C.R.K. Marrian

Interatomic Auger analysis of the oxidation of thin Ba films: II. Applications to impregnated cathodes

Abstract The oxidation and thermal desorption of evaporated Ba films on different substrates have been characterized using low energy interatomic AES techniques. The films, which were deposited on W〈100〉 (as well as Ir and Os-W) varied from a fractional layer to ∼ 10 layers. Oxidation states, i.e. Ba: O x , ranged from x =0 to x ≈2. Comparison of these data to those from actual impregnated cathodes allows one to infer the thickness and state of oxidation of the cathodes. The results indicate that the active state for “B-type” impregnated cathodes can be reproduced by a near monolayer of the stoichiometric BaO on the W surface. The density of Ba atoms (or O atoms) in this “monolayer” should be ∼ one half that of the W atoms based on size consideration of the Ba compared to the W. This was confirmed by LEED. Measurements of the effusion of Ba compounds from the pores, the substrate interaction and the thermal evaporation indicate that on the W substrate, the monolayer BaO stoichiometric ratio is not only a stable configuration, but also is the one that has the lowest work function (i.e. ∼ 2 eV). The substrate work function is lowered because of the dipole formed from the adsorbed BaO. Studies of the electron interactions between the Ba and O indicate that (in agreement also with surface plasmon results) a more complete electron transfer exists between the Ba and O for the BaO layer on the Ir substrate than on the W substrate. These results are based on the observation that O 2p states on the W substrate appear to be filled not only by Ba 6s but also by W valence electrons.

Interatomic Auger analysis of the oxidation of thin Ba films: I. Characterization of the low energy Auger spectrum

The oxidation states of thin Ba films on Ir have been studied using an analysis of interatomic Auger lines. From published XPS data, some of the low energy AES lines were described in terms of various possible core, core, valence transitions where the core states were Ba states but the valence state could be either due to Ba or O. The magnitude of all AES lines, whose energy indicated an O valence state, was observed to increase with oxidation while those ascribed to Ba valence states decreased. The Ba 4d, 5p, valence transitions in the 65–75 V range appeared relatively free from clutter of other Ba and substrate lines and were the ones primarly used in determining the Ba-O bonding. The increase with oxidation of the interatomic Ba 4d, 5p, O 2p line at ∼ 68 V was found to be correlated with the increase in the O 2p loss peak. Furthermore, shifts in energy of this peak for various film thicknesses varied in accordance with corresponding shifts in O 2p binding energy with respect to the Fermi level. Also, the decrease with oxidation of the Ba 4d, 5p, 6s line at ∼ 73 V was found to be correlated with the decrease in conduction electrons contributing to the surface plasmon peak. The manner in which these lines can be used to determine the extent of electron transfer in the formation of compounds other than BaO (such as BaC2) is also given.

Preparation and oxidation of a thin Ba film

A method of determining the thickness of vapor-deposited thin Ba films is described. The thickness is expressed in terms of numbers of Ba monolayers which were defined and characterized on W(100) and Ir(100) substrates using Low Energy Electron Diffraction (LEED) and Auger Electron Spectroscopy (AES). The changes in the electronic properties of a 10-layer Ba film on Ir(100) during oxidation were studied using Low Energy Electron Reflectivity (LEER) measurements. When the oxidation of Ba was nearly complete (as shown by AES), the LEER patterns were similar to that found for a randomly oriented BaO layer with perhaps a partial layer of Ba on the surface. Additional O2 exposure, however, resulted in the formation of a BaO crystal whose characteristic LEER patterns indicated a Fermi level position which suggested a high density of oxygen-vacancy donors and whose large electron affinity indicated non-stoichiometry at the surface. Further O2 exposure resulted in a BaO-type surface dipole and oxygen vacancy donor densities more typical of BaO lattices. At this level of oxidation, a minimum work function of 1.4 eV was attained, which was identical to that of an active thin BaO film. Furthermore, AES indicated a Ba to O ratio identical to that of a BaO film. Any change from this stoichiometry through the addition of either oxygen or barium resulted in an increase in the work function.

The characterization of the surfaces of tungsten-based dispenser cathodes

Abstract The “CD”, “M” and scandate cathodes all have a better thermionic emission performance than the “B” type of impregnated tungsten based dispenser cathode. Modern surface analysis techniques (AES, ISS, SIMS and SLEEP) have been used to examine the surfaces of the four cathode types. The cathode surfaces have been studied in their active state, during reactivation following a sputter cleaning of their surfaces and whilst poisoned by, and reactivating from, an exposure to 10 L O 2 . The results are discussed in terms of surface models which would be consistent with the observed data. The differences in the surfaces of the different cathode types are examined and the value of the various surface analysis techniques discussed.

Interactions of BaO with mixed-metal substrates: WIr

Abstract Surface studies of BaO on WIr alloy substrates of different composition show work functions lower for the alloys than on either pure W or pure Ir. In particular, the lowest work function (∼ 1.8 eV) appears near the 50% W, 50% Ir concentration. Peak height characterizations and energy shifts of interatomic Auger lines indicate an adsorbate-substrate interaction is responsible for the function lowering. The interaction consists of the O of the adsorbed Ba+O− molecule receiving electrons from the W and the Ba from the same molecule contributing electrons to the Ir. The resulting increase in surface dipole (as observed from interatomic O 2p and Ba 6s peak heights) is in very good agreement with the actual work function lowering. If the substrate consists of significant “patches” of W and Ir, then the Ba and O from the same molecule cannot interact with the different substrate components. When this happens, neither is the adsorbate-substrate interaction seen in the interatomic Auger spectra nor is any work function-lowering observed. The minimum in work function near 50% W, 50% Ir is the point where the dipole lowering (i.e., charge transfer) is most efficient because at thatconcentration there is one Ir atom available for each Ba atom and one W atom available for each O atom on the surface.

Fast turn-on characteristics of tungsten-based dispenser cathodes following gas exposures

Abstract The problems associated with the reactivation following shelf storage of different types of tungsten-based dispenser cathodes have been investigated. Reactivation times were found to be severely limited by repoisoning processes, which have been isolated and identified. Data are presented, indicating the finite times (in the absence of repoisoning), which are required to reactivate the cathodes following exposure to various gases. Of the gases studied, exposure to oxygen and water vapour caused the slowest reactivations. Water vapour was the component of the vacuum system ambient responsible for the poisoning caused by exposure to the ambient. Following exposure to each of the gases studied, the “M” type cathode reactivated slower than the “B” type cathode. The results have shown that both the choice of cathode and the design of the microwave tube are important if a fast turn-on following shelf storage is to be achieved.

Interpretation of AES data of impregnated cathodes

Abstract Many surface studies of impregnated cathodes involve a measurement of the Ba590, O510 and W170 Auger peak heights. This paper shows how these measurements plotted in terms of O Ba versus W Ba concentration ratios can be used to interpret experimental results on impregnated B-type cathodes. Data are presented from actual cathodes and simulated cathode surfaces using well defined BaO films on W. These results as well topographic data will be compared to various calculated models. It will be shown that an active B-type cathode consists of nearly a monolayer of BaO on W with a slight excess of O compared to Ba. This excess O, however, is associated with W rather than the BaO. Ageing increases the O Ba ratio and this takes place principally by the removal of Ba from the BaO surface, again with the excess O being associated with the W rather than the BaO. The work function increase during ageing corresponds only to the area-decrease of BaO. If the surface accumulates so much O that it exceeds the available empty Ba sites (i.e., all W is covered by Ba or O), then any additional O appears to sit on top of the remaining Ba and the surface will be poisoned. These conclusions are not only the result of work function observations but are also consistent with interatomic Auger analysis of BaO interactions as well as surface plasmon and ISS results. Toporgaphic data obtained from pore/impregnant regions are also in good agreement with calculated values of partial coverage of typical tungstate and impregnant residues.

Reaction products and subsequent thermal decomposition of Ba films exposed to CO, CO2, H2O and O2

The reactions involving thin Ba films (monolayer as well as 10-layer films on Ir) and CO, CO2, H2O and O2 have been studied using Auger Electron Spectroscopy (AES) and thermal desorption. Common to these reactions is the formation of BaO which is evident from the growth of the interatomic 68 eV (Ba 4d, 5p, O 2p) Auger line and decreases in 73 eV (Ba 4d, 5p, 6s) line. From CO and CO2 exposures, BaC2 has also been observed to form as indicated by the Auger peak shape in the carbon Auger lines. A quantitative analysis has been used to determine the chemical equation governing each reaction. The presence of BaC2 in BaO has been observed to increase the work function to 2.0 eV (compared to 1.4 eV for an active BaO film). Although this poisoning is moderate compared to that resulting from BaCO3 or Ba(OH)2 formation, it requires higher temperature for reactivation. In addition to BaC2, BaS has also been analyzed in terms of interatomic Auger transitions. The results showed that BaS like BaO exhibited such a transition indicating a charge transfer from Ba to S in the molecular bond. BaC2 on the other hand showed no such transition.

Study of pyrolytic carbon films for non-emitting grid applications

Abstract Pyrolytic carbon films were prepared on a hot substrate inside UHV system back-filled with 6 × 10 −5 Torr of CH 4 . The resulting films showed two different chemical states of carbon; carbides formed on W and Ta substrates and graphitic carbon layers were yound on ?. Ir substrate. Alter BaO deposition, the carbides showed a work function too low to be useful as grids. On the other hand, the work function of pyrolytic carbon films were high both before and after Ba or BaO coverage. Study of the degradation mechanisms of carbon films, formed by e-bean techniques, however, showed that they also reacted rapidly with W and T 2 resulting in a carbide formation. Thus it suggested that in the application of pyrolytic graphite grids, contact with carbide-forming metals should be avoided. Interaction of the pyrolytic carbon with BaO also resulted in the removal of carbon. However, this “leeching” phenomenon projects a life for pyrolytic graphite grids much longer than actually observed in the tube tests. The diffusion of Ba and BaO through pyrolytic carbon layer was evident in this workm but its precise relation to life failure rate is not yet understood.

High-resolution lithography with a vacuum STM

Abstract The scanning tunneling microscope (STM) is demonstrated to be a useful low-energy e-beam lithographic tool and a valuable probe of resist materials. Lithography with a vacuum STM and a 10 nm 50 kV e-beam has been performed on identically prepared and processed films of a state-of-the-art high-resolution negative resist (SAL-601 from Shipley). On bulk substrates (Si and GaAs), resist films up to 50 nm thick have been patterned, developed and observed in a scanning electron microscope. A resist thickness of 50 nm is sufficient to withstand a reactive ion etch. On Si, the minimum feature size observed with the 50 kV e-beam was 95 nm. In contrast, the STM lithography defined features observed in the developed resist to have linewidths down to 23 nm. The variation of feature size with exposure dose, exposure voltage and resist thickness has been studied. The STM lithography has shown that SAL-601 is inherently capable of sub 25 nm resolution and that low voltage e-beam lithography can produce smaller minimum feature sizes than lithography with a 10 nm 50 kV e-beam.