Polyethylene Nanocomposites for the Next Generation of Ultralow

May 20, 2016 - KTH Royal Institute of Technology, School of Chemical Science and Engineering, Fibre and Polymer Technology, SE-100 44 Stockholm, Swede...
0 downloads 14 Views 4MB Size
Subscriber access provided by UCL Library Services

Article

Polyethylene Nanocomposites for the Next Generation of Ultra-low Transmission-loss HVDC Cables: Insulations Containing Moisture-resistant MgO Nanoparticles Amir Masoud Pourrahimi, Love K. H. Pallon, Dongming Liu, Tuan Anh Hoang, Stanislaw Gubanski, Mikael S. Hedenqvist, Richard Tobias Olsson, and Ulf W. Gedde ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b04188 • Publication Date (Web): 20 May 2016 Downloaded from http://pubs.acs.org on May 26, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Submitted to ACS Applied Materials & Interfaces Revised 2016–05–17

Polyethylene Nanocomposites for the Next Generation of Ultra-low Transmission-loss HVDC Cables: Insulations Containing Moisture-resistant MgO Nanoparticles

Amir Masoud Pourrahimi,a Love K. H. Pallon,a Dongming Liu,a Tuan Anh Hoang,b Stanislaw Gubanski,b Mikael S. Hedenqvist,a Richard T. Olssona and Ulf W. Geddea*

a

KTH Royal Institute of Technology, School of Chemical Science and Engineering,

Fibre and Polymer Technology, SE-100 44 Stockholm, Sweden b

Chalmers University of Technology, Department of Materials and Manufacturing

Technology, High Voltage Engineering, SE-412 96 Göteborg, Sweden

* Corresponding author. Tel.: +46 8 790 7640. Fax: +46 8 208856. E-mail address: [email protected] 1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The use of MgO nanoparticles in polyethylene for cable insulation has attracted a considerable interest, although in humid media the surface regions of the nanoparticles undergo a conversion to hydroxide phase. A facile method to obtain MgO nanoparticles with a large surface area and remarkable inertness to humidity is presented. The method involves (a) low temperature 400 °C thermal decomposition of Mg(OH)2, (b) a siliconeoxide coating to conceal the nanoparticles and prevent inter-particle sintering on exposure to high temperatures, and (c) heat treatment at 1000 °C. The formation of hydroxide phase on these silicone oxide-coated MgO nanoparticles after extended exposure to humid air was assessed by thermogravimetry, infrared spectroscopy and X-ray diffraction. The nanoparticles showed essentially no sign of any hydroxide phase compared to particles prepared by the conventional single-step thermal decomposition of Mg(OH)2. The moisture-resistant MgO nanoparticles showed their improved dispersion and interfacial adhesion in the LDPE matrix with smaller nano-sized particle clusters than conventionally prepared MgO. The addition of 1 wt. % moisture-resistant MgO nanoparticles was sufficient to decrease the conductivity of polyethylene 30 times. The reduction in conductivity is discussed in terms of defect concentration on the surface of the moistureresistant MgO nanoparticles at the polymer/nanoparticle interface.

Keywords: MgO nanoparticles, thermal decomposition, surface coating, humidityresistance, HVDC cable.

2 ACS Paragon Plus Environment

Page 2 of 30

Page 3 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1. Introduction Remote regions endowed with carbon-dioxide-neutral energy sources (solar, wind or hydro) need to be connected to densely populated areas to ensure renewable and costeffective electric power.1 High-voltage direct-current (HVDC) cables are the most feasible solution in view of geography (transmission across sea is often required) and to avoid the use of overhead transmission lines.2 The key to reduce losses in the transmission of electric power through cables is to use very high DC voltages. Current state-of-the-art extruded cables insulated with crosslinked polyethylene reach a transmission voltage level of 525 kV with a power rating range of up to 2.6 GW.3 The need for larger, longer and more efficient transmission capacity is expected to require a maximum voltage of 1 MV by 2030.4 The increase in DC voltage from the present 525 kV to 1 MV requires a reduction in the electrical conductivity of the insulating material by a factor of five. One feasible strategy to reach this target is to incorporate a low concentration of dispersed metal oxide (e.g. MgO) nanoparticles in polyethylene. MgO nanoparticles with high surface activity create trapping sites for charged species, suppress space charge accumulation in the insulating polymer matrix, and reduce the conductivity in composites by one order of magnitude in a high electric field at low filler loadings (typically lower than 5 wt.%).5-12 Various methods have been used to prepare MgO nanoparticles, including aqueous precipitation of Mg(OH)2 followed by thermal decomposition at temperatures above 400 °C, which yields uniform nanoparticles with a high surface activity. For the first time, Pallon et al5 reported highly porous foams of Mg(OH)2 nanoparticles obtained by a freezedrying method in order to avoid particle agglomeration and contamination by subsequent grinding. The calcination of Mg(OH)2 at 300–400 °C resulted in 90% conversion of Mg(OH)2 to MgO13 and the formation of polycrystalline MgO nanoparticles containing nanopores (< 3 nm).5, 14-17 The nano-sized crystallites had a high surface concentration of hydroxide groups, which at elevated temperatures (> 400 °C) could potentially react with hydroxide groups situated on adjacent crystallites. Heat treatment at an even higher temperature transforms this structure by sintering into highly pure and larger cubic MgO particles (known as dead- or hard-burned magnesia) with a low specific surface area.13, 15

3 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

In the cable-manufacturing industry, the insulation materials including nanoparticles are exposed to a humid environment.8, 18 The accumulation of water molecules around the nanoparticles provides excess charge carriers thus increasing the electrical conductivity of insulating nanocomposites.18 High-surface-activity MgO nanoparticles readily adsorb water and are converted into Mg(OH)2.14, 19-24 Commercially available MgO nanoparticles contain a significant fraction of hydroxide (see Figure S1, Supporting Information), which in turn can facilitate the formation of hard nanoparticle agglomerates in the composites that reduce the electrical breakdown strength,8, adsorb

impurities.27

Although

several

25-26

papers

and the hydroxide phase may also mention

that

the

LDPE/MgO

nanocomposites have a high potential as DC insulation materials, high-purity MgO nanoparticles with a large surface activity and no hydroxide phase have hitherto not been reported. By applying a thin silica coating to metal oxide nanoparticles (Fe3O4, ZnO), it has previously been demonstrated that heat treatment can purify the nanoparticles (i.e. removing counter ions and surface water) without undesirable inter-particle sintering and with preserved surface activity.28-30 This approach has not however been evaluated for MgO nanoparticles intended for HVDC insulation materials. This paper presents a novel method using a nanoparticle coating prior to a high temperature heat treatment in order to achieve highly pure, uniform and nano-sized MgO particles, which seems to be ideal for ultra-high-voltage insulation materials. The new method is herein after referred to as the modified two-step heat treatment of Mg(OH)2, involving three steps: a low temperature heat treatment at 400 °C for thermal decomposition into ca. 90% pure MgO, the application of a hydrocarbon functional silicone oxide coating, and finally a heat treatment at 1000 °C for complete Mg(OH)2 transformation. The method yields large surface area MgO nanoparticles with a remarkable inertness to humidity. The partial silica coverage of these MgO nanoparticles provides a high interfacial adhesion between nanoparticles and low-density polyethylene matrix. These large surface area nanoparticles reduce the electrical conductivity of pristine polyethylene by more than an order of magnitude.

4 ACS Paragon Plus Environment

Page 5 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

2. Experimental Section 2.1. Materials Magnesium chloride hexahydrate (MgCl2·6H2O, ACS Reagent, Sigma-Aldrich), sodium hydroxide (≥ 98 wt.%, Sigma Aldrich), Milli-Q water (18.2 MΩ cm at 25 °C), octadecyl trimethoxy silane (OTMS, ≥ 90 wt.%, Sigma Aldrich), n-heptane (≥ 99 wt.%, VWR), ethanol (≥ 96 wt.%, VWR), potassium bromide (KBr, ≥ 98 wt.%, FTIR grade, Sigma Aldrich), and Irganox 1076 (Ciba Specialty Chemicals, Switzerland) were used as received. Low-density polyethylene (density (23 °C) = 922 kg m-3; melt flow index = 2 g (10 min)-1 (190 °C; 2.16 kg; ISO 1133)) was supplied by Borealis AB, Sweden.

2.2. Preparation and coating of MgO nanoparticles 2.2.1. Synthesis of Mg(OH)2 nanoparticles Mg(OH)2 nanoparticles were prepared by an aqueous precipitation method described by Pallon et al.5 1 L of 0.75 M magnesium chloride aqueous solution was added to 1 L of 1.5 M sodium hydroxide solution and allowed to react for 30 min at 23 °C under vigorous stirring. The precipitate was washed thrice with milli-Q water under ultrasonication (Bandelin Sonorex RK 100H, volume = 3 L, ultrasonic peak output = 320 W, frequency = 35 kHz) in order to remove residual counter ions.31 The washed precipitate of Mg(OH)2 nanoparticles was divided into two fractions: one fraction was dried in air at 90 °C and ground into a fine powder using a pestle and mortar (sample referred to as Mg(OH)2-CD), and the second fraction was freeze-dried (sample referred to as Mg(OH)2-FD). The freezing was carried out in liquid nitrogen for 10 min using 5 mL batches (5 wt.% aqueous suspension), after which the batches were dried for 12 h in a CoolSafe freeze drier (Scanvac) operating at 6 Pa and –96 °C. 2.2.2. Single-step heat treatment of Mg(OH)2 The Mg(OH)2 powder was heated in dry air from 23 °C to 1000 °C at a rate of 10 °C min-1 and held at 1000 °C for 1 h in a H14-GAXP furnace (Micropyretic Heaters International Inc.). The thermally treated particle samples designated MgO-CD1000 and 5 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

MgO-FD1000 were stored in a desiccator at reduced pressure to prevent the uptake of CO2 and H2O. 2.2.3. Modified two-step heat treatment of the Mg(OH)2 The Mg(OH)2-CD sample was first treated in ambient air at 400 °C for 1 h, yielding particles designated MgO–CD400. MgO–CD400 (2 g) was dispersed in 600 mL n-heptane and 3.66 mL octadecyl trimethoxy silane (OTMS) was added to the suspension under rapid stirring and allowed to react for 24 h at 23 °C. The coated nanoparticles, designated MgO–C18, were heated from 23 °C to 1000°C at a rate of 10 °C min–1 and held at 1000 °C for 1 h yielding particles designated MgO–C18–1000. The thermal treatments were carried out in dry air. 2.3. Nanoparticle characterisation A field emission scanning electron microscope (SEM; Hitachi S–4800) and a transmission electron microscope (TEM; Hitachi HT7700) were used to assess the particle shape and size distributions. For the SEM analysis, the powder samples were coated with a conductive layer of Pt/Pd (60/40) by a 20 s sputtering using a current of 80 mA in a Cressington 208 HR. For the TEM analysis, the particles were deposited onto 400 mesh copper grids from a suspension of pure ethanol containing nanoparticles at a concentration of 0.45 ± 0.05 g L–1. The samples were then dried and examined in the microscope operating at an acceleration voltage of 100 kV. The size distributions of the different nanoparticle samples were assessed by measuring 250 particles per sample using Image J (National Institute of Health, Maryland, USA). The Brunauer-Emmett-Teller (BET) method based on nitrogen adsorption/desorption with a Micromeritics ASAP 2000 at 77 K was used to determine the specific surface area and pore size distribution. Before measurement, the samples were degassed at 200 °C until the pressure reached 0.3 hPa. The different nanoparticle samples were exposed to humidity at 23 °C and 50% RH. X-ray diffractograms of the powder samples were recorded at 23 ± 2 °C using a PANalytical X’pert Pro MPD diffractometer with a Cu-Kα source using a 2θ step size of 0.017°. The crystal size was obtained from the Scherrer equation:

6 ACS Paragon Plus Environment

Page 6 of 30

Page 7 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

D=

kλ β cos(θ )

(1)

where D is the crystal size (diameter) in nm, k is a shape factor equal to 0.94, λ is the Xray wavelength (0.154178 nm), θ is the Bragg angle and β is the peak width in radians at half-height. Thermogravimetry (TG) was carried out in a Mettler Toledo TG/DSC 1. Powder samples (5 ± 1 mg) were placed in 70 µL aluminium oxide crucibles and heated from 30 °C to 1000 °C at a rate of 10 °C min-1 in dry oxygen (flow rate = 50 mL min–1). Infrared spectra were recorded with a Perkin-Elmer Spectrum 2000 FTIR spectrometer and analysed with Perkin-Elmer Spectrum software (Norwalk, CT, USA). The samples were prepared by mixing 3 mg of particles and 300 mg of ground KBr and pressing the mixture under a load of 10 kN for 1 min in a die to form a pellet. The spectral range was 450 to 4000 cm-1 with a resolution of 4 cm–1. Each spectrum was based on 32 scans. 2.4. Preparation of composite samples MgO particles – MgO-CD400, MgO-CD1000 and Mg-C18-1000; further information about their preparation are presented in Section 2.2 – (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 ultrasonicated 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 mixture was dried at 80 °C overnight, after which it was shaken for 1 h. The powder obtained was melt-compounded in a Micro 5cc Twin Screw Compounder (DSM Xplore) at 150 °C for 6 min with a screw speed of 100 rpm. The extruded nanocomposite was compression-moulded under a load of 200 kN into a 80 µm thick film 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 10 °C min-1 while maintaining the compressive load. 2.5. Characterization of composite samples The crystallisation and melting of the nanocomposites were studied in a Mettler-Toledo differential scanning calorimeter DSC 1. The samples (5.0 ± 0.5 mg; enclosed in 100 µl aluminium pans) were heated at a rate of 10 °C min-1 from 25 to 150 °C and kept at this temperature for 5 min in nitrogen (flow rate = 50 mL min-1) in order to erase the effects of 7 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

the previous thermal history. The samples then were cooled to -50 °C at a rate of 1 °C min-1 to record the crystallization temperature and finally heated to 150 °C at a rate of 10 °C min-1 to record the melting temperature. The mass crystallinity (wc) of the polymer was determined according to the total enthalpy method:32 wc =

∆H f  w p  ∆H 0f −  



Tm0

T1

(c p , a

(2)

 − c p,c )dT   

where ∆Hf is the measured melting enthalpy, ∆Hf0 is the melting enthalpy for 100 % crystalline polyethylene (293 J g-1)

32

at the equilibrium melting point, T1 is an arbitrary

temperature below the melting range, Tm0 is the equilibrium melting temperature, wp is the mass fraction of polymer in the nanocomposite33 and cp,a and cp,c are respectively the specific heat capacities of the amorphous and crystalline components, which were obtained from Wunderlich and Baur.34 The crystallinity at a given temperature (wc(T)) within the melting temperature range was obtained according to the expression: wc (T ) = wc ×

∆H f (T )

(3)

∆H f

where ∆Hf (T) is the melting enthalpy associated with melting above temperature T, which was obtained by truncating the broad melting peak at this particular temperature. The crystal thickness (Lc) and amorphous thickness (La) associated with the melting peak temperature were calculated according to the Thomson-Gibbs equation:35 Lc =

La =

2σ  T ∆H 0f ρ c 1 − m0  T m 

(4)

   

ρ c Lc (1 − wc ) ρ a wc

(5)

where Tm is the melting peak temperature, ρc = 1003.0 kg m-3 and ρa = 851.9 kg m-3 are the densities respectively of the crystalline and amorphous components and σ = 93 mJ m-2 is the fold surface free energy for linear polyethylene.36-37 Stress-strain data for unfilled LDPE and its nanocomposites were obtained using an Instron 5944 tensile testing machine equipped with a 50 N load cell. The tests were performed at 23 ± 1 °C and 50% RH using a strain rate of 50% min-1. The cuboid specimens tested were 5 mm wide and 80 µm thick. The gauge length was 30 mm.

8 ACS Paragon Plus Environment

Page 9 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

2.6. Electrical conductivity measurements The electrical conductivity measurements were performed by applying a 2.6 kV DC voltage from a power supply (Glassman FJ60R2) across the 80 µm thick film sample and measuring the current using a Keithley 6517A electrometer. The electric field across the film was 32.5 kV mm-1. The detected current signal was recorded by LabVIEW software incorporated in a personal computer and the data were stored for further analysis. An oven was used to control the temperature, and an overvoltage protection secured the electrometer from damage due to possible overshoots and a low-pass filter removed any 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, and the guard ring eliminated surface currents. Good contact between 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 4 × 104 s (11.1 h). 3. Results and Discussion 3.1. Morphology of Mg(OH)2 and MgO nanoparticles Figures 1a–c shows the morphology of the Mg(OH)2 conventionally dried at 90 °C (Mg(OH)2-CD) and of the MgO after the first and second step heat treatment at 400 °C (MgO-CD400) and 1000 °C (MgO-CD1000), respectively. The Mg(OH)2 nanoparticles showed a hexagonal character with an average size of 43 nm within the (001) plane and a thickness of 10–20 nm along the [001] direction (Figure 1a). The Mg(OH)2 nanoparticles transformed into ca. 90% pure MgO platelets via dehydration and rearrangement of the crystal lattice during the treatment at 400 °C, i.e. to cubic close-packed (CCP) crystallites with a cuboid shape inside the platelets (Figure 1b). Naono13 reported a contraction of 49 % along the [001] direction of Mg(OH)2 as a result of a similar heat treatment at 400 °C. During the dehydration associated with the contraction, open slit-shape mesopores formed between the cuboid MgO crystallites, while the hexagonal framework was retained. The open slit-shaped mesopores can be observed inside the platelets in Figure 1b. The crystallites overlap and interact by covalent bonds between the MgO and the 9 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

residual layered Mg(OH)2 phase, which has been reported to result in a framework strength of ca. 50 MPa.16 The residual Mg(OH)2 transformed completely into MgO during the second heat treatment at 1000 °C, while undesirable simultaneous sintering of the platelets occurred into cubic particles with an average size of ca. 150 nm (Figure 1c).

Figure 1 Transmission and scanning electron micrographs of (a) Mg(OH)2 nanoparticles (Mg(OH)2-CD), MgO particles after dehydration at (b) 400 °C (first step, MgO-CD400) and (c) 1000 °C (second step, MgO-CD1000) heat treatments.

Figure 2 a–b and d–e present scanning electron micrographs of the conventionally and freeze-dried Mg(OH)2 nanoparticles and of the ultimately formed MgO particles after their heat treatment at 1000 °C, respectively. The conventionally dried Mg(OH)2 nanoparticles showed compact and hard agglomerates in contrast to the freeze-dried sample, which showed a porous and fragile foam structure of weakly associated nanoparticles (Figures 2a,b). Heat treatment of the conventionally dried sample at 1000 °C resulted in 50 % larger MgO particles with a broader size distribution than the freezedried sample heated in the same manner at 1000 °C (Figures 2d–e). Previously, it was 10 ACS Paragon Plus Environment

Page 11 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

reported that a residual stress remains among oriented Mg(OH)2 nanoparticles after the freeze-drying procedure due to the consolidation of the nanoparticles into an open porous network.5 Pourrahimi et al

28

suggested that a repulsive interaction energy between

oriented nanoparticles inhibited the particle growth during the high temperature heat treatment. It is here suggested that the very fast quenching of the Mg(OH)2 nanoparticle suspensions in liquid nitrogen (prior to the freeze drying) prevented stacking of the nanoparticles and therefore had the same effect of inhibiting particle growth. The heat treatment at 1000 °C reduced the specific surface area of the Mg(OH)2 nanoparticles regardless of the drying method into 17–18 m2g-1 (Table 1). Although the freeze-drying before heat treatment at 1000 °C retained the small particle size, this method was not successful for obtaining particles with a larger specific surface area. The freeze-dried nanoparticles were not therefore further investigated.

Figure 2 Scanning electron micrographs of (a) Mg(OH)2-CD, (b) Mg(OH)2-FD, (c) MgO-CD400, (d) MgO-CD1000, (e) MgO-FD1000 and (f) MgO-C18-1000. The insets show the particle size distribution.

The modified two-step heat treatment was developed in order to obtain MgO particles with a large surface area without the risk of undesirable sintering during the hightemperature heat treatment. The first step was carried out at a lower temperature of 400 °C in order to minimize the particle growth (Figure 2c). The temperature was selected as the lowest possible temperature that allowed extensive conversion of the Mg(OH)2 nanoparticles into 90% pure MgO platelets.5 The silicone oxide coating of these platelets 11 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

prior to the second heat treatment at 1000 °C resulted in 2.35 times larger specific surface area and 62 % smaller particles compared to that obtained by the conventional single-step heat treatment (Figures 2d and f, Table 1). The MgO nanoparticles prepared by the modified two-step heat treatment had an average size of 58 nm and a specific surface area of 40 m2g-1. These values were close to the values of the Mg(OH)2 nanoparticles, 43 nm and 44 m2g-1. It is suggested that the modified two-step procedure, which resulted in fullytransformed MgO, hindered undesirable sintering and preserved the surface area of the Mg(OH)2 nanoparticles during the final heat treatment. Table 1 Characteristics of Mg(OH)2 and MgO Particles. Sample Particle size BET surface area

Crystal size a

Volume of unit-cell

(m g )

(nm)

(Å3)

2

(nm)

-1

Mg(OH)2-CD

43±14 b

44

-

-

Mg(OH)2-FD

43±14

b

52

-

-

72±16

c

167

9

75.6

MgO-CD400

c

MgO-CD1000

149±60

17

39

75.2

MgO-FD1000

99±32

c

18

23

75.2

58±16

c

40

29

75.2

MgO-C18-1000 a

Crystal size was calculated by the Scherrer equation (Eq. (1)) applied to the X-ray diffraction peak at 42° originating from the (200) planes. b Nanoparticle size measured over the (001) plane as revealed by TEM (Figure 1a). c Particle size as revealed by SEM (Figure 2).

12 ACS Paragon Plus Environment

Page 13 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 3 Nitrogen adsorption (red filled squares)-desorption (blue open squares) isotherms for (a) Mg(OH)2-CD, (b) MgO-CD400, (c) MgO-CD1000, (d) MgO-C18-1000. Inset: the corresponding BJH (Barret-Joyner-Halenda) pore size distribution.

Figures 3a–c show nitrogen desorption/adsorption isotherms together with the corresponding Barret-Joyner-Halenda (BJH) pore size distributions (inset graphs) of the Mg(OH)2 nanoparticles (Figure 3a), and of the MgO particles formed after the heat treatments at 400 and 1000 °C (Figures 3b and c). All the samples conformed to the type III isotherm with small H1 hysteresis loops at P/P° > 0.85, suggesting that the particles had a uniform size.28, 38 A small H3 hysteresis loop at 0.4 < P/P° < 0.7 was observed in the case of MgO prepared at 400 °C, indicating that small mesopores with a slit shape remained after the thermal annealing (Figure 3b).14, 24, 39 These data were consistent with the observation of open and parallel pores between the primary MgO crystallites inside the hexagonal platelets (Figure 1b). The small and open mesopores, which resulted in a large specific surface area of 167 m2g-1, coexisted with residual layered hydroxide,13 which had to be removed in order to obtain a fully-dehydrated MgO particles. The single-step heat treatment of Mg(OH)2 at 1000 °C showed no sign of mesopores but the specific surface area dropped sharply from 44 to 17 m2g-1 (Figure 3c). Figure 3d shows that the modified

13 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

two-step heat treatment successfully removed the mesopores and residual hydroxide phase while retaining the specific surface area of the Mg(OH)2 nanoparticles. 3.2. Thermal decomposition of Mg(OH)2 into MgO Figure 4a shows the normalized mass plotted as a function of temperature for the conventionally dried Mg(OH)2 and for the MgO nanoparticles coated with the silane after the 400 °C thermal treatment. The total mass loss of 32.8 % in the Mg(OH)2-CD thermogram was slightly larger than the theoretical decomposition value of 30.9 %. This discrepancy was due to the presence of residual water in the Mg(OH)2 sample. Residual water can be loosely bound on the exterior surface of the Mg(OH)2 nanoparticles (< 200 °C) or tightly bound with the surfaces of the cuboid crystallites (> 400 °C).13 A more detailed analysis of the Mg(OH)2-CD sample thermogram showed three main mass loss regimes that could be associated with a step-wise dehydration of magnesium hydroxide into the oxide phase: 275–375 °C (26.2 wt.%), 375–750 °C (3.8 wt.%) and 750–825 °C (1 wt.%). For the silane-coated MgO sample (MgO-C18) that had been pre-heated at 400 °C, only two mass losses occurred at 200–475 °C (8.1 wt.%) and 750–875 °C (2.7 wt.%). These mass losses were due to transformation of the condensed silane coatings into a silica layer after degradation of the hydrocarbon functional units, and thermal decomposition of the small amount of remaining layered Mg(OH)2, respectively. Since the mass loss at 375–750 °C was not observed in the case of the MgO-C18 sample (with reference to Mg(OH)2-CD), it is suggested that the silica layer shifted the removal of residual water and layered hydroxide to a higher temperature (> 750 °C).

14 ACS Paragon Plus Environment

Page 15 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 4 (a) normalized mass plotted as a function of temperature as revealed by TG, (b) IR spectra and (c) X-ray diffractograms of Mg(OH)2 and different formed MgO particles; (d) normalized mass plotted as a function of temperature, (e) IR spectra and (f) X-ray diffractograms of differently prepared MgO particles after exposure to humid air (RH 50%, 1 week). The arrows show the layered hydroxide phase.

The transformation of Mg(OH)2 into MgO by thermal treatment was confirmed for all the samples by infrared spectroscopy (Figure 4b). The intense absorbance band at 3660–3800 cm-1 was assigned to the O–H stretching of the free surface hydroxyls. The broad absorbance bands at 3200–3600 and 1635 cm-1 were assigned to O–H stretching of physically and chemically sorbed water, respectively.28 The MgO particles obtained by a single-step heat treatment at 1000 °C (MgO-CD1000) showed none of the absorbance band assigned to the hydroxide phase. The silane-coated MgO (MgO-C18) showed three absorption peaks between at 2850–2990 cm-1 assigned to C–H stretching,30 and one peak at 1468 cm-1 assigned to CH2 units not covalently bonded to a silicon atom40 (within the C18H37– terminal alkyl group of the silane). The heat treatment at 1000 °C resulted in the removal of the alkyl group absorbance bands; see MgO-C18-1000. The two intense peaks at 1033 and 1076 cm-1 were due to antisymmetric stretching of the Si–O–Si groups in the 15 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

silica layer formed by the heat treatment at 1000 °C.40 The schematic drawing of molecular conformation for silane coatings with relatively long alkyl chain (C18H37–) revealed that silicon oxide bridges was not formed due to hindrance of carbon moieties.30 After the heat treatment of the C18-coated MgO nanoparticles at 1000 °C and further removal of carbon moieties the silica layer therefore covered partially on the MgO particles surfaces’. The absence of absorption at 1120 cm–1 (Si–O–Si band for crosslinked silicon oxide structure) indicated that the silica layer covered only a part of the particle surface.33 In contrast to the MgO-C18 particles with a trace peak at 3660–3800 cm-1 (shown by the arrow, Figure 4b), the MgO-C18-1000 particles showed no absorbance bands assigned to the layered hydroxide, which indicated that the MgO nanoparticles were completely dehydrated. The X-ray diffraction results showed that the brucite crystal structure of Mg(OH)2 was completely transformed into the cubic crystal structure of MgO (Figure 4c) in MgOCD1000. The small amount of layered Mg(OH)2 in the MgO-C18 particles (indicated by the arrow in Figure 4c) was also erased as a result of the heat treatment at 1000 °C (MgOC18-1000). The cubic lattice parameters, i.e. the crystal interplanar distances (d) and the lattice constants a = b = c were calculated from the Bragg equation (λ = 2d sin θ) and the plane lattice geometry (Miller index of (hkl)) according to: 41 1 h2 + k 2 + l 2 = d2 a2

(6)

A small peak shift towards higher diffraction angles, i.e. to shorter interplanar distances and thus a smaller unit cell volume (V = a3) was observed in all the samples after the heat treatment at 1000 °C (Table 1). The shrinkage of the unit cell was attributed to the relaxation of defects inside the crystal unit cell e.g. magnesium and oxygen vacancies.5, 28, 42

The effect of these surface defects on DC conductivity of LDPE is further discussed in

Section 3.5.

16 ACS Paragon Plus Environment

Page 17 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 5 Scanning electron micrographs of (a-c) Mg(OH)2 and (d-f) OTMS-coated MgO-CD400 (MgO-C18) nanoparticles after heat treatment for 1 h at different temperatures. The inset micrographs were obtained by TEM.

Prior to the silane coating procedure, the crystallites of the MgO prepared at 400 °C were 9 nm in size, which was equal to the size of the cuboid crystallites inside the hexagonal platelets (Figure 1b). The coating of the platelets prior to the second heat treatment at 1000 °C resulted in crystals with an average size of 29 nm (X-ray diffraction), which were contained inside the 58 nm MgO-C18-1000 particles. In comparison, the single-step heat treatment of Mg(OH)2 at 1000 °C yielded MgO crystals size 39 nm in size inside 150 nm particles as revealed by SEM (Table 1). It was therefore concluded that the nanoparticles obtained by the modified two-step heat treatment were dominantly mono/dual domain particles, whereas the conventional heat treatment resulted in more polycrystalline particles (containing in ca. 4 crystal domains in each particle). Figure 5 shows electron micrographs of Mg(OH)2-CD thermally decomposed into MgO at 400, 600 and 800 °C (top), in comparison with the coated MgO-C18 nanoparticles (bottom). As noted before, most of the Mg(OH)2 was decomposed into highly porous MgO platelets at 400 °C, resulting in a hexagonal framework. The intra-platelet mesopores among the MgO crystallites were removed at 400–600 °C, while the hexagonal platelet shape (framework) remained intact (Figures 5a and b). The hexagonal framework fragmented partially into sub-nanoparticles ca. 20 nm in size at 800 °C (Figure 5c). Itatani 17 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 30

et al16 attributed the collapse of the platelet framework to the removal of residual layered hydroxide, which had created bonds between cuboid MgO crystallites. After fragmentation when the thermal decomposition was complete (> 800 °C, according to TG, Figure 4a), the individual sub-nanoparticles began to rearrange and sinter into submicron particles at 1000 °C (Figure 2d). Figure 5e shows that the silica coating encapsulated the platelets after exposure to 600 °C (inset TEM image), and hindered inter-platelet sintering (shown by arrows). From the micrographs it is evident that the fragmentation into ca. 20 nm particles which was observed during the conventional single-step heat treatment at 800 °C never occurred when the silica coating was present (Figure 5 c and f). It was also confirmed by TG that the removal of residual layered hydroxide shifted to higher temperatures (> 750 °C) due to presence of the silica coating (Figure 4a). It is suggested that the silica coating on MgO platelets increased the strength of the platelet, and thereby inhibited extensive fragmentation of the platelets (removal of residual layered hydroxide) in the temperature range 600–800 °C. The concealing of the platelets probably also had a major influence on the merging of the crystallites (that occurred inside the silica shells), into mono/dual-domain nanoparticles.

3.3. Resistance of silica-coated MgO particles to humidity The moisture-resistant character of high-temperature-treated dehydroxylated silica as a highly hydrophobic surface was previously reported.43 Figure 4d shows the normalized mass plotted as a function of temperature for the different MgO particles after exposure to humid air for 1 week (50 %RH at 23 °C). The MgO nanoparticles prepared by heat treatment at 400 °C (MgO-CD400) showed three main mass losses at 25–200 (8.5 wt.%), 275–375 (11 wt.%) and 750–875 °C (2 wt.%). The relatively large mass loss at 25–200 °C was due to loosely bound water, which was previously shown to occur as a result of the immediate adsorption of water molecules on the large surface area MgO nanoparticles.2223

The mass loss at 275–375 °C was assigned to thermal decomposition of Mg(OH)2

formed via the re-hydration. According to the stoichiometry of the re-hydration reaction, 24.4 % of the MgO nanoparticles transformed back to Mg(OH)2 during exposure for one week to air at 50 % RH. The particles prepared by the one-step heat treatment at 1000 °C (MgO-CD1000) showed a mass loss of 0.8 % at 300–400 °C. This rather low mass loss 18 ACS Paragon Plus Environment

Page 19 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

was equivalent to a transformation of only 1.8 % of MgO into Mg(OH)2. The fullydehydrated MgO showed a considerable resistance towards re-hydration once the complete condensation into pure MgO had occurred. The nanoparticles with a partial silica coverage prepared by the modified two-step heat treatment (MgO-C18-1000) showed essentially no sign of a hydroxide phase after exposure to humid air. The transformation of MgO back to Mg(OH)2 was also studied by infrared spectroscopy and X-ray diffraction (Figures 4e and f). The MgO nanoparticles prepared by heat treatment at 400 °C (MgO-CD400) showed a narrow peak at 3700 cm-1 due to the O–H stretching of Mg(OH)2 formed by the re-hydration (Figure 4e, indicated by arrow).39 The absorbance at 3200–3600 cm-1 was attributed to O–H stretching of adsorbed water, and clearly increased after the exposure to humid air (Figures 4b and e). The peaks centred at 850, 1460 and 1650 cm-1 originated from different forms of dissociation of CO2 and water into carbonate and hydrogen carbonate species on the MgO nanoparticle surfaces.44 The intensity of the XRD-peak at 2θ = 31.6°, corresponding to the (100) crystal plane of the Mg(OH)2, was not changed after exposure to moist air of the sample prepared at 400 °C (Figures 4c and f, indicated with arrow). Since, the amount of the hydroxide phase formed was according to TG results larger than the X-ray detection limit (ca. 3 nm),31 the Mg(OH)2 formed on MgO nanoparticles during the humidity exposure must have had an amorphous structure. The two MgO samples prepared at 1000 °C showed only very weak absorbance bands in the 3200–3600 cm-1 and 1500 cm-1 regions, assigned to the formation of Mg(OH)2 and adsorbed water and CO2 molecules in the humid environment (Figure 4e). Hence, the heat treatment at 1000 °C was essential to obtain MgO particles resistant to humidity and CO2. 3.4. LDPE/MgO composites Figures 6a–c shows scanning electron micrographs of composites based on the different MgO particles (3 wt.%). The composites which contained MgO particles prepared by the single-step heat treatment at 400 and 1000 °C (MgO-CD400 and MgO-CD1000) showed some agglomerates 1–3 µm in size (indicated by arrows in Figures 6a and b). It has previously been reported that the incompatibility between LDPE and hydrophilic surface of the MgO particles results in major aggregation.5 The composite based on the MgO 19 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 30

nanoparticles obtained by the modified two-step heat treatment (MgO-C18-1000) showed very uniform particle dispersion; the agglomerates always being smaller than 100 nm (Figure 6c). The shear force applied during extrusion was sufficient to break the MgO agglomerates into solitary particles when the silica coatings were present. The dispersion of the MgO particles in LDPE matrix was evaluated by measuring the interparticle distance which was further discussed in Section 3.5. In order to gain more information about the interfacial adhesion between nanoparticles and LDPE matrix, tensile tests were performed (Figure 6d). The addition of particles to LDPE had only a moderate impact on the Young’s modulus: 0.36 GPa (pristine LDPE), 0.38 GPa (composite based on MgOCD400), 0.33 GPa (composite based on MgO-CD1000) and 0.31 GPa (composites based on MgO-C18-1000). The strain-at-break of the composites based on the MgO-CD1000 particles (150 %) was lower than that of the pristine LDPE and other composites (300 – 400 %). The first yield for the composites occurred at lower stress value with reference to that of pristine LDPE. The first yield strain values were similar for the different nanocomposites (15 %). All the samples showed a drop in the stress-strain curves after the first yield point. The strain of the second yield point was 85 %, except for the composite based on MgO-CD1000 that showed a lower yield strain, 60 %. It was reported that the cavitation was observed around particles in the LDPE/Al2O3 nanocomposites at the second yield point.33,

45

Since the formation of cavities at the polymer/particle interface

indicates the weak interfacial adhesion between the MgO phase and LDPE matrix, the second yield point (which occurred simultaneously with cavitation) can be used as a method to assess the interfacial adhesion strength. Here, a higher cavitation strain (second yield strain value) was observed in LDPE nanocomposites based on MgO-CD400 and MgO-C18-1000 nanoparticles, which suggested that the interfacial adhesion between nanoparticles and LDPE matrix was strongest in these samples. The MgO-CD400 nanoparticles made micron-sized agglomerates in the LDPE matrix, but the interfacial surface area between the nanoparticles and LDPE matrix was large due to the much larger specific surface area than in the MgO-C18-1000 nanoparticles (Table 1). The partial silica coverage of the MgO-C18-1000 nanoparticles increased the hydrophobicity of the surface, and this enhanced both nanoparticles compatibility and dispersion within the LDPE matrix. 20 ACS Paragon Plus Environment

Page 21 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Figure 6 Scanning electron micrographs of cryo-fractured surfaces of LDPE composites filled with 3 wt. % of (a) MgO-CD400, (b) MgO-CD1000 and (c) MgO-C18-1000. The arrows showed the micron-sized agglomerates. (d) tensile stress-strain curves, (e) X-ray diffractograms and (f) DSC crystallization thermograms of pristine LDPE and its composites filled with 3 wt. % differently formed MgO particles.

Figures 6e and f shows the X-ray diffractograms and crystallization thermograms of the LDPE composites filled with 3 wt.% of MgO particles and of the unfilled LDPE. At room temperature, all the samples showed an orthorhombic crystalline phase with peaks at 21.3, 23.6 and 36.0° corresponding to the (110), (200) and (001) crystal planes together with an amorphous phase with a broad halo at 10-30° (Figure 6e).46 The crystallization thermograms showed two exothermic peaks; the most intense peak at 97 – 99 °C and a less intense peak at 62 °C. The important feature is highlighted in the inset graph of Fig. 6f: the onset of crystallization occurred at 107 °C for the composites based on MgOCD– 400 and MgO–C18–1000 and at a much lower temperature (102–103 °C) for pristine 21 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 30

LDPE and the composite based on MgOCD–1000. These data thus indicate that two of the nanofillers (MgOCD–400 and MgO–C18–1000) had a nucleating effect on crystallization. The nanofillers had essentially no significant impact on the temperature positions of the low and high temperature peaks. This is very expected, because the nucleating effect of the added nanoparticles is maximized in the absence of competing nucleating compounds such as the LDPE crystals themselves. In the melting thermogram (See supporting information, Fig. S2) there is only one peak (peak temperature = 111 °C) which was used for calculation of crystal thickness (Eq. 4). The LDPE retained an invariable crystal thickness (Lc) even after inclusion of the MgO nanoparticles (Table 2). The LDPE filled with MgO particles showed higher crystallization onset temperatures than unfilled LDPE (see inset graph of Figure 6f), indicating the nucleation effect of MgO particles on LDPE crystallization. This effect was more pronounced in the MgO samples with a large specific surface area (MgO-CD400 and MgO-C18-1000) due to the reduced free energy for nucleation at the melt/nanoparticle interface.47 These nanoparticles also showed the highest interfacial adhesion in the LDPE matrix. Table 2 Thermal characteristics of pristine LDPE and LDPE nanocomposites based on 3 wt. % MgO particles. Sample Tonset a Tm b wc c Lc d La e (°C)

(°C)

(nm)

(nm)

LDPE

106

111

0.43

8.5

13.5

LDPE/MgO-CD400

109

111

0.44

8.5

12.6

LDPE/MgO-CD1000

106

111

0.45

8.5

12.5

LDPE/MgO-C18-1000

109

111

0.46

8.5

11.9

a

The onset temperature for crystallisation during the cooling (1 °C min-1), Figure 6f (inset graph). The melting peak temperature during the second heating (10 °C min-1), see supporting information, Figure S2. c Mass crystallinity at 60 °C calculated from DSC data (see supporting information, Figure S2) using Eqs. (2) and (3). d Crystal thickness corresponding to the melting peak temperature calculated from DSC data using Eq. (4). e Amorphous thickness at 60 °C calculated according to Eq. (5). b

22 ACS Paragon Plus Environment

Page 23 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

3.5. DC conductivity of the composites The electrical insulation capacity of the LDPE composites was compared to that of unfilled LDPE (Figure 7). The charging current for the composite systems was always lower during the entire measurement than that for unfilled LDPE. The addition of 1 wt. % moisture-resistant MgO nanoparticles reduced the conductivity of the LDPE by ca. 30 times (1 min value), which is a smaller effect than was obtained by adding a similar amount of MgO-CD400 nanoparticles; the latter yielding a reduction in conductivity by ca. 60 times (Table 3). The conductivity-suppression (resistivity) level for all composite systems was decreased when the loading of MgO particles was increased from 1 to 3 wt.%. This behavior was more severe for the composites based on high-temperature heat-treated MgO particles (MgOCD1000 and MgO-C18-1000). It was stated in section 3.2 that the perfect and pure MgO crystals were obtained during high temperature heat treatment. It was reported that charge carriers in O- states were dissociated from O2- sublattice of MgO in the bulk and diffused to the surface during the heat treatment.47 After fast cooling of MgO particles to room temperature, the charge carriers mainly remained at cation vacancies on the crystal surfaces’ which increased the electrical conductivity of the MgO phase. On the other hand, the MgO particles in an LDPE matrix act as additional electron traps, and increase the tunneling barriers/distance for the transport of charge carriers thus reducing the charge mobility and DC conductivity within the LDPE matrix.6 According to the competition between these two phenomena, there is a limit to the loading of relatively conductive MgO particles in order to reduce the conductivity of pristine LDPE. This observation is in accordance with another study,48 that the conductivity of the nanocomposites increased to the level of LDPE after addition of 6–9 wt. % MgO nanoparticles which were heat-treated at 400 °C (MgO-CD400). This limit of loading MgO particles was herein shifted to the lower value (3 wt. %) in the case of MgO heattreated at 1000 °C (MgO-CD1000 and MgO-C18-1000).

23 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 30

Figure 7 Charging current of pristine LDPE and its composites based on different weight fractions of MgO particles obtained at 2.6 kV (E= 32.5 kV mm-1) at 60 °C.

The conductivity (charging current) of all the systems decreased with time by a powerlaw relationship, i(t) ∝ t-n (Figure 7). The power factor (n) was obtained from the slope of the fitted line of log i vs. log t. The unfilled LDPE showed a conductivity reduction with a single slope, whereas the composite systems showed a transient slope at 60 s (Table 3). Lau et al 49 related this transient behaviour to a change in the charge transport mechanism. The low power factor, n ≤ 1, indicated the formation of trapping sites for space charge carriers in the semi-crystalline LDPE matrix, whereas the high power factor, n ≥ 1, was consistent with a reduction in charge carrier mobility (tunneling and hopping) at the interface of nanofiller and polymer matrix.49 Since the semi-crystalline nature was retained after the addition of MgO particles (Figure 6e, Table 2), the conductivity suppression of the composite system was controlled only by the MgO particles/polymer interface. Power factors larger than unit were observed at regimes shorter than 60 s (denoted n1) for the composite systems, which indicated that the effect of MgO particles on the conductivity suppression occurred in this regime. The larger power factor for the nanocomposites based on MgO-CD400 suggested that the reduction in charge mobility and the trapping of charge carriers were more effective in the case of nanocomposites based on moisture-resistant MgO nanoparticles (MgO-C18-1000).

24 ACS Paragon Plus Environment

Page 25 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Table 3 Electrical characteristics of pristine LDPE and LDPE nanocomposites obtained at 2.6 kV (E= 32.5 kV mm-1) at 60 °C. Sample Conductivity (×10 -14 S m-1) a n1 b n2 b MgO fraction

1 wt.%

LDPE

3 wt.%

1 wt.%

69.6

3 wt.%

1 wt.%

0.5

3 wt.%

0.5

LDPE/MgO-CD400

1.1

1.7

2.2

2.2

0.5

0.5

LDPE/MgO-CD1000

3.1

10.0

2.2

1.5

0.6

0.3

LDPE/MgO-C18-1000

2.4

8.3

2.2

1.5

0.1

0.1

a

Electrical volume conductivity at 60 s. Power factor calculated from the slope of fitted lines in charging current-time curves at two regimes shorter (n1) and longer (n2) than 60 s (Figure 7). b

We have shown in another study that the reduction in LDPE conductivity was attributed to the dispersion state of particles in the LDPE matrix and to the surface reactivity of the MgO particles.48 The dispersion of the nanoparticles in the LDPE matrix was assessed by determining the mean centre-to-centre distance to the 51st MgO phase obtained from an examination of scanning electron micrographs (Figure 6 a–c). The details of the method have been presented by Pallon et al.48 The lower distance value for the nanocomposites based on moisture-resistant MgO nanoparticles (3678 nm) indicated a better particle dispersion of the moisture-resistant MgO nanoparticles than of the MgOCD400 nanoparticles in the LDPE matrix (5433 nm). This finding suggests that the surface activity was the dominant factor influenced the conductivity. Charge trapping on the nanoparticles with a large surface activity was facilitated due to the presence of surface defects.50 It should be noted that the specific surface area of MgO-CD400 was reduced 4.2 times by the modified two-step heat treatment. Therefore, the moistureresistant MgO nanoparticles in LDPE had higher DC conductivity due to the fact that the heat treatment at 1000 °C removed the concentration of favourable surface defects and reduced the surface area. However, this DC conductivity is still within the range of electrical conductivity needed for HVDC insulation cables. 4. Conclusions Highly insulating nanocomposites based on low-density polyethylene and moistureresistant MgO nanoparticles (≤ 3 wt.%) have been synthesized for potential use as an 25 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 30

insulating material in HVDC cables. The key to obtaining moisture-resistant MgO nanoparticles was a silane coating as an intermediate step after a low-temperature thermal decomposition of Mg(OH)2 prior to a high-temperature heat treatment. In this way, the specific surface area of Mg(OH)2 was successfully preserved after the high-temperature heat treatment. The moisture-resistant MgO nanoparticles retained their phase/structure even after extended exposure to a humid medium, which has not to the best of our knowledge been reported before. The LDPE nanocomposites based on the moistureresistant MgO nanoparticles showed high interfacial adhesion and well-dispersed particles compared to composites based on single-step heat-treated MgO particles. Although the moisture-resistant MgO nanoparticles had a low concentration of crystal surface defects, the addition of these nanoparticles (1 wt.%) into LDPE matrix resulted in a significant reduction in the electrical conductivity. These nanoparticles thus showed their high potential in ultra-low electrically conductive nanocomposites where the absence of the adsorbed water/hydroxide species is of paramount importance. Acknowledgements The financial support from the Swedish Foundation for Strategic Research (SSF; grant EM11-0022) is gratefully acknowledged. Borealis AB, Stenungsund, Sweden is acknowledged for supplying the low-density polyethylene.

References: 1. Muller, B.; Arlt, W.; Wasserscheid, P., A new concept for the global distribution of solar energy: energy carrying compounds. Energy Environ. Sci. 2011, 4 (10), 4322-4331. 2. van der Hoeven, M., Energy Technology Perspectives 2014: Harnessing Electricity's Potential. International Energy Agency: 2014. 3. Gustafsson, A.; Saltzer, M.; Farkas, A.; Ghorbani, H.; Quist, T.; Jeroense, M. The new 525 kV extruded HVDC cable system; ABB Grid Systems: 2014. 4. Sellerholm, H. The megavolt challenge; 14:21; Elforsk AB: Stockholm, 2014. 5. Pallon, L. K. H.; Olsson, R. T.; Liu, D.; Pourrahimi, A. M.; Hedenqvist, M. S.; Hoang, A. T.; Gubanski, S.; Gedde, U. W., Formation and the structure of freeze-dried MgO nanoparticle foams and their electrical behaviour in polyethylene. J. Mater. Chem. A 2015, 3 (14), 7523-7534. 6. Lewis, T. J., Charge transport in polyethylene nano dielectrics. IEEE Trans. Dielectr. Electr. Insul. 2014, 21 (2), 497-502. 7. Simin, P.; Jinliang, H.; Jun, H.; Xingyi, H.; Pingkai, J., Influence of functionalized MgO nanoparticles on electrical properties of polyethylene nanocomposites. IEEE Trans. Dielectr. Electr. Insul. 2015, 22 (3), 1512-1519. 26 ACS Paragon Plus Environment

Page 27 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

8. Yang, J.; Wang, X.; Zhao, H.; Zhang, W.; Xu, M., Influence of moisture absorption on the DC conduction and space charge property of MgO/LDPE nanocomposite. IEEE Trans. Dielectr. Electr. Insul. 2014, 21 (4), 1957-1964. 9. Takada, T.; Hayase, Y.; Tanaka, Y., Space charge trapping in electrical potential well caused by permanent and induced dipoles for LDPE/MgO nanocomposite. IEEE Trans. Dielectr. Electr. Insul. 2008, 15 (1), 152-160. 10. Tanaka, T.; Kozako, M.; Fuse, N.; Ohki, Y., Proposal of a multi-core model for polymer nanocomposite dielectrics. IEEE Trans. Dielectr. Electr. Insul. 2005, 12 (4), 669-681. 11. Zhang, L.; Zhou, Y.; Tian, J.; Sha, Y.; Zhang, Y.; Wu, H.; Wang, Y., Experiment and simulation of space charge suppression in LDPE/MgO nanocomposite under external DC electric field. J. Electrostatics 2014, 72 (4), 252-260. 12. Fillery, S. P.; Koerner, H.; Drummy, L.; Dunkerley, E.; Durstock, M. F.; Schmidt, D. F.; Vaia, R. A., Nanolaminates: Increasing Dielectric Breakdown Strength of Composites. ACS Appl. Mater. Interfaces 2012, 4 (3), 1388-1396. 13. Naono, H., International Symposium on AdsorptionMicropore formation due to thermal decomposition of magnesium hydroxide. Colloids Surf. 1989, 37, 55-70. 14. Yu, J. C.; Xu, A.; Zhang, L.; Song, R.; Wu, L., Synthesis and Characterization of Porous Magnesium Hydroxide and Oxide Nanoplates. J. Phys. Chem. B 2004, 108 (1), 64-70. 15. Green, J., Calcination of precipitated Mg(OH)2 to active MgO in the production of refractory and chemical grade MgO. J. Mater. Sci. 1983, 18 (3), 637-651. 16. Itatani, K.; Koizumi, K.; Howell, F. S.; Kishioka, A.; Kinoshita, M., Agglomeration of magnesium oxide particles formed by the decomposition of magnesium hydroxide. J. Mater. Sci. 1988, 23 (9), 3405-3412. 17. Mironyuk, I. F.; Gun’ko, V. M.; Povazhnyak, M. O.; Zarko, V. I.; Chelyadin, V. M.; Leboda, R.; Skubiszewska-Zięba, J.; Janusz, W., Magnesia formed on calcination of Mg(OH)2 prepared from natural bischofite. Appl. Surf. Sci. 2006, 252 (12), 4071-4082. 18. Hui, L.; Schadler, L. S.; Nelson, J., The influence of moisture on the electrical properties of crosslinked polyethylene/silica nanocomposites. IEEE Trans. Dielectr. Electr. Insul. 2013, 20 (2), 641653. 19. Delplancke-Ogletree, M. P.; Ye, M.; Winand, R.; de Marneffe, J. F.; Deltour, R., Influences of thermal annealing and humidity exposure on surface structure of (100) single-crystal MgO substrate. J. Mater. Res. 1999, 14 (05), 2133-2137. 20. Utamapanya, S.; Klabunde, K. J.; Schlup, J. R., Nanoscale metal oxide particles/clusters as chemical reagents. Synthesis and properties of ultrahigh surface area magnesium hydroxide and magnesium oxide. Chem. Mater. 1991, 3 (1), 175-181. 21. Langel, W.; Parrinello, M., Hydrolysis at stepped MgO surfaces. Phys. Rev. Lett. 1994, 73 (3), 504. 22. Shin, H.-J.; Jung, J.; Motobayashi, K.; Yanagisawa, S.; Morikawa, Y.; Kim, Y.; Kawai, M., Stateselective dissociation of a single water molecule on an ultrathin MgO film. Nat. Mater. 2010, 9 (5), 442447. 23. Geysermans, P.; Finocchi, F.; Goniakowski, J.; Hacquart, R.; Jupille, J., Combination of (100), (110) and (111) facets in MgO crystals shapes from dry to wet environment. PCCP 2009, 11 (13), 22282233. 24. Lee, M. H.; Park, D. G., Preparation of MgO with high surface area, and modification of its pore characteristics. Bull. Korean Chem. Soc. 2003, 24 (10), 1437-1443. 25. Fabiani, D.; Montanari, G. C.; Testa, L., Effect of aspect ratio and water contamination on the electric properties of nanostructured insulating materials. IEEE Trans. Dielectr. Electr. Insul. 2010, 17 (1), 221-230. 26. Javerberg, N.; Edin, H.; Nordell, P.; Nawaz, S.; Hillborg, H.; Azhdar, B.; Gedde, U. W., Dielectric properties of alumina-filled poly (ethylene-co-butyl acrylate) nanocomposites Part II- wet studies. IEEE Trans. Dielectr. Electr. Insul. 2012, 19 (2), 391-399. 27 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 30

27. Degen, A.; Kosec, M., Effect of pH and impurities on the surface charge of zinc oxide in aqueous solution. J. Eur. Ceram. Soc. 2000, 20 (6), 667-673. 28. Pourrahimi, A. M.; Liu, D.; Ström, V.; Hedenqvist, M. S.; Olsson, R. T.; Gedde, U. W., Heat treatment of ZnO nanoparticles: new methods to achieve high-purity nanoparticles for high-voltage applications. J. Mater. Chem. A 2015, 3 (33), 17190-17200. 29. Piao, Y.; Kim, J.; Na, H. B.; Kim, D.; Baek, J. S.; Ko, M. K.; Lee, J. H.; Shokouhimehr, M.; Hyeon, T., Wrap-bake-peel process for nanostructural transformation from [beta]-FeOOH nanorods to biocompatible iron oxide nanocapsules. Nat. Mater. 2008, 7 (3), 242-247. 30. Liu, D.; Pourrahimi, A. M.; Pallon, L. K. H.; Andersson, R. L.; Hedenqvist, M. S.; Gedde, U. W.; Olsson, R. T., Morphology and properties of silica-based coatings with different functionalities for Fe3O4, ZnO and Al2O3 nanoparticles. RSC Adv. 2015, 5 (59), 48094-48103. 31. Pourrahimi, A. M.; Liu, D.; Pallon, L. K. H.; Andersson, R. L.; Martinez Abad, A.; Lagaron, J. M.; Hedenqvist, M. S.; Strom, V.; Gedde, U. W.; Olsson, R. T., Water-based synthesis and cleaning methods for high purity ZnO nanoparticles - comparing acetate, chloride, sulphate and nitrate zinc salt precursors. RSC Adv. 2014, 4 (67), 35568-35577. 32. Gray, A. P., Polymer crystallinity determinations by DSC. Thermochim. Acta 1970, 1 (6), 563579. 33. Liu, D.; Pourrahimi, A. M.; Olsson, R. T.; Hedenqvist, M. S.; Gedde, U. W., Influence of nanoparticle surface treatment on particle dispersion and interfacial adhesion in low-density polyethylene/aluminium oxide nanocomposites. Eur. Polym. J. 2015, 66 (0), 67-77. 34. Wunderlich, B.; Baur, H., Heat capacities of linear high polymers. Springer: 1970. 35. Gedde, U. W., Polymer Physics, 1995. Chapman & Hall, London. 36. Gaur, U.; Wunderlich, B., Heat capacity and other thermodynamic properties of linear macromolecules. II. Polyethylene. J. Phys. Chem. Ref. Data 1981, 10 (1), 119-152. 37. Huseby, T. W.; Bair, H. E., Dissolution of Polyethylene Single Crystals in Xylene and Octadecane. J. Appl. Phys. 1968, 39 (11), 4969-4973. 38. Gregg, S. J.; Sing, K. S. W., Adsorption, surface area, and porosity. Academic Press: 1991. 39. Tian, P.; Han, X.-y.; Ning, G.-l.; Fang, H.-x.; Ye, J.-w.; Gong, W.-t.; Lin, Y., Synthesis of Porous Hierarchical MgO and Its Superb Adsorption Properties. ACS Appl. Mater. Interfaces 2013, 5 (23), 1241112418. 40. Grill, A., Porous pSiCOH ultralow-k dielectrics for chip interconnects prepared by PECVD. Annu. Rev. Mater. Res. 2009, 39, 49-69. 41. Suryanarayana, C.; Norton, M. G., X-ray diffraction: a practical approach. Springer: 1998. 42. Sterrer, M.; Berger, T.; Diwald, O.; Knözinger, E., Energy Transfer on the MgO Surface, Monitored by UV− Induced H2 Chemisorption. J. Am. Chem. Soc. 2003, 125 (1), 195-199. 43. Blesa, M.; Weisz, A.; Morando, P.; Salfity, J.; Magaz, G.; Regazzoni, A., The interaction of metal oxide surfaces with complexing agents dissolved in water. Coord. Chem. Rev. 2000, 196 (1), 31-63. 44. Cornu, D.; Guesmi, H.; Krafft, J.-M.; Lauron-Pernot, H., Lewis Acido-Basic Interactions between CO2 and MgO Surface: DFT and DRIFT Approaches. J. Phys. Chem. C 2012, 116 (11), 6645-6654. 45. Cobo Sánchez, C.; Wåhlander, M.; Taylor, N.; Fogelström, L.; Malmström, E., Novel Nanocomposites of Poly(lauryl methacrylate)-Grafted Al2O3 Nanoparticles in LDPE. ACS Appl. Mater. Interfaces 2015, 7 (46), 25669-25678. 46. Watanabe, S.; Sano, N.; Noda, I.; Ozaki, Y., Surface Melting and Lamella Rearrangement Process in Linear Low Density Polyethylene. J. Phys. Chem. B 2009, 113 (11), 3385-3394. 47. Freund, F.; Freund, M. M.; Batllo, F., Critical review of electrical conductivity measurements and charge distribution analysis of magnesium oxide. J. Geophys. Res.: Solid Earth 1993, 98 (B12), 2220922229. 48. Pallon, L. K. H.; Hoang, T. A.; Pourrahimi, A. M.; Hedenqvist, M. S.; Nilsson, F.; Gubanski, S.; Gedde, U.; Olsson, R. T., The impact of MgO nanoparticle interface in ultra insulating polyethylene nanocomposites for high voltage DC cables. J. Mater. Chem. A 2016.

28 ACS Paragon Plus Environment

Page 29 of 30

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

49. Lau, K. Y.; Vaughan, A. S.; Chen, G.; Hosier, I. L.; Holt, A. F., Absorption current behaviour of polyethylene/silica nanocomposites. J. Phys.: Conf. Ser. 2013, 472 (1), 012003. 50. König, T.; Simon, G. H.; Rust, H. P.; Pacchioni, G.; Heyde, M.; Freund, H. J., Measuring the Charge State of Point Defects on MgO/Ag(001). J. Am. Chem. Soc. 2009, 131 (48), 17544-17545.

29 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 30

The modified two-step heat treatment of Mg(OH)2 in order to obtain moisture-resistant MgO nanoparticles with preserved specific surface area aimed for ultra-low conductivity LDPE nanocomposites in HVDC cable insulations.

30 ACS Paragon Plus Environment