A real observation kicked this off: wiping laser-coloured titanium with isopropyl alcohol left it permanently duller — not shifted to a different hue. That single, specific detail rules some explanations out and points straight at others. This page works through the physics, six competing theories for what's actually happening, and the tests that would tell them apart.
The Observation
Every titanium interference colour comes from the same nanometre-thin oxide film — see Titanium Colour Physics for how anodizing and laser marking both grow it. So why did three different surface treatments do three completely different things to that same film?
| Observation | Implication | Confidence |
|---|---|---|
| Clear varnish preserved some colour but reduced the metallic, vivid look. | A conventional clear coat likely adds a thick optical layer and surface reflection that mutes thin-film interference. | High |
| Wax changed colours dramatically — gold shifted toward blue/purple. | A residual wax layer can become part of the optical stack and change the perceived hue outright. | High |
| Isopropyl alcohol (IPA) permanently dulled a laser-coloured piece. | The laser finish may include a fragile texture, removable residue, or porous/mixed oxide sensitive to solvent + wiping. | High (observation) medium (mechanism) |
| The IPA effect was dulling, not a major hue jump. | Loss of gloss/specular reflection, or removal of an index-matching residue, is more likely than a large change in oxide thickness alone. | Medium |
Foundation
Titanium oxide colours are structural, not pigment — there is no dye to protect. That's exactly what makes them fragile in a way people don't expect.
Competing Explanations
None of these is confirmed. They're listed in no particular order of likelihood — the point of writing them out explicitly is that each one predicts a different, checkable outcome, which is what the diagnostic guide below is built on.
IPA removes a thin transparent or semi-transparent residue that was adding optical thickness to the stack.
The residue fills microscopic surface valleys and raises specular reflection, making the colour look glossier and more saturated without changing the hue itself.
Wiping — not the solvent — abrades, collapses, or rearranges a fragile laser-formed surface layer or nanostructure.
IPA enters pores, extracts soluble residues, then redeposits a dull film — or changes wetting behaviour in a persistent way — even with no wiping at all.
Laser processing creates TiO₂ plus suboxides, or a graded oxide microstructure, whose response to solvent and handling varies by exactly how it was made.
The mirror image of Theory A: instead of an existing residue being removed, something new — a fingerprint, a protective oil, a wax — gets deposited on the surface, adding its own optical thickness.
Telling Them Apart
The practical question underneath all six theories is simpler than it looks: was the lost residue mainly smoothing the surface (theory B), mainly adding optical thickness (theory A), or some mix with mechanical damage (C) or solvent chemistry (D)? These five diagnostics separate the two ends of that spectrum.
| Diagnostic | If residue was mainly smoothing | If residue was mainly optical thickness |
|---|---|---|
| Point-light reflection | Becomes fuzzier / less intense, similar hue. | Stays sharp, but hue changes noticeably. |
| Viewing-angle sweep | Angle dependence stays similar; saturation drops. | The whole colour-vs-angle behaviour shifts. |
| Reapply removable liquid | Mineral oil, glycerin, or Krytox restores brilliance while present. | A liquid shifts hue according to its own thickness/index, not simply restoring gloss. |
| Spectrometer | Peak intensity drops; peak position barely moves. | Reflectance peak moves in wavelength. |
| Microscopy | Surface looks rougher, hazier, or pitted after cleaning. | Surface looks similar; optical response still changes. |
If you ran one of these tests on your own scrap and got a result below, click it — this maps straight back to the theories above and says what to try next.
Options
Every candidate below is judged on the same axis: does it protect, and how much does it cost you optically? As a rule, the thinner the layer relative to the oxide film itself (tens to hundreds of nm), the less it disturbs the interference colour underneath.
| Option | Typical scale | Likely optical effect | Fit for this project |
|---|---|---|---|
| No coating | None | Preserves original colour best; no added optical layer. | Best baseline and often best appearance. |
| Fluorosilane monolayer / anti-fingerprint coating | ~1–5 nm | Least visible change expected — molecularly thin, low-index. | Most promising, if obtainable and compatible. |
| PFPE oil (e.g. Krytox GPL 105) | Nanometre-scale if wiped nearly dry | Possible slight hue shift; may improve gloss by filling roughness. Not a cured hard coating. | Good reversible experiment on scrap. |
| Microcrystalline wax (Renaissance Wax) | Tens–hundreds of nm | Moderate hue shift possible; can reduce sparkle. | Useful comparison — but observations above suggest visible colour change. |
| Automotive ceramic coating | Thicker than a monolayer; formulation-dependent | May still alter hue, gloss, or saturation. | Worth testing only on coupons; many need solvent prep. |
| Acrylic / polyurethane varnish | Multi-micron to tens of microns | Significant reduction in brilliance; plastic-like reflection. | Generally not recommended. |
| Cerakote clear / Elite / Micro Slick | Multi-micron | Likely mutes metallic interference; may scatter light. | Only if durability matters more than exact colour. |
| ALD alumina, thin SiO₂, or optical DLC | Nanometre-scale, professionally controlled | Can preserve optical response better than varnish if uniform. | Best industrial route, not the easiest hobby route. |
Product examples mentioned in the source discussion: Daikin OPTOOL DSX/DSX-E and Shin-Etsu KY-series (industrial optical anti-fingerprint coatings — strong conceptual fit, not typically consumer-available); FOTS and FDTS silanes (lab-grade self-assembled monolayer reagents, need careful handling); Krytox GPL 105 (easiest to buy in small amounts, but many listings are repackaged for keyboard hobbyists); Renaissance Wax (accessible baseline, likely to shift interference colours); consumer ceramic coatings like CarPro CQuartz or Gtechniq EXO (worth testing, not equivalent to a true fluorosilane monolayer).
Handling
Method
Make several matching titanium coupons at the same laser settings, ideally spanning a colour gradient or multiple colour bands (see the reference chip-card approach used on the Titanium Colour Physics page). Keep one untouched control, then divide the rest into masked test strips.
| Strip | Treatment | Question answered | Expected result if theory is correct |
|---|---|---|---|
| A | Untreated control | Baseline colour, gloss, angle response. | Highest-fidelity reference for every comparison. |
| B | Dry optical wipe only | Is mechanical contact alone enough to dull it? | Dulling here supports Theory C — fragile texture / mechanical disturbance. |
| C | IPA drop, no wiping | Does solvent alone alter the surface? | Dulling or rings support Theory D — solvent extraction / redeposition. |
| D | IPA wipe, controlled pressure | Does solvent + wiping reproduce the original dulling? | Dulling only here supports a combined solvent + mechanical mechanism (C + D). |
| E | Krytox GPL 105, wiped nearly dry | Can a very thin PFPE layer restore brilliance? | Restored gloss supports Theory B; a hue shift instead supports Theory F — the Krytox itself is adding thickness, not just filling roughness. |
| F | Renaissance Wax, buffed | How large is the wax-induced hue/gloss change? | Visible shift supports wax becoming part of the optical stack (Theory F). |
| G | Consumer ceramic coating, uncleaned scrap | Can an accessible cured coating preserve enough brilliance? | Good result: durable, only slightly shifted. Bad result: muted / plastic. |
| H | Fluorosilane / optical anti-fingerprint coating, if obtained | Does a molecularly thin coating preserve colour while resisting fingerprints? | Best candidate: minimal hue/saturation change, improved repellency. |
Where This Leaves Us
The most promising protection route is a true optical anti-fingerprint fluorosilane or similarly molecularly thin coating, provided it can be applied without a damaging solvent-prep step. The most accessible low-risk experiment is Krytox GPL 105 on scrap, applied extremely thinly and buffed with minimal pressure. The strongest mechanical coatings — Cerakote and conventional clear coats — are likely to change the appearance substantially, and should only be used when durability matters more than exact colour.
Unresolved
Terms used across this page, in the order they're most useful to know.
| Term | Meaning |
|---|---|
| Structural colour | Colour produced by physical interference of light, not pigment or dye — everything on this page and on Titanium Colour Physics. |
| Optical thickness | Physical thickness × refractive index — the single number that actually determines hue in a thin-film interference system. |
| Optical stack | The full sequence of layers light passes through and reflects off — air, any coating, the oxide film, the metal. Adding a coating always creates a new stack, not a passive layer on an unchanged colour. |
| Index matching | Filling microscopic surface roughness with a liquid or film of similar refractive index, restoring specular reflection without necessarily changing the underlying colour. |
| Redeposition | Solvent picks up dissolved material and leaves it behind elsewhere as it evaporates, often as a haze or ring rather than a clean removal. |
| PFPE (perfluoropolyether) | A class of chemically inert, low-surface-energy oils (e.g. Krytox) used here as a thin, reversible test layer. |
| Fluorosilane monolayer | A molecularly thin (1–5 nm), covalently bonded anti-fingerprint coating — the category with the best chance of protecting without disturbing the interference colour. |
| ΔE (colour difference) | A standard numeric measure of how different two colours appear — see the Colour Stability Simulator for how researchers track this over time on laser-marked steel. |
Continue Learning