ANNEALING BEHAVIOUR OF CuTi ALLOYS AT DIFFERENT STAGES OF MECHANICAL ALLOYING

HAL Id: jpa-00230781

https://hal.archives-ouvertes.fr/jpa-00230781

Submitted on 1 Jan 1990

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

ANNEALING BEHAVIOUR OF CuTi ALLOYS AT DIFFERENT STAGES OF MECHANICAL ALLOYING

G. Cocco, L. Schiffini, I. Soletta, M. Baricco, N. Cowlam

To cite this version:

G. Cocco, L. Schiffini, I. Soletta, M. Baricco, N. Cowlam. ANNEALING BEHAVIOUR OF CuTi AL-

LOYS AT DIFFERENT STAGES OF MECHANICAL ALLOYING. Journal de Physique Colloques,

1990, 51 (C4), pp.C4-175-C4-180. �10.1051/jphyscol:1990421�. �jpa-00230781�

COLLOQUE DE PHYSIQUE

Colloque C4, supplbment au n014, Tome 51, 15 juillet 1990

ANNEALING BEHAVIOUR OF CuTi ALLOYS AT DIFFERENT STAGES OF MECHANICAL ALLOYING

G. COCCO, L. SCHIFFINI, I. SOLETTA, M. BARICCO* and N. COWLAM"'

Fipartimento di Chimica dell'UniversitA, Via Vienna, 2 Sassari, Italy INGF,I-10125 Torino, Italy

.

Resume

Les alliages de poudre de Cu et de Ti purs ont et6 realises mecaniquement et traites de facon thermique dans des conditions de moulinage croissantes. Nous avons Btudie la dependance des,produits intermBdiaires en fonction de la composition, du temps et de La temperature du moulinage. Nous avons obtenu ainsi de renseignements sur le mecanisme des reactions. A basse concentration de cuivre, la cinBtique de l'amorphisation est liee B l'apparition competitive de CuTiz et/ou CuTi compos6s qui, a leur tour, dependent de la temperature de moulinage.

Abstract

Mixtures of elemental Cu and Ti powders were mechanically alloyed and thermally treated at increased levels of the ball-milling process. We have investigated the dependence of intermediate products of annealings as a function of composition, milling time and milling temperature. Useful hints are gained relevant to the interdiffusion mechanisms operative in the alloying reactions. At low Cu concentration the kinetic of amorphization is constrained by the competitive occurrence of CuTiz and/or CuTi compounds which in turn depend on milling temperature.

1

-

In a precedent paper /l/, here after referred to as I, the Mechanical Alloying (MA) process of the CuTi system was presented. A central issue in that study was the impossibility of amorphizing the system completely at compositions lower than 50 Cu at%, unless the milling container was cooled to 250 K. At room temperature, after a very early stage where the amorphous phase actually forms by an interdiffusion mechanism, the reaction path switches to the formation of CuTiz and CuTi crystalline compounds, which coexist with the amorphous fraction. It is apparent that a possible explanation for the different alloying paths must lie in different kinetic mechanisms which become operative by changing the reaction temperature.

Unfortunately, the experimental check of kinetic phenomena is a quite difficult task owing to the particular morphology of the processed powders which hinder the local probe of their structural conditions at the interphase. Nevertheless, as it will be shown in the present work, some useful hints on the various diffusive mechanisms which sustain the alloying reactions in the CuTi system, can be gained from the intermediate products which develop during annealing treatments after different stages of MA.

Accordingly, the present paper is specifically addressed to detect and to identify the particular phases which form first and that, in their turn, depend on the nature of the initial kinetic events at the continuously renewed clean interface.

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1990421

COLLOQUE DE PHYSIQUE

2- EXPERIMETAL METHODS

The reader is referred to paper I for details on alloy synthesis, apparatus employed, X-ray characterization and thermal analysis. We remind briefly that several compositions were synthesized by MA of pure crystalline Cu and Ti elemental powders quoted as Cu,,, Cu5,, Cu6,, and Cu7, which evidence the Cu at% content. Moreover, the CU,, composition was milled by keeping the vial at room temperature and at 250 K respectively. Samples milled at progressive times were sealed in glass vials, under an argon atmosphere, and thermally treated 1 h in isothermal conditions at different temperatures.

3

-

A general behaviour was found common to all the examined compositions:

short milling times (1-2 hours) produce sufficient intermixing to complete the alloying reactions when the specimens are isothermally treated l h at 823 K. In fact, Cu and Ti diffraction lines, when present in the as-milled powders, are almost completely lost in the fully crystallized samples.

Therefore composition modulated structures develop from the beginning, pointing out that each powder particle has been already involved in several collisions in agreement with the model of Davis et al. /2/. Moreover, the equilibrium crystalline intermetallics depend on the experimental milling condition besides the mixture composition. Let us now discuss these items in detail.

As far as the Cu7,Ti3, sample, milled as long as 16 h and treated l h at 823 K, the following intermetallics can be indexed in the X-ray pattern:

orthorhombic Cu,Ti, Pnma, JCPDS 20-370 /3/, traces of tetragonal Cu4Ti3, I4/mmm, JCPDS 18-460 /3/ and the cubic form of CuTiz, CCoW3 type /4/, which is probably stabilized by oxygen. The absence of equilibrium tetragonal CuFi,, JCPDS 18-459 /3/ is noticeable. The crystalline phases which develop at intermediate milling times and/or different annealing temperatures can be reconducted to the compounds quoted above even if some weak lines remain not indexed.

A similar behaviour was observed for the Cu6,Ti,, sample after 16 h of milling and 1 h at 823 K. An overlapping of crystalline peaks characterizes the X-ray pattern and a spread of intermetallics, almost encompassing the entire amorphization range, as Cu4Ti, traces of Cu4Ti3, tetragonal CuTi P4/mmm, delta-phase JCPDS 7-113 /3/, tetragonal CuTi, Si02-type, 14/mmm, JCPDS 15-717 /3/ and the cubic CuTi, can be indexed. No heat treatments were performed at intermediate milling stages.

The structural transformations induced by various isothermal annealings of Cu5,Ti5, milled 2, 4, 8 and 16 h are depicted in Figure 1. The heat treatment appears of little relevance up to 493 K, then (at 573 K), different structural conditions are detectable depending on the milling time. Samples milled 2, 4 and 8 h still show Cu and Ti lines, and in the angular range between their main signals two new peaks arise relevant to the tetragonal gamma-CuTi compound, P4/nmm1 JCPDS 7-114 /3/ which contains an excess of Ti respect to the equiatomic composition /5/. Moreover an in- creasing fraction of amorphous alloy is also present which becomes the predominant phase in the sample milled 16 h. It can be also noted that the maximum quantity of gamma-CuTi is present after 4 h and, conversely, gamma- CuTi is absent in the pattern of 16 h milled sample. The presented behaviour is significative of various reactions. In fact we infer that the observed gamma-CuTi does not form, at this temperature, from the amorphous fraction, i.e. it is not a crystallization product, rather it develops by an interdiffusion reaction in intermixed particles. The population of these

"multilayered-like particles" must achieve a maximum since the refinement process evolves to the full amorphous condition: this maximum is observed

between 4 and 8 h. This interpretation is substantiated by the behaviour of specimens treated at 823 K.

823 K

573

K 473

as

2 THETA

Figure 1. CuKa of X-ray patterns of Cu,,Ti,, samples ball milled for (from the left) 2, 4, 8 and 16 hrs and treated isothermally 1 hr at the quoted temperatures. Symbols: , gamma- CuTi; 0 , delta-CuTi; m , CuTi,; C1 , Cubic CuTiZ;

*

At this annealing temperature the patterns of the samples milled up to

8 h are similar and essentially characterized by gamma-CuTi. The 823 K pattern relevant to powders processed 16 h is different. The CuTi crystalline compound is now identified as the tetragonal Cu enriched delta- phase /5/. Moreover Cu,Ti, and traces of CuTi, are also present. These results can be explained by considering the Cu atoms as the faster moving species so that at the beginning, titanium rich modulated structures are formed. In this initial stage the alloying reaction is still incomplete and its accomplishment occurs during the annealing treatments because of thermal assisted diffusion: the final crystallization products are therefore reminiscent of the initial Ti rich condition at the interface of the composite powders. After 16 h of milling, the refined powders look X-ray

(24-1 78 COLLOQUE DE PHYSIQUE

amorphous which means that the alloying reaction is almost completed during the milling treatment. The thermodynamic driving force for the alloying is no more present and the transformation path during heating is now relevant to only the long range order establishment. So the observed crystalline phases reflect compositional fluctuations existent in different regions of the particles /6-7/.

DSc analysis strongly supports this view. As an example we compare in Figure 2 the heat flow from the samples milled 2 and 16 h. For the former (left curve) it is clearly observed that the transformation path is a two- stage process: the first exothermal signal is attributed to the formation of the amorphous phase in incompletely reacted parts of the sample /7/, the second peak is due to crystallization. Only the crystallization step is discernible in the sample milled 16 h. It is worth while to remark the close similarity of the reaction path occurring in the sample ball milled 2 h with the solid-state reactions observed in multilayered thin-film of elemental metals /8/ as far as thermal patterns are concerned.

CuSo milled 4h CuSo milled 16h

*

*

100 200 300 400 500 100 200 300 400 500

TEMPERATURE ('C) TEMPERATURE ('C)

Figure 2. DSc traces of CuS0Ti,, mixture. Left: partially unreacted; right: milled 16 h and X-ray amorphous.

Figure 3 shows the annealing behaviours of Cu4,,T& milled 2 h (left) 4 h (centre) and 16 h (right). Two parallel paths are presented relevant to milling conditions carried out at room temperature (A) and at 250 K (B) respectively as anticipated in paper I.

As it can be observed in Figure 3, left side, after 2 h of milling, the as-milled powders (bottom curves) show different X-ray patterns pointing out that the alloiing reaction is more advanced in the specimen processed at room temperature. This difference is enhanced by the subsequent annealings:

at 623 K a mixture of CuTi, and gamma-CuTi compounds is already observed in the powders processed at room temperature whereas, at 250 K, elemental Cu and Ti are still present with unresolved peaks. At 823 K (upper curves) the crystallization process leading essentially to CuTiz and CuTi is accomplished for both samples, but in the former (A)

CUT^^

Coming back to the as-milled powders it was inferred that the diffusive processes are damped down at 250 K (Figure 3, bottom curves) and this reduction primarily hampers the formation of CuTi2 compound. Under these conditions the amorphization can progress. If this last reaction occurs with a process of nucleation and growth, the formation rate of the amorphous

phase becomes at least equal to the formation rate of the competing crystalline compound that now is essentially the gamma-CuTi phase.

2 THETA

Figure 3. CuKa X-ray diffraction patterns of powders milled 2 h (left), 4 h (centre) and 16 h (right) and isothermally annealed l h at the quoted temperatures. A comparison is presented relevant to milling processes carried out at room temperature (A) and at 253 K

(B). Symbols : , gamma CuTi; , CuTi,; v , Ti; A , Cu.

I I

0 20 4 0 00 00 100

ATOMIC PERCENT COPPER

Figure 4. Schematic free energy diagram of metastable CuTi at 300 K. Full line represents the free energy of the amorphous CuTi; broken curves are relevant to the terminal solid solutions. The free energy curves of CuTi, and CuTi intermetallic compounds are also shown as dotted line. Point 1 and 2 in the common tangents quote the thermodynamic conditions during milling processes at different temperatures.

We do not know if the crystalline CuTi nuclei actually form as it can probably be inferred from the observed annealing behaviour; in any case they must be in undercritical conditions and must dissolve within the observed

C4-180 COLLOQUE DE PHYSIQUE

growing amorphous phase. This situation is better described by aid of Fig- ure 5 where the free energy curve of metastable CuTi is depicted together with the free energies of the crystalline intermetallic compounds /4/ of interest here. A detailed discussion on the evaluation of these curves has been given in /lO/. When both CuTi, and CuTi intermetallics are present in the early stage of milling, as it occurs in the milling process carried out at room temperature, the amorphous phase is greatly disfavoured (point 1 in the figure) and the amorphization can occur through a process similar to that observed during grinding of an intermetallic mixture /g/. This needs long milling times in order to introduce enough chemical and structural disorder to rise the free energies of intermetallics up to that of amorphous phase. As presented in paper I, the amorphous phase actually forms in sequence to the intermetallics after 32 h of milling.

When the formation of CuTi, is hindered or kinetically limited, as it is sustained by the present results at low milling temperature, a different condition occurs. This can be evaluated from point 2 in the Figure. It ap- pears that the amorphous phase formation become thermodynamically more competitive with the crystalline one when a diffusion reaction starts at the clean interface.

ACKNOWLEDGEMENT

This work was finacially supported by ENEA under contract 3965 and by CNR under contract 89.00623.68.

REFERENCES

COCCO G., SOLETTA I., ENZO S., MAGINI M., COWLAM N., J-Phys. Paris, this issue.

DAVIS R.M., McDERMOTT B., KOCH C.C., Met. Trans., (1988) 2867.

Joint Committee on Powder Diffraction Standards, Powder Diffraction Files (Swarthmore, Pennsylvania: 1nt.Center for Dif. Data).

MURRAY L., Bull. Alloy Phase Dia., 4 (1983) 81.

KARLSSON N., J.Inst. of Metals, 79 (1951) 391.

BATTEZZATI L., ENZO S., SCHIFFINI L., COCCO G., J.Less-Common Met., (1988) 301.

BATTEZZATI L., COCCO G., SCHIFFINI L., ENZO S., Mat. Sci. Eng., 97 (1988) 121.

COTTS J., MENG W.J., JOHNSON W.L., Phys.Rev.Letters, 57 (1986) 2295 LEE R.Y., JANG J., KOCH C.C., J.Less-Common Met., 140 (1988) 73.

BATTEZZATI L., BARICCO M., RIONTINO G., SOLETTA I., J.Phys. Paris, this issue.