SiC and Ga2O3 are promising wide band gap semiconductors for applications in power electronics because of their high breakdown electric field and normally off operation. However, lack of a suitable dielectric material that can provide high interfacial quality remains a problem. This can potentially lead to high leakage current and conducting loss. In this work, we present a novel atomic layer deposition process to grow epitaxially MgxCa1−xO dielectric layers on 4H-SiC(0001) and β-Ga2O3$\left( {\bar 201} \right)$ substrates. By tuning the composition of MgxCa1−xO toward the substrate lattice constant, better interfacial epitaxy can be achieved. The interfacial and epitaxy qualities were investigated and confirmed by cross-sectional transmission electron microscopy and X-ray diffraction studies. Mg0.72Ca0.28O film showed the highest epitaxy quality on 4H-SiC(0001) because of its closest lattice match with the substrate. Meanwhile, highly textured Mg0.25Ca0.75O films can be grown on β-Ga2O3$\left( {\bar 201} \right)$ with a preferred orientation of (111).

We deposit films of tin–calcium sulfide by atomic layer deposition (ALD) and demonstrate the metastability of this material. Rough and spiky films are obtained by using Sn and Ca precursors with different ligands, whereas compact and smooth films are obtained when the two metal sources share the same ligands. Compositional and quartz crystal microbalance results indicate that part of the underlaying SnS film is replaced and/or removed during the CaS ALD cycle during the ternary film deposition, possibly via a temperature-dependent cation exchange mechanism. The crystal structure transforms from orthorhombic to cubic as the calcium content increases. Furthermore, resistivity increases with calcium content in the alloy films, whereas optical band gap only depends weakly on Ca content. After annealing at 400 °C in an H2S environment, the cubic alloy film undergoes a phase transition into the orthorhombic phase and its resistivity also decreases. Both phenomena could be explained by phase separation of the metastable alloy.

We use cyclic charge-discharge experiments to evaluate the capacity retention rates of two quinone-bromide flow batteries (QBFBs). These aqueous QBFBs use a negative electrolyte containing either anthraquinone-2,7-disulfonic acid (AQDS) or anthraquinone-2-sulfonic acid (AQS) dissolved in sulfuric acid, and a positive electrolyte containing bromine and hydrobromic acid. We find that the AQS cell exhibits a significantly lower capacity retention rate than the AQDS cell. The observed AQS capacity fade is corroborated by NMR evidence that suggests the formation of hydroxylated products in the electrolyte in place of AQS. We further cycle the AQDS cell and observe a capacity fade rate extrapolating to 30% loss of active species after 5000 cycles. After about 180 cycles, bromine crossover leads to sufficient electrolyte imbalance to accelerate the capacity fade rate, indicating that the actual realization of long cycle life will require bromine rebalancing or a membrane less permeable than Nafion® to molecular bromine.

The growth of of metallic copper by atomic layer deposition (ALD) using copper(I) di-sec-butylacetamidinate ([Cu(sBu-amd)]2) and molecular hydrogen (H2) on SiO2/Si surfaces has been studied. The mechanisms for the initial surface reaction and chemical bonding evolutions with each ALD cycle are inferred from in situ Fourier transform infrared spectroscopy (FTIR) data. Spectroscopic evidence for Cu agglomeration on SiO2 is presented involving the intensity variations of the SiO2 LO/TO phonon modes after chemical reaction with the Cu precursor and after the H2 precursor cycle. These intensity variations are observed over the first 20 ALD cycles at 185°C.

A new process was developed for deposition of the silicates and oxides of metals such as zirconium and hafnium at low substrate temperatures (100–300°C). The silicon and oxygen source is tris(tert-butoxy)silanol, (tBuO)3SiOH, and the metal precursors are metal amides. A typical reaction is

ZrL4 + 2 (tBuO)3SiOH → ZrSi2O6 + 4HL + 6H2C=C(CH3)2 + 2 H2O

in which the ligand L is ethylmethylamide, -NEtMe. The precursor vapors were alternately pulsed into a heated reactor, yielding about 0.3 to 0.7 nm of metal silicate film for each cycle. Replacing the silanol pulses with water pulses yields pure metal oxides with a thickness of about 0.1 to 0.15 nm per cycle. The silicon content of the films can be adjusted to any desired value by replacing some of the silanol pulses by water pulses. This new process has a number of advantages over previous methods for depositing metal silicates. Uniformity of thickness and stoichiometry are readily achieved. The deposition atmosphere is non-oxidizing, so that formation of low-k interfacial oxides between the deposited layer and a silicon substrate is minimized. The new halogen-free precursors avoid halogen contamination of films and corrosion of deposition systems. This process is a promising method for forming the next generation of ultra-thin high-k gate dielectrics in silicon-based microelectronics.

A new process was developed for deposition of the tungsten nitride at moderate substrate temperatures (350-400 °C). The tungsten source bis (tert-butylamido)bis(tertbutylimido) tungsten, (tBuNH)2(tBuN)2W, reacts with ammonia in a low-pressure CVD reactor. Growth rates ranged from about 0.6 to 4.1 nm per minute. The stoichiometry of the films varied from WN1.2O1.7 to WN1.6o0.25, depending mainly on deposition temperature. The films are amorphous by X-ray diffraction, and smooth by scanning electron microscopy. Step coverage is nearly 100% in vias with an aspect ratio of 6:1 for films deposited at 400 °C or lower. Barriers 45 nm thick resist diffusion of copper up to temperatures of 600 °C. Adhesion is strong to all substrates tested, including silicon, silicon dioxide, soda-lime glass, glassy carbon, aluminum and stainless steel. This new halogen-free process avoids halogen contamination of films and corrosion of equipment. Uniformity of thickness and stoichiometry are readily achieved. This process is a promising method for forming copper diffusion barriers in future generations of microelectronics.

Transparent, electrically conductive films have been prepared from a wide variety of materials. These include semiconducting oxides of tin, indium, zinc, and cadmium, and metals such as silver, gold, and titanium nitride. In this article, the physical properties of these materials are reviewed and compared.

A new CVD process is described for depositing conformal layers containing tungsten, tungsten nitride or tungsten oxide. A film of tungsten metal is deposited by vaporizing liquid tungsten(0) pentacarbonyl 1-methylbutylisonitrile and passing the vapors over a surface heated to 400 to 500 °C. This process can be used to form gate electrodes compatible with ultrathin dielectric layers. Tungsten nitride films are deposited by combining ammonia gas with this tungsten-containing vapor and using substrates at temperatures of 250 to 400 °C. Tungsten nitride can act as a barrier to diffusion of copper in microelectronic circuits. Tungsten oxide films are deposited by adding oxygen gas to the tungsten-containing vapor and using substrates at temperatures of 200 to 300 °C. These tungsten oxide films can be used as part of electrochromic windows, mirrors or displays. Physical properties of several related liquid tungsten compounds are described. These low-viscosity liquids are stable to air and water. These new compounds have a number of advantages over tungsten-containing CVD precursors used previously.

New precursors were developed for the chemical vapor deposition (CVD) of aluminum nitride (AIN) and gallium nitride (GaN) at low temperatures. Synthetic methods for the new materials will be reported, along with their analyses and spectral characterization. The precursors are volatile (180 mTorr at 45-55 °C), low-viscosity (10 centipoise) liquids, so they are more convenient as vapor sources than previously available solid sources. They are thermally stable to temperatures well above their vaporization temperatures, so their vaporization is highly reproducible and leaves no residue. Unlike previously available liquid precursors, they are not pyrophoric, so they are safer to handle. AMN films formed by reaction with ammonia at around 200 °C are amorphous, transparent insulators that are good barriers to diffusion of water, oxygen, and other materials.

The first volatile, liquid compounds of alkali metals were synthesized and used for the CVD of materials containing alkali metals. Amides of the type MNR1(SiMe2R2) and MN(SiMe2R2)2 [M = Li, Na, K; R1 = t-butyl, t-amyl; R2= ethyl, n-propyl, i-propyl, n-butyl, i-butyl, n-hexyl, n-octyl] were made and characterized. The lithium amides were prepared via the deprotonation of the parent amine using butyl lithium. The sodium and potassium amides were formed by transamination of sodium amide and potassium bis(trimethylsilyl)amide with the parent amines. For example, lithium bis(ethyldimethylsilyl)amide was prepared from butyl lithium and bis(ethyldimethylsilyl)amine and was distilled as a clear, colorless liquid at 122 °C (0.2 Torr) having a viscosity of 37 cP at 40 °C. These alkali metal amides can be used as convenient liquid sources for CVD of mixed metal oxides containing alkali metals, such as the non-linear optical material lithium niobate, lithium-containing materials for battery electrodes, electrochromic tungsten bronzes, and the pyroelectric and ferroelectric material potassium tantalate.

The first room-temperature liquid compounds useful for the CVD of alkaline earth metalcontaining oxides were prepared by reacting metal (Mg, Ca, Sr, and Ba) beta-diketonates with novel polyamine ligands. The compounds are monomeric and can be completely flash-vaporized without leaving any non-volatile residue detectable at the parts-per-million level. A stable, solvent-free liquid mixture was formed by mixing new liquid barium, strontium and titanium compounds. CVD experiments using direct liquid injection of this liquid mixture deposited films of barium strontium titanate. This approach should also be applicable to the deposition of many other multicomponent oxides containing alkaline earth metals: ferroelectrics (strontium bismuth tantalate), metallic conductors (strontium vanadium oxide, lanthanum strontium cobalt oxide), phosphors (calcium tungstate), non-linear optical materials (beta-barium borate), magnetic oxides (barium ferrite), colossal magnetoresistive materials (lanthanum strontium manganese oxide), high Tc superconductors (yttrium barium copper oxide, bismuth calcium strontium copper oxide) and microwave dielectrics (barium magnesium tantalate).

We demonstrate a novel approach to improving the step coverage of thin films made by chemical vapor deposition (CVD). Titanium nitride (TiN) films were deposited by atmospheric pressure CVD using tetrakis(diethylamido)titanium vapor (TDEAT) and ammonia gas (NH3) carried in nitrogen gas. Trimethylamine (NMe3) gas was added during some of the depositions. The substrates were patterned silicon wafers having holes with aspect ratio of 3.5 through a silicon dioxide layer. We discovered that the step coverage was significantly increased for TiN films made with NMe3. At 320 °C, the step coverage was increased from 70% to nearly 100%. Within the range of deposition temperatures used in our study, 320 °C to 370 °C, the amount of improvement increased as the deposition temperature decreased. The trimethylamine did not increase the resistivity or the impurity levels in the films, but it did reduce the growth rate slightly. We suggest that the trimethylamine adsorbs onto the surface, temporarily blocking some of the sites on which growth could take place. Thus the effective sticking coefficients for the precursors are decreased, and the step coverage is increased.

New precursors have been found for chemical vapor deposition (CVD) of many metal oxides. Each precursor is a mixture formed by randomly attaching a selected set of organic groups, such as the isomers of the butyl group, to a metal 2,4-pentanedionate (also known as acetylacetonate) in place of the methyl groups of the 2,4-pentanedionate ligand. Most of these new mixed metal beta-diketonates are liquids at room temperature, whereas the corresponding metal 2,4-pentanedionates are solids. In the cases where they were solids or viscous liquids, small amounts of organic solvents were added to reduce the viscosity. We have so far prepared mixed beta-diketonate precursors for barium, strontium, calcium, magnesium, aluminum, indium, tin, lead, bismuth, titanium, zirconium, vanadium, niobium, tantalum, chromium, molybdenum, manganese, iron, ruthenium, cobalt, nickel, copper, zinc, yttrium, lanthanum and cerium.

Liquid sources are much more convenient for CVD than solid sources. These liquid mixtures or solutions were vaporized by ultrasonically nebulizing the liquid into a flow of hot nitrogen carrier gas preheated to 150–250 °C. These vapor mixtures were mixed with air or oxygen and flowed over substrates heated typically to 350–450 °C. Films of the corresponding metal oxide (or carbonate, in the case of barium, strontium and calcium) were deposited on substrates of silicon or glass. Gas pressures from 20–760 Torr were used.

Because a common set of ligands is used for each of these metal precursors, they can be mixed as liquids or vapors without any precipitation due to ligand exchange reactions. To demonstrate their use in forming mixed metal oxides, we have prepared films of ferroelectric barium titanate. This method should be applicable to other mixed metal oxides of current interest, such as high dielectric constant strontium titanate, ferroelectric bismuth strontium tantalate, superconducting yttrium barium copper oxide, refractory yttrium zirconium oxide, second-harmonic generating barium borate, metallic lanthanum strontium cobalt oxide and magnetoresistive lanthanum strontium manganate.

Near stoichiometric titanium nitride (TiN) was deposited from tetrakis(dimethylamido)titanium (TDMAT) and ammonia using atmospheric pressure chemical vapor deposition. Experiments were conducted in a belt furnace; static experiments provided kinetic data and continuous operation uniformly coated 150-mm substrates. Growth rate, stoichiometry, and resistivity are examined as functions of deposition temperature (190−420 °C), ammonia flow relative to TDMAT (0−30), and total gas-flow rate (residence time 0.3−0.6 s). Films were characterized by sheet resistance measurements, Rutherford Backscattering Spectrometry, and X-Ray Photoelectron Spectrometry. Films deposited without ammonia were substoichiometric (N/Ti, 0.6−0.75), contained high levels of carbon (C/Ti = 0.25−0.40) and oxygen (O/Ti = 0.6−0.9), and grew slowly. Small amounts of ammonia (NH3/TDMAT ≥ 1) brought impurity levels down to C/Ti, 0.1 and O/Ti = 0.3−0.5. Ammonia increased the growth rates by a factor of 4−12 at temperatures below 400 °C. Films 500 Å thick had resistivities as low as 1600 μΩ-cm when deposited at 280 °C and 1500 μΩ-cm when deposited at 370 °C. Scanning electron micrographs indicate a smooth surface and poor step coverage for films deposited with high ammonia concentrations.

Films of amorphous aluminum oxide were deposited from vaporized alkylaluminum alkoxide compounds and oxygen. For example, a gaseous mixture of triethyldialuminum tri-sec-butoxide vapor and dry air deposits amorphous aluminum oxide films on substrates heated to temperatures around 400 °C. Transparent, smooth, adherent films of aluminum oxide were formed on silicon, glass and metal substrates. The new precursors display a number of advantages over previously used sources for aluminum oxide. They are non-pyrophoric, low-viscosity, low-cost liquids. High deposition rates, over 0.2 μm/min, were observed. The coatings have high purity, high electrical resistivity and high transparency to light. They are excellent barriers to diffusion of water and of ions, such as sodium. These films are useful in optical coatings, as wear-resistant hard coatings, and as diffusion barriers that protect flat panel displays, computer microcircuits, solar cells and metals from corrosion and degradation by impurities.

Transparent, electrically conductive films have been prepared from several different metal oxides, including those of tin, indium and zinc. Deposition methods for these materials are reviewed, and their properties summarized and compared. A figure of merit for a transparent conductor may be defined as the ratio of the electrical conductivity to the optical absorption coefficient of the film. The figure of merit for fluorine-doped zinc oxide is shown to be larger than that of other transparent conductors, such as boron-doped zinc oxide, fluorine-doped tin oxide, and tin-doped indium oxide. Physical, chemical and thermal durability, deposition temperature, and cost are other factors which may also influence the choice of material for a particular application.

Titanium nitride films were formed by chemical vapor deposition from titanium tetrabromide and ammonia at atmospheric pressure and substrate temperatures from about 400 to 600 °C. Although titanium tetrabromide is a highly hygroscopic solid at room temperature, it can be handled conveniently as a very concentrated liquid solution (85 weight percent) in bromine as a solvent. This solution can be vaporized by a direct liquid injection system. Alternatively, the solution can be pumped into a bubbler, from which the bromine solvent is then removed by fractional distillation. Rutherford Backscattering Spectroscopy was used to determine that the bromine content of TiN deposited at 500°C was about one atomic per cent. Growth rates were about 17 nm/min, and electrical resistivity was found to be about 200 μΩ-cm.

Titanium nitride (TiN) films serve an important function as an adhesion layer and diffusion barrier in contact and via plug structures of ULSI devices. In this work, TiN films were deposited by chemical vapor deposition (CVD) at atmospheric pressure using the tetrakis (dimethylamido) titanium (TDMAT) and tetrakis (diethylamido) titanium (TDEAT) organic precursors over a range of temperatures and ammonia partial pressures. The step coverage and morphology of these films was evaluated at these various conditions by SEM of cross-sectioned contacts. In addition, material properties such as resistivity. surface roughness, and stoichiometry were assessed at some of the operating conditions.

Hydropyridine is introduced as a new ligand for use in constructing precursors for chemical vapor deposition. Detachment of hydropyridine occurs by a low-temperature reaction leaving hydrogen in place of the hydropyridine, and a very stable byproduct, pyridine vapor. Hydropyridine ligands can be attached to a variety of elements, including main group metals, such as aluminum and antimony, transition metals, such as titanium and tantalum, semiconductors such as silicon, and nonmetals such as phosphorus and arsenic.

Use of metal-organic precursor materials has permitted thermal CVD of many metal nitrides at remarkably low temperatures and without corrosive by-products. Metallic TiN, VN, Nb3N4 and Mo2N3; semiconducting GaN and Sn3N4; and insulating AIN, Zr3N4, Hf3N4 and Ta3N5 can be deposited at temperatures typically in the range 100 to 400 C. Deposits free of carbon are obtained by transamination reactions of metal dialkylamido precursors with a sufficiently large excess of ammonia. The resulting TiN films are good diffusion barriers, and provide low contact resistance between metals and silicon.

Transparent semiconducting oxide films, such as SnO2, ZnO and TiO2, are often made by MOCVD for use in solar cells, energy-efficient window coatings and electro-optical displays. These wide band-gap semiconductors can be doped to n-type conductivity by a variety of dopant elements. Fluorine is the dopant which produces materials with the highest electron mobility, conductivity and transparency, by substituting for oxygen. Certain organic fluorine compounds have been found to be very effective fluorine dopants for these CVD reactions, yielding films with very shallow donors having nearly 100 % electrical activity.

Precursors for the CVD of alkaline earth metal oxides, particularly barium, lack the volatility and stability needed for reproducible deposition of superconducting, ferroelectric or magnetic oxides. Use of ammonia or volatile amines as carrier gases greatly enhances the volatility and stability of betadiketonates of barium, strontium and calcium, providing high and stable transport rates for source temperatures below 100 C, even for non-fluorinated ligands.