The Microstructure Of Electrodeposited Nanocrystalline Copper And Its Deformation Mechanism

A mass of useful models were summarized from the experimental results of traditional coarse-grained metals and alloys. Two of the most excellent models are Hall-Petch relationship and Coble creep equation. The fist one exactly describes the relationship between the material's grain size and its yield strength; the last one shows us that plenty of factors have effect on the grain diffusion rate and how these factors affect the grain diffusion rate. According to the Hall-Petch relationship, it is expected that the nanocrystalline (nc) materials should have ultra-high strength, tens time higher than that of coarse-grained materials. Based on the Coble creep equation, nc material would make it amenable to high creep rates and large-scale deformation at much lower homologous temperatures, so that ductile ceramics and diffusion creep of pure metals would be possible even at room temperature. Thereupon, the nanocrystalline materials should have ultra-high strength and, meanwhile, very good ductility. They should be excellent materials for engineering application. But how is the truth? What is the mechanism for nanocrystalline materials deformation? These confused questions are so interesting that a lot of scientists have devoted themselves into this topic for past decades.Consequently, five different nanocrystalline copper specimens were synthesized by electrodepostion techniques in this study. These microstructures were characterized by XRD and TEM. The grain refinement mechanism is summarized. The tensile properties of two specimens whose grain size are below 100 nm are tested by MTS. The deformation mechanism was discussed according to the tensile test results. We also synthesized a copper with large angle grain boundary, equiaxed grain and grain size in 25 nm by electro brush-plated technique, and conducted compressive creep on this materials. Based on these experimental results, we have arrived at the following conclusion:1. In the electrodeposited copper specimens, we identified two different grain refinement mechanisms, individually or mutually, operating. Firstly, twin-twin intersection is a crucial point for grain subdivision process in the present specimens. It directly creates grains with large-angle boundaries. These grains can be as small as 10 nm. While the growth twins play an important role in grain refinement, the dislocations also contribute. When dislocation configurations like dislocation walls or sub-boundaries are well developed in twins, they cut the twin lamella into small piece; and the dislocations can be absorbed by twin boundaries to make the twin block curved and round like the ordinary grains. As the dislocation tangles transform into sub-boundaries and highly misoriented grain boundaries, they can also refine the microstructure individually. The process of grain refinement to nanometer scale in electrodeposited copper is very similar to that of severe plastic deformation metals, but twin play a coequal role as dislocations in electrodeposited copper, as copper has low stacking fault energy. So we classify the electrodeposition technique into up-down type.2. A ns Cu with a mixture of nc grains and nanoscale growth twins in sub-micrometer grains was synthesized by electrodeposition. Its strength and ductility depended on both the mean grain size and twin lamellar thickness. The ns Cu showed two-stage strain rate sensitivity, i.e., a high value of 0.054 at strain rates less than 10-3 s-1 and a low value of 0.016 at strain rates higher than 10-3 s-1. Accordingly, the activation volume showed reverse variation. The ns Cu also showed very different surface and fracture morphologies and different fracture directions when deformed at low and high strain rates. These results demonstrated a deformation mechanism transition from dislocation deformation at higher strain rates in larger grains to both dislocation deformation dominated in larger grains and GB diffusion and GB sliding dominated in small grains at low strain rates. It is generally accepted that for nc and ns metals there is a critical grain size across which a transition from dislocation motion to GB diffusion- and sliding-dominated deformation happened. The mean grain size and twin lamellar thickness of the ns Cu is in the range where the dislocation deformation should dominate. However, the GB diffusion and sliding are proved to take place in low strain rate region. So it is deduced that the critical grain size is strain rate dependent.3. A nc Cu with grain size of 72nm was obtained through a direct current electrodeposition technique. This nc cu showed a broad grain size distribution ranging from 10 to 290nm. Strain rate sensitivity and activation volume were estimated to be 0.085 and 12b3, respectively, from tensile flow stress at a fixed strain of 1%. Such high m value and small activation volume suggested that besides dislocation deformation, GB activities like sliding and diffusion, would take place. And both the m andυvalues may be the average contributed by both dislocation deformation and GB activities. The nc Cu exhibited a elastic-nearly perfectly plastic behavior at specific low strain rates. No apparent necking and shear band were found in our results. The high m would be responsible for the phenomena. The strength and ductility were optimized at strain rate of 1×10-4 s-1 with ultimate strength of 620MPa and a ductility of 22%, which was attributed to the combined action of lattice dislocation slip and GB activeties. A high strength of about 1.04GPa was observed at the high strain rate of 0.1s-1.4. In summary, the linear strain-rate dependency on applied stress (n=m=1.0) and an apparent activation volume of 1.5b3 were observed for a 25 nm brush-plating Cu, which proves that Coble creep dominates the deformation of the nc Cu at very low strain rate. A transition from Coble creep to power-law creep with n=10 or m=0.1 occurred at strain rate higher than 4.5×10-7s-1. Both the stress level and the m value at high creep strain rate were connected smoothly to those obtained by compressive test at corresponding strain rate, so a panoramic view of stress-strain rate behavior of this nc Cu showed the deformation mechanism transition from dislocation activity at very high strain rate to Coble creep at very low strain rate with a wide intermediate strain rate range where both dislocation motion and GB activities operate.5. After a mass of twin boundaries were introduced into the ultra fine grain, the yield stress of NT Cu was much enhanced. The twin boundaries are effective on blocking dislocation motion and improve tensile strength. The work hardening capacity of NT Cu is losing as the intersection of dislocation hardly happen in the nano-scale lamella between twin boundaries. But NT Cu still has an excellent ductility as UFG Cu, for it has a much higher m value than that of UFG Cu. This high m value is helpful in improving ductility. The activation volume of NT Cu is reduced compared to UFG Cu, for the region for dislocation motion in NT Cu is smaller than that of UFG Cu and, some times, only partial dislocation can penetrate the twin boundary.