Salvatore Siano1 , Marcello Miccio2, Laura Bartoli1, Mark Daymond3, Winfried Kockelmann3

1) Istituto di Fisica Applicata “Nello Carrara”, Consiglio Nazionale delle Ricerche, Firenze, Italy,2) Soprintendenza per i Beni Arheologici della Toscana. 3)Rutherford Appleton Laboratory, ISIS Neutron Facility, Chilton, England

ACHAEOMETALLURGICAL STUDIES BY NEUTRON DIFFRACTION

The discovery of an ancient object typically originates a variety of problems, which can be broadly grouped into the following categories: 1) the correct determination of the historical and cultural frame where it was realised, 2) the restoration, 3) the selection of protection treatments and microclimatic conditions providing its preservation. In this concern, a fundamental role is played by material characterisation analyses. These can address the selection of the most suitable preservation strategy, as well as to investigate ancient manufacturing techniques and to reconstruct commercial and human exchanges among contemporaneous civilisations.The whole of chemical, physical, and microstructural techniques presently employed in the practice of the archaeometallurgical characterisation is rather wide. Along with the metallographic approaches, various elemental and structural analysis techniques are usually applied, such as for examples Absorption Spectrophotometry, SEM-EDX (WDX), X-Ray diffraction, and other. All these techniques are subjected to some limitations because of the very localised chemical and structural information and destructivity. Actually, the most significant analyses are those based on invasive sampling such as coring or cutting. Thus, in case of high value artefacts, a thorough scientific investigation is often impracticable. This problem stimulated many researches aiming at the development of novel diagnostic techniques not requiring material sampling or, at least, strongly reducing the amount of material to be removed. The non-destructive ones based on X-Ray Fluorescence (XRF), Proton Induced X-Ray Emission (PIXE), Synchrotron Radiation (SR), and Neutron Diffraction (ND), underwent significant technological and methodological developments along the last years. Among these, ND is the only one which can provide fundamental information on bulk composition and structure of a metal object, through the whole thickness of its walls and on meaningful volumes. In the present work, we investigated and demonstrated for the first time, a number of peculiar diagnostic potentials of the ND in the study of ancient bronzes. This was achieved through a systematic analysis carried out on standardised bronze specimens an d a number of original Etruscan and Roman artefacts.

Binary bronze for a(%Sn) calibration, eutectoid and peak shape investigations

Ternary bronzes to check quantitative phase analysis

Mechanically and thermally treated specimens to investigate identifying features

RATIONALE

Aim

Experimental evaluation of the Time of Flight Neutron Diffraction (TOF-ND) potentials to characterise archaeological bronzes

Alloy and mineralisation phase analyses Residual stress measurements Texture analysis

Systematic approach

Production and investigation of standardised bronze specimens simulating possible working processes

Suitable selection and investigation of archaeological artefacts with different expected features

Study of archaeometrical problems

STANDARDISED SPECIMENS

ETRUSCAN AND ROMAN ARTEFACTS

Courtesy of the National Archaeological Museum of Chiusi (Dr. M. Iozzo)

Jug (IV-III BC)

9.6

cm

Roma (III AC)2.7 cm

15 cm

Vessel base (IV BC)

12

cm

Mirror (II BC)

12

.5 c

m

Olpe (V-IV BC)

Razor (IX BC)14 cm

13.5 cm

Fibula (VIII BC)

IDENTIFICATION OF THE WORKING PROCESSES

Effect of thermal and mechanical treatments on peak shape

Neutron beam

Bank 3

Following homogenisation, peak broadening is essentially due to residual stresses

TEXTURE (PREFERRED ORIENTATIONS) ANALYSIS

Neutron beam

Incomplete experimental pole figures

ODF calculationalgorithm

Reconstructed pole figures by MAUD code

MIRROR: absence of preferred orientation (multiple treatments)

COIN: weak fiber texture demonstrating it was produced by striking

Goss {011}<001>

JUG (111)

MEASUREMENT OF RESIDUAL STRESSES BY ENGIN-X

• One dimensional spatial resolution up to 0.5 mm (typical gauge volume 0.50.55 mm3)

N (// planes)

S ( planes)

Slit

Lattice parameter associated with planes // to the coin surface

N

27 mm

0.7 mm

Neutron beam

Vessel base (200)

ROTAX

Neutron beam

Bank 3Bank 2

Ban

k 1

Sampletank

- Large cylindrical sample tank: diameter 40 cm, height 60 cm

- Three detector banks providing a d-spacing coverage between 0.3-20 Å

- A collimator cuts out reflections occurring outside a 90 mm radius around the centre

HOMOGENISED BINARY BRONZES

Strong texture (preferred orientations) Linear d-shift up to 14% tin content Detection of the eutectoid peaks at 16% tin content

Experimental linear dependence of the lattice parameter on Sn content, as derived by Le Bail full pattern refinements

a=23/2rCu+23/2(rSn-rCu)CSn

AS-CAST SPECIMENS

Evidence of “dendritic peak broadening” Strong texture Linear d-shift up to 6% tin content Detection of the eutectoid peaks at 8% tin content

QUANTITATIVE PHASE ANALYSIS

Check on DIII(81:6:13): Cu 79.9 %, Sn 6%, Pb 14.1

Olpe: pattern splitting at the repair patch

Repair patch

Setups

Rietveld refinement by GSAS code

Jug

Vessel base

Bottom wall(o.b.w.)

Lateral wall(o.l.w.)

Handle(o.h.)

Repair patch

[Å]

Medium resolution powder diffractometer at the pulsed spallation source ISIS, Rutherford Appleton Laboratory, UK.

TOF-ND diagnostic techniques were designed and successfully applied to investigate archaeological bronze artefacts through the comparison with suitable reference specimens. The demonstration of non-destructive analyses, such as phase quantification through relatively thick or multiple bronze walls, determination of composition and homogeneity of the alloy, residual stresses evaluations by reflection peak broadening (ROTAX) or shift (ENGIN-X), and eventually texture analysis provide a peculiar and powerful characterisation set. This proves a real application perspective for this novel characterisation approach, which is ready for a large-scale employ in archaeometallurgical studies.

AcknowledgementsThis work was carried out within the activity frame of the CNR Agenzia 2000 Project “Development of neutron diffraction techniques for non-destructive analysis of metal archaeological findings”. The authors wish also to thank Dr. Renzo Salimbeni and Dr. Marco Zoppi for having encouraged and supported this research.

Some references on the topicW. Kockelmann, E. Pantos, A. Kirfel, in: Radiation in Art and Archaeometry, Ed. D.C. Creagh, D. Bradley (Elsevier, Amsterdam, 2000) pp. 347-377.W. Kockelmann, A. Kirfel, E. Haehnel, J. Arch. Sci. 28, (2001) 213.S. Siano, W. Kockelmann, U. Bafile, M. Celli, M. Iozzo, M. Miccio, O. Moze, R. Pini, R. Salimbeni, M. Zoppi, Appl. Phys. A 74 (2002) S1139.S. Siano, L. Bartoli, M. Zoppi, W. Kockelmann, M. Daymond, J.A. Dann, M G. Garagnani, M. Miccio, in: Archaeometallurgy in Europe, Associazione Italiana di Metallurgia, Milano (2003) 319-329.S. Siano, L. Bartoli, W. Kockelmann, M. Zoppi, and M. Miccio, in press in Physica B (2004).