Surface treatments — anodising, plasma spraying, vapour deposition: Question bank

Before we start, one anchor so nothing here is used before it is defined:

Anode = the electrode where oxidation (loss of electrons) happens. In anodising the part itself is the anode, so the part is what gets oxidised — that is why we grow oxide on it.
Passivation = a metal protecting itself with a thin, stuck-on oxide skin (see Corrosion and Passivation).
Refractory = melts only at very high temperature (ceramics like zirconia, ~2700 °C).

Definition Every symbol used on this page — defined once, up front

So no formula below uses an unexplained letter:

$I$ = current (amperes = coulombs of charge per second flowing through the bath).
$t$ = time the current flows (seconds). Together $Q = I t$ = total charge (coulombs).
$F$ = Faraday constant $= 96485 C/mol$ : the charge carried by one mole of electrons. It is the bridge from "coulombs" to "moles of electrons".
$η$ = current efficiency: the fraction (0 to 1) of charge that actually builds oxide rather than being wasted on side reactions.
$M$ = molar mass: mass of one mole of the substance (kg/mol).
$ρ$ = density: mass per unit volume (kg/m³) — turns a mass into a thickness once you know the area.
$A$ = area of the surface being treated (m²).
$c_{p}$ = specific heat: joules to raise 1 kg by 1 K without changing phase.
$L_{f}$ = latent heat of fusion: joules to melt 1 kg at constant temperature (Latent Heat and Phase Change).
$Φ$ = atom flux: number of atoms arriving per m² per second in vapour deposition.
$N_{A}$ = Avogadro's number $= 6.022 \times 1 0^{23}$ per mol: how many atoms make one mole. It converts a count of atoms into moles.

The three formulae, built visually

Before testing traps, look at the three pictures that make the formulae obvious. Each panel shows what quantity flows and how it becomes a coating.

Figure — Surface treatments — anodising, plasma spraying, vapour deposition

The middle panel is the plasma-spray energy budget: heat the particle ( $m c_{p} Δ T$ ) then melt it ( $m L_{f}$ ). The flat step in the temperature curve is where all the incoming energy goes into $L_{f}$ at constant $T$ — that is why the latent-heat term is not optional.

True or false — justify

TF1. "Anodising deposits a fresh layer of aluminium oxide onto the surface, like paint."

False — the oxide is grown from the metal itself by oxidising the aluminium, so it is integral and rooted into the substrate rather than sitting on top like a deposited film.

TF2. "A thicker anodised layer always means better corrosion protection, right up to any thickness."

False — the grown oxide is porous, and beyond a point the outer pores and stress cracks reduce protection unless the layer is sealed; thickness alone is not the whole story.

TF3. "Because plasma-sprayed droplets arrive molten, the coating is fused (welded) to the metal."

False — bonding is mostly mechanical interlocking onto a grit-blasted rough surface plus minor local fusion; it is not a metallurgical weld, so adhesion depends heavily on surface prep (Adhesion and Surface Roughness).

TF4. "CVD involves a chemical reaction at the surface; PVD does not."

True — in PVD the material is only physically evaporated or sputtered (no chemical change), whereas CVD relies on a gaseous precursor decomposing or reacting on the hot part.

TF5. "Latent heat of fusion is optional in the melt-energy budget because it is small compared to heating the particle."

False — for zirconia

L_{f} \approx 7 \times 1 0^{5} J/kg

is a large chunk of the total

\sim 2.25 \times 1 0^{6} J/kg

; ignore it and you underestimate the energy needed to actually melt (Latent Heat and Phase Change).

TF6. "In anodising, hydrogen gas is produced at the same electrode where the oxide grows."

False — oxide grows at the anode (the part) by oxidation, while

H_{2}

evolves at the cathode by reduction; they are opposite electrodes.

TF7. "Current efficiency

η = 1

in a real anodising bath."

False — some charge is lost to side reactions like

O_{2}

evolution, so

η < 1

(often ~0.6); this is why real deposition takes longer than the ideal Faraday calculation predicts.

TF8. "PVD can build a 300 µm thermal-barrier coating just as easily as plasma spraying can."

False — PVD grows sub-µm/hr, so a 300 µm layer would take an impractically long time; thick refractory coats are the job of plasma spraying, not vapour deposition.

TF9. "Raising the anodising voltage indefinitely just grows a thicker, better oxide."

False — voltage sets pore size and cell wall thickness, and above a threshold the oxide dielectrically breaks down ("burning"), producing localised burnt spots and powdery oxide instead of a uniform film.

Spot the error

SE1. "The oxidation half-reaction is

Al \to A l^{3 +} + 3 e^{-}

, so 3 electrons make one mole of

A l_{2} O_{3}

Error — one mole of

A l_{2} O_{3}

contains two Al atoms, so the balanced reaction

2 Al + 3 H_{2} O \to A l_{2} O_{3} + 6 H^{+} + 6 e^{-}

needs 6 electrons per mole of oxide, not 3 — this is exactly the

/6

in the thickness formula.

SE2. "Since anodising protects the metal, the aluminium is being reduced and acts as the cathode."

Error — protection here comes from controlled oxidation growing oxide, which happens at the anode; the metal loses electrons, so it is the anode ("anodising").

SE3. "To find deposition thickness from flux we use

r = Φ M N_{A} / ρ

Error — you must divide by

N_{A}

, giving

r = Φ M / (N_{A} ρ)

;

Φ M / N_{A}

is mass per area per second, and dividing by density

ρ

converts it to thickness per second.

SE4. "In the melt-energy formula

Q = m c_{p} (T_{m} - T_{0}) + m L_{f}

, we can drop

T_{0}

since it is only room temperature."

Error — dropping

T_{0}

overstates

Δ T

and therefore the heating term; the particle starts at

T_{0} = 25^{\circ} C

, so

Δ T = T_{m} - T_{0}

, not

T_{m}

SE5. "The CVD reaction for TiN is

TiC l_{4} + N_{2} + H_{2} \to TiN + HCl

Error — it is unbalanced; the balanced form is

TiC l_{4} + \frac{1}{2} N_{2} + 2 H_{2} \to TiN + 4HCl

(four Cl must leave as four HCl).

SE6. "Higher current efficiency

η

makes the oxide grow slower because more charge is used up."

Error — higher

η

means more of the same charge becomes oxide instead of gas, so the target thickness is reached faster, not slower.

SE7. "Plasma spraying needs an electrolyte bath just like anodising."

Error — plasma spraying is a thermal-spray (dry) process using a hot plasma jet; no electrolyte is involved, unlike the electrolytic anodising bath.

SE8. "Electrolyte temperature and pH don't matter — only current and time set the anodised layer."

Error — a warmer or more acidic bath dissolves oxide faster, widening pores and lowering hardness (and

η

); the same

I t

can give very different films at different temperature/pH.

Why questions

WHY1. Why is the natural ~2–3 nm oxide on aluminium not enough for aerospace service?

It is far too thin to resist wear and aggressive corrosion; anodising deliberately thickens it to 5–25 µm to give a hard, durable barrier.

WHY2. Why can anodised aluminium be dyed but sealed anodised aluminium holds the colour permanently?

The freshly grown oxide is porous, so dye soaks into the pores; sealing then swells/closes the pores, locking the dye in and completing the corrosion barrier.

WHY3. Why do we express the melt-energy of a spray particle per kilogram rather than per particle?

A per-kg value lets us compare directly against the plasma's enthalpy flux (also per kg), independent of the specific particle size chosen — as the middle figure's energy budget makes visual.

WHY4. Why is surface roughening (grit-blasting) done before plasma spraying?

Because bonding is mainly mechanical interlocking; a rough surface gives the flattened splats more grip, dramatically improving adhesion (Adhesion and Surface Roughness).

WHY5. Why is PVD/CVD used only for thin (sub-10 µm) films and not thick coats?

Growth is atom-by-atom at sub-µm/hr rates (the right figure shows why:

r

scales with the tiny flux

Φ

), so building a thick layer would be prohibitively slow; the payoff is exceptional hardness and smoothness in a thin skin.

WHY6. Why is plasma spraying, not electrolysis or vapour deposition, the practical route for a thick zirconia thermal-barrier coat?

Zirconia melts near 2700 °C and has no useful electrolytic route; the plasma jet (10,000–15,000 K) can melt it and stack thick splats quickly for 100–500 µm heat shields (Thermal Barrier Coatings).

WHY7. Why does the anodised-thickness formula

d = M I t η / (6 F ρ A)

divide by area

A

The charge

Q = I t

produces a fixed amount of oxide; spreading that same oxide over a larger area gives a thinner layer, so thickness falls as

A

rises — the last "divide by area" arrow in the left figure.

WHY8. Why does the

/6

appear in the anodising thickness formula and not, say,

/3

Because one mole of

A l_{2} O_{3}

holds two Al atoms, and each Al releases 3 electrons, so 6 moles of electrons build 1 mole of oxide; the electron count per mole of product is what the formula needs.

Edge cases

EC1. What happens to the growth if the current efficiency

η \to 0

With

η = 0

effectively all charge goes to side reactions (

O_{2}

evolution), so no oxide grows and

d \to 0

no matter how long you run the bath.

EC2. What if you try to anodise a metal whose oxide is non-adherent or soluble in the acid?

The oxide flakes off or dissolves as fast as it forms, so no protective integral layer builds up — anodising only works when the oxide is stable and clings to the substrate.

EC3. In plasma spraying, what happens to a particle too large to fully melt during its flight time?

It arrives partly solid, does not splat and interlock properly, and leaves a defect/void — this is why particle size and residence time are optimised together.

EC4. What is the limiting behaviour of deposition rate

r = Φ M / (N_{A} ρ)

as the arriving flux

Φ \to 0

r \to 0

— with no atoms arriving there is no growth; rate scales linearly with flux, so halving the flux halves the growth rate.

EC5. What if a CVD precursor does not decompose at the chosen surface temperature?

No reaction occurs on the surface, so no film forms; CVD requires a hot enough substrate to drive the surface chemistry, unlike PVD which just condenses physical vapour.

EC6. Degenerate case: bulk aluminium is already fully corroded before treatment — can anodising still help?

Not meaningfully — anodising grows oxide from sound metal; if the substrate is already consumed there is no clean aluminium to oxidise into a protective layer.

EC7. What is the limiting/threshold behaviour as anodising voltage is pushed past the breakdown field?

The oxide can no longer sustain the electric field, so current arcs through weak spots ("burning") — instead of thicker film you get pitted, powdery, non-uniform oxide; there is a maximum useful voltage per electrolyte.

Recall One-line summary of the traps

Anodising oxidises the part (anode, integral oxide, porous then sealed, voltage-limited by breakdown); plasma spraying is a mechanical thick refractory coat, not a weld; vapour deposition is thin atom-by-atom, PVD physical vs CVD chemical. Every misconception above comes from confusing one process's rule for another's.

Connections

Corrosion and Passivation — anodising is engineered passivation
Faraday's Laws of Electrolysis — the charge-to-thickness reasoning behind SE1/SE2
Thermal Barrier Coatings — why plasma spraying wins for thick zirconia
Adhesion and Surface Roughness — the grit-blasting logic in WHY4/TF3
Latent Heat and Phase Change — the two-term melt budget in TF5/SE4
Aluminium Alloys in Aerospace — the substrate being protected