S Shrestha a, A Sturgeon a, P Shashkov b and A Shatrov b
Paper presented at Magnesium Technology 2002, Seattle, WA, 17-21 February 2002 and published by TMS (The Minerals, Metals & Materials Society)
This paper describes the results of a study on the corrosion behaviour of die-cast magnesium alloy type AZ91D coated using the Keronite TM
process. This process is a commercially available and environmentally friendly electrolytic coating method applicable to all types of magnesium alloys. The process involves creation of a hard ceramic oxide layer on the surface of a light substrate alloy by plasma electrolytic oxidation in a low concentration alkaline solution.
Corrosion performance was assessed in a 3.5% NaCl solution using an accelerated electrochemical potentiodynamic polarisation method, and a salt spray exposure. Parallel tests were also carried out on an uncoated AZ91D alloy. The study demonstrated an improvement in the corrosion resistance of Keronite coated AZ91D magnesium alloy.
Magnesium-based alloys are of growing interest for many industrial applications, which include components in automotive 
, aerospace [2,3]
, electronics 
, textile machinery and printing machines due to their favourable strength to weight ratio. Thin walls, high casting rates and long die life are typical advantages that can be achieved by using die-cast magnesium. The most commonly used alloy is AZ91D (9%Al, 1%Zn), which exhibits good mechanical and physical properties with excellent castability 
. But the relative poor corrosion and wear resistance limits its industrial use to just a few applications 
. An example of a potential application of the Keronite coating to improve the corrosion resistance of automotive pump casing made of AZ91D alloy is shown in Fig.1a
Fig. 1. 1a) Pump case made of AZ91D and coated with Keronite TM
1b) Plasma discharge surrounding the surface of a component immersed in an electrolytic bath
Various chemical conversion surface treatments such as chrome pickling or anodising, in combination with alkali resistant organic coating systems, have been used for magnesium alloys to give limited corrosion protection in mild corrosive environments. Due to environmental concerns these traditional processes are being replaced with more environmentally safe surface treatment processes. Techniques for improving the wear resistance of magnesium alloys such as laser surface alloying  and thermal spraying  have been reported. But these applications also have limitations and cannot be used for components with restricted access areas and smaller dimensions.
Thus, there exists potential for new electrolytic processes such as Keronite TM licensed by Keronite Ltd (UK), Anomag offered by Magnesium Technology Licensing Ltd (New Zealand) and Magoxid-Coat licensed by AHC Oberflachentechnic GmbH & Co. OHG (Germany). Unlike traditional acidic chromate and/or fluoride-based processes, these new processes employ alkaline-based chromate-free solutions.
This report presents results obtained from an evaluation of the corrosion performance of die-cast magnesium AZ91D alloy without any coating and with a coating deposited using the Keronite TM process.
Keronite TM Process
Coating of magnesium alloy using the Keronite TM process involves creation of a plasma discharge around a component immersed in an electrolyte. A modulated AC voltage is applied to the component, creating intense plasma due to micro-arc ( Fig.1b) generation at the component surface. This results in oxidation of the component surface (plasma electrolytic oxidation) as well as elemental co-deposition from the electrolyte solution. The electrolyte is a low concentration alkaline solution of proprietary composition and can be disposed of without further treatment. A bath with a capacity of 200 litres and a power rating of 80kW can coat a substrate up to 50dm 2 surface area at a rate of 1-2µm per minute. Substrate preparation requires only degreasing of the component and the process is operated between 20-50°C.
Material and Test Specimens
The magnesium alloy considered in this work was a high purity die cast type known as AZ91D (Bal Mg, 9Al, 1Zn, 0.2Mn, max. 0.001Fe, max. 0.001Ni and max. 0.015Cu). A Keronite coating to a thickness of about 35µm was deposited on as-cast alloy samples of size 150x100x1.5mm. For comparison purposes in terms of performance, parallel tests were carried out on uncoated AZ91D alloy.
A coated sample was carefully sectioned to a size of 10x20mm using an abrasive cutting wheel and the cut specimen was mounted in araldite in cross section. Standard metallographic preparation techniques of grinding on abrasive papers and polishing to 1µm diamond finish were used to prepare the cross section. The obtained cross section was gold sputtered in an evaporative coater and observed under a scanning electron microscope (SEM).
Determination of the various crystalline phases present in the coating was undertaken using the X-ray diffraction (XRD) technique. The XRD technique comprised of collecting spectrum using CuK α radiation from the coating material ground to a fine powder.
Coating hardness was measured on the polished cross section by Vickers micro-indentation with a load of 25g using the procedure described in ASTM E384-99.
Free Corrosion Test
Specimens of size 35x35mm were cut from the larger samples and painted with lacquer 'Electromask' so that the coating/substrate interface was sealed during the exposure period. The specimen was immersed in a 3.5% NaCl solution. The temperature was maintained at 25±1°C and the conductivity of the solution was mainted at 53mS.cm -1. Observations were made on a daily basis over a one-month period and any change on the coating surface was noted.
Salt Spray Test
Specimen size for the salt spray test was 150x100x2mm. The salt spray exposure was carried out as described in the ASTM standard B117-97 'Standard practice for operating salt spray (fog) apparatus'. This test comprised of exposing the coating surface to a 5wt% NaCl solution atomised to create a fog within an enclosed chamber. The coating was supported 15-30° from the vertical. The exposure duration was 1000 hours and temperature was maintained at 35°C. Changes to the coating surfaces were recorded following periodic observations every 24 hours. On completion of the test after 1000-hour exposure, the surface quality was given a ranking number in accordance with ASTM D1654-92 'Evaluation of painted or coated specimens subjected to corrosion environments' using the rating for unscribed areas - procedure B.
Electrochemical Polarisation Test
Specimens were cut to a size of 35x35mm for the electrochemical corrosion tests. A small area at the rear face of the coated specimen was scratched and micro-welded with titanium rod to provide electrical connection for the electrochemical polarisation test. The coated area on the front face exposed to the electrolyte was 1cm 2
in size. Testing of the coating was performed in the as-received condition after degreasing with alcohol. Testing of the uncoated specimen was performed on a surface abraded with 1200-grit sandpaper.
Electrochemical corrosion tests were performed using a three-electrode Avesta cell arrangement attached to a computer controlled potentiostat as shown in Fig.2. Experiments were carried out according to the guidelines described in the ASTM standard G61. The electrolyte was a 3.5% NaCl solution (made from distilled water) with a pH value of about 8.1. The temperature was maintained at 25±1°C. Electrical conductivity was measured at about 53mS.cm -1 at 25°C and the solution was deaerated by continuous purging with nitrogen.
Fig. 2. Schematic diagram of the Avesta cell
The specimens were electrically connected to the Avesta cell. The coated surface was exposed to the electrolyte and the rest potential 'E corr' (also known as free corrosion potential) was stabilised for one hour prior to the anodic polarisation. This potential was measured using a reference saturated calomel electrode (SCE). The area exposed to the electrolyte was then anodically polarised from its rest potential to more positive potential at a rate of 10mV.min -1. This was continued until a current density of 5mA.cm -2 was measured in the external circuit between the working electrode (specimen) and a platinum auxiliary electrode. On reaching a corrosion current density of 5mA.cm -2 the scan was reversed and the potential reduced in the negative direction back to E corr. A plot of the corrosion current density as a function to the polarisation potential was recorded.
The collected polarisation plots were used to compare the corrosion behaviour of the magnesium alloy in the uncoated condition with that having the Keronite coating. The following features were used to compare the corrosion behaviour:
- Rest potential 'E corr' - a more negative E corr representing greater susceptibility to corrosion attack.
- Breakdown potential 'E b' - At potentials above E b, the current rapidly increases with further increase in potential. The constant current region below Eb is referred to as a passive region where little or no corrosion occurs. At potentials above E b, rapid corrosion is occurring. The presence of a passive region and a more positive value for E b indicates better resistance to corrosion attack.
A back-scattered SEM image of a typical 35µm Keronite coating is shown in Fig.3
, which displays a relatively low porosity coating. A few light grey and dark contrast features appeared in the back-scattered SEM image due to atomic number difference.
Fig. 3. Cross-section SEM image of as-deposited 35µm Keronite coating on die cast AZ91D
Figure 3 shows a multi layer coating structure, which comprised a top porous layer, a middle main coating layer and a very thin interfacial layer adjacent to the substrate. The existence of similar features has been reported for the Keronite coatings on aluminium alloys  .
An SEM image of the deposited Keronite coating surface displays an abundance of agglomerated particles formed during the micro-arc oxidation process ( Fig.4). X-ray diffraction of the coating layer confirmed spinel (MgAl 2O 4) as the major constituent together with existence of stishovite (SiO 2) and silicon phosphide (SiP 2). This is illustrated by the XRD pattern shown in Fig.5.
Fig. 4. SEM image of the as deposited surface of the Keronite coating on AZ91D
Fig. 5. X-ray diffraction patterns collected from the Keronite coating
Micro-indentation hardness values measured on the polished cross section ranged between 345 to 446 HV for the 35µm Keronite coating and about 95 for the AZ91D alloy. Average hardness values of the AZ91D alloy and the 35µm Keronite coating are compared in Fig.6.
Fig. 6. Vickers microhardness values measured on die cast AZ91D and 35µm Keronite coating under a load of 25g.
The magnesium alloy coated with Keronite did not show any sign of significant corrosion during the one-month immersion. A small amount of white corrosion product began to appear on the surface after two weeks of immersion. However, there was no significant further increase of this corrosion product during the one-month test period. The uncoated magnesium alloy showed evidence of corrosion after 24 hours of immersion and severe pitting was observed after one month of immersion. The SEM image in Fig.7 shows the surface of the Keronite coating after one-month immersion in salt water.
Fig. 7. SEM image of the exposed area of the Keronite coated specimen after one-month exposure to 3.5% NaCl solution
Fig. 8. Salt spray (ASTM B117) endurance of the uncoated AZ91D and AZ91D with the 35µm Keronite coating
Salt Spray Test
Salt spray corrosion test showed little visual evidence of corrosion attack on the surface of AZ91D coated with a 35µm Keronite coating after 1000 hours of exposure. In contrast, uncoated AZ91D displayed significant corrosion of the exposed surface after 24 hours of exposure. Figure 8 shows endurance of the coated and uncoated test coupons to the rating of 9 in accordance with ASTM D1654-92, procedure B. Surface appearance of the uncoated and Keronite coated test coupons after 1000 hours of salt spray exposure are shown in Fig.9.
Fig. 9. Optical image of the exposed surface of AZ91D (left) and AZ91D coated with 35µm Keronite coating (right) after 1000 hours of salt spray (ASTM B117) test
Electrochemical Polarisation Test
Accelerated electrochemical test (anodic polarisation) results of AZ91D uncoated and coated with 35µm Keronite are shown in Fig.10. The results demonstrate contrasting behaviour of the coated specimens compared to uncoated alloy. The coated specimen displayed a more noble rest potential 'E corr' compared to that shown by the uncoated specimen. A more noble rest potential in deaerated electrolyte generally signifies less susceptibility to corrosion attack.
Fig. 10. Anodic polarisation (ASTM G61) curves of uncoated AZ91D and AZ91D coated with Keronite in deaerated 3.5% NaCl solution at 25°C.
Moreover, the coated specimen displayed greater resistance to initiation of corrosion attack shown by the presence of a breakdown potential 'E b
' and a passive potential range 'E b
'. A metal surface, which is not protective and corrodes actively, displays a curve similar to uncoated AZ91D with significant current within a small range of positive potential from its rest potential value.
This work shows that the coating formed on the magnesium alloy mainly consists a ceramic layer of magnesium-aluminium oxide compound (Spinel). The coating microstructure in cross section also shows that the main layer of Keronite coating possesses little porosity. The bottom layer is relatively thin (1-2µm) but higher magnification SEM revealed that it is dense. Existence of multiple layers has been reported also for Keronite coatings on aluminium alloy substrates 
and also on coatings 
deposited with similar processes (micro-arc oxidation) on aluminium alloys. The presence of a dense ceramic layer on magnesium alloy would be expected to improve its resistance to corrosion attack. This was confirmed by the electrochemical test result with a presence of a passive region over a range of potential more positive to 'E corr
'. This was further supported by the resistance to corrosion attack during 1000 hours of salt spray exposure and one-month immersion in salt water. These results indicate the ability of the Keronite coating to act as a barrier coating to stop penetration of corrosive species. The surface of uncoated magnesium displayed both general type of corrosion as well as localised pitting.
Further work is now underway to look at the corrosion behaviour of Keronite coated magnesium alloys in a galvanic couple with steel.
- Microscopic observation showed that the Keronite coating has a relatively homogeneous coating structure with little porosity and multiple layers.
- The coating formed on a soft magnesium alloy is primarily a ceramic layer of Spinel (MgAl 2O 4)
- Electrochemical results suggest that the Keronite coating can improve the corrosion resistance of magnesium alloy AZ91D.
- The Keronite coating can survive one-month immersion in a saline solution and 1000 hours in a salt spray environment without significant visual evidence of corrosion attack.
- M. Brettmann, B. Reinhold, and H. Schnattinger, 'The use of magnesium in cars - Aspects of corrosion,' Magnesium Alloys and Their Applications, ed. K. U. Kainer, (Wiley-VCH, Verlag GmbH, 2000), 271-273
- B. Landkof, 'Magnesium application in the electronic industries,' Proc. 2 nd Israeli Intl Conf. on Magnesium Science & Technology, ed. E. Aghion and D. Eliezer, (Magnesium Research Institute Ltd, Israel, 2000), 50-56.
- G. B. Evans, 'Applications of magnesium in aerospace,' Magnesium Technology, Proc. London Conf., November 1986, (Institute of Metals, London, 1987), 103-109.
- R. L. Edgar, 'Global overview on demand and applications for magnesium alloys,' Magnesium Alloys and Their Applications, ed. K. U. Kainer, (Wiley-VCH, Verlag GmbH, 2000), 3-8.
- G. Govender and J. H. Ferreira, 'Erosion corrosion of magnesium AZ91D in the as-cast condition,' Proc. 2 nd Israeli Intl. Conf. on Magnesium Science & Technology, ed. E. Aghion and D. Eliezer, (Magnesium Research Institute Ltd, Israel, 2000), 371-376.
- R. Galun, A. Weisheit, and B. L. Mordike, 'Properties of laser alloyed surface layers on magnesium base alloys,' Magnesium Alloys and Their Applications, ed. B. L. Mordike, and K. U. Kainer, (Werkstoff-Informationsgesellsschaft mbH, Germany 1998), 439-444.
- U. Lenz, A Weisheit, and B. L. Mordike, 'Thermal spraying on the magnesium alloys AZ91D,' Magnesium Alloys and Their Applications, ed. B. L. Mordike and K. U. Kainer, (Werkstoff-Informationsgesellsschaft mbH, Germany 1998), 445-450.
- P. Shashkov, 'Environment-friendly ceramic surface for aluminium,' Aluminium Today, June-July 2000, 44-45.
- J. Patel, 'Micro-arc Oxidation Process Protects Aluminium Alloys,' Materials Progress, (August, 2000).