Jack Flotte and Jean Dommermuth
Richard Lippold’s sculpture Variations Within a Sphere, #10: The Sun, was commissioned in 1953 by the Metropolitan Museum of Art. It consists of approximately two miles of gold-plated wire spot-welded together to form a skeletal sphere within a conceptualized cube of about twenty-two feet in each dimension. Despite its relatively recent manufacture, the sculpture has had to be removed from display because of the formation on its surface of hundreds of thousands of small green pustules, which have totally destroyed the artist’s intended brightly reflective visual effect. The authors analyzed the wire to determine the structure of the metals and to identify the nature of the corrosion. We then sought a method to treat the sculpture and return it to its original brightness.
Analysis of the wire and its corrosion products
Small portions of the wire were embedded in epoxy resin, cut into cross- and taper-sections, and examined under the metallographic microscope. This examination revealed that the wire is composed of no less than five layers of metal. Elemental analysis with the SEM/EDS revealed the make-up of these layers. The 350 micron core of the wire is silicon bronze (94.1% Cu, 3.6% Si, small amounts of Mn, AI, and Zn); over this is a 0.5 micron layer of silver. Next is a 5 micron layer of brass (93% Cu, 6% Zn), an 8 micron layer of nickel, and finally a layer of gold 1.5 microns thick. The reason for all these layers is unclear. The artist has said that due to the inadequacy of the Met’s commission, he was forced to use cheap materials; perhaps this is wire that was converted from other purposes. It has also been suggested that some of the layers may have been lubricants used in drawing the wire (Lippold, Epner, 1993).
Metallographic analysis also revealed the structure of the corrosion. Typically, the green pustules consist of mounds of loose material surmounting pits that penetrate through all four layers of plating and deep into the silicon bronze core. The copper alloys have been attacked preferentially, the nickel to a lesser degree, and the noble metals least of all, in many cases forming small pinpricks in the layers of gold and silver over much larger gaps in the underlying layers. SEM/EDS analysis of the loose corrosion products demonstrated that chlorine was present, and X-Ray diffraction confirmed the presence of cuprous chloride, or nantokite, the root cause of “bronze disease”. It is possible that the chlorine may have been introduced into the wire during the plating process, during Lippold’s welding of the support wires, or during cleaning processes in the museum.
A series of experiments were conducted to determine the best treatment of The Sun, both in terms of its aesthetic appearance and the underlying cause of the corrosion. Several stripping agents were tested: alkaline Rochelle salts (sodium potassium tartrate in an aqueous solution of 5% NaOH) H2S04, NaCO4, EDTA (ethylenediamine tetraacetic acid) and BTA (benzotriazole). Samples of the corroded wire were then suspended in a container of each solution for specified periods of time, and then examined under the stereomicroscope. The most successful of the treatments was the alkaline Rochelle salts, which removed the corrosion from the wire within half an hour, leaving a bright specularly reflective surface. Because of the stratified structure of the pits, very little of the brightness of the gold was lost, since the diameters of the pits in the gold underneath the mounds of corrosion products were so small Once the success of this treatment was established, an additional element was added by testing corroded wire in a bath of alkaline Rochelle salts subjected to bursts of ultrasonic vibration. This addition was even more successful, taking only minutes to remove the corrosion.
Examination of a section of the treated wire under the SEM and the metallographic microscope showed that the superficial corrosion had indeed been stripped off, leaving the substrate metals essentially intact. Further examination of the wire to determine whether the nantokite layer has been totally removed, as well as tensile strength tests and accelerated aging tests of the wire in conditions of elevated humidity, remain to be completed. If these tests are successful, it is planned to test an entire section of the sculpture to determine the practicality of the large-scale operations necessary to treat the complete piece.
Although the use of stripping agents such as alkaline Rochelle salts has been in disfavor recently because of the danger of etching the substrate metal (Merk, 1978), the results of this experiment suggest that the use of stripping agents such as alkaline Rochelle salts may be advantageous to the conservator faced with a gold- or possibly silver-plated object with underlying corrosion. Certain limits, however, must be observed. First, the treatment should not endanger the exposed surface of the object or undermine the structure of the object. Second, it must not alter the intended aesthetic effect of the object. Finally it should not be used when less drastic solutions are applicable.
In conclusion, this project illustrates some of the problems that arise when dealing with art of the industrial age, when new materials and processes have been used whose properties the conservator as well as the artist may not entirely comprehend. Added to this factor is the effect of enlarged scale which magnifies the problems inherent in the materials. In cases such as this one, it may be necessary for conservators to re-evaluate their methods, and adapt old treatments to new situations.
Conversation with Richard Lippold and David Epner at the Metropolitan Museum of Art, November 23, 1993.
L.E. Merk, “A Study of Reagents Used in the Stripping of Bronzes,” Studies in Conservation, 1978, Vol. 23, pp. 15-2.2.