A Short History of Thermionic Cathodes B. Vancil 1 1 e beam, inc., 21070 SW Tile Flat Rd., Beaverton, OR 97007, USA bernie@ebeaminc.com By thermionic cathode, we mean a device that, when heated, releases electrons into the vacuum for some useful purpose. We trace their history from Edison s original discovery (see Figure 1) in 1883, to Fleming s valve in 1903, and the nearly simultaneous invention of the oxide cathode by Arthur Wehnelt. Also, near the same time, Richardson published his famous 3 kt equation j ATe. He did this at least 25 years before quantum mechanics provided a rigorous derivation. Then Schottky postulated barrier lowering and added V to Richardson s equation. Let us not forget Child s explanation in 1911 of space charge limited flow 3/2 kv j, 2 d along with adjustments by Irving Figure 1. Edison s discovery Langmuir in 1913. Thus, the entire gamut of thermionic emission was explained and demonstrated in a ten-year window of time from 1903 to 1913. The electron the subject of all this activity was not discovered until 1896 (J. J. Thompson). Many more discoveries were made in the 1920 s and 30 s. One of these was the beneficial effect of thorium oxide on tungsten cathode emission. This lowered the temperature by 400º C and made possible long-life power tubes. We switch forward now to 1948, when Philips invented the L cathode and ushered in the era of barium dispenser cathodes 1. This was the first reservoir cathode. It contained a sealed reservoir behind a porous tungsten matrix. The reservoir contained barium carbonate. The cathode was prone to long periods of outgassing as the carbonate converted to oxide. A few years later, Katz at Telefunken AEG in Germany invented the M-K (metal capillary) reservoir cathode 2. It employed barium oxide in the reservoir along with tungsten wool as the activator. It was tested on the Crane life test (see Figure 2) in the 1990 s and achieved over 100,000 hours of operation with no discernible knee point migration. Thus, as early as 1953, the world had a cathode with virtually perpetual emission. But reservoir cathodes are complex and expensive, and for most applications, the world doesn t need eternal life. In 1955, Philips introduced the impregnated dispenser cathode a tungsten matrix with barium-aluminate infiltrating its pores 3. It had the drawback that barium evolution diminishes with time, which means that the temperature must be raised to maintain barium
coverage. But this drawback was overlooked and these cathodes became the mainstay of the microwave tube industry in the 1960 s. In 1966, Philips struck again with Zalm s invention of the so-called M cathode 4. He added a layer of osmium on the surface of an impregnated cathode. This dramatically lowered the work function. Knee temperature (the transition between space charge and temperature-limited flow) dropped from 1030º C b (W) to about 960º C b (W) at 5 A/cm 2. These cathodes came into nearly universal use for microwave tubes. Figure 2. Crane life test results We should not forget the important work on sources for micro-focus electron beams, such as are used in SEMS, TEMS, micro-focus x-rays and electron lithography. Single-crystal lanthanum hexaboride was chosen for this purpose. It was ground into truncated cones with flats as small as 25 micrometers, with the 100 plane as the emitting surface. Lynn Swanson and his team at Oregon Graduate Institute were key players in this work. Recently, Schottky emitters have emerged as the dominant source for micro- and nano-focus. They are field enhanced thermionic emitters. Back to barium-based cathodes. Things did not remain static for long. Varian in the late 1980 s added a barrier layer behind the osmium coating on M cathodes. This prevented interdiffusion of tungsten into the osmium. Thus, the alloy of tungsten-osmium could be kept at its optimum ratio for lowest work function. This dropped the temperature from 960º C b (W) to 910º C b (W). At the same time, Varian developed the RV reservoir cathode. It employed the new osmium-tungsten layer and several other innovations. RV cathodes were deployed in the Crane life test and operated 100,000 hours with no knee point migration (see Figure 2). They emulated the M-K
cathode at an operating temperature about 200º C lower. Michael Green was largely responsible for these breakthroughs. In the 1980 s and 90 s, Philips was back again with groundbreaking work on scandia-doped impregnated cathodes. This led to record-breaking emission levels at 400 A/cm 2 (Georg Gaertner, 1997) 5. Then came another breakthrough, this time by workers at Beijing University of Technology. Workers there discovered a method for distributing nano-crystalline scandium oxide in the tungsten powder, which is subsequently pressed and sintered into cathodes. These cathodes achieved emission levels of 100 A/cm 2 at temperatures below 970º C b (W). A seminal paper by Professor Wang appeared in 2007 6. Other groups doing important early work on nano-scandia-doped impregnated cathodes were Beijing Vacuum Electronics Research Institute (BVERI) under Ji Li and University of California at Davis under Neville Luhmann. E beam inc. was able to produce a nano-scandia cathode in 2014. Figure 3 summarizes the important advances in impregnated cathode technology between 1955 and 2005. Figure 3. Advances in impregnated cathode technology Miniaturization of impregnated cathodes came largely from the oscilloscope industry. Tektronix, Inc. needed a way to improve writing speed on its high-performance cathode ray tubes. This required higher cathode emission and very small, low-power, low-capacitance cathodes. This led to one of the first miniature dispenser cathodes 7. Later, as sweep speeds and pixel counts increased on computer monitors and high-definition television, a number of miniature designs were mass-produced. E beam inc. produced a miniature reservoir cathode in 2002 8. And So We Come to the Present Although cathode ray tubes are gone, there is still a need for miniature cathodes, especially in space (CubeSats) and for millimeter and sub-millimeter wave linear beam oscillators and amplifiers. E beam inc. recently produced a 0.034-inch diameter cathode dissipating only 0.6
watt, see Figure 4. It also produced a miniature sheet beam cathode dissipating only 0.1 watt, see Figure 5. Figure 4. e beam inc. 0.034-inch diameter Figure 5. e beam inc. miniature sheet beam cathode cathode Efficient miniature cathodes are also needed for micro-electric propulsion of CubeSats and other small satellites. From microwave ovens to klystrons, x-rays to SEMS, pentodes to traveling wave tubes, gyrotrons to e beam lithography, magnetrons to mass spectrometers, ion thrusters to free electron lasers, backward wave oscillators to clinical accelerators, thermionic cathodes continue to play a critical role in the functioning of the modern world. What is the Future? In space, microwave tubes are far more efficient and compact than solid state for amplifying high-frequency radio waves. A broad-band traveling wave space tube might have overall efficiency of 60%, while its counterpart solid state power amplifier might achieve only 20%. That is a lot of extra heat to get rid of. And the only way to get rid of heat in space is by 4 radiation. Q A T. But transistors can t operate at high temperature, so their radiation efficiency is very low. A traveling wave tube with a 0.6-watt scandate cathode is 35 times more efficient at radiating heat than an SSPA. The demands on efficiency and compactness become even steeper for CubeSats. They have less surface for radiating heat, they are very small (10 cm x 10 cm x 10 cm) and they have limited prime power (about four watts). With constellations of hundreds of CubeSats now contemplated, the demand for high-performance thermionic cathodes will increase rapidly. Furthermore, CubeSats flying in formation need a means to adjust position and attitude. This will require at least two micro-hall thrusters. Each of these requires two cathodes, one for ionization
and one for neutralization. Also, extremely small high-output cathodes are needed for millimeter and sub-millimeter sources and amplifiers. Scandate cathodes will probably fill this need. There is a great need for cheap terrestrial amplifiers on a dollar per watt basis from 100 to 1,000 watts, depending on frequency. If the price could be brought down to one dollar per watt, tube technology would have a great advantage over solid state. Solid state can t easily reach these power levels without expensive and inefficient power combiners. Conventional microwave tubes are too expensive. An inexpensive traveling wave tube would be a great boon. This should be achievable. Remember, the OEM price for a 1500-watt microwave oven magnetron is under $10. References 1. H. J. Lemmens, U.S. Patent No. 2543728, published 2/27/1951. Priority The Netherlands, 26/11/1947. 2. H. Katz, U.S. Patent No. 2750527, published 12/6/1956, Priority Germany, 19/11/1951. H. Katz claims earlier priority: French Patent Specification 903976, published 10/13/1946, priority Germany 19/8/1942. 3. R. Levi, R. Hughes, U.S. Patent No. 2,700,000, appl. 02/27/1952, granted 01/18/1955. 4. P. Zalm and A. J. A. van Stratum, Osmium dispenser cathodes, Philips Technical Review, Vol. 27, 69-75 (1966). 5. G. Gaertner, P. Geittner, H. Lydtin, and A. Ritz, Emission properties of top-layer scandate cathodes prepared by LAD, Appl. Surf. Sci., Vol. 111, No. 3, pp. 11-17, February 1997. 6. Y. Wang, J. Wang, W. Liu, K. Zhang, and J. Li, Development of High Current-Density Cathodes with Scandia-Doped Tungsten Powders, IEEE Trans. On Electron Dev., Vol. 54, No. 5, May 2007. 7. B. Vancil, Cathode Structure, U.S. Patent No. 4954745, filed 03/22/1989, granted 4/9/1990. 8. B. Vancil and E. Wintucky, Miniature Reservoir Cathode An Update, Appl. Surf. Sci., Vol. 215, 18-24 (2003).