PatAR You indicate:
Quote:
most handheld XRF devices do not have sufficient precision to detect trace amounts of gold whether present in the coin or not.
First you need to know the power of the unit and the range of elements it can detect. Many hand held devices are only accurate for scrap metal applications while others are used in the field to disclose hazardous materials. The low power scrap metal XRF units may be of very little use for coins.
So my remarks need to be evaluated with regards to handheld units capable of detecting gold at a level of 0.01% - that is 100 ppm. These typically cost in the $3-5000 up range.
The reservations I have regarding the use of these types of handheld XRF testers is not with their ability to disclose trace gold but rather with their ability to penetrate deep enough into a coin's surface to avoid distortions in the ratio between silver and gold caused by silver enrichment at the surface. In addition the detection of very low density elements with most handheld units is impossible. The value of these units ends with gold and platinum. They will not give accurate readings for copper and silver and therefore authentication using those results Silver and copper will not be accurate.
Use of higher power XRF devices is far better than handheld units when it comes to identification of very small trace contaminants and to get the silver to copper ratio accurate enough to be used for authentication. However, the cost is often prohibitive ($50-500). So as a concession to cost, it makes sense to use the better grade handheld XRF as a screening tool to detect gold as a trace contaminant. Most decent handheld XRF devices will detect gold if it is present at 100 ppm or greater. This is a level well below the MINIMUM amount of gold that must be present in all Mexican coins struck before 1880. So a ZERO reading for gold on a handheld XRF can determine a coin is not a genuine strike from 1805. The same test will not produce a reliable ratio of silver to copper. Silver will always be OVERSTATED by an XRF gun and copper will be understated. These are just the facts of how an XRF gun works.
If you want to use a low power XRF the only value would be to determine very high levels of gold in excess of 1000 ppm. This does occur but so infrequently that the tests is almost a waste of money even at $5 a test (junk yard rates).
The lower the power used by the XRF device the shallower the signal return. However, the denser the element involved the stronger the signal return. These two factors are inverse properties. They are very critical when looking at results. Gold being the heaviest of the three object elements will return a signal from a point near 100 microns which is deep enough for authentication. Copper the least dense element will return a signal from only a fraction of that depth. This is too shallow to penetrate the silver enrichment layer which is found in all older issues that were acid washed at the mint. Silver is of moderate density but penetration past the enrichment layer is rather unlikely.
As an example using theoretical data the two XRF tests might return an alloy signature as follows for three Portrait 8Rs:
Coin 1:
Handheld XRF 94.11% Ag 5.21% Cu 0.01% Au
Laboratory XRF 90.143 Ag 9.337% Cu 0.015% Au
Results indicate POSSIBLY BUT UNLIKELY REAL - further tests advisable. What you see is a 94% silver 5% copper coin with only a 100 PPM gold level. The silver to copper ratio should not represent a problem to authenticity since acid washing or sea water exposure will remove surface copper while retaining the silver. There is some gold detected but not enough to be 100% certain the coin uses the correct alloy. A density test in this case would disclose a Sheffield plate counterfeit far more reliably. Looking at other smaller trace contaminants reported in a laboratory level XRF test might disclose a modern heavy metal that should not be present which would condemn a coin as a Numismatic Forgery.
Coin 2:
Handheld XRF 95.12% Ag 4.81% Cu 0.00% Au
Laboratory XRF 90.043 Ag 9.437% Cu 0.001% Au
Results indicate DEFINATELY NOT GENUINE The handheld test in this case shows a similar incorrect silver to copper ratio as seen with Coin 1 but no gold at all is detected at the 100 ppm level. The laboratory test is conclusive but redundant. The deeper penetration of the XRF beam almost corrects the silver copper ratio but it only confirms the limited amount of gold in the alloy a purity level that could not be achieved until the late 1800's. The inescapable conclusion is that the coin can not be a genuine strike from Mexico made before 1880. This determination can be based on the handheld test alone thereby saving the cost of a lab test.
Coin 3:
Handheld XRF 96.11% Ag 3.71% Cu 0.03% Au
Laboratory XRF 90.343 Ag 8.937% Cu 0.028% Au
Results indicate Alloy is CORRECT for period. COIN MAY BE GENUINE In this case the silver enrichment distortion is very evident in the handheld XRF result the worst silver to copper ratio of the three coins. The key point is the Gold result of 300 ppm. This is within the minimum level set for a genuine alloy seen in coins from Mexico before 1880. Once again the lab test in this case is confirming but unnecessary. The conclusion is ONLY that the alloy is correct - NOT that the coin is in fact genuine. Other analysis is needed to complete the authentication. Evidence like die style, method of manufacture etc.
Note the overall information to take away.
1. XRF tests must be accurate to 0.01 % to be genuinely useful. That level represents a 100 PPM base line. Absolute minimum level based on hundreds of lab tests is 200 ppm. Because extremely few coins test at this level some caution is advised in authentication of any coin that tests below 400 ppm. The average of all results to date is roughly 500 ppm. The highest to date (only seen in the Republican and War of Independence issues) is 3,800 ppm seen on an 1838 Guanajuato 8R.
2. The silver enrichment layer, typical on all coins that have been dipped or pickled in acid will render most silver to copper ratios inaccurate. The lower the power the greater the error. Lab tests can now be done to 10 ppm levels of accuracy but the test is extremely costly (ca $500 per test) so it would be reserved only for extremely valuable coins. Authentication based on the silver to copper ratio is best done using Specific Gravity.
3. The use of handheld XRF testers as an inexpensive screening tool is valid. The only components likely to be accurate are dense elements like gold and platinum. That is all that is needed to exclude most modern forgeries and many Class 2 Silver counterfeits made for the China Trade.
I hope that clears up any issues you may have.