Reference · plating, anodising, surface finishing

Plating

Electroplating is not chemistry in a bucket but four numbers: current density, temperature, time and distance. Get them right and it hardly matters which recipe you follow. Get them wrong and no bath in the world will help. This document is therefore as much about current as about liquid — and about anodising, where the voltage alone decides what colour the metal takes.

Something come out wrong? The troubleshooting table is arranged by what the surface looks like, not by cause.

Version 1.0 Baths 11 Sections 13 SEEN Units A/dm², °C, µm

The voltage is the colour

Tap a field · titanium in trisodium phosphate electrolyte

No dye is involved. The voltage decides how thick the oxide layer grows, the thickness decides which wavelength is cancelled by interference, and what is left is the colour. The same physics as an oil film on a puddle. The scale therefore comes round twice — gold appears both at 40 and at 85 volts, at different layer thicknesses.

01

First principles

Six things that decide whether the result is a mirror or a grey sheet of sandpaper. None of them is about which bath you chose.

1 · Current density, not current

The only number that means anything is amperes per square decimetre of the work's surface. Two amperes on a washer the size of a coin is a burnt disaster; two amperes on a bicycle handlebar is barely noticeable. Work out the area first, decide the current density second, and set the supply accordingly. That is why a supply with current regulation matters more than one with high power.

I [A] = j [A/dm²] × A [dm²]

Too low a current density gives dark, dull or patchy areas and a deposit that does not adhere. Too high gives burnt, grey-black, powdery edges. The window between them is often only a factor of three wide.

2 · Faraday decides the time, not your judgement

How much metal is deposited is exactly proportional to the charge. That means the thickness can be worked out in advance, and that half the current for twice the time gives the same thickness — but not the same quality. Nickel deposits at about 12.3 µm per A/dm² and hour at full current efficiency. The rest of the metals are in section 12, where they can also be calculated.

3 · The current takes the shortest path

Whatever is nearest the anode gets the most current, and therefore the thickest layer. Edges and corners stick out into the field and get most of all — which is why they burn first. Cavities, insides and shadowed areas get least, and can end up with no deposit at all.

  • Anodes on both sides if the work is flat. One anode gives thick on one side and thin on the other.
  • The distance should be at least the work's largest dimension, preferably more. A close anode gives an uneven result.
  • A thief cathode — a sacrificial piece of sheet beside a sharp edge — draws off the excess current and takes the burn instead.
  • A shield of plastic between anode and work evens out the field where it is too strong.

A bath's ability to coat even what lies in shadow is called throwing power. Cyanide baths have excellent throwing power, which is the whole reason industry has used them. The baths in this document have moderate throwing power, and deep cavities need a help anode inside them.

4 · Agitation is not cosmetic

At the cathode surface the metal ions run out faster than they can migrate there. Without agitation the depleted layer grows thicker and thicker, and you get burnt edges at a current density that would otherwise be entirely normal. Agitation — an air pump, a magnetic stirrer, or moving the work about — raises the current density you can run at, often to double.

5 · Adhesion is decided before the current is switched on

Almost every failure is a preparation error, not a bath error. A deposit that flakes, blisters or can be scraped off with a fingernail was doomed before the power supply was touched. See section 2.

6 · The substrate shows through

Plating does not level. Scratches, casting pores and grinding marks are still there afterwards, often more visible than before, because a bright deposit reflects. Either polish first, or lay down a levelling copper layer and polish that. Brighteners in the bath help somewhat but cannot hide a scratch.

02

Preparation

Ninety per cent of all failures sit here. The dullest step is also the only one that decides whether the rest was worth doing.

The water break test

Rinse the work and hold it up. If the water runs off as an unbroken film the surface is clean. If it draws up into droplets there is grease left, and then there is no point going further. The test takes two seconds and is the only quality measure you need before the bath.

The steps, in order

  1. Mechanical preparation. Grind, polish, blast. Everything that has to go should go now. After this step you do not touch the surface with your fingers again.
  2. Degreasing. An alkaline hot bath: 30 g sodium carbonate plus 20 g sodium metasilicate per litre, 60 °C, 5–10 minutes. The silicate both dissolves grease and protects light alloys. A solvent such as IPA does for light soiling but leaves a film on anything heavier.
  3. Electrolytic degreasing if you want to do it properly. The same bath, the work as cathode, 3–5 A/dm² for a minute. The hydrogen bubbles tear the last of it away mechanically.
  4. Rinsing. Thoroughly, in at least two stages. The final rinse in deionised water.
  5. Pickling and activation. 5–10 % sulphuric acid or 10 % hydrochloric acid for 30–60 seconds removes the oxide layer and makes the surface reactive.
  6. Rinse again, and straight into the bath. Steel starts to oxidise again within minutes. If the work is left to dry between pickling and bath, you have thrown step 5 away.

A rinse cascade

Two or three rinse tanks in a row, the work passing through them in turn and the fresh water added to the last. The dilution then becomes multiplicative instead of additive: three tanks at tenfold dilution each give a thousandfold overall, with a fraction of the water. It is also how you keep the volume of metal-bearing waste down.

Substrates that need special handling

  • Stainless and nickel alloys have a passive chromium oxide layer that the deposit will not stick to. They need a Wood's nickel strike — an acidic nickel bath with a high chloride content, run briefly at high current density, which tears the oxide off and lays down nickel in the same operation.
  • Aluminium oxidises in seconds. It needs a zincate treatment, meaning the aluminium oxide is exchanged for a thin zinc layer that the plating can grip.
  • Zinc die castings do not tolerate acidic baths. They need alkaline copper first.
  • Plastic has to be made conductive with graphite or copper paint, and etched chemically first so the deposit has something to hold on to.
03

Equipment

What is actually needed, and what merely looks good on the bench.

What is needed
ItemRequirementWhy
Power supply0–15 V, 0–5 A, CC/CV Must be able to limit the current, not just the voltage. A bench supply in constant-current mode is the right tool; a battery charger is not.
Power supply for anodising0–120 V DC Titanium colouring needs up to 110 V. Aluminium anodising manages on 15–25 V but needs more current.
TankPP or HDPE Takes both acid and alkali. Glass works but cracks with heat. Never metal.
HeatAquarium heater, 25–60 °C Most baths want 40–55 °C. A titanium immersion heater for aggressive baths.
AgitationAir pump or magnetic stirrer Raises the usable current density, often by a factor of two.
RackingTitanium or copper wire Titanium does not plate and can be reused. Copper is cheap but plates itself.
AnodesThe same metal as the bath Dissolves and replenishes the bath. Exception: lead or mixed metal oxide titanium as an insoluble anode.
FilteringAquarium pump with wadding Particles in the bath give roughness and pinholes. Underrated.
pH measurementStrips are enough Nickel baths drift upwards in pH while running and need adjusting.
Scales0.1 g, 2 kg Bath recipes are by weight.
Several of the chemicals require a permit in Sweden

The EU explosives precursors regulation, 2019/1148, makes the following prohibited for private individuals without a permit from MSB:

  • Sulphuric acid above 15 wt% — which means both the anodising bath and the electropolishing bath in this document.
  • Nitric acid above 3 wt% — which means the desmutting step.
  • Hydrogen peroxide above 12 wt%.

Permits are applied for at MSB and granted restrictively. Buying, possessing or using them without one is an offence, and the retailer is obliged to report suspicious transactions. Plan the work around what you may actually keep at home — or leave those steps to someone who holds a permit.

Protection
  • Ventilation. Hydrogen forms at the cathode and oxygen at the anode — together a flammable mixture. Acid baths also give off an acid mist that rides up with the bubbles. Work outdoors or under extraction, never in a closed room.
  • Close-fitting goggles and nitrile gloves throughout.
  • Anodising runs at high voltage. 110 V direct current through a salt bath is lethal in a different way from 12 V. One hand at a time, dry floor, and never reach into a live bath.
  • Nickel is a sensitiser and nickel salts are classified as suspected carcinogens by inhalation. Weigh them out of doors, wear a mask for powder, and never let children near the bath.

Two things that are not in this document

Cyanide baths. The industry standard for copper on zinc, for silver and for gold, and unbeatable in throwing power. They are also lethal, and mixed with acid they give hydrogen cyanide. They have no place in a home workshop and are not described here.

Hexavalent chromium. Hard chrome and yellow chromating rest on Cr(VI), which is carcinogenic by inhalation and heavily regulated in the EU. Decorative chrome plating also needs a current density and a degree of bath control that are not realistic at home. Use trivalent chromates over zinc instead — they are legal, they work, and they are in section 6.

04

Copper

The foundation under almost everything else. The copper layer is what levels the surface and gives the following layers something bright to sit on.

Acid copper bath Acid copper · copper sulphate and sulphuric acid

Cu · n = 2
Current density2–5 A/dm²
Temperature20–30 °C
pH< 1
AnodePhosphorised copper
Rate13.2 µm/A/dm²/h
AgitationNecessary

The bath

IngredientPer litreRole
Copper sulphate, pentahydrate200 gMetal source
Sulphuric acid, concentrated30 mlConductivity, throwing power
Hydrochloric acid 30 %0.2 mlChloride, about 60 mg/l — needed for the anode to dissolve evenly
Deionised waterto 1 lCarrier
Acid into water

Dissolve the copper sulphate in most of the water first, then pour the acid into the solution, slowly and with stirring. Never the other way round. The mixture gets noticeably hot.

About the chloride

Those two tenths of a millilitre look like a typo but are decisive. Without chloride the phosphorised copper anode passivates, stops dissolving evenly, and the bath loses copper. Too much chloride gives a pitted, matt deposit instead. The window is roughly 40–80 mg/l.

Brighteners

Without additives the copper comes out dull and coarse-grained. Three components are used industrially: a carrier (polyethylene glycol), a brightener (SPS, bis-sodium sulphopropyl disulphide) and a leveller (Janus Green B). Buy a ready-made additive set for circuit boards — mixing them yourself is not worth the trouble.

An honest home alternative: a few drops of molasses or a pinch of gelatin per litre gives a fine-grained, semi-bright deposit. Not a mirror, but far better than nothing.

Use for

  • A levelling base layer under nickel and chrome. Copper is soft and fills microscopic unevenness.
  • Building up worn parts that are then turned or ground to size.
  • Electroforming — building a free-standing copper shell on a model that is then removed.
  • A conductive layer on plastic after graphiting.
Not directly on steel or zinc

In an acid copper bath, copper deposits chemically on iron and zinc before the current has done anything — a loose, brown-red film with no adhesion. Lay down alkaline copper or nickel first.

Base layerLevellingElectroforming
05

Nickel

The workhorse. Hard, corrosion-resistant, bright or matt to taste, and the layer that carries almost every decorative finish.

Watts nickel The standard bath since 1916 · sulphate, chloride, boric acid

Ni · n = 2
Current density2–5 A/dm²
Temperature45–55 °C
pH3.8–4.5
AnodeNickel, activated
Rate11.8 µm/A/dm²/h
Efficiencyabout 96 %

The bath

IngredientPer litreRole
Nickel sulphate, hexahydrate280 gMetal source
Nickel chloride, hexahydrate45 gDissolves the anode, raises conductivity
Boric acid40 gBuffer at the cathode surface
Deionised waterto 1 lCarrier

The boric acid is not optional

Hydrogen ions are consumed at the cathode surface, so the pH rises locally even if the bath as a whole is correct. Without a buffer, nickel hydroxide precipitates into the deposit and you get a grey, brittle, poorly adhering layer. The boric acid holds the pH down exactly where it counts. It is the ingredient most often left out of home recipes and the commonest reason they do not work.

Bright or matt

  • Without additives the nickel comes out semi-matt and silver-grey, softer and more ductile. That is often what you want on a functional part.
  • Bright nickel plating needs two additives: saccharin or sodium saccharinate as the primary brightener, and a secondary one such as butynediol. The result is mirror-bright but harder and more brittle, and it contains sulphur — which lowers the corrosion resistance.
  • A wetting agent, sodium lauryl sulphate at a very low concentration, lowers the surface tension so hydrogen bubbles release from the surface rather than leaving pinholes.

Maintenance

  • The pH rises while running. Bring it down with dilute sulphuric acid, up with nickel carbonate. Check it every time you run.
  • The anode should be bagged in polypropylene. Nickel anodes shed a sludge that otherwise ends up in the deposit.
  • Filter between runs. Pinholes are nearly always down to particles.
  • Iron contamination from dropped work makes the bath cloudy. Raise the pH to 5, aerate, filter off the precipitated iron hydroxide, and bring the pH back down.
Nickel allergy

About one woman in ten and a few per cent of men react to nickel, and once developed the allergy is lifelong. Do not nickel-plate jewellery, watch parts, spectacle frames or anything else that sits against skin — REACH restricts nickel release from such articles, and the reason behind the rule applies just as much to what you make for yourself.

Corrosion protectionUnder chromeHardSensitiser

Electroless nickel Electroless nickel · nickel-phosphorus, with no current at all

Ni-P
CurrentNone
Temperature85–90 °C
pH4.5–5.0
Rate15–25 µm/h
Phosphorus content6–12 %
Bath lifeLimited

Why it is interesting

Sodium hypophosphite reduces the nickel ions chemically, without external current. That means the layer comes out exactly the same thickness everywhere — in threads, in drilled holes, inside tubes, on the back. No current density, no throwing power, no anode geometry. For a part of complex shape it is often the only thing that works.

IngredientPer litreRole
Nickel sulphate, hexahydrate25 gMetal source
Sodium hypophosphite25 gReducing agent
Sodium citrate15 gComplexing agent, keeps the nickel in solution
Sodium acetate10 gBuffer
Stabiliser, bismuth or antimonytraceStops the whole bath precipitating
Do not use a lead stabiliser

Older recipes call for lead or cadmium in trace amounts, and they work very well — but RoHS and ELV restrict both, and neither belongs in a home workshop. Bismuth, antimony or iodate are today's alternatives and are in every commercial lead-free bath. Buy a ready-made concentrate rather than mixing from scratch: the stabiliser content is in milligrams per litre and is what decides whether the bath survives, and it cannot be weighed out on a kitchen scale.

The bath is short-lived and temperamental

An electroless nickel bath is consumed while it works and becomes unstable as it ages. Without analysis equipment you get a handful of runs out of a fresh bath, and then it starts precipitating on the tank walls instead of on the work.

Heat treatment at 400 °C for an hour precipitates nickel phosphide and raises the hardness from about 550 to over 1000 HV. That makes it harder than hard chrome.

Even thicknessThreads and holesHard after heat treatment
06

Zinc and passivation

The genuinely useful rust protection, and the only one in this document that sacrifices itself for the steel underneath.

Acid zinc bath Chloride zinc · the simple home bath

Zn · n = 2
Current density1–3 A/dm²
Temperature20–30 °C
pH4.5–5.5
AnodeZinc
Rate16.2 µm/A/dm²/h
Typical thickness8–12 µm

The sacrificial anode principle

Zinc is less noble than iron. If the coating is scratched the zinc goes on corroding instead of the steel, and the exposed steel is protected electrochemically, not merely mechanically. That is the difference from nickel and chrome, which protect only as long as they are unbroken — a scratch through nickel accelerates the rust underneath.

IngredientPer litreRole
Zinc chloride70 gMetal source
Potassium chloride200 gConductivity
Boric acid25 gBuffer
Brightener for zincper manufacturerWithout it the zinc goes grey and powdery

Passivation — the step that actually gives the rust protection

Bare zinc oxidises to white rust within weeks. Passivation lays a thin chromate film over the zinc, and that film is what multiplies the life. Use trivalent chromium, Cr(III), sold ready-made in the hobby trade. Hexavalent Cr(VI) gives the classic yellow rainbow but is carcinogenic and requires an authorisation.

Passivations over zinc
TypeAppearanceProtectionNote
BlueClear blue, almost colourless24–72 hThin, decorative. The commonest on screws.
YellowYellow iridescent72–150 hTrivalent yellow exists but the colour is weaker than the old Cr(VI) yellow.
BlackDeep black50–100 hNeeds silver ions in the bath. The dearest.
OliveOlive green, matt150–300 hBest protection, worst appearance. Military standard.
SealingOver any of them+50–100 %A silicate-based topcoat. The cheapest way to double the life.

The hours refer to salt spray to ISO 9227 until first white rust, and should be read as a ranking rather than as a promise about years outdoors.

Passivate within the hour

The chromate film adheres best to fresh zinc. Do not leave the plated work overnight — the protection is markedly worse the next day.

Hydrogen embrittlement

Steel with a tensile strength from about 1000 MPa, which corresponds to roughly 31 HRC, takes up hydrogen during pickling and plating and can crack spontaneously days later. Springs, high-tensile bolts and hardened parts are the ones at risk.

ASTM B850 specifies 190–220 °C for 8 to 22 hours, with the time increasing with strength — the hardest steels sit at the upper end. Bake within a few hours of plating and before any passivation. Shorter cycles appear in manufacturers' procedures, but they are qualified for a specific part and process, not a general rule.

This is not a theoretical risk — it is the classic reason a plated spring breaks in the hand. And the treatment promises no complete protection, only a greatly reduced risk.

Rust protectionSacrificialScrews and fittings
07

Tin

For solderability and for things that touch food. Soft, base, and with a curious defect that takes years to appear.

Acid tin bath Methanesulphonic or sulphuric acid · Sn(II)

Sn · n = 2
Current density1–3 A/dm²
Temperature20–30 °C
AnodePure tin
Rate28.8 µm/A/dm²/h
Typical thickness5–15 µm
Food contactApproved
IngredientPer litreRole
Tin(II) sulphate40 gMetal source
Sulphuric acid, concentrated100 mlConductivity
Phenolsulphonic acid20 mlStops Sn(II) oxidising to Sn(IV)
Additive for tinper manufacturerWithout it the layer comes out coarse-grained

The bath ages by oxidation

Divalent tin is oxidised by atmospheric oxygen to tetravalent, which precipitates as a white sludge and cannot be plated. A tin bath left open for a few weeks is spent. Keep the lid on, stir no more than necessary, and count on a shorter life than the nickel bath.

Whiskers

A pure tin deposit grows spontaneous single-crystal filaments over time, a micrometre or so thick and up to millimetres long. They can short out electronics, and it is the reason lead was in solder to begin with. A slight alloying with lead or bismuth, or a thicker and duller layer, suppresses them.

SolderabilityFood contactWhiskers
08

Silver and gold

Expensive baths in small volumes. Here the cyanide question is unavoidable, and the answer is to choose the baths that avoid it.

Silver, cyanide-free Thiosulphate or succinimide based

Ag · n = 1
Current density0.3–1 A/dm²
Temperature20–30 °C
AnodePure silver
Rate38.4 µm/A/dm²/h
Typical thickness3–10 µm
NoteBuy ready-made

Silver has the highest electrical conductivity of all metals, which is the whole reason it is plated onto RF conductors and contacts rather than for looks.

The cyanide bath is the classic and gives the best result, but is ruled out here. Cyanide-free alternatives exist as ready-made concentrates — thiosulphate- and succinimide-based — and work well on a small scale. Mixing them from scratch is rarely worth the trouble: the silver salts dominate the cost, and the ready-made bath contains complexing agents that are hard to substitute.

Directly onto copper does not work

Silver deposits chemically onto copper and brass without current, which gives a loose layer with no adhesion. Put a thin nickel layer in between — it doubles as a diffusion barrier, without which copper migrates up into the silver and yellows it over time.

Silver tarnishes from sulphur in the air. Follow with a tarnish inhibitor, or accept it and clean it with recipe 18 in Chemicals, which restores the silver without abrading it away.

ConductivityRFTarnishes

Gold, sulphite-based Cyanide-free · for contacts and decoration

Au · n = 1
Current density0.1–0.5 A/dm²
Temperature50–60 °C
AnodePlatinised titanium
Rate38.1 µm/A/dm²/h
Typical thickness0.1–2 µm
NoteBuy ready-made

Gold is plated in fractions of a micrometre — a connector typically has 0.2 to 0.8 µm. The cost is in the metal, so the layer is kept as thin as the wear allows.

A nickel barrier underneath is mandatory on copper. Without it, copper diffuses through the thin gold layer within months and the surface goes dull and discoloured. Electronics often uses ENIG — electroless nickel with thin immersion gold — for exactly that reason.

ContactsDoes not oxidiseThin layer
09

Anodising aluminium

Plating in reverse. Here you do not build metal up but grow an oxide layer out of the work itself — and that layer is porous, which is the whole point.

Aluminium always has a thin natural oxide layer. Anodising makes it thicker, harder and — crucially — shot through with millions of vertical pores. The pores can then be filled with dye, and sealed afterwards. The colour therefore sits inside the metal's own surface rather than on top of it. That is why an anodised part cannot flake — there is nothing lying on the surface to let go.

2 Al + 3 H2O → Al2O3 + 6 H+ + 6 e−

The work is the anode, unlike in plating. The cathode is lead or aluminium and should have roughly twice the area.

Sulphuric acid anodising Type II · the one that takes colour

Al2O3
Electrolyte180 g/l H₂SO₄
Current density1.5 A/dm²
Voltage15–20 V
Temperature20–22 °C
Time45 min ≈ 20 µm
CathodeLead or aluminium

The rule that makes the thickness predictable

The American rule of thumb is called the 720 rule and is reckoned in amperes per square foot: current density times minutes, divided by 720, gives the thickness in thousandths of an inch. Converted to metric units it becomes

thickness [µm] ≈ j [A/dm²] × time [min] × 0.33

1.5 A/dm² for 45 minutes therefore gives about 22 µm, which is a normal thickness for type II. Double the time gives double the thickness. The relationship holds as long as the bath is kept at the right temperature — if it gets too warm the oxide dissolves as fast as it forms and the rule stops applying.

The sequence

  1. Degrease alkaline, rinse.
  2. Etch in 50 g/l sodium hydroxide, 50 °C, 1–3 minutes. Gives the matt satin finish. Skip it if you want to keep a polished surface.
  3. Desmut in 10 % nitric acid, 1 minute. Removes the black alloy layer the caustic leaves behind.
  4. Anodise as in the table. Bubbles rise from the work throughout.
  5. Rinse in deionised water.
  6. Dye in a 60 °C dye bath, 5–15 minutes. The pores draw it in.
  7. Seal in boiling deionised water, 20–30 minutes. The oxide converts to boehmite, swells and closes the pores.
The alloy decides the result

6061 and 6082 anodise excellently. The 5000 series works well. The 2000 series with its high copper content comes out patchy and dark, and cast alloys with silicon go grey to almost black whatever you do. If the part is a casting: expect grey, not colour.

The temperature is critical

Above 25 °C the sulphuric acid dissolves the oxide as fast as it forms, and you get a soft, powdery layer that takes no dye. The bath heats itself while running — a five-litre bucket rises several degrees in an hour. Cool it with bottles of ice in a water bath outside the tank, or run shorter sessions.

Dyeing

  • Textile dye for polyester works surprisingly well and is the cheapest. RIT and similar. The lightfastness against sun is poor, though.
  • Anodising dyes bought for the purpose are lightfast and give a deeper colour.
  • Black is the simplest and most durable. Red and yellow fade fastest in UV.
  • Electrolytic colouring with a tin salt and alternating current gives bronze shades that are completely lightfast — it is how architectural facades are done.
Sealing is not optional

Unsealed anodising takes up grease, fingerprints and dirt straight into the pores, and the dye washes out. Twenty minutes in boiling deionised water is what turns a coloured part into a finished one.

AluminiumTakes colourHard-wearingCannot flake

Hard anodising Type III · cold and thick

Al2O3
Temperature0–5 °C
Current density2–4 A/dm²
Voltage40–80 V
Thickness25–75 µm
Hardness400–600 HV
ColourGrey to brown-black

The same electrolyte, but close to zero degrees. The cold stops the acid dissolving the oxide while it forms, so the layer comes out much denser and harder. The price is that it turns grey or dark brown of its own accord and can barely be dyed in light shades.

Mind the dimensions. Roughly half the layer thickness grows outwards from the original surface. 50 µm of hard anodising makes the part 50 µm larger in diameter. On a fit, that matters.

The cooling is the real difficulty at home. Without active cooling of the bath you get type II with delusions of grandeur.

Wear protectionDimensional changeNeeds cooling
10

Titanium and niobium

The document's prettiest chapter, and the simplest. No dyes, no additives, no temperature control — just a voltage and ten seconds.

How it works

Titanium builds a transparent oxide layer when anodised. Light striking the surface reflects both from the top of the oxide and from the metal beneath, and the two reflections interfere. Which wavelength cancels depends on the thickness. There is no pigment. The same phenomenon that colours an oil film on a puddle and a heat-tinted weld.

The layer thickness is governed almost entirely by the voltage, and barely at all by time or current. That makes the process self-limiting: once the layer has reached the thickness that voltage can drive, growth stops and the colour is fixed.

layer thickness ≈ 2.5 nm per volt  ·  the colours cycle twice up to 110 V

Because interference is cyclic the colours recur. Gold is found both at about 40 and at 85 volts, but at different layer thicknesses and in slightly different tones. The scale at the top of the page shows the whole series.

Titanium anodising Also niobium, tantalum and zirconium

TiO2
Electrolyte50 g/l TSP
Voltage10–110 V DC
CurrentLow, self-limiting
Time10–60 s
TemperatureRoom temperature
CathodeTitanium or stainless

Electrolyte

Trisodium phosphate, 50 g per litre, is the standard and gives clean colours. Alternatives that work: bicarbonate 50 g/l, borax, or Coca-Cola — the last is not a joke but a consequence of its phosphoric acid content, though it gives muddier shades. Avoid chloride electrolytes such as table salt: the chlorides attack the oxide and give patches.

The sequence

  1. Clean thoroughly. Degrease with alkali, rinse, and do not touch the surface afterwards. Fingerprints become permanent markings in the colour.
  2. Pickling in dilute hydrofluoric acid gives the cleanest colours, but hydrofluoric acid has no place in a home — the burns are not felt for hours. Mechanical cleaning and thorough degreasing is enough.
  3. Set the voltage before you dip. The colour is decided the moment the work goes in.
  4. Dip for 10–60 seconds until the current has fallen towards zero.
  5. Rinse and dry. No sealing is needed — the layer is dense.
Two tricks

Gradients come from lowering the work in gradually while the voltage is raised, or from dipping it in stages. The result is bands of colour that fade into one another. Patterns can be painted with a brush dipped in electrolyte and connected to the positive — only the parts the brush touches take colour.

It can be redone, but not undone

To change the colour: raise the voltage, and the layer grows on to the next shade. Going downwards is not possible — the layer does not shrink. The oxide then has to come off mechanically or with a pickle, and you start again.

110 volts direct current

This is the only process in the document where the voltage itself can kill. A salt bath conducts extremely well. Keep one hand in your pocket, never touch the tank and the work at the same time, and switch the current off before you lift anything out. A supply with current limiting at a hundredth of an amp or so makes it considerably safer without affecting the result.

What it is used for

  • Jewellery — titanium and niobium are hypoallergenic, unlike nickel.
  • Surgical implants and instruments, where the colour is used for size marking.
  • Bicycle and motorcycle parts, screws and bolts.
  • Marking tools that have to withstand solvents — the colour is oxide and cannot be wiped off.
No dye involvedSelf-limitingHypoallergenicHigh voltage
11

Electropolishing

Plating backwards. Metal is dissolved off the surface, and because the peaks dissolve faster than the valleys the surface becomes smoother and finally a mirror.

Electropolishing stainless Phosphoric and sulphuric acid

Anodic dissolution
Current density10–25 A/dm²
Temperature50–70 °C
Time3–10 min
The workAnode (+)
CathodeLead or stainless
Removal5–25 µm
IngredientPer litreRole
Phosphoric acid 85 %700 mlForms the viscous layer at the surface
Sulphuric acid, concentrated200 mlConductivity
Deionised water100 mlCarrier

The mechanism is that a viscous, metal-saturated layer settles in the valleys and insulates them, while the peaks stick up through it and dissolve. The result is a bright surface and passivation into the bargain.

Hot concentrated acid mixture

This is the most dangerous bath in the document: almost neat acid at 60 °C, with a high current density and vigorous gassing that carries acid mist with it. Face shield, apron, gloves and extraction. Always add acid to water, and let it cool between steps.

Works best on stainless. Carbon steel needs other electrolytes and rarely gives a true mirror. Aluminium is electropolished in a different mixture entirely.

Mirror finishPassivatesRemoves material
Electrolytic rust removal belongs in the other document

Removing rust with soda and direct current is the same family of processes but needs no plating equipment at all. It is in Chemicals, recipe 16. That is also where the warning sits that is easy to miss: never a stainless anode, because the chromium dissolves into the bath as hexavalent chromium.

12

Work out the bath

Current and time can be worked out exactly. It is the only part of plating that requires no judgement at all.

How it is calculated

Faraday's law says the deposited mass is proportional to the charge. Rearranged for time it becomes:

t [s] = (ρ × d × n × F) / (M × j × η)

where ρ is the density, d the thickness, n the number of electrons per metal ion, F Faraday's constant 96 485 C/mol, M the molar mass, j the current density and η the current efficiency. The efficiency is the only term that is not a constant of nature — part of the current goes to evolving hydrogen instead of metal, and how much depends on the bath and how it is kept.

Measuring the area

Area is what people are careless with, and an error here goes straight through. Remember both sides of flat parts, and that a screw has considerably more surface than it looks, because of the thread — count on roughly double a plain cylinder. If you are unsure, guess high. Too low a current density gives a slower but usable result; too high gives burnt edges that cannot be saved.

Typical thicknesses

What is usually laid down
ApplicationLayerThickness
Screws and fittings indoorsZinc + blue passivation5–8 µm
Outdoors, weather-exposedZinc + olive or sealing12–25 µm
Decorative, under chromeCopper + nickel15 + 10 µm
Wear surfaceElectroless nickel, hardened25–50 µm
Solderable surfaceTin5–10 µm
ConnectorsNickel + gold3 + 0.2–0.8 µm
RF conductorSilver5–10 µm
Anodising type IIAl₂O₃15–25 µm
Anodising type IIIAl₂O₃25–75 µm
13

Troubleshooting

Arranged by what the surface looks like, because that is what you have in front of you. The cause comes second.

Plating
SymptomLikely causeRemedy
The deposit flakes or can be scraped offA preparation error — grease left, or an oxide layer that reformedStart again from degreasing. Less time between pickling and bath.
Blisters under the layerTrapped hydrogen, or grease left in poresLower the current density, degrease electrolytically, bake to ASTM B850.
Dark, burnt edges and cornersCurrent density too high locallyA thief cathode at the edge, a greater anode distance, more agitation.
Grey, powdery surface everywhereCurrent density too high overall, or too little metal in the bathRecalculate the area. Check the bath concentration.
Matt where it should be brightThe brightener is used upTop it up. They are consumed in proportion to charge, not to time.
Pinholes, small pitsHydrogen bubbles that stuck, or particlesWetting agent in the bath, filter, increase the agitation.
No deposit inside cavitiesPoor throwing powerA help anode inside the cavity, or switch to electroless nickel.
Uneven thickness side to sideOnly one anodeAnodes on both sides, turn the work halfway.
The deposit does not grow at allBad contact, a passivated anode, or reversed polarityCheck that the work is on the negative. Brush the anode.
A loose brown-red film immediately on dippingChemical displacement — acid copper bath on iron or zincLay down alkaline copper or nickel first.
The bath goes cloudy and greenIron contamination from dropped workRaise the pH to 5, aerate, filter, bring it back down.
Streaks following the current directionToo little agitationAir pump, or move the work while running.
A plated spring breaks days laterHydrogen embrittlementBake at 190–220 °C for 8–22 hours immediately after plating, the longer the harder the steel. It cannot be repaired afterwards.
Anodising
SymptomLikely causeRemedy
A soft, powdery layer that rubs offThe bath is too warm — the acid dissolves the oxide as fast as it formsCool to 20 °C. Check throughout the run, not only at the start.
The colour runs out during rinsingInsufficient or omitted sealingBoil for 20–30 min in deionised water.
A white haze after sealingLime from the tap waterDeionised water in the last step.
Patchy or dark colourThe alloy — copper in the 2000 series, silicon in castingsCannot be remedied. Choose 6061 or 6082 if the colour matters.
Burnt or eaten away at the racking pointCurrent concentration at the contact pointA titanium spring with a larger contact area, or several contact points.
No colour at all on titaniumThe oxide layer was already there, or a chloride electrolytePickle the surface. Change from table salt to trisodium phosphate.
The titanium colour came out uneven or patchyFingerprints, or uneven immersionDegrease thoroughly. Dip in one movement, not gradually.
The wrong colour on titanium despite the right voltageVoltage drop in leads and connectionsMeasure across the work, not on the power supply display.
Waste and regulations
  • None of this may go down the drain. Nickel, copper, zinc and tin are hazardous to the aquatic environment at low concentrations, and a treatment works cannot remove them.
  • Precipitate the metals out of spent baths and rinse water: raise the pH to 9–10 with soda or lime, let the hydroxides settle, decant the water and keep the sludge.
  • Hand it in as hazardous waste at the recycling centre, clearly labelled with the contents. Ring ahead if it is more than a few litres.
  • A rinse cascade reduces the waste volume more than any other single measure.
  • Kemikalieinspektionen has requirements on storage and labelling for private individuals too. Nickel salts and chromium compounds are the most regulated.