Compressibility and solid-state sintering behavior of W-Cu composite powders

1 Introduction

Tungsten (W)-copper (Cu) composites are widely used in various kinds of electrical contacts. According to the contact application, the weight ratio of W to Cu differs in different composites. For example, the composites with 75% wt Cu are used as current-carrying contacts, while the ones with 15% wt Cu are mainly used as vacuum switches, arcing tips, and oil circuit breakers. The conventional methods for fabrication of W-Cu composites are infiltration of Cu-to-W skeleton, press/sinter/repress (PSR), or press/sinter of W-Cu powder mixtures [1]. However, it has been shown that, in the latter process, W-Cu composite powders show better sinterability rather than W-Cu powder mixtures [2]. Several methods such as mechanical milling [3], mechanochemical [4], and coprecipitation [5–7] have been applied to synthesize W-Cu composite powders. In the case of mechanical milling process, different kinds of powder mixtures such as W-Cu [8], W-CuO [9, 10], and WO₃/CuO [11, 12] can be used as starting materials. Due to the brittleness of the oxide powders, a finer dispersion of the constituents can be achieved at a relatively short milling time. Many researchers have investigated the effect of sintering temperature on the density of the sintered composite powders. However, the compaction behavior and the influence of cold pressing magnitude on the sintered density of W-Cu composite powders have not been investigated properly.

In this research, the correlation between cold compaction pressure and green density of W-15% and W-75% wt Cu composite powders was investigated. Also, the effect of uniaxial compaction pressure on the density of the solid-state sintered samples at 1000°C was studied.

2 Materials and methods

Extra pure WO₃ (Merck, Whitehouse Station, NJ, USA) and CuO (Merck) were used as starting powders. Two different WO₃ and CuO powder mixtures were prepared in order to synthesize W-15% and W-75% wt Cu composite powders. The powders were milled in a planetary ball mill at the milling speed of 250 rpm for 10 h in air atmosphere. The milling vessel and balls were made of tungsten carbide. The milling media to powder weight ratio was 15:1. The milled powders were reduced in a tube-type electrical furnace at 900°C for 1.5 h in hydrogen atmosphere. The milled and reduced powders were characterized by X-ray diffraction (XRD) and scanning electron microscopy (SEM). The reduced powders were uniaxially compacted by 100, 200, and 300 MPa in a 13 mm die. The green compacts were sintered at 1000°C for 1 h in hydrogen atmosphere. The green and sintered densities were determined from mass, diameter, and thickness of the cold pressed and sintered compacts, respectively. The theoretical density of the composites was determined using the following formulae [13]:

where ρ_Cu is the theoretical density of Cu, ρ_W is the theoretical density of W, and W_Cu and W_W are mass fractions of Cu and W, respectively.

3 Results and discussion

Figure 1A and B shows the SEM micrograph of the milled powders. The milled particles do not have a definite shape and their dimension is mainly below 1 μm. Also, the agglomeration of the powder particles is obvious in this figure.

Figure 1

SEM micrograph of the milled powders with relatively high CuO (A) and WO₃ (B) contents.

The lattice strain and crystalline size of WO₃ in both groups of samples were determined by Debye-Scherrer’s formula [14, 15]:

where d is the crystalline size, λ is the wavelength of CuKα radiation (0.15404 nm), θ is the Bragg angle, K is Scherrer’s constant (0.9), η is the lattice strain, and β_s is the full-width at half-maximum (FWHM) of a diffraction peak, which can be determined by the following relation:

where β_e and β_i are the measured FWHM and instrumental broadening for each diffraction peak, respectively.

The measured lattice strain and crystalline size of WO₃ for both groups of samples were 0.63% and 33.2 nm, respectively. This result is due to similar mechanical milling process on both groups of powder mixtures. The lattice strain and crystalline size of CuO could be measured just for the samples with relatively high volume percent of this constituent. In this group of samples, due to the high intensity of CuO XRD peak, the FWHM could be measured with high accuracy. The measured microlattice strain and crystalline size of CuO were 0.737% and 17.8 nm, respectively. However, according to the obtained results for WO₃, it can be concluded that the measured quantities of lattice strain and crystalline size for CuO would be similar for both groups of samples.

XRD patterns of the reduced powders are shown in Figure 2. As it is expected, the intensity of the corresponding Cu and W peaks is not similar in the illustrated patterns, which is due to different weight percent of the constituents in the reduced powders.

Figure 2

(A) XRD pattern of W-15% wt Cu composite powders. (B) XRD pattern of W-75% wt Cu powders.

In order to evaluate the cold compaction behavior of the composite powders, Heckel and Panelli-Ambrosio equations were used [16]:

where D is the relative density, P is the compaction pressure, and K, L, B, and C are constants.

Figure 3A and B shows the curve of In(11-D) versus P and P^1/2 for both groups of samples. As it is shown, the amount of correlation factor (R²) shows higher accuracy of Heckel equation rather than Panelli-Ambrosio equation for W-75% wt Cu composite powders. On the contrary, the R² value, which is related to Panelli-Ambrosio, is closer to unity compared with Heckel equation for the samples with 15% wt Cu. Based on the obtained results, it can be declared that the Heckel equation is more suitable than the other one for evaluating the compaction behavior of ductile powders. However, it can be said that the Panelli-Ambrosio equation has higher accuracy than the Heckel equation for predicting the cold compaction behavior of brittle powders.

Figure 3

(A) In11−D versus P (cold compaction pressure). (B) In11−D versus (P) for W-15% and W-75% wt Cu composite powders. R² value represent the correlation coefficients.

The effect of cold compaction pressure magnitude on the relative density of the green and solid-state sintered specimens is shown in Figure 4. As it can be seen, by increasing the compaction pressure, the relative densities of the green compacts were increased. However, increasing the compaction pressure led to higher consolidation of W-75% wt Cu rather than W-15% wt Cu composite powders, which was due to higher volume fraction of Cu within the microstructure of this group of powders. According to this figure, by increasing the cold compaction pressure, the difference between green and sintered densities was increased for W-75% wt Cu samples. On the contrary, the samples with 15% wt Cu show a different trend, and by increasing the cold compaction pressure, the difference between green and sintered relative densities was decreased.

Figure 4

Effect of cold compaction pressure on the relative density of the green and solid-state sintered samples.

In order to increase the density and elimination of pores within the microstructure, the atoms diffuse from the bulk to the necking area between the powder particles [18] (Figure 5A). However, it can be declared that diffusion of Cu atoms was the dominant phenomenon, which was due to consolidation of the compacts. The reason of this matter is the relatively high melting point of W, which led to low diffusion coefficient of this agent at sintering temperature (i.e., 1000°C). The driving force for diffusion of Cu atoms is the stress that acts upon the neck and also the difference of vacancy concentration between the bulk of the powders and the neck-like zone at the sintering temperature. The mentioned stress is calculated by the following formula [17, 18]:

Figure 5

Effect of cold compaction magnitude on the curvature radius of the neck-like zone.

where γ_sv is the interface energy between the solid and gas phase and S is the radius of neck curvature (Figure 5A).

At the initial stage of sintering process, x>>S, the stress acts as a tensile force (σ=-γsvs), which is the driving force for atomic diffusion.

Figure 5A and B shows the effect of cold compaction magnitude on the x parameter. As it can be seen, by increasing the cold compaction pressure, σ, and as a consequence, the driving force of atomic diffusion is decreased. On the contrary, by increasing the temperature of green compacts during sintering, the equilibrium concentration of vacancies is increased significantly. However, the increase of equilibrium concentration of vacancies in the neck-like zone is higher than the bulk of the particles [18]. Due to the lower melting temperature of Cu, the formation of vacancies in the microstructure of Cu is too higher than W. Therefore, it can be concluded that, by increasing the volume percent of Cu within the microstructure of the composite powders, the latter mechanism, which is based on the difference of vacancies’ concentration at the different regions of the powders, has a more effective role on the densification of powder compacts.

Based on the above discussion, it can be concluded that, in the case of W-75% wt Cu composite powders, the dominant mechanism of densification was the vacancy concentration gradient. Furthermore, it seems that, for the powders with W-15% wt Cu, due to the relatively low weight percent of Cu, the stress gradient has a higher effect on Cu atom diffusion. However, in both groups of samples, the relative density of the compacts was not increased significantly during solid-state sintering at 1000°C and the density of the samples may be increased by subsequent compaction. Also, the relative sintered density of W-15% wt Cu sintered samples may be increased by liquid-phase sintering of the cold compacted powders above the melting point of Cu.

4 Conclusion

In this study, the compaction and solid-state sintering behavior of W-15% and W-75% wt Cu composite powders were investigated and the following results were obtained:

1. For both groups of milled powders with different WO₃ to CuO weight ratios, the crystalline size and lattice strain of WO₃ were identical.

2. Heckel and Panelli-Ambrosio equations are the preferred equations for evaluating the compaction behavior of W-15% and W-75% wt Cu composite powders, respectively.

3. Increasing the cold compaction pressure of W-75% wt Cu composite powders was due to higher densification during solid-state sintering.