Apparent “three-dimensional” Fermi surface of transition-metal monolayers

One of the fundamental properties of an electronic system is its dimensionality. Novel photoemission apparatuses enable a direct determination of the Fermi surface of metals in all three dimensions but the data obtained from transition-metal monolayers appears to defy established experimental and theoretical insights. Our experimental approach resolves all of the existing conflicts surrounding Ni/Cu(100) and similar systems. We suggest the measurement of Fermi-surface projections to optimize spin-dependent transport properties.

Figure 1

(Color online) Angle-resolved photoemission in normal emission. $\text{Ni}d$ states do not disperse with $h\nu$ or ${\mathbf{k}}_{\perp }$ for 1 ML (two-dimensional) but disperse strongly for 6 ML Ni (three-dimensional). For 1 ML, ticks mark ${\overline{\Gamma }}_1$ states at 0.9 eV and ${\overline{\Gamma }}_5$ states at 0.3 eV. For 6 ML, solid ticks mark ${\Delta }_1$ states and dashed ticks mark ${\Delta }_5$ states.

Figure 2

(Color online) Angle distribution of photoemission intensity at $E_{\text{F}}$, $h\nu =60\text{eV}$. Predominantly $\mathbf{k}\perp \mathbf{E}$ (top) and $\mathbf{k}\parallel \mathbf{E}$ (bottom) (Ref. 12). Fermi-surface contours are similar for 1 and 6 ML Ni and very different from those of Cu.

Figure 3

(Color online) Angle-dependent photoemission spectra, $h\nu =60\text{eV}$ (Ref. 12). Dispersions with ${\mathbf{k}}_{\parallel }$ are less pronounced for 1 ML than for 6 ML Ni while they intersect $E_{\text{F}}$ at similar ${\mathbf{k}}_{\parallel }$ vectors.

Figure 4

(Color online) Angle-resolved constant-initial-state spectroscopy at $E_{\text{F}}$. Characteristic features C of 1 ML Ni and E of bulk Ni appear in Cu band gaps which prevent coupling to the substrate.

Figure 5

(Color online) Calculated orbital character at $E_{\text{F}}$ of Cu and Ni bands compiled using Ref. 17. The fourth minority-spin band of Ni ($\text{Ni}\downarrow$ band 4) corresponds to feature E in Fig. 4f which develops for 6 ML Ni and not for 1 ML Ni. It is of pure $d$ character.