ARHEOMET - Materials - Reports

Report phase V

The elaboration of a procedure for the use of portable XRF spectrometer in archaeological application

X-MET 3000TX analyzer

The X-MET 3000TX is a portable analyzer used for analyzing the elemental composition of a very large number of material having as principal application the alloy, soils, geological/ore sample or artefacts analysis. The analyzer is based on X fluorescence (the emission of characteristic lines spectra for each element presented in the sample) and uses has excitation source a X-tube with Rh - anode. It can use external power supply or accumulators, and a PDA computer controls it. X-ray detector is a PIN silicon diode, Peltier cooled, placed in the back of a Kapton window. To perform a measurement, the sample is placed perpendicular on that window.

In the following we will particularize this, using gold, silver and bronze archaeological objects. In the three cases we used the fundamental parameters method - the preset version for alloys. We have measured the 13 pieces of the Persinari Thesaurus (bronze age about 500 B.C.) - 11 gold daggers one gold ceremony sword, and a fragmentary gold dagger at the National Museum of History, Romania. The results are shown below:

(Figures 1-7, see Illustrations)

Dagger 23238

1. Au=63.96%; Ag=31.13%; Cu= 4.91%
2. Au=63.74%; Ag=30.05%; Cu= 4.91%
3. Au=64.09%; Ag=30.34%; Cu= 5.69%

Dagger 23239

1. Au=57.21%; Ag=38.24%; Cu= 4.56%
2. Au=55.45%; Ag=37.40%; Cu= 6.50%

Dagger 23241

1. Au=60.37%; Ag=33.99%; Cu= 5.64%
2. Au=59.92%; Ag=33.70%; Cu= 5.73%

Dagger 23240

1. Au=55.03%; Ag=35.80%; Cu= 8.29%
2. Au=56.81%; Ag=35.97%; Cu= 7.21%
3. Au=54.58%; Ag=35.83%; Cu= 8.99%

Dagger 23242

1. Au=56.96%; Ag=33.13%; Cu= 9.14%
2. Au=57.37%; Ag=32.88%; Cu= 8.85%

Dagger 23235

1. Au=64.41%; Ag=29.82%; Cu= 5.06%
2. Au=64.80%; Ag=29.90%; Cu= 5.30%

Dagger 23244

1. Au=58.07%; Ag=32.16%; Cu= 8.99%
2. Au=59.72%; Ag=30.70%; Cu= 9.08%

Dagger 23243

1. Au=56.06%; Ag=36.70%; Cu= 7.24%
2. Au=55.82%; Ag=36.90%; Cu= 7.28%

Dagger 23236

1. Au=63.66%; Ag=30.70%; Cu= 5.63%
2. Au=64.18%; Ag=30.41%; Cu= 5.41%

Dagger 23237

1. Au=62.79%; Ag=31.28%; Cu= 5.06%
2. Au=63.14%; Ag=30.20%; Cu= 5.95%

Dagger 23234

1. Au=63.43%; Ag=30.64%; Cu= 5.26%
2. Au=65.61%; Ag=29.51%; Cu= 5.88%

Fragmentary dagger C 938 (9 pieces)

1. Au=63.66%; Ag=30.38%; Cu= 5.18%
2. Au=62.96%; Ag=30.91%; Cu= 5.49%

Sword

1. Au=64.02%; Ag=27.54%; Cu= 7.83%; Fe= 0.62%
2. Au=88.46%; Ag=9.30%; Cu= 2.24%
3. Au=88.55%; Ag=9.18%; Cu= 2.27%
4. Au=55.67%; Ag=27.50%; Cu= 6.89%; Fe= 5.75%; Zn=2.71%; Pb=1.33%
5. Au=63.23%; Ag=26.82%; Cu= 9.54%; Fe= 0.38%
6. Au=61.91%; Ag=29.83%; Cu= 8.06%

the spectrometer was placed in a vertical position so the object for analyze to be placed above;
a number of spectra was collected with 30 - 300 seconds measurement time, in two or three different points from dagger and in six points for the sword (blade, case grip);
the objects can be measured also straight in the show case ( of course after the glass is removed), but only with the condition that the surface to be measured to be bigger than the spectrometer window; in contrary case the spectrometer will measure the show case background (plastic, canvas, plush) and is impossible to obtain a quantitative results (the instrument soft is tricked by the background composition);
we've establish that for a more precisely determination, after quantitative results are obtain the check also the x-ray spectra of that determination;
the sensibility for copper determination in the case of gold objects is very poor, about 0.5%; but using a measurement time of 300 seconds to collect a spectra the copper ratio can be evaluate straight from the XK_α(8.05 keV) pick surface;
the zinc can not be determinate in the gold object because of the superposition of zinc XK_α with copper XK_β and zinc XK_α with gold XK_l X-ray.
because of the small distance between the window detector and the sample surface arise many pile-up and escape picks of the gold X-ray, a compromise between the efficacy (the minimum distance between the detector and the source) and the absence of the real picks interferences - electronic picks (large distance between the detector and the source) must be made.

In conclusion this spectrometer is adequate for gold title object determination and their lesser components (bigger than 1%). If one wants to determinate more precisely the trace of elements (below 1%) is required to use the classical spectrometer Am-241excitation source type, with Si(Li) or Ge(I) detector.

Regarding this results we establish the presence of two electrum alloy type categories at the daggers (Au=65%, Ag=30%, Cu=5% and Au=55%, Ag=35%, Cu=10%), and for the sword, the blade is Au=89%, Ag=9%, Cu2%, the case from electrum (Au=65%, Ag=27%, Cu8%), and from the original grip only some traces remained in which we have identified Pb, Zn, Fe, Cu suggesting a bronze - brass alloy.

1. grip head, Φ cca 2cm, h=2cm polished - Ag=82%; Au=15%; Cu=2.5%; Fe=0.15%; Figure 8
2. axe blade - Ag=75%; Au=22%; Cu=2.7%; Fe=traces;
3. piece - Ag=74%; Au=20%; Cu=5.3%; Fe=0.05%;
4. piece - Ag=80.8%; Au=14.7%; Cu=4.3%; Fe=0.20%;
5. a grip piece - half - Ag=82.5%; Au=15.5%; Cu=1.9%; Fe=0.10%;
6. piece - Ag=74.8%; Au=22.3%; Cu=2.5%; Fe=0.20%;
7. the entire axe blade ( not solder) - Ag=82.5%; Au=16.2%; Cu=0.85%; Fe=0.25%;
8. blade axe (one time solder) - Ag=80.8%; Au=18.5%; Cu=0.45%; Fe=without;
9. small piece- Ag=81.6%; Au=16.5%; Cu=1.7%; Fe=0.10%;
10. big piece solder - Ag=83.2%; Au=16.2%; Cu=0.35%; Fe=0.08%;
11. blade axe broken and solder - Ag=81.7%; Au=15%; Cu=2.8%; Fe=traces;

The measurement procedure was the same like in the gold objects case. The only difference is that, because of the smaller gold quantity the sensibility of copper detection may decrease up till 0.3%. In the following it remain the tin (Sn) problem, and from here the necessity to use a classical spectrometer to detect it.

Regarding the result we have to deal with two types of electrum very rich in silver ( Ag=75%, Au=22%, Cu=3%) and (Ag=82%, Au=15%, Cu=3%). The presence of the electrum in daggers as well as in axes suggest the Anatolian origin (Asia Minor) of the alloys, being possible to arrived, in this region in Muntenia, in block shape or like tumbling barrel.

Regarding the bronze pieces we have measured Late Roman pieces funded at Tropaevm Traiani and in Banat (Ancient Moldavia). The measurement procedure is the same like the one used for gold and silver objects. Has particular characteristics we observe:

because of the zinc XK_α with copper XK_β ray superposition it is necessary to use the zinc XK_β for quantitative determination.
the Silver being present like traces in bronze, the detection of tin is not longer a problem;
a special attention must be paid to the lead XK_β and arsenium XK_α (10,5 keV) superposition, and that does necessary to use the arsenium XK_β (12,7keV) and lead XK_β (12,7keV) rays for quantitative determinations.

1. ring, N.III.10.16 - Cu=62%; Pb=23%; Sn=12%; As=15; Sb=0.2%; Ag=0.3%;
Fe=0.3%; Zn=0.8%
2. key, N.III.10.20 - Cu=80%; Pb=6%; Sn=6%, Zn=7%; Sb=0.15%; Ag=0.10%;
Fe=traces; As=traces;
3. lug, N.III.10.21 - Cu=84%; Zn= 10%; Sn=3%; Pb=2.5%; Ag=0.20%;
Fe=0.25%
4. lug with head horse protoma , N.III.10.22 - Cu=66%; Zn=27%; Pb=3%; Sn=3%;
Ag=0.10%; Sb=0.10%;Fe=0.30%
5. fibula with buttons..., N.IV.B.10.1 - Cu=84%; Zn=12%; Pb=1.5%; Sn= 2%;
Ag=traces; Sb=traces; Fe=0.15%

7. clasp, N.V.10.22 - Cu=63%; Pb=2.5%; Zn=14%; Sn= 20%; Ag=0.15%;
Ke=0.10%
8. small ring bell, N.V.10.25 - Cu=20%; Pb=25%; Zn=45%; Sn= 10%; Fe=dirt!
9. key , N.V.10.28 - Cu=85%; Zn=3%; Pb=5%; Sn= 6%; Ag=0.15%; Sb=0.25%;
Fe=0.25%; As=traces;
10. key, N. V.10.30 - Cu=76%; Zn=0.8%; Pb=12%; Sn= 10.5%; Ag=0.20%;
Sb=traces; Fe=0.25%; As=0.15%;
11. ornament, N.V.10.30 - Cu=89%; Zn=0.5%; Pb=1%; Sn= 9%; Ag=0.3%;
Sb=0.10%; Fe=0.25%;
12. pyxis tree, N.V.10.34 - Cu=80.8%; Zn=18%; Pb=0.2%; Sn= 0.6%; Ag=traces;
Fe=0.10%;
13. fibula with twisted leg, N.VI.A.10.2 - Cu=86.5%; Zn=8%; Pb=1.3%; Sn= 3.5%;
Sb=0.15%; Ag=0.30%; Fe=0.25%;
14. fibula with twisted leg, N. VI.A. 10.3 -Cu=88%; Zn=3%; Pb=traces; As=0.3%;
Sn= 8%; Ag=0.20%; Fe=0.4%;
15. clasp, N. VI. A. 10.4 - Cu=68%; Zn=5%; Pb=1.4%; Sn= 25%; Ag=0.20%;
Fe=0.10%; - Figure 9
16. fibula with twisted leg, N.VI.B.10.3 - Cu=85.5%; Zn=10%; Pb=1.6%; Sn= 2.3%;
Ag=0.15%; Fe=0.20%;
17. fibula with twisted leg, N. VI. B. 10.4 - Cu=97%; Zn=2%; Pb=0.2%; As= 0.15%;
Ag=0.15%; Fe=0.10%;
18. bronze bracelet, N.VI.B.10.5 - Cu=97%; Sn=1.7%; Pb=traces; Ag=0.20%;
Fe=0.10%;
19. bronze ring with cu protoma, N.VI.B.10.9 - Cu=97.8%; Sn=5%; Pb=0.15%;
Ag/Cu=1/4; Fe=traces; SILVERED bronze!
20. chain let, N. VI.B.10.10 - Cu=85.8%; Zn=7%; Pb=3.5%; Sn= 3%;
Sb=0.10%; Ag=0.15%; Fe=0.25%;
21. crochet, passim - Cu=75%; Zn=23%; Pb=traces; Sn= 1.2%; Sb=traces;
Ag=0.20%; Fe=0.4%;
22. armour small nappe squamata, A.S.5, 2006 - Cu=70%; Zn=28%; Pb=traces;
Sn=1.6%; Ag=0.25%; Fe=dirt;
23. clapper clasp, A.9, 2006 - Cu=73%; Zn=12%; Pb=6%; Sn= 8%;
Sb=0.20%; Ag=0.15%; Fe=0.30%;
24. sandal nail, bas. D, 2007 - Cu=82%; Zn=3%; Pb=0.45%; Sn= 14%;
Sb=0.15%; Ag=0.10%; Fe=dirt;
25. pen head, bas.D,2007 - Cu=82%; Zn=17%; Pb=0.2%; Sn= 0.3%;
Sb=0.15%; Ag=0.15%; Fe=dirt; As=0.15%;

Craddock states that from the I^st-II^nd centuries in Roman Empire the brass (Cu + Zn) it was uses instead of bronze (Cu + Sn) on a large scale, due to the economical reasons - the rarity of the Tin deposits (Spain, United Kingdom) that make that the price of zinc to be ten times smaller, and also from technological reasons - brass is easier to be decorated in niello style and polished they will have a beautiful golden look. Being a little resistant at corrosion, the brass was not used for the objects with a important behalf - see aqvila - standard symbol measured in summer, small medicinal spoon from Resita, some fibulas or the sandal nail. The objects, that contain zinc and also tin were intermediary qualitative, and those made only from brass (zinc with a small quantity of tin or not at all) were poor qualitative and the cheap. The lead was used in the bronze to do the melting of the alloys easier by diminishing the melting point. To note that it was used the remould of the bronze and also of brass objects on a large scale, thing what explain the mixed composition tin-zinc.

1. Fibula, no. Inv. 8593 / R 1504, sec. III-IV, Gradinari;
Cu=85.5%; Sn= 8%; Pb=2.5%; Ag=traces; Zn=0.3%; Fe=0.5%; Cr=3%
Bronze - high quality;
2. Small plate, no. Inv. 73169 / R 1725, sec. IV, Gornea;
Cu=70%; Zn=5.5%; Pb=22%; Sn=1%; Fe=0.5%; Ni=traces;
Brass with a lot of lead - to facilitate the ornament stamp (mild alloy);
3. Lug, no. Inv. 73173 / R 1725, IV^th century , Gornea;
Cu=83%; Sn=8%; Pb=7%; Zn=traces; Ag=0.10%; Sb=0.15%; Fe=0.10%; Mn=0.5%
Bronze- very high quality;
4. Fibula, no. Inv. 3429 / R 1260, III^rd - IV^th century, Ancient Moldavia;
Cu=89%; Zn=9%; Pb=1.3%; Ag=0.10%; Mn=0.5%; Ni=traces; WITHOUT TIN
Simple brass;
5. Fibula, no. Inv. 3466 / R 1259, III^rd - IV^th, Ancient Moldavia ;
Cu=81.5%; Zn=0.10%; Pb=11.2%; Sn=6.2%; Sb=0.15%; Ag=0.15%; Fe=0.15%; Ni=traces;
Bronze - high quality;
6. Fibula, no. Inv. 3483 / R 1257, III^rd - IV^th century, Ancient Moldavia;
Cu= 65.6%; Zn=3.8%; Sn=0.5%; Pb=28.5%; Ag=0.10%; Fe=0.5%; V=traces; Mn=traces;
Brass with lead - to obtain a malleable alloy
7. Fibula, no. Inv. 3488 / R 1258, III^rd - IV^th century, Ancient Moldavia;
Cu=83%; Zn=13%; Sn=0.4%; Pb=2.7%; Sb=0.10%; Ag=0.10%; Fe=0.2%;
Brass.
8. Fibula, no. Inv. 4209/ R 1071, IV^th century, Ancient Moldavia;
Cu=90.5%; Zn=3.7%; Sn=3.3%; Pb=0.7%; Sb=0.10%; Fe=0.4%; Ni=traces; Cr=0.6%; As=0.5%, SILVERED
Brass and silvered bronze mixture
9. Small medicinal spoon, no. Inv. 3730/R1088, III^rd - IV^th century, Ancient Moldavia
Cu=92%; Sn=6.5%; Pb=0.25%; As=0.10%; Fe=0.20%;
Bronze - high quality;

One can assume that the bronze objects rise from an important urban centre (external), and the ones of brass are the products of some local foundries that used the remould of old objects on a large scale.

In conclusion this portable XRF spectrometer it can be very useful for a quick determination and especially in-situ (straight in the museums) of the principal metal alloys presented in the archaeological discoveries: Au+Ag+Cu, Ag+Cu+Au, Cu+Sn+Pb, Cu+Zn+Pb.

^ top

latest

Report phase V

The elaboration of a procedure for the use of portable XRF spectrometer in archaeological application

X-MET 3000TX analyzer