Supplementary Information 1) Surface alloy stability tests Surface segregation stability tests are performed by considering all possible segregation events that could occur, for each alloy, within our ( 3 R o x 3) 30 unit cell, both with and without adsorbed hydrogen. Using a Grand Canonical free energy formalism 1, we evaluate the free energy change for each such event by assuming a hydrogen chemical potential equivalent to aqueous protons and electrons at zero potential vs. the reversible hydrogen electrode and ph=0, and we determine the most thermodynamically favorable event for each alloy. If the free energy change (normalized to the number surface unit cells) for this event is negative, then the alloy is likely to be unstable with respect to surface segregation. Intrasurface instabilities, such as surface dealloying, are evaluated by calculating the free energy change per surface unit cell, in both the presence and the absence of adsorbed hydrogen, associated with several possible rearrangements for each surface alloy. The considered rearrangements include phase separation of the alloys into multiple product phases and are limited to structures permitted by the ( 3 R o x 3) 30 unit cell in our calculations. We note, in passing, that by including pure solute overlayers in the product structures described above, it is possible, even with small unit cells, to simulate the formation of islands (the edge energy of the islands is, however, neglected in these analyses). For all stability tests involving intrasurface transformations, the same hydrogen chemical potential value as for the surface segregation tests is used.
The surface alloy stability test for oxygen adsorption considers only O adsorption on the unreconstructed surface alloys. We note that combined stability tests involving both reconstruction and adsorption (for example, surface segregation followed by oxygen adsorption) could also be envisioned. This could remove alloys where a metal that binds oxygen weakly completely covers a metal that binds oxygen strongly, such as Rh, from consideration. Free energies of metal stripping (dissolution) are evaluated using data from the electrochemical series 2. For each surface alloy, the surface element with the most negative free energy of dissolution is used in the stability analyses; the indicated dissolution free energies are used without further modification. 1. Bollinger, M. V., Jacobsen, K. W. & Nørskov, J. K. Atomic and electronic structure of MoS2 nanoparticles. Physical view B 67, #085410 (2003). 2. CRC Handbook of Chemistry and Physics (CRC Press, New York, 1996). 2) Captions for supplementary figures Figure S1. mputational high-throughput screening for ΔG H on 256 pure metals and surface alloys. The rows indicate the identity of the pure metal substrates, and the columns indicate the identity of the solute embedded in the surface layer of the substrate. The solute coverage is (a) 2/3 ML and (b) a full ML. The adsorbed hydrogen coverage is 1/3 ML in both cases. The diagonal of the plot corresponds to the hydrogen adsorption free energy on the pure metals. Figure S2. Cyclic voltammetry in the H-UPD region, indicating exposed surface area of the sample after 3 stages of synthesis. We measured multiple CVs for the pure sample, but measured only one CV for each step involving in order to preserve the state of. For the sake of consistency, only the initial CVs are plotted in the main figure, with the inset exhibiting a voltammogram of an identically prepared pure after the furnace treatment, and after sweeps in a wide potential window. Sweep rate: 50 mv/s.
Fe Solute Element Rh Pd Ag Cd Sb Fe Host Element Rh Pd Ag Cd Sb G H >0.5 0.4 0.5 0.3 0.4 0.2 0.3 0.1 0.2 0 0.1 2/3 ML Supplementary 1(a) Nørskov
Fe Solute Element Rh Pd Ag Cd Sb Fe Host Element Rh Pd Ag Cd Sb G H >0.5 0.4 0.5 0.3 0.4 0.2 0.3 0.1 0.2 0 0.1 Full ML Supplementary 1(b) Nørskov
rrent (ma) 0.10 0.05 0.00-0.05 pure after annealing - ir surface alloy 0.2 0.1 0.0-0.10-0.1-0.2 0.0 0.1 0.2 0.3 0.4 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Voltage vs. SHE Figure S2. Nørskov et al