Questions about XRF analysis

[kirispupis]

Hello everyone,

Recently, I acquired (though am still waiting on shipping) a very interesting rare coin. I plan to post something much more detailed when I receive it, but I believe the city it's been assigned to is incorrect. I'll get into the gory details later, but my argument boils down to the following:

• The reverse has the first two letters of the city and per Barrington Atlas there were about 30 cities beginning with those letters
• Neither the obverse nor the reverse resembles any other issues from the city it's attributed to
• This coin is unpublished and I was unable to find any resources mentioning finds
• But, the reverse is exactly the same (except for the city name) as two other cities' issues
• Those cities are only a few kilometers from each other, and there happens to be another city that starts with the same two letters next to them

I also acquired the similar type from another of these cities. The reason I'm curious about an XRF scan is that - since these cities were only 4km from each other, I would expect them to have used the same source for their bronze. Therefore, if I'm correct, I would expect their XFR analysis to be similar, but an XFR analysis of other coins from the currently attributed city (I own several) - which lies 2000 km away - to be different.

If this theory is correct, then there's another coin - from a very different area - that I'm debating to buy. I'm hesitant because it's very expensive. It contains the full name of the city and there are only three cities in Barrington that match. Two are known to have minted coins and none of their issues resembles this at one at all. For the third city, two of their next door neighbors issued bronzes with identical reverses save for the city name. The weight is also identical. However, I've read the respected works on this area of the world and none mention any coinage from this city until Roman times. Unlike the previous coin where I found other examples sold, this is the only example of the type I could find.

If the XFR theory is correct, then a similar test could suggest whether this is the first known issue from that city during this time, or if it's a fantasy coin.

Therefore, I'm wondering the following.

1. Can XRF analysis provide the information I seek?
2. So far, the cheapest XRF I've seen was $23k (but...free shipping!) Are there more economical options?
3. Are there any labs that have XRF machines that I can take the coins to in lieu of buying a machine myself?

Note that all the coins I'm interested in right now are bronze, though I would eventually be interested in silver too.

Edited September 21, 2023 by kirispupis

[David Soknacki]

When I wanted to have some coins tested with XRF equipment, I went to a local lab that tests metals for mining. I found that these labs have current, precise equipment, and typically the service is quite cost effective.

[Rand]

A university lab may be of help as well. I plan to analyse my coins. Last time I checked, the price was £50 per sample. I am sure it would be cheaper to analyse, say, 100 coins. 

The trouble with XFR is that it analyses surfaces - bare metal must be exposed for accurate data. This has stopped me so far as I do not want to remove patinas, even though there is not much patina on the gold coins I am interested in. 

I discussed the possibility of using metal content to help coin attribution by linking to metal sources with an archaeologist. Recycling is a major issue for gold coins for such analyses. This may be different for bronze, which I do not know.

[kirispupis]

40 minutes ago, David Soknacki said:

When I wanted to have some coins tested with XRF equipment, I went to a local lab that tests metals for mining. I found that these labs have current, precise equipment, and typically the service is quite cost effective.

Thanks! I called around and found that there's one local lab that can do XRF testing. However, they want $200 per test - so it would be $600 just for this one validation. :( 

[John Conduitt]

I'm not sure you'd get an answer anyway. Would the alloy used be that consistent to provide a match, even if they went to the same copper mine each time and never melted down old coins?

[robinjojo]

My understanding of XFR analysis is that it can only penetrate around 100 microns (0.1 mm), so the compositional data might include impurities on the coin's surface.  I don't know if this analysis would be helpful in you attempt to link the two coins. It's an interesting idea though.

Edited September 21, 2023 by robinjojo

[Dwarf]

54 minutes ago, kirispupis said:

Can XRF analysis provide the information I seek?

I doubt whether this will work for bronze.
It is a mixture of different metals, coming from different mines around "the world".
Even older coins or broken objects may have been melted to produce the flans.

But you may try

[kirispupis]

32 minutes ago, Dwarf said:

I doubt whether this will work for bronze.
It is a mixture of different metals, coming from different mines around "the world".
Even older coins or broken objects may have been melted to produce the flans.

But you may try

Thanks. I hadn't thought of this, but you're right. It was common in antiquity to recycle statues and other bronze artifacts for coinage. Therefore, a similar XRF reading would "suggest" a similar location, but a different one would say little.

[CPK]

That's a very interesting idea @kirispupis. Now I'm curious about your mystery coins!

[kirispupis]

1 hour ago, CPK said:

That's a very interesting idea @kirispupis. Now I'm curious about your mystery coins!

As am I. 🙂 Hopefully the seller will get around to putting it in the mail and I can post it here!

In the meantime, I've thought of another idea, though it's a long shot. The city in question is currently being excavated. I read that they found hundreds of bronze coins, but I haven't seen any report on which types. I may be able to find the email address of one of the archeologists and send a photo of my coin and ask if he/she's seen any at the site like it.

If that by some odds works, then it would be pretty difficult for any numismatist to disagree with my attribution...

[Deinomenid]

4 hours ago, kirispupis said:

Can XRF analysis provide the information I seek?

I'm not sure it can. If the information you seek is that the  bronze was produced from materials that are so  far apart (the 2k km you mention)  that they have highly  distinct markers usually they don't. For example to try to solve this a  load of Roman Republican SILVER coins were lead  isotope tested (not xrf) and the results were almost incoherent even to the specialists. Whether silver or bronze  isn't  key I believe as lead is the component sought and its present in both types. There  have been some papers on it  just recently where after endless  complex equations and PhD's  pouring over the data some  method  sort of emerges but it's way over a generalist's  (aka  my) head. (I only know this because one way to attempt to work out the age of Aegina staters is to check for Laurion silver in them rather  than Siphnos  silver on the  assumption that Laurion  didn't start production until x, and Siphnos suffered disaster in y etc). 

And this is all  before  Dwarf's extremely valid  point about melting/recycling etc.

This is one cleaned up "good" example from a  Roman coin, presumed location. They concluded though that 99 samples was far too small to be  of statistical use.

 

 

 

Please no Mark Rubio "I'm not a scientist but" memes  directed at me! I know little of this...I just wanted to  know the age of my Aegina turtle  🙂

[ela126]

There are different levels of preciseness and outputs, of XRF as with any analyzer. I have gone to simple gold buyer shops with my coins, and they freely did XRF scans on them (literally takes 5 seconds) and can give a basic Gold, Silver, Copper output, with percentages of each. This is because it's all the spectra (might not be the right term, that's for IR sourced analyzers) loaded into the XRF due to their interest. This one situation was an Olympus model XRF.

On Oak Island as many of us may have seen, they are able to get considerably more outputs with their unit. Lead, nickel, mercury, etc. As you mention, these are likely university level analyzers which are considerably more expensive. (i'm a analyzer salesman for gaseous compounds for my day job) So, if you're ok with very basic percentage copper value on your bronze, you may be able to do it for free.. if you want a good breakdown, it sounds like you've found it's expensive. (dont even toy with buying a university level resolution analyzer, likely $100k+)

Edited September 22, 2023 by ela126

[Ed Snible]

Many pawnshops and "we buy gold" places have these machines.

I visited more than 20 of them.  None of them would analyze for a fee, or let me operate the machine.  (Three shops would do it for free for one or two coins.)

I had several coins analyzed in pawnshops and a couple of dozen more at the house of a colleague who has one of the machines (who lives 5 hours away from me).  I ended up with XRF results for about thirty coins with no cash outlay.

The machines produced consistent numbers for the silver coins (front vs back, different parts of the coin).  As expected, placement made a big different with fourrees.  I only scanned one bronze coin.

I don't want to give away when I am writing about next, but here is a chart of XRF results:

I took some silver coins and measured copper, gold silver, and lead content and calculated ratios and plotted them on this triangle graph.  Each shape/color is another kind of coin.  I wanted to see if the coins in blue and the coins in black are from the same mint.  I was able to show that the black examples have more gold, but it isn't super convincing that they have different origins, because there are some black points with low gold content and some blue points with high gold content.

With two coins you will have a hard time being sure of the results.  I didn't have enough coins to get good results.  The idea is to get a lot of coins and use statistics.  If you are mathematically inclined read this paper by professor Kenneth Sheedy: https://www.academia.edu/44438852/Studies_in_Athenian_Silver_Coinage_Analysis_of_Archaic_Owl_Tetradrachms_MiN6_pp_207_214

He got a grant to take an XRF gun to various museums around the world and point it at their Athenian tetradrachms and write papers like these with the results.

 

Edited September 22, 2023 by Ed Snible

[kirispupis]

Thank you @Ed Snibleand @Deinomenidfor the specific answers. In summary:

• It's uncertain how reliable the XRF scan will be because it only measures the surface of the coin, which may have a thick patina
• Even if I get through the patina, bronze coinage was notoriously created from melted down objects that could have come from anywhere
• Even if the bronze sources were the same, a "standard" metal tester used at pawn shops isn't going to provide the precision necessary
• Even if I have the precision necessary, this requires a large number of coins to make any sense

I'm therefore not going to pursue the XRF angle. 

Both of these coins are very rare. For coin #1 (the one I bought), there are two examples on ACSearch - mine and one from the Plankenhorn collection. For coin #2, that is the only example.

I'll be very excited for the seller to send out my coin #1 and for others here to lend their opinions on my attribution. I received my coin from the neighboring city yesterday so I can put them side-by-side.

Edited September 22, 2023 by kirispupis