Alloy Formation at the CoAl Interface for Thin Co Films Deposited on Al001 and Al110 Surfaces at Roo

1
Alloy Formation at the Co-Al Interface for Thin
Co Films Deposited on Al(001) and Al(110)
Surfaces at Room TemperatureN.R. Shivaparan,
M.A. Teter, and R.J. Smith, Physics Dept.,
Montana State University, Bozeman MT 59717

• Goal Understand epitaxial growth of metals on
  metals
• Contributing factors include lattice matching,
  surface energies, formation energies,
  strain energy
• Related work "small" atoms (Ni, Fe, Pd,) form
  alloys at the Al surface "large" atoms ( Ag,
  Ti ) tend to form an overlayer at room
  temperature (kinetic barrier)
• What about Co ? Alloy or overlayer?
• Co-Al phase diagram very similar to Ni-Al
• 15 lattice mismatch with Al
• metallic radii of Co and Ni both smaller than Al
• surface energy of Co larger than for Al
• negative formation energy -54 kJ/mole for CoAl
• Motivation recent interest in GMR and tunneling
  devices
• Co/Al2O3/Co structures may have diffuse
• interfaces which affect electron (spin)
  transport
• Present work XPS, HEIS, LEED for Co deposition
  using
• resistively heated Co wires.

• Figure 1. (left) HEIS Channeling spectra for
  0.84 ML Co
• Measure an increase of Al surface peak, so Co
  atoms cause
• Al atoms to move off lattice sites (alloy
  formation)
• Coverage of Co is determined from area of Co
  peak
• Figure 2. (right) HEIS Al surface peak vs. Co
  coverage
• Number of visible Al atoms increases up to 3 ML
• Slope of 21 suggest a stoichiometry of Al2Co
  for the
• interface but Al2Co is not found in the bulk
  phase diagram
• Interface formation stops after 3 ML and Co
  metal covers
• the interface

• Figure 3. (left) HEIS Co surface peak (aligned)
  vs. Co coverage
• No decrease in Co surface peak area for aligned
  geometry as compared to a random incident
  direction
• Thus no ordered Co structure aligned with the
  substrate (100) axis.
• Figure 4. (right) XPS Intensity for Al 2p and Co
  2p3/2 vs Co coverage
• Al intensity is normalized to clean surface
  value Co intensity is normalized to measured
  value at 7.4 ML
• Al yield is attenuated, but more slowly than
  expected for growth of a
• Co metal adlayer Co yield grows rapidly,
  non-linearly
• Lines are from model calculations for growth of
  CoAl compound with (100) orientation, followed
  by Co metal island growth after 3 ML
• Co binding energy shifted to larger values in
  interface, 0.1 eV
• Figure 5. (far right) LEED patterns for Co on
  Al(001) at 42.8 eV
• (a) Clean Al(001)
• (b) 0.5 ML Co destroys pattern completely
• (c) 7.6 ML A hint of some long range order
  (1x1) is seen.

• Figure 6. HEIS Al yield vs. Co coverage on
  Al(110)
• Number of visible Al atoms increases up to 5 ML
• Slope of 2.31 suggests a stoichiometry of Al2Co
  or Al5Co2.
• Al5Co2 exists in bulk phase diagram, with
  formation
• energy of -41 kJ/mole, but a more complex
  crystal
• structure, so less likely to form at room
  temperature.
• Interface formation stops after 5 ML and Co
  metal covers
• the interface.
• Figure 7. (right) XPS Intensity for Al 2p and Co
  2p3/2 vs Co coverage on Al (110)
• Al intensity normalized to clean surface value
• Co intensity normalized to thick film value
• Al yield is attenuated very slowly at first, but
  rapidly after
• 5 ML suggests compound growth followed by Co
  metal in a
• layer-by -layer mode
• Lines are from model calculations for growth of
  CoAl
• compound with (110) orientation, then Co
  metal growth in a
• layer-by-layer mode

• Conclusions
• Co deposition leads to interface alloy formation
  on both
• Al surfaces, similar to that seen for Ni on
  Al(110)
• V. Shutthanandan, et al., Surf. Sci 450 (2000)
  204
• Behavior seen here for Co-Al differs from that
  for Ni-Al
• in that the Co-Al interface is more abrupt
  (thinner) and
• has a single growth stage, i.e. a single
  compound.
• No strong evidence observed for an ordered
  epitaxial
• compound on either surface. Hint of order
  on (100).
• XPS intensities are fit well using a CoAl
  stoichiometry but HEIS data suggests more Al
  rich interface. May
• be related to dechanneling below the interface.
• No diffusion of Co into Al is observed at room
• temperature.

Work supported by NSF Grant DMR-9409205 and
0077534, and by NASA EPSCoR grant NCCW-0058. A
copy of this poster is available at
http//www.physics.montana.edu/Ionbeams/ionbeams.h
tml