Electrochemical dating of archaeological gold based on refined peak current determinations and Tafel analysis

Highlights

• Electrochemical dating of archaeological gold samples is based on the voltammetry of immobilized particles.

• Samples from the Mapungubwe Gold Collection (1200–1290 CE) and Santa Llúcia archaeological site (600-550 BCE) were studied.

• Peak potentials and peak currents for gold oxidation in 0.10 M HCl are used.

Abstract

This report proposes a refined method to date archaeological gold samples based on the Tafel analysis of the ascending part of voltammetric curves corresponding to the oxidation of surface flakes of gold in contact with hydrochloric acid aqueous electrolyte. This allows estimating the equilibrium potential of gold oxidation correcting for irreversibility effects. This equilibrium potential can be correlated with the coverage of adsorbed oxygen species and hence the estimated age of the gold samples. A satisfactory potential/time calibration graph was constructed from a set of archaeological samples including two sets of samples from the Mapungubwe Gold Collection, South Africa (1200–1290 CE) and Santa Llúcia, Alcalà de Xivert in Spain (dated back 600-550 BCE) sites. The dated samples come not only from different continents, but their ages also cover more than 3000 years.

Introduction

Dating of ancient metals is an important analytical aim in archaeology, historical and conservation science, arts and other related disciplines. This is a real challenge due to the unavailability of the majority of techniques (e.g., radiocarbon, uranium decay, thermoluminescence, obsidian hydration) to date metals [1,2]. Recently, methods for dating lead using the Meissner superconducting effect [3], and gold based on He, U, and Th analysis [4,5] have been described. However, these methods suffer from the need of rather large amounts of samples (some hundred milligrams), thus requiring an invasive (destructive) sampling procedure. Here a minimally invasive sampling and electrochemical analysis method is proposed, which requires just a few nanograms.

Aimed to complement the existing techniques for metal dating, the application of solid state voltammetry for dating lead [6], copper/bronze [7], and gold [8] samples has been previously described. These methods exploit the high sensitivity of the voltammetry of immobilized particles (VIMP) [[9], [10], [11]], a solid state electrochemistry technique that provides sensitive responses restricting the sampling to only a few nanograms collected on the metal surface by means of graphite leads [12,13]. Based on this methodology, widely used in the analysis of archaeological, historic and art objects in the science of conservation and restoration field [14,15], and primarily proposed as applicable to date ceramic materials [16], chronological information on lead-bronze statuary [17] and coinage [18], gold altarpieces [19] and embroidery [20] has been made available.

As all other ‘chemical’ dating methods, e.g., the obsidian hydration and amino acid racemization [21,22], the application of these voltammetric methodologies requires the assumptions that the chemical conditions were rather constant over time. A second aspect to be accounted for is that the composition of the corrosion layers of metals depends in general on the depth with a concomitant effect on the voltammetric responses [23].

In this context, electrochemical dating of archaeological gold is of particular interest because its voltammetric behaviour is rather well known [[24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39]], and, due to its noble character, the aging of this metal is much less sensitive to the ‘local’ physico-chemical conditions and their possible fluctuations over the entire time span to be dated. However, the electrochemical response of archaeological gold artefacts can be sensitive to the composition of the base metal (silver content in particular) and the manufacturing process. Archaeological gold alloys (gold-silver, gold-copper, gold-copper-silver) deteriorate in different ways which enhances the difficulties of preservation and their dating. As a result, previous data using a set of gold coins and different archaeological objects, provided a calibration graph with relatively high dispersion [8]. Here, an improved electrochemical methodology for archaeological gold dating is reported based on: i) refined peak current measurements, and ii) the Tafel analysis of the voltammetric curves.

In the current work, a series of samples from the Museum of Fine Arts of Castelló (Spain), Museum of Mannheim (Germany), and University of Pretoria Museums (South Africa) were studied. These include two sets of particularly relevant samples: the Mapungubwe Gold Collection (Republic of South Africa), dated back to 1200–1290 CE [40,41], and the collection from the Santa Llúcia, Alcalà de Xivert (Spain), dating back to 600-550 BCE [42]. These correspond to objects recovered from burials having in principle the same age thus prompting the study of the influence of possible differences in manufacturing and/or composition of the objects in the dating.

It is pertinent to underline that, by reasons of conservation, the sampling (consisting of pressing a graphite electrode onto the surface of the objects, as previously applied [[14], [15], [16], [17], [18], [19], [20]]) was limited to three spots in each archaeological object. As a result, we obtained three replicate voltammetric measurements for each sample/object. The sampling was carried out in the museums by the researchers of the respective institutions and the electrodes were mailed to the University of Valencia where the voltammetric measurements were performed. Due to the limited number of accessible samples, we used the three-step voltammetric protocol in 0.10 M HCl solution already described [8] in order to maximize the analytical signals able to be processed. In order to test the suitability of the proposed dating methodology, one of the samples, one wire ring (MA01, “im Grafenwald” Hermeskeil) from the Rheinische Landesmuseum Trier was taken as a problem sample. This sample is representative of a frequent problem in archaeology: the disposal of samples coming from non-professional excavations or corresponding to sites with ill-defined stratigraphy. As judged by its archaeological context, the estimated age of sample MA01 was 30–150 BCE. However, as far as the relation of the object with the context was not entirely known, there was uncertainty about its age. Accordingly, this sample was treated as a problem sample of unknown age by the electrochemistry research team.

The detailed description of the samples is provided in Table 1. Voltammetric data are combined with Raman spectroscopy and focusing ion beam-field emission scanning electron microscope (FIB-FESEM) and high resolution field emission scanning electron microscopy (HRFESM-EDX). By the aforementioned reasons of conservation, these techniques were applied only to contemporary gold objects taken as a reference.