Corrosion Mitigation Coatings for RF Sources and Components

Accelerator facility upgrades, new accelerator applications, and future design efforts are leading to novel klystron and IOT device concepts, including multiple beam, high-order mode operation, and new geometry configurations of old concepts. At the same time, a new simulation capability, based upon finite-difference “cut-cell” boundaries, has emerged and is transforming the existing modeling and design capability with unparalleled realism, greater flexibility, and improved accuracy. This same new technology can also be brought to bear on a difficult-to-study aspect of the energy recovery linac (ERL), namely the accurate modeling of the exit beam, and design of the beam dump for optimum energy efficiency. We have developed new capability for design calculations and modeling of a broad class of devices which convert bunched beam kinetic energy to RF energy, including RF sources, as for example, klystrons, gyro-klystrons, IOT's, TWT’s, and other devices in which space-charge effects are important. Recent advances in geometry representation now permits very accurate representation of the curved metallic surfaces common to RF sources, resulting in unprecedented simulation accuracy. In the Phase I work, we evaluated and demonstrated the capabilities of the new geometry representation technology as applied to modeling and design of output cavity components of klystron, IOT's, and energy recovery srf cavities. We identified and prioritized which aspects of the design study process to pursue and improve in Phase II. The development and use of the new accurate geometry modeling technology on RF sources for DOE accelerators will help spark a new generational modeling and design capability, free from many of the constraints and inaccuracy associated with the previous generation of “stair-step” geometry modeling tools. This new capability is ultimately expected to impact all fields with high power RF sources, including DOE fusion research, communications, radar and other defense applications.

All solid-state ion layers has an ever-growing interest in order to accomplish on-chip energy storage devices. As such, lithium-ion synaptic transistor to build artificial neural networks, integrated area-enhanced capacitors and lithium-ion microbatteries for energy supply are suitable candidates fulfilling energy-efficient devices. Atomic layer deposition (ALD) is one of the most promising techniques for manufacturing ultrathin layers for lithium-ion devices with exceptional capabilities to enable conformal growth at atomic scale. In this area, in situ diagnostics has become of vital importance for the growth process understanding especially on view of the air-sensitivity character of lithium ultra-thin layers.The purpose of this work is to present an in situ spectroscopic ellipsometry (SE) study of ALD thin films used on the fabrication of lithium-ion devices. We will consider several aspects of ALD ultrathin films growth, such as film thickness, saturation curves and initial films nucleation. We shall compare in situ SE results with those of ex situ diagnostics such as XRR (x-ray reflectometry) and XPS (x-ray photoelectron spectroscopy).We focus on the characterization of 10-100 nm thin lithium phosphate electrolyte with and without nitrogen doping deposited at ALD temperature range 300-350°C using LiHMDS-TMPO and LiHMDS-DEPA precursors, respectively. On the other hand, the realization of silicon integrated lithium ionic devices requires the implementation of essential barrier films to prevent Li-ion into the Si substrate. We will explore here the ability of titanium nitride depositions to act as lithium-ion diffusion barriers. TiN deposition (10-20 nm thin) by ALD process (350°C-400°C) uses TiCl4 and NH3 precursors.First, we investigate the growth process of LiPO, LiPON and TiN intrinsic layers deposited using thermal ALD. Secondly, we apply in situ and ex situ diagnostics on exemplary LiPO/LiPON deposited on top of the 10 nm ALD TiN film (Figure 1: Monitoring of LiPON thickness vs deposition time using in-situ Ellipsometry). The film thickness deduced from in situ SE measurements was corroborated by thickness measurements with XRR technique (Figure 2: Evolution of LiPON thickness vs ALD cycles using Ellipsometry and SE characterization techniques). Ex situ SE was used most often for WiW uniformity evaluating (Figure 3: 8” Wafer thickness mapping of ALD LiPON layers (a: 10nm TiN film; b: 17nm LiPON film). The preliminary results underline the fact that ALD growth rate and nucleation of lithium electrolyte and titanium nitride obviously depends on the ALD process parameters (temperature, pulse and purge times, number of cycles) but on the substrate nature (silicon, thermal oxide, and metallic). In situ and ex situ diagnostics highlight that initial film growth of lithium electrolytes and TiN layers strongly depends on the precursors kind used in the ALD process and the substrate material employed. Figure 1

Review: 2454 refs.

Atomic layer deposition (ALD) processes for the growth of ZrO2 and TiO2 were developed using novel precursors. The novel processes were based on cycloheptatrienyl (CHT, -C7H7) – cyclopentadienyl (Cp, -C5H5) compounds of Zr and Ti, offering improved thermal stability and purity of the deposited oxide films. The CpMeZrCHT/O3 ALD process yielded high growth rate (0.7–0.8 Å/cycle) over a wide growth temperature range (300–450 °C) and diminutive impurity levels in the deposited polycrystalline films. Growth temperatures exceeding 400 °C caused partial decomposition of the precursor. Low capacitance equivalent thickness (0.8 nm) with low leakage current density was achieved. In the case of Ti, the novel precursor, namely CpTiCHT, together with ozone as the oxygen source yielded films with low impurity levels and a strong tendency to form the desired rutile phase upon annealing at rather low temperatures. In addition, the thermal stability of the CpTiCHT precursor is higher compared to the usually applied ALD precursors of Ti. The introduction of this new ALD precursor family offers a basis for further improving the ALD processes of group 4 oxide containing thin films for a wide range of applications.

Operating pressure influences over micro trenches in exposure time introduced atomic layer deposition

Atomic layer deposition (ALD) of lithium (Li)-based thin films has aroused significant interest in recent years. Promising applications are Li-ion thin-film batteries (TFBs), protective particle coatings, interface model systems, and neuromorphic computing [1,2]. Here, we demonstrate high footprint capacities and excellent high-rate performance of Li4Ti5O12 (LTO) fabricated by ALD for three-dimensional (3D) solid-state TFBs to power upcoming energy-autonomous sensor systems.The simultaneous increase of power and energy density of full-cell 3D TFBs by coating the battery layer stack over microstructured substrates was recently demonstrated [3]. The required conformal, pinhole-free deposition, and stoichiometric control of nanometer-thin films on highly structured surfaces are only accessible via ALD. The vapor-phase technique based on sequential, self-limiting surface reactions is well understood. However, the direct deposition of Li-based anodes remains challenging [1].In previous studies, we developed a thermal three-step ALD process for Li-containing mixed oxides on 200 mm silicon wafers [4, 5]. Lithium-tert-butoxide (LTB) and lithium hexamethyldisilazide (LiHMDS) were proven as suitable precursors, forming high-quality spinel LTO with low impurities after rapid thermal processing. The excellent electrochemical behavior of ALD LTO with LiHMDS was examined and linked to the film texture [5].In this work, we optimize the LTO ALD process with LTB towards high-capacity 3D TFBs and evaluate the electrochemical performance for the first time. A process with tetrakis (dimethylamino) titanium (TDMAT) and H2O at 300 °C was developed, leading to a saturated growth per cycle of 1.06 Å cycle-1 for 7 s LTB pulses. An ALD temperature window between 240 and 320 °C was identified. The effect of the substrate on the initial growth and crystallization behavior was investigated. To overcome the crystallization-inhibiting effect of the titanium nitride current collector, we introduced an ultrathin AlOx interlayer.Planar ALD LTO films with various thicknesses of up to 75 nm were manufactured, increasing the footprint capacity from 1.5 to 4.3 μAh cm-2. We observed a higher surface capacity contribution due to an increased roughness and an inferior C-rate performance with increasing film thickness. 50 nm LTO films demonstrate the optimum energy and power density. A footprint capacity of 1.95 μAh cm-2, corresponding to 67 % of the initial capacity, was achieved at 50 C with an average Coloumb efficiency of 99.9975 %.Next, we evaluate ALD LTO films as 3D TFB anodes for the first time. The conformality of the ALD process on microstructured substrates was improved by extending the LTB pulse and purge times. The purge time is the key factor enabling a step coverage of 70 % for holes with an aspect ratio of 10:1. The 3D samples with an area-enhancement factor (AEF) of 9 coated with 50 nm LTO obtained a footprint capacity of 20.2 μAh cm-2 at 1 C. The significant capacity enhancement of 6.9 compared to planar samples is according to the conformality. The remarkable high-power capability of 3D LTO is demonstrated with 7.75 μAh cm-2 at extreme currents of 5 mA cm-2. A capacity retention of 97.4 % after 500 cycles at 1 mA cm-2 shows the high-rate cyclability. The superior high-rate performance of ALD LTO compared to other 3D anode materials illustrates the enormous potential for realizing high-energy and high-power on-chip 3D TFBs.

Our laboratory is developing new chemical precursors for the growth of electropositive metal and element thin films by atomic layer deposition (ALD). We are also interested in processes that exhibit area selective growth. ALD has many current applications in copper metallization, diffusion barriers, liners, and transistor fabrication. Thermal ALD is often preferred because plasmas can afford low conformal coverage due to radical recombination on the walls of deep and narrow features. There has been extensive progress in the thermal ALD of copper and noble metal films in recent years, because the positive electrochemical potentials allow relatively easy reduction of precursor ions to the metals. Thermal ALD approaches to most other metals and elements in the periodic table are not well developed, due to the negative electrochemical potentials of the ions and a current lack of ALD co-reagents that can convert the ions to the metals or elements. Herein, we will describe the thermal ALD growth of electropositive metals such as nickel, cobalt, aluminium, and others. The ALD of nickel and cobalt metal films has been achieved using precursors containing diazadienyl ligands (precursors 1 and 2). These precursors enable the deposition of cobalt and nickel metal films at temperatures below 200 °C and use alkylamines as benign co-reagents. Growth rates are high (0.60 Å/cycle for nickel, 0.98 Å/cycle for cobalt), high purity, low resistivity metal films are obtained, and the films have low rms roughnesses. The processes exhibit inherent selective growth on metal substrates such as platinum, ruthenium, and copper. By contrast, no growth is observed on insulating substrates. We will also overview a new thermal ALD process for the growth of aluminum metal films. This process entails treatment of surface-bound AlCl3 with a thermally stable, volatile aluminium hydride co-reagent. The growth rate for the aluminium metal ALD process is high, and high purity, low resistivity aluminium metal films are obtained. Prospects for the area selective growth of aluminum metal films will be presented. These examples demonstrate that ALD processes can be enabled for electropositive metals through careful design of precursors and chemistry. Moreover, many of the metal ALD processes exhibit area selective growth. Figure 1

This paper presents the results of a research program set up to evaluate atomic layer deposition (ALD) conformal coatings as a method of mitigating the growth of tin whiskers from printed circuit board assemblies. The effect of ALD coating process variables on the ability of the coating to mitigate whisker growth were evaluated. Scanning electron microscopy and optical microscopy were used to evaluate both the size and distribution of tin whiskers and the coating/whisker interactions. Results show that the ALD process can achieve significant reductions in whisker growth and thus offers considerable potential as a reworkable whisker mitigation strategy. The effect of ALD layer thickness on whisker formation was also investigated. Studies indicate that thermal exposure during ALD processing may contribute significantly to the observed whisker mitigation.

During the past decade, atomic layer deposition (ALD) has become an important thin-film deposition method in microelectronics industry, and it has also gained a lot of interest in many other areas, such as nanotechnology. The success of ALD is built on proper surface reactions. In this paper, in situ reaction mechanism studies on ALD processes are reviewed with the aim of building a general understanding on similarities and differences exhibited by various processes and process groups. Also, levels of understanding reaction mechanisms in ALD are discussed. The main methods used to study ALD chemistry in situ under typical process conditions are quadrupole mass spectrometry (QMS), quartz crystal microbalance (QCM), and infrared (IR) spectrometry. These are presented in detail in the review. Various other optical methods, ellipsometry in particular, have been used to study ALD processes too, but they provide little information about the reaction mechanisms. Competent in situ investigations solve the ALD reaction mechanism as balanced equations for the ALD half-reactions. The majority of ALD processes are exploiting ligand exchange reactions, where the ligands of the metal precursor are eliminated by bonding to the Lewis acids of the nonmetal precursors, most commonly hydrogen. These volatile byproducts are usually released during both precursor pulses, and determining their relative amounts is the important task in reaction mechanism studies of such processes. These processes are mechanistically fairly well understood, though some of these also display side-reactions to the ligand exchange reactions. There are also whole groups of processes that are using chemistry almost entirely different from the ligand exchange reactions. The most important such processes involve combustion chemistry, with oxygen, oxygen plasma, or ozone as a precursor or co-reactant. The mechanisms of these processes are complicated and less understood compared with the mechanisms of the ligand exchange reaction processes.

Due to various applications in electronic and optical fields, considerable efforts have been made to develop ZnO based semiconducting materials. ZnO has a wide and direct band gap of 3.37 eV, large exciton binding energy (60 meV), high mechanical and thermal stabilities, and good electrical conductivity. Therefore, ZnO thin films are intensively investigated and applied in solar cells, piezoelectrical devices, gas and biological sensors, catalysts, and high-efficiency short-wavelength optoelectronic nanodevices. To further enhance the properties of ZnO, researchers have been synthesizing ZnO films doped with aluminum or indium. Al doped ZnO (AZO) thin films have been highly developed featuring non-toxicity, low material cost, good thermal stability, high optical transmittance, and good electrical conductivity. [1, 2, 3] Many techniques have been used to deposit AZO thin films currently, such as sol gel processing, magnetron sputtering, pulsed laser deposition (PLD), chemical vapor deposition (CVD), and atomic layer deposition (ALD). Among these film deposition methods, ALD has been extensively investigated for the deposition of thin films of metals, alloys, semiconductors, and oxides. Compared with other techniques, ALD is considered as a novel technique for the deposition of AZO films due to its characteristic feature of self-limiting surface saturating reactions in each ALD cycle. ALD technology consequently and accurately controls the layer thickness with unprecedented atomic level precision, as well as the stoichiometry, composition, uniformity of large areas and maintains excellent conformity over complex surface nanostructures. The other advantage of ALD technology is to introduce each ALD chemical precursor sequentially separated by time to react on the surface of the structure. The concentration of dopant in the ZnO films can be precisely controlled by the number of ALD cycles, so that the physical properties of AZO films, such as carrier density and conductivity, can be adjusted to meet the requirements of different applications. In this study, a Savannah 100 cross-flow thermal ALD reactor was used for the deposition of Al doped ZnO thin films. The precursors of aluminium, zinc and oxygen used in the ALD process for the synthesis of Al doped ZnO thin films were trimethylaluminium (TMA: Al2(CH3)6), diethylzinc (DEZ: (C2H5)2Zn) and DI water. The growth temperature of AZO thin films was maintained at 250 oC. The flow rate of N2, used as a carrier gas, was accurately controlled at 20 sccm. In order to deposit ALD AZO films with various percentages of aluminium, TMA cycles were introduced once in every 30 pulses of DEZ and H2O injection. After deposition, AZO films were annealed in air at 500 oC, 600 oC, and 700 oC for 30 min, respectively. X-ray diffraction (XRD) was performed to characterize the crystal structure of ALD ZnO and AZO films. Field emission scanning electron microscopy (FE-SEM) was used to inspect the morphology of AZO films before and after annealing. The surface roughness was analyzed by atomic force microscopy (AFM). The compositions of AZO thin films were analyzed by energy dispersive X-ray spectroscopy (EDS). The sheet resistivity of AZO films was determined by using four-point probe measurements. Raman spectroscopy and photoluminescence (PL) were used to investigate the properties and detect structures of ALD AZO films. The band gaps of AZO films prepared by the ALD process were analyzed from PL spectroscopy results. The materials property and defect structure of ALD AZO films were investigated by Raman spectroscopy.

A liquid injection atomic layer deposition (ALD) process has been developed for conformal deposition of lead zirconate titanate thin films on three dimensional structured silicon substrates. Future applications of these structures address the field of nonvolatile ferroelectric memory devices. PZT films were prepared by mixing the binary ALD processes of PbOx, TiOx, and ZrOx. It was found that the type of precursors as well as the stacking sequence of the binary oxide layers is crucial for the control of film stoichiometry and for a saturation of the film growth behaviour characteristic for ALD processes. Using an optimized ALD process the molar ratios of Pb/(Ti+Zr) as well as Zr/(Ti+Zr) in the films show a saturation behaviour against an increasing precursor injection volume. Pb(ZrxTi1-x)O3 films deposited on 3D structured silicon exhibited a uniform film thickness and a homogeneous cation composition over the complex structure.

Atomic layer deposition (ALD) processes were demonstrated for lead-titanate (PbTiO3) and lead-zirconium-titanate (PZT) films. The metal organic precursors were tetraethyl lead Pb(Et)4 reacted with ozone; along with tetrakis ethylmethylamino zirconium and tetrakis dimethylamino titanium reacted with either water or ozone. These precursors were selected based on compatibility with ALD processes for the component oxides. Single oxide films of PbO, ZrO2, and TiO2 were easily deposited using the selected precursors. ALD of the complex oxide films was done by combining the ALD processes for the component oxide films. The compositions of Pb, Zr, and Ti in the films could be predictably controlled by the relative ratio of Pb, Zr, and Ti precursor doses during the ALD process, and controlled composition for both PbTiO3 and PZT films was demonstrated.

We fabricated nanosheet oxide semiconductor FETs with InGaZnO x (IGZO) channel grown by thermal atomic layer deposition (ALD) process. The effects of annealing, channel thickness, and composition on performance and reliability were systematically investigated. IGZO FETs achieved normally-off operation with positive threshold voltage (V th) while maintaining high mobility compared to InGaO x (IGO) FETs. The devices showed thermal stability up to 400 °C post-deposition annealing, compatible with BEOL process. Excess oxygen is a unique issue in ALD by oxygen-rich chemical reaction process compared to sputtering. The positive bias stress reliability issue in thin channel IGO FETs caused by excess oxygen was mitigated by adding Zn. This study can provide an insight for designing IGO-based FETs in terms of composition and channel thickness optimization for 3D LSI applications.

NiO thin films have attracted tremendous attention owing to their outstanding chemical stability, good magnetic and catalytic properties1. It is also considered an efficient electrocatalyst for water oxidation, and is one of the few p-type semiconductive oxides with optical transparency due to its wide bandgap of 3.6 eV. Furthermore, NiO films can control high and low conductive states under an external electric field, which is very useful for developing resistance-switching random-access memory devices2.NiO films have been prepared by different deposition techniques, including sputtering, sol-gel, pulsed laser deposition, spray pyrolysis, and chemical vapor deposition (CVD), atomic layer deposition (ALD). Among these, ALD is a powerful technique to grow thin conformal films with fine-tuning of both composition and thickness. The commonly used nickel precursors for ALD process are Ni(Cp)2, Ni(MeCp)2, Ni(EtCp)2, Ni(dmamp)2, Ni(dmamb)2, Ni(acac)2, Ni(apo)2, Ni(dmg)2, Ni(thd)2, and Ni(amd)2 in combination with ozone, water, hydrogen peroxide, or oxygen plasma1. However, ALD of NiO is not well developed yet. The major disadvantages associated with these nickel precursors are low volatility, thermal instability, and poor reactivity or narrow ALD window. Some of these precursors are expensive or very difficult to synthesize. The aforementioned situations indicate that developing a new ALD process for the NiO film with new precursors is pertinent.In this work, we explored different Ni-precursors such as Ni ketoiminates, Ni(ipki)2 and Ni(eeki)2 (their synthesis has been proposed earlier3), and Alanis, provided by Air Liquide. These different reactants combined with different oxygen sources (H2O and O3) resulted in high-quality NiO thin films. Surface analyses indicate NiO films are stoichiometric. They are uniform and exhibit a columnar morphology as determined by TEM (Fig.1a). From the growth characteristics, it is found that the Ni precursors have a strong effect on the GPC. Similarly, the oxygen source affects GPC but it also influences the morphology of the layer leading to significant roughness variations (observed by XRR and AFM). The use of H2O can also result in a nucleation delay. Polycrystalline films are observed when the reaction temperature reaches 125°C, and the crystallinity is further enhanced with deposition temperature (Fig.1b). After comprehensive optimization of the ALD process with Alanis precursor, it has been possible to reach one of the highest growth rate of 1.5 Å/cy, which is nearly seven times higher than that obtained with Ni(EtCp)2 based reaction with the same ALD reactor and having a wide temperature window (100 to 175 °C) with ozone. The same precursor also reacted well with water, giving a GPC of 0.9 Å with an initial nucleation delay. With the use of Ni-ketoiminates, the growth rate could be enhanced considerably with a GPC of 0.7 Å in combination with O3 at a higher temperature.The NiO layers were subsequently used as catalysts for water photooxidation in alkaline medium. The electrochemical properties are better when compared to layers grown by PVD. References Y. Zhang, L. Du, X. Liu, and Y. Ding, A high growth rate atomic layer deposition process for nickel oxide film preparation using a combination of nickel (II) diketonate-diamine and ozone, App. Surf. Sci., 2019, 481, 138-143.H.L. Lu, G. Scarel, X.L. Li, M. Fanciulli, Thin MnO and NiO films grown using atomic layer deposition from ethylcyclopentadienyl type of precursors, J. Cryst. Growth, 2008, 310, 5464-5468.3. D. Zywitzki, D.H. Taffa, L. Lamkowski, M. Winter, D. Rogalla, M. Wark, and A. Devi, Tuning coordination geometry of nickel ketoiminates and its influence on thermal characteristics for chemical vapour deposition of nanostructured NiO electrocatalysts, Inorg. Chem., 2020,59, 14, 10059-10070. Figure 1

For atomic layer deposition (ALD) of doped, ternary, and quaternary materials achieved by combining multiple binary ALD processes, it is often difficult to correlate the material properties and growth characteristics with the process parameters due to a limited understanding of the underlying surface chemistry. In this work, in situ Fourier transform infrared (FTIR) spectroscopy was employed during ALD of zinc-oxide, tin-oxide, and zinc-tin-oxide (ZTO) with the precursors diethylzinc (DEZ), tetrakis(dimethylamino)tin (TDMASn), and H2O. The main aim was to investigate the molecular basis for the nucleation delay during ALD of ZTO, observed when ZnO ALD is carried out after SnO2 ALD. Gas-phase FTIR spectroscopy showed that dimethylamine, the main reaction product of the SnO2 ALD process, is released not only during SnO2 ALD but also when depositing ZnO after SnO2, indicating incomplete removal of the ligands of the TDMASn precursor from the surface. Transmission FTIR spectroscopy performed during ALD on SiO2 powder revealed that a significant fraction of the ligands persist during both SnO2 and ZnO ALD. These observations provide experimental evidence for a recently proposed mechanism, based on theoretical calculations, suggesting that the elimination of precursor ligands is often not complete. In addition, it was found that the removal of precursor ligands by H2O exposure is even less effective when ZnO ALD is carried out after SnO2 ALD, which likely causes the nucleation delay in ZnO ALD during the deposition of ZTO. The underlying mechanisms and the consequences of the incomplete elimination of precursor ligands are discussed.