Impact of Bidirectional Charge Transfer and Molecular Distortions on the Electronic Structure of a Metal-Organic Interface

Interface energetics are of fundamental importance in organic and molecular electronics. By combining complementary experimental techniques and first-principles calculations, we resolve the complex interplay among several interfacial phenomena that collectively determine the electronic structure of the strong electron acceptor tetrafluoro-tetracyanoquinodimethane chemisorbed on copper. The combination of adsorption-induced geometric distortion of the molecules, metal-to-molecule charge transfer, and molecule-to-metal back transfer leads to a net increase of the metal work function.

Figure 1

(a) F4-TCNQ on Cu(111) in the $3\sqrt{3}\times 5$ surface unit cell as seen from the side (bottom) and from above (top). The calculated (experimental) atomic distances to the surface are given to the right. (b) Total photoemission and reflectivity measurements as a function of photon energy relative to the Bragg energy. The fits (solid lines) to the experimental data (symbols) yield the average coherent positions $P_{\mathrm{eff}}$ and coherent fraction $f_{\mathrm{eff}}$ [13] used to determine the bonding distances for F ($P_{\mathrm{eff}}=0.59$, $f_{\mathrm{eff}}=0.43$), N ($P_{\mathrm{eff}}=0.30$, $f_{\mathrm{eff}}=0.28$), and C ($P_{\mathrm{eff}}=0.33$, $f_{\mathrm{eff}}=0.15$) atoms. A coherent fraction of the C signal of roughly one-third compared to the F signal is in excellent agreement with what can be derived from the calculated positions in the bent molecule.

Figure 2

UPS spectrum (circles) and DFT calculated DOS (solid line) of the organic layer. The experimental spectrum is obtained by subtracting the UPS signal of the bare Cu(111) surface from that of a covered surface with a nominal layer thickness of 4 Å (ca. 1 monolayer) $\mathrm{F}4\mathrm{\text{−}}\mathrm{TCNQ}/\mathrm{Cu}(111)$. Spectra at nominally 2 Å and 1 Å layer thickness look virtually identical. In the calculations, photoemission cross sections are not taken into account and the energy scale has been shifted ($-0.2\mathrm{eV}$) after expanding it (by a factor of 1.34), which is common practice when comparing DFT data within the generalized gradient approximation to UPS data [28]. The inset illustrates the measured work-function modification, $\Delta \Phi$, as a function of film thickness $\Theta$.

Figure 3

Top: DOS projected onto selected MOs. For clarity, the curves have been displaced on the vertical axis and the $y$ scale has been multiplied by a factor of 20 for energies greater than 1 eV above $E_F$. Bottom: Occupation (in percent) of the lowest 60 MOs. The full (open) circles and solid (dashed) lines correspond to the orbitals which are occupied (unoccupied) in the isolated molecule.

Figure 4

Charge rearrangement $\Delta {\rho }_e$ and related potential energy of an electron (black lines). $q_e$ is the charge of an electron. $\Delta {\rho }_e$ is calculated by subtracting the noninteracting electron density from the electron density of the interacting system and averaging in the $x\mathrm{\text{−}}y$ plane. The gray curve represents the $xy$-averaged potential energy of an electron for the noninteracting F4-TCNQ monolayer (see the text).