I haven't seen much about colloidal or electrospray thrusters besides that Wikipedia article and a few questions here. The idea is similar to ion thrusters in that propellant mass is broken into small bits, ionized, and accelerated electrostatically. The difference is in the nature of the small bits.

Ion thrusters can achieve high ionization efficiencies, most propellant atoms that leave are successfully ionized and accelerated.

However If I understand correctly, a colloidal thruster only breaks the propellant into tiny droplets and ionized and accelerates those. The maximum charge you can put on a tiny droplet or solid particle is only a small fraction of the number of electrons because the Coulomb force is so great. I don't know exactly, but perhaps on the order of one charge per one thousand atoms give or take a power of ten, though it will depend on the droplet's or particle's size and nature.

With such a low ionization efficiency, would colloidal thrusters have any advantage over ion thrusters? Simplicity perhaps? Lower weight? Or is my understanding of the technology flawed? Perhaps electrospray thrusters and colloidal thrusters not the same thing?

Related questions:

  1. Relationship between part dimensions and performance of ion electrospray thruster
  2. Generating an electric field for a colloid thruster

NASA News: MIT SPL delivers the Scalable ion Electrospray Propulsion System (S-iEPS) for CubeSats to NASA


1 Answer 1


One significant advantage, or disadvantage, depending on the way you look at it, is the minuscule thrust, way lower than in case of classic ion engines.

Another advantage, is the electrospray process produces ionized droplets out of liquid without the need for fairly complex electromechanical plumbing that first releases xenon from a pressurized tank, then ionizes the stream of it. This allows both for miniaturization (no moving parts, reduced number of parts) and assures 100% of the spray is charged (and to a fairly constant charge) - it's the charging process of the liquid surface that causes the separation, overcoming the surface tension, and the threshold of separation is quite stable (per propellant), and rapid, fine control over the rate of the process.

And as power consumption, similarly to ion engines, is proportional to specific impulse and mass flow, with mass flow this minuscule, it's quite moderate.

That means, that electrospray thrusters don't seem to be possible to scale up in performance like, say, VASIMR, nor will they allow near the same level of thrust, they are compact, they allow for very precise (and lossless) control of thrust (you just shut down the electrodes that cause the electrospray process and the droplets cease to separate), they don't put much of a strain on the craft's power source, and in effect they make wonderful RCS thrusters of extreme precision.

So, electrospray thrusters are not a very good idea for main propulsion - they could work with nanosatellites in a pinch (considering craft life time and mission duration makes them rather unattractive) they are THE option for station-keeping of ultra-precisely aligned constellations. eLISA, the gravitational wave observatory with 2.5 million km separation between the satellites, will maintain a couple millimeter, or smaller scale of positioning imprecision of the satellites (the remainder of required precision will be achieved through mechanical alignment of sensors withing the craft - more out of need to conserve the electrospray RCS fuel than out of need for better alignment precision using the RCS) but other proposed projects involve particle accelerators utilizing vacuum of space as medium of the "corridor", or telescopes of extreme focal length, where separate lenses or mirrors are mounted on separate satellites aligned to the optical path.

In short, they are a completely different kind of animal than typical ion thrusters - they are the designated ultra-precise RCS thrusters, and trying to compare their performance to dedicated main engines is a misguided endeavor focusing on entirely wrong set of properties.

  • $\begingroup$ Thanks for the great answer! For equal maximum thrusts, Ion thrusters probably have similar excellent precision, right? You can vary the current and/or acceleration voltage (energy) continuously or pulse them. So isn't the advantage actually just weight? For smaller satellites, wouldn't precision ion thrusters simply be heavier than electrospray thrusters? $\endgroup$
    – uhoh
    Commented Jul 14, 2017 at 14:54
  • $\begingroup$ @uhoh: You can't vary the mass flow nearly as precisely for ion thrusters though. You may change the amount of ionization / energy of the xenon rapidly, but the amount of xenon being injected is not as easily/rapidly controlled. This is lossy - you don't vary mass flow, you vary specific impulse (reducing it below optimum). And yes, size-wise you'd be worse off. Especially if you want to provide at least 12 thrusters for all degrees of freedom. $\endgroup$
    – SF.
    Commented Jul 14, 2017 at 15:02
  • $\begingroup$ Read my question again. If electrospray engines use droplets, then every single charge accelerated "costs" hundreds or thousands of atoms, compared to one atom for an ion thruster. Wouldn't you say this is much much more lossy? $\endgroup$
    – uhoh
    Commented Jul 14, 2017 at 15:05
  • $\begingroup$ @uhoh: 1: what percentage of xenon escapes the engine un-ionised? 2: the electrospray process is exactly the effect of "overcharging" the liquid - the droplets split into smaller ones until they can hold the charge; I was unable to find the droplet size, but the flow can be of order of 10^12 particles per minute, so the droplets will likely be closer to dozens of atoms than hundreds of thousands of them. $\endgroup$
    – SF.
    Commented Jul 14, 2017 at 15:22
  • 1
    $\begingroup$ @uhoh: follwoing some example data I'm coming up with ~19 coulombs per kilogram of propellant. Volt for volt, gram for gram, specific impulse will differ only with charge per unit of propellant mass. I'd still want to see how much charge does a kilogram of xenon receive when ionized, can't find any sources for that though. $\endgroup$
    – SF.
    Commented Jul 14, 2017 at 20:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.