* Ezzeddine Triki, Ne ´bil Souissi Unite ´ de Recherche “Corrosion et Protection des Me ´talliques”, Ecole Nationale d’Inge ´nieurs de Tunis, 1002 Tunis (Tunisia) Latifa Bousselmi Laboratoire Eau & Environnement, Institut National de Recherche Scientifique et Technique, 2050 Hammam-Lif (Tunisia) Slim Khosrof Laboratoire de Conservation-Restauration, Institut National du Patrimoine, 2000 Bardo (Tunisia) Electrochemical behaviour of an archaeological bronze alloy in various aqueous media: New method for understanding artifacts preservation Elektrochemisches Verhalten eines archa ¨ ologischen Bronzelegierung in ver- schiedenen wa ¨ ssrigen Medien: Eine Methode zum Versta ¨ ndnis der Erhaltung von Kunstwerken N. Souissi, L. Bousselmi, S. Khosrof and E. Triki* The aim of this work is to investigate the electrochemical beha- viour of an archaeological bronze in several aqueous electrolytes (Na 2 SO 4 , NaCl, KOH, NaHCO 3 and Na 3 PO 4 ) at 0.01 M. The anodic polarization revealed the dissolution of the material independently from the corrosive medium, but the damaging action is stronger for sulfate and chloride where the pH was below 7. Electrochemical impedance spectroscopy permitted to establish a damage ranking based on the pH of the electrolytes: SO 4 2– > Cl 3 OH > HCO 3 > PO 4 3– . Phosphate was tested successfully as inhibitor for archaeological bronzes against the corrosive action of sulfate and chloride. Das Ziel dieser Arbeit ist die Untersuchung des elektrochemi- schen Verhaltens einer archa ¨ologischen Bronze in verschiedenen wa ¨ssrigen Elektrolyten (Na 2 SO 4 , NaCl, KOH, NaHCO 3 und Na 3 PO 4 ) bei 0,01 M. Die anodische Polarisation zeigte, dass die Auflo ¨sung des Materials unabha ¨ngig vom korrosiven Medium ist, dass aber die scha ¨digende Wirkung fu ¨r Sulfat und Chlorid, wo der pH-Wert unterhalb 7 lag, sta ¨rker ist. Die elektrochemische Impedanzspektroskopie erlaubt es eine Reihenfolge hinsichtlich der Sta ¨rke der Scha ¨digung zu erstellen, die auf dem pH-Wert der Elektrolyten basiert: SO 4 2– > Cl 3 OH > HCO 3 > PO 4 3– . Phosphat wurde erfolgreich als Inhibitor fu ¨r archa ¨ologische Bronzen gegen- u ¨ber der korrosiven Wirkung von Sulfat und Chlorid getestet. 1 Introduction Several investigations were carried out on archaeological bronzes [1 – 9]. Two types of alteration structures were en- countered [10]. The authors distinguished for their phenom- enological model two types of aggressive media: SO 2 4 ; CO 2 3 ðHCO 3 Þ and PO 3 4 ðHPO 2 4 and H 2 PO 4 Þ. These large size anions with low electro-negativity and com- plexing proprieties may be considered as precursors for type I alteration structure. As Robbiola et al. [10] demonstrated, chloride environment leads to type II morphology; but what these authors considered as type I may be encountered only under special conditions. Very few investigations have been concerned with the elec- trochemical behaviour of archaeological bronzes in aqueous media. However, Robbiola et al. [11] when simulating the burying soil by means of a 0.01 M sodium sulfate solution, reveals a decuprification phenomenon at the patina layers. Tommesani et al. [12] tested three 0.01 M corrosive electro- lytes: NaCl, (NH 4 ) 2 SO 4 and NaNO 3 . The authors concluded that the damaging action is in the sequence: (NH 4 ) 2 SO 4 NaNO 3 > NaCl. All these works were carried out on synthetic bronzes, thus direct applications could not be drawn for further understanding and asserting neither the corrosion pro- cess nor any specific preservation procedure to be applied on real ancient bronze. This paper reports the primary investigation work on the corrosion phenomenology of ancient bronzes during burial. Electrochemical techniques were used to study the corrosion behaviour of an original archaeological bronze alloy in var- ious aqueous electrolytes. 2 Experimental conditions The bronze archaeological alloy was a sample obtained from a fragmented Punic “œnochoe ´” exhumed from the ex- cavation of Utica (North East of Tunisia) [13]. The bad state of conservation of this artifact does not allow its use for mu- seographic exhibit, and make it eligible for such studies. A sample has been taken from the artifact providing an electrode corresponding to an active area of 0.2 cm 2 , the alloy composi- tion is given in Table 1. Experiments were carried out, on the same metallic sample polished previously to each immersion in a different media, at room temperature with 0.01 M aerated solutions (Table 2) prepared from analytical grade reagents. 318 Souissi, Bousselmi, Khosrof and Triki Materials and Corrosion 54, 318–325 (2003) 0947-5117/03/0505-0318$17.50þ.50/0 F 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim