Field Emission Electric Propulsion (FEEP)

and Colloid thrusters

FEEP and colloid thrusters are particularly suitable for missions requiring thrusts in the sub-millinewton level with precise control abilities. FEEPs feature unmatched performance at the µN thrust level, and is unique for some highly demanding applications like drag-free control of scientific spacecraft and disturbance reduction of microgravity platforms. Colloids provide more thruster with their lower specific impulse.


In FEEP and colloid space rockets, unlike most ion electric propulsion engines, ions and droplets are directly extracted from the liquid phase. The thrusters can accelerate a large number of different liquid metals: For FEEPs, indium is usually selected for its high atomic weight, low ionization potential, relatively low melting point (m.p.= 156 °C), and good wetting capabilities on the emitter substrate. Colloid thrusters use polyorganics with much lower electrical conductivity, higher specific mass and lower and surface tension. The thrusters' main features are:




Specific impulse (seconds)



Thrust (N)

30 µN – 50 mN

1 µN - 1 mN

Common propellant

Organic liquids


Thruster mass (g)



Impulse bits (nN-sec)

1,000- ~1 million


Prop. electrical conductivity (S/m)



· Absence of moving parts - no valves, no pressurized gases
· Self-contained propellant reservoir

They generate thrust by application of a strong electric field to pull liquid propellant off a tungsten needle. The potential on the accelerating electrode is normally -1,000 to -6,000 V, generating a field at the tip of about 1 V/nm. The thrust generated can vary significantly depending on whether ions or droplets are emitted from the needle tip. Fig. (1) displays two main types of thruster designs, a capillary and needle setup.

Figure 1. Field Emission Electric Propulsion Capillary [1] and Needle [2] Thrusters

For the capillary design (1c), the emitter module consists of two metallic plates with a small propellant reservoir. A sharp blade is accurately machined on one side of each plate. A thin layer of material is sputter-deposited on the other three sides of one of the plates, to act as a spacer; when the two emitter halves are tightly clamped together, a slit of about 1 µm is left between the blades. The propellant flows through this tiny channel, forming a free surface at the exit of the slit with a radius of curvature in the order of 1 µm. Under a strong electric field generated by the application of a voltage difference between the emitter and an accelerator electrode located directly in front of it, the free surface of the liquid metal approaches a condition of local instability, due to the combined effects of the electrostatic force and the surface tension. A series of protruding cusps, or "Taylor cones", are created. When the electric field reaches a value of about 109 V/m, the atoms at the tip spontaneously ionize and a thrust-producing ion jet is extracted by the electric field, while the electrons are rejected in the bulk of the liquid. For the needle design (1d), the emitter is a single needle with a pool of indium melted in a reservoir with an external heater. The accelerating electrode is the circular ring above the needle.

Figure 1c.) Experimental slit FEEP. Figure 1d.) Experimental needle FEEP

Current and future uses

A number of satellites for scientific missions, for example SMART-2, Darwin, and GAIA, will require FEEP technology to enable these craft to be positioned with mm and 10 m-arcsec accuracy [3]. An upcoming mission designed around field emission propulsion is the Laser Interferometer Space Antenna (LISA) [4]. The primary objective of the LISA mission is to detect and observe gravitational waves from massive black holes and galactic binary stars in the frequency range 10-4 to 10-1 Hz [5]. Useful measurements in this frequency cannot be made on the ground because of the unshieldable background of local gravitational noise.

Ongoing research issues

Sufficiently high electric fields applied to the liquid meniscus tend to pull it into a pointed shape, or Taylor cone. For suitable values of the applied voltage and liquid flow rate (over 109 V/m and 10 ng/s respectively), a steady jet is emitted from the cone apex. Cone formation times vary from 10-8 to 10-6 seconds for low impedance needles (inertial effects dominate) to 0.1-1s for high impedance needles (flow effects dominate) [6]. Figures 2 and 3 show schematic and microscopic views of Taylor cones.

In a Liquid-Metal Ion Source (LMIS) thruster, it is desirable to extract a high current of over 100 µA. A feature of these ion sources is that at relatively large current levels, over 15 µA, instabilities occur and micro-droplets are emitted in addition to the desired ion current [7]. These instabilities are deleterious since mass efficiency decreases as the propellant is consumed at a higher rate and the droplets will deposit on the extractor electrode, limiting thruster lifetimes.

A critical element of our goal to create an accurate model of the plasma cone droplet size and relative charge.


Ongoing Research Approach

Simulating droplet snap-off is difficult because the region around the tip possess characteristics that cause most numerical methods to fail. Examples of these characteristics are the presence of very large dynamic electrostatic gradients, rapid temporal change, highly non-symmetric geometry and extreme fluid curvature. To partially address these issues, a simulation method known as level sets in employed to track the fluid-vacuum interface front.

In the work below, the level set model considers an incompressible, isothermal, viscous liquid. The propellant is treated as a perfect conductor and its atoms accelerated with ring electrodes, producing thrust. The governing 2D equations are:




Surface tension (γ) and viscosity (μ) are functions of the liquid temperature. Surface curvature (κ) and the stress tensor (Dliquid) drive the interfacial pressure balance. Using these equations, a mathematical approach called the boundary integral method rapidly and accurately computes the normal electric field (En) by only considering the edge of the problem, thus allowing the surface to be advected forward.

Some simulations of propellant on a needle FEEP being accelerated upward are below. More details are in the papers on the earlier page.


[1] Centrospazio (Italian) capillary FEEP thrusters, see here for description of concept; and here for operation principles.

[2] Austrian (ACRS) FEEP thrusters, see here for rough description.
[3] Franks, A. et al. "A design study of a microthrust balance for space applications," 5th Joint Tokyo-Warwick Biennial Nanotechnology Symposium. 3-5 Sept. 1997, Noda, Japan.
[4] JPL website
[5] Saccoccia, G. and Berry, W., "European Electric Propulsion Activities and Programmes," Acta Astronautica, Vol. 47, 2000, pp. 193-203.
[6] Mair, G.; Aidinis, C.; Bischoff, L. and Ganetsos, Th. “On the Dynamics of Liquid Metal Ion Sources,” Journal of Physics D: Applied Physics. Vol. 35, 2002, pp. 1392-1396.
[7] Tajmar, M. and Genovese, A. “Experimental Validation of a Mass Efficiency Model for an Indium Liquid Metal Ion Source,” Applied Physics A, submitted.


Funding for this work provided in part by NASA's Jet Propulsion Center (JPL).


VanderWyst, A.; Christlieb, A.; Sussman, M. and Boyd, I.D., "Simulation of Charge and Mass Distributions of Indium Droplets Created by Field Emission,", 37th AIAA Plasmadynamics and Lasers Conference, San Francisco June 2006

VanderWyst, A.; Christlieb, A.; Sussman, M. and Boyd, I.D., "Boundary Integral Formulation of Electric Fields in Level Set Simulations of Charged Droplets,", 36th AIAA Plasmadynamics and Lasers Conference, Toronto June 2005