Figure shows the XRD patterns of MgCu samples after ball milling for different times. It is found that some of Mg firstly reacts with Cu, forming the Mg2
Cu alloy in the primary stage of ball milling. As the milling time increases, the intensities of Mg, Mg2
Cu, and Cu peaks gradually decrease, and the corresponding widths broaden, which is mainly attributed to the grain refinement and accumulation of microstrain during the ball milling. The evolution of grain size and microstrain in the Mg and Cu is estimated using the single-line method of diffraction line-broadening analysis and illustrated in Figures
and , respectively (this estimation was only applied to the short-time milled samples for the reason that the diffracted peaks of Mg and Cu become unobvious when the ball milling time is longer than 5
h). One can see that the grain sizes of Mg and Cu both decrease, while their microstrains increase significantly as the ball milling proceeds. When ball milling time reaches 18
h, all the peaks of Mg, Mg2
Cu, and Cu cannot be recognized, and a broad halo peak appears in the XRD pattern. In previous studies [9
], it was suggested that the halo peak originated from the amorphous phase formed during the ball milling. However, considering the asymmetric shape of this broad halo characteristic of amorphous materials, another possibility that this halo peak results from the very small size of Mg2
Cu and Cu crystals is proposed in the present work: As the ball milling time increases, the sizes of Mg2
Cu and Cu become smaller, and consequently, the diffracted peaks of Mg2
Cu and Cu between 39° and 45° become so broad that they could overlap each other, finally forming the amorphous-like halo peak. Other peaks of Mg2
Cu and Cu cannot be found mainly due to the small size of Mg2
Cu and Cu nanocrystals (as discussed above, the size of grains have already decreased below 20
nm) after ball milling.
XRD patterns of MgCu sample after ball milling for different times.
Evolution of grain size and microstrain in Mg during ball milling of the MgCu sample.
Evolution of grain size and microstrain in Cu during ball milling of the MgCu sample.
Based on the above discussion, what is the real origin of the broadened halo peak in the milled sample? Are nanocrystals or amorphous phase present in the ball-milled MgCu sample? To answer these questions, it is needed to further investigate the ball-milled sample with the aid of microstructure observation and annealing experiment.
and show some representative transmission electron microscopy (TEM) images of different particles in the 18-h milled MgCu sample. According to the dark-field image and the corresponding selected area electron diffraction (SAD) pattern in Figure , the particle consisted of nanocrystals with a size of less 20
nm. Based on the SAD pattern, the nanocrystals are identified as Cu and Mg2
Cu. The result is consistent with the XRD patterns in Figure , in which the peaks of Cu and Mg2
Cu phase both gradually become broader and finally unrecognized due to the formation of Cu and Mg2
TEM dark-field image (a) and corresponding SAD pattern (b) of one 18-h milled MgCu particle.
TEM dark-field image (a) and corresponding SAD pattern (b) of another 18-h milled MgCu particle.
The microstructure of the particle in Figure seems different from the particle in Figure . The rings in the SAD pattern (see Figure b) are more diffused, and no clear ring can be observed. Only a halo is present in the SAD pattern. This type of pattern is always identified as amorphous in the literature. However, in the corresponding dark-field image of this particle (see Figure a), the presence of bright dots indicates that many nanocrystals still exist in this particle (about 10
nm). Observing Figure b more carefully, it can be found that the halo is located in almost the same position of the rings belonging to Cu and Mg2
Cu (see Figure b). Hence, it is speculated that the diffraction of a large amount of very small Cu and Mg2
Cu nanocrystals possibly results in the amorphous-like pattern in the Figure b. Different microstructures in the Figures
and from the same milled sample also imply that the size distribution of grains after ball milling is not uniform.
The DSC curve of the 18-h milled MgCu sample during the heating process is present in Figure . There are two exothermal peaks appearing in the DSC curve. In order to determine the origin of these two exothermal peaks, annealing experiments on 18-h milled MgCu samples were carried out at different temperatures, and the phase composition of annealed samples was studied by X-ray diffraction. One can see that no obvious change can be found in the XRD patterns of the as-milled sample and sample annealed at 130°C (the ending point of the first exothermal peak) except that a minor peak at about 40° appears (on the halo peak) in the XRD pattern of the sample annealed at 130°C (see Figure ). This peak was identified as the main peak of the Mg2Cu phase. It is explained that as annealing is performed, Mg2Cu nanocrystals start to grow, and they become so large that they can be detected by X-ray diffraction. Hence, the first exothermal peak seems to be associated with the growth of the nanocrystals and also the relaxation of stress and should not result from the crystallization of the amorphous phase. On the other hand, one can see that the diffracted peaks of MgCu2 appear in the XRD pattern of the milled MgCu sample after annealing at 240°C (the peak point of the second exothermal peak). Moreover, as annealing temperature increases, the peak intensities of MgCu2 and Mg2Cu become stronger, and finally, they are the main crystalline phases in the sample annealed at 350°C (the ending point of the second exothermal peak). Combined with the above results, the second exothermal peak is related to the reaction between nanocrystalline Mg2Cu and Cu, forming MgCu2. There is no obvious exothermic peak corresponding to the crystallization of the amorphous phase in the whole DSC curve of milled sample.
Measured DSC curves of 18-h milled MgCu sample. The sample was heated from 50°C to 450°C with a heating rate of 20°C/min.
XRD pattern of 18-h milled MgCu sample annealed at different temperatures.