Structures and reactivities of oxides under non-UHV conditions:

 A New Type of Pressure Gap

            We are using STM and other surface science techniques to probe the behavior of ordered oxide surfaces and oxide-supported metal particles at pressures above ultra-high vacuum (UHV; P ≤ 10^-8 Torr).   These studies are being carried out in close collaboration with electronic structure and molecular dynamics calculations by Dwight Jennison of Sandia National Laboratories/Albuquerque 

            The structures and reactivities of oxide surfaces are of critical importance to catalysis, micro- and nano-electronics, bio-engineering, and to corrosion inhibition.  In such applications, the environment at the oxide surface is typically either a liquid, or a gas with pressures >> 10^-8 Torr.  Our understanding of oxide structure and reactivity, however, is generally derived from experiments carried out under rigorously controlled UHV conditions.  While such UHV studies have added much to our understanding, higher pressure conditions can give rise to reaction pathways which are not observed in UHV.  Such qualitatively different behavior at higher pressure than at UHV is often termed a pressure gap.  We have recently identified such behavior for H₂O molecules at the surfaces of transitional phase Al₂O₃.[1], at P_H2O> 10^-4 Torr, at 300 K.

            A high resolution STM image of the vicinal surface of a 7 Å thick Al₂O₃ film grown by oxidation of a Ni₃Al(110) substrate[2], is shown in figure 1 (left), together with,  with corresponding LEED data[3] (right).  The STM image and LEED data were acquired at 300 K under UHV conditions.

Fig. 1  (Left)Near atomic resolution STM image of a 7 Å thick Al₂O₃/ Ni₃Al(110) film in UHV, 300 K, showing a 10 Å repeat distance in the surface structure.  (Right) LEED data showing scattering attributable to the substrate, interface and oxide.  The 10 Å repeat distance is characteristic of the kappa phase.  Details in references [1] and [2]

First-principles Density Functional Theory (DFT) results for the structure of a fully-relaxed κ-phase alumina film are shown in figure 2.  The theoretically predicted structure is in excellent agreement with experiment (fig. 1), exhibiting a 10 Å repeat distance and unit cell which matches the LEED results, indicating that the actual film (fig. 1) is indeed κ phase.

Figure 2:  Results of DFT calculations (Jennison) for a fully relaxed kappa film.  The lattice distances are in excellent agreement with STM and LEED data (fig. 1). The calculations also predict the presence of interstitial vacancies (arrow), which may store atomic hydrogen

Exposure of such ultrathin Al₂O₃ films to H₂O vapor at pressures > 10^-5 Torr, 300 K results in drastic reconstruction of the surface, which is initiated at defect sites, and gradually overwhelms the entire surface.  This effect is shown in figure 3 for to 7Å thick Al₂O₃ films.  The top film is grown on the Ni₃Al(110) substrate, and is kappa-phase, while the bottom film is grown on the Ni₃Al(111) surface and is gamma-like.[4] At H₂O pressures of ~ 10^-7  Torr or below, the films are inert.  Interestingly, the surface reconstruction (fig. 3) occurs without the formation of a UHV-stable surface hydroxide phase, the latter being observed only upon exposures to H₂O partial pressures > 1 Torr[5], in common with α-phase surfaces[6,7]

Fig. 3. 300 nm x 300 nm STM constant current images of: (a) a 7 Å thick as grown Al₂O₃/Ni₃Al(110)film (b)  a Al₂O₃/Ni₃Al(110) film exposed to H2O at 10^-4 Torr, 45 min, 300 K; (c) Al₂O₃/Ni₃Al(110) exposed to H2O at 10^-4 Torr, 90 min, 300 K; (d) a 7 Å thick as grown Al₂O₃/Ni₃Al(111) film; (e) Al₂O₃/Ni₃Al(111) after exposure to H₂O at 10^-4 Torr; 45 min, 300 K; (f) the Al₂O₃/Ni₃Al(111) exposed to H₂O at 10^-4 Torr, 90 min, 300 K. Scanning parameters: gap voltage=2 V, current=0.1 nA.

That both transitional-phase films react toward H₂O indicates that this process is not specific to a specific phase or oxide/metal interface.  Other data (not shown) indicate that :

• The surface reconstruction begins at the oxide/vapor interface, not at the oxide/metal interface, as is therefore NOT due to H₂O diffusion.

• The reconstruction process is specific to H₂O. (Similar exposures to CO have no effect.)

• The surface reconstruction process is pressure-dependent, rather than exposure-dependent , demonstrating that the reaction is cooperative; involving interaction between two or more H₂O molecules.

The cooperative nature of this phenomenon explains why it is not observed under UHV conditions, and indicates its potential importance to catalysis over alumina supports, since an impurity of only 1 ppm in a 100 Torr environment of, e.g., CO results in the partial pressures of water vapor at which this effect can be observed.  Additionally, recent experimental and theoretical results of Jennison and co-workers [8] strongly suggest that H₂O-alumina interactions result in the generation of atomic hydrogen, with potentially far-reaching impact for hydrogen and hydrocarbon catalysis, microelectronics processing, corrosion, and MEMS fabrication.

We are currently testing the possibility of hydrogen generation and storage in alumina, and plan to extend these studies to other oxides and oxide-supported metal nanoparticles, all under non-UHV conditions.

References:

1. F. Qin, N. P. Magtoto and J. A. Kelber, Surface Science  565, L277

2. F. Qin, N. P. Magtoto, J. A. Kelber, and D. R. Jennison, J. Molec. Catalysis A (in press)

3. F. Qin, N. P. Magtoto and J. A. Kelber, Materials at High Temperature (in press)

4.  S. G. Addepalli, B. Ekstrom, N. P. Magtoto, J.-S. Lin, and J. A. Kelber, Surface Science 442, 385-99(1999)

5. M. Garza, N. P. Magtoto and J. A. Kelber,  Surface Science 519, 259-68 (2003)

6. P. Liu, T. Kendelewicz, J. G. Brown, E. G. Nelson and S. A. Chambers, Surface Science 417, 53-65 (1998)

7. C. Niu, K. Shepherd, D. Martini, J. Tong and J. A. Kelber, and D. R. Jennison and A. Bogicevic,  Surface Science 465, 163-76 (2000)

8. D. R. Jennison, P. A. Schultz and J. P. Sullivan, Physical Review B 69, 041405(R) (2004)