Recently, Greiner et al. [ Nat. Mater. 2012, 11 , 76 ] published a survey of the level alignment of about 40 metal oxide/organic molecule interfaces. They observed a striking regularity in the electronic level alignment of the highest occupied molecular orbital (HOMO) and the Fermi level that depends solely on the difference between the substrate work function and the ionization energy of the molecule independent of the details of the electronic structure of the oxide. The authors could reproduce their data under the assumption of thermodynamic equilibrium occupation of the HOMO using four adjustable parameters. A model that quantifi es well-established concepts in heterojunction physics and achieves the same result without any adjustable parameters is presented here. This approach explains why the level alignment is rather independent of the experimental details, such as the electronic structure of the oxide, defects in the oxide, and the thickness of oxide and overlayer. are not included in the spectroscopically determined HOMO and LUMO energies as expressed through IE org and EA org , respectively. The unusually large polaronic binding energy is, according to the authors, due to the fact that one is dealing here with interface polarons. We shall return to this question later.Recently, Greiner et al. [ 29 ] expanded the results of Salaneck and co-workers [ 26 ] to encompass small molecules as well. They too observed the upper part of the Zorro curve for a great number of organic/metal oxide molecule interfaces, albeit measured slightly differently. Instead of the work function, they measured the energy of the HOMO orbital relative to E F , (i.e., the hole-emission barrier Φ Bp ), and plotted Δ E H = Φ Bp = E F -E HOMO vs. Φ sub -IE org . This plot is redrawn here as Figure 2 . Substrates and interfaces were prepared in UHV and the substrate work function, the ionization energy of the organic molecules, and Δ E H were measured in situ by photoemission spectroscopy. In this sense, Greiner et al.'s [ 29 ] procedure follows more closely that of Kahn and co-workers, [ 27 ] except for using metal substrates with a well-defi ned oxide layer on top. Nevertheless, the experimental results comply perfectly with those of Salaneck and co-workers and the interpretation is based on the same concept, namely the transition from the Schottky-Mott regime to pinning when the substrate work function equals the ionization energy:The transition at Φ sub -IE org = 0, as suggested by the dashed line, indicates that the relevant ICT + is indeed the HOMO. However, spectroscopically the HOMO is measured about 0.3 eV below E F . Greiner et al. [ 29 ] were able to describe their considerable body of data with an expression that contains four adjustable parameters. In what follows, we present a the electron affi nity and ionization energy of the organic semiconductor, respectively, as they are measured for the isolated constituents of the interface. This, of course, presupposes that the LUMO and HOMO are the transport levels for electrons a...

Surface sensitive C1s core level photoelectron spectroscopy was used to examine the electronic properties of C60F48 molecules on the C(100):H surface. An upward band bending of 0.74 eV in response to surface transfer doping by fluorofullerene molecules is measured. Two distinct molecular charge states of C60F48 are identified and their relative concentration determined as a function of coverage. One corresponds to ionized molecules that participate in surface charge transfer and the other to neutral molecules that do not. The position of the lowest unoccupied molecular orbital of neutral C60F48 which is the relevant acceptor level for transfer doping lies initially 0.6 eV below the valence band maximum and shifts upwards in the course of transfer doping by up to 0.43 eV due to a doping induced surface dipole. This upward shift in conjunction with the band bending determines the occupation of the acceptor level and limits the ultimately achievable hole concentration with C60F48 as a surface acceptor to values close to 1013 cm−2 as reported in the literature.

No abstract

We demonstrate a novel doping mechanism of silicon, namely n-type transfer doping by adsorbed organic cobaltocene (CoCp2*) molecules. The amount of transferred charge as a function of coverage is monitored by following the ensuing band bending via surface sensitive core-level photoelectron spectroscopy. The concomitant loss of electrons in the CoCp2* adlayer is quantified by the relative intensities of chemically shifted Co2p components in core-level photoelectron spectroscopy which correspond to charged and neutral molecules. Using a previously developed model for transfer doping, the evolution in relative intensities of the two components as a function of coverage has been reproduced successfully. A single, molecule-specific parameter, the negative donor energy of -(0.50±0.15)  eV suffices to describe the self-limiting doping process with a maximum areal density of transferred electrons of 2×1013  cm-2 in agreement with the measured downward band bending. The advantage of this doping mechanism over conventional doping for nanostructures is addressed.

We demonstrate that the intrinsic electron doping of monolayer epitaxial graphene on SiC(0001) can be tuned in a controlled fashion to holes via molecular doping with the fluorinated fullerene C60F48. In situ angle-resolved photoemission is used to measure an upward shift of (0.6 ± 0.05) eV in the Dirac point from −0.43 eV to +0.17 eV relative to the Fermi level. The carrier density is observed to change from n ∼ (1 × 1013 ± 0.1 × 1013) cm−2 to p ∼ (2 × 1012 ± 1 × 1012) cm−2. We introduce a doping model employing Fermi-Dirac statistics which explicitly takes temperature and the highly correlated nature of molecular orbitals into account. The model describes the observed doping behaviour in our experiment and readily explains why net p-type doping was not achieved in a previous study [Coletti et al., Phys. Rev. B 81, 8 (2010)] which used tetrafluorotetra-cyanoquinodimethane (F4-TCNQ).