Anatase and rutile TiO 2 films deposited by arc-free deep oscillation magnetron sputtering
Explore the source record for details and available documents.
SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Explore the source record for details and available documents.
A thin layer of Al 2 O 3 at the back of CdSe x T e1-x /CdTe devices is shown to passivate the back interface and drastically improve surface recombination lifetimes and photoluminescent response. Despite this, such devices do not show an improvement in open-circuit voltage (VOC.) Adding a p + amorphous silicon layer behind the Al 2 O 3 bends the conduction band upward, reducing the barrier to hole extraction and improving collection. Further optimization of the Al 2 O 3 , amorphous silicon (a-Si), and indium-doped tin oxide (ITO) layers, as well as their interaction with the CdCl 2 passivation process, are necessary to translate these electro-optical improvements into gains in voltage.
Talk for AVS-Tampa Retarding field energy analyzer data on MoS2
Explore the source record for details and available documents.
High Power Impulse Magnetron Sputtering (HiPIMS) is a coating technology that combines magnetron sputtering with pulsed power concepts. Furthermore, by applying power in pulses of high amplitude and a relatively low duty cycle, large fractions of sputtered atoms and near-target gases are ionized. In contrast to conventional magnetron sputtering, HiPIMS is characterized by self-sputtering or repeated gas recycling for high and low sputter yield materials, respectively, and both for most intermediate materials. The dense plasma in front of the target has the dual function of sustaining the discharge and providing plasma-assistance to film growth, affecting the microstructure of growing films. Many technologically interesting thin films are compound films, which are composed of one or more metals and a reactive gas, most often oxygen or nitrogen. When reactive gas is added, non-trivial consequences arise for the system because the target may become “poisoned,” i.e., a compound layer forms on the target surface affecting the sputtering yield and the yield of secondary electron emission and thereby all other parameters. It is emphasized that the target state depends not only on the reactive gas' partial pressure (balanced via gas flow and pumping) but also on the ion flux to the target, which can be controlled by pulse parameters. This is a critical technological opportunity for reactive HiPIMS (R-HiPIMS). The scope of this tutorial is focused on plasma processes and mechanisms of operation and only briefly touches upon film properties. It introduces R-HiPIMS in a systematic, step-by-step approach by covering sputtering, magnetron sputtering, reactive magnetron sputtering, pulsed reactive magnetron sputtering, HiPIMS, and finally R-HiPIMS. The tutorial is concluded by considering variations of R-HiPIMS known as modulated pulsed power magnetron sputtering and deep-oscillation magnetron sputtering and combinations of R-HiPIMS with superimposed dc magnetron sputtering.
The aim of this project was to leverage the low area specific resistance (ASR) of Redox’s GDC electrolyte-based cell architecture and increase solid oxide fuel cell (SOFC) efficiency (i.e., open circuit voltage, or OCV) without significantly increasing cell resistance. The key to achieving the increased SOFC efficiency is the introduction of a thin sputtered yttria stabilized zirconia (YSZ) electron-blocking layer and a thin sputtered gadolinia doped ceria (GDC) barrier layer on top of the half-cell substrate (i.e., anode and GDC electrolyte). A key initial effort of the project was to improve the quality of the half-cell substrate surface so that any remaining defects were significantly smaller than the desired film thickness. During lab-scale trials, we had to overcome challenges with film cracking during post-sputtering treatments (e.g., thermal anneals) as well as damage to the sputtering targets. After the initial lab-scale trials, the focus shifted toward the use of commercial scale sputtering equipment with a sputtering equipment manufacturer in the microelectronics industry. Using the commercial sputtering equipment, we deposited films with different variations of power, PO2, sputtering time, platen speed, etc. We then determined the combinations of sputtering conditions and post-sputtering treatments that yielded high-quality films without cracks or other significant defects. While thin films were deposited on cells as large as 10 cm by 10 cm, the processing was optimized using 4 cm by 4 cm cells. Cell performance was evaluated in stainless-steel test fixtures between 500 °C and 700 °C with hydrogen fuel fed to the anode and air fed to the cathode. Extensive studies allowed us to determine that modified cathode and cathode contact firing processes were required to achieve theoretical OCV. Moreover, use of the new firing processes resulted in the need for a modified cathode contact to achieve a low ASR. In summary, the project demonstrated the performance of high OCV (1.13 V at 650 °C) from sputtered layers and a low ASR (~0.25 Ohms-cm2 at 650 °C) resulting from a modification of cathode/contact processing and the introduction of alternative contact layers that are sufficient to yield a Gen-1 cell with a maximum power density of approximately 1.2 W/cm2. At an operating voltage of 0.74 V, this would yield a cell power density of ~1.1 W/cm2. While not utilized in this project, a Redox Gen-2 cell has a catalyst-infiltrated porous anode that reduces the ASR by more than 50% from that of the Gen-1 cells used in this project. Therefore, if the sputtered YSZ electron-blocking layer and GDC barrier layer are added to a Gen-2 half cell with a similar increase in OCV to the theoretical value of ~1.13 V at 650 °C, and if the same improvement in ASR (from that demonstrated in this project) is achieved when using the Gen-2 half-cell architecture as a sputtered cell substrate, then the power density at 0.74 Vop could be as high as ~2.6 W/cm2. The impact of such power density gains, while still maintaining high cell efficiency, and thus high system efficiency, is a dramatic decrease in system cost because the stack represents ~30-40% of the SOFC system cost.
Astronomical measurements of S abundances in the diffuse interstellar medium (ISM) indicate ionized S is a dominant species with little (< 5%) S residing in grains (e.g. Jenkins 2009). This is an enigmatic result, given that abundant Fe-sulfide grains are observed in dust around pre- and post-main sequence stars and are also observed as major components of primitive meteoritic and cometary samples. These disparate observations suggest that the lifetime of sulfide grains in the ISM is short because of destruction processes. Our previous work has shown that FeS and MgS retain their crystallinity and do not amorphize during radiation processing, whereas enstatite and forsterite are readily amorphized. We have extended this study to measure the relative sputtering rates of FeS and MgS compared to enstatite and forsterite. Irradiation of FeS with 4 keV He+ results in preferential sputtering of S and the formation of a thin 2-3 nm, compact Fe metal layer that armors the surface. The zone of S loss extends to a depth of ~8-10 nm below the exposed surface. Despite this S loss, the FeS retains its crystallinity and shows no sign of incipient amorphization. Irradiation of FeS with 5kV Ga+ in a focused ion beam (FIB) instrument resulted in preferential sputtering of S and the formation of a 5-8-nm thick surface layer of nanophase Fe metal. X-ray mapping shows that the zone of S sputtering extends to a depth of nearly 20 nm, but there is no evidence for FeS amorphization, consistent with our previous work. The irradiation experiments show that the relative sputtering rate of FeS and MgS are much higher than olivine or enstatite. Sputtering experiments utilizing 30 kV and 5 kV Ga ions in the FIB produced volume loss in troilite that was ~4X greater than in enstatite or forsterite. The sputter yield under these conditions is such that for every Si atom sputtered from enstatite, ~14 S atoms are sputtered from FeS. We have performed similar sputtering experiments on Fe-bearing niningerite (MgS) and co-existing enstatite from the ALH 84170 EH3 chondrite. MgS also sputters much more rapidly than enstatite with a relative Si:S sputter yield of 1:8. For MgS, sulfur is highly depleted at the surface and the S-depletion zone extends to a depth of ~15 nm (using 5 kV Ga+). There is a corresponding zone of Mg and especially Fe enrichment that extends from 5 to ~10 nm below the surface, respectively. The dominant grain destruction mechanism in the ISM is sputtering from passage of supernova-generated shock waves. This process also results in the amorphization of crystalline silicates in the ISM. Our results indicate that FeS and MgS grains produced in evolved stars and injected into the ISM will be destroyed more rapidly than crystalline silicates. This process may account for the lack of significant depletion of S from the gas phase in the ISM. However, rare nanophase FeS grains occur as inclusions in circumstellar amorphous silicate grains found in comet dust particles analyzed in the laboratory. These results show that a finite amount of S in the ISM is sequestered in solid grains.
Previous literature has used several monocrystalline sputtering targets with various crystalline planes, respectively, to investigate the variations of the sputter yield of materials in different crystalline orientations. This study presents a method to measure the sputtered yields of Mo for the three low-index planes (100), (110), and (111), through using an easily made polycrystalline target. The procedure was firstly to use electron backscattered diffraction to identify the grain positions of the three crystalline planes, and then use a focused ion beam to perform the micro-milling of each identified grain, and finally the sputter yields were calculated from the removed volumes, which were measured by atomic force microscope. Experimental results showed that the sputter yield of the primary orientations for Mo varied as Y{sub (110)} > Y{sub (100)} > Y{sub (111)}, coincidental with the ranking of their planar atomic packing densities. The concept of transparency of ion in the crystalline substance was applied to elucidate these results. In addition, the result of (110) orientation exhibiting higher sputter yield is helpful for us to develop a Mo target with a higher deposition rate for use in industry. By changing the deformation process from straight rolling to cross rolling, the (110) texture intensity of the Mo target was significantly improved, and thus enhanced the deposition rate. - Highlights: • We used EBSD, FIB and AFM to measure the sputter yields of Mo in low-index planes. • The sputter yield of the primary orientations for Mo varied as Y{sub (110)} > Y{sub (100)} > Y{sub (111)}. • The transparency of ion was used to elucidate the differences in the sputter yield. • We improved the sputter rate of polycrystalline Mo target by adjusting its texture.
This paper primarily reviews the potential of using the sputtering process as a deposition technique; however, the manufacturing and sputter etching aspects are also discussed. Since sputtering is not regulated by classical thermodynamics, new multicomponent materials can be developed in any possible chemical composition. The basic mechanism for dc and rf sputtering is described. Sputter-deposition is described in terms of the unique advantageous features it offers such as versatility, momentum transfer, stoichiometry, sputter-etching, target geometry (coating complex surfaces), precise controls, flexibility, ecology, and sputtering rates. Sputtered film characteristics, such as strong adherence and coherence and film morphology, are briefly evaluated in terms of varying the sputtering parameters. Also described are some of the specific industrial areas which are turning to sputter-deposition techniques.
The sputter-deposition on surfaces exposed to plasma plays an important role in the erosion behavior and overall performance of a wide range of plasma devices. Plasma models in the low density, low energy plasma regime typically neglect micron-scale surface feature effects on the net sputter yield and erosion rate. The model discussed in this paper captures such surface architecture effects via a computationally efficient view factor model. The model compares well with experimental measurements of argon ion sputter yield from a nickel surface with a triangle wave geometry with peak heights in the hundreds of microns range. Further analysis with the model shows that increasing the surface pitch angle beyond about 45° can lead to significant decreases in the normalized net sputter yield for all simulated ion incident energies (i.e., 75, 100, 200, and 400 eV) for both smooth and roughened surfaces. At higher incident energies, smooth triangular surfaces exhibit a nonmonotonic trend in the normalized net sputter yield with surface pitch angle with a maximum yield above unity over a range of intermediate angles. The resulting increased erosion rate occurs because increased sputter yield due to the local ion incidence angle outweighs increased deposition due to the sputterant angular distribution. The model also compares well with experimentally observed radial expansion of protuberances (measuring tens of microns) in a nano-rod field exposed to an argon beam. The model captures the coalescence of sputterants at the protuberance sites and accurately illustrates the structure's expansion due to deposition from surrounding sputtering surfaces; these capabilities will be used for future studies into more complex surface architectures.
Reducing damages caused by sputtering of transparent conductive oxide (TCO) electrodes is critical in achieving highly efficient and stable perovskite/silicon tandem solar cells. Here we study the sputter caused damage to bathocuproine (BCP), which is widely used in highly efficient p-i-n structure single junction perovskite solar cells. While BCP buffer layer protects the underneath layers from damage, itself can be damaged by sputtering of TCOs at a wide range of target-substrate distances, supported by molecular dynamic simulation. More intriguingly, it is observed that TCO easily peeled off after sputtering when the sputtering target is close to substrate. This is ascribed to formation of stress during cooling down process after sputtering due to different thermal expansion coefficients of the layers. Our studies explain why tin oxide (SnO2) made by atomic layer deposition can replace BCP for a much better tandem device performance. SnO2 has high affinity with sputtered TCO electrode to suppress peeling-off issue and has higher bond energy to resist sputter induced damage, thus it allows a wider window of target-substrate distances than BCP during TCO sputtering. Ultimately, we demonstrate an efficient perovskite/silicon monolithic tandem solar cell with efficiency of 26.0% to illustrate the beneficial effects of reduced stress and damage.
Particle accelerators are considered as an important device that has wide applications in cancer treatment, sterilizing waste, preserving foods, ion implantation in semiconductor industry, and in production of isotopes for medical applications. Superconducting radiofrequency (SRF) cavities are the building blocks of a linear particle accelerator. Current particle accelerators use niobium (Nb) superconductors as the sheet material to fabricate a single SRF cavity for particle acceleration. With better superconducting properties (critical temperature Tc ~ 18.3 K, superheating field Hsh~ 400 mT), Nb3Sn is considered a potential candidate in SRF technology. Magnetron sputtering is a promising deposition method to fabricate Nb3Sn thin films inside SRF cavities. Superconducting Nb3Sn films were fabricated on Nb and sapphire substrates by magnetron sputtering from a single stoichiometric Nb3Sn target, by multilayer sputtering of Nb and Sn followed by annealing, and by co-sputtering of Nb and Sn followed by annealing. The variation of morphological and superconducting properties was investigated for different substrate temperatures, annealing temperatures, annealing durations, and thicknesses. The film properties were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), atomic force microscopy (AFM), and energy dispersive X-ray spectroscopy (EDS). The films had crystalline Nb3Sn structure without any presence of poor superconducting Nb6Sn5 and NbSn2 phases. The highest Tc of the films fabricated from the stoichiometric target, multilayer sputtering and co-sputtering were 17.44, 17.93, and 17.66 K respectively. Finally, a cylindrical sputter coater with two identical magnetrons was designed and commissioned to fabricate Nb3Sn films inside a 2.6 GHz SRF cavity. The magnetrons were installed facing opposite to each other in a custom designed vacuum chamber and multilayers of Nb and Sn films on 1 cm2 Nb substrates replicating the beam tubes and equator locations of the cavity and the coated multilayered films were annealed at 950 °C for 3 h. The XRD of the as?deposited and annealed films confirmed the formation of Nb3Sn after the annealing. The dissertation discusses the fabrication process, characterized results of the fabricated films, the design of the cylindrical sputter coater and the preliminary data obtained from the sputter coater.
Initial studies have been made of the sputtering and deposition phenomena in a 30-cm thruster. Sputtering rates, of the cathode baffle, one of the main sources of sputtered material in a thruster, have been measured by weight loss as a function of several thruster parameters. Sputtering rates were found to increase with both cathode flow rate and beam current when constant discharge voltage of 37 volts and power loses of 185 ev/ion were maintained. Sputtering rates were reduced 24% as discharge voltage was decreased from 37 to 33 volts while keeping discharge power constant. Qualitative agreement was found between sputtering rates obtained by the weight loss and those implied by spectroscopically observed line intensities of the excited iron sputtered atoms. After the completion of the sputtering tests, deposition and sputtering sites inside the thruster were identified.
Initial studies have been made of the sputtering and deposition phenomena in a 30-cm thruster. Sputtering rates, of the cathode baffle, one of the main sources of sputtered material in a thruster, have beem measured by weight loss as a function of several thruster parameters. Sputtering rates were found to increase with both cathode flow rate and beam current when constant discharge voltage of 37 volts and power losses of 185 ev/ion were maintained. Sputtering rates were reduced 24% as discharge voltage was decreased from 37 to 33 volts while keeping discharge power constant. Qualitative agreement was found between sputtering rates obtained by the weight loss and those implied by spectroscopically observed line intensities of the excited iron sputtered atoms. After the completion of the sputtering tests, deposition and sputtering sites inside the thruster were identified.
Surface sputtering is explained theoretically by means of a 3-body sputtering mechanism involving the ion and two surface atoms of the solid. By means of quantum-statistical mechanics, a formula for the sputtering ratio S(E) is derived from first principles. The theoretical sputtering rate S(E) was found experimentally to be proportional to the square of the difference between incident ion energy and the threshold energy for sputtering of surface atoms at low ion energies. Extrapolation of the theoretical sputtering formula to larger ion energies indicates that S(E) reaches a saturation value and finally decreases at high ion energies. The theoretical sputtering ratios S(E) for wolfram, tantalum, and molybdenum are compared with the corresponding experimental sputtering curves in the low energy region from threshold sputtering energy to 120 eV above the respective threshold energy. Theory and experiment are shown to be in good agreement.
The potential of using the sputtering process as a deposition technique is reviewed; however, the manufacturing and sputter etching aspects are also discussed. The basic mechanism for dc and rf sputtering is described. Sputter deposition is presented in terms of the unique advantageous features it offers such as versatility, momentum transfer, stoichiometry, sputter etching, target geometry (coating and complex surfaces), precise controls, flexibility, ecology, and sputtering rates. Sputtered film characteristics, such as strong adherence and coherence and film morphology, are briefly evaluated in terms of varying the sputtering parameters. Also discussed are some of the specific industrial areas which are turning to sputter deposition techniques.
The effects of sputtering n-type (100)GaAs with 1 to 5 keV argon ions have been studied in ultrahigh vacuum. Auger electron and X-ray photoelectron spectroscopies have been used to characterize the surface before and after sputtering. Gold Schottky contacts were prepared in situ to avoid contamination at the metal/semiconductor interface. The data show that sputtering depletes the surface in As, and the extent of depletion increased as the ion energy was increased. However, angular-resolved photoelectron data showed As was enriched on the outer atom layer by 5 keV but depleted by 1 keV Ar(+) sputtering. Current and capacitance versus voltage data showed that sputtering always lowered the Schottky barrier height and generally increased the ideality factor. However, the barrier height after 3 keV sputtering was higher than after 1 keV sputtering. Those effects are explained based on a combined effective work function model and creation of a donorlike surface damage layer. The donor layer was correlated with As depletion by sputtering. Deep level trap formation and annealing of sputtering effects were studied.
Sputter erosion in ion thrusters has been measured in lifetests at discharge voltages as low as 25 V. Thruster operation at this discharge voltage results in component erosion rates sufficiently low to satisfy most mission requirements. It has been recognized that most of the internal sputtering in ion thrusters is done by doubly charged ions. Knowledge of the sputtering threshold voltage of a xenon molybdenum system would be beneficial in understanding the sputtering process as well as making more accurate calculations of the sputtering rates of ion thruster components. Sputtering threshold energies calculated from various formulations found in the literature results in values ranging from 28 to 200 eV. It is evident that some of these formulations cannot be relied upon to provide sputtering thresholds with any degree of accuracy. This paper re-examines the threshold energies measurements made in the early sixties by Askerov and Sena, and Stuart and Wehner. The threshold voltages as derived by Askerov and au have been reevaluated by using a different extrapolation method of sputter yields at low ion energies. The resulting threshold energies are in general similar to those measured by Stuart and Wehner. An empirical relationship is derived,for mercury and xenon ions for the ratio of the sputtering threshold energy to the sublimation energy as a function of the ratio of target to ion atomic mass.