A Comparison of the Mixing Rates for Fe-Ti and Ni-Ti Bilayers
- PDF / 310,837 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 71 Downloads / 173 Views
A COMPARISON OF THE MIXING RATES FOR FE-TI AND NI-TI BILAYERS. and H. H. Johnsona
P. Borgesena, D. A. Lilienfeldb, R. E. Vistroma, a) Dept.of Materials Science and Engineering b) National Nanofabrication Facility Cornell University Ithaca, NY 14853. USA ABSTRACT
Previous comparisons of the room temperature mixing of Ni-Ti and Fe-Ti multilayers have shown significant differences in the mixing rates in spite of closely similar heats of mixing as well as ballistic properties. This result suggests contributions from mechanisms other than ballistic and thermal spike effects, or a break-down of present thermal spike models for these materials. In order to identify the mixing mechanisms involved, quantitative measurements were made at various temperatures between 80K and 350K using bilayer samples. Surprisingly, the mixing rates were found to disagree significantly with those estimated for multilayer samples. INTRODUCTION The extent of ion beam mixing of a two-component system at an interface is generally expressed as the variance, a, of the atomic displacement distribution. This quantity usually varies linearly with irradiation fluence, *, and we define the mixing rate as the ratio 6e /6+. Room temperature ion beam mixing of multilayer samples near the 50-50 composition has been seen to be much more efficient in the Ni-Ti than in the Fe-Ti system [1]. In both cases, the intermixed region is known to remain amorphous 12-4]. Since the nuclear masses, and thus the damage energies FD, in the two systems are very similar, the difference cannot be due to ballistic mixing effects [5,6]. Thermal spike mixing is generally expected to scale in a simple manner with deposited nuclear energy per unit path length, c, and heat of mixing, AH•,x [7,8]. These parameters are very similar in the two systems [9], which also does not explain the difference. Recently, the multilayer mixing rates were quantified [9], although with large error bars, and the temperature dependence of the Ni-Ti mixing briefly investigated. Room temperature irradiations with 600 keV Xe-ions were estimated to lead to mixing rates of -7.5x10 4 A4 and -1.4x10 4 A4 for Ni-Ti and Fe-Ti multilayers, respectively. However, at 80K the Ni-Ti rate was seen to agree with the Fe-Ti rate to within -45% (the experimental uncertainty) which would suggest that the difference between the two systems at room temperature is caused primarily by additional contributions from a temperature dependent mechanism, such as radiation enhanced diffusion. Johnson et al. 18] have proposed the expression c2
2AoL2/A+ = K1
bHImi x
- -- --- -- --- 5
P /
3
(A'coh )
2
(1 + K2
-
)
(1)
6.coh
for thermal spike mixing of a bilayer sample with large layer thicknesses. p is the average atomic density at the depth in question, and fHcoh is the cohesive energy. In general, we may expect the mixing efficiency to vary with film thickness 18,10], decreasing with increasing thickness for positive H*.ix (de-mixing during the spike counteracting ballistic mixing effects
Mat. Res. Soc. Symp. Proc. Vo
Data Loading...
