Tuning the work function in transition metal oxides and their heterostructures

The development of novel functional materials in experimental labs combined with computer-based compound simulation brings the vision of materials design on a microscopic scale continuously closer to reality. For many applications interface and surface phenomena rather than bulk properties are key. One of the most fundamental qualities of a material-vacuum interface is the energy required to transfer an electron across this boundary, i.e., the work function. It is a crucial parameter for numerous applications, including organic electronics, field electron emitters, and thermionic energy converters. Being generally very resistant to degradation at high temperatures, transition metal oxides present a promising materials class for such devices. We have performed a systematic study for perovskite oxides that provides reference values and, equally important, reports on materials trends and the tunability of work functions. Our results identify and classify dependencies of the work function on several parameters including specific surface termination, surface reconstructions, oxygen vacancies, and heterostructuring.

Figure 1

Plane averaged electrostatic potential of a symmetric ${\mathrm{SrRuO}}_3$ slab consisting of six SrO layers, five ${\mathrm{RuO}}_2$ layers, and a 20 Å vacuum. The electrostatic potential is defined with respect to Fermi energy $E_F$, and converges to a constant value in the vacuum region that is the work function $\mathrm{\Phi }$ of ${\mathrm{SrRuO}}_3$ with SrO surface termination. $\mathrm{\Phi }$ therefore indicates the required energy to remove an electron at Fermi level from the material to a state at rest in the vacuum nearby the surface.

Figure 2

DFT calculated work functions of the perovskite ${\mathrm{ABO}}_3$ series with A=Sr and $B$ being an element of the $3d$ (red) or $4d$ (blue) period, and with AO (circle with solid lines) and ${\mathrm{BO}}_2$ (square with dashed lines) surface termination. Experimental work function of ${\mathrm{SrTiO}}_3$ with SrO and ${\mathrm{TiO}}_2$ termination by Susaki et al. [34] as well as ${\mathrm{SrRuO}}_3$ with unknown termination by Fang et al. [33] are shown as stars.

Figure 3

DFT calculated work functions of the ${\mathrm{ATiO}}_3$ (red), ${\mathrm{AZrO}}_3$ (black), and ${\mathrm{ARuO}}_3$ (blue) series with A=Ca, Sr, and Ba, and with AO (circle with solid lines) and ${\mathrm{BO}}_2$ (square with dashed lines) surface termination.

Figure 4

DFT calculated density of states for ${\mathrm{SrVO}}_3$ capped with one unit cell of ${\mathrm{SrTiO}}_3$ (upper panel) and ${\mathrm{SrNbO}}_3$ capped with one unit cell of ${\mathrm{SrTiO}}_3$ (lower panel). We show only the partial density of states for V, Nb, and Ti $3d$ states. $t_{2g}$ and $e_g$ states are indicated by solid and dashed lines, respectively. The figure shows a clear material dependence of the charge transfer between base and capping material.

Figure 5

Plane averaged electrostatic potential for a ${\mathrm{LaAlO}}_3/{\mathrm{SrRuO}}_3$ heterostructure calculated with DFT: three layers of ${\mathrm{LaAlO}}_3$ grown on a ${\mathrm{SrRuO}}_3$ substrate with AO termination (upper panel) or ${\mathrm{BO}}_2$ termination (lower panel). The internal field of the polar layers tunes the work function depending on its direction.

Figure 6

Work function plotted versus thickness of polar capping material. As already indicated in Fig. 5 the additional field produced by polar layers can be used to tune the work function in either way. Here we show how either termination can be tuned up or down by choosing the appropriate capping material ${\mathrm{LaAlO}}_3$ or ${\mathrm{KTaO}}_3$ which have opposite effects due to their opposite polarity.