These two good answers (one and two) to the question "Is there a maximum Isp for 'exothermic chemical reaction rockets?'" put the limits around 500-550 seconds for the limits of practical chemical reaction engines and in the 700's for beyond practical.

So when I saw the last line in the table below, I become very interested. This is not an "Electric Solid Propellant Rocket" as described in this excellent answer to How does an Electric Solid Propellant rocket work?, and it flies in the face of Elon Musk's oft-quoted all modes of transport will become fully electric with ironic exception of rockets. (see comments)

So to which kind of alternative engine design might the last line of this table refer? Is it something that has ever been tried, or just hypothetical? Does it really refer to an electrical arc passing through monatomic hydrogen? That's 13.6 eV to ionize each proton, which is quite a lot of work in the physics sense of the word, therefore it's not really likely to be a conventional ion propulsion engine (is it)?

Can it be miniaturized? For example, a 3U cube sat with plenty of solar panels, some space-rated batteries, and a small tank of hydrogen gas (diatomic) - could this kind of propulsion be implemented in that setting?

        Fuel          Isp(seconds)
gun powder               350 
kerosene + LOX**         360
LH2 + LOX**              462
H2 + Nuclear             800 
H + electrical arc      1600 
  *Measured on the surface of the Earth
  **LOX: Liquid Oxygen

table of some unusual Isp values from California State University Long Beach's course ENGR 370I, Astronautics and Space

above: Table from this page from California State University Long Beach's course ENGR 370I, Astronautics and Space.

  • 1
    $\begingroup$ Could it be en.wikipedia.org/wiki/Arcjet_rocket? Ion engines are scalable to cubesat formfactors. $\endgroup$
    – Antzi
    Commented Oct 27, 2016 at 1:14
  • $\begingroup$ @Antzi I see hydrogen and 1600, so that paragraph may be coming from the same place (a non-technical blurb has been archived by the Wayback Machine), but I don't think it should be called an ion engine. The propulsion comes from gas expansion from heat I believe (though not sure yet), not from electrostatic acceleration. $\endgroup$
    – uhoh
    Commented Oct 27, 2016 at 1:24
  • 1
    $\begingroup$ @uhoh: Wikipedia article on Arcjet quotes University of Stuttgart's Institute of Space Aviation Systems Hydrogen-based arcjet with exhaust speed of 16km/s. Divide that by g and you have 1600s of ISp. Also, the article has sources links. $\endgroup$
    – SF.
    Commented Oct 28, 2016 at 8:02
  • $\begingroup$ @SF. the link is dead! The Wayback Machine Archived (and so untraceable) text contains a different number, a different material (solid teflon not monatomic hydrogen) and describes an ion engine, not an arcjet. $\endgroup$
    – uhoh
    Commented Oct 28, 2016 at 10:14
  • 3
    $\begingroup$ Regarding Musk's oft-quoted "all modes of transport will become fully electric with ironic exception of rockets", that once again is a case of people reading far too much into far few words. Musk was referring to the kinds of rockets needed to launch a vehicle into orbit (or beyond) from the surface of the Earth. Electric propulsion is a non-starter in this regard because all existing and theorized electric propulsion schemes only operate in vacuum and have ridiculously low thrust-to-weight ratios. $\endgroup$ Commented Oct 28, 2016 at 10:33

4 Answers 4


The specific entry of the table appears to be the HIPARC-R hydrogen arcjet thruster developed by Space Travel Institute of University of Stuttgart.

enter image description here

The concept of Arcjets is to use the propellant as conductor between two electrodes, creating intense electric arc, and exciting the propellant into superheated plasma. This allows to infuse it with more energy than a chemical reaction would provide - turning electricity into heat.

Since there is no inherent limit on how much electricity one can push through the plasma, there is no theoretical limit on the energy of the gas and as result, the exhaust velocity and specific impulse. The practical limit is of course of engineering nature - the electricity won't heat only the plasma, and superheated plasma is very aggressive against any structures it passes through. This both necessitates cooling of the whole thruster and limits the applicable energies to levels at which the thruster won't sustain critical damage from operation.

Regardless, the energies used are massive - HIPARC-R operates around 100 kilowatt, so even disregarding the scale of the thruster itself comparing to cubesats (about 3U) or the propellant forbidden on cubesats there is no way to keep it powered with a cubesat.

The Space Travel Institute developed also other, smaller arcjets. For example, ATOS operating on 750W, with 480s of specific impulse, 480 gram of own weight, 24 mg/s mass flow, 115 mN of thrust, and running on ammonia, might be applicable on 6U cubesats with extendable solar panels,

  • $\begingroup$ Excellent - experimental data with Isp > 2000 seconds! Is there a way to credit the photo? $\endgroup$
    – uhoh
    Commented Oct 28, 2016 at 13:44
  • $\begingroup$ @uhoh: source. Also look around their site. That's where ATOS data comes from too. $\endgroup$
    – SF.
    Commented Oct 28, 2016 at 13:52
  • $\begingroup$ Great - it might be good to include more permanently in the answer than in a comment - if you happen to be editing sometime. That looks like really helpful information! $\endgroup$
    – uhoh
    Commented Oct 28, 2016 at 13:57

This answer is supplementary information I've collected in the past few days. I've accepted the other answer because it's excellent and identifies the actual source of the "1600" mystery value.

note: Cubesat-Friendly Developments are discussed at the end.

The Principle

The idea is to heat a propellant using electrical current instead of chemical reaction producing even higher temperatures and therefore higher specific impulse.

For a fixed temperature, higher thrust velocity comes from lower mass species, so hydrogen-rich propellants are desireable. Hydrazine is populare both because of its hydrogen content and good space experience and often pre-existing availability within spacecraft. Ammonia is another option.

The random kinetic energy of the hot species is converted to cold, directed kinetic energy upon expansion. For a given temperature, lower mass results in higher veolocity.

$$ E = \frac{1}{2}m v^2$$

$$v = \sqrt{2E/m}$$

Propellant at a rate of the order of 100 mg/sec passes through a restriction of the order of 1 millimeter in diameter where an electrical current of order 10-100 Amperes passes through the dense, almost completely ionized gas. The ohmic heating of the plasma through electron-ion (and electron-atom) collisions transfers roughly a third of the power into direct heating of the ions to temperatures of 10,000 to 20,000 K. The hot plasma then expands, cools, and passes out the nozzle as directed thrust.

Because of the complex interaction of the plasma, current, and heat flow, a "sweet spot" exits in the 1kw to 30 kW range, and this is where most of the work has been done. The engine requires significant (and heavy) power conditioning and thermal isolation from the spacecraft, since more than half of the electrical power results in heat of the engine components.

enter image description here

image above and text below: from Status and Prospects on Nonequilibrium Modeling of High Velocity Plasma Flow in an Arcjet Thruster, Hai-Xing Wang, Su-Rong Sun, Wei-Ping Sun, Plasma Chem Plasma Process (2015) 35:543–564 DOI 10.1007/s11090-015-9610-4

"The key physics of the constricted arc discharge in arcjet are depicted in Fig. 1. An arc is stuck between the central, conical-tipped cathode and the coaxial, nozzle shaped anode. Working gas, injected with high swirl velocity near the cathode tip, passes through the constrictor region and is ohmically heated by the arc. The energy transferred to the gas is thought to result predominantly from electron–ion or electron–neutral collisions as elec- trons are the dominant current carriers. Extremely small constrictor size, extremely high gas velocity at the nozzle exit and operation at relatively low arc current are a few of the primary features of these kinds of thrusters.

Typical low power arcjets have conical converging–diverging nozzle with constrictor diameter on the order of 0.5 mm, expansion half angle of 20°, and exit diameter of 3.5 mm [8–10]. The physical characteristics of the arcjet flow field vary from a nearly fully ionized plasma with temperature in excess of 20,000 K near the cathode tip to a relatively cold plasma (1,000–2,000 K) at the anode wall. Moreover, velocities vary from approximately 10 km/s on centerline to zero at the wall.*" from here (paywalled)

There are many examples of arcjet thrusters in the 500 to 2000 Watt range so I'll talk about those first.

enter image description here

above: From Performance Computation of a Low-Power Hydrogen Arcjet Kazuhisa Fujita & Yoshihiro Arakawa, J. Propulsion. and Power v15, n 1, Jan-Feb 1999 (paywalled)

NASA started the Arcjet Thruster Research and Technology (ATRT) program in 1984. For example:

enter image description here

above: illustration of the arcjet principle from here.

enter image description here

above: Illustration of an arcjet start-up process, including spiral-fed propellant. Figure 3-6 here.

The Wikipedia article List of spacecraft with electric propulsion currently lists a few spacecraft that carried arcjet thrusters. The Telstar 401 is the first commercial use, launched in 1993. Four arcjet thrusters were used for North-South and East-West station keeping in Geosynchronous orbit.

Telstar 401,                MR-508, Hydrazine, Comms (AS-7000)
Telstar 402R (Telstar 4),   MR-508, Hydrazine, Comms (AS-7000)
A2100,                      MR-510, Hydrazine, Comms
ARGOS (P91-1),              ESEX,   Ammonia,   Experimental Military 
AMSAT-Phase 3-D (OSCAR-40), ATOS,   Ammonia,   OSCAR Sat cold gas mode

According to Performance and preliminary life test of a low power hydrazine engineering design model arcjet (Tang et al. 2015, Aerospace Science and Technology, v 15, n 7, Oct–Nov 2011, 577–588)

Telstar 401 (a geostationary communication satellite of Lockheed Martin Corporation) equipped with PRIMEX Aerospace Company’s (originally Rocket Research Company, RRC) MR-508 hy- drazine arcjet propulsion system was successfully launched in 1993, which is the first application of thermal arcjet propulsion systems (11). From then on, hydrazine arcjet propulsion systems have operated successfully on more than 29 spacecrafts, showing substantial performance and reliability (12). (my emphasis)

MR-508, -509, -510 series Hydrazine Arcjets

The MR-508, -509, -510 series of Hydrazine arcjet thrusters can be seen at Aerojet Rocketdyne - here is a set of four MR-510 and the Power Conditioning Unit (PCU). They provided an Isp above 500 seconds using hydrazine, and therefore provided more efficient use of the hydrazine mass.

However, in this power range, the power supply and control electronics can be much heavier than the engine itself!

enter image description here

Flight Qualification of the 2.2 kW MR-510 Hydrazine Arcjet System

ARC-1, ARC-2, ARC-3 (use Lockheed Martin A2100 bus)

ESEX Ammonia Arcjet



ESEX Arcjet report (ARGOS)

AIAA 99-2706, An Overview of the On-Orbit Results from the ESEX Flight Experiment

enter image description here

above: ESEX Arcjet from here

enter image description here

above: Propellant supply system for ESEX (Electric Propulsion Space Experiment) from here on ARGOS.

Cubesat-Friendly Developments

Discharge Characteristics of a Very Low-Power Arcjet, Hideyuki Horisawa, Hotaka Ashiya and Itsuro Kimura, presents results of an experiment to understand the issues of scaling an arjet thruster down to the 1 to 10 Watt range and about 4 centimeters in length. A quartz window was included so that the plasma in the restriction region could be observed as experimental parameters were varied. In these early studies, "typical thrust was 1.5 ~ 2.0 mN, Isp: ~ 100 sec for input power of 1 ~ 5 watts and propellant (N2) mass flow rate of 0.6 ~ 2 mg/sec", and based on these results the possibility of higher Isp in this power range can be explored.

enter image description here

enter image description here

Direct Drive

In Testing of an Arcjet Thruster with Capability of Direct-Drive Operation (A. K. Martin et al., NASA-Marshall Space Flight Center, American Institute of Aeronautics and Astronautics) an arject engine capable of being driven directly from a solar array (without substnatial power conditioning) has been demonstrated.

The testing indicated that an operating point exists within the I-V characteristics that is compatible with direct-drive solar-electric operation; for a flow rate of 20 mg/s (argon) the arc could be sustained at a voltage of about 20 V and a current of 25 A (500W).

enter image description here

enter image description here

Cubesats have several restrictions on chemical propellants and storage of chemical energy. One step towards addressing this problem has been discussed in Performance Characteristics of Low-Power Arcjet Thrusters Using Low Toxicity Propellant HAN Decomposed Gas Matsumoto et al. IEPC-2013-095, (33st International Electric Propulsion Conference, The George Washington University, USA October 6 – 10, 2013).

HAN or Hydroxylammonium nitrate is a potential new propellant. However early results showed significant corrosion with conventional materials.

There has been some work on production of hydrogen-containing vapor from solid Teflon [citation needed].

  • 1
    $\begingroup$ How do you get hydrogen containing vapor out of Teflon which is hydrogen free? $\endgroup$
    – ikrase
    Commented Dec 13, 2020 at 10:14
  • 1
    $\begingroup$ @ikrase excellent point! This is from four years ago and I have no idea where that came from. I'll strike it through but leave it for historical purposes (I love to highlight my mistakes). I wouldn't have written that if there wasn't something there, but we have to assume that they either were going to use carbon or fluorine ions, or that they were going to use a different solid like polyethylene... Hmm... $\endgroup$
    – uhoh
    Commented Dec 13, 2020 at 22:30
  • $\begingroup$ @ikrase maybe I'd read that the use of teflon to avoid Hydroxylammonium nitrate's issue with "significant corrosion with conventional materials" and misconstrued it as being used as the propellant. Not sure right now. $\endgroup$
    – uhoh
    Commented Dec 13, 2020 at 22:34

Here is a supplementary answer addressing the last part of the OP's question about the possibility of miniaturization and use specifically in cubesats.

Cubesats have several restrictions on chemical propellants and storage of chemical energy. One step towards addressing this problem has been discussed in Performance Characteristics of Low-Power Arcjet Thrusters Using Low Toxicity Propellant HAN Decomposed Gas, Matsumoto et al. IEPC-2013-095, (33st International Electric Propulsion Conference, The George Washington University, USA, October 6 – 10, 2013).


Although hydrazine (N2H4) is used as a propellant of spacecraft thrusters, it is high toxicity liquid. Spacecraft researchers need a low toxicity propellant. Hydroxyl Ammonium Nitrate (HAN: NH3OHNO3) is proposed as a low toxicity propellant. In this study, the possibility of HAN as a propellant of direct-current arcjet thrusters is examined. Using the HAN-simulated-decomposed gas of H2O, CO2 and N2 mixture, the hydrazine-simulated-decomposed gas of N2 and H2 mixture, and N2 itself, an arcjet thruster was operated, and the basic characteristics were obtained and compared. The thrust and the thrust efficiency with the HAN decomposed gas were 182.6 mN and 10.3 %, respectively, at a specific impulse of 156.4 sec with 1.36 kW input power, although severely-eroded electrodes was observed, and their performance was slightly low compared with the hydrazine decomposed gas.

  • $\begingroup$ What is a source of CO2 from NH3OHNO3 ??? $\endgroup$ Commented Dec 14, 2020 at 8:42

Antzi's correct, that's an arcjet. Like ion engines, specific impulse is really good, but thrust is extremely low. This article covers some of the state of the art as of 2013 (spoiler: 100 milliNewton at 800 watts.)

Looking around, I see references to hydrogen arcjets achieving between 1200 and 2000s specific impulse.

Aerojet has some small arcjet thrusters that would fit in a cubesat (linked datasheet has a couple of arcjets in addition to Hall effect and other units) -- but their power conditioning units (scaled for ~2kW operation) wouldn't fit. Those are lower-efficiency units using hydrazine at ~600s specific impulse.

Anode/nozzle erosion seems to be the biggest drawback, with run times limited to the order of 1000 hours -- that seems like a long time, but you need it at those thrust levels, and ion engines can run for at least 10 times as long.

  • $\begingroup$ Any idea where the Isp of 1600 might be coming from? Do you think this technology is potentially 3U-cubesattable? I'm not asking if that is a good thing or not here, being atomic hydrogen. Oh, does the 1600 number depend specifically on obaining H i.e. atomic hydrogen? This link only discusses a technical increment, but doesn't really explain much. $\endgroup$
    – uhoh
    Commented Oct 27, 2016 at 5:56
  • $\begingroup$ Actually, I've spent the day at the library, so I'll post a substantial answer tomorrow. $\endgroup$
    – uhoh
    Commented Oct 27, 2016 at 10:31
  • 3
    $\begingroup$ You can put an arcjet in a cubesat, but you can't get 800 watts into it. 10cm x 30cm of solar panels gets you at most about 39 watts. If thrust scales with wattage that's 4.8 mN. $\endgroup$ Commented Oct 27, 2016 at 13:09
  • $\begingroup$ Sticking with Russell Borogove's last comment, if you have a battery on board you could manage higher wattage pulses, i.e. higher thrust pulses, which might suit some folks but they'd still be limited to the same assumption of 39 Watts average electrical power. $\endgroup$
    – Puffin
    Commented Oct 27, 2016 at 17:36
  • 2
    $\begingroup$ Pulses simply aren't useful for electric thrusters; they have to work over very long duty cycles to yield any significant delta-v. $\endgroup$ Commented Oct 27, 2016 at 18:55

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.