Example: The Process of Analyzing Two (Modern) Coins

Above: A Viking silver penny circa 920... Or is it a fake?
Ancient coins were one of the first artifacts to be analyzed using instrumental techniques that arose in the 1950s and 1960s. One of these early researchers was C. M. Kraay, who used neutron activation analysis to measure silver and gold in Greek coins in the hope that differences would be detected between different mints. In 1958, his article in the journal Archaeometry describes his findings. For instance, the data showed the “owl” coins of Athens (circa 500-480 BCE) have uniformly low percentages of gold and copper, whereas earlier Athenian coins (circa 570 BCE) have less uniform compositions and usually greater amounts of gold and copper. Historical sources reveal that Athenian coins between 500 and 480 BCE were made from deep-mined silver from Laurium in Attica. Kraay hypothesizes the more variable composition of those earlier coins was due to either less homogenous surface deposits at Laurium or use of silver from a different source. The latter possibility is likely “the north Greek mining districts in Macedonia and Trace, with which Athens had been in touch during the 6th century, came under Persian control" around 512 BCE. He also found that Sicilians added copper to their coins, possibly to increase durability.
Kraay's work was influential. A few years later, Wyttenbach and Hermann (1966) explained why the composition of ancient coins should be of interest. They assert the major elements “will answer questions as to the standard of the coins, to the policy followed in setting these standards, and to the economic history of the society.” Minor and trace elements, on the other hand, may be able to "shed light on the origin of the metals used, on the development of the refinery processes used, and on the authenticity of coins.” Archaeologists took notice of such useful applications of coinage analyses, and other archaeological research projects soon followed.
For instance, Carter (1971) suggests that an “ancient coin bears a record, more or less determinable, of the ore from which the metal was obtained, the method by which the metal was won and alloyed, the economics... of the period in which it was made, and its time and place of manufacture.” He utilized X-ray fluorescence to examine copper-based Roman coins minted from 200 BCE to 400 CE. He contends “it is important to determine whether the composition of copper coins change from one period to another” and that it is also significant to determine “the policy of the emperor towards copper coinage in the colonies.” Carter sought to answer: “Were minting procedures in colonies similar to or difference from those used in Rome at about the same time?” His results suggested that “Augustus instituted radical changes in the styles, compositions, denominations, and policies” of Roman coinage and “Lugudunum was the most important mint outside of Rome” to Augustus and Tiberius. The data from his analyses also allowed him to state that "compositions of coins definitely may be correlated with time and mints; . . . individual groups of coins minted in 1 to 6 year periods indeed do have similar compositions and each group has a unique composition.”
In Science and Archaeology, Gordus (1971) contends “of all of the artifacts of ancient and medieval cultures that are extant today, coins are by far the most prevalent.” He also claims that “in spite of the abundance of coins, historians have tended to neglect this source of data and rely almost completely on written records.” Gordus argues that most numismatic studies have been limited to “the examination of the designs, legends, mintmarks, dates, denominations, and weights of the coins,” but "what is in the coin... determines its value as a monetary unit.”
Showroneck and Houck (1990) used SEM-EDS to identify unrecognizable coins from the Marquette Mission site, a North American contact-period archaeological site. They explain conservators "identified two otherwise featureless metallic disks from the Marquette Mission site as coins,” and the chemical analyses "revealed coins to be sixteenth-century douzains -- the oldest identified French coins in North America.” They ask: “How many other tightly datable artifacts exist in uncleaned and unconserved contact era collections?” How many outwardly featureless coins have “been dismissed as capable of yielding no further information”?
As you've no doubt noticed, none of the studies listed here used an electron microprobe (although Showroneck and Houck did use a scanning electron microscope equipped with a energy-dispersive X-ray spectrometer). It turns out, though, coins were among the first artifacts to be studied using a microprobe.
In 1960, G. Roberts reported in the journal Archaeometry that an electron microprobe was under construction at the Oxford instituted the Research Laboratory for Archaeology and the History of Art and soon would be complete. The microprobe, he wrote, would first be used to study surface enrichment in coins. In that same journal two years later, Hornblower discussed this work on coins and other metal artifacts, and he hoped their microprobe would be able to identify ore sources, manufacturing methods, and even individual smiths.
Today, however, coins have symbolic value, not value based on any precious metal content. This, of course, is also valid for paper money (or banknotes). The "silver dollar" hasn't had any silver in it since 1971. Most coins are made from common alloys, and the face value of the coin is higher than the worth of its metal content. Banknotes are only paper and ink. Value comes from a guarantee by the issuing government.
This brings us to the samples at hand.
Example #1 is a replica Viking silver penny (also known as St. Peter of York money). This replica coin appears at the top of this post. How do I know it is a replica, not genuine? This is an easy case with two major clues.
First, the dark "tarnish" in the grooves of this coin appear to be artificially corroded:

Above: A secondary-electron (SE) image show topographic details of the sword of the replica coin.
Below: A backscattered-electron (BSE) image show compositional variations within the same area.

Second, and more conclusively, the EDS spectrum shows the metal is tin (Sn), not silver:

This is a simple example of identifying a modern replica or forgery -- it is not always this straightforward.
Example #2 is also a modern, not archaeological, coin:

This contemporary coin is composed of two different metal alloys. Such a coin is known as "bimetallic" by collectors although each alloy contains several metallic elements. There are a few ways to manufacture two-component coins. The most common is simply smashing an outer ring and an inner disc together as they pass through a coining press which also stamps a design onto the surface -- this process is known as "striking" a coin.
Let's use this modern coin as a proxy for an archaeological coin (or other metal artifact, like a bronze axe head -- but you might actually see analyses of such an axe head here soon). Using this coin, I can show the basic processes of identifying and analyzing an alloy artifact using an electron microprobe. A notable difference in examining a modern coin (which makes this example simpler) is that we know there is no alteration of this coin's surface. A coin that had been buried for some time would likely have diagenetic (post-depositional) surface alteration. This coin, though, will have an unaltered surface (Or will it?) so this demonstration avoid this issue (but I might be allowed to post the data from those bronze axe heads here, and that project is focused on this problem).
I first collected an X-ray spectrum using the EDS (energy-dispersive spectrometry) system in order to determine the basic composition of the inner disc. The resulting spectrum is below:
Below: EDS spectrum of inner disc; Cr and Fe characteristic X-ray peaks are apparent

The only apparent X-ray peaks correspond to chromium and iron. This suggests that the inner disc is stainless steel (also called "chromium steel"). Carbon, a key element in steel, is too light to be seen in an EDS spectrum. We must use our WDS (wavelength-dispersive spectrometry) system to measure the carbon content within steel. Manganese commonly is added to steel, but one can see below the chromium and iron X-ray peaks could easily obscure a small manganese peak in the EDS spectrum. Such peak overlaps are often a problem with EDS analysis, and WDS scan must be used to determine if a small amount of manganese is present or not.
Below: EDS spectrum of inner disc; are the Cr and Fe peaks hiding a small Mn peak?

Below are scans using our five WDS spectrometers -- each spectrometer covered a different portion of the X-ray spectrum. The largest X-ray peaks belong to iron and chromium, just as in the EDS spectrum. A number of tiny, previously unobserved peaks are also visible because WDS has better resolution and detection limits than EDS:
Below: WDS spectrum of inner disc; minor and trace elements are now apparent

The WDS spectrum above shows that nickel, vanadium, silicon, carbon, and a little oxygen are present. Oxygen is present due to a thin oxidized layer, and the other four elements are common within steel. We will keep all of these elements in mind for later when we conduct quantitative analyses of this alloy.
Now, let's look at an EDS spectrum collected on the outer ring:
Below: EDS spectrum of outer ring; Cu and Zn characteristic X-ray peaks are obvious

The copper and zinc peaks indicate this alloy is some type of brass. Quantitative analyses will be needed in order to determine the exact type of brass. There are also two other small X-ray peaks:
Below: EDS spectrum of inner disc; Cl and Ca peaks

The chlorine and calcium peaks probably represent calcium chloride (CaCl2) that is bound to its surface. This salt is water-soluble and has a wide variety of applications: road salt, desiccant, additive in plastic and fabric softener, food preservative, electrolyte in sports drinks, and more. Calcium chloride will corrode metals, especially brass, so it isn't surprising to find a small amount bound to the surface. The coin must have come in contact with calcium chloride at some point during its life: in a pocket, on the street, through the laundry, etc.
As with the outer ring, WDS analysis can reveal minor and trace elements in the inner disc:
Below: WDS spectrum of inner disc; minor and trace elements are now apparent

These scans reveal the presence of nickel, sulfur, potassium, silicon, aluminum, and magnesium. Aluminum makes brass stronger and more resistant to corrosion. Silicon is used to reduces wear, nickel reduces corrosion, and it isn't surprising to find traces of sulfur, potassium, and magnesium in brass. These six elements will also be added to our checklist of elements to be measured during our quantitative analyses on this coin.
Next I wanted look for evidence of how the two alloys were joined. It was most likely that the steel inner disc and the brass outer ring were flattened together by a coining press. Electron imaging shows no evidence of welding, riveting, brazing or soldering, or any joining technique other than flattening in a coining press:

I also used the X-ray spectrometers to look for chemical evidence of brazing. Brazing joins two materials by using a filler metal or alloy between them. Brazing alloys, like solder, must have low melting points and are composed of tin, lead, silver, antimony, and other metals. I first used the EDS system to analyze the interface between the brass and stainless steel and looked for any elements common to brazes and solders:
Below: EDS spectrum at the inner disc-outer ring interface; no new elements are apparent

The EDS spectrum showed no elements that hadn't already been detected in the brass and steel -- if there was any solder or braze, there wasn't much. To look for lower concentrations, I used the WDS system:
Below: WDS spectrum at the inner disc-outer ring interface; no new elements are apparent

Again no new elements are found at the interface. We only have the elements found in the brass and steel -- zinc, potassium, sulfur, silicon, iron, calcium, aluminum, copper, chromium, magnesium, nickel, vanadium, chlorine, and carbon. These are the elements that I included in the quantitative analyses I conducted:

These quantitative data allow us, for instance, to identify the type of brass used for the ring. A copper content of 64% and a zinc content of 35% indicates this is "high brass," a type of high-tensile brass used in screws and springs. The grade of stainless steel c also be identified using its composition: 430-grade.
What can we tell from these data? What if we were treating this modern coin like an archaeological one? First, we'd know what materials the makers used: high brass and 430-grade stainless steel. Based on that information, we'd be able to calculate the actual value of the metals within the coin. Most modern coins have face values greater than the value of their metal contents, so we could deduce a minimum possible value for this coin. The face value of the coin should not be less than the value of its metal (at the time the coin was made).
There are additional things we can deduce from the makers' choice of materials. Grade-430 stainless steel has good corrosion resistance, formability, and mechanical characteristics that allow it to be rolled into a sheet. High brass has high tensile strength and abrasion resistance, and it is often used to make screws, springs, and rivets. Based on this information, we can tell the coin makers chose alloys that not only would not easily ware or corrode but also have the necessary properties to be rolled flat, stamped out, and used in a coining press. We may also suspect that the brass and steel could possibly have been recycled based on their trace elements.
Lastly, we know that this particular coin came into contact with calcium chloride at some point. This salt has a broad variety of applications: road salt, desiccant, additives in plastic and fabric softener, food additive, electrolyte in sports drinks, and more. The coin might have been exposed to this salt during its manufacture or its circulation. Therefore, calcium chloride must be one of the chemicals used by people in this nation.
So this example might not have been as insightful as Kraay's research, but one can still see the basic processes of identifying and analyzing an alloy artifact using an electron microprobe.

Works Cited:
Carter, Giles F. 1971. Compositions of some copper-based coins of Augustus and Tiberius. In Science and Archaeology (Robert H. Brill, editor). Symposium on Archaeological Chemistry, 4th, Atlantic City, 1968. Boston: Massachusetts Institute of Technology Press. pp. 114-130.
Gordus, Adon A. 1971. Rapid nondestructive activation analysis of silver in coins. In Science and Archaeology (Robert H. Brill, editor). Symposium on Archaeological Chemistry, 4th, Atlantic City, 1968. Boston: Massachusetts Institute of Technology. pp. 145-55.
Hornblower, A.P. 1962. Archaeological applications of the electron probe microanalyser. Archaeometry, volume 5. pp. 108-112.
Kraay, C.M. 1958. Gold and copper traces in early Greek silver. Archaeometry, volume 1. pp. 1-5.
Roberts, G. 1960. X-ray microanalyser. Archaeometry, volume 3. pp. 36-37.
Showroneck, Russel K. and Max M. Houck. 1990. The nondestructive identification of worn coins from the Marquette Mission site, St. Ignace, Michigan. American Antiquity, volume 55, number 2. pp. 336-347.
Wyttenbach, A. and H. Hermann. 1966. A quantitative nondestructive analysis of silver coins by neutron activation. Archaeometry, volume 9. pp. 139-147.

Additional Reading:
Barrandon, J.N., J.P. Callu, and C. Brenot. 1977. The analysis of Constantinian coins (AD 313-340) by non-destructive californium-252 activation analysis. Archaeometry, volume 19. pp. 173-186.
Beauchense, F., J.N. Barrandon, L. Alves, F. Gil, and M. Guerra. 1988. Ion beam analysis of copper and copper alloy coins. Archaeometry, volume 30, number 2. pp. 187-197.
Bluyssen, H. and Ph. B. Smith. 1962. Determination of the silver content of Greek coins by neutron activation. Archaeometry, volume 5. pp. 113-118.
Carter, Giles F. 1977. Reproducibility of x-ray fluorescence analyses of Septimius Severus denarii. Archaeometry, volume 19. pp. 67-73.
Carter, Giles F. 1978. Precision in the x-ray fluorescence analysis of sixty-one Augustan quadrantes. Journal of Archaeological Science, volume 5. pp. 293-300.
Carter, Giles F., E.R. Caley, J.H. Carlson, et al. 1983. Comparison of analyses of eight Roman orichalcum coin fragments by seven methods. Archaeometry, volume 25. pp. 201-213.
Cesareo, R., M. Ferretti, and M. Marabelli. 1982. Analysis of silver objects by scattering and by x-ray fluorescence of monoenergetic gamma rays. Archaeometry, volume 24. pp. 170-180.
Condamin, J. and M. Picon. 1964. The influence of corrosion and diffusion of the percentage of silver in Roman denarii. Archaeometry, volume 7. pp. 98-105.
Emeleus, Vera M. 1958. The technique of neutron activation analysis as applied to trace element determination in pottery and coins. Archaeometry, volume 1. pp. 6-15.
Ferreira, G.P. and F.B. Gil. 1981. Elemental analysis of gold coins by particle induced x-ray emission (PIXE). Archaeometry, volume 23. pp. 189-198.
Fleet, R.J. 1975. The use of specific heat in the non-destructive analysis of silver/copper alloy coins. Archaeometry, volume 17. pp. 101-106.
Fleet, R.J. 1976. The application of specific heat in the detection of debasement in ancient silver/copper alloy coins. Archaeometry, volume 18. pp. 117-120.
Gordus, A.A. 1967. Quantitative non-destructive neutron activation analysis of silver in coins. Archaeometry, volume 10. pp. 78-86.
Hawkes, S.C., J.M. Merrick, and D.M. Metcalf. 1966. X-ray fluorescent analysis of some Dark Age coins and jewelry. Archaeometry, volume 9. pp. 98-138.
Hughes, M.J. and W.A. Oddy. 1970. A reappraisal of the specific gravity method for the analysis of gold alloys. Archaeometry, volume 12. pp. 1-12.
Kraay, C.M. 1958. The composition of electrum coinage. Archaeometry, volume 1. pp. 21-23.
Kraay, C.M. 1959. Gold and copper traces in early Greek silver II. Archaeometry, volume 2. pp. 1-16.
Mancini, C. and P. Petrillo Serafin. 1976. Identification of ancient silver-gold coins by means of neutron absorption. Archaeometry, volume 18. pp. 214-217.
Metcalf, D.M. and F. Schweizer. 1971. The metal contents of the silver pennies of William II and Henry I (1087-1135). Archaeometry, volume 13. pp. 177-190.
Meyers, P. 1969. Non-destructive activation analysis of ancient coins using charged particles and fast neutrons. Archaeometry, volume 11. pp. 67-83.
Mommsen, H. and T. Schmittinger. 1981. Test analysis of ancient Au and Ag coins using high energy PIXE. Archaeometry, volume 23. pp. 71-76.
Oddy, W.A. 1972. The analysis of gold coins -- a comparison of results obtained by non-destructive methods. Archaeometry, volume 14. pp. 109-118.
Radcliffe, C.D., B. Angle, E.S. Macias, and P.P. Gaspar. 1980. Gold analysis by differential absorption of gamma-rays. Archaeometry, volume 22. pp. 47-56.
Reimers, P., G.J. Lutz, and C. Segebade. 1977. The non-destructive determination of gold, silver, and copper by photon activation analysis of coins and art objects. Archaeometry, volume 19. pp. 167-172.
Schweizer, F. and A.M. Friedman. 1972. Comparison of methods of analysis of silver and gold in silver coins. Archaeometry, volume 14. pp. 103-107.
Thompson, Margaret. 1960. Gold and copper traces in late Athenian silver. Archaeometry, volume 3. pp. 10-15.
Yao, T.C. and F.H. Stross. 1965. The use of analysis by x-ray fluorescence in the study of coins. American Journal of Archaeology, volume 69. pp. 154-156.

1/16/08

Next >

< Previous

Main Page

Electron Microprobe Analysis in Archaeology

Electron microprobe analysis (EMPA), also known as electron probe microanalysis (EPMA), is an analytical technique that combines scanning electron microscopy (SEM) and compositional analysis using x-ray spectrometry. The ability to determine structure and chemistry of samples makes EMPA very versatile. This is a dominant analytical technique in geology, but it is not as commonly used in archaeology despite similar materials in studied both fields. Here I will post about topics in EMPA, artifacts I have analyzed, archaeological studies that use EMPA, etc. If there is a topic you'd like to see posted here, please let me know.

Ellery Frahm

Doctoral Candidate, Archaeology

Research Fellow, Geology & Geophysics

University of Minnesota - Twin Cities

Main Page & CV

Obsidian Sourcing

Electron Microprobe Analysis in Archaeology

Petroarchaeology

Geo/Arch/Sci Blog

Geoarchaeology Funding

Everything Else