Cleaning

The step where the most specimens are improved and the most are ruined, frequently in the same session.

Least aggressive first

Start with the gentlest method that could work and escalate only if it does not.

Every method removes something. The question is always whether it removes only what you meant.

And test on the worst specimen from a group, never the best.

Water, and which water

A soak and a soft brush handles most field dirt.

Room temperature, and let cold material warm up first: a cold specimen in warm water cracks.

Rinse finally in distilled or deionized water. Tap water in a hard-water area deposits its own dissolved solids as it dries, which is a salt problem you added.

And know which species cannot be washed at all: soluble sulfates, halides, and anything already showing decay. Water accelerates pyrite oxidation.

Mechanical

Soft brush, then dental picks, then ultrasonic, then air abrasive, in that order of aggression.

Ultrasonic is efficient and it will shake apart anything with a cleavage, a fracture or a delicate attachment. Not for fragile specimens, and not for anything repaired.

Air abrasive is a specialist tool that removes matrix precisely and will erase a crystal face in a second if it drifts.

Chemical, and the label rule

Dilute acids for carbonate coatings, oxalic acid for iron staining, and each will destroy some species while cleaning another. Identify the specimen before choosing a reagent.

Protective equipment is not optional, and neither is ventilation.

Then the rule this whole section exists to state: the label does not go to the sink. The single commonest way locality data is lost is a specimen going to be cleaned while its label stays on the bench, and three specimens coming back to two labels.

Work one specimen at a time, or keep each label in the labeled tray its specimen came from.

References