Eddy current testing methods were used to characterize model gilding layers on bronzes and to make measurements on gilded bronze objects in the Freer Gallery of Art. Swept frequency eddy current measurements were able to discern differences between gilding layers of varying thickness, and also between gilding layers of varying composition, i.e. conductivity, on a series of reference samples. Measurements made on the Freer Gallery objects were consistent with reference sample results for mercury amalgam gilding.

The critical depth of penetration of the eddy currents was calculated for gold and bronze for each of the four frequency ranges used. A simple model was developed to describe the influence of the different types of gilding layers on the change in probe impedance resulting from the interaction of the sample and probe. The model treats gilding layers of three different thickness ranges separately. In the first case, the gilding layer thickness is less than the critical depths for gold or bronze, tAu layer < tcr(Au). , tcr(bronze). In the second case, the gilding layer thickness is greater than the critical depth for gold, yet less than the critical depth of the bronze, tcr(Au) < tAu layer < tcr(bronze), and in the third case, the gilding layer thickness is greater than the critical depth for both gold and bronze, tcr(Au) , tcr(bronze) < tAu layer.Variations in the bronze substrate composition and microstructure influence the

Cross section of a leaf gilded sample. Bar indicates 50 microns.

measurements. Thus a differential measurement was used in this study for the gilded surfaces, normalizing the raw data by a measurement on a bare bronze surface. Surface roughness was found to influence results at high frequencies where the roughness is much less than the eddy current critical depth for the metal. Experiments using nonconducting Mylar were undertaken to model the effect of surface and subsurface corrosion, and lift-off. It was concluded that surface corrosion, and similarly lift-off, will decrease the eddy current penetration depth into the sample. Subsurface corrosion does not affect eddy current penetration into the gilding layer, but does decrease penetration into the bronze substrate. In both cases, the relative effect of the substrate and gilding layer on the probe impedance change due to the presence of the corrosion.

By making measurements at two frequency ranges, one at low frequency and one at high frequency, it was possible to discriminate between samples gilded using several different gilding techniques, including electrochemical methods, leaf gilding, mercury amalgam gilding, and foil gilding.

A second eddy current technique that uses a pulsed excitation source was briefly explored. The technique was able to discriminate between samples with 24K gold leaf, 23.5K gold leaf and 24K foil gilding.

This research was made possible through Grant MT-2210-9-NC-18 from the National Center for Preservation Technology and Training (NCPTT).

National Center for Preservation Technology and Training
645 University Parkway
Natchitoches, LA 71457

Email: ncptt[at]nps.gov
Phone: (318) 356-7444
Fax: (318) 356-9119