CNC Machined Products: Does Gold Electroplate Tarnish?
Gold electroplate tarnish is real, but it is almost never the gold itself failing. This page explains what actually happens on machined parts, which thickness and substrate combinations hold up, and when gold plating is the wrong choice for your application.
In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
What gold electroplate actually is
Gold electroplating deposits a thin gold layer on a conductive surface using a gold salt bath and direct current. On CNC machined parts the layer is typically 0.05 to 5 μm thick. That is thin. A human hair is roughly 70 μm across, so a 0.5 μm coating is about 1/140 of that.
Pure gold does not oxidize in air at room temperature. It does not form a sulfide film the way silver does. So when a buyer says their gold electroplate tarnish appeared, the film is usually not gold oxide. It is nickel oxide, copper sulfide, residual plating salt, or an organic contaminant sitting on top of a porous or too-thin gold layer.
That distinction matters for how you fix it. Re-plating thicker gold on the same contaminated substrate will fail again. Fixing the barrier layer and the rinse sequence usually solves it.
Why gold electroplate tarnish shows up on machined parts
Electroplated gold grows in a columnar grain structure. The columns do not always meet perfectly. Where they meet, you get micro-pores and grain boundaries that run all the way down to the layer below. On a 0.1 μm gold flash over copper, that porosity can reach the copper in hundreds of spots per square centimeter.
Copper under a porous gold layer reacts with sulfur and chlorine in the air. The reaction product migrates up through the pores and spreads sideways across the gold surface. What you see is a brown or dark film. The gold is still there. The copper is what changed color.
Nickel barriers behave differently. Nickel oxidizes slowly and forms a thin passive oxide. If the gold is thinner than about 0.1 μm, that nickel oxide can show through as a hazy, slightly yellow-grey discoloration. It is not corrosion in the structural sense, but it reads as tarnish on a visual inspection line.
Coating thickness sets the ceiling on service life
Thickness is the single most controllable variable. Below 0.1 μm, gold is a flash coat. It provides solderability and a gold color. It does not provide a continuous barrier. Pores are statistically guaranteed across a part of any practical size.
From 0.1 to 0.5 μm, pore density drops but does not reach zero. This range works for indoor electronics that see clean, dry air. It is common on connector shells and RF housings where the gold is mainly there for contact resistance.
Above 1 μm, and especially 2 to 5 μm, the layer closes up. Hard gold at 2–5 μm with a cobalt or nickel co-deposit is the standard for edge connectors and sliding contacts. It survives thousands of mating cycles and resists the wear that exposes the barrier layer.
Substrate and pre-plate finish decide the adhesion
Gold does not plate well directly onto aluminum or titanium. Both form a tenacious oxide within seconds of exposure to air. Platers use a zincate or a nickel strike to create a bondable surface first. If that step is rushed, the gold layer can lift or blister, and lifted edges trap electrolyte that later bleeds out as a stain.
Stainless steel 304 and 316 need an activation step, usually a Wood's nickel strike, before any gold goes down. Skip it and the coating may look fine on the rack and fail a tape test a week later.
The as-machined surface finish also matters. A Ra 0.2–0.8 μm polished surface plates more uniformly than a Ra 3.2 μm as-machined surface. Deep machining marks create recessed valleys where the local current density drops and the gold comes out thinner.
What the service environment does to the coating
Humidity, sulfur, and chlorine drive the reaction. An indoor climate-controlled enclosure at 40–60% RH is mild. A coastal plant with airborne chloride, or an environment near rubber and vulcanization equipment, is aggressive.
Sulfur is the specific problem for copper-based substrates. Rubber gaskets, some packaging foams, and certain lubricants release sulfur compounds over time. A part stored in a closed box with the wrong foam can show discoloration in weeks.
Temperature cycling adds mechanical stress. Gold and the underlying nickel expand at different rates. Over many cycles, the gold can crack at grain boundaries and open new paths for the substrate to react. This is why automotive under-hood electronics rarely use thin gold as the only protection.
When gold plating is the wrong answer
Gold is expensive and it is soft in its pure form. If the part is a large machined housing that only needs corrosion protection, anodizing on aluminum or electroless nickel on steel costs far less and will outlast a thin gold layer outdoors.
If the requirement is wear resistance on a sliding surface, hard gold at 2–5 μm works, but so does a hard anodized or a DLC coating at a lower cost per part. Choose gold when you also need low and stable contact resistance.
If the part will see salt spray, sulfur, or chlorine, no reasonable gold thickness is a substitute for a proper sealed enclosure. Gold is a contact material, not a marine coating.
How we control gold plating on machined parts
- 1Deburr and refine the surfaceBreak all sharp edges and hold pre-plate finish at Ra 0.8–1.6 μm or finer where the gold must be uniform. Sharp edges plate thick and can crack.
- 2Clean and activateAlkaline clean, acid pickle, then a strike layer matched to the substrate. Aluminum gets zincate, stainless gets a Wood's nickel strike.
- 3Deposit the barrierElectroless or electrolytic nickel at 2–5 μm. This is the layer that stops copper or steel from reaching the surface.
- 4Plate the goldSet current density and time to hit the target thickness. We measure with X-ray fluorescence on the actual part, not a coupon.
- 5Rinse and dryThree-stage rinse plus a hot DI rinse. Trapped salt at blind holes is a common source of later staining.
- 6Inspect and packVisual check, tape test on sample parts, then sulfur-free packaging with a desiccant where the part ships to a humid climate.
Gold plating thickness vs. application and tarnish risk
Use this as a starting point, then confirm against your test standard.
| Gold thickness | Typical use | Pore risk | Service environment |
|---|---|---|---|
| 0.05–0.1 μm flash | Solderability, cosmetic gold color | High | Dry indoor, sealed enclosure |
| 0.1–0.5 μm | Connector shells, RF housings | Moderate | Clean indoor air only |
| 0.5–1.0 μm | Signal contacts, low mating cycles | Low | Indoor, occasional humidity |
| 1–2 μm | General contact surfaces | Very low | Industrial indoor |
| 2–5 μm hard gold | Edge connectors, sliding contacts | Negligible | Wear plus humidity |
| 5 μm+ selective | Harsh chemical or marine | Negligible | Aggressive, with barrier |
Symptom, cause, and fix for gold electroplate tarnish
Match your symptom to the row before changing the spec.
| Symptom | Likely cause | Fix |
|---|---|---|
| Brown film after weeks | Copper sulfide through pores | Add nickel barrier, raise Au to 1 μm+ |
| Hazy grey surface | Nickel oxide, gold under 0.1 μm | Raise gold to 0.3 μm minimum |
| Dark spots in clusters | Trapped electrolyte at edges | Improve rinse, deburr before plate |
| Blister or flake | Weak strike on Al or Ti | Zincate or Ni strike, verify tape test |
| Discolor only on marks | Thin gold in machining valleys | Refine pre-plate Ra to 0.8 μm |
| Color shifts in storage | Sulfur from foam or rubber | Switch to sulfur-free packaging |
The short version
If you need stable contact resistance, specify a nickel barrier plus 1–3 μm gold. If you only need a gold color or solderability, a 0.1 μm flash is enough, but do not expect it to resist tarnish in a humid or sulfurous environment.
Gold electroplate tarnish questions
Does pure gold itself tarnish?
Bulk pure gold does not oxidize or sulfide at room temperature in normal air.
What changes color on a plated part is almost always the layer underneath, or residue on the surface, not the gold.
How thick does gold need to be to stop tarnish?
There is no thickness that guarantees zero porosity, but pore density falls sharply above 1 μm.
For a humid or mildly sulfurous environment, 1–3 μm over a 2–5 μm nickel barrier is a practical working range.
Can tarnished gold plating be cleaned instead of re-plated?
Light surface film can sometimes be removed with a mild non-abrasive cleaner, and the part may look acceptable again.
If the discoloration comes from the substrate through pores, cleaning only removes the visible film. It will return, and the part needs re-plating with a better barrier.
Why did the same spec tarnish on one batch and not another?
Rinse quality, bath contamination, and part geometry vary between runs. Blind holes and tight corners trap electrolyte.
Check the rinse sequence and the rack position of the failing parts before changing the thickness spec.
Does gold plating work on aluminum and titanium?
Yes, but not directly. Both need a zincate or nickel strike first, or the gold will not bond reliably.
Without that step the coating may pass a visual check and fail a tape test later.
What test confirms the gold thickness?
X-ray fluorescence on the actual part is the standard non-destructive method.
We measure at multiple points, including edges and recesses, because thickness varies with local current density.
Send us your part and we will check the plating spec
Upload your drawing and we will return a quotation and a free DFM analysis within 12 hours, including a surface finish and plating recommendation.
12-hour quote100% inspectionNDA on request
More machining and finishing notes
We publish setup notes, tooling trials and inspection data from the factory floor.
