Cu alloy synthesized by hydrogen reduction

0 downloads 0 Views 3MB Size Report
advantages of synthesizing this alloy by the hydrogen reduction route were investigated ... on the other hand, showed the presence of two phases, pure W and pure Cu. The SEM ..... would be expected from the phase diagram. The particle.
Morphology studies of a W/Cu alloy synthesized by hydrogen reduction U. Tilliander,a) H. Bergqvist, and S. Seetharaman Department of Materials Science and Engineering, Division of Materials Process Science, Royal Institute of Technology (KTH), SE - 100 44 Stockholm, Sweden (Received 20 October 2005; accepted 9 March 2006)

Because of the applications for W/Cu composite materials in high technology, the advantages of synthesizing this alloy by the hydrogen reduction route were investigated, with special attention to the properties of the product that was formed. Kinetic studies of reduction indicated that the mechanism changes significantly at 923 K, and the product had unusual properties. In the present work, morphological studies of the W/Cu alloy with 20 wt% Cu, produced at 923 K, were carried out by x-ray diffraction (XRD), scanning electron microscopy (SEM), and transmission electron microscopy (TEM) analyses. The structural studies performed by XRD indicated that, at 923 K, Cu dissolved in W, forming a metastable solid solution in the nanocrystalline state. The samples produced at higher as well as lower temperatures, on the other hand, showed the presence of two phases, pure W and pure Cu. The SEM results were in agreement with the XRD analysis and confirmed the formation of W/Cu alloy. TEM analysis results confirmed the above observations and showed that the particle sizes were about 20 nm. The structure of the W/Cu alloy produced in the present work was compared with those for pure Cu, produced from Cu2O produced by hydrogen reduction under similar conditions. This indicated that the presence of W hinders the coalescence of Cu particles, and the alloy retains its nano-grain structure. The present results open up an interesting process route toward the production of intermetallic phases and composite materials under optimized conditions.

I. INTRODUCTION

The demands for materials with unique properties are increasing in a number of high-technology areas, for example in electronics. Processing multiphase composite materials using the powder metallurgy approach is rapidly becoming an important fabrication route for advanced materials production. A promising process route for the production of alloys in powder form is hydrogen reduction of metal oxides. Because metal powders can be produced from the respective oxides by reduction using a suitable gaseous reductant, the alloy formation can be achieved at fairly low temperatures, resulting in fine-grained alloys or materialmatrix composites. In the present work, a hydrogen reduction route was used, and it represents a direct production of the W/Cu material of unusual morphologies and with uniform composition. This route offers a unique advantage as the

a)

Address all correspondence to this author. e-mail: [email protected] DOI: 10.1557/JMR.2006.0181 J. Mater. Res., Vol. 21, No. 6, Jun 2006

materials produced under well-defined conditions have enough active centers to achieve good chemical bonding between the dispersed material and the matrix during the production process. As the number of unit processes is kept to a minimum and no environmentally dangerous effluence occurs in the process, the hydrogen reduction process offers a green route toward the production of alloy powders and intermetallics. Composite materials of W/Cu are used today in a variety of applications where the thermal and electrical properties of Cu and the very low expansion characteristics of W can be advantageously utilized. The combination of these two elements even leads to the optimization of other alloy properties such as ductility, mechanical strength, corrosion, and wear resistance at elevated temperatures. Primary applications for W/Cu composite materials are electrical contacts and electrodes, welding and electro-forging dies, heat sinks, and packaging materials. Because W and Cu show an extremely limited solid solubility,1–4 only a few processes have been known to be commercially successful. The most common methods for the fabrication of W/Cu composite material are the © 2006 Materials Research Society

1467

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

infiltration process, where liquid Cu is filled in a coldpressed porous W skeleton; liquid-phase sintering of hotpressed blends of Cu and W powders; and the mechanical alloying process, which has been used to synthesize W/Cu nanocomposite powders.5–12 Unfortunately, the latter process causes problems with impurities in the final material, introduced during processing. The abovementioned processes have been successful only on a limited scale, and there is a strong need for the development of new process routes toward the production of W/Cu composites with tailor-made properties.9,13–18 The samples used in this work were reduced under isothermal conditions at 923 K by using thermogravimetric analysis (TGA).19 The product’s morphology and chemical composition could then be characterized by xray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM). II. EXPERIMENTAL DETAILS

The nanostructured W/Cu alloy powder, with a composition of W–20 wt% Cu, was produced by a hydrogen reduction route. Complete reduction and alloying in one step was accomplished by optimized process parameters, namely, experimental temperature, reduction gas flow rate, and sample quantity. A. Materials

The raw materials used were a mixture of commercial and pure Cu2O (Sigma-Aldrich, Sweden) and WO3 (Atlantic Equipment Engineers) powders with average particle sizes of 1.3 and 2 ␮m, respectively [determined by dynamic light scattering (DLS) analysis]. The Cu2O and WO3 powders were mixed together in an agate mortar and ground well before use. A SEM image of the Cu2O–WO3 mixture (×5000) is presented in Fig. 1. The mixture was subjected to SEM analysis to examine the particle size and shape as well as to make sure that the oxides were homogeneous. The figure shows that the oxides are mixed together well and that there are some differences in both particle shape and size. The WO3 has a larger particle size and a cubic shape, and the Cu2O is smaller and is more spherical. During the heating sequence of the reduction experiments, argon (plus grade) was used as a protective gas. Hydrogen gas (scientific grade) was used as the reductant. Both the argon gas and the hydrogen gas were supplied by AGA GAS, Stockholm, Sweden. B. Characterization of the starting materials

In the course of isothermal reduction experiments, the reaction showed a decrease in reduction rate in the temperature range 923–973 K. During the reduction experiments, the starting oxide mixture is heated to the targeted temperature and maintained in argon atmosphere before 1468

FIG. 1. SEM image of the surface of the powdered raw oxide material.

the reduction reaction is initiated by letting in the hydrogen gas. During the heating period, the oxide mixture could have possibly reacted to form a ternary compound accompanied by melting at temperatures higher than 873 K. If a complex oxide has been formed before the reduction takes place or if the reactants melt, the nature of the reduction reaction would be quite different. Because it was felt that the Cu2O–WO3 mixture could undergo a chemical reaction and/or sintering at these temperatures, it was necessary to characterize the starting material used in the present work. Toward this aim, differential thermal analysis (DTA) experiments were performed in argon atmosphere on the starting material mixture. The DTA experiment revealed a strong endothermic peak at 904 K.19 This suggests that a compound is formed by an endothermic reaction between the two component oxides or that a first-order transformation—for example, formation of a liquid phase—has occurred. This necessitated further examination of the heated oxide mixture. To gain a deeper understanding of the state of the Cu2O–WO3 mixture during heat treatment, the mixture was heat-treated under controlled conditions in argon atmosphere. The heat-treated Cu2O–WO3 mixture was investigated by XRD.19 The result showed clearly that the mixture has formed a new compound during the heattreatment. However, identification of the new phase was not possible as the XRD-peaks did not correspond to the standard patterns in the database. Further attempts are being made to identify the compound. The morphology of the heat-treated Cu2O–WO3 mixture was investigated by SEM. Figure 2 shows the oxide material after the heat treatment at two different magnifications: (a) ×1000 and (b) ×5000. In Fig. 2, it can be seen that the powder mixtures have two different particle shapes, one shape with round or faceted and equiaxed particles with a mean diameter of 0.2 ␮m, the other with needlelike particles. The partly transformed part with a

J. Mater. Res., Vol. 21, No. 6, Jun 2006

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

Scanning Electron Microscope accompanied by electron probe microanalysis (EPMA) and energy dispersive spectrometry (EDS) for local chemical analysis. XRD analysis was also preformed at room temperature using a Philips PW 1830 diffractometer (X’pert System). An operating voltage of 40 kW and a emission current of 30 mA were used to a monochromatized x-ray beam with nickel-filtered Cu K␣ radiation (␭ ⳱ 1.5418 Å). A continuous scan mode was used to collect data from 30° to 90° with a 0.02 sample pitch and a 4° min−1 scan rate. The observations of the microstructure and the chemical composition on the reduced Cu/W powder were made by the JEOL/JSM-840 SEM and a field emission gun SEM (FEG-SEM; JEOL 8900) accompanied by wavelength dispersive spectroscopy (WDS). Both backscattered electron- and secondary electron imaging were performed (BEI and SEI). The nanostructured powder samples were prepared in two different ways before they were examined by a 2000EX JEOL TEM operating at 200 kV. In the first case, the W/Cu powder particles were dispersed in ethanol and then dispersed on a carbon supporting film on a 200 mesh aluminium grid (supplied by Agar Scientific). In the second case, the W/Cu powder was compressed into compact disks with a diameter of 3 mm and a height of approximately 100 ␮m. The disks were then dimpled to electron transparency in an ion miller, a precision ion polishing system (PIPS; Gatan). The average chemical composition in the powder samples were analyzed by energy-dispersive x-ray spectroscopy (EDXS; INCA Energy TEM: microanalysis system for TEM). III. RESULTS AND DISCUSSION FIG. 2. SEM images of the phase morphology in the oxide powder mixture after heat treatment at 973 K in 48 h: (a) magnification of ×1000 and (b) magnification of ×5000.

A. Isothermal reduction

To further illustrate the phenomena with reaction rates that decrease as temperature increases, some selected

needlelike structure is a characteristic structure for materials crystallized from a liquid phase. C. Apparatus and procedure

The reduction experiments of the Cu2O–WO3 mixture were carried out in a temperature interval of 673–1073 K by means of a Setaram TGA 92 thermoanalyzer. The experiments were monitored at intervals and the Setaram computer software for the TGA 92 thermobalance enabled the handling of the data from the experiments. A detailed description of the apparatus is given in earlier publications.20,21 The reduction experiments were performed in the TGA reaction chamber. The experimental details are given in an earlier publication.19 Preliminary observations of the microstructure and the chemical composition were made by a JEOL/JSM-840

FIG. 3. The fraction of reduction X as a function of time t for the isothermal reduction experiments; mass of sample: 15 mg; flow rate of H2: 0.6 nl/min.

J. Mater. Res., Vol. 21, No. 6, Jun 2006

1469

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

isothermal reduction curves, obtained in the present study, are presented in Fig. 3. In this figure, the dimensionless mass change (X) is plotted against time and used to describe the extent of the reaction. X represents the

FIG. 4. XRD patterns of the isothermal reduction experiments, carried out at 773, 923, and 973 K. The standard patterns for W and Cu is shown as dotted lines.

FIG. 5. SEM images of the powder surface after the isothermal reduction experiments performed (a) at 973 K and (b) at 923 K. 1470

FIG. 6. SEM images, at different magnifications, of the powder surface after the isothermal reduction experiments performed at 923 K: (a) ×10k, (b) ×70k, and (c) ×100k.

J. Mater. Res., Vol. 21, No. 6, Jun 2006

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

FIG. 7. TEM images of the W/Cu powder, at a magnification of ×250k, with (a) equiaxed grains, together with (b) corresponding diffraction pattern and (c) elongated banded grain structure together with (d) the corresponding diffraction pattern.

ratio of an instant mass loss to a theoretical final mass loss of all oxygen atoms from the oxide mixture. From Fig. 3, it can be seen that the reduction rate was found to increase gradually with temperature in the interval of 823–923 K. On the other hand, the rate of reduction was found to decrease between 923 and 973 K. At temperatures higher than 973 K, the reduction rate was once again found to increase. The reduction reaction slows down at 923 K, and the reduction rates in the temperature region 973–1073 K are lower than the rate at 923 K in this region, as can be seen in Fig. 3. The activations energies calculated from the reduction curves in Fig. 3 show two distinct values, namely, 97 kJ/mol below 923 K and 27 kJ/mol above 973 K and that there is a distinct change in the mechanism of reduction around 923 K.19 An examination of the precursor revealed the formation of a possible compound phase, and the mechanism of reduction has been explained in detail in an earlier publication.19 B. X-ray diffraction

The isothermally reduced Cu2O–WO3 mixture was also subjected to XRD analysis, and the diffraction patterns obtained are presented in Fig. 4. The XRD patterns correspond to the fully reduced samples from the isothermal experiments at 773, 923, and 973 K, and the standard patterns for W and Cu are shown as dotted lines. The occurrence of high and sharp peaks corresponding to elemental Cu is clearly seen in the XRD pattern

TABLE I. Calculated hkil values for the diffraction in Fig. 7(b). (hkil)

dcal (Å)

101¯ 0 2000 101¯ 1 101¯ 2 112¯ 0 112¯ 1 101¯ 3 112¯ 2 202¯ 1 112¯ 3 101¯ 4 202¯ 3 123¯ 0 202¯ 0

2.186 2.221 1.961 1.560 1.262 1.214 1.226 1.097 1.061 0.953 0.975 0.873 0.826 1.093

corresponding to the reduction at 773 and 973 K. On the other hand, the pattern corresponding to the intermediate temperature, 923 K, shows no clear elemental Cu peaks. These may indicate the alloy formation between Cu and W. In the pattern corresponding to 923 K, the peaks showing elemental W are found to be broader and exhibit a significant shift in comparison to the fully reduced sample at 773 K. This strongly suggests that the Cu formed by reduction might have dissolved in the formed W-phase. At 923 K, these peaks are also much smaller. These XRD results indicate that the alloy formed at 923 K is nearer to the nanocrystalline state. In the present

J. Mater. Res., Vol. 21, No. 6, Jun 2006

1471

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

FIG. 8. (a) TEM image of the W/Cu powder at a magnification of ×250k. (b) Corresponding dark field image.

work, the results at 923 K might suggest the formation of a metastable solution of Cu in W in solid state. Similar possibility has been indicated by Subramanian et al.4 C. SEM investigation

Figure 5 shows a SEM image of the powder surface, at a magnification of ×5000, after the isothermal reduction experiments performed at (a) 973 K and at (b) 923 K. The significant difference that can be seen in these two figures is that the product formed at 923 K has a structure 1472

FIG. 9. (a) TEM image of a particle with elongated grain structure (×250k). (b) Corresponding diffraction pattern showing a hexagonal structure.

totally different from that of the sample produced at 973 K. The material, shown in Fig. 5(b), appears to have a fractal growth of facets. The analysis carried out in the imaging modes (BEI and SEI) and the FEG-SEM on the product strongly indicates the formation of a W/Cu alloy with significant Cu solubility. Further, three SEM images are presented with different magnifications in Fig. 6(a)

J. Mater. Res., Vol. 21, No. 6, Jun 2006

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

FIG. 10. TEM images of (a) W/Cu particle with a magnification of ×70k, (b) decomposed particle with corresponding diffraction pattern (×10k), (c) W/Cu grains dispersed in the vicinity of a particle (×50k) with an enclosed diffraction pattern, and (d) grains formed at longer distance from the decomposed particle (×250k).

×10k, Fig. 6(b) ×70k, and Fig. 6(c) ×100k. With higher magnification, the pictures reveal that the powder contains of two different structures. In Fig. 6(b), the fractal growth of facets is clearly seen, and in Fig. 6(c), the material seems to be a brittle agglomerate consisting of small spherical particles. D. TEM investigation

In light of all the experiments performed in the TGA and all the analysis carried out with XRD and SEM, the experiment carried out at 923 K resulted in the most fine-grained W/Cu material. On the basis of these results, all the experiments performed by TEM were samples produced at isothermal reduction conditions at 923 K. In Fig. 7(a), the structure in a powder particle is shown with a magnification of ×250k. The particle structure consists of equal axed grains with an average grain size well below 20 nm. A number of the grains exhibit a lamellar contrast, and the grain boundaries are affected by a shear. This phenomenon is often seen at twinning. In a greater detail, the grain boundaries show a rotation in points where the lamellae intersect grain boundaries. Figure 7(b) shows the electron pattern corresponding to the lamellar structure in the grains [Fig. 7(a)]. The diffraction pattern reveals the presence of preferred texture orientation of the crystal lattice in the lamellar structure. Furthermore, it can be seen that the pattern shows a broad spread in lattice space regarding the diffraction rings.

This is, to some extent, related to the occurrence of Cu spots for the (111) and (200) planes, which indicates the existence of Cu in the powder material. Referring to the diffraction pattern in Fig. 7(b), the crystal structure was identified as a hexagonal lattice with a ⳱ 2.52 and c/a ⳱ 1.76. The calculated values are presented in Table I. Figure 7(c) shows a particle structure different from that of the same powder sample, as in Fig. 7(a). The corresponding diffraction pattern is presented in Fig. 7d and gave the same information about the crystal structure as in Fig. 7(b). Figures 8(a) and 8(b) represent the elongated grain structure in a bright field and a dark field, respectively. The bright-field image [Fig. 8(a)] shows the diffraction contrast structure. The dark field image, Fig. 8(b), was created from one of the diffracted spots in the inner ring of the diffraction pattern. The bright areas represent the hexagonal phase in the lamellae as well as in a large part of the grains in this image. Further examination of the elongated structure was carried out by exposing mainly one grain, located at the edge of the foil, to the electron beam. Figure 9(a) shows an image of the grain and the corresponding diffraction pattern is presented in Fig. 9(b). The diffraction spots, originating from the single grain, reveal a hexagonal structure, whereas the diffraction rings originate from the poly crystalline area. Figure 10 describes the phenomena in the TEM

J. Mater. Res., Vol. 21, No. 6, Jun 2006

1473

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

as it occurred during the experimental observation. Figure 10(a) shows a grain after it was exposed to the electron beam and dispersed. Material from the grain was condensed on the carbon film, resulting in the presence of a large number of small particles with a size inversely proportional to the distance from the bigger original particle [Figs. 10(b), 10(c), and 10(d)]. The phases present in the transformation are not identified. However, it was found that no W or Cu diffraction spots were present in the diffraction patterns. It would be very interesting to compare the morphology observations in the case of the reduction product of the Cu2O–WO3 mixture with those obtained by the reduction of pure Cu2O.20 In the latter case, the reductions could be carried out at lower temperatures, namely, 553– 673 K, whereas, in the case of the Cu2O–WO3 mixture, the reduction temperatures are slightly higher, 673– 1073 K.19 It would be reasonable to expect sintering and growth of the particles if they are produced at a higher temperatures. The SEM micrographs of Cu powder produced by the reduction of Cu2O at 553, 613, and 673 K are presented in Fig. 11. It is clearly seen that the particles coalesce into each other as the reduction temperature is increased. The estimated average particle sizes are 0.2 ␮m for the product at 553 K, 0.5 ␮m for the product at 613 K, and 0.8 ␮m for the product at 673 K. The grain growth is approximately linear with increasing temperature. This observation is in agreement with the reasoning in the earlier paragraph. It would be interesting to compare these results with the SEM micrograph in Fig. 6(c) for the product of Cu2O–WO3 reduction at 923 K. The average particle size of the reduction product can be estimated to be around 20 nm. This would mean that the co-reduction of the two oxides results in the production of nanoalloys. As mentioned earlier, the co-reduction experiments were carried between 673 and 1073 K. It has been mentioned earlier that the W/Cu alloy is produced only around 923 K. Both at lower as well as higher temperatures, namely, 773 and 973 K, the XRD analysis has shown that W and Cu exist as separate phases,19 which would be expected from the phase diagram. The particle sizes for Cu estimated from the SEM pictures for the products at these temperatures are around 0.2 ␮m. Thus, it appears that the alloy formation and a metastable solution appear to be important factors in the formation of nanoalloys. It is quite interesting in this respect to reexamine Figs. 10(c) and 10(d), where the W/Cu alloy grains are dispersed. The dispersion is caused by the impact of the electron beam. While the impact energy causes dispersion, no decomposition of the alloy into W and Cu takes place as revealed by EDS analysis. The particles, which are about 10 nm are still W/Cu alloy particles. Because 1474

FIG. 11. SEM micrographs showing the fully reduced powder with the effect of different reduction temperatures: (a) 553 K, (b) 613 K, and (c) 673 K (at the same magnification, ×15k).

the electron beam impact is a transient event, it is possible that fast cooling after the impact does not bring the system to an equilibrium structure. The above observations open up a number of possibilities in reduction processing to produce nanoalloys and composites. The hydrogen reduction process parameters can be suitably manipulated to tailor the product

J. Mater. Res., Vol. 21, No. 6, Jun 2006

U. Tilliander et al.: Morphology studies of a W/Cu alloy synthesized by hydrogen reduction

structure. This would require extensive experimentation as well as theoretical calculations to model the processes toward desired properties. Such attempts are currently being made in the present authors’ laboratory.

4.

5.

IV. SUMMARY AND CONCLUSIONS

6.

The present investigation deals with the characterization studies of morphology and chemical composition of a nanostructured W/Cu alloy powder, with a composition of W-20 wt% Cu. The powder was produced by a hydrogen reduction route and characterized by XRD, SEM, and TEM. The sample with optimum properties, synthesized at 923 K, is highlighted. The XRD pattern shows no clear elemental Cu peak and peaks showing W are found to be both broader as well as smaller and exhibit a significant shift from the peaks for pure W. This strongly suggests that the Cu formed by reduction might have dissolved in the W-phase that was formed. The analysis further indicates that the alloy formed at 923 K is nearer to the nanocrystalline state. In the present work, the results might suggest the formation of a metastable solution of Cu in W in the solid state. The TEM analysis revealed that the particle of the W/Cu powder consists of two different structures, one with equiaxed grains (with an average size well below 20 nm) and one with elongated grains. Both of the crystal structures were identified as hexagonal lattices. When a grain of the powder sample was exposed to the electron beam, during the experimental observations in the TEM, material from the grain was condensed. The result was a large number of small particles having a size inversely proportional to the distance from the bigger original particle.

7. 8. 9. 10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

REFERENCES 1. T.B. Massalski: Binary Alloy Phase Diagrams, 2nd ed. (ASM, Materials Park, OH, 1990), Vol. 1, pp. 1503–1504. 2. M. Hansen: Constitution of Binary Alloys, 2nd ed. (Mc.Graw-Hill Book Company, Inc., 1958), p. 649. 3. O. Knacke, O. Kubachewski, and K. Hasselman: Thermochemical

20. 21.

Properties of Inorganic Substances, 2nd ed. (Springer-Verlag, Berlin, Germany, 1991). P.R. Subramanian, D.J. Chakrabarti, and D.E. Laughlin: Phase Diagrams of Binary Copper Alloys, 1st ed. (ASM, Materials Park, OH, 1994), pp. 475–477. Metals Handbook, 9th ed. (American Society for Metals, OH, 1984), Vol. 7, pp. 560–561, 630–634. Concise Encyclopedia of Composite Materials, 1 ed. (Pergamon Press, Oxford, UK, 1989), p. 189. K. Brooks: Making finer powders. Metal Powder Report 57, 34 (2002). J.P. Wittenauer, T.G. Nieh, and J. Wadsworth: Tungsten and its alloys. Adv. Mater. Proc. 142(3), 28 (1992). J.L. Sepulveda and L.A. Valenzuela: Brush Wellman advances Cu/W technology. Metal Powder Report 53, 24 (1998). L.P. Dorfman, D.L. Houck, M.J. Scheithauer, J.N. Dann, and H.O. Fassett: Solid-phase synthesis of cupric tungsten. J. Mater. Res. 16, 1096 (2001). L.P. Dorfman, D.L. Houck, M.J. Scheithauer, and T.A. Frisk: Synthesis and hydrogen reduction of tungsten-copper composite oxides. J. Mater. Res. 17, 821 (2002). L.P. Dorfman, D.L. Houck, and M.J. Scheithauer: Consolidation of tungsten-coated copper composite powder. J. Mater. Res. 17, 2075 (2002). T.H. Kim, J.H. Yu, and J.S. Lee: The mechanism of hydrogen reduction synthesis of nanocomposite W-Cu powders. Nanostruct. Mater. 9, 213 (1997). T. Raguim, R. Sudaresan, P. Ramakrishnan, and T.R. Rama Mohan: Synthesis of nanocrystalline copper-tungsten alloys by mechanical alloying. Mater. Sci. Eng. A 304–306, 438 (2001). C.S. Xiong, Y.H. Xiong, H. Zhu, and T.F. Sun: Synthesis and structural studies of the Cu–W alloys prepared by mechanical alloying. Nanostruct. Mater. 5, 425 (1995). T.H. Kim and J.S. Lee: Fabrication of nanocomposite W-Cu powders, Editions Phys. (France), 1777 (1994). D-G. Kim, B-H. Lee, S-T. Oh, Y.D. Kim, and S-G. Kang: Mechanochemical process for W–15 wt% Cu nanocomposite powders with WO3–CuO powder mixture and its sintering characteristics. Mater. Sci. Eng. A 395, 333 (2005). D-G. Kim, G-S. Kim, S-T. Oh, and Y.D. Kim: The initial stage of sintering for the W-Cu nanocomposite powder prepared from W–CuO mixture. Mater. Lett. 58, 578 (2004). U. Tilliander, R.E. Aune, and S. Seetharaman: Hydrogen reduction of a WO3–Cu2O mixture, Royal Institute of Technology, Stockholm, Sweden, Metall. Mater. Trans. B (2005, in press). U. Tilliander, R.E. Aune, and S. Seetharaman: Kinetics studies of hydrogen reduction of Cu2O. Z. Metallkd. 97, 72 (2006). I. Arvanitidis, D. Sichen, and S. Seetharaman: A study of the thermal decomposition of BaCO3. Metall. Mater. Trans. 27B, 409 (1996).

J. Mater. Res., Vol. 21, No. 6, Jun 2006

1475