Various methods of testing gold (and other precious metals)

Various methods of testing gold (including advanced methods used at the Birmingham Assay Office)

I visited the Birmingham Assay office for their special training course for touch acid testing.

They started by explaining that to test the fineness (carat) of precious metals for the purpose of hallmarking, the Birmingham Assay Office uses XRF Testing and the referee cupellation method (both explained below). Both of these methods are very accurate. They explained that before the introduction of XRF testing, they used Touch Acid testing to give an approximate indication of fineness of a precious metal and the homogeneity of a batch.

They then introduced the 'industry standard' tester: the Troytest 4 (which is manufactured by us, Quicktest).

After half an hour of theory (going through the instructions), we had a long practical session, testing dozens of samples. This was interesting. When we manufacture the testing sets here at Quicktest, we check the acids against a dozen known samples. On the course, I had the opportunity to test the acids against 45 known samples.

The people running the course spend their working lives testing gold (and other precious metals); the Birmingham Assay Office hallmarks between five and six million items per year, most of which are tested. And so I was curious to know if they used any techniques I had not come across. They did. And so here are some 'advanced' instructions, and also a list of other testing methods used by the Birmingham Assay Office for testing precious metals.


THE PRINCIPLE: if an item is gold plated and you test the surface, you will, quite correctly, get a 'gold' reading - because the surface is gold. So you must file / scrap the surface so that you are testing whatever might be underneath.
OUR METHOD: We give away a steel needle file with each set, you can file thoroughly over a very small area, which causes the minimum of damage. Also to be recommended is a diamond-impregnated file, it's harder, cuts deeper, and is more durable.
ASSAY OFFICE METHOD: they remove the surface with a 'scraper' - a three-edged steel tool (the edges of the scraper are sharp, it's used like a potato peeler). Because they are scraping a relatively large area (unlike using a fine needle file) there is no resulting 'notch', it's much more 'gentle', in skilled hands you can hardly see that the item has been 'scratched' *. The downside is that it takes practice to use, it's not as easy a needle file. However, I'm impressed and we now stock scrapers.
* This is the origin of the expression, "to come up to scratch".


THE PRINCIPLE: instead of placing the acid directly onto the item, you rub the item firmly across a special stone (touchstone) so that that it leaves a streak of gold, then test the streak with acid.
 a sample of gold is selected from several sets of sample 'needles' (each set has ten samples of different carats, each set is a slightly different colour). Select a set that best matches the colour of the unknown item, then guess what the carat might be. Starting with the lowest carat in the set of needles, rub it across the touchstone so that it leaves a gold streak; below this, streak the item to be tested so that you now have two streaks parallel to each other. Now select an acid from a set of 9 different strengths (mixed in the laboratory at the Assay Office) put an acid across both streaks (starting with the weakest acid) and see if either dissolves. The process is repeated, using increasingly stronger acids (and cleaning the touchstone between each test) until you see that the 'unknown' streak dissolves at the same rate as the known sample, then you know the two are exactly the same carat. The girl in the photo is the apprentice, she has been doing the job for six months and is making good progress.
SIMPLE METHOD: our method is the same as above but you will  have a choice of 6 test needles instead of 100, and two acid mixtures instead of 9, and you will not be supervised for six months while you learn. 
The testers we manufacture (including the TROYTEST, recommended by the Assay Office for retail use) are much simpler, they are designed to be used without the touchstone method, but you are welcome to buy a touchstone and set of test-needles if you wish.  
WARNING: some think that using a touchstone does not require the item to be filed (to remove any plating). Before you carry out a test you must 'streak' the item across the touchstone a few times to remove any plating, then wash the touchstone clean, then carry out the tests as above.


Gold plated items can be 'flash' plated (very thin) or 'hard' plated (very durable); gold can be rolled (rolled gold) by rolling two very thin sheets of gold around a central sheet of copper; white gold can be plated with rhodium to improve the colour (or to speed up the polishing process at the manufactures) - in all cases you must file through the surface to test the 'true' metal underneath. While this is all in my book, The Gold & Silver Buyer's Handbook, I came away with a curious cluster of extra facts.

Filing through very heavy gold plating with a steel file is hard work compared with using a diamond-impregnated file; XRF testers (below) cannot test through most plating (unless the plating is very thin) but they can be programmed to 'ignore' rhodium plating; large silver items covered in a very thick layer of silver are made by 'electroforming' - a wax shape is coated with silver nitrate then heavily plated with silver. In all of these cases you must file through the surface so that you are testing the metal underneath.



ICP is inductively coupled plasma mass spectrometry.

A tiny sample is removed, weighed and dissolved into the appropriate solution. The solution is then sprayed through an ultra hot flame where until it is destroyed in a burst of plasma and the resulting spectrum is analysed. This is the most accurate method of analysing precious metals to discover their exact chemical composition.

As you see from the pictures, these are not 'testing machines', these are laboratories and cost tens of thousands of pounds.


This is the referee method of assaying gold alloys. A small gold sample is weighed, silver added, and then wrapped in a lead envelope. The sample is melted at 1200C in a cupel. This drives off the base metals and leaves a ball of fine silver and gold. The ball is flattened and put into acid to dissolve the silver away, leaving a "cornet" of pure (24ct) gold. Then it's weighed again and the carat (percentage of pure gold) is calculated. Each test takes two and a half hours.


XRF is x-ray fluorescence, those who can't remember "XRF" say, "the x-ray machines".

XRF (cabinet). This is an example of a 'cabinet style' XRF tester but not one used by the Birmingham Assay Office. The Birmingham Assay Office's XRF instruments cost £50,000.00 each [that was a few years ago, today you can get small versions for as little as £15,000.00]. As with ICP (see above) it gives a detailed analysis of the chemical composition but, unlike ICP, it is totally non-contact. However, it will not usually test through plating (the best clue to plated items is inconsistent or unexpected readings) so you still have to file the item first.

XRF (handheld). Scrap dealers and antiques traders tend to use these 'cheap' XRF machines (handheld 'scanner' type), which you can now get find for as little as  £10,000.00 (these are not used by the Birmingham Assay Office).

Experienced users acknowledge that, even with these, care must be taken to ensure you are not testing surface plating (or rolled gold) or you will get incorrect readings.



It is not only the testing of precious metals that takes place at the Birmingham Assay Office, they test for nickel (high levels on nickel can trigger allergies, especially jewellery used for piercings); they test for lead (not found in precious metals but it used to be common in paints and in children's toys); they test for cadmium (in production, molten cadmium is very toxic); they mainly test costume jewellery, toys and fabrics, further information.