skip to content

Composites and Coatings Group

Department of Materials Science & Metallurgy

Plasma electrolytic oxidation (PEO) is a surface modification process which confers wear and corrosion resistance to light alloys such as aluminium, magnesium and titanium [1]. PEO has been a commercial process for a number of years, developed extensively by Keronite Ltd. However, much of the commercial research is driven by customer needs, with the process being optimised on a case by case basis.

This project is aimed at understanding the fundamental processes involved in PEO, with a view to improve the industrial process. Understanding the process further will allow a degree of control over the microstructure and properties of coatings, and potentially improve the efficiency of the process. Currently the process is very energy intensive, limiting its widespread use. Oxidation is an exothermic process, so should liberate energy on conversion of the substrate metal to its oxide. However, PEO requires large amounts of energy to drive the formation of plasma discharges. The formation of such plasmas is thought to provide mechanism for exposing the substrate metal to oxygen in some form. The plasmas are short lived and the heat generated by the plasmas is quickly dissipated in the electrolyte and wasted.

Previous work in the group has developed methodologies for monitoring the electrical characteristics of discharges during PEO. [2-4] This has been coupled with optical emission spectroscopy (OES) to determine the species present, the electron densities and the temperatures in the plasmas.

This project will involve characterisation of individual discharges by high speed photography with time resolutions of ~5 microseconds, see Figure 1 [5]. Work has shown that discharges occur in cascades in well-defined physical locations, and extend over several periods of the applied voltage. Figure 1(a) shows a superimposition of all the frames captured in a 100 ms of video. There is just one very bright spot, corresponding to a discharge location. Figure 1(b) shows the integration of the light intensity for each frame as a function of time - showing that there are multiple peaks in light intensity due to discharge events over at least five periods of the applied voltage - and Figure 1(c) is the same as (b) but summed only from a circle around the bright spot in (a). This confirms that all the discharges did occur just at this location. Figure 1(d) shows a higher resolution plot of (c), showing multiple discharge events at the same location within one anodic half-period, with a typical voltage profile superimposed. These results are important in understanding the mechanisms by which discharges occur.


Cascade figure 

Figure 1. Information relating to a sequence of 17,500 images (exposure time ~ 5.5 μs, pixel size 12 μm) from a PEO coating ~ 50 μm in thickness, showing (a) a superimposition of the complete sequence, (b) total summed light intensity for each frame, as a function of time, (c) as for (b), but taken only from the bright area (circle of 288 μm diameter) in (a), and (d) higher resolution plot of part of (c), approximately covering an anodic half-cycle, with a typical measured voltage profile during such a period superimposed.


High speed camera video has been synchronised with small area electrical monitoring - this allows us to make correlations between discharge electrical profiles and the optical characteristics of discharges. Strong illumination has been used in order to see the dynamics of the gas bubble generated by a discharge. This information, combined with electrical data, allowed a semi-quantitative energy audit of an individual discharge to be carried out. A schematic representation of this is shown in Figure 2. The energy required to generate the plasma is actually only a fraction of the input electrical energy, and we concluded that the main energy absorption mechanism was actually the transient vaporisation of electrolyte surrounding the discharge site. When the discharge current stops the gas bubble quenches and heats up the surrounding electrolyte. Further information on this work can be found in [6].  


Figure 2. Semi-quantitative plot showing how the total (electrically-injected) energy changes during formation of the individual discharge under consideration here, and how it is converted between different forms during and immediately after the discharge period.


Frequency effects of PEO processing under current controlled conditions have also been investigated. Microstructural effects of discharge cascades were investigated by applying short (1 s) PEO treatments to small samples (1 mm diameter) which had PEO coating pre-deposited on them. The initial coatings were imaged in the SEM, before being PEO processed again with electrical monitoring and high speed video monitoring. The only microstructural changes occurred in regions of light emission. A discharge channel became visible, surrounded by an approximately circular region (~100 μm diameter) in which there were significant changes. These changes are highlighted in Figure 3 and 4 by red dotted circles for 50 Hz and 2500 Hz respectively. Further details can be found in [7]. 



Figure 3. SEM micrographs of a small area sample, with 100 μm PEO coating, showing (a) complete surface in initial state, (b) magnified top central region, (c) complete surface after further PEO processing at 50 Hz for 1 s and (d) magnified top central region after the 1 s treatment (with the regions in which the discharges were localised indicated by circles).


Figure 4. Microstructural effects of a discharge cascade on a small area sample (with 100 μm PEO coating) at 2,500 Hz for 1 s, showing (a) SEM micrograph of the surface in the initial state, (b) superimposed set of (∼200,000) video images taken during the process, (c) magnified SEM of the region indicated in (b) as the cascade location and (d) the same area after PEO processing (with the region in which the discharges were localised indicated by circles).


It was also found that at high frequency, discharges are able to form during the cathodic half-cycle. Cathodic discharges are only able to form after around 30-40 μm of coating has be produced. This was attributed to a kinetic effect, based on the ability of hydrogen and hydroxyl ions to move through the coating and carry the imposed current density. At high frequency, there is less time for the movement of hydrogen ions (to carry the cathodic current). As the coating thickens, this process becomes more difficult, so the cathodic voltage rises. Eventually, the cathodic voltage is sufficient to initiate dielectric breakdown under cathodic polarisation, in a similar manner to anodic discharges. However, the cathodic discharges carry higher current, emit more light, and are higher energy. They form large craters in the coating microstructure, and coating mass gain and growth rate drop off after the cathodic discharges are initiated. Examples of the defects caused by cathodic discharges are shown in Figure 5. 

Sam_Fig 5-3  Sam_Fig 5-2 

Figure 5. Micrographs from a sample processed for 30 minutes at 2500 Hz, showing (a) part of the free surface and (b) a transverse section through a large crater of the type generated by cathodic discharges.


[1] TW Clyne, SC Troughton "A review of recent work on discharge characteristics during plasma electrolytic oxidation of various metalsInternational Materials Reviews, (2018) 1-36

[2] CS Dunleavy, IO Golosnoy, JA Curran, TW Clyne "Characterisation of discharge events during plasma electrolytic oxidationSurface and Coatings Technology, 203, 22 (2009): 3410-3419

[3] CS Dunleavy, JA Curran, TW Clyne "Self-similar scaling of discharge events through PEO coatings on aluminiumSurface and Coatings Technology, 206, 6 (2011): 1051-1061

[4] CS Dunleavy, JA Curran, TW Clyne "Time dependent statistics of plasma discharge parameters during bulk AC plasma electrolytic oxidation of aluminiumApplied Surface Science, 268 (2013): 397-409

[5] A Nomine, SC Troughton, AV Nomine, G Henrion, TW Clyne "High Speed Video Evidence for Localised Discharge Cascades during Plasma Electrolytic Oxidation" Surface & Coatings Technology, 269 (2015) 125-130

[6] SC Troughton, A Nominé, AV Nominé, G Henrion, TW Clyne "Synchronised electrical monitoring and high speed video of bubble growth associated with individual discharges during plasma electrolytic oxidation" Applied Surface Science, 359 (2015) 405-411

[7] SC Troughton, A Nominé, J Dean, TW Clyne "Effect of individual discharge cascades on the microstructure of plasma electrolytic oxidation coatings" Applied Surface Science, 389 (2016) 260-269