Abstract
The embrittlement of copper by small amounts of bismuth has been known for many years. Voce and Hallowes proposed recently that the agent responsible for embrittlement is bismuth present in the form of thin films at copper grain boundaries. In order to examine this picture from a different angle, internal friction and elastic modulus measurements were made on copper specimens free from and containing bismuth up to 0.01 percent with a frequency of transverse vibration of about 1000 cycles per second from −50°C to 550°C. It has been found that: (1) In the bismuth‐bearing copper heat‐treated to be brittle, there is no abrupt change in elastic modulus in passing through the melting point of bismuth (271°C). (2) The internal friction peak (around 500°C) associated with the stress relaxation across copper grain boundaries was reduced by the presence of bismuth, indicating that some bismuth did go to the grain boundaries. (3) Each bismuth‐bearing specimen gives an internal friction peak around 290°C and the height of this peak varies approximately linearly with the bismuth content. This ``bismuth peak'' is correlated to the embrittlement of copper. It was subsequently utilized in studying the effect of cold‐work, heat treatment and the rate of cooling upon the embrittlement of copper by bismuth. The effect of previous cold‐work was found only to speed up the precipitation of bismuth to the grain boundaries the rate of which is very slow at low temperatures. The data presented above corroborate the proposition that embrittlement of copper is caused by bismuth present at copper grain boundaries. However, it appears that the bismuth is distributed in a highly heterogeneous manner. Across most of the boundaries no appreciable amount of bismuth is precipitated, at least not enough to influence the grain boundary viscosity. It is these grain boundaries which give rise to that internal friction peak characteristic of pure copper. Across a few of the grain boundaries enough bismuth is precipitated to markedly decrease the grain boundary viscosity. Since no discontinuous behavior is observed as the temperature passes through the melting point of pure bismuth, it is concluded that the film of bismuth along these boundaries is not of sufficient thickness as to acquire the properties of crystalline bismuth.