Thin Films: Preparation, Characterization, Applications

The potential-dependent atomic structures of a Cu(111) electrode surface exposed to dilute sulfuric acid have been studied by means of in-situ STM.

At a positive electrode polarization the specific adsorption of sulfate anions induce a reconstruction of the copper substrate characterized by an expansion of the topmost substrate layer. A direct proof for the adsorbate-induced reconstruction is given by a kind of spectroscopic STM measurement where not only the adsorbate but also the underlying reconstructed copper substrate is imaged. Sweeping the potential in negative direction sulfate desorbs from the surface accompanied by the lifting of the reconstruction. The bare copper surface, however, does not remain stable under these conditions. An adsorption of solvent species takes place leading to the formation of a highly ordered, hexagonal Moiré pattern which undergoes an electrocompression process starting with a c(4 x 4) and ending with a c(5 x 5) superstructure with decreasing potentials. These adsorbate layers remain stable even under massive hydrogen evolution. The dependence of the imaging properties of this adlayer on the tunneling conditions has been systematically studied.

It is well-known that the physical and chemical properties of an ultra thin metal layer on a foreign substrate are different from those of the bulk metal.^1,2 The establishment of the preparation method of the ultra thin metal layer with an ordered structure and the understanding of the origin of its unique physical and chemical properties are very important both for fundamental science and industrial applications. The epitaxial growth of a well-defined thin layer of metals has been achieved by vapor deposition, molecular beam epitaxy (MBE), and metalorganic chemical vapor deposition (MOCVD) under condition.^3,4 Compared to the metal deposition by these techniques in vacuum, electrochemical metal deposition is economical and easy because expensive vacuum equipments are not necessary for the electrochemical deposition. Unfortunately, however, the quality of the electrodeposited metal layers is usually low. Recent development of electrochemistry of single crystal electrode and in situ surface characterization techniques such as scanning tunneling microscopy (STM) and surface X-ray scattering (SXS) of atomic resolution make the growth of metal layer with an ordered structure under electrochemical control possible.⁵

The use of well-defined single crystal surfaces is at the origin of the development of Surface Electrochemistry. In particular, the discovery of the flame annealing technique^1,2 has been crucial for the use of electrocatalytic metals in fundamental studies. As a consequence, structure sensitivity effects have been pointed out, emphasizing the role of the surface structure in reactivity. An increasing number of studies deal now with these differences in reactivity, not only of the electrodes having surfaces corresponding to the basal planes³ but also of stepped electrode surfaces with well-defined step densities⁴. The most recent results in this particular field have been obtained by using kinked surfaces, which have shown unexpected properties, such as chiral reactivity.⁵

The under-potential deposition (UPD) of metallic and semiconducting species on noble-metal substrates has been studied over the past two decades using numerous electrochemical and spectroscopic techniques as well as scanning probes. Among various substrate-adsorbate systems, the electrodeposition of silver on single-crystal and polycrystalline Pt electrodes has attracted a lot of attention owing to its simplicity. Specifically, the UPD of Ag is a single-electron process, the substrate (Pt) and adsorbate (Ag) have the same cubic face centered (cfc) crystallographic structure, and possess similar lattice constants (a_Ag = 4.09 Å and a_Pt = 3.92 Å, respectively). The difference of the lattice constant of Ag and Pt is only 4.3%, yet the difference of the volumes of the Ag and Pt crystallographic units is 13.6% (a_Ag ³ = 68.42 ų and a_Pt ³ = 60.24 ų). Thus, for a purely crystallographic reason, the overlayer of Ag_UPD on the Pt( 111) substrate is always compressed. The UPD Ag on Pt(hkl) electrodes from various aqueous electrolytes has been studied using a variety of techniques such as cyclic-voltammetry (CV), in-situ scanning tunnelling microscopy (STM), atomic force microscopy (AFM), and surface X-ray scattering (SXS)^1–20. The first investigation oi the UPD Ag on an ordered Pt(111) electrode as well as on other low-index faces of Pt was reported by El Omar et al.². They observed that the deposition of Ag_UPD resulted in a two-layer deposit, but did not report any surface structure. A few years later, Franc et al. using angular distribution Auger microscopy (ADAM) concluded that the Ag_UPD overlayer was commensurate with the Pt(111) substrate¹¹, but this structural assignment later generated a great deal of controversy. Subsequently, it was observed that in order to accomplish a complete suppression of the under-potential deposition of H (UPD H) an overlayer of ca. 1.5 equivalent monolayers (ML) of Ag was required ¹².

Underpotential deposition (upd) provides a convenient method for plating an atomic layer of one metal onto the surface of another more noble metal.¹ Examples include the underpotential deposition of metals such as copper and silver onto the more noble surfaces of gold and platinum.^{1, 2} The structure and composition of these and other adlayers—and their dependence on factors such as electrochemical potential, crystal structure of the electrode metal, and ions in the electrolyte—have been the topic of numerous investigations.^1–5 In constrast, little work has been directed toward modifying the upd surfaces with other agents, most likely due to a perception of extreme fragility for upd adlayers. We find that some upd systems—notably, upd adlayers of silver and of copper on gold—offer sufficient stability under conditions where they can undergo chemical and/or electrochemical modification.^6–14 Figure 1 schematically illustrates these processes of modification that are the topic of this chapter. Specifically, the upd layers can be derivatized by the adsorption of ligating species—examples include alkanethiols,^6–9 alkylphosphonic acids,¹⁰ and halides^12–14—and remain as a stable adlayer on the gold surface during this process. Notably, the presence of the monoatomic copper and silver adlayers alter the chemisorptive properties of the gold surface, thereby providing new approaches for modification. In addition, for some metal combinations, the upd adlayer can be modified electrochemically by the upd of a second metal onto the parent electrode surface to produce a two-component adlayer of monoatomic thickness on an electrode surface.¹¹

Compound semiconductors are an important group of materials, used in a wide variety of optoelectronic devices, including detectors, displays and photovoltaics. They are generally used in the form of thin films, deposited by molecular beam epitaxy (MBE), chemical vapor deposition (CVD), or one of a variety of low-tech methods, such as chemical bath deposition (CBD) or electrodeposition. Compound electrodeposition has been well reviewed by a number of workers^1–5.

The Electrochemical Atomic Layer Epitaxy methodology was employed for the growth of cadmium and zinc chalcogenides on Ag(111). The different compounds were obtained by depositing alternate layers of chalcogen and metal at underpotential. The first layer, i.e. the layer deposited on the bare silver substrate, was the chalcogen. The principal features of S, Se and Te UPD layers on Ag(l 11) together with the experimental conditions for the attainment of the different chalcogenides of cadmium and zinc are described.

Whenever examined by chronocoulometric or voltammetric analyses, the chalcogen or metal UPD layers were found to grow with a two-dimensional mechanism. This mechanism is consistent with layer-by-layer growth which is a good prerequisite for epitaxial deposits. The amount of metal and chalcogen deposited in a given number of cycles was determined by stripping the deposits first anodically and then cathodically. It is a function of the number of cycles employed, again suggesting a layer-by-layer growth.

Table 1 summarizes the charges associated with each layer of metal and chalcogen for the different compounds. A ratio metal/chalcogen very close to unity is generally obtained.

The preparation and design of materials with controlled nanoarchitectures has emerged as an active research area which is of fundamental as well as technological importance. In particular materials with tailor-made pore sizes and shapes in the mesoporous domain (2–50 nm) have received considerable attention because of their potential value in applications such as shape-selective catalysis, molecular sieving, chemical sensing and selective adsorption^1–4.

In recent years monopolar electrodeposition in thin-layer cells (ECD) under constant electric fields has been the subject of renewed interest as a paradigmatic model for the study of pattern formation during growth^1–42. A closely related problem in bipolar electrochemistry under constant electric fields, known as Spatially Coupled Bipolar Electrochemistry (SCBE) has been recently introduced^43–46. In SCBE electrodissolution and electrodeposition in an applied electric field can be exploited to create directional growth of copper deposits between copper particles that are not connected to an external circuit. The study of monopolar and bipolar electrochemistry in thin-layer cells under alternating voltage conditions and its effects upon ion transport has very recently been addressed^47–48; electric field and ion transport constitute crucial factors in the characteristics of the connection.

The fabrication of arrays of nanometer scale particles has recently assumed enormous relevance in the study of the magnetic, optical and electronic properties of materials having dimensions comparable to the length scales where size quantization and surface phenomena determine the properties of interest. Large area arrays of identical particles in fact not only provide means for the study of the properties of nanoparticles by macroscopic methods, but also model systems for the implementation of devices exploiting these properties.

Bismuth telluride and its alloys are currently the best thermoelectric materials known at room temperature and are therefore used for portable solid-state refrigeration. If the thermoelectric figure of merit ZT could be improved by a factor of 3 or more, quiet and rugged solid-state devices could eventually replace conventional compressor-based cooling systems. In order to test the theoretical prediction that low dimensional materials could enhance ZT due to reduced thermal conductivity,¹ we are developing solution processing methods to make two-dimensional materials. Bismuth telluride and its p-type and n-type alloys have layered structures consisting of 5 atom thick Te-Bi-Te-Bi-Te sheets. Lithium ions are intercalated into the layered materials using liquid ammonia. Lithium intercalated Bi₂Te₃ has a higher conductivity and lower Seebeck coefficient than pristine Bi₂Te₃ likely due to electron transfer from the lithium. The intercalated materials can be exfoliated in water to form colloidal suspensions with relatively narrow particle size distributions. The layers are then deposited onto substrates, which after annealing at low temperatures, form highly c-axis oriented thin films. The low dimensional materials are characterized with powder X-ray diffraction, scanning electron microscopy, inductively coupled plasma and dynamic light scattering.

Interactions of surface-active agents with metals and semiconductors have attracted considerable attention of researchers in such fields as modified electrodes, photocatalysis, hydrometallurgy, metal corrosion, and mineral flotation. In the latter field, a special place has found ethyl xanthate (O-ethyl dithiocarbonate), whose interactions with various minerals have intrigued researchers from the early 1950’s [1]. The potassium ethyl xanthate (KEtX) has long been utilized as a collector in flotation of various minerals [1‐3], including chalcocite (Cu₂S), bornite (Cu₅FeS₄), chalcopyrite (CuFeS₂), and galena (PbS), and is now commonly used in the industrial recovery of metals from ores containing these minerals. The collecting efficiency of EtX^- ion is ascribed to its interaction with mineral surface, possibly through a chemisorption bond [3‐6]. Similar interactions of EtX- ions with various transition metals and Pb have been found [1,3‐7]. Self-assembling monolayers of EtX^- ions on silver, mercury, gold, platinum, and copper electrodes were investigated. For some metals, e.g. silver, EtX^- adsorption may cause a reverse piezogravimetric effect due to the hydrophilic/hydrophobic transitions on a surface with high roughness factor. Such an effect has recently been observed by Jeffrey and Woods [8] and incorrectly generalized to all adsorption systems. We have confirmed the appearance of this effect on a rough silver electrode using the EQCN technique and also, by measuring the high frequency admittance magnitude and phase shift using the quartz crystal immittance (QCI) technique.

The Grignard reagent, RMgX, where R is a hydrocarbon group and X is a halogen, is one of the more important and versatile reagents for organic synthesis [1]. It is formed in a heterogeneous reaction between magnesium and an organic halide in an appropriate organic solvent [2‐5]

The class of π-conjugated polymers as solid state materials has been investigated ranging from its use as electrically conducting polymers to electroluminescent (EL) and laser-active materials.¹ Synthesis using mainly aromatic and heteroaromatic polymer structures has been investigated.² In particular, the study of polyfluorenes and its derivatives are of recent interest as photoactive fluorescent and laser-generating materials.^{3, 18} The systematic synthesis of microstructured polyfluorene materials with fluoren-2,7-diyl units either by Pd⁴ or Ni-catalyzed ⁵ coupling of 2,7-dibromofluorenes has found applications in the fabrication of blue polymer light emitting diode (PLED) devices. The chemical oxidation of fluorenes with FeCl₃ ⁶ or electrochemical oxidation of fluorenes has also been reported.⁷ Substitution on the 9-fluorenyl position enhances the solubility and processability of the resulting polyfluorene polymers.⁸

Surface modification techniques are widely used to tailor a polymer’s interfacial properties such as permeability, wettability, adhesion, and biocompatibility, transforming inexpensive raw materials into highly valuable finished products.^1–3 These surface modification processes most commonly involve either wet chemical or vacuum based treatment strategies. Over the past decade, vacuum based methods, which include plasma based processing, ^4–8 noble gas sputtering and metallization⁹ have become the preferred pathway to modification. Compared to wet chemical processes, vacuum based technologies are fast, easily controlled and environmentally benign.¹⁰ Despite their technological importance, a molecular level understanding of vacuum based surface modification processes at polymeric interfaces is not well understood. This lack of mechanistic understanding can be ascribed to the complexity of the typical reaction medium coupled with the heterogeneity of polymeric substrates.

Layer-by-layer (LBL) self-assembly technique has been widely used to fabricate artificial thin films for a number of potential applications with several advantages (Decher, 1997; Lenahan et al., 1998; Kaschak et al., 1999). First, many materials with charges can be chosen as building blocks for this simple approach to thin-film fabrication; this includes conducting polymers, optical chromophores, metal complexes, or semiconductor nanoparticles. Second, LBL self-assembly processes are much simpler when compared to the Langmuir-Blodgett technique, or chemical/physical vapor depositions (Liu et al., 1997; Li et al., 1998; Lütt, et al., 1998; Dante et al., 1999; Kim et al., 1999; Liu et al., 1999; Ostrander et al., 2001). Finally, by controlling length scale at the nanometer level, LBL approach can be used to construct materials with designed properties for applications in electronics, photonics, and optoelectronics. Since the interfaces between each layer become the key linkages for multilayers and therefore interface properties are crucial to the self-assembled systems (Li et al., 1998; Schlenoff et al., 1998).

Chemically derivatized silica nanoparticles have been used in our laboratory for the study of the effects of restricted diffusion on the pyrolysis mechanisms of the organic molecules bound to the particle surface.¹ The principle undergirding these studies is that immobilization of radicals in controlled environments leads to different reaction mechanisms and product slates, relative to the pyrolysis of a substance in solution or the gas phase. The silica surface models some features of the pyrolysis of complex fossil fuels wherein diffusion limitations can play a large role and may lead to atypical reaction behavior.

An atomic-scale understanding of electronic, morphological and chemical structure of materials is a necessary prerequisite for tailoring nanostructured materials for catalytic applications. Scanning tunneling microscopy (STM) and spectroscopy (STS) are surface sensitive tools than that can be used to systematically probe morphological and electronic structure that can affect catalytic properties. STS, which can be used in tandem with STM, can give information on the densities of both filled and unfilled states at the nanometer scale by probing the local density of states (DOS) underneath the tip. This mapping is accomplished by varying the applied voltage and measuring the tunneling current while holding the tip at a constant position over an area of interest in the ST micrograph. A current-to-voltage (I-V) spectrum providing information regarding the chemical environment of a single atom can thus be produced. Figure 1 shows a diagram showing the band gap (E_g) between the conduction (E_c) and valence (E_v) band edges of metal clusters adsorbed onto a conductive support. Electrons (injected from the tip to the surface) occurs between the Fermi levels of the tip and sample, with electrons tunneling out of the more negative source. Fully metallic clusters exhibit no band gap (denoted by the length of the plateaus at the zero current); but an increase is observed with a decrease in size as the admetal clusters adopt a more non-metallic character. The length of the plateaus (in eV) is an effective band gap measurement of the supported adclusters.

Methanol is a promising fuel for direct methanol oxidation fuel cell technology. Traditionally, platinum group metals have been extensively used as catalysts for methanol oxidation fuel cells^1–4. Recent efforts are expanded to explorations of Pt-based bimetallic (e.g., Pt-Ru), multi-component ^5–8 and nanometer-sized metal particles as active catalysts ^9–13. The use of nanostructured catalysts promises to enhance our ability to design and control the catalytic activity and stability, and to cut down the catalyst cost as well. Two critical issues are however the propensity of poisoning at traditional platinum group catalysts by adsorbed CO-like intermediate species and the tendency of nanoparticle aggregation. The exploration of gold nanoparticles as catalysts is driven by recent discovery of high catalytic activity of nanosized gold towards hydrocarbons and recent theoretical findings of unique properties of gold ^9–10,¹². It is clear that to achieve highly active and stable gold catalysts for CO oxidation requires a fortunate combination of factors and conditions. Although there is growing experimental evidence showing the linkage of the catalytic activity to core sizes less than 5 nm, why, precisely, gold particles in such a restricted size range are catalytically active is not yet clear.

Thin films of hard amorphous carbon continue to gain importance in an increasing range of technologies and applications. Solid state ¹³C magic angle spinning (MAS) nuclear magnetic resonance (NMR) spectroscopy has proven to be a powerful tool in the investigation of the local structure in amorphous carbon films. The majority of these ¹³C MAS NMR investigations have focused on hydrogenated diamond-like carbon films.^1–13 Investigations of natural and synthetic diamonds,^14–17 as well as graphite and graphite intercalation compounds have also been reported.^{18, 19} Recently, more unusual carbon forms have been probed using ¹³C MAS NMR, including studies of pressure/temperature-treated C₆₀,²⁰ two-dimensional polymerized C₆₀ phases,²¹ and nanodiamonds produced during explosive compression.²²

The formation of octadecyltrichlorosilane (OTS) self-assembled monolayers (SAM) was investigated with contact angle measurements using water, hexadecane, and squalane as probe liquids. Results show that water and hexadecane contact angles of partial OTS SAMs do not provide reliable information about surface composition based on Cassie’s equation. However, squalane contact angles correlate linearly with surface coverages of partial OTS SAMs in the range of 0.4 and above. The behavior of contact angles as a function of surface coverages is discussed in terms of molecular interactions between probe liquid and surface OTS molecules.

Ultraviolet photoelectron spectroscopy (UPS) has been instrumental in developing modern molecular electronic structure theory. The energy difference between the neutral molecule ground state and the low-lying cationic states, measured as ionization energies by UPS, is the closest measure of orbital energies described by molecular orbital theory.¹ In addition, the ionizations independently provide information on electron configurations, charge potentials, bond strengths and other properties that relate to chemical reactivity in molecular systems.^{1, 2} Historically, high-quality photoelectron spectroscopy has been most informative for molecular species in the gas phase. However, many molecular systems of interest are not sufficiently stable or volatile for gas-phase UPS investigations. In an effort to provide high quality electronic structure data for such systems we are investigating methods for obtaining photoelectron spectra of molecules in thin films in which the surface-molecule and intermolecular interactions are relatively weak; thus approaching the gas phase limit. Previously we reported the ionizations of molecules containing metal-metal quadruple bonds³ and the first ever valence photoeleciron spectrum of C₆₀,^{4, 5} and found that the valence ionizations of these molecules in thin films closely resemble their gas-phase photoeleciron spectra. The differences (and similarities) between thin film and gas-phase data provide additional information about intermolecular interactions and electron relaxation in bulk materials. Here we present the UPS of ferrocene tethered to a gold surface via an alkanethiol chain. Ferrocene-terminated alkanethiols are known to form stable monolayers and have been characterized extensively.^{6, 7}

Self-assembled organic monolayers (SAMs) can be used to alter and control the chemical nature of surfaces. Self-assembly is simple, relatively low cost and widely applicable in areas such as lubrication, templating, optoelectronics and microelectromechanical systems (MEMS)¹. In addition, SAMs are potentially useful as base substrates for construction of model-biomembranes and protein attachment. For this purpose the monolayer should be very stable and ideally chemically bonded to the substrate.

The application of micro-electromechanical or micromechanical systems to the biomedical arena (BioMEMS) has tremendous potential in terms of developing new diagnostic and therapeutic modalities. Micro- and Nano-fabrication techniques are currently being used to develop implants that can record from, sense, stimulate, and deliver to biological systems. Micromachined neural prostheses, drug delivery micropumps/needles, microfabricated immunoisolation biocapsules, and retinal implants have all been fabricated using precision-based silicon technologies^1–6.

Increasing demands for more sophisticated and more biocompatible artificial medical devices in the recent years have stimulated researchers to either develop new technologies or adopt the established technologies from other scientific fields. Microfabrication technology (MEMS or micro-electrical-mechanical systems), mainly used in IC and microelectronics industry, has elicited increased interest among researchers due to the ability to control microstructure, topography, and feature size on the micro- and nano- scale; and control of surface chemistry in a precise manner through biochemical coupling and photopatterning processes. However, the applications of this technology are limited by the restricted choice of the materials (e.g. insulators, semiconductors and some metals) that can be used for micromachining. The most common materials used for biomedical applications are silicon and silicon dioxide. Though these materials are reasonably inert and biostable, several studies^1–6 on silicon implants have shown fibrosis and scar tissue formation, thus limiting the long term functioning of the microdevice. Hence, in order to develop efficient implantable microdevices with minimal chances of immunorejection, it is extremely desirable to develop surface modification strategies to augment the biocompatibility of silicon before using it for biomedical applications.

Liquid droplets residing on solid substrates are a familiar sight in daily life. For droplets smaller than the capillary length, which is on the order of 1mm for common liquids, the equilibrium shape on a macroscopic scale can be described as a spherical cap. This result is obtained by minimizing the total free energy of the droplet, as given by the sum of the interfacial energies (σ_lv: liquid-vapor, σ_sl: solid-liquid and σ_sv: solid-vapor interface). The equilibrium contact angle θ _Y is related to the interfacial energies by the well-known Young equation:

Microcantilevers comprise an emerging sensor platform.^1–24 The sensing mechanism is straightforward. Molecular adsorption on a resonating cantilever shifts its resonance frequency and changes its surface forces (surface stress).^7–9 Adsorption onto microcantilevers comprised of two chemically different surfaces results in a differential stress between the top and bottom surfaces of the cantilever and induces microcantilever bending.^10–12 The bending can be measured with angstrom resolution using optical reflection, piezoresistive, capacitance, and piezoelectric measurement methods commonly used in atomic force microscopy.