Документ взят из кэша поисковой машины. Адрес оригинального документа : http://hbar.phys.msu.ru/gorm/dating/coins2.pdf Дата изменения: Wed Apr 26 00:00:00 2000 Дата индексирования: Mon Oct 1 20:28:09 2012 Кодировка: Windows-1251

Nuclear Instruments and Methods in Physics Research B 161+163 (2000) 748+752

www.elsevier.nl/locate/nimb

Characterization of Dyrrhachium silver coins by micro-PIXE method
I. Uzonyi
a

a,* ,

R. Bugoi b, A. Sasianu c, A.Z. Kiss a, B. Constantinescu b, M. Torb agyi

d

Department of Electrostatic Accelerators, Institute of Nuclear Research of the Hungarian Academy of Sciences (ATOMKI), P.O. Box 51, Bem ter 18/c, 4001 Debrecen, Hungary b Cyclotron Laboratory, Institute of Atomic Physics, Bucharest, Romania c Department of Numismatics, Cris County Museum, Oradea, Romania d Department of Numismatics, Hungarian National Museum, Budapest, Hungary

Abstract Ancient silver coins (drachms) issued by the Greek city Dyrrhachium during 68+43 years BC were analysed nondestructively by micro-PIXE method. The selected 27 drachms, including four imitations, belong to the numismatic collection of the Hungarian National Museum (HNM). Nine elements (Fe, Cu, Zn, Br, Ag, Sn, Au, Pb and Bi) were determined quantitatively. Samples are characterised with a uniformly low $92% Ag concentration. Debasement is supposed. У 2000 Elsevier Science B.V. All rights reserved. PACS: 82.80.E; 29.30.Kv Keywords: Archaeology; Greek silver coins; Micro-PIXE

1. Introduction Ag drachms issued by the two Greek cities of Illyria, Apollonia and Dyrrhachium, in the second and Rrst centuries BC have been frequently found in the Carpathian Basin and in the Balkan region. From the archaeological point of view, the main problem is to classify them in order to explain related economical and political aspects such as the impetuous increasing of their amount in the

Corresponding author. Tel.: +36-52-417-266; fax: +36-52416-181. E-mail address: uzonyi@atomki.hu (I. Uzonyi).

*

Rrst century BC, especially before and during the Roman civil war between Pompeius and Caesar [1+3]. The Rrst modern classiRcation of Apollonia and Dyrrhachium coins discovered in Albania was accomplished by Ceka [4], who found a connection between elemental composition and the minting period. Four years ago, the Romanian co-authors of the present paper also started a comprehensive and systematic study of the coin-hoards discovered in Romania [5]. The existence of similar artefacts at the Hungarian National Museum (HNM) led to this collaboration in order to reveal connections between drachms found in the neighbouring countries.

0168-583X/00/$ - see front matter У 2000 Elsevier Science B.V. All rights reserved. PII: S 0 168-5 83X(99 ) 0 0967 -2


I. Uzonyi et al. / Nucl. Instr. and Meth. in Phys. Res. B 161+163 (2000) 748+752

749

2. The archaeological samples Out of 153 pieces Dyrrhachium drachma collection of the Coin Cabinet of HNM [3] 27 were selected for analysis including four ancient imitations (see Fig. 1 and Table 1). Imitations are characterised by misspelled legends and the lack of reRnement of decorations, while the original drachms have readable inscriptions. All analysed coins were minted during the 68+43 years BC. The exact Rnding spot of this hoard is still unknown but, supposedly, it should have been found in southern or south-eastern part of Hungary.

energy of 2.0 MeV, beam current of $200 pA, total collected charge of $0.2 lC and a beam spot size of $3 ? 3 lm2 . For the selective Rltering of the most intense Ag La line, a 10 mg/cm2 Al foil was placed between sample and the Si(Li) detector. In all cases three dierent areas were scanned in ?at and bright regions of the coins where the surface is more likely free from patina due to the continuous rubbing and cleaning eect of use. Spectrum Rtting and calculation of elemental composition was carried out by using the PIXYKLM programme package [7]. In the spectra of coins, altogether, the characteristic lines of nine elements such as Fe, Cu, Zn, Br, Ag, Sn, Au, Pb and Bi could be detected.

3. Experimental The analysis of drachms was performed at the Debrecen scanning nuclear microprobe [6]. The experimental conditions were as follows: proton 4. Results and discussion In this study, X-ray intensity maps were generated in all measurements for all analysed elements. The scan size was 1 ? 1 mm2 . This way it was possible to exclude deposited or heavily corroded areas as well as inclusions from the evaluation. For example, Fig. 2 shows Cu enrichment on the surface of a Ag coin. The unaected area is characterised by 94% Ag and 5% Cu concentrations, while the corresponding values at the Cu-rich spot (diameter $1 mm) are completely dierent: 74% Ag and 25% Cu. X-ray lines of Fe have often been observed in the spectra of Ag coins [8]. Here again, the excellent lateral resolution of the microprobe helps to interpret their origin. As it can be seen from Fig. 3, Fe Ka counts are concentrated in a few extremely protruding peaks with a diameter of <20 lm. A possible explanation for this fact is that Fe-peaks originate mainly from either surface contamination or inclusions present in the sample.

Fig. 1. Photo of some analysed Ag drachms (front and reverse sides). Catalogue numbers are 52, 82, 128.

Table 1 IdentiRcation of the analysed Dyrrhachium coins (minted 68+43 years BC) Group I (n 15) II (n 4) III (n 4) IV (n 4) Catalogue number 24, 26, 27, 28, 31, 33, 34, 36, 40, 42, 43, 52, 53, 59, 61 79, 82, 89, 106 127,128, 129, 130 150, 151, 152, 153 Type MENIRKOR-DIO-NY-RIOY Ceka 320 [4] MENIRKOR-KY-KIR-KOY Ceka 325 NENXN-DA-MH-NOR Ceka 357 Imitations


750

I. Uzonyi et al. / Nucl. Instr. and Meth. in Phys. Res. B 161+163 (2000) 748+752

Fig. 2. Elemental maps showing Ag Ka and Cu Ka intensity distributions on a corroded layer of a Ag coin. Scan size: 2500 ? 2500 lm2 .

Fig. 3. A typical Fe Ka intensity map. Scan size: 1000 ? 1000 lm2 .

Depth proRling was carried out on an ``average quality'' area by using 2.0, 2.5, 3.0 and 3.5 MeV beam energies. As Ag Ka radiation may emerge from Rve times deeper layer ($40 lm) of the surface than Cu Ka ($8 lm), inhomogeneities in depth distributions of Cu or Ag should be re?ected in changes of calculated concentrations when energy increases (similar statement can be made for minor elements such as Au or Pb). At the above energies, the determined concentrations varied less than 0.5% (absolute) for the major elements Cu and Ag. The measure of this eect is comparable

with the accuracy of the used fundamental parameter method, therefore, within these limits, the homogeneity hypothesis was accepted. Table 2 summarises the micro-PIXE results for the Hungarian drachms. Following the archaeological classiRcation, coins were divided into four groups and analytical data were subjected to statistical evaluation. Each coin was represented by averaged concentrations from three independent measurements. From the individual values group averages as well as standard deviations were calculated, which can be seen in Table 2. Compositional data for series of similar type Dyrrhachium coins from Romanian Viisoara's hoard are also presented for comparison. They were measured by 3 MeV external PIXE and XRF methods in the Cyclotron Laboratory of the Institute of Atomic Physics, Bucharest, Romania [5]. As a Rrst step, one way ANOVA analysis was carried out to test the hypothesis of equality of group averages. In statistical sense, at a probability level of P < 0.05, there are no signiRcant differences between the elemental compositions of the Hungarian groups. In contrary, the Romanian results were statistically dierent from the Hungarian ones in some cases (see Table 2). It is worth mentioning that, except Cu and Ag, the dierences are not serious. They arose from either biases in the analytical techniques used (e.g., in case of Au) or the varying degree of surface contamination (Fe, Br). Namely, Br can be associated to air pollution of the coin as it was pointed out by


Table 2 Elemental composition of Dyrrhachium silver coinsa Zn (ppm) 210 180 120 520 670 n.d. 560 n.d. 460 780 70 20 400 270Г 70 30 450 220Г 50 10 130 130 90.4 85.3 91.8 86.8 92.6 90.1 2.1 7.1Г 1.4 6.4 1.7 5.4 n.d. 180 80 n.d. 230 210 n.d. 7700 5800ГГ 0.35 0.45 0.36 0.40 0.33 0.43 0.03 0.08Г 0.05 0.15 0.16 0.06 0.77 0.83 0.87 1.02 0.58 0.42 Br (ppm) Ag (%) Sn (ppm) Au (%) Pb (%) 0.5 0.4 0.3 0.8 0.2 0.2 Bi (%) 0.12 0.10 0.09 0.03 0.13 0.14 0.06 0.09 0.06 0 0.07 0.02 Mass (g) 3.23 3.08 3.21 3.03 3.28 2.98 0.2 0.2 0.2 0.3 0.1 0.2

Group 2.2 7.3Г 1.1 6.6 1.4 4.8

Fe (ppm)

Cu (%)

I (n 15) I(R) (n 16) II (n 4) II(R) (n 11) III (n 4) IV (n 4)

830 50 2200 1800Г 580 200 1000 800 740 100 1200 600

8.24 13.1 6.74 11.3 6.26 8.21

I. Uzonyi et al. / Nucl. Instr. and Meth. in Phys. Res. B 161+163 (2000) 748+752

a

For comparison, analytical results characterising the Viisoara's (Romania) coin-hoards are involved [5]. Abbreviations and remarks: R: Romanian results; n.d.: not detectable. * SigniRcant dierence (P < 0.05) between the corresponding Hungarian and Romanian results. ** One of the imitations (no. 151) does not contain Sn.

751


752

I. Uzonyi et al. / Nucl. Instr. and Meth. in Phys. Res. B 161+163 (2000) 748+752

Brissaud et al. [9]. In case of Cu and Ag, the large variances of Romanian results are remarkable. It is worthwhile to mention that the Au concentration is practically constant ( 0.35%) for all analysed coins. This metal may be a Rngerprint of the Ag ore. Pb was supposedly used to lower the melting point of the alloy. It is also interesting to note, that there are strong correlations between Cu (Ag)+Zn (r 0.86) and Cu+Ag (r )0.99) and practically no other signiRcant correlations. The high negative correlation between Cu and Ag means the substitution (probably intentional) of Ag by Cu. Measurements have shown that the quality of Ag drachms degraded from an initial $98% Ag value down to 70+90% until the middle of the Rrst century BC [5]. This degradation may explain the signiRcant dierences in Cu and Ag concentration between the Hungarian and Romanian coins. A characteristic element for imitation is Sn, which was found only in three of the suspected samples. With the permission of HNM one of these coins (no. 150) was Rnely scratched at its edge in order to study its structure with the microprobe. PIXE analysis revealed a low quality Ag core (Cu 46%, Zn 0.3%, Ag 50.5%, Sn 2.5%, Au 0.2%, Pb 0.4%, Bi 0.06%) below the surface. This way, the supposition that these samples are coated counterfeits was proven. The fourth coin (no. 151) does not dier in composition from the normal ones, which suggests a misidentiRcation of this coin. 5. Conclusion This investigation contributes to the extensive Romanian studies and helps to establish an exact

chronological classiRcation of Dyrrhachium Ag coins. The results of both laboratories present the same phenomenon of degradation of the Rneness of Ag by introducing increasing quantity of Cu ($8+10%) as compared to the coins minted in the second century BC. Acknowledgements This work was carried out under the framework of the European COST Action G1 and supplied by the Hungarian National Research Fund (OTKA No. T 025771). A short term scientiRc mission from COST Action G1 for one of the authors (R. Bugoi) to the laboratory in ATOMKI is gratefully acknowledged.

References
[1] A. Maier, Numismatische Zeitschrift 1 (1908) 1. [2] A. Sasianu, Ancient coinage in western and north-western Romania. Nr. 45 Oradea (1980) 111. [3] M. Torb agyi, Numizmatikai Kozlony LXXXVIII+ LXXXIX (1989+1990) 3 (in Hungarian). [4] H. Ceka, in: Proceedings Actes du Ier Congres International des Etudes Balkaniques et sud-est Europeenes II. SoRa, Bulgaria 1969. p. 277. [5] R. Bugoi, B. Constantinescu, F. Constantin, D. Catana, D. Plostinaru, A. Sasianu, J. Radioanal. Nucl. Chem. 242 (1999) 777. [6] I. Rajta, I. Borb ely-Kiss, Gy. Morik, L. Bartha, E. Koltay, A.Z. Kiss, Nucl. Instr. and Meth. B 109+110 (1996) 148. [7] Gy. Szabo, I. Borb ely-Kiss, Nucl. Instr. and Meth. B 75 (1993) 123. [8] M.A. Meyer, G. Demortier, Nucl. Instr. and Meth. B 49 (1990) 300. [9] I. Brissaud, P. Chevallier, C. Dardenne, N. Deschamps, J.P. Frontier, K. Gruel, A. Taccoen, A. Tarrats, J.X. Wang, Nucl. Instr. and Meth. B 49 (1990) 305.