomposites based on the C8 and C18-coated nanoparticles (610-660) was higher than that of the pristine LDPE and nanocomposites based on C1-coated nanoparticles (420-460). The nanocomposites based on uncoated nanoparticles showed much lower strain-at-break with reference to pristine LDPE, indicating their poor interfacial adhesion due to presence of big aggregates and voids. All samples showed a stress drop in stress-strain curve after the first yield point. Liu et al., Eur. Polym. J., 2015, 66, 67-77 observed the cavitation around particles in the LDPE/Al.sub.2O.sub.3 nanocomposites at the strain of the second yield point. Here, the higher strain for cavitation was observed for LDPE nanocomposites based on C8- and C18-coated nanoparticles, which suggested the strongest interfacial adhesion occurred between nanoparticles and LDPE matrix in these nanocomposite samples. The long chain of these silanes on the surface of the nanoparticles increased the surface hydrophobicity, which enhanced their compatibility with the LDPE matrix.

DC Conductivity of the LDPE/ZnO Nanocomposites

[0432] The electrical insulation capacity of the LDPE nanocomposites based on coated ZnO nanoparticles was compared to unfilled LDPE: FIG. 19a shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C1, and unfilled LDPE (Reference), FIG. 19b shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C8, and unfilled LDPE (Reference), and

[0433] FIG. 19c shows nanocomposites comprising 0.1, 1.0 and 3.0 wt. %, respectively, of LDPE/ZnO-25-C18, and unfilled LDPE (Reference). The charging current was always lower for all nanocomposite systems during the entire measurement, as compared with unfilled LDPE. The level of conductivity decreased by addition of coated ZnO nanoparticles from 0.1 to 3 wt. %. The addition of 3 wt. % C8-coated ZnO nanoparticles decreased the conductivity of the LDPE by 2-3 orders of magnitude (11 h value), which is much larger effect than obtained by adding a similar amount of C1 and C18-coated nanoparticles. It was reported that the LDPE with different crystallisation behaviour show different conductivity due to contrasting conductive crystallites and resistive amorphous regions, see T. J. Lewis, IEEE Trans. Dielectr. Electr. Insul., 2014, 21, 497-502. DSC was used to study the crystallisation and melting of the nanocomposites. No significant changes in crystallinity (0.4-0.45) or melting peak temperature (111 C.) were observed on addition of the nanoparticles to the LDPE matrix. The addition of nanoparticles thus had no significant effect on the crystallization of the polymer. However, the ZnO nanoparticles with their large surface areas act as additional electron traps, and reduce the average hopping distance for the charge carriers with reference to that of the LDPE matrix see K. Y. Lau et al., J. Phys.: Conf. Ser., 2013, 472, 012003 and T. J. Lewis, IEEE Trans. Dielectr. Electr. Insul., 2014, 21, 497-502. The conductivity suppression level was decreased by decreasing the specific surface area of ZnO particles from 34 to 13 m.sup.2 g.sup.1, A. M. Pourrahimi, et al., Mater. Chem. A, 2015, 3, 17190-17200 (FIGS. 19a-c and 20). The charge transport on the nanoparticles with large surface activity was facilitated due to the presence of surface defects, see T. Konig et al., J. Am. Chem. Soc., 2009, 131, 17544-17545. Here, these defects became inactive while the nanoparticles were covered by large amount of silane (C1 and C18, See Table 3). The C8-coated nanoparticles showed greatest reduction in conductivity due to the low amount of silane coverage, which resulted in accessible defects for the charge transport. Another important characteristic of the C8-coating is their high porosity compared to other silane coating which provided new trap sites for charge carriers.

[0434] The dispersion of the nanoparticles was analysed to identify relations between nanoparticle dispersion state and conductivity values in the LDPE/ZnO nanocomposites.

[0435] FIG. 21 shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C1 with a nanoparticle content of 3 wt. % ZnO.

[0436] FIG. 22 shows the corresponding bitmap of FIG. 21 with ZnO nanoparticles as the black phase.

[0437] FIG. 23a shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C8 with a nanoparticle content of 3 wt. % ZnO.

[0438] FIG. 23b shows scanning electron micrograph of nanocomposite of LDPE and ZnO-25-C18 with a nanoparticle content of 3 wt. % ZnO.

ZnO Phase Distribution in the Fractured Surfaces 2-Dimensions (2D) (See Above the Corresponding Paragraph for LDPE/MgO Nanocomposites for Details)

[0439] FIGS. 24 and 25 show the overall dispersion of the ZnO phase in the PE matrix for the nanocomposites with different weight fractions of nanoparticles, i.e. regarding the aggregated ZnO nanoparticles as discrete phases (in accordance with FIG. 22).

[0440] FIG. 24 shows centre-to-centre distance for the 1.sup.st neighbour as a function of volume percentage of 25 nm ZnO nanoparticles surface modified with C1, C8 and C18 alkyl chain on the silane at 1 wt. % and 3 wt. %.

[0441] FIG. 25 shows centre-to-centre distance for the 51.sup.st neighbour as a function of volume percentage of 25 nm ZnO nanoparticles surface modified with C1, C8 and C18 alkyl chain on the silane at 1 wt. % and 3 wt. %.