The Opal That Must Not Be Heated
A client brings in an inherited ring for restoration: a 12 × 8 mm oval light opal cabochon, 3.8 mm high, in a worn 18 kt yellow bezel whose wall is split, on a shank worn down to 0.6 mm at the base. She wants the ring three sizes smaller, the bezel repaired and the whole piece re-polished before a wedding in ten days. The workshop has a laser welder, an ultrasonic tank, a steam cleaner, polishing mops and a full bench. Under a 10× loupe there is a hairline fracture running in from the girdle. The client says the last jeweller "just tightened it up" and it came back with a dull patch and a click when she taps it, and she is now anxious about the stone and asks that it never leave the ring.
1. State what may and may not be done to this stone, in terms of the material's properties, and explain the significance of the dull patch and the click.
Read the worked answer
Opal is hydrated amorphous silica, roughly 3–10% water, hardness about 5.5–6.5, no cleavage but low toughness and a conchoidal fracture habit. That dictates the limits:
- No torch heat. Water trapped in the structure expands and migrates on heating; combined with the low thermal conductivity of silica this produces heat-shock crazing. Even a short, localised flame can craze or cloud a stone that looks fine on the day.
- No ultrasonic. Vibration drives existing fractures and, on some stones, treatment oils or water out of the pores. It is the single most common way a fractured opal is turned into a broken one.
- No steam cleaner. Heat plus pressure plus an existing fracture.
- No pushing or hammering near the girdle. Stress goes straight into the fracture line.
- Careful with solvents and detergents on porous or treated material; never soak a doublet or triplet.
The dull patch and the click. Together they suggest the previous "tightening" either closed the bezel with enough force to start or extend a fracture, or heated the stone - the dull patch is consistent with crazing or with polishing compound driven into a surface-reaching fracture. The click is the sound of a loose or fractured piece moving. The stone must be treated as already damaged, and that must be recorded and shown to the client before any work begins.
2. The client insists the stone never leaves the ring. Sequence the job and justify whether you accept that condition.
Read the worked answer
You cannot accept it as stated, and the reason is arithmetic: sizing down three sizes means cutting and re-soldering the shank at 18 kt working temperatures, well above 700 °C. No technique in this workshop - torch, laser or otherwise - lets that happen safely with an opal 3 mm away. The only way to honour the request literally is to refuse the sizing.
Sequence I would propose:
- Document the stone, its dimensions, the fracture and the dull patch, with photographs, and have the client sign that the stone is accepted as already fractured.
- Remove the stone by cutting the bezel wall away from the outside with a fine saw or a graver, not by rolling or pushing the wall back. Pushing a bezel off a fractured cabochon is how the click becomes a break. Accept that the bezel will need rebuilding or replacing.
- Hold the stone in a padded box, in the client's sight if she wishes - the concern is about substitution, so transparency, witnessed sealing and a numbered receipt solve it. Offer to let her watch the removal.
- Do all metalwork: shank reduction, new or repaired bezel, solder, pickle, emery, polish. No stone anywhere near the bench.
- Re-set last, with a polished pusher, progressively, checking the fracture after every pass under the loupe.
- Final polish around the setting, not over it.
If the client still refuses removal, the correct professional answer is to decline the sizing in writing rather than risk the stone.
3. Describe how you would size the ring down and what quality checks and written statements you would make before the ring is returned.
Read the worked answer
Sizing down with the stone out. Three sizes is a reduction in internal circumference of roughly 3 mm, so a wedge of that length plus a small allowance for the saw kerf and for closing is removed from the base of the shank - not squeezed shut, which distorts the bezel seat. Because the base is already worn to 0.6 mm, the honest answer is a half-shank replacement in matching 18 kt, not a cut-and-close on metal that thin. Solder with a hard 18 kt solder, anneal, round on the mandrel, file, emery through the grades, polish. Check the size on a ring gauge at the correct point, not with calipers across a distorted ring.
Bezel rebuild: close or replace the split wall, then fit to the actual stone before any setting.
Quality checks before return:
- Size confirmed on a ring gauge.
- Bezel wall even, stone seated with no rocking; push test and loupe check that the fracture has not extended.
- No pits, porosity or solder lines visible at the joint.
- Weight recorded before and after; photographs before and after.
- No compound residues around the stone.
Written statements: the pre-existing fracture and dull patch are recorded as pre-existing; the stone was removed and re-set with the client's knowledge; the ring has been re-polished, so any original surface detail or engraving is reduced; the sizing changed the hallmark position, so the piece must go for re-assay. Deliver with that document signed by both parties.