Plasma/Electrode Interactions in High Current Density Environments

Transcription

1 AFOSR Program Review on Materials and Processes Far From Equilibrium Sept , 2012 Plasma/Electrode Interactions in High Current Density Environments Edgar Choueiri Electric Propulsion & Plasma Dynamics Laboratory (EPPDyL) Jay Polk Jet Propulsion Laboratory California Institute of Technology Alp Sehirlioglu Case Western Reserve Program Manager: Dr. Ali Sayir

2 Practical Motivation Critical Air Force missions will rely on high current-density ( A/cm 2 ) devices, such as high-power ( kw) plasma thrusters compact pulsed power systems and plasma switches. Electrodes, and specifically cathodes, are the most highly stressed components of these high current-density devices (HCDD). Presently cathodes were derived from thermionic emitters developed for vacuum tube applications in the 1950 s. The next generation of cathodes will be based on a deeper understanding of Plasmaelectrode interactions (PEI).

3 Fundamental Motivation Cathodes provide a rich area for investigation of fundamental plasma-material interaction problems Plasma discharges create a harsh environment for emitter materials (erosion) High temperature Ion bombardment and sputtering Chemical erosion processes The plasma has an enormous effect on material transport (where erosion products go) Ionization of evaporated emitter material Electric field effects on transport Suppression of chemical reactions due to buildup of products Surface modification impacts device performance (effects of erosion) Increases operating temperature or Decreases current emission capability Makes ignition difficult Ultimately limits life of cathode

4 An Overview of Thermionic Cathode Types Dispenser Cathodes Cathode Tube Insert Heater Orifice Plate A low work function emitter surface is formed by a refractory metal substrate with a monolayer of electropositive atoms (e.g. Ba) Electropositive activator atoms are lost by desorption from the hot surface The activator atoms must be replenished by a supply source in the interior Bulk Emitter Cathodes The emitter surface is a bulk material with a relatively low work function (e.g. lanthanum hexaboride)

5 Operation of Dispenser Cathodes Cathode Tube Insert Emitter is a porous tungsten matrix impregnated with barium calcium aluminate Heater Orifice Plate Key to low work function surface is the equilibrium coverage of the adsorbed complex: Ba + O - on W Diffusion-limited reactions between the tungsten and the impregnant in pores produce Ba and BaO vapor Vapor flow through pores replenishes Ba and O lost by evaporation

6 Bulk Emitters Have Some Potential Advantages Operating temperature of dispenser and LaB6 cathodes Bulk emitters generally have higher work functions, therefore higher temperatures (a possible exception is ceramic electrides: <1 ev!) Example: LaB 6 operates at about C, but still has a lower evaporation rate LaB 6 is much more resistant to poisoning by reactive impurities LaB6 has lower evaporation rate, even At higher operating temperature LaB6 can tolerate 1-2 orders of magnitude higher concentrations of water and oxygen

7 Impact of Lack of Knowledge There is no systematic design methodology for gas discharge cathodes because fundamental processes are not understood State-of-the-art hollow cathodes exploit only a small part of the emitter area Xenon Plasma Density, n e /10 19 (m -3 ) Critical gaps in knowledge Fundamental material properties of promising emitter materials Processes that control the plasma-surface contact area Emitter Temperature and Electron Current Density Material transport processes that determine cathode life Plasma modification of emitter surfaces

8 Tools We re Applying to Fill The Knowledge Gaps State-of-the-art surface and thermochemistry analysis techniques to determine fundamental properties LaB 6 evaporation rates and products LaB 6 cathode surface stoichiometry and work function Barium depletion depth in dispenser cathodes Surface state of advanced dispenser cathodes such as Ba on O on W-Ir or W-Os (mixed metal matrix cathodes) State-of-the-art hollow cathode plasma codes to understand plasma contact area on the emitter Plasma transport models and surface kinetics models to understand the flow of erosion products through the plasma and the effect that has on the emitter surface state Barium transport in xenon plasmas La and B transport in xenon for LaB 6 hollow cathodes Tungsten transport in xenon dispenser cathodes Integration of the various models into design codes to develop advanced hollow cathodes

9 Ba Evaporation Rate (mg/cm 2 hr) Vacuum Dispenser Cathodes Are Easy! Time (hrs) Control parameters: Temperature Emitter Area Extraction Voltage Geometry (Electrostatics) Barium transport: Ba and BaO generated in pores flows out by surface diffusion and Knudsen flow Once it evaporates from the surface, it is gone (deposits on downstream surfaces) End of life: Ba supply rate drops as reaction front proceeds into the cathode (longer, more resistive path through porous tungsten) When supply no longer compensates for desorption, coverage drops and work function increases

10 Plasma Discharge Cathodes are Fundamentally Different from Vacuum Cathodes Barium transport in hollow cathodes is much more complex than that in vacuum cathodes Experiments and modeling have revealed 3 key features: Deposits Blocking Barium Flow: Erosion and redeposition of tungsten at the downstream end of the insert result in formation of a dense tungsten shell, which inhibits barium flow from the interior. Despite this, cathodes perform normally. Barium Recycling: Barium ions created in the xenon plasma are pushed back down to the emitter surface. This gas phase transport process resupplies the downstream emitter surface. Reaction Suppression: Buildup of barium vapor in the discharge results in a higher barium partial pressure compared to the equilibrium pressure of the decomposition reaction As a result, barium is supplied from a narrow region between the tungsten shell and the region upstream where the barium partial pressure suppresses the impregnant decomposition reactions

11 Green Cathodes: Reduce, Reuse, Recycle Ba Is Ionized in Xe Plasma Ba Vapor Evaporates From Surface Ba Ions Return To Surface Plasma reactions and plasma-surface interactions have a major (and beneficial, if properly exploited) effect on cathode operation

12 Plasma Discharge Cathodes are Also Operated in a Very Different Way We control current, not temperature Plasma cathodes are generally self-heating, with no direct control over the temperature Emitting area is not effectively controlled Much of the potential emitting area is not used Current is emitted where the plasma density is high Until recently, the plasma distribution was not well-understood Extraction voltage cannot be controlled Electric field at the surface is determined by the plasma sheath characteristics The plasma mitigates space-charge effects Complexities of transport processes have so far limited our ability to influence mass loss rates Improved understanding will give us the capability to control more of these parameters, to effectively design plasma discharge cathodes

13 Previous Work on Dispenser Cathodes Shows Barium is Effectively Recycled Xenon Plasma Solution Barium Transport in Xenon Plasma

14 Comparison of Simulation and Experimental Results Depletion depth corresponding to operation in vacuum for 8200 hours is plotted as the lower dashed line. Values are much larger than the measured values. Barium partial pressure in the discharge plasma suppresses depletion. The predicted magnitude of the depletion depth is larger than the measured values by a factor of up to 2.5. Temperature affects both the magnitude of the depletion depth and the width of the profile. Agreement between model and experiment is good assuming the temperature of the cathode is 70 C lower than the measured values (solid blue line). A lower temperature reduces the rate of depletion and the equilibrium barium pressure over the reaction front.

15 Barium Recycling Can Be Influenced by the Cathode Design Small Orifice Cathode Large Orifice Cathode Barium Ion Density, n Ba /10 14 (m -3 ) Barium Ion Density, n Ba /10 14 (m -3 ) Greater Ba loss out of the larger orifice because: The internal gas pressure is lower, and the plasma penetrates further The peak plasma density is lower However, the cathode temperature is lower! Multi-dimensional design space

16 Cross Sections and Rate Coefficients for La and B IP = IP = No data on ionization cross sections for La and B La values based on Drawin cross-section; should be refined B values based on Kim model, which agrees well with other models and data for Al, Ga, In (other elements in the same column) Ionization Rate Coefficients (m 3 /s) Electron Temperature (ev) Barium Lanthanum Boron Xenon B cross section is surprisingly large because of autoionization from excited states La expected to behave like Ba, not clear for B

17 Current Focus is on Lanthanum Hexaboride Cathodes Xenon Plasma Solution La and B Transport in Xenon Plasma Plasma density, n e / 1e19 (m -3 ) Lanthanum Ionization Mean Free Path (m) Electron temperature, T e (ev) Plasma potential, f (V) Boron Ionization Mean Free Path (m) Lanthanum MFP is <1 mm; likely to be ionized and very effectively recycled in discharge Boron MFP is 4 mm; most evaporated boron will be lost through orifice Currently modifying Ba transport code to model details of La and B transport

18 Characterization of an Insert from a High Current LaB6 Cathode

19 Comparison of a Pristine LaB 6 Surface with the Emitter Surface after ~250 Hours Fresh Fracture 5000X Inner Diameter 5000X

20 Comparison of a Pristine LaB 6 Surface with the Emitter Surface after ~250 Hours Fresh Fracture 13KX Inner Diameter 13KX

21 EDS Spectra Comparison Indicates Coating is Rich in O and La B Grain Coating O La

22 Element Maps Show Clearly That Coating is Predominantly La and O Boron Carbon Oxygen Lanthanum The coating could be vapor-deposited lanthanum that was oxidized after exposure to air

23 Surface State and Vapor Composition Surface state depends on a balance between diffusion in the interior and vaporization from the surface Lanthanum-rich vapor is associated with lanthanum-rich surfaces

24 Effect of Surface State on Emission Boron-rich phases have a higher work function Shouldn t need to worry about work function changes with La enrichment

25 Vaporization Behavior Depends on Temperature Vaporization dominates Diffusion dominates

26 Diagnostic Needs Depth profiling capability to determine composition underneath suspected LaO layer Ability to perform model experiments: Synthesis of relevant surfaces by deposition of La on hot LaB6 surface Surface analysis tools to determine composition as surface evolves Diagnostics such as UPS to measure work functions of synthesized surfaces Mass spectrometry to determine vaporization rates and vapor content of synthesized surfaces Resulting data would be extremely valuable in refining model boundary conditions

27 High-Risk Cathode Concepts In addition to applying the tools we develop to optimize hollow cathodes with a more conventional design, we intend to explore high-risk but potentially high-payoff solutions. Among these are the following radical concepts, based on distilling our decades-long experience in electric propulsion. Main feature is control of the current attachment dynamics. Current Focus: The RF-Controlled Hollow Cathode Potential Advantages Controlling the mass flow and the RF field allows controlling the penetration of the RF-created plasma downstream into the hollow cathode. Controlling the downstream plasma penetration would allow controlling the current penetration depth. Potential Challenges Power/complexity cost

28 RF-Controlled Hollow Cathode Two Candidate Configurations: Electric Field Magnitude ~96% of incident RF power absorbed, the majority within 2 mm of tip Models of RF power deposition in the coaxial RF power injection approach demonstrate: High efficiency power deposition, but mostly near the center conductor Currently studying alternate geometries to deposit more power downstream RF Power Absorbed

29 Modeling of RF Power Deposition EM wave propagation in 3D; RF Module in COMSOL Multiphysics Conductivity profile from classic resistivity and baseline parameters from Goebel and Chu Inputs from profiles of plasma density, neutral density, and T e Calculations validated against a second cathode type. RF wave behavior and RF power absorption in single-stage configuration stinger or power projection?

30 Model Regions for Coaxial RF Injection

31 Meshing Sensitivity Smaller elements result in less than 0.1% change in absorbed power

32 Power Absorption Profile Electric Field Magnitude ~62% of power absorbed within 2 mm of the inner coaxial conductor tip RF Power Absorbed

33 Results: RF Power Absorption Localized, high absorption: 96% of incident RF power absorbed, majority is within 2 mm of tip Absorption is dependent on specific conductivity profile (cathode specific) and distance from orifice (i.e. end of inner conductor) Does not project RF waves, but acts as a stinger

34 Results: RF Heating of Insert Low absorption: 3% of RF power absorption without a plasma Significant RF heating prior to plasma ignition is unlikely

35 Results: RF Breakdown Electric field maximum stays within a relatively small range for widely varying incident powers Strong electric fields at tip (> 1.5e5 V/m), Higher neutral pressures (> 1 Torr), And microwave frequencies (2.45 GHz) at reasonable incident powers (100 W) allow the RF-CHC to breakdown xenon reliably (Lisovski, Technical Physics, 1999)

36 RF-CHC Summary Results to date: Localized, high absorption possible with single-stage configuration Inner conductor acts as a stinger RF heating of insert prior to plasma ignition is unlikely due to low power absorption RF breakdown of xenon gas is feasible within typical baseline parameters Areas to be investigated: Erosion of inner conductor? Higher frequencies allow for power projection? Different RF-CHC configurations Different cathode geometries

37 Conclusions Plasma-materials interactions are central to the operation of cathodes, a key component in many devices of interest Fundamental studies of the material properties, transport processes, and surface kinetics can yield the understanding required to properly design cathodes Combined experiments and modeling are required Novel cathode concepts and new materials can improve cathode life by exploiting our evolving picture of plasma-material interactions Recycling of lanthanum in LaB6 cathodes can be exploited to reduce erosion of the high temperature emitter Localized, high power absorption is possible in cathodes with a coaxial RF feed, showing potential for RF-control of the plasma contact area RF heating of insert prior to plasma ignition is unlikely due to low power absorption Inner conductor acts as a stinger--rf ignition of xenon discharge is feasible within typical baseline parameters Planned work for remainder of FY12: Detailed modeling of La and B transport in LaB6 cathode plasmas and identification of methods to minimize erosion Experimental proof-of-concept demonstration for the RF-controlled cathode