The Effect of Hydrogen Content on Ion Beam Mixing
- PDF / 390,481 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 20 Downloads / 230 Views
THE EFFECT OF HYDROGEN CONTENT ON ION BEAM MIXING R. E. WISTROM*, P. BORGESEN*, H. H. JOHNSON*, and D. A. LILIENFELD** *Department of Materials Science and Engineering, Bard Hall Nanofabrication Facility Cornell University, Ithaca, NY 14853
"**National
ABSTRACT A qualitative study of ion beam mixing of multilayers has shown that the presence of hydrogen in the sample slows mixing considerably for Fe/Ti, and slightly for Ni/Ti. We have quantified this effect and extended the study to four more systems (Co/Ti, Pd/Ti, Ti/Cu, and Ti/Al) and to lower temperatures. The degree to which H charging reduced the mixing rate varied substantially with multilayer system. H was lost during mixing. H was lost fastest from those films for which the H effect was the smallest. Our data are consistent with a model that relates mixing inhibition to H-vacancy binding energy in the non-Ti component.
INTRODUCTION Ion beam mixing of multilayer samples effectively produces a wide range of stable and metastable alloys [1,2]. Impurities can influence mixing rates [3-5], but hydrogen contamination is often overlooked since it does not appear in Rutherford backscattering (RBS) spectra. Hirvonen et al. [6] qualitatively compared mixing rates of Ni/Ti and Fe/Ti multilayers, and reported that H charging slightly inhibited mixing of Ni/Ti and strongly inhibited mixing of Fe/Ti. They suggested that H may bind to vacancies to reduce the vacancy mobility, and thus the mixing rate. This would explain why the mixing inhibition was more pronounced for Fe/Ti than for Ni/Ti. We will refer to this mixing inhibition as the "H effect". The technological importance of alloys containing H getters such as Ti [7-10] makes the H effect of immediate interest. We have quantified mixing results for Fe/Ti, Ni/Ti, Co/Ti, and Pd/Ti multilayers and for Ti/Al and Ti/Cu bilayers. Our mixing rate measurements confirm Hirvonen et al.s results for Fe/Ti and Ni/Ti. Possible explanations for the effect are discussed. EXPERIMENT Multilayer and bilayer samples consisting of Ti and another metal (Ni, Fe, Co, Pd, Cu, or Al) were electron beam deposited in an ion pumped system. Pressures were typically 2x10"7 torr before and 5x10-7 torr during deposition. To minimize oxidation, Ti was never the top layer. Multilayer compositions and thicknesses are listed in Table I. Electrochemical charging was performed in deaerated 0.1 N NaOH or 0.001 N H 2 SO 4 at about 2mA/cm 2 for times ranging from 0.5 to 2 hours using a Pt
Mat.Rea.Soc. Symp. Proc. Vol. 128. 91989 Materials Research Society
220
anode. Hydrogen content was measured by Forward Recoil Energy Spectrometry (FRES) [11]. The H/Ti ratio was approximately one for all charged samples with a wide range of charging times, showing that the H content reached a saturation level. Cu/Ti and Al/Ti multilayers peeled apart when charged, therefore, bilayer samples of Ti on top of Al or Cu were deposited. For these samples, the H/Ti ratio was less than one after charging, probably due to Ti oxidation. Thin coatings (50A) of Pd were therefo
Data Loading...
