Obtaining of W/Cu nanocomposite powders by high energy ball milling process

The morphology of the particles is important in the process of obtaining alloys based on W/Cu, thus this investigation is focused on the influence of the copper content on the properties of W/Cu nanocomposites powders obtained after 20 hours of high energy ball milling. The experimental results regarding the obtaining of W100-x/Cux nanocomposites (x between 20 and 45 wt. %) are presented. Composition of the mixtures influenced the particle size distribution namely, the higher is Cu content the larger dimensions of the particles will be attained. After 20 hours of high energy ball milling the crystallites size was about 30 nm for copper respectively 12 nm for tungsten and Cu atoms entered in the W structure. Introduction One of the specific classes of materials which are suitable for elaboration by Powder Metallurgy (PM) consists in pseudo-alloys based on W-Cu due to their mutual insolubility. These materials are very important due to their wide field of applicability such as: welding electrodes, nozzle liners for rockets and missiles, heat sink materials, high power electrical contacts, fission reactors and so on, applications that require high mechanical properties conferred, in this case by tungsten, combined with high electrical and thermal conductivity which are conferred by copper [1-6]. The properties of these materials are in correlation with their composition and morphology and because of that is very important to choose the right composition function of the application [7]. In the field of high power electrical contacts, W-Cu materials must have high arc erosion resistance high temperature strength and high tribological properties to ensure an as long as possible lifetime [8, 9]. Particle size of the component elements plays an important role in the final properties of WCu alloys [10]. One of the techniques used for fabrication of the W-Cu materials is the infiltration one, which consists in the formation of a porous skeleton by tungsten which will be filled with molten copper. Vacuum pulse carburisation was reported [11] to be an infiltration method that leads to the formation of W-30wt.%Cu material with core-shell structure which presents high electrical conductivity (46.55%IACS) compared to the national standard (GB/T8320-2003 – 42%IACS) and a friction coefficient μ=0.64. To improve the sinter ability of W-Cu materials it can be introduced some activators such as Ni, Fe or Co which can be grain growth inhibitors [12]. Using of this activators can lead to a decreasing of electrical and thermal conductivity of W-Cu Powder Metallurgy and Advanced Materials – RoPM&AM 2017 Materials Research Forum LLC Materials Research Proceedings 8 (2018) 173-181 doi: http://dx.doi.org/10.21741/9781945291999-20 174 materials [13, 14]. A technique used for mass production which is suitable to produce W-Cu complex shape parts is Metal Injection Moulding (MIM) [15]. Another method to produce W-Cu materials is Mechanical Alloying (MA) which ensures obtaining of the nanocomposite powders and structural homogeneity which leads to an improvement of the sintering process by reducing the sintering activation energy [16-19]. Compared with micron powders, by using submicron powders (400nm) in the case of infiltrated W-25wt%Cu alloys can be obtained better properties such as: microstructural homogeneity, relative density (98.9%), hardness (230 HB) [10]. By MA, in the case of W-Cu system, Cu phase can dissolute in the W phase [20]. In most of cases the MA process is carried out at room temperature [21, 22]. In the present work nanocomposite powders W100-x/Cux (x=20-45 wt. %) were prepared by mechanical milling (MM) process having as starting powders W nanometric and Cu micrometric powders. The influence of the MM times and composition of the mixture on the morphology, phase transformations and particle size distribution were investigated. Experimental work As raw materials were used tungsten nanopowders prepared by MM (35 hours of milling) and copper micrometric powders (type SE Pometon). The morphologies of the initial powders at the same magnification (1500x) are presented in (Fig. 1). Fig. 1. SEM images of initial powders: a) W nanopowders; b) Cu powders. For MM process of the six mixtures W100-x/Cux (x=20, 25, 30, 35, 40, 40 and 45 wt. %) a Pulverissete 4 Vario planetary ball mill made by Fritsch was used. The parameters used for MM were: bowls volume 250 ml; material of the bowls stainless steel; balls diameter 10 mm; balls material stainless steel; milling type dry; milling medium – argon (type 5.0, purity 99.999%); material/ball weight ratio 1/2; speed: 400 rpm for the main disk and -800 rpm for the planets; milling time 20 hours; from 5 to 5 hours were taken samples to be analysed. In order to establish the cycles for MM, which means to not have higher temperature and pressure inside the grinding bowls, a GTM system made by Fritsch was used. In (Fig. 2) is presented the evolution of temperature and pressure for a MM of 10 minute with a break of 2 minutes. Powder Metallurgy and Advanced Materials – RoPM&AM 2017 Materials Research Forum LLC Materials Research Proceedings 8 (2018) 173-181 doi: http://dx.doi.org/10.21741/9781945291999-20 175 Fig. 2. Evolution of temperature and pressure inside the grinding bowl As it can be seen in (Fig. 2), the temperature didn`t exceed 50 C and the pressure is almost constant due to the dry milling type. These parameters must be controlled because if the temperature is higher, it can be damage the bowl and the equipment too. For morphological aspects of the powder mixtures was used a JEOL microscope JSM-5600 LV. Evolution of particle size distributions and polydispersity were studied by dynamic laser scattering (DLS) using a 90Plus particle size analyser, Brookhaven Instruments Corporation, USA, equipped with 35 mW solid state laser, having 660 nm wavelength. The temperature was 25 C and the scattering angle was 90. The dilutions of the powders mixtures were made in water and the solutions of each sample were subjected to ultrasonic treatment for 5 minutes to avoid flocculation of the particles. The investigation by X-ray diffraction has been performed using an Inel diffractometer, model Equinox 3000 working in reflection and using Co radiation. The 2theta investigated interval was 20-110 degree. Results and discussion In (Fig. 3) are presented morphologies of the as initial homogenous mixture and for the samples milled for 20 hours, respectively. From (Fig. 3 a, c, e, g, i, k) it is obvious that the initial homogenised mixtures have tungsten particles with lower particle size (nanoscale) compared with those of copper. Also, in (Fig. 3 k) it is observed higher covering degree of tungsten nanopowders on the copper particles, which is in accordance with composition of the mixture (80W/Cu). After 20 hours of MM (Fig. 3 b, d, f, h, j, l) all the samples elemental individual morphology is changed to a homogenous state. Most probably a mixing and welding of different particles being realized. Also the powders tend to agglomerate. Fig. 4 shows the XRD pattern of the mixtures in different stages. In the figure are presented the X-ray diffraction patterns of the W-Cu mixtures not milled samples and samples milled for 5, 10, 15 and 20 h. Alongside of these patterns, in the same figure are presented the evolution of the mean crystallite size of the W and Cu upon increasing the milling time. In the X-ray diffraction patterns of the not milled samples on can observe the peaks characteristic for the W bcc structure Powder Metallurgy and Advanced Materials – RoPM&AM 2017 Materials Research Forum LLC Materials Research Proceedings 8 (2018) 173-181 doi: http://dx.doi.org/10.21741/9781945291999-20 176 from Im-3m space group and Cu fcc structure from Fm-3m space group according to JCPDS files 04-0806 and 04-0836 respectively. Fig. 3. SEM images of the: a) 55W/Cu homogenised mixture; b) 55W/Cu after 20 hours of MM; c) 60W/Cu homogenised mixture; d) 60W/Cu after 20 hours of MM; e) 65W/Cu homogenised mixture; f) 65W/Cu after 20 hours of MM; g) 70W/Cu homogenised mixture; h) 70W/Cu after 20 hours of MM; i) 75W/Cu homogenised mixture; j) 75W/Cu after 20 hours of MM; k) 80W/Cu homogenised mixture; l) 80W/Cu after 20 hours of MM; The ratio of the W and Cu peaks intensities differ upon varying the amount of W and Cu in the mixtures, as expected. One can also observe is the broadening of the tungsten diffraction peaks due to the nanocrystalline state of the tungsten powder used in the mixture. The mean crystallite size of the tungsten powder (computed with the Scherrer method) is at 12 ± 2 nm. After 5 h of milling it can be noticed that the diffraction peaks of the copper are also broadened indicating the copper crystallite decreases. Further increase of the milling time lead also to the enlargement of the diffraction peaks of the copper and up to the final milling time no other peaks are identified in the diffraction patterns for all the ratios between W and Cu. It can be observed that independent on the amount of Cu in the material its peaks are observed in the diffraction pattern. The fcc Cu-based structure is present in the material after 20 h of milling. Known that in equilibrium condition there is no solubility between W and Cu and known also that by mechanically milling solid solution between immiscible elements can be obtained, it can be assumed that after 20 of milling some Cu atoms entered in the W structure. At the end of milling time independent on W to Cu ratio according to X-ray investigation the material is a nanocomposite one consisting in a W-Cu solid solution and Cu nanocrystallites. It can be remarked that the tungsten crystallites does not have significant variation upon milling together with copper. It remains at about 12 ± 2 nm. The mean Powder Metallurgy and Advanced Materials – RoPM&AM 2017 Materials Research Forum LLC Materials Research Proceedings 8 (2018) 173-181 doi: http://dx.doi.org/10.21