Matisse bronze sculptures

Until now, little attention has been given to the analysis of ancient copper alloys with LA-ICP-MS. This type of material is usually analyzed with fast or instrumental neutron activation analysis (FNAA or INAA), particle induced X-ray emission (PIXE), X-ray fluorescence (XRF), inductively coupled plasma-atomic emission spectrometry or inductively coupled plasma-atomic absorption spectrometry (ICP-AES or ICP-AAS). Some of these techniques are destructive and involve extensive sample preparation, some measure only surface compositions, and some require access to a cyclotron or a reactor. LA-ICP-MS is riot affected by any of these inconveniences. We propose here an analytical protocol for copper alloys using LA-ICP-MS and present its application to the study of Matisse bronze sculptures. 

The Baltimore Museum of Art hosts the Cone collection including many works by Matisse. Bronze sculptures by Matisse were cast using different methods (lost wax and sand cast) and in different foundries. Ann Boulton of the Baltimore Museum of Art initiated a project aimed at determining whether or not different compositions of Matisse bronze sculptures could be correlated to different manufacturing techniques or locations. The project started while the author was a post-doctoral fellow at the Smithsonian Center for Materials Research and Education, now Museum Conservation Institute, and involved other analytical techniques (12). The results presented here were obtained at the Field Museum of Natural History. 

Figure 3 compares the results obtained for the Matisse bronze samples with LA-ICP-MS and ICP-MS for six elements zinc, tin, lead, arsenic, bismuth and silver. The best correlations between ICP-MS and LA-ICP-MS occur for silver, arsenic and tin. For zinc, lead and bismuth, the results by ICP-MS and LA-ICP-MS concur quite well for some samples. For the other samples, the values are either over or underestimated. Immiscibility of lead in copper, even at low concentrations, could explain the non-agreement for some samples between the ICP-MS and the LA-ICP-MS results (13). However, zinc and bismuth are miscible in copper at the concentrations encountered in the Matisse bronze sculptures. 

The conclusions, derived from the results obtained by LA-ICP-MS and ICP-MS are identical. This preliminary study shows that our parameters of analysis and more especially our parameters of ablation for LA-ICP-MS were found suitable to study Matisse bronze sculptures. These results are promising and further study should hopefully demonstrate the usefulness of LA-ICP-MS for the characterization of other types of copper alloys. 

I am grateful to Ann Boulton from the Baltimore Museum of Art for the samples from the Matisse bronze sculptures and to Jia-Sun Tsang from the Museum Conservation Institute for having involved me in the Matisse bronze sculpture project. 

Indian hemp [N.A. plant] fibers, charred and uncharred, 61-64/ Inductively coupled plasma-mass spectrometry (ICP-MS) and ICP-OES chemical analyses, Philistine potteiy, 402-411 correlation to LA-ICP-MS elemental analysis, Matisse bronze sculptures, 343-346 Olivella biplicata shell compositional chemistry, geographic patterning, 170-180 prehistoric textiles, 29,35 Roman Carthage curse tablets, 319, 332,333/... 

We investigated eight Matisse bronze samples that were either small pieces of metal or shavings saved from the drilling of mounting bolts. All sculptures were made using lost wax casting except for The Serf which was sand cast (Table III). 

---