Loss and Recovery of Hydrophobicity of Polydimethylsiloxane after Exposure to Electrical Discharges

TABLE OF CONTENTS

1. INTRODUCTION
1.1 Purpose of this study
1.2 The author's contribution

2. BACKGROUND
2.1 Properties of polydimethylsiloxane
2.1.1 General description
2.1.2 Production of the polymer
2.1.3 Production of a crosslinked network
2.1.4 Rubber formulations for high voltage outdoor insulation
2.2 Silicon rubber as outdoor high voltage insulation
2.2.1 General applications
2.2.2 Log-term Performance
2.3 Loss and recovery of hydrophobicity-electrical discharges
2.3.1 Loss of hydrophobicity
2.3.2 Hydrophobic recovery
2.4 Thermal degradation of polydimethylsiloxane
2.5 Influence of environmental factors on hydrophobicity
2.5.1 Water immersion
2.5.2 UV radiation

3. EXPERIMENT AL
3.1 Materials
3.1.1 HTV silicone rubber (Paper I)
3.1.2 'Exact' PDMS networks (Papers II-IV, VI)
3.1.3 Field aged silicone rubber formulation (Paper VII)
3.2 Conditions during exposure to electrical discharges
3.2.1 Corona discharges
3.2.2 Radio-frequency plasma (RF)
3.2.3 Microwave plasma (MW)
3.3 Characterisation techniques
3.3.1 Contact angle measurements
3.3.2 Wilhelmy balance
3.3.3 Reflection infrared spectroscopy (IR)
3.3.4 X-ray photoelectron spectroscopy (XPS)
3.3.5 Scanning electron microscopy (SEM)
3.3.6 Atomic force microscopy (AFM)
3.3.7 Neutron reflectivity measurements
3.3.8 Optical microscopy
3.3.9 Size exclusion chromatography (SEC)
3.3.10 Gas chromatography-mass spectrometry (GC-MS)

4. EXPOSURE TO ELECTRICAL DISCHARGES
4.1 HTV silicone rubber: corona (Paper I)
4.2 'Exact' PDMS networks: corona and air-plasma (MW) (Paper III)
4.3 'Exact' PDMS networks: hydrophobic long term
4.4 'Exact' PDMS networks: oxygen plasma (RF and MW) (Paper II)
4.5 Characterisation of migrating low molar mass PDMS (Paper IV)

5. FIELD AGING OF AC/DC ENERGISED HTV SILICONE RUBBER FORMULATIONS (Paper VII)
5.1 Background to the field test
5.2 Field observations and electric measurements
5.3 Surface characterisation
5.3.1 Scanning electron microscopy
5.3.2 Reflection infrared spectroscopy
5.3.3 X-ray photoelectron spectroscopy
5.4 Characterisation of siloxanes within the pollution layer
5.4.1 Size exclusion chromatography
5.4.2 Gas chromatography-mass spectrometry
5.5 Mechanisms of surface ageing

6. CONCLUSIONS
6.1 Accelerated testing using corona or plasma
6.1.1 Oxidised surface layers
6.1.2 Low molar mass siloxanes
6.2 Field ageing using electric stress as accelerating factor
6.3 'Life-time' of hydrophobic recovery
6.4 Regeneration of low molar mass PDMS

7. ACKNOWLEDGEMENTS

8. REFERENCES

1. INTRODUCTION

1.1 Purpose of this study

This thesis deals with the loss and recovery of hydrophobicity of silicone rubber surfaces after exposure to electrical discharges. The project was directed to form a link between electrical engineering and polymer science, focusing on the understanding of the mechanisms of deterioration of silicone rubber used as high voltage outdoor insulation.

Silicone rubber is currently replacing glass or porcelain as outdoor insulation. One of the benefits is the hydrophobic surface properties of polydimethylsiloxane (PDMS); the polymer used in these silicone rubber formulations. On glass and porcelain insulation, water readily forms a continuous film on the hydrophilic surface. In the presence of contamination, leakage currents develop, which may result in a flashover of the insulator. The hydrophobic surface of PDMS prevents the formation of such continuous water films. If the polymer is exposed to electrical discharges this can cause a loss of hydrophobicity due to surface oxidation. PDMS, however, has the unique ability to recover hydrophobicity after the exposure. When the project was initiated, a hypothesis of the lifetime of the hydrophobic recovery of PDMS was formulated: the intrinsic surface properties of PDMS are essentially preserved during Period 1 (Fig. 1.1). After a certain incubation time (or dose of energetic species) a glassy, amorphous skin is formed which severely retards the rate of hydrophobic recovery. The 'end of the life of hydrophobicity' is then reached (Period 2). The goal of the project was to test the hypothesis that the lifetime of hydrophobic recovery of PDMS could be divided into two different periods.

illustration not visible in this excerpt

Figure 1.1 Schematic representation of the time-dependence of hydrophobic recovery rate of two different silicone rubber formulations.

In order to answer the question, the main mechanisms for the deterioration of PDMS used for outdoor insulation must be established. This would allow the development of proper evaluation techniques of silicone rubber outdoor insulation. It was decided to use electric discharges (corona, air-plasma) as accelerated testing of both commercial silicone rubbers as well as 'exact' PDMS networks.

The thesis is organised as follows:

Chapter 2 starts with a general description of PDMS. This is followed by a description of silicone rubbers in outdoor insulation applications. Furthermore, the mechanisms for the loss and recovery of the hydrophobicity of PDMS is reviewed.

In Chapter 3 the experimental methods are described.

Chapter 4 describes the surface properties of PDMS after exposure to electrical discharges. The rate of hydrophobic recovery is correlated to the properties of oxidised, 'silica-like', surface layers. The hypothesis of lifetime of hydrophobic recovery is tested. Finally the nature of the diffusing low molar mass siloxanes is analysed and their role in the hydrophobic recovery is discussed.

In Chapter 5 chemical analysis of field- tested silicone rubbers, aged under alternating current (AC) or direct current (DC) voltage, is presented and related to the results obtained from the laboratory tests (Ch. 4). Chapter 6 summarises the results, including the properties of the 'silica- like' surface layer on hydrophobic recovery rate. The difference in deterioration mechanism between corona/air-plasma exposed and field- aged PDMS is discussed.

Chapter 7 discusses the practical implications of the results on the long- term properties hydrophobicity; a revised lifetime model of hydrophobicity upon exposure to corona is suggested.

Most of the experimental details are not included in Chapters 4 and 5, but can be found in the appended papers.

1.2 The author's contribution

The author was responsible for and carried out the major part of the work on papers I and III-VI.

In paper II the author performed contact angle measurements, SEM and analysis of XPS data, summarised the results and wrote the paper.

Tomas Gustavsson was responsible for paper VII. The author performed the major part of the chemical analysis of the field-aged samples and assisted in the analysis of the data and the writing of the paper.

The entire work was supervised by Ulf W. Gedde.

2. BACKGROUND

In this chapter a general description of silicone rubber, based on Polydimethylsiloxane, and its use as high voltage outdoor insulation is presented. The mechanisms responsible for the loss and recovery of hydrophobicity of silicone rubbers are summarised. The presentation is not intended to be extensive, but rather wishes to illustrate the width of the subject. The aim is to familiarise the reader with some of the chemical and physical aspects of hydrophobicity loss and recovery of silicone rubbers and its relevance to high voltage outdoor insulation applications.

2.1 Properties of polydimethylsiloxane

2.1.1 General description

Polydimethylsiloxane (PDMS) is the polymer commonly used in silicone rubber formulations for high voltage outdoor insulation applications. PDMS consists of an inorganic backbone of alternating Silicon and oxygen atoms. Methyl groups are attached to the Silicon atoms forming the repeating unit in the polymer (Fig. 2.1).

illustration not visible in this excerpt

Figure 2.1 Repeating unit of PDMS.

The low surface free energy - 16-21 mN m"1; the precise value depends on the molar mass,1'2 is due to closely packed methyl groups in the surface. Four structural characteristics of PDMS account for its surface properties: (1) the low intermolecular force between methyl groups, (2) the unique flexibility of the siloxane backbone (Si-O-Si: 143°), (3) the high strength of the siloxane bond and (4) the partial ionic nature of the siloxane bond1. The Pauling electronegativity difference of 1.7 between Silicon and oxygen results in a 40-50% polar character of the siloxane bond3. The polar nature of the siloxane bond makes, however, PDMS susceptible to hydrolysis, especially during acidic or basic conditions. The positively polarised Silicon atom is electron withdrawing, thus polarising the methyl group and makes it less susceptible to radical attack. Thus the methyl groups in PDMS have higher thermal and oxidative stability compared with a methyl group within a hydrocarbon polymer (e.g. polypropylene)1. Another reason for the excellent thermal stability of PDMS is the high dissociation energy of the siloxane bond (445 kJ mol"1)4. The energy barriers for torsion around the main chain bonds are very low compared to that of other polymers: < 4 kJ mol-1 for PDMS5, compared to 15 kJ mol"1 for Polyethylene6. The segmental mobility and the high free volume of PDMS is reflected in its low glass transition temperature, - 127°C7 and very high permeability and diffusivity to gases. The oxygen permeability at room temperature is 21 times greater for PDMS than for natural rubber and 170 times greater than for low-density Polyethylene8. The regulär chain structure makes PDMS a crystallisable polymer. The equilibrium melting point is - 54°C and the enthalpy of fusion is very low, 2.75 kJ/(mole repeating unit)9. The entropy of fusion is 6.28 J/(mole of flexible main chain bonds) which is lower than for Polyethylene (9.9 J/(mole of flexible main chain bonds)10. Hence, the low melting point of PDMS is due to its very low enthalpy of fusion.

2.1.2 Production of the polymer

Silicon dioxide (Si02) is reduced to Silicon by reaction of a carbon electrode. A SiO intermediate is formed, followed by the production of a chemical grade of powdered Silicon. Linear PDMS is usually manufactured from dimethyldichlorosilane, which is produced by the reaction of powdered Silicon with methyl chloride. A mixture of cyclic (c) and linear (l) PDMS oligomers is obtained by hydrolysis of the dichlorosilane followed by condensation reactions (Reaction 1):

illustration not visible in this excerpt

High molar mass polymers are then produced by anionic or cationic ring- opening polymerisation of the cyclic oligomers or by polycondensation of the silanol end-blocked linear oligomers11.

2.1.3 Production of a crosslinked network High temperature vulcanisation

In high temperature vulcanising (HTV) silicone rubbers crosslinking of the polymer chains takes place through decomposition of peroxides at temperatures above 100°C. The peroxides decompose into free radicals which react with unsaturated bonds and/or by abstraction of hydrogen atoms, depending on type of peroxide used. The free radicals then recombine and form crosslinks between the siloxane chains. Residual volatile decomposition products within the HTV silicone rubbers are generally removed by a post—curing step (storage at elevated temperatures).

Room temperature vulcanisation

Two different crosslinking methods are generally used for room temperature vulcanising (RTV) silicone rubbers for outdoor insulation applications. One uses a condensation reaction of silanol groups to form siloxane bonds with the liberation of water (Reaction 2). The reaction involves water and is an equilibrium process, catalysed by acid or base.

illustration not visible in this excerpt

The hydrosilylation reaction involves the addition of a Silicon hydrogen (Si-H) to an unsaturated carbon bond, usually a vinyl group (-CH=CH2), catalysed by a noble metal, e. g. a platinum complex (.Reaction 3).

illustration not visible in this excerpt

This second reaction is very specific and the crosslink density can be controlled very accurately by this method.

2.1.4 Rubber formulations for high voltage outdoor insulation

To the uncrosslinked PDMS, reinforcing silane-treated fillers are added, e.g. 10-20% of amorphous silica. Silica can also be used for rheological control of the Compound. Typically -50% aluminum trihydrate (ATH) is added as a flame-retardant because unfilled PDMS is too flammable for these applications. ATH decomposes to aluminium oxide and water when heated to temperatures above 200°C. The liberation of water is endothermic and the surface is cooled, which may extinguish an electric arc (Reaction 4). Furthermore, the resistance against tracking ('electrical erosion') is improved.

illustration not visible in this excerpt

Typical formulations also contain smaller proportions of silicone oil for process control, pigments, and chemicals used for vulcanisation (cross­linking) reactions.

2.2 Silicone rubber as outdoor high voltage insulation

2.2.1 General applications

Silicone rubbers have, since their introduction in the 1960's, steadily gained market share from porcelain and glass as outdoor high voltage insulation {e.g. transmission line insulators, bushings, surge arresters and cable terminations). The silicone rubbers are used in composite insulators as housing material on a load-bearing core, e.g. glass-fibre-reinforced plastics. An example of a composite insulator (line insulator type) is shown in Fig. 2.2a. The main functions of the silicone rubber housing are to protect the core from the outdoor environment and to provide a suitable profile for the insulator, i.e. minimise leakage currents between the energised end and ground. An example of installed line insulators is shown in Fig. 2.2b.

Silicone rubbers have a property profile — hydrophobic surface properties (water repellence), low surface and bulk conductivity, fracture toughness over a wide temperature ränge - which gives them distinct advantages compared to ceramic materials. Low weight, vandalism resistance and contamination resistance are other important characteristics of composite insulators with silicone rubber housings12'13. An alternative rubber material used for housings is EPDM (Ethylene- Propylene-Diene Monomer).

illustration not visible in this excerpt

However, silicone rubber insulators have been reported to perform better compared to porcelain, glass and EPDM both in heavily polluted areas and in severe marine sites14"18 and also during laboratory tests19'21. Using porcelain or glass high voltage insulators, water readily forms a continuous film on the hydrophilic surface. In the presence of severe contamination, leakage currents develop, which may result in a flashover of the insulator. The initially hydrophobic surface properties of EPDM are permanently lost when exposed to electrical discharges or pollution22. The hydrophobic surface properties of silicone rubber prevents the formation of continuous water films, instead water droplets are formed which simply bead off the surface. This hydrophobicity can, however, be temporarily lost during exposure to electrical discharges (corona, dry band arcing)22"24, or by rapid build-up of a continuos layer of pollution, i.e. salt or dust25"28. After a subsequent period of rest, hydrophobicity is, however, regained. Silicone rubbers thus have the ability to recover hydrophobicity. This will be further discussed in Section 2.3-2.4.

2.2.2 Long-term Performance

A central question to assess is the long-term Performance of composite insulators. Currently there are no well-defined specifications for accelerated ageing of composite polymer insulators or polymeric insulating materials29. Several national and international organisations attempt to develop such Standards. These include IEEE, IEC, CIGRE, American National Standards Institute (ANSI) and National Electric Manufacturers Association (NEMA)29. However, most of the existing tests for accelerated ageing are primarily useful for ranking of materials, thus only tests in field stations and actual Performance on power lines and outdoor substations can yield realistic results on outdoor Service Performance29. Different electric tests of composite insulators and materials under both field and laboratory conditions have recently been reviewed by Fernando and Gubanski30, Bärch et al.31 and Sebo and Zhao32.

A central concern of polymeric composite insulators is the surface degradation of the housing, which may decrease its ability to prevent water ingress to the glass-fibre reinforced core. In the case of silicone rubber this question is closely correlated to the surface hydrophobicity. In 1974 Niemi and Orbeck suggested that a failure of polymeric insulators was a result of progressive tracking or by direct flashovers, initiated by large leakage currents and dry band arcing33. This mechanism was further developed by Gorur et al. who suggested that ageing of silicone rubber housings in an outdoor environment started with the loss of hydrophobicity due to dry band arcing34. The arcing erodes the surface, by depolymerisation and clustering of the exposed filier particles, which eventually lead to initiation of tracking and a subsequent failure of the insulator34.

However, dry band arcing only occurs when the hydrophobicity is reduced, since a hydrophobic surface will have no leakage currents. Thus on a hydrophobic surface, such as a silicone rubber, corona was suggested to be the initiating mechanism for electric ageing35. Corona and dry band arcing are two very different phenomena, since corona is initiated by high electric fields, whereas dry band arcing is related to leakage currents. High electric fields, exceeding the ionisation level of air, can be caused by field enhancement by water droplets (water droplet induced corona) or by improper design of the geometry of the insulator. The effects of corona have been verified by observing surface degradation of silicone rubber housings before the detection of any leakage current pulses exceeding 0.5 mA36. It has been shown that surface erosion can be avoided when surfaces parallel to the electric field are not exposed to fields higher than 0.4-0.6 kV/mm36,37. This can be achieved by modifying the field distribution, i.e. the geometry of the insulator or by using Corona rings. If surface hydrophobicity is lost due to the corona, water forms a conducting film by dissolving the pollution always present in an outdoor environment, followed by localised drying and subsequent dry band arcing. This will cause erosion/ablation of the silicone rubber surface. The processes will continue during the wet period, but cease upon drying. Hydrophobic recovery of the silicone rubber will then occur. This cycle will be repeated during wet periods and finally a flashover of the composite insulator may

illustration not visible in this excerpt

Figure 2.3 (a) Illustration of filament formation and spot discharges on a hydrophobic silicone rubber insulator surface40, (b) Water filaments on an aged silicone rubber insulator41.

Shah et al. proposed a flashover mechanism initiated by interactions between water droplets and contamination on the insulator surface, forming conductive hydrophilic regions39,40. Low leakage currents cause an ohmic (resistive) heating of the surface between these regions, making them to coalesce into larger conductive water filaments. Spot discharges (self-quenching discharges) between conductive filaments are then initiated by enhancement of the electric field. This finally led to insulator failure due to flashover along the wetted filaments4The mechanism is illustrated in Fig. 2.3a. An example of an aged silicone rubber insulator with inhomogeneous hydrophobicity is shown in Fig. 2.3b. High voltage outdoor insulation using silicone rubber housings has been used for over 2o years12. However, glass and porcelain is often preferred as a first choice, due to the limited knowledge of the long-term Performance and life expectancy of silicone rubber housings15'42.

2.3 Loss and recovery of hydrophobicity: electrical discharges

The loss and recovery of hydrophobicity caused by exposing silicone rubber to electrical discharges (corona or plasma) have been investigated by many researchers. Plasma treatments have been used to increase the wettability of silicone rubber for improved compatibility to other materials, e.g, in biomedical applications and in printing technology. In these fields the hydrophobic recovery is considered as a problem and much effort has been directed to reveal the fundamental underlying mechanisms. The complexity of corona/plasma exposure is partly due to the fact that the polymer is simultaneously subjected to a mixture of energetic species and radiation, e.g. electrons, ions, UV and ozone. A great number of reactions take place. In addition, the effects of the treatments are highly dependent on the material structure and composition. The main effects of corona/plasma treatment on silicone rubber can be summarised as follows: (1) an increase of the oxygen content at the surface by the formation of silanol and carbonyl groups (2) oxidative crosslinking, (3) degradation of the network structure resulting in the formation of low molar mass cyclic Compounds and medium to high molar mass PDMS. In order for oxidation reactions to occur oxygen needs to be present, either in the surrounding atmosphere, or dissolved in the polymer.

2.3.1 Loss of hydrophobicity

The hydrophobic surface properties of silicone rubber have been shown to be temporary lost by surface oxidation during exposure to corona discharges in air22"24'43'44, radio frequency (RF) and microwave (MW) plasma treatments45"51 or salt-fog tests52"55. The oxidation resulted in the formation of an inorganic, silica-like (SiOx) structure, i.e. Silicon atoms bonded to more than two oxygen atoms23'24'53'56'. Polar silanol groups were also introduced in the surface region49'57. Both of these formed structures increase the hydrophilicity of the surface. The mechanisms are briefly summarised below.

[...]