Page 54 - Copper and Bronze in Art: Corrosion, Colorants, Getty Museum Conservation, By David Scott
P. 54

the Getty Conservation Institute Museum Research Laboratory as part of a technological study
            of  the  material  (Scott  1999).  These  copper  fragments  date  from  about  4000  B.C.E.,  making
            them six thousand years old.  They are smelted copper, relatively pure, with only a few parts  per
            million of typical elemental impurities, such  as silver, arsenic, gold, antimony, and lead; yet the
            corrosion penetration through the copper  averages only about  30  μιη.  The corrosion consists of
            a thin layer of cuprite and a thinner layer of malachite. This patina translates into a corrosion
            rate of 0.005 μπι per year, and in some cases, less than that. Even in the most benign soils, this
            is  a substantially lower rate than would  be inferred from  the  recent  experimental work.  One
            possible explanation is that, in undisturbed contexts, the rate of corrosion of unalloyed  copper
            progressively decreases as the cuprite layer develops  and impedes further attack on the  copper,
            slowing the corrosion processes to levels far below those experimentally predicted, particularly
            for  soils low in chloride-ion  content.
                Groundwater  flowing  through  igneous  bedrock  accumulates  insignificant quantities of
            chloride ion but  may  gain  appreciable  concentrations  of Si0 2 ,  F~, Na ,  HC0 3 ~,  and  Ca .
                                                                                      ++
                                                                      +
            Such groundwater may be slightly acidic. In sedimentary and metamorphic rocks, the ions pres­
            ent  may  include  substantial  Cl~, 0 4 ,  Ca ,  C 0 3 ,  Fe 2 ,  A l 3 ,  and N0 3 ~. Here  the  water
                                        S
                                                      =
                                                           +
                                                               +
                                                ++
                                           =
            may be neutral or alkaline (Baas Becking, Kaplan, and Moore i960).  Plants  and bacteria  may
            contribute about 550-2750 ppm of chloride ions in regions of dense vegetation. Chloride ions in
            groundwater may also be derived from the atmosphere,  from evaporites such as halite, and from
            the decomposition of certain micas.
               Highly organic soils or soils over calcareous bedrock have high carbon dioxide contents  and
            may be chemically very aggressive  because the carbon dioxide may react with water to form car­
            bonic acid (Garrels 1954), which may attack metals directly and prevent the formation of a pro­
            tective  film  on the metal surface  (Wilkins  and Jenks 1948).  Calcareous  soils may  also  act in a
            quite benign fashion, however, especially f carbon dioxide and water produce  the soluble cal­
                                              i
            cium bicarbonate. This may act to protect the bronze  from  corrosion: since calcium bicarbon­
            ate is a salt of a weak acid, its aqueous solution is alkaline, and by binding with carbon dioxide,
            it  prevents  the  extensive  dissolution of copper (I)  ions  (Geilmann 1956). At values of pH >  8,
            calcium bicarbonate  precipitates  as  carbonate,  and, in subsequent acidic conditions, this may
            dissolve instead of copper (II)  compounds. The overall pH in dilute natural groundwater is prin­
            cipally controlled by this CaC0 3 — H 2 0 —  C 0 2  equilibrium (Garrels and Christ 1965).
               The groundwater in soils may have a significant phosphate content, which may be  delete­
            rious  to  buried bronzes,  although  this  is frequently  controlled by  the  aluminum content of
            the  soil  (Lindsay  1979).  The  phosphate  activity  may  also  be  dependent  on  the  speciation
            of  the  phosphorous.  For  example,  the  chemical nature  of dissolved phosphate  may  be  prin­
            cipally P0 4 ~  at pH  >  12 or H 2 P0 4 ~,  between pH 2 and  7.5, so  even in this  case, it is not
                     3
            possible to predict all  the effects  of soil phosphate content simply by knowing how much phos­
            phate is in the soil.



                                                    C O R R O S I O N  AN D  E N V I R O N M E N T
                                                                       37
   49   50   51   52   53   54   55   56   57   58   59