Chemical Testing

Ammonia vapor can cause beautiful blue crystals to form on the surface of bronze. To the eye these crystals look like azurite, which is often found on ancient corroded bronze. A quick chemical test can tell the difference. Solvents can distinguish between authentic lacquer, made from a tree sap, and modern synthetic lacquer. Simple chemical tests are useful in determining what further action to take when examining artifacts.

Chemical tests, performed on site at TK, are largely based on techniques practiced by mineralogists. They are often performed in conjunction with related tests of physical properties, such as streak color and hardness. A small number of chemicals can be used to detect the presence of numerous elements in a sample. The tests are not quantitative, though the relative proportions of the detected elements may be roughly estimated based on the reactions of a series of tests. With minute samples, impurities may not be detected. In addition to the normal range of chemicals used by mineralogists, solvents may be used to check for the presence of certain adhesives and resins, which are frequently incorporated in applied surface layers on repairs and forgeries. Tests are generally performed on samples of the corrosion crust found on metal objects. Occasionally, a small metal sample will be dissolved for testing. This is generally done when the alloy is unusual or suspicious. In the case of mineral samples, these simple and inexpensive tests generally eliminate the need for XRD analysis. Where the results are remarkable, more advanced testing may proceed.

Additional Information

One of the most important tests is for the presence of chloride in the corrosion on copper and its alloys. This anion is linked to the aggressive and detrimental corrosion phenomenon known as bronze disease. This corrosion can produce pitting, as well as internal expansion and subsequent weakening of the metal, leaving it more vulnerable to mechanical damage. In humid conditions, the copper chloride nantokite will absorb water from the atmosphere in a process known as deliquescence. This is the same process that causes salt crystals to stick together, or dissolve altogether, in sufficiently high humidity. Through hydrolysis and oxidation, the nantokite alters to an alkaline copper chloride such as atacamite or paratacamite.1 Subsequent dry conditions result in the opposite of deliquescence, known as efflorescence, whereby a substance loses excess water. Paratacamite effloresces to a powdery green form. An object with bronze disease displays this powdery corrosion, which may flake off. The chloride-containing material may then come in contact with another copper-based object, which may also become "infected" with the disfiguring corrosion.

When chloride is detected the artifact must be treated and stabilized, before it can be turned over to the collector. Salt in the burial environment is usually the source of chloride contamination of Chinese bronzes. Human sweat also contains chloride, and it is inadvisable to handle copper and its alloys with bare hands. Though the effect of chloride on other metals varies, ideally, no metal artifact should be touched with bare hands. While it is preferable to wear gloves, at the very least, hands should be thoroughly washed before touching artifacts.

Case Study

Perhaps the most interesting use of chemical testing at TK to date involved a pair of Tang Dynasty lead statues of recumbent horses. On the underside of each horse, corrosion of the metal was advanced, though the surface appeared minimally effected, to the naked eye. On Chinese artifacts, most lead-based corrosion encountered in large quantities is white or pale off-white shades. On the horses, some streaks of pale lead carbonate and chloro-carbonate, found on the animal's backs, were chemically identified. However, the samples from the undersides were a deep, rich, blood red. As expected, chemical testing yielded a strong positive for lead. Several oxides of lead can appear red, but the tests cannot detect oxygen. Preliminary physical tests, such as streak color and hardness, were consistent with minium, a lead oxide seldom reported on Chinese artifacts. As it was not possible to test for oxygen, an attempt was made to identify the sample by eliminating other lead-containing red minerals. Testing discarded nearly one hundred possibilities. It became apparent that the sample was not minium, but a mix of minerals. Unfortunately, chemical testing was inadequate for the purpose of identification, but it had proved that the sample was not common, and was thus worthy of further study. XRD analysis verified the presence of several forms of lead oxide, including minium.2 Minute impurities in the minerals resulted in slight variations on the normal colors for some of the identified substances. (The gems ruby and sapphire are both the mineral corundum, but with minute impurities causing the dramatic variation in color from red to blue.) The XRD report noted that impurities are common in naturally formed minerals, while synthesized samples tend to be purer. This was additional evidence to support the classification of the statues as naturally corroded antiquities. Had chemical testing not lead to XRD analysis, this important characteristic would have gone undetected.

Conclusion

Chemical analysis of materials, as performed at TK, generally serves to identify corrosion products on metal, and to detect soluble modern adhesives and coatings. The presence of chloride corrosion on metals, especially copper and copper alloys, is considered reason to treat the metal to prevent further potentially destructive corrosion. Chemical tests are usually conducted in conjunction with physical tests, as practiced by mineralogists. The findings may prompt further investigation, using more advanced analytical techniques.

Notes

1. Atacamite and paratacamite are not solely products of the alteration of nantokite, but can form independently as corrosion products.


2. The XRD analysis of sample 040102-11V was performed by H & M Analytical Services, Inc., Allentown, NJ. Chemical analysis, at H&M, was performed with a Quanx X-ray fluorescence unit operating in air. The X-ray diffraction pattern was collected with a Huber G670 Guinier diffractometer operating with a Cu radiation at 40 kV/30mA in a transmission mode. Total data collection time was 4 hours. Phase identification was by comparison to the Powder Diffraction File.

References

Gettens, Rutherford J. 1963. The corrosion products of metal antiquities. In: Annual Report. Smithsonian Institution, 1963. Publication 4588. Washington D.C.: Smithsonian Institution. p 547-568. Pages 552-54 pertain to copper chlorides.

Gettens, Rutherford J. 1969. The Freer Chinese Bronzes: Volume 2: Technical Studies. Washington, D.C.: Smithsonian Institution, Freer Gallery of Art. Chapter 8, Patina and corrosion. p 171-195. Pages 182-86 pertain to basic copper chlorides and the phenomenon of bronze disease.

Plough, Frederick H. 1988. A Field Guide to Rocks and Minerals. Boston, New York: Houghton Mifflin Company. 396 p.

Horie, CV; Kenyon, D. 1996. Chemistry for conservators. London: International Academic Projects, Distance Learning Programme. Block 4, Unit 11, Effects of water. p 1-11. Pagination within each unit begins with 1.

Scott, David A. 2002. Copper and bronze in art: corrosion, colorants, conservation. Los Angeles: Getty Conservation Institute. Chapter 4, Chlorides and basic chlorides. p 122-44. Pages 122-34 pertain to copper chlorides and information relating to bronze disease.