Definition first: what micro-arc oxidation is, and where it parts company with anodising
Micro-arc oxidation (MAO; in the international literature also plasma electrolytic oxidation, PEO, and earlier anodic spark deposition, ANOF) is a surface conversion process in which a valve-metal workpiece - aluminium, magnesium, titanium, niobium, tantalum - acts as the anode in a weakly alkaline electrolyte. When the applied voltage is raised past the dielectric breakdown threshold of the growing oxide film (typically 200-600 V), micro-plasma discharges ignite, local transient temperatures reach roughly 2000 degrees C, molten substrate reacts with oxygen and is quenched by the electrolyte, and a ceramic conversion layer built from high-temperature phases (alpha-Al2O3, Mg2SiO4, periclase, spinel) grows in situ. The dividing line with conventional anodising is not the power supply but whether micro-plasma discharge occurs: sulphuric acid anodising works at 12-60 V inside an electrochemical dissolution-deposition equilibrium and yields amorphous or gamma-phase films, whereas MAO produces corundum (alpha-Al2O3), a phase that can only form at high temperature - which is why hardness and wear resistance change by an order of magnitude.
Key figures at a glance
- Working voltage 200-600 V (Mg alloys commonly in this band); conventional anodising 12-60 V.
- Film thickness 20-100 um on Mg, 30-80 um typical engineering range on Al; up to 200-230 um reported.
- Microhardness 500-1500 HV routinely; 2134 HV at 120 min and a peak of 4300 HV reported at ja = 15 A/dm2 with jc/ja = 0.7.
- Neutral salt spray 500-1000 h for MAO; above 200 h for hard anodising; 100-200 h for Mg anodising; 24-72 h for chemical conversion films (GB/T 10125-2021).
- Growth rate rises slowly up to ja of about 3 A/dm2, then accelerates sharply beyond 5 A/dm2; typical treatment 20-60 min.
- Dual-layer structure loose outer layer dominated by gamma-Al2O3, dense inner layer dominated by alpha-Al2O3; the dense layer accounts for 20-80% of total thickness and can exceed 66% under optimised settings.
- Electrical insulation insulation resistance above 100 Mohm; breakdown voltage is a primary selection criterion for electrical and electronics structures.
- Thermal stability long-term service above 300 degrees C; electrolytes are weakly alkaline silicate systems without strong acids, alkalis or heavy-metal complexants.
1. Growth mechanism: what happens past the breakdown point
1.1 The four stages
- Conventional anodic stage. At low voltage a thin, dense barrier layer forms under the balance between electrochemical oxidation and film dissolution in the electrolyte. No visible discharge.
- Spark stage. Once the breakdown threshold is crossed, transient sparks appear at weak points - small, fast-moving, low-intensity. The film converts from barrier to porous type. This is the threshold signal: from here on the microstructure is governed by breakdown.
- Micro-arc stage. Sparks mature into brighter, stable but migrating arc channels. Molten metal-oxygen melt is ejected from each channel and freezes around its mouth as a disc-shaped splat - one disc per channel. Most of the film thickness is built in this stage.
- Arc stage. The arc anchors, energy concentrates, local overheating burns and powders the film. This is the failure boundary and must be suppressed through waveform control, current-density ceilings, electrolyte circulation and cooling.
1.2 Why the phase assemblage is not simply "aluminium oxide"
The channel process is "instantaneous high temperature plus electrolyte quench", so metastable high-temperature phases are preserved:
- Al substrates: gamma-Al2O3 and amorphous matter in the loose outer layer, alpha-Al2O3 (corundum) in the dense inner layer; mullite forms in silicate electrolytes, with aluminium supplied by the substrate and silicon by the bath.
- Mg substrates: periclase (MgO), forsterite (Mg2SiO4) and spinel (MgAl2O4) - silicate baths favour forsterite, aluminate baths favour spinel.
- Ti substrates: anatase and rutile TiO2.
- Current density steers the phases: at low j the alpha-Al2O3 share rises markedly; above roughly 10 A/dm2 the film turns gamma-dominated. In short, "high current buys thickness, low current buys corundum" - a trade-off that must be taken together with the hardness target.
2. Process window: current density, ratio and waveform
2.1 Electrical parameters
- Anodic current density ja 0.5-15 A/dm2. Together with jc/ja it sets growth rate, dense-layer share and pitting resistance; ja = 15 A/dm2 with jc/ja = 0.7 gives the strongest pitting resistance and the highest microhardness in published work.
- Cathodic/anodic ratio jc/ja - the core variable of a bipolar asymmetric waveform. Across its range the dense-layer share can be tuned between 20% and 80%; the reverse half-cycle provides cooling and throttles over-growth.
- Waveform and frequency line frequency, medium frequency, high-frequency pulse and DC all produce films; frequency and duty ratio control energy per unit time and heat diffusion, hence roughness and micro-crack density. For equipment selection prioritise switchable constant-voltage/constant-current, bipolar asymmetry and live data acquisition.
- Dynamic resistance forward and reverse resistance diverge during growth (forward usually much higher); their time curves are the practical online indicator of densification during scale-up.
- Time 20-60 min typical. Growth rate peaks and then falls beyond about 60 min, while hardness climbs fast for the first 45 min and flattens - so extra time mostly buys thickness and roughness, not performance.
2.2 Chemical and thermal parameters
| System | Typical main salt | Film character | Engineering bias |
|---|---|---|---|
| Silicate | Na2SiO3 5-20 g/L + NaOH 2-10 g/L | Mg2SiO4 / mullite, thicker and more uniform film, wide process tolerance | Default starting point for Mg alloys, low environmental load |
| Aluminate | NaAlO2 5-15 g/L | Favours spinel / alpha-Al2O3, higher hardness and dense-layer share | Wear-driven components |
| Phosphate | Na3PO4 / Na2HPO4 5-20 g/L | Faster growth, phosphate-bearing phases, commonly blended with tungstate | Where rate and corrosion resistance must both be met |
| Tungstate composite | Na2WO4 1-5 g/L blended with silicate or phosphate | Better density and corrosion resistance, higher cost | Aggressive corrosion specifications |
| Particle / sol doping | Published example: 3 g/L Cr2O3 + 1 g/L graphene | Higher alpha/gamma-Al2O3 share; 25.3 um thickness, 763.01 HV, mean friction coefficient 0.4781 | Combined anti-friction and densification route |
- Bath temperature and circulation most lines hold 20-40 degrees C with forced cooling; loss of thermal control is the leading cause of arc transition and film powdering.
- Geometry and shielding edges concentrate current, deep cavities and blind holes shield the field, producing burnt or bare areas. Design with R0.2-0.5 mm radii and anticipate limited internal coverage.
- Pre-treatment degrease, rinse, (etch and desmut where needed), rinse, dry. Die-spray residue and silicon segregation on die-cast parts show up directly as mottling and pore clusters.
3. Film structure and acceptance criteria: dual layer, porosity, roughness
3.1 Dual layer and dimensional compensation
The interface between the two layers sits close to the original workpiece surface: the loose layer grows outward, the dense layer inward. Two consequences matter commercially:
- Dimensions once the loose layer is ground off, the part is essentially back at its original size - so counting the loose layer as functional thickness in an assembly calculation causes interference.
- Hardness gradient microhardness increases from surface to substrate; a cross-section test is required, and the load, position and preparation must be stated in the report.
3.2 Porosity - the weakness you must design around
Discharge channels leave micro-pores (sub-micrometre to a few micrometres) together with disc-shaped splats and micro-cracks. A typical industrial part is reported at porosity below 1% (AZ91, 15-20 um, HV 600-700), while unoptimised films are clearly more open. Boiling-water data make the cost explicit:
| Immersion / h | Thickness / um | Hardness / HV | Porosity / % | Corrosion current / uA cm-2 |
|---|---|---|---|---|
| 0 | 18.5 | 680 | 0.8 | 1.5 |
| 24 | 18.3 | 665 | 1.0 | 2.0 |
| 72 | 17.9 | 650 | 1.5 | 3.5 |
| 168 | 17.2 | 630 | 2.0 | 5.2 |
After 168 h the film lost about 7% of its thickness, porosity rose from 0.8% to 2.0% and the corrosion current grew roughly 3.5-fold, yet the layer stayed coherent with no delamination. The conclusion: the corrosion ceiling of a MAO film is set not by the ceramic phases but by whether the pores and micro-cracks are sealed.
3.3 Roughness and appearance
As-grown surfaces are comparatively rough, coloured grey-white to dark grey or brown, and cannot be dyed reliably. That limits their use as decorative finishes - and is an advantage when the part is destined for paint, because the micro-nano roughness and open porosity give an organic coating mechanical keying.
4. Sealing: the step that decides the project
| Method | Typical conditions | Mechanism | Strength | Limit |
|---|---|---|---|---|
| Hydrothermal / boiling water | 90-100 degrees C water or steam, 10-60 min | Hydration products (boehmite-like) fill pores | Cheap, heavy-metal free | Limited on large through-pores; may extend micro-cracks |
| Silane / fluoro-silane sol | Dip or spray hydrolysed silane, cure 80-150 degrees C | Sol condenses into a gel barrier inside pores | Adds low surface energy and hydrophobicity | Excess solids skin the surface and hurt topcoat adhesion |
| Resin / clear seal | Dilute epoxy or polyurethane dip, ambient or low-temp cure | Low-viscosity resin wicks into the pore network | Large corrosion gain; works as a tie coat | Adds an organic layer - assess weathering and heat |
| Wax / stearic modification | Stearic acid in alcohol, dip | Long-chain adsorption on pore walls | Clear short-term water and appearance gain on Mg | Heat-limited, sacrificial |
| Second-pulse self-sealing | Reduce ja and lengthen the reverse half-cycle at the end of treatment | Re-melting and re-freezing of the loose layer at low energy | No extra station | Narrow window; needs live voltage/current curves |
4.2 Judging seal quality
- Admittance / sodium-sulphite methods (following GB/T 8753.1 and ISO 2143 practice) give a quantified degree of sealing for incoming and batch control.
- Dye-spot test: alkaline dye uptake after wipe-down indicates open-pore density - fast but semi-quantitative.
- Electrochemical re-test: the drop in corrosion current density before and after sealing is the metric customers accept; composite treatments in the literature reduce the corrosion current of AZ31 by up to several orders of magnitude.
- Combine salt spray with boiling water: a 24 h salt-spray pass hides hydrothermal failure. Parts exposed to hot water, steam or condensation need 72 h and 168 h boiling-water checkpoints.
5. Pairing MAO with organic coatings (duplex systems)
Treating MAO as the final finish is the most common cause of failure; its correct role in a corrosion package is that of an inorganic conversion primer.
5.1 Three workable combinations
- MAO + sealing + topcoat - close the through-pores with a sol or dilute resin, then apply polyurethane or epoxy for colour, levelling and weathering. This is the standard route for magnesium housings and avoids pinholes and loss of gloss from painting straight onto a porous inorganic layer.
- Chemical conversion + paint (no MAO) - if the requirement is only adhesion promotion and in-process protection, a 1-5 um phosphate or permanganate conversion film achieves it without high-voltage plant and energy cost.
- MAO + powder coat - viable when thick-film hiding and takt time dominate, but the powder cure temperature (commonly 180-200 degrees C) must be validated against film hydrothermal stability and against migration of residual bath ions.
5.2 Four parameters to fix before production
- Target roughness - MAO raises Ra; where the topcoat must be orange-peel free, agree a measurable Ra ceiling that balances keying against levelling.
- Dimensional compensation - calculate fits and thread pitch diameter from total thickness and dense-layer share, not the nominal micron figure on the drawing.
- Recoat window after sealing - silane and sol sealers have an optimum time band before topcoating; too wet or too dry both pull down pull-off adhesion.
- Acceptance basis - state method and pass criteria by standard number for adhesion, salt spray and boiling water in the purchase agreement; never accept "appearance" as a test item.
5.3 When not to choose MAO
MAO is usually the wrong answer when colour and appearance grade are the primary requirement (grey, rough, hard to dye); when thin walls are sensitive to residual stress and micro-cracking; when takt time must sit in single-digit minutes; or when the budget cannot absorb a high-voltage supply and cooling plant, whose capital cost is several times that of conventional pre-treatment. In those cases thicker hard anodising, a sealed conversion film plus paint, or a full organic system is more rational.
6. Selection comparison
| Technology | Thickness / um | Hardness | NSS / h | Main limitation |
|---|---|---|---|---|
| Chemical conversion (Mg) | 1-5 | 50-100 HV | 24-72 | No wear resistance; primer or transport protection only |
| Conventional Mg anodising | 5-20 | 200-300 HV | 100-200 | Limited density and hydrothermal resistance; dyeable |
| Hard anodising (Al, Type III) | 25-80 | 400-600 HV | 200+ | Cold sulphuric bath, clear growth in size, few colours |
| Micro-arc oxidation (MAO/PEO) | 20-100 (up to 200-230) | 500-1500 HV, higher possible | 500-1000 | Through-porosity and micro-cracks demand sealing; rough, not reliably dyeable; high energy and capital cost |
| Organic system (epoxy primer + PU topcoat) | 60-200 total DFT | Pencil scale, not HV | 1000+ with proper spec | Wear and hardness far below ceramic; depends on application control |
7. Standards and test methods
| Number | Scope | Use in this system |
|---|---|---|
| GB/T 39233-2020 | Metal and other inorganic coatings - Magnesium alloy micro-arc oxidation films | China's product/method standard for Mg MAO: thickness, hardness, corrosion, insulation |
| T/TMAC 219-2025 | Technical specification for magnesium alloy surface treatment (issued and effective 2025-08-11) | Spans passivation, MAO, electroless/plating and painting: requirements, flow, inspection, HSE |
| GB/T 38216.2-2019 | Magnesium alloys surface treatment - Part 2: anodic oxidation films | Boundary of the adjacent non-breakdown process, used for comparison |
| GB/T 10125-2021 | Corrosion tests in artificial atmospheres - Salt spray tests | Method basis for every NSS hour quoted in contracts |
| GB/T 4340.1 | Metallic materials - Vickers hardness test | Microhardness testing and reporting; load and section must be stated |
| GB/T 5210 / GB/T 9286 | Tensile adhesion / cross-cut | Quantified acceptance of sealer-to-topcoat adhesion |
| ISO 2143 / GB/T 8753.1 | Anodic oxidation coatings - sealing quality | Admittance and sodium-sulphite methods, transferable to MAO batch control |
A caveat: existing standards cover sealing quality and boiling-water resistance weakly, yet these two govern real service life for magnesium parts in humid and hot-water duty. Add 72 h and 168 h boiling-water checkpoints - hardness, porosity and corrosion current - to the technical agreement, and require the supplier to hand over the raw curves for the same batch.
Frequently asked questions
Q1: Micro-arc oxidation or hard anodising - how do I choose?
Three filters: substrate, failure mode, downstream process. For aluminium parts dominated by abrasive wear where size stability and colour matter, hard anodise. Where a Mg or Ti part must resist corrosion and wear together, or needs insulation and service above 300 degrees C, use MAO. If a topcoat follows, MAO's rough dual layer keys the paint better - but seal it first.
Q2: Is higher hardness always better?
No. Hardness reflects alpha-Al2O3 share, dense-layer fraction and test load; a figure taken through the loose layer understates the functional film, while hardness bought with high ja usually comes with more roughness and micro-cracking. Anchor acceptance on dense-layer fraction per micron plus measured wear and corrosion.
Q3: Why does an unsealed film still pass salt spray?
Salt spray probes short-term barrier behaviour against chloride; the film's own passivation and re-deposition at pore mouths carries it a few hundred hours. Boiling water, condensate and humid heat drive water through the pores to the substrate and age the film from underneath - hence the 3.5-fold corrosion-current rise at 168 h above. Re-verify sealing with boiling water whenever condensation or steam is in scope.
Q4: Is the bath environmentally demanding?
Most systems are weakly alkaline silicate, aluminate or phosphate with no strong acids or alkalis, no hexavalent chromium and no heavy-metal complexants - a clear advantage over chromic acid anodising. Long-run maintenance is about aluminium and magnesium ion build-up and suspended solids: provide side filtration and periodic adjustment, and treat spent liquor per local metal-finishing requirements.
Q5: Can MAO parts be dyed or colour matched?
Reliably, no. The film is porous but rough, grey-white to dark grey or brown, takes dye weakly and varies batch to batch. If appearance leads, let MAO carry function (corrosion, wear, insulation) and put colour in an organic topcoat - do not ask an inorganic film to deliver the colour.
Q6: What about cavities, deep holes and threads?
Field strength collapses inside cavities and high aspect-ratio blind holes, giving uneven or absent films; thread pitch diameter shrinks as the film grows. Add process holes or auxiliary conductivity at design stage, mask threads, or plan a re-tap and go/no-go check afterwards.
Q7: As a buyer, which three records should I insist on?
First, a cross-section SEM image with the dense-layer fraction - it predicts real performance. Second, the sealing method plus a measured degree of sealing (admittance or dye spot). Third, a combined salt-spray and boiling-water report showing corrosion-current and hardness change. These three tell you far more than a nominal micron figure.
Sources: GB/T 39233-2020; T/TMAC 219-2025; GB/T 38216.2-2019; GB/T 10125-2021; Song Renguo, Development and applications of micro-arc oxidation technology, Journal of Materials Engineering 2019, 47(3): 50-62; Yang Han, Feng Jun et al., Effect of doped chromium oxide and graphene on microstructure and properties of tritium-blocking coatings on micro-arc alumina, Journal of Materials Engineering 2024, 52(12): 151-158, DOI 10.11868/j.issn.1001-4381.2023.000618; study on preparation and characterisation of MAO ceramic films on Al and Ti alloys (source of the ja = 15 A/dm2, jc/ja = 0.7 and HV 4300 data); industry comparisons of conversion film, anodising and MAO for magnesium alloys together with published third-party test data.
Last updated: 2026-10-20
Prepared by the technical team of Kexin New Materials (Guangdong) Co., Ltd. as a selection reference for industrial coatings and surface engineering. All figures are attributed to a standard or publication; validate against your own service conditions.