- Email: [email protected]
Mechanical milling of Fe–Li and Cu–Li systems and their nitrogen absorption properties
Materials Science and Engineering A 449–451 (2007) 1067–1070
Mechanical milling of Fe–Li and Cu–Li systems and their nitrogen absorption properties K.N. Ishihara ∗ , K. Irie, F. Kubo, E. Yamasue, H. Okumura Graduate School of Energy Science, Kyoto University, Yoshida Honmachi, Sakyo, Kyoto 606-8501, Japan Received 22 August 2005; received in revised form 8 December 2005; accepted 15 February 2006
Abstract It is known that lithium reacts with nitrogen to produce nitride at room temperature. Lithium alloys are one of the candidates for nitrogen absorption materials. On the other hand, it is also known that the mechanical milling is generally used to create various non-equilibrium alloys, which are usually in active powder forms. In this work, in order to activate the nitrogen absorption properties of lithium, the mechanical milling of lithium systems is performed. As the alloying elements, iron and copper are chosen, since both systems of Fe–Li and Cu–Li have no intermediate phase in equilibrium. For mechanically alloyed powders, the atomic structure and the reactivity with nitrogen are investigated using X-ray diffractometry, thermal analysis, etc. In the case of Fe–Li, the mechanically milled sample shows high reactivity with nitrogen for a short milling time, compared with lithium in bulk form. While, Cu–Li forms supersaturated solid solution which hardly reacts with nitrogen at room temperature. © 2006 Elsevier B.V. All rights reserved. Keywords: Cu–Li; Fe–Li; Immiscible system; Mechanical alloying; Nitriding
1. Introduction Mechanical alloying (MA) is one of the most suitable methods to produce alloys with various elements such as refractory, low melting and high vapor pressure materials [1,2]. There are many works to report the formation of various materials utilizing the MA method. For the case of lithium base alloys, the lithium ion battery is very popular and MA has been investigated for several systems including C–Li [3], Li–Mg [4], Cu–Li [5] alloys. It is known that lithium reacts with nitrogen to form nitride even at room temperature. If lithium becomes more reactive than the bulk form, it may be useful for nitrogen absorption materials. We reported that the activity of an element in metastable solid solution produced by mechanical alloying is enormously increased for the case of immiscible systems [6,7]. Moreover, the lithium element is the lightest metal. It is interesting to know whether a metastable solid solution including lithium is obtained, and whether a new light alloy can be obtained. Nevertheless, the systematic study on lithium-based alloys has not been performed [8]. Our recent work on the mechanical alloy-
∗
Corresponding author. Tel.: +81 75 753 5464; fax: +81 75 753 5464. E-mail address: [email protected] (K.N. Ishihara).
0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.02.273
ing of Li systems [9] shows that for Fe–Li and Si–Li systems, the reactivity with nitrogen at room temperature becomes higher than that of lithium in bulk form. But, samples which formed intermetallic compounds (Al–Li and Si–Li) do not show the reactivity with nitrogen. The purpose of this work is to show how lithium forms alloys during MA with copper and iron, since Fe–Li and Cu–Li do not form intermediate phase in equilibrium. Especially, the characteristics for Cu–Li system, which is less immiscible than Fe–Li system, will be discussed.
2. Experimental procedure Small pieces of Li metal and powders of Cu (97%, 60 mesh) and Fe (99.9%, 53 m) were used as starting materials. The high energy vibrating ball mill was performed under an argon atmosphere at room temperature. In total, about 10 g or 5 g of samples was used. No process control agent (PCA) was used. The ball to power ratio was kept to 10. For the nitrogen gas absorption, the sample was put into the test tube under an argon atmosphere. Then, the test tube is evacuated and is filled with nitrogen gas. The amount of nitrogen absorption is measured by change of pressure and sample weight.
1068
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
Fig. 2. SEM image of Fe–15.3% Li sample milled for 1 h. Layered structures are seen.
Fig. 1. The change of pressure by the formation of lithium nitride for pure lithium and mechanical alloyed Fe–15% Li samples.
The microstructure of the specimen is observed using scanning electron microscopy (SEM, JSM-5800) with energy dispersive X-ray analysis (EDAX). Differential scanning calorimetry (DSC, Perkin-Elmer lambda 700) and differential thermal analysis with thermal gravity (DTA/TG, Rigaku) were used for thermal analysis. 3. Results and discussions For the Fe–Li system with the Li concentration higher than 20%, the samples could not be obtained in powder form after milling. Then, in this work, 15.3% Li sample will be reported. The results of the nitrogen absorption experiment are shown in Fig. 1. In the case of bulk lithium, it takes about 40 min to start the reaction with nitrogen gas. Whereas, the sample milled for 10 min exhibits a much faster reaction. The longer-milled samples, however, do not show the clear reaction. The rates based on the mid-height slope for these reaction curves are summarized in Table 1. Fig. 2 shows the SEM image of the sample milled for 1 h. It is difficult to detect lithium atoms under SEM/EDX measurements since the atomic number is small. We then focused on the oxygen atoms, which cover the lithium phase as lithium oxide, since oxidation occurred during handling the sample in the air for the SEM/EDX. The formation of a laminar structure consisting of
oxygen rich areas and iron rich areas was confirmed by the EDX analysis. The so-called ‘kneading effect’ is clearly seen in this sample as it is often observed for a general milling process. In the Cu–Li system, the solid solutions were obtained up to 30 at.% Li for 20 h milling. The lattice parameters determined by X-ray diffraction are shown in Fig. 3 with published data. It indicates that the complete solid solution is formed for the 30 at.% Li sample. For the contents higher than 40% Li, the powder form samples were not obtained up to 10 h milling. The samples including solid solution hardly react with nitrogen at room temperature. The DSC thermogram (Fig. 4) in nitrogen atmosphere shows a distinct exothermic peak at around 680 K as well as broad peak at around 370 K indicating nitrification of lithium. Other two broad exothermic peaks may be attributable to recovery. The lattice constants measured by X-ray diffraction patterns without heat treatment and after heat treatment at 473 K, 573 K, 673 K and 773 K are shown in Fig. 5. It is suggested that the solid solution reacts with nitrogen to form lithium nitride and pure copper.
Table 1 Rates based on mid-height slope for pure lithium in bulk form and mechanical alloyed of Fe–15.3% Li Sample
Rate × 103 (s−1 )
Li 10 min milling 60 min milling 600 min milling
0.06 1.5 1.1 0.005
Fig. 3. Lattice parameters of milled samples 16.7% Li for 10 h and 30% Li for 20 h (open squares) are plotted with published data (closed squares).
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
1069
Fig. 4. DSC thermogram of Cu–30% Li.
Fig. 5. Lattice constants of Cu–30% Li samples heat treated at 473 K, 673 K and 773 K.
Fig. 7. Schematically drawn activities of Li in fcc and bcc solid solutions. Dashed curves indicate regions of metastable state.
Fig. 6 shows the calculated activity of Li in hypothetical Fe–Li solid solution using Miedema’s model [10,11], where the structural enthalpy is neglected. It is suggested that if the solid solution is formed at the investigated composition, the order of activity is about 1012 . As proceeding of the mechanical alloying process, the lithium becomes very active. Thus, for the short-time-milled samples, the fast reaction with nitrogen
gas may have occurred. Further milling, however, leads to oxidation of lithium by supplying oxygen from the initial contents of iron powder (1.46 wt.% O). In order to avoid the oxygen contamination, the other iron powder (99.9%, 150 m, 0.04 wt.% O) was used. In this case, the proceeding of mechanical alloying is slow and the powder from sample is difficult to be obtained for the sample higher than 14.3% Li up to 10 h milling. For the 9.1% Li sample, powder form sample was partially obtained and it does not react with nitrogen. Balls taken from the vessel, however, react with nitrogen (rate: 0.07 × 10−3 s−1 ). It is suggested that almost all lithium powder adheres to balls and iron powder remains. In contrast to Fe–Li, the immiscibility of Cu–Li system is not so high. The estimated activity of Li in Cu–Li solid solution is shown in Fig. 7. In this figure, solid lines indicate the equilibrium state. For the fcc solid solution, the activity coefficient [12] and lattice stability for pure fcc Li [13] are used for estimation. It is noted that the shape of the phase diagram in Ref. [14] is controversial and it may include bcc solid solution as intermediate phase [15]. The activities in Fig. 7 are based on this speculation. The activity of Li in solid solution is always less than unity, that of pure lithium, in our samples. Then, the reactivity with nitrogen is also sluggish compared to the Fe–Li system, where the reaction with nitrogen is attributable to the
Fig. 6. Activity of Li and Fe, and excess enthalpy of mixing for hypothetical solid solution of Fe–Li system, calculated by Miedema’s model.
1070
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
fine pure lithium. In the case of Cu–Li system, it is reported that the solid solution (18% Li) is produced by electrodeposition [16] and the sample reacts with nitrogen, in air and in nitrogen, at around 680 K [17] to form Li3 N and solid solution (3% Li). The sample produced by mechanical alloying for 30% Li shows similar DSC thermogram in nitrogen atmosphere around 680 K. However, the milled sample reacts with nitrogen even at room temperature. It is suggested that the intensive defects and strains introduced to samples during mechanical milling decrease the reaction temperature. 4. Conclusion The mechanical alloying was performed for Fe–Li and Cu–Li systems. It is concluded as follows: (i) As for the Fe–Li system, which is immiscible even at liquid state, the evidence of alloying is not observed. The results can be explained in terms of the heat of mixing and the mutual diffusivity in these systems. (ii) In the case of Cu–Li system, which has limited solid solution up to 20% Li in published phase diagram, the solid solution is formed up to 30% Li, which is supersaturated. The reaction with nitrogen takes place even at room temperature, but the rate is sluggish compared with pure lithium in bulk form. Acknowledgement This research was partially supported by Ministry of Education, Culture, Sports, Science and Technology, Grant-in-Aid for Scientific Research (B), 17360369, 2005.
References [1] F. Zhou, Y.T. Chou, E.J. Lavernia, Mater. Trans. 42 (2001) 1566– 1570. [2] K. Uenishi, K.F. Kobayashi, S. Nasu, H. Hatano, K.N. Ishihara, P.H. Shingu, Z. Metallkd. 83 (1992) 132–135. [3] D. Guerard, R. Janot, J. Phys. Chem. Sol. 65 (2004) 147–152. [4] J. Huot, S. Bouaricha, S. Boily, J.-P. Dodelet, D. Guay, R. Schulz, J. Alloys Compd. 266 (1998) 307–310. [5] C. Camurri, M. Ortiz, C. Carrasco, Mater. Character. 51 (2003) 171– 176. [6] K.N. Ishihara, S. Diaz de la Torre, T. Hosoe, A. Yamamoto, P.H. Shingu, in: H. Hennen, T. Oki (Eds.), First International Conference on Processing Matereials for Properties, Hawaii, 1993, pp. 667–670. [7] S.D. De la Torre, A. Yamamoto, K.N. Ishihara, P.H. Shingu, T. Hirato, Y. Awakura, Mater. Sci. Forum 237 (1995) 179–181. [8] C. Guminski, H.U. Borgstedt, Arch. Metall. Mater. 49 (2004) 529– 534. [9] K.N. Ishihara, K. Irie, F. Kubo, K. Shichi, E. Yamasue, H. Okumura, ISMANAM 2005 (2005). [10] F.R. de Boer, R. Boom, W.C.M. Mattens, A.R. Miedema, A.K. Niessen, Cohesion in Metals Transition Metal Alloys, North-Holland, 1988. [11] H. Bakker, Enthalpy in Alloys, Miedema’s Semi-Empirical Model, Trans Tech Publications, 1998. [12] A.D. Pelton, Bull. Alloy Phase Diagr. 7 (1986) 142–144. [13] A.D. Pelton, Bull. Alloy Phase Diagr. 7 (1986) 223–228. [14] T.B. Massalski (Ed.), Binary Alloy Phase Diagrams, second ed., ASM, Materials Park, 1990. [15] A. Van Walle, Z. Moser, W. Gasior, Arch. Metall. Mater. 49 (2004) 535– 544. [16] O.A. Lambri, A. Penaloza, A.V. Moron Alcain, M. Ortiz, F.C. Lucca, Mater. Sci. Eng. A212 (1996) 108–118. [17] O.A. Lambri, J.I. Perez-Landazabal, J.A. Cano, V. Recarte, J. Campo, A. Penaloza, M. Ortiz, C.H. Worner, Powder Technol. 152 (2005) 24–30.
Mechanical milling of Fe–Li and Cu–Li systems and their nitrogen absorption properties K.N. Ishihara ∗ , K. Irie, F. Kubo, E. Yamasue, H. Okumura Graduate School of Energy Science, Kyoto University, Yoshida Honmachi, Sakyo, Kyoto 606-8501, Japan Received 22 August 2005; received in revised form 8 December 2005; accepted 15 February 2006
Abstract It is known that lithium reacts with nitrogen to produce nitride at room temperature. Lithium alloys are one of the candidates for nitrogen absorption materials. On the other hand, it is also known that the mechanical milling is generally used to create various non-equilibrium alloys, which are usually in active powder forms. In this work, in order to activate the nitrogen absorption properties of lithium, the mechanical milling of lithium systems is performed. As the alloying elements, iron and copper are chosen, since both systems of Fe–Li and Cu–Li have no intermediate phase in equilibrium. For mechanically alloyed powders, the atomic structure and the reactivity with nitrogen are investigated using X-ray diffractometry, thermal analysis, etc. In the case of Fe–Li, the mechanically milled sample shows high reactivity with nitrogen for a short milling time, compared with lithium in bulk form. While, Cu–Li forms supersaturated solid solution which hardly reacts with nitrogen at room temperature. © 2006 Elsevier B.V. All rights reserved. Keywords: Cu–Li; Fe–Li; Immiscible system; Mechanical alloying; Nitriding
1. Introduction Mechanical alloying (MA) is one of the most suitable methods to produce alloys with various elements such as refractory, low melting and high vapor pressure materials [1,2]. There are many works to report the formation of various materials utilizing the MA method. For the case of lithium base alloys, the lithium ion battery is very popular and MA has been investigated for several systems including C–Li [3], Li–Mg [4], Cu–Li [5] alloys. It is known that lithium reacts with nitrogen to form nitride even at room temperature. If lithium becomes more reactive than the bulk form, it may be useful for nitrogen absorption materials. We reported that the activity of an element in metastable solid solution produced by mechanical alloying is enormously increased for the case of immiscible systems [6,7]. Moreover, the lithium element is the lightest metal. It is interesting to know whether a metastable solid solution including lithium is obtained, and whether a new light alloy can be obtained. Nevertheless, the systematic study on lithium-based alloys has not been performed [8]. Our recent work on the mechanical alloy-
∗
Corresponding author. Tel.: +81 75 753 5464; fax: +81 75 753 5464. E-mail address: [email protected] (K.N. Ishihara).
0921-5093/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.msea.2006.02.273
ing of Li systems [9] shows that for Fe–Li and Si–Li systems, the reactivity with nitrogen at room temperature becomes higher than that of lithium in bulk form. But, samples which formed intermetallic compounds (Al–Li and Si–Li) do not show the reactivity with nitrogen. The purpose of this work is to show how lithium forms alloys during MA with copper and iron, since Fe–Li and Cu–Li do not form intermediate phase in equilibrium. Especially, the characteristics for Cu–Li system, which is less immiscible than Fe–Li system, will be discussed.
2. Experimental procedure Small pieces of Li metal and powders of Cu (97%, 60 mesh) and Fe (99.9%, 53 m) were used as starting materials. The high energy vibrating ball mill was performed under an argon atmosphere at room temperature. In total, about 10 g or 5 g of samples was used. No process control agent (PCA) was used. The ball to power ratio was kept to 10. For the nitrogen gas absorption, the sample was put into the test tube under an argon atmosphere. Then, the test tube is evacuated and is filled with nitrogen gas. The amount of nitrogen absorption is measured by change of pressure and sample weight.
1068
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
Fig. 2. SEM image of Fe–15.3% Li sample milled for 1 h. Layered structures are seen.
Fig. 1. The change of pressure by the formation of lithium nitride for pure lithium and mechanical alloyed Fe–15% Li samples.
The microstructure of the specimen is observed using scanning electron microscopy (SEM, JSM-5800) with energy dispersive X-ray analysis (EDAX). Differential scanning calorimetry (DSC, Perkin-Elmer lambda 700) and differential thermal analysis with thermal gravity (DTA/TG, Rigaku) were used for thermal analysis. 3. Results and discussions For the Fe–Li system with the Li concentration higher than 20%, the samples could not be obtained in powder form after milling. Then, in this work, 15.3% Li sample will be reported. The results of the nitrogen absorption experiment are shown in Fig. 1. In the case of bulk lithium, it takes about 40 min to start the reaction with nitrogen gas. Whereas, the sample milled for 10 min exhibits a much faster reaction. The longer-milled samples, however, do not show the clear reaction. The rates based on the mid-height slope for these reaction curves are summarized in Table 1. Fig. 2 shows the SEM image of the sample milled for 1 h. It is difficult to detect lithium atoms under SEM/EDX measurements since the atomic number is small. We then focused on the oxygen atoms, which cover the lithium phase as lithium oxide, since oxidation occurred during handling the sample in the air for the SEM/EDX. The formation of a laminar structure consisting of
oxygen rich areas and iron rich areas was confirmed by the EDX analysis. The so-called ‘kneading effect’ is clearly seen in this sample as it is often observed for a general milling process. In the Cu–Li system, the solid solutions were obtained up to 30 at.% Li for 20 h milling. The lattice parameters determined by X-ray diffraction are shown in Fig. 3 with published data. It indicates that the complete solid solution is formed for the 30 at.% Li sample. For the contents higher than 40% Li, the powder form samples were not obtained up to 10 h milling. The samples including solid solution hardly react with nitrogen at room temperature. The DSC thermogram (Fig. 4) in nitrogen atmosphere shows a distinct exothermic peak at around 680 K as well as broad peak at around 370 K indicating nitrification of lithium. Other two broad exothermic peaks may be attributable to recovery. The lattice constants measured by X-ray diffraction patterns without heat treatment and after heat treatment at 473 K, 573 K, 673 K and 773 K are shown in Fig. 5. It is suggested that the solid solution reacts with nitrogen to form lithium nitride and pure copper.
Table 1 Rates based on mid-height slope for pure lithium in bulk form and mechanical alloyed of Fe–15.3% Li Sample
Rate × 103 (s−1 )
Li 10 min milling 60 min milling 600 min milling
0.06 1.5 1.1 0.005
Fig. 3. Lattice parameters of milled samples 16.7% Li for 10 h and 30% Li for 20 h (open squares) are plotted with published data (closed squares).
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
1069
Fig. 4. DSC thermogram of Cu–30% Li.
Fig. 5. Lattice constants of Cu–30% Li samples heat treated at 473 K, 673 K and 773 K.
Fig. 7. Schematically drawn activities of Li in fcc and bcc solid solutions. Dashed curves indicate regions of metastable state.
Fig. 6 shows the calculated activity of Li in hypothetical Fe–Li solid solution using Miedema’s model [10,11], where the structural enthalpy is neglected. It is suggested that if the solid solution is formed at the investigated composition, the order of activity is about 1012 . As proceeding of the mechanical alloying process, the lithium becomes very active. Thus, for the short-time-milled samples, the fast reaction with nitrogen
gas may have occurred. Further milling, however, leads to oxidation of lithium by supplying oxygen from the initial contents of iron powder (1.46 wt.% O). In order to avoid the oxygen contamination, the other iron powder (99.9%, 150 m, 0.04 wt.% O) was used. In this case, the proceeding of mechanical alloying is slow and the powder from sample is difficult to be obtained for the sample higher than 14.3% Li up to 10 h milling. For the 9.1% Li sample, powder form sample was partially obtained and it does not react with nitrogen. Balls taken from the vessel, however, react with nitrogen (rate: 0.07 × 10−3 s−1 ). It is suggested that almost all lithium powder adheres to balls and iron powder remains. In contrast to Fe–Li, the immiscibility of Cu–Li system is not so high. The estimated activity of Li in Cu–Li solid solution is shown in Fig. 7. In this figure, solid lines indicate the equilibrium state. For the fcc solid solution, the activity coefficient [12] and lattice stability for pure fcc Li [13] are used for estimation. It is noted that the shape of the phase diagram in Ref. [14] is controversial and it may include bcc solid solution as intermediate phase [15]. The activities in Fig. 7 are based on this speculation. The activity of Li in solid solution is always less than unity, that of pure lithium, in our samples. Then, the reactivity with nitrogen is also sluggish compared to the Fe–Li system, where the reaction with nitrogen is attributable to the
Fig. 6. Activity of Li and Fe, and excess enthalpy of mixing for hypothetical solid solution of Fe–Li system, calculated by Miedema’s model.
1070
K.N. Ishihara et al. / Materials Science and Engineering A 449–451 (2007) 1067–1070
fine pure lithium. In the case of Cu–Li system, it is reported that the solid solution (18% Li) is produced by electrodeposition [16] and the sample reacts with nitrogen, in air and in nitrogen, at around 680 K [17] to form Li3 N and solid solution (3% Li). The sample produced by mechanical alloying for 30% Li shows similar DSC thermogram in nitrogen atmosphere around 680 K. However, the milled sample reacts with nitrogen even at room temperature. It is suggested that the intensive defects and strains introduced to samples during mechanical milling decrease the reaction temperature. 4. Conclusion The mechanical alloying was performed for Fe–Li and Cu–Li systems. It is concluded as follows: (i) As for the Fe–Li system, which is immiscible even at liquid state, the evidence of alloying is not observed. The results can be explained in terms of the heat of mixing and the mutual diffusivity in these systems. (ii) In the case of Cu–Li system, which has limited solid solution up to 20% Li in published phase diagram, the solid solution is formed up to 30% Li, which is supersaturated. The reaction with nitrogen takes place even at room temperature, but the rate is sluggish compared with pure lithium in bulk form. Acknowledgement This research was partially supported by Ministry of Education, Culture, Sports, Science and Technology, Grant-in-Aid for Scientific Research (B), 17360369, 2005.
References [1] F. Zhou, Y.T. Chou, E.J. Lavernia, Mater. Trans. 42 (2001) 1566– 1570. [2] K. Uenishi, K.F. Kobayashi, S. Nasu, H. Hatano, K.N. Ishihara, P.H. Shingu, Z. Metallkd. 83 (1992) 132–135. [3] D. Guerard, R. Janot, J. Phys. Chem. Sol. 65 (2004) 147–152. [4] J. Huot, S. Bouaricha, S. Boily, J.-P. Dodelet, D. Guay, R. Schulz, J. Alloys Compd. 266 (1998) 307–310. [5] C. Camurri, M. Ortiz, C. Carrasco, Mater. Character. 51 (2003) 171– 176. [6] K.N. Ishihara, S. Diaz de la Torre, T. Hosoe, A. Yamamoto, P.H. Shingu, in: H. Hennen, T. Oki (Eds.), First International Conference on Processing Matereials for Properties, Hawaii, 1993, pp. 667–670. [7] S.D. De la Torre, A. Yamamoto, K.N. Ishihara, P.H. Shingu, T. Hirato, Y. Awakura, Mater. Sci. Forum 237 (1995) 179–181. [8] C. Guminski, H.U. Borgstedt, Arch. Metall. Mater. 49 (2004) 529– 534. [9] K.N. Ishihara, K. Irie, F. Kubo, K. Shichi, E. Yamasue, H. Okumura, ISMANAM 2005 (2005). [10] F.R. de Boer, R. Boom, W.C.M. Mattens, A.R. Miedema, A.K. Niessen, Cohesion in Metals Transition Metal Alloys, North-Holland, 1988. [11] H. Bakker, Enthalpy in Alloys, Miedema’s Semi-Empirical Model, Trans Tech Publications, 1998. [12] A.D. Pelton, Bull. Alloy Phase Diagr. 7 (1986) 142–144. [13] A.D. Pelton, Bull. Alloy Phase Diagr. 7 (1986) 223–228. [14] T.B. Massalski (Ed.), Binary Alloy Phase Diagrams, second ed., ASM, Materials Park, 1990. [15] A. Van Walle, Z. Moser, W. Gasior, Arch. Metall. Mater. 49 (2004) 535– 544. [16] O.A. Lambri, A. Penaloza, A.V. Moron Alcain, M. Ortiz, F.C. Lucca, Mater. Sci. Eng. A212 (1996) 108–118. [17] O.A. Lambri, J.I. Perez-Landazabal, J.A. Cano, V. Recarte, J. Campo, A. Penaloza, M. Ortiz, C.H. Worner, Powder Technol. 152 (2005) 24–30.