Old coin which fluoresces under short wave UV.

Since copper seems to be the key issue here I was fortunate to find a 10 Yen Japanese coin which also exhibits fluorescence under shortwave UV.
The composition is 95% copper, 3.5% zinc and 1.5% tin.

Cleaning with concentrated sodium chloride had very little effect in removing whatever causes the fluorescence.
The next step involved the effect of electrolysis.
Using a battery charger and the concentrated sodium chloride solution, the coin was used as the cathode and partially submerged in the solution.
The anode was stainless steel.

The chemistry of the electrolysis of sodium chloride is shown.

The fluorescence no longer occurs where the coin was submerged.

The theory behind cleaning coins using electrolysis is when the coin is the cathode, the production of hydrogen gas results in a mechanical action of lifting the patina off the surface.
Since the mechanical action of cleaning of the coin after soaking with sodium chloride solution had no effect I’m not convinced hydrogen gas removed whatever is causing the fluorescence.

It’s possible this mysterious coating was dissolved in NaOH (sodium hydroxide) produced during electrolysis or was plated over.
Whatever it is I’m not any closer to identifying it.

SelfSim said:

This material seems to be in the depressed areas of the pattern, yes?
I wonder whether it might have some closer relationship with whatever is being done with the dies used to create the imprint?

I think dies are typically made from hardened steel, which apparently become decarbonised when reannealed (see decarburization). This process alters the chemistry of the steel, too:

I see the dies are also polished every so often (maybe using the mysterious substance .. which then becomes embedded in the coin, once minted?)

I realise this may not help much .. but perhaps the substance has more to do with the die ageing (and carbon chemistry) than with the coin's?

Click to expand...

I agree with you the build up of the material in the depressed areas is related to the manufacturing process.
The question is in what part of the process and what is it?

sjastro said:

I agree with you the build up of the material in the depressed areas is related to the manufacturing process.
The question is in what part of the process and what is it?

Click to expand...

The 2000 year old Roman coin is the outlier here.
Coin production was probably no different to the ancient Greeks.

I wonder why fluorescence occurs mainly along the edge of the coin.

Not a necro-thread but a resurrected thread.
To recap some of the questions raised in this thread.
(1) Why does a 1947 Australian florin fluoresce under 254nm UV light but not 365nm UV. (Why does it fluoresce in the first place)?

The conventional answer is organic coatings or a mineral oil residue but in forensic engineering we use UV at 365nm as a spot check so this doesn't seem to be the answer.
Back in 2021 we didn't have AI as an aid, today GPT-5 gave the following explanation.

What you are seeing is real fluorescence, but it is not coming from the bulk Ag–Cu–Ni–Zn alloy itself. It is a surface‐controlled, short-wave–excited phenomenon, which explains both the colour and the wavelength dependence.

---

1. Why the coin fluoresces at ~254 nm (short-wave UV)​

At 254 nm you are delivering high-energy UV photons (~4.9 eV). That energy is sufficient to excite electronic transitions in surface compounds that exist on old silver coins:

Likely fluorescent species on the florin​

These are not metallic silver or copper, but thin surface films formed over decades:

• Silver oxide (Ag₂O) and sub-oxide states
• Copper oxides (Cu₂O, CuO)
• Copper carbonates (malachite-like traces)
• Zinc oxide (ZnO)
• Adsorbed organic residues (skin oils, waxes, cleaning residues)

Several of these have wide band gaps that only absorb deep UV:

Compound	Band gap / absorption
ZnO	~3.3 eV (strong SW-UV response)
Cu₂O	~2.1 eV, defect-related fluorescence
Ag₂O	defect-state luminescence
Organics	π→π* transitions <300 nm

At 254 nm, these materials absorb strongly and re-emit visible light, producing the green fluorescence you see.

---

2. Why it does NOT glow at 365 nm (long-wave UV)​

At 365 nm, photon energy drops to ~3.4 eV, which is below the excitation threshold for most of the above materials.

Key points:

• Pure metals do not fluoresce
• Silver–copper alloys quench fluorescence efficiently
• Oxide defect states need higher excitation energy
• ZnO is weak or non-emissive under 365 nm unless nanocrystalline

So at 365 nm:

• No electronic transition is efficiently excited
• Any weak emission is overwhelmed by absorption and surface quenching
• Result: no visible glow

This is why the coin appears “dead” under long-wave UV.

---

3. Why the glow is green​

Green emission (~520–560 nm) is characteristic of:

• ZnO defect fluorescence
• Cu²⁺-related defect centers
• Silver oxide surface states
• Organic residue fluorescence

The uniformity of the glow strongly suggests a thin surface film, not inclusions inside the metal.

---

4. Why alloy composition matters (but does not fluoresce itself)​

Your alloy:

• 50% Ag
• 40% Cu
• 5% Ni
• 5% Zn

Effects:

• Silver & copper → strong non-radiative quenching
• Nickel → suppresses fluorescence
• Zinc → forms ZnO, which can fluoresce

So the alloy:
✔ Creates the oxide chemistry
✘ Does not fluoresce intrinsically

---

5. Diagnostic conclusion​

The fluorescence is due to surface oxide and contamination layers excited by short-wave UV only.

This explains:

✔ Strong response at 254 nm
✔ No response at 365 nm
✔ Green colour
✔ Surface-confined appearance

---

6. How to confirm (if you want to go further)​

If you want laboratory confirmation:

• XPS – identifies Ag₂O, CuO, ZnO surface layers
• SEM-EDS – maps Zn and Cu enrichment at surface
• UV-Vis photoluminescence spectroscopy – confirms excitation threshold
• Gentle chemical cleaning – fluorescence will diminish or disappear

---

In short:​

The coin glows at 254 nm because short-wave UV excites defect states in surface oxides (especially ZnO and Cu oxides). At 365 nm the photons are not energetic enough, and the metal alloy quenches any weak emission.

Click to expand...

If you want, I can also explain why this effect is commonly used to authenticate silver coins and detect cleaning or coatings.

Click to expand...

I agree with its summation except organic coatings are generally more sensitive to fluorescence at 365 nm than 254 nm, so the conclusion is the coin had no organic coating or residue on it, and the florescence is mainly due to the patina or oxide layer.

The next question for the following post is:
(2) What about the Roman coin of the emperor Trajan which is 90% silver and 10% copper.
Why does it have an yellow glow around the edge?

2) What about the Roman coin of the emperor Trajan which is 90% silver and 10% copper.
Why does it have an yellow glow around the edge?

The yellow rim on the Trajan denarius under short-wave UV is a very characteristic and well-understood effect for high-silver Roman coins, and—again—it is not fluorescence from the silver–copper alloy itself.

---

Key facts to anchor the explanation​

• Alloy: ~90% Ag, 10% Cu
• Response:
  • Short-wave UV (~254 nm) → yellow glow at the edge
  • Long-wave UV (365 nm) → little or no response
• Glow location: edge / rim, not the fields

This combination immediately points to surface-segregated copper chemistry.

---

Why the edge glows yellow​

1. Copper migrates to the surface—especially at edges​

Over ~1900 years, Roman silver coins undergo:

• Selective silver enrichment of the surface
• Copper diffusion toward grain boundaries
• Preferential oxidation of copper, not silver

The rim and edge:

• Have higher defect density
• Are more exposed to burial chemistry
• Accumulate copper-rich corrosion products

So even a “90% silver” coin can have a copper-dominated surface layer at the edge.

---

2. Copper corrosion products fluoresce yellow under short-wave UV​

Several copper compounds are known to fluoresce yellow–yellow-green under 254 nm:

Compound	UV behaviour
Cu₂O (cuprite)	Yellow–orange defect fluorescence
CuO	Weak, but contributes
Copper carbonates (malachite-like)	Yellow-green
Copper chlorides (from burial salts)	Strong SW-UV response

These compounds:

• Absorb strongly below 300 nm
• Do not respond at 365 nm
• Are commonly concentrated at edges and cracks

This exactly matches your observation.

---

3. Why silver does not glow​

• Metallic Ag does not fluoresce
• Ag₂S (silver sulfide toning) is non-fluorescent
• Silver actually quenches fluorescence

So any emission you see must come from non-silver surface phases.

---

Why the glow is confined to the rim​

This is the most diagnostic part:

✔ Edge = more copper exposure
✔ Edge = more corrosion
✔ Edge = more defect states
✔ Edge = less silver quenching

The fields remain dark because they are silver-enriched and optically quenching.

---

Why it disappears at 365 nm​

• Copper corrosion compounds require higher excitation energy
• 365 nm photons (~3.4 eV) are below the excitation threshold
• Result: no visible emission

This is why the coin looks inert under long-wave UV.

---

What this tells you about the coin​

Importantly, this is a good sign:

• ✔ Consistent with ancient metallurgy
• ✔ Consistent with long burial
• ✔ Consistent with genuine Roman silver
• ✘ Not a modern coating
• ✘ Not lacquer or resin (those glow blue/white instead)

Numismatists sometimes use this exact edge fluorescence to distinguish ancient silver from modern replicas.

---

One-sentence summary​

The yellow glow around the edge of the Trajan denarius under short-wave UV comes from copper-rich corrosion products concentrated at the rim, which fluoresce at 254 nm but are quenched at 365 nm, while the silver itself remains non-fluorescent.

Click to expand...

If you want, I can also explain how this differs from modern silver cleaning residues or how burial soil chemistry changes the colour of the fluorescence.

Click to expand...

Well evidently this is a well known mechanism...... to an expert.
I had to ask it why stamping would cause copper to especially migrate to the edges which is in the next post.

(3) Why does copper especially migrate to the edges.

Copper does not “choose” the edge in a conscious sense — the edge lights up because physics, metallurgy, and corrosion kinetics all bias copper toward rims and boundaries over long timescales. On an ancient silver coin, the edge is where every relevant process overlaps.

Below is the clean, mechanism-by-mechanism explanation.

---

1. Metallurgical reason: grain boundaries intersect the edge​

Ancient silver coins were:

• Cast or hot-worked
• Cold struck between dies
• Never homogenised by modern heat treatment

This leaves a microstructure with:

• Silver-rich grains
• Copper preferentially segregated to grain boundaries

Grain boundaries:

• Have higher free energy
• Are fast diffusion pathways
• Terminate at the surface, especially at the edge

➡ The edge is where the most grain boundaries are exposed, so copper is already enriched there from the start.

---

2. Mechanical reason: striking concentrates defects at the rim​

When a coin is struck:

• Metal flows radially outward
• The rim experiences:
  • Highest plastic strain
  • Highest dislocation density
  • Microcracking and work hardening

These defects:

• Lower activation energy for diffusion
• Act as pipelines for copper atoms

➡ Copper migrates faster where the lattice is already damaged — the rim.

---

3. Electrochemical reason: copper is sacrificed first​

In a silver–copper alloy:

• Copper is anodic
• Silver is cathodic

In moisture, soil water, or air films:

• Copper preferentially oxidises:
  [
  \text{Cu} \rightarrow \text{Cu}^{2+} + 2e^-
  ]
• Silver is protected

Edges:

• Retain moisture longer
• Are thinner and more exposed
• Have higher oxygen access

➡ Copper at the rim is continuously drawn to the surface and removed as corrosion products, pulling more copper from below.

This is dealloying, not simple diffusion.

---

4. Chemical reason: silver enrichment pushes copper outward​

As copper is removed:

• The surface becomes silver-rich
• This creates a chemical potential gradient
• Copper from deeper regions migrates outward to re-establish equilibrium

Edges have:

• Shortest diffusion distance
• Highest corrosion rate

➡ Copper supply is preferentially drained from the rim.

---

5. Geometric reason: curvature accelerates corrosion​

The rim has:

• Higher curvature
• More exposed surface per unit volume
• More frequent micro-damage

Corrosion rates scale with:

• Surface energy
• Curvature
• Defect density

➡ Even tiny differences become dominant over centuries.

---

6. Why this takes centuries (and why modern coins don’t show it)​

This process is:

• Thermodynamically slow
• Diffusion-limited
• Corrosion-driven

It requires:

• Repeated wet/dry cycles
• Oxygen access
• Long burial or atmospheric exposure

Modern silver coins:

• Are homogenised
• Have cleaner alloys
• Have fewer defects
• Haven’t had 1,900 years

---

7. Why this produces fluorescence specifically at the edge​

Because the edge ends up with:

✔ Copper oxides
✔ Copper carbonates
✔ Defect-rich corrosion phases
✔ Minimal silver quenching

These phases fluoresce under short-wave UV, giving the yellow rim.

---

One-sentence summary​

Copper accumulates and expresses itself at coin edges because grain boundaries, strike-induced defects, electrochemical dealloying, and geometric exposure all peak at the rim, making it the fastest pathway for copper migration and oxidation over centuries.

Click to expand...

If you want, I can also:

• Sketch a cross-section model of a Roman silver coin
• Explain why cracks glow before fields
• Compare this with surface-silvered fourrées

Just say the word.

Click to expand...

Okay it has a LaTex related problem in expressing the reaction Cu -> Cu² + 2e⁻, I will have to do more research on this (on the mechanism not LaTex).
The next post is about I theory I had at the time about how electrolysis removed florescence from a Japanese Yen coin.

Doing electrolysis on a partially immersed Yen coin as the cathode in a salt solution removed the yellow fluorescence.



My theory was the hydrogen gas released at the cathode stripped the patina or oxide layer.
GPT-5 assessed this as being relevant but incomplete.

Yes — your understanding is essentially correct, and we can sharpen it with the electrochemistry involved.

Short version:

When the Yen coin was made the cathode, cathodic reduction plus hydrogen evolution chemically reduced and physically detached the oxide patina that was responsible for the luminescence. Once that layer was gone, the fluorescence disappeared.

Click to expand...

Here’s the precise breakdown.

---

1. What was fluorescing before electrolysis​

On a Cu–Zn–Sn alloy (brass/bronze-like), the fluorescence does not come from the metal itself but from the surface patina, typically:

• Cu₂O / CuO
• ZnO
• Mixed Cu–Zn oxides
• Possibly carbonate phases

Many of these compounds:

• Absorb strongly in short-wave UV
• Emit green–yellow luminescence
• Are only a few microns thick

This layer is electrically fragile.

---

2. What happens when the coin is the cathode​

At the cathode, two critical processes occur simultaneously.

(a) Electrochemical reduction (chemical destruction)​

Copper oxides are reduced back to metallic copper:

[
\text{CuO} + H_2O + 2e^- \rightarrow \text{Cu} + 2OH^-
]

[
\text{Cu}_2O + H_2O + 2e^- \rightarrow 2\text{Cu} + 2OH^-
]

ZnO and mixed oxides are similarly destabilised in alkaline conditions.

➡ The electronic structure required for fluorescence is destroyed.

---

(b) Hydrogen evolution (mechanical delamination)​

At sufficiently negative potential:

[
2H_2O + 2e^- \rightarrow H_2\uparrow + 2OH^-
]

Hydrogen gas forms:

• At the metal–oxide interface
• Inside pores and cracks in the patina

This causes:

• Local pressure buildup
• Blistering
• Lifting and flaking of the oxide layer

➡ This is a mechanical stripping mechanism, not just chemical.

You correctly identified this effect.

---

3. Why luminescence does not return afterward​

Once electrolysis is complete:

• The patina is thinner, reduced, or gone
• The surface is metallic copper-rich
• Fresh copper quenches fluorescence

Even if a new oxide forms later:

• It will be thin
• Less defect-rich
• Less emissive until aged again

So the loss of luminescence is permanent on historical timescales.

---

4. Why this effect is especially strong for Cu–Zn–Sn alloys​

Compared to silver alloys:

• Copper oxides are easily reducible
• ZnO is alkali-sensitive
• Tin oxides (SnO₂) are stable but non-fluorescent
• Hydrogen overpotential on copper is low → lots of H₂

➡ The patina is particularly vulnerable under cathodic treatment.

---

5. Important diagnostic point​

This is why conservators avoid electrolysis on coins unless absolutely necessary:

• It removes corrosion
• But also removes historical surface chemistry
• And destroys UV-diagnostic features

What you observed is exactly what professionals expect.

---

One-sentence summary​

Making the Yen coin the cathode chemically reduced its copper–zinc oxide patina and mechanically lifted it via hydrogen evolution, eliminating the defect-rich surface layer responsible for UV luminescence.

Click to expand...

If you want, I can also explain:

• Why anodic treatment behaves differently
• Why some patinas survive electrolysis better than others
• How long it takes for luminescence to re-develop naturally

Just say the word.

Click to expand...

Once again correcting the LaTex issues, I did not take into consideration the chemical reduction of the patina occurring at the cathode.
CuO + H₂O + 2e⁻ -> Cu + 2OH⁻
Cu₂O + H₂O + 2e⁻ -> 2Cu + 2OH⁻

Also at the cathode but in a different reaction is 2H₂O + 2e⁻ -> H₂ + 2OH⁻.
An electrode at the anode will corrode.