Surface Effects and the Plasticity of Zinc Crystals

The effects of various surface treatments on the plastic properties of zinc crystals are examined. Roscoe’s observation of the hardening effect of a surface film of oxide is confirmed ; electrolytically polished specimens have critical shear stresses of about 33 gm mm-2 whereas oxidized specimens have values up to 65 gm.mzr2 . Similar effects are observed on the rate of creep under constant stress The Rehbinder effect-softening crystals by immersing them in paraffin and oleic acid-can be produced on oxidized specimens but not on others. Oxidized specimens do not respond instantly to immersion and the time of response is proportional to the viscosity of the liquid. I t is suggested that the liquid penetrates the oxide film and reduces its ability to harden the metal; the action of the oxide film itself is not yet clear. 5 1 I N T R O D U C T I O N H E success of Griffith’s theory of cracks in explaining the low breaking strength of brittle solids has encouraged occasional attempts to explain T the low yield strength of plastic crystals in a similar manner, although interest has moLtly been diverted in recent years to the theory of dislocations. One result of efforts to obtain evidence for the action of cracks has been the observation that the plastic properties of some metal crystals are affected by the surface condition of the specimens. Roscoe (1936) showed that oxide films of about 1,000 atoms thickness on cadmium crystals raised the critical shear stress of these to 2.4 times that of freshly cleaned crystals. This has been confirmed recently by Cottrell and Gibbons (1948). In explanation Roscoe suggested that molecules of oxide sealed up surface cracks and strengthened the outer layers of the crystal. Using crystals of zinc and tin, Rehbinder, Lichtmann and Maslenikov (1941) observed that the yield strength was the same for specimens tested in air and in a non-polar paraffii; oi!, but that, when 0.2y0 of oleic acid was added to the oil, the strength was halved. Similarly, the rate of flow in a creep test could be increased five to ten times by the addition of oleic acid. These effects were accompanied by an increase in the number of glide lamellae and a decrease in their size. I t was suggested that the oleic acid acted by penetrating surface cracks, causing them to open and expand. 332 S. Harper and A. H . Cattrell Although these observations show that plastic properties are affected by surface conditions they do not go far towards indicating the mechanism involved. I t was felt that further investigations were needed and the work on zinc crystals reported below was started. While this was in progress the results of two other investigations became known. Andrade and Randall (1948) reported experiments in which clean cadmium crystals were heated in air or in vucuo, or were immersed in solutions of cadmium and other salts. Heating in air for 20 hours at 2 0 o o C c . produced hardening and this was attributed to the formation of a surface film, in agreement with Roscoe's work. This hardening was produced more rapidly when the specimen was first contaminated locally with cadmium nitrate solution, In general, immersion in salt solutions caused the specimen to soften, although in the case of cadmium nitrate this was followed by a considerable hardening effect. Kemsley (1949) has recently attempted to reproduce the Rehbinder effect on tin crystals ; although the technique used in the original work was followed he was unable to obtain an observable response to treatment in oleic acid solution. 5 2. EXPERIMENTAL D E T A I L S All experiments were made on a sample of spectroscopically pure zinc supplied by Messrs. Johnson Matthey and Co., in the form of 1 mm. wire. A modification of the Kapitza method due to Andrade and Roscoe (1937) was used to grow crystals. By cleaning each wire before growth, using baked silica quills to support the wires while molten and an atmosphere of argon to reduce oxidation and volatilization, crystals could be prepared which had smooth, clean surfaces and which could be removed easily from their quills. Each crystal was cut into three pieces with the aid of a fine gas flame, care being taken to localize the heating and to minimize distortion. Two of the pieces, each 4 cm. long, were used for mechanical tests and the remaining one for determining the crystal orientation. Barrett and Levenson's (1940) etch-pit method was used for orientation measurements, with an accuracy within about 1". Some of the specimens for mechanical testing were electropolished, using the chromic acid solution suggested by Rodda (1943). These were each immersed in the solution along the axis of a nickel tube, which served as the cathode and an E.M.F. of 6 volts was applied for 10 seconds. This gave a highly polished surface free from irregularities and reduced the diameter by about 0.1 mm. In the early part of the investigation, the effect of the surface condition on the critical shear stress was studied. Most of these experiments were made on an apparatus of the type described by Andrade and Roscoe (1937). The loadextension curve was recorded photographically and it was possible to extend specimens at fairly fast rates, about lo-* cm.sec-1. A few experiments were also made at lower rates of extension, lo-*10-5 cm.sec-1, with an apparatus similar to that described by Polanyi (1925). As is usual in critical shear stress measurements a large scatter was observed, amounting occasionally in extreme cases to 20% of the average value. However, more reproducible results were obtained among specimens taken from the same crystal than from different crystals, suggesting that the scatter was due to some uncontrolled variation, for example in the gas content, among the crystals. To reduce the effect of the scatter repeated tests on several specimens were always made and their results averaged. Later experiments were made by creep testing, using an apparatus which enabled constant shear stress to be maintained on the glide plane in the glide Surface EJjCects and the Plasticity of Zinc Crystals 333 direction (Cottrell and Aytekin 1947, 1950). Creep testing had the advantage that the surface condition of a specimen could be changed while the test was in progress, so that difficulties due to variations in the properties of different ,pecimens were avoided. 5 3 C O M P A R I S O N O F V A R I O U S S U R F A C E C O N D I T I O N S Table 1 summarizes the measurements of the critical shear stress after applying various surface treatments. Untreated specimens (Series A) possessed light-grey oxidized surfaces and an average critical shear stress of 56 gm.mm-2. Series B and C represent various treatments, the common feature of which was a light attack by dilute hydrochloric acid. Apart from the C5 specimens, a critical shear stress of about 43 gm.mm-2 was observed in all cases, independently of whether the specimen was tested while immersed in the acid, made the electrode in a cell, or removed from the acid Table 1. Critical Shear Stresses of Zinc after various Surface Treatments (Rate of Strain2:10-2 sec-1) Number Average stress for Serial number Surface treatment of 0.5% extension tests gm.mm-a Freshly grown, untreated Tested in 10% HCI solution As B1, with specimen as anode of electrolytic As B1, with specimen as cathode Etched in HCI, washed, dried, exposed to As C1, exposed for 30 minutes As C1, exposed for 1 hour As C1, exposed for 18 hours As C1, exposed for several days Electrolytically polished, washed, dried, exposed to air for 5 minutes As D1, exposed for 1 hour As D1, exposed for 18 hours As D1, exposed for several days Polished, immersed in FeSO, solution, dried, As El, exposed for 1 hour Polished, immersed in HCI, dried, exposed As E3, exposed for 30 minutes Polished, dried without washing, exposed to Polished, immersed in ZnC1, solution, dried, cell