Mechanically induced gas–solid reaction for the synthesis of nanocrystalline ZrN powders and their subsequent consolidations

Journal of Alloys and Compounds 313 (2000) 224–234

L

www.elsevier.com / locate / jallcom

Mechanically induced gas–solid reaction for the synthesis of nanocrystalline ZrN powders and their subsequent consolidations a, b M. Sherif El-Eskandarany *, A.H. Ashour a

Mining and Petroleum Engineering Department, Faculty of Engineering, Al-Azhar University, Nasr City, 11371 Cairo, Egypt b National Center for Radiation Research and Technology, Nasr City, Cairo, Egypt Received 11 July 2000; accepted 15 August 2000

Abstract Equiatomic nanocrystalline ZrN powders with an average grain size of less than 8 nm in diameter have been fabricated by high energy ball-milling elemental powders of Zr under nitrogen gas-flow at room temperature. The ductile powders of Zr tend to agglomerate during the first stage of the reactive ball milling (RBM; ,11 ks) to form powder particles with larger diameters. The powders are then intensively disintegrated into smaller particles during the second stage of milling (11–43 ks). These disintegrated particles that have fresh or new surfaces begin to react with the milling atmosphere (nitrogen) during this stage of milling to form a cubic phase of ZrN powder coexisting with unreacted hcp-Zr powder. Towards the end of milling (86–173 ks), a single phase of nanocrystalline ZrN (NaCl-structure) is obtained. The powders of this end-product have spherical like morphology with average particle size of about 0.4 mm in diameter. Cold and hot pressing techniques were employed to consolidate the powders at the several stages of the RBM. The as-milled and as-consolidated powders were characterized as a function of the RBM time by means of X-ray diffraction, transmission electron microscopy, scanning electron microscopy, optical metallography and chemical analyses. The results have shown that the consolidated ZrN compact of the end product (173 ks of RBM) still maintains its unique nanocrystalline characteristics with an average grain size of less than 80 nm. Density measurements of these consolidated samples of the end-products (86–173 ks of the RBM) show that they are essentially fully dense (above 99% of the theoretical density for ZrN). The dependences of the hardness on the grain size and the consolidation temperatures were investigated.  2000 Elsevier Science B.V. All rights reserved. Keywords: Nitride materials; Gas–solid reactions; Mechanical alloying; Nanofabrications

1. Introduction Amongst the nonequilibrium materials (amorphous, quasicrystals and solid solutions), nanocrystalline materials [1] that are defined as materials with grain sizes less than 100 nm have received as much attention as advanced engineering materials with improved physical and mechanical properties [2,3]. A wide variety of techniques can be used successfully for preparing such unique materials. Mechanical alloying (MA) [4], which is a well known process for preparing several amorphous [5–10], metal nitrides [11,12], metal carbides [13,14] alloys, and nanocomposite materials [15], has been considered a powerful technique for synthesizing numerous nanocrystal*Corresponding author. Present address: Exploratory Research for Advanced Technology, Japan Science and Technology, Inoue Superliquid Glass Project, The Research Institute for Electric and Magnetic Materials, 2-1-1 Yagiyama-Minami, Taihaku-ku, Sendai, Miyagi 982-080, Japan. Fax: 181-22-243-0262. E-mail address: [email protected] (M.S. El-Eskandarany).

line materials. Recently, the MA process has become a popular method to fabricate nanocrystalline materials due to its simplicity and relatively inexpensive equipment [16]. Nitrides possess unique properties which are highly desirable for a variety of applications. They are technologically important materials because of their hardness, stability at high temperatures and electrical and optical properties [17]. Zirconium nitride (ZrN) alloys are used for cutting tools, tool coatings, solar-control films and microelectronics applications. ZrN is chemically stable with respect to most etching solutions, has a low receptivity and provides an excellent diffusion barrier against metals [17]. The cubic form of ZrN can be obtained by several methods such as activated reactive evaporation [18], the self propagating combustion method [19] under high nitrogen pressure (10 5 atm) and high temperature (1500 K) and the plasma spray method [20] under nitrogen gas-flow. Moreover, a chemical vapor deposition technique has been used also for ZrN preparation by reacting the zirconium tetra chloride (ZrCl 4 ) with ammonia (NH 3 ) at tempera-

0925-8388 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0925-8388( 00 )01175-0

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

tures above 1300 K. In addition, the low-temperature physical vapor deposition, atmospheric pressure chemical vapor deposition [18] and reactive sputtering [21] techniques are used widely for the preparation. The high costs of preparation and the excess of contamination in the end-product are the disadvantages of these methods. In the present study, elemental Zr powders are directly reacted with nitrogen gas at room temperature to form a single phase of NaCl type-ZrN powder containing about 50 at.% N 2 , using the reactive ball-milling technique (RBM) [22]. The reacted powders were then consolidated to form full-density bulk ZrN, using cold and hot pressing sintering techniques. This consolidation method has been successfully used to obtain nanocrystalline fully dense ceramics materials such as ZrC [23], WC and nanocomposite WC– Co [24] materials. One of the significant potential technological attractions of the present work is that it proposes a relatively inexpensive and simple process for the industrial application of the RBM method to produce large amounts of an advanced material of great interest to engineers and material scientists.

2. Experimental details Pure elemental Zr powder (99.5 at.%) and purified nitrogen gas (99 wt.%) were used as the starting reacting materials. The powders were charged into a stainless steel (SUS 316) vial (250 ml in volume) and sealed together with 25 stainless steel (SUS 316) balls (10 mm in diameter) in a glove box under a purified argon (99.99 wt.%) atmosphere. The ball-to-powder weight ratio was 10:1. The RBM process took place by mounting the vial in a high-energy ball-mill (Fritsch P6) equipped with a rotary pump and a gas-flow system. The vial was evacuated for about 4 ks and then a flow (1.0 ml s 21 ) of nitrogen gas was passed through the inlet of the vial via a plastic pipe. The outlet of the vial was connected with a oil bubbler. Once the gas bubbles were observed, the RBM experiment started with a milling rate of 4.2 s 21 at room temperature. The RBM was stopped periodically after selected milling times and the powders were completely discharged from the vial in the glove box. The powders were then consolidated into compact samples using two steps. In the first step, the powders were cold-compacted at room temperature under a purified argon atmosphere, using a tungsten carbide die of 13.0 mm in diameter at a pressure of 1.0 GPa for 70.2 ks. The green-compacts that were obtained were then transferred to another tungsten carbide die of 15 mm and then hot-pressed in vacuum under a pressure of 1.5 GPa at 1873 K for 43 ks without the addition of any binding materials. In order to study the effect of the consolidation temperature on the grain size of the consolidated powders, different samples of the end-product (173

225

ks of the RBM time) were individually sintered at various selected temperatures (1073, 1273, 1473 and 1673 K). The as-milled and as-consolidated samples were characterized by means of X-ray diffraction (XRD) with CuKa radiation, transmission electron microscopy (TEM) (using a 200 kV microscope), scanning electron microscopy (SEM) (using a 20 kV microscope) and optical microscopy. In order to determine the concentrations of Zr and the level of Fe contamination for the final product, the samples were analyzed by the induction coupled plasma emission method. On the other hand, the concentrations of nitrogen and oxygen contamination were determined by the helium carrier fusion-thermal conductivity method. The hardnesses of the compacted samples were determined using a Vickers indenter with a load of 50 kg. In addition, densities of the consolidated samples were determined by Archimedes’ principle using water immersion.

3. Results

3.1. Formation of ZrN powders by reactive ball-milling XRD analysis was employed to follow the progress of the gas–solid reaction during milling elemental Zr powders under flow of nitrogen gas (Fig. 1). At the starting stage of the RBM time (0 ks), the XRD pattern displayed sharp Bragg-peaks reflections corresponding to polycrystalline hcp-Zr powders, as shown in Fig. 1a. After 11 ks of the RBM time, a new phase that corresponded to NaCl-type ZrN (closed symbols) appeared (Fig. 1b). The Bragg-peaks corresponding to the ZrN became sharp and pronounced after 22 ks of RBM time (Fig. 1c), indicating an increasing ZrN fraction in the milled powder. Increasing the RBM time to 43 ks lead to further increases of the ZrN, and the reflections that come from the unreacted elemental Zr powders were hardly seen (Fig. 1d). After 86 ks of RBM time, all the Bragg-peaks corresponding to the metallic Zr disappeared and a single phase of NaCl-type ZrN appeared Fig. 1e), indicating the completion of the RBM process. The lattice parameter, a 0 , of this phase was calculated from the (111) and (200) reflections and was found to be 0.457748 nm, which is in good agreement with the reported value for pure ZrN (0.457756 nm) [25]. TEM analyses were used as powerful tool to observe the local structural changes during the RBM process. The bright field image (BFI) and the corresponding selected area diffraction pattern (SADP) for the powders milled for 4 ks of RBM time are shown in Fig. 2. Obviously, the powders contain a high dense network of imperfections (Fig. 2a). The corresponding SADP (Fig. 2b) showed a sharp ring-spot pattern that is characteristic of polycrystalline hcp-Zr. Increasing the milling time lead to increasing the shear and impact forces generated by the milling media and this caused mechanical deformation on the powder particles

226

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

gas–solid reaction and the formation a single phase of ZrN. There can be no more fundamental characteristics of a powder than the size and the shape of the individual particles. The SEM technique was employed here to follow the morphological changes of the Zr powders that were milled under nitrogen gas. Fig. 6 displays the SEM micrographs of the powders after selected RBM times. The powders at the early stage of RBM (4 ks) were agglomerated due to the cold welding and formed larger particles with an average diameter of nearly 500 mm, as presented in Fig. 6a. These powders were disintegrated into smaller powders (nearly 100 mm in diameter) upon further milling (22 ks) and the product of this stage of milling looked very heterogeneous in shape and size, as show in Fig. 6b. Towards the end of the RBM processing time (86 ks) the powders were dramatically disintegrated to form fine particles of nearly spherical-like morphology with a wide size distribution, ranging from 0.7 to 5 mm, as shown in Fig. 6c. The morphological properties of the powders were improved by increasing the RBM time to 173 ks (Fig. 6d) so that the powders become uniform equiaxed spheres with smooth surfaces. In addition, they revealed ultrafine characteristics being about 1 mm (or less) in diameter.

3.2. Consolidation of the mechanically reacted ZrN powder

Fig. 1. XRD patterns of Zr powders milled under a flow of nitrogen after (a) 0 ks, (b) 11 ks, (c) 22 ks, (d) 43 ks and (e) 86 ks of RBM time.

which translates into numerous faults and dislocations which appeared clearly in the powders taken after 7.2 ks of RBM time, as shown in Fig. 3a. It is worth noting that there is no evidence of the formation of the ZrN phase and the powders contain a rather fine structure of hcp-Zr, as suggested by the SADP in Fig. 3b. Fig. 4a shows the BFI and Fig. 4a and d show the corresponding SADPs for the powders that were milled for 18 ks of RBM time. The sample was classified into two regions, I and II (Fig. 4a). It can be reported that the structure of the milled powders at this stage of milling was extremely heterogeneous, as shown by the different features of the SADPs taken at regions I (hcp-Zr, Fig. 4b) and II (ZrN coexisted with unreacted elemental Zr, Fig. 4c). The dark field image (DFI) and the corresponding indexed SADP of the powders of the end-product (173 ks) are shown together in Fig. 5. The powders are composed of cell or lens-like structure with an average grain size of less than 8 nm in diameter (Fig. 5a), suggesting the formation of nanocrystalline materials. Moreover, the SADP indicates a clear NaCl-type ZrN with the absence of elemental Zr (Fig. 5b), indicating the completion of the

A sintering step is applied to consolidate the powders after selected RBM times. Fig. 7 shows a SEM micrograph of the cross-sectional view of as-consolidated powder where the RBM time was 22 ks (Fig. 7a) and for 173 ks (Fig. 7b). As was presented in the previous section (see Fig. 1c), after 22 ks of RBM time, the as-milled powders consisted of two phases: unreacted metallic Zr and fully reacted ZrN powders. Since the consolidation process took place at 1873 K [just below the melting point of pure Zr (2128 K) and far below the melting point of ZrN (3233 K)], the ZrN particles in the mixed powders (the agglomerated particles in the center of Fig. 7a) are embedded in the diffused matrix of metallic Zr to form a composite Zr–ZrN compact. Thus, the consolidated samples for this stage of RBM (11–43 ks) were either rich or poor in ZrN. During the Vickers hardness measurements, the sample hardness varied from a few GPa to nearly 16 GPa, as shown in Fig. 8. In addition to the sample hardnesses found from the early and intermediate stages of RBM, the samples that were milled for 86 and 173 ks and then consolidated into compacts contained a single phase of ZrN with fine equiaxed grains structure (less than 10 mm in diameter), as shown in Fig. 7b. It should be noted that the irregular grain edges in some cases (see Fig. 7b) arise from the long etching time used during the sample preparation. Moreover, the hardness value of these samples for the final-product has a narrow distribution range of

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

227

Fig. 2. (a) BFI and (b) SADP of Zr powders after 4 ks of RBM time.

about 18 GPa, suggesting the formation of a homogeneous single phase of ZrN (Fig. 8). The BFI and the corresponding SADP of the endproduct (173 ks) that was hot-pressed at 1873 K is shown in Fig. 9. Comparing this micrograph with the one in Fig.

5, we can say that the consolidation procedure for ZrN powders lead to moderate grain growth. Since the average grain size of this consolidated sample was less than 100 nm (about 80 nm in diameter) one can say that the sintered sample maintained its nanocrystalline character. We should

Fig. 3. (a) BFI and (b) SADP of Zr powders after 7.2 ks of RBM time.

228

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

Fig. 4. (a) BFI and (b) SADP of Zr powders after 18 ks of RBM time.

emphasize that this consolidation step did not lead to any structural changes, and the ZrN material maintains the NaCl-structure, as shown by the SADP in Fig. 9b.

Fig. 10 shows the grain size distribution which was determined from at least 100 representative grains taken from several BFI and DFI micrographs of the end-product

Fig. 5. (a) BFI and (b) SADP of Zr powders after 173 ks of RBM time.

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

229

Fig. 6. SEM micrographs of the Zr powders milled under a flow of nitrogen for (a) 4 ks, (b) 22 ks, (c) 86 ks and (d) 173 ks of RBM time.

(173 ks) that was hot-pressed at 1873 K. The average grain size was estimated to be 79.5 nm. It is worth noting that there are very few larger grains (much larger than 100 nm, i.e. submicrons) if compared with those grains that are less than 100 nm in diameter. The bulk density of the as-consolidated samples was plotted as a function of RBM time in Fig. 11. During the first and intermediate stages of RBM, the bulk density increased monotonically with increasing RBM time, suggesting the existence of a highly dense phase (ZrN). At the final stage of RBM (86–173 ks), the densities of the consolidated samples were in the range of 7.28–7.30 g / cm 3 . Comparing these values with the theoretical density of the ZrN compound (7.35 g / cm 3 ) [25], this indicates that the consolidated ZrN samples were fully dense. Fig. 12 presents the correlation between the consolidation temperature, the relative density (closed symbols) and

grain size of the consolidated samples (open symbols). The relative density of the powders that were consolidated at room temperature had a density of about 48% (green compact), with an average grain size (8 nm in diameter) which does not differ markedly from the milled samples (5–10 nm in diameter). Increasing the consolidation temperature to 1073 K, lead to a remarkable increase in the relative density to about 66%. Accordingly, the grains were grown in size to be nearly 20 nm in diameter. The bulk sample that was consolidated at 1273 K had a higher density of about 80%, with an average grain size of 30 nm. This relative density increased to more than 90% upon consolidation at 1473 and 1673 K so that the grain sizes increased to values ranging from 50 to 65 nm. Nearly fully-dense (above 99.5%) samples could be obtained at consolidation temperatures ranging between 1873 and 1973 K. Unfortunately, consolidation of the powders at

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

230

Fig. 6. (continued)

such high temperature lead to a strong grain growth so that the bulk compact sample consisted of larger grains of about 80–120 nm in diameter. The well-known empirical dependence of strength and hardness on porosity and grain size for ceramics materials is given by [26–28]: H 5 Kd

2a

e

2bp

where H is the hardness value, d is the grain size, P is the specimen porosity and K, a and b are empirical constants. The constant a can be approximated to be 0.5 [23,26,29]. Fig. 13 shows the correlation between the P of the consolidated samples that were hot-pressed at various temperatures and the combined effects of H and d. It can be notified that the hardness increased with increasing density (low porosity values) and grain size. A linear

relation was obtained with a slope, b55.029, which is similar to that found for a variety of ceramic materials (see for example Ref. [29]).

3.3. Chemical analysis Fig. 14 displays the nitrogen content in the as-milled (closed symbols) and as-consolidated (open symbols) powders as a function of the RBM time. This figure can be used to illustrate the progress of the gas–solid reaction between Zr and nitrogen that took place in the ball-mill. After few kiloseconds of RBM time (22 ks), the powders contained a considerable amount of nitrogen (21 at.%.). Increasing the RBM time lead to an increase in the nitrogen content in the milled powders to about 40 at.% after 43 ks. The nitrogen content of the end-product was about 49 at.%, being in a good agreement with the reported

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

231

value of NaCl-ZrN [25]. It is worth noting that the hotpressing step does not lead to any compositional changes of the samples, as indicated by the values for the nitrogen content for the as-milled and as-consolidated samples. The contamination content (iron and oxygen) in the as-milled powders was shown as a function of the RBM time in Fig. 15. Iron contamination was introduced to the milled powders due to using stainless steel milling tools. In order to decrease the iron contamination during RBM, the metallic Zr powders were milled before charging the reactant materials to form an iron-resistance wear on the surface of the milling tools. However, the iron contamination content increased monotonically with increasing RBM time. The concentration of iron in the end-product of the ZrN was less that 0.29 at.%. Comparing this value with the one for the as-received Zr powders (0.12 at.%) indicates that the iron contamination that was introduced into the powder during the RBM is 0.17 at.%. The as-received Zr powders already contained a small amount of another contamination, i.e. oxygen (0.08 at.%). As it is shown in Fig. 15, the concentration of the oxygen increases with increasing RBM time. This may be attributed to the existence of oxygen in the nitrogen and / or argon gas used in the present study, or oxygen may be introduced into the powders during handling outside the glove box. The oxygen contamination content does not seem to present a serious problem because it below the level of 0.30 at.% in the end-product (173 ks).

Fig. 7. SEM micrographs of the cross-sectional view of as-consolidated Zr powders which were milled under a flow of nitrogen for (a) 22 ks and (b) 173 ks of RBM time.

4. Discussion We have employed a high-energy ball-mill operated at room temperature under nitrogen gas-flow for preparing a single phase NaCl structure ZrN powder. The as-milled powders were consolidated into bulk samples using a hot-pressing method. The mechanism of the gas–solid reaction via the RBM technique and the consolidation procedure of these powders will be discussed. Results have shown that the RBM process can be classified mainly into four stages. The end-product of each stage varied widely in terms of structure and morphology.

4.1. The early stage of RBM

Fig. 8. Dependence of Vickers hardness of the as-consolidated mechanically reacted Zr powders on the RBM time. The measurements of the hardness were done using a load of 50 kg. The symbols that have low hardness values (less than 5 GPa) correspond to the Zr matrix in the composite Zr–ZrN compacts (see text and Fig. 7a).

In this stage of RBM (0–4 ks) – referred to as the first milling stage – the powder particles of the Zr agglomerate, as a result of cold-welding, form larger particles (Fig. 6a). The as-received Zr powders contain about 0.08 at.% oxygen (Fig. 15) and, hence, the powders are already coated with a thin layer of zirconium oxide. These layers are likely to prevent the Zr powders from reacting with the surrounding milling atmosphere (nitrogen) or even with the other gas contamination (e.g. hydrogen and oxygen gas). During this stage, the surface area of the particles is too

232

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

Fig. 9. (a) BFI and (b) SADP of the end-product of NaCl-structure ZrN powders after consolidation at 1873 K.

small to react simultaneously with the nitrogen gas which is continuously introduced to the vial. The hardness of the consolidated samples taken from this stage (Fig. 8) is almost the same as pure metallic Zr and the bulk density is like that of elemental Zr (Fig. 11). All of these observations, and the XRD patterns as well, signify the absence of ZrN.

4.2. The second stage of RBM The most important RBM method for producing metal nitrides is a gas–solid reaction propagated between the Zr

Fig. 11. Correlation between the bulk density of the consolidated ZrN powders and the RBM time.

Fig. 10. Grain size distribution of as-consolidated ZrN (173 ks of RBM) at 1873 K (1.5 GPa) as determined from several TEM micrographs.

Fig. 12. Correlation between the relative density and the average grain size of as-consolidated ZrN (173 ks of RBM time) at several temperatures of hot-pressing.

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

Fig. 13. Dependence of the Vickers hardness on porosity and grain size of bulk ZrN (173 ks of RBM time).

233

surface of powder particles and a nitrogen gas-flow. The previous stage of RBM is followed by a second stage (4–22 ks) in which the agglomerated particles of Zr are shattered and disintegrated into several particles apparently irregular in shape and size (Fig. 6b). The disintegration of the powder occurred as a result of a continuous shear and impact forces generated by ball–powder–ball collisions. These new surfaces of metallic Zr are active and able to react with the nitrogen gas which flows continuously into the vial to form a nitride surface layer over the unreacted Zr powders. At this stage of milling the powders differ widely in structure from particle to particle and within the particle itself, containing N-rich and N-poor regions (Fig. 4). Thus, the consolidated samples of this stage are a composite of Zr–ZrN, as shown in Fig. 7a. Accordingly, the hardnesses of these samples have two values: a low one that correlates to the unreacted metallic Zr and a high one which corresponds to ZrN (Fig. 8). Moreover, the bulk density of this sample increases due to the existence of ZrN (Fig. 11).

4.3. The third stage of RBM

Fig. 14. Dependence of the nitrogen content in the as-milled Zr powders on the RBM time.

This stage of milling (22–86 ks) can be defined as the intermediate stage of RBM, in which the unprocessed Zr particles continuously disintegrate and react with nitrogen in a typical gas–solid reaction. The further milling at this stage enhances the reaction of Zr surfaces with the nitrogen gas and increases the volume fraction of ZrN against the unreacted Zr core in the individual particles. This increasing mole fraction of NaCl-ZrN at this stage (Fig. 1d) causes an increasing bulk density of the consolidated sample (Fig. 11). Since, the consolidated sample of this stage contains a considerable amount of the unreacted metallic Zr (Fig. 1c), no marked change in the hardness value could be notified (Fig. 8), and the sample still contains ZrN rich and poor zones.

4.4. The fourth stage of RBM

Fig. 15. Relation between the contamination content in the milled powders and the RBM time.

During this final stage of milling (86–173 ks), all of the metallic Zr particles react with nitrogen to form a singlephase of ZrN (Fig. 1e). Increasing the milling times (173 ks) leads to increase the shear and impact forces which cause a dramatic in the grain size of the end-product and the formation of nanocrystalline ZrN materials (Fig. 5). At this stage of milling, the powders are uniform with respect to shape and size (Fig. 6d). The nitrogen content of the material saturates in this stage (Fig. 14), and the bulk density of the consolidated samples is indicative of fulldense ZrN (Fig. 11). In addition, the hardness of the as-consolidated samples is uniform throughout (Fig. 8). The formation of these ZrN nanocrystalline materials can be attributed due to the plastic deformation that is usually produced in the crystal lattices during the high-energy ball milling process and this occurs by slip and twinning in the

234

M.S. El-Eskandarany, A.H. Ashour / Journal of Alloys and Compounds 313 (2000) 224 – 234

lattice of the milled powders. Accordingly, this deformation leads to shear bands containing a high dense network of dislocations. Due to the successive accumulations of the dislocation density, the crystals are disintegrated into subgrains that are initially separated by low angle grain boundaries. The formation of these subgrains is attributed to the decrease of the atomic level strain. Further ballmilling time leads to further deformation and the formation of shear bands located in the unstrained parts of the powders. This leads to subgrain size reduction so that the orientation of the final grains becomes random in the crystallographic orientations of the numerous grains and, hence, the direction of slip varies from one grain to another. Reduction in grain size is a very important factor for consolidation because it increases the sinterability of the powders and improves the mechanical and physical properties of the sintered materials. In the present work, a hot-pressing technique has been used to consolidate the mechanically reacted nanocrystalline ZrN powders into a full-dense bulk compact. This consolidation step is necessary for most industrial applications. One can say that the grains of the milled ZrN powders were grown with a value about ten times this consolidation step. This grain growth is still acceptable since the sintered sample of ZrN is nearly fully-dense (Fig. 12) and still maintain its advanced nanocrystalline characteristics (about 80 nm in diameter, Fig. 9c) with an extremely high hardness value (Fig. 8).

5. Conclusions Whereas ZrN is usually prepared under high nitrogen pressures and high temperatures, the present study demonstrates a powerful technique for preparing a single-phase of NaCl-structured ZrN at room temperature. In the present work, elemental Zr powder is milled under flowing nitrogen at room temperature, using a high-energy ball-mill. Single phase ZrN (containing 49 at.% nitrogen) with nanocrystalline grains (about 8 nm in diameter) was obtained after 86–173 ks of reaction ball-milling time. A hot-pressing technique was applied to consolidate the mechanically reacted powders into fully-dense bulk sam-

ples. This consolidation did not lead to extensive grain growth, and the compacted samples consisted of nanoscale grains of about 80 nm in diameter. The densities and the hardness of the consolidated samples were measured as a function of the reaction ball-milling time.

References [1] H. Gleiter, Progr. Mater. Sci 33 (1990) 4. [2] H. Hahan, Nanostructured Mater. 9 (1997) 3. [3] M. Sherif El-Eskandarany, M. Omori, T.J. Konno, K. Sumiyama, T. Hirai, K. Suzuki, Metal. Trans. A 29 (1998) 1973. [4] J.S. Benjamin, Metal. Trans A 1 (1970) 2943. [5] C.C. Koch, O.B. Cavin, C.G. McKamey, J.O. Scarbourgh, Appl. Phys. Lett. 43 (1983) 1017. [6] R. Schwarz, C.C. Koch, Appl. Phys. Lett. 49 (1986) 146. [7] M. Sherif El-Eskandarany, J. Alloys Comp. 23 (1996) 467. [8] M. Sherif El-Eskandarany, K. Sumiyama, K. Suzuki, Acta Metall. 36 (1997) 1175. [9] M. Sherif El-Eskandarany, K. Aoki, K. Suzuki, Appl. Phys. Lett. 70 (1997) 1679. [10] M. Sherif El-Eskandarany, J. Alloys Comp. 284 (1999) 295. [11] M. Sherif El-Eskandarany, K. Sumiyama, K. Aoki, K. Suzuki, J. Mater. Res. 9 (1994) 2891. [12] M. Sherif El-Eskandarany, J. Alloys Comp. 203 (1994) 117. [13] M. Sherif El-Eskandarany, K. Sumiyama, K. Suzuki, J. Mater. Res. 10 (1995) 659. [14] M. Sherif El-Eskandarany, Metal. Trans. A 27 (1996) 2374. [15] M. Sherif El-Eskandarany, J. Alloys Comp. 279 (1998) 263. [16] C.C. Koch, Nanostructured Mater. 9 (1997) 13. [17] J.N. Musher, R.G. Gordon, J. Mater. Res. 11 (1996) 989. [18] K. Nakamura, K. Inagawa, K. Tsuruoka, S. Komiya, Thin Solid Films 40 (1977) 155. [19] S. Deevi, Z.A. Munir, J. Mater. Res. 5 (1990) 2177. [20] Y. Tsunekawa, H. Hirai, M. Okumiya, I. Niimi, J. Thermal Spraying Soc. Jpn. 26 (1989) 1150. [21] Y. Inoue, S. Tanimoto, K. Tsujimura, Y. Yamashita, Y. Ibara, Y. Yamashita, K. Yoneda, J. Electrochem. Soc. 141 (1994) 1056. [22] M. Sherif El-Eskandarany, J. Alloys Comp. 203 (1994) 117. [23] A.A. Mahday, M. Sherif El-Eskandarany, H.A. Ahmed, A.A. Amer, J. Alloys Comp. 299 (2000) 244. [24] M. Sherif El-Eskandarany, A.A. Mahday, H.A. Ahmed, A.A. Amer, J. Alloys Comp. (2000) in press. [25] JCPDS-ICDD Card No. 38-753. [26] E. Ryshkewitch, J. Am. Ceram. Soc. 36 (1953) 65. [27] W. Duckworth, J. Am. Ceram. Soc. 36 (1953) 68. [28] F.P. Knudsen, J. Am. Ceram. Soc. 42 (1959) 376. [29] R.W. Rice, Mater. Sci. Eng. 73 (1985) 215.