AVS2014 Session SS+EM-FrM: Semiconductor Surfaces and Interfaces 2

Organic-based electronics are attractive because they have potential manufacturing advantages such as mechanical flexibility and simpler processing (solution-based, low temperature, and atmosphere conditions). Molecular-based semiconductors offer a nearly limitless range of possibilities in tailoring the chemical composition and structure for a desired electronic, optical, or film-processing property. Probing and understanding molecular surfaces and interfaces is essential for the further development of organic-based photovoltaics, light emitting diodes, and field-effect transistors. Organic-organic interfaces are key in some of those devices, and understanding the impact of a self-assembled monolayer (SAM) has when an organic semiconductor is on top of it, is a complex issue.¹ This strategy is commonly implemented as a way to modify the hole injection barrier between an organic material and an inorganic substrate.

Here, we have investigated the interaction between a pi-conjugated organic semiconductor (tris-(8-hydroxyquinoline) aluminum, Alq₃) on SAM’s of different tail and backbone composition. We have used ultraviolet and X-ray photoelectron spectroscopies to monitor the energy level alignment and chemical structure at the interface. The SAM’s strongly interact with the Au substrate, where an interface dipole can down shift or up shift the surface work function. After Alq₃ is deposited onto the SAM-coated substrates, we find that the highest occupied molecular orbital of Alq₃ is relatively constant (with respect to the substrate Fermi level) on all surfaces, suggesting Fermi level pinning.² However, the composition of the SAM’s did strongly influence the growth and chemical structure of the Alq₃ at the interface. The photoemission signal arising from the Au substrate is least attenuated when the SAM/Au surface is hydrophobic when compared to a hydrophilic SAM/Au or bare Au surface. The difference in substrate attenuation suggests that that the early growth of the Alq₃ layer strongly depends on this surface property. This finding is corroborated with microscopy of the same samples. In addition, Alq₃ chemically reacts with a fluorinated SAM at the organic-organic interface as indicated by the shifting and asymmetric broadening of Al and N core levels. These results will be discussed in context of painting a comprehensive picture of the organic-organic interface formation that influences the chemical composition, electronic structure and physical structure at the interface.

[1] F. Rissner et al, ACS Nano 3, (2009) 3513.

[2] L. Lindell et al., Appl. Phys. Lett. 102, (2013) 223301.

Quinhydrone dissolved in methanol has long been know to react with hydrogen terminated silicon surfaces to passivate electronic defects where photo-excited carriers recombine non-radiatively. The mechanism of this passivation is not well understood. We have shown that benzoquinone, C₆O₂H₄ rather than hydroquinone, C₆O₂H₆, both components of the quinhydrone mixture, is the active component. Benzoquinone reacts to abstract a hydrogen and then itself bonds with the surface. We have shown that the hydrogen can be abstracted from the solvent and that incident light is necessary for this reaction to take place. X-ray photoelectron spectroscopy and Fourier Transform Infrared Spectroscopy were used to show that the benzoquinone reacted with the surface. Photo-excited carrier lifetime is a good measure of the extent of the passivation of the surface. Density functional theory supports the proposed reaction mechanism.

The use of dielectric films in device passivation is complicated by the fact that they are typically deposited on processed material surface that bear little resemblance to that of the virgin growth surface. This is particularly evident in technologically important device structures employing antimonide-based compound semiconductor (ABCS) superlattices, where the exposed mesa sidewalls may be comprised of four or more atomic species and their complex oxides. Physically, the etched surface presents a different crystallographic orientation, and may have additional structure due to variation in etch rate of superlattice layers. Since the nature of the dielectric/semiconductor interface directly impacts the density of surface states, it is critical to understand how processed, multilayer semiconductor surfaces may be modified during the initial phase of the atomic layer deposition (ALD) process.

A significant effort has been focused on surface preparations prior to ALD that removes the native oxide and passivates the III-V atoms in order to ensure the best possible interface. Current approaches typically rely upon wet-chemical etches to remove the defect-prone native oxide layer prior to dielectric deposition; however, this technique typically suffers from a lack of reproducibility, as well as potential interface contamination between processing steps. Therefore, we studied the effectiveness of using the ALD precursor, TMA, in conjunction with wet and dry pre-treatments, in removing carbon and etch precipitates, scavenging the various oxide species, and residues of excess group III and V elements on (100) surfaces of ABCS superlattices as a function of precursor choice, sequence (i.e. TMA vs oxidizer first), exposure time, as well as substrate temperature. Furthermore, surface passivation stability was investigated as a function of temperature and time. Surfaces were analyzed using XPS, AFM, and SEM both before and after ALD treatments. Results indicate that a completely oxide free surface may not be necessary to produce a good electrical interface.

Etching of Si in oxidant + HF solutions can lead to a self-limiting reaction that spontaneously produces nanocrystalline porous Si – a process known as stain etching. The presence of a metal catalyzes and localizes etching such that ordered arrays of pores or nanowires can be formed depending on the structure of the metal – a process known as metal assisted etching (MAE). Ag, Au, Pd and Pt were deposited from solution onto H-terminated Si to act as catalysts for MAE. The metals all catalyzed the injection of holes into the Si. They all increased the rate of hole injection by approximately a factor of 5. The stoichiometry of MAE in V₂O₅ + HF solutions depended on the chemical identity of the metal. The stoichiometry when etching with Ag and Au was the same as for stain etching in V₂O₅ + HF solutions. However, for Pd and Pt, the stoichiometry differed significantly, consuming more V₂O₅ and producing less H₂ per mole of Si etched. This indicates that the metal catalyst can change the mechanism of etching. Etching in V₂O₅ + HF solutions was well behaved and gave consistently reproducible kinetic results. The behavior is much different when HOOH is added instead of V₂O₅. In the absence of deposited metal, no reaction occurs with HOOH. When HOOH was added to metal-coated Si samples immersed in HF(aq), etching was immediate in all cases. In contrast to V₂O₅, we were unable to obtain well-behaved stoichiometric results for HOOH + HF solutions. This is related to heightened sensitivity on reaction conditions compared to the V₂O₅ system as well as nonlinearities introduced by side reactions.

The mechanism of Si etching changes based on the presence of a metal catalyst during metal assisted etching and depends on the chemical identity of the metal. A valence 2 path dominates the formation of photoluminescent nanoporous Si in stain etching as well as MAE with Ag and Au. A valence 4 path dominates the formation of photoluminescent nanoporous Si in MAE with Pt. However for MAE with Pd, no nanoporous Si is formed initially and a mixture of valence 4 and valence 2 processes is observed. The nature of the electron transfer process and its dependence on the electronic structure of the metal/Si interface will be discussed.

We report about accurate monitoring of ultra-low La doses inserted in advanced high-k/metal gate stacks for threshold voltage tuning purposes. Three characterization techniques are implemented for precise and reproducible lanthanum quantification. LEXES (Low energy Electron X-ray Emission Spectrometry) capabilities are highlighted in terms of sensitivity and accuracy thanks to a comparison with reference results obtained by Rutherford Backscattering Spectrometry (RBS). The capabilities of state-of-the-art Auger nanoprobes for depth profiling in the sub-nanometer range are also illustrated.