Page 247 - Copper and Bronze in Art: Corrosion, Colorants, Getty Museum Conservation, By David Scott

X-ray diffraction studies also showed that some green-colored layers present on objects
from the Thames sites consisted of antlerite, Cu 3 (S0 4 )(OH) 4 , and the unusual sulfate mineral
guildite, CuFe(S0 4 ) 2 (OH)-4H 2 0. No other occurrences of guildite have been reported. This
mineral may form from a more common mixed copper-iron sulfide, such as chalcopyrite, that
is transformed into the sulfate by oxidation after the object is removed from the burial environ­
ment. Until other examples of guildite are found, however, there is insufficient evidence to say
exactly how it does form. The mechanism by which the chalcopyrite or pyrrhotite develop is
also not known. The copper in the chalcopyrite may be a codeposit with iron when the elements
are precipitated from solution, or the copper may have diffused from the object into the surface
layers and substituted for iron in a defective film.
I "LAKE" AND "LAND" PATINAS Schweizer (1994) described
work on bronzes from human settlements around lakes near Neuchâtel, Switzerland, during
the late Bronze Age (i050-870 B.C.E.) Compositional analyses showed that the objects were
tin bronzes with about 8-9% tin, with an expected array of trace elements. Two primary
patina types were identified: "lake patinas" (on objects excavated from lake sites), which
had smooth, dense, brown-yellow surfaces; and "land patinas" (on objects from land sites),
which had thick green-blue layers incorporating some quartz grains. The land patinas were
identified as various mixtures of malachite, Cu 2 C0 3 (OH) 2 ; antlerite, Cu 3 S0 4 (OH) 4 ; and pos­
njakite, Cu 4 S0 4 (OH) 6 -H 2 0. Posnjakite is normally associated with antlerite, brochantite, and
atmospheric corrosion, and it is rarely mentioned in the context of excavated bronzes.
The identification of the true nature of the lake patina proved troublesome. In a prelimi­
nary publication, Schweizer concluded that sinnerite, Cu 6 As 4 S 9 , was present, but later quan­
titative studies showed that this was erroneous; chalcopyrite was later confirmed as the phase
actually present. On another bronze, a mixture of chalcocite, Cu 2 S, and djurleite, Cu 1 9 6 S, was
identified. Schweizer postulated that the chalcopyrite formed because of the preferential dis­
solution of copper, leaving a tin-enriched interior zone. The copper ions, iron, and sulfur then
combined and precipitated as a uniform layer of chalcopyrite.
Schweizer divides the objects from this site into three types of patina with the possible dis­
tribution into "lake" and "land" previously described: (1) objects with sulfide crusts, (2) objects
with both sulfide and carbonate crusts, and (3) objects with carbonate and sulfate crusts. It is
possible to make some inferences regarding these objects by examining the Pourbaix diagram
for the copper-sulfate-water system, shown in FIGURE 6.1. The sulfides, which are the primary
corrosion product, were formed under anaerobic conditions in a soil rich in organic matter and
in the presence of sulfate-reducing bacteria. The sulfide crust is about 100 μιη thick. The objects
with a sulfide corrosion layer followed by a carbonate layer represent the transitional survivors
from a reducing environment, where the sulfide layer is only partially oxidized to sulfates and
carbonates. The carbonate and sulfate crust formed in an aerated soil in contact with air and
grew from the oxidation of the primary sulfides.

C H A P T E R SI X
230