On videos of Soyuz docking to ISS we can see exhaust from the control engines (if background is dark enough). For example here

Soyuz TMA-19M docked to ISS at December 15 2015. Exaust plume is circled at screenshot:

enter image description here

The textbook chemical formula of the NDMH and nitrogen tetroxide reaction is:

H2NN(CH3)2 + 2*N2O4 -> 2*CO2 + 3*N2 + 4*H2O

My question is:

How are visible ice crystals are forming in the exhaust? The molecules of exhaust products should be too hot and moving too fast... Or any other products occur in the reaction?

Also I wonder about the contamination of spacecraft by hypergolic exhaust products. The Hubble telescope has no reaction engines because of this concern. The ISS has limitations for use of reaction engines in close proximity. But if exhaust products are nitrogen, carbon dioxide and water - can they cause a long-term contamination, not just icing? Or they can react with metals because of their high speed?


I have read "Rocket propulsion elements" by Sutton and Biblarz (google "rocket propulsion elements pdf"). A lot of interesting info, including chapter 18 about exhaust plumes. But not so much for my question.. About contamination of spacecraft the book only mentions "hydrazinum nitrate and other materials were found".

So, to reformulate the question:

  • What are the visible products of reaction engines' exhaust in the video? I suppose they should be in solid or liquid phase, because a gas could not scatter enogh light.


Summarised some points in self answer

  • 1
    $\begingroup$ On videos of Soyuz docking to ISS we can see .... Can you add a link to one of these videos? $\endgroup$ Jun 4, 2018 at 21:54
  • 1
    $\begingroup$ Yes it will be helpful to add a link to an example video. Also there's no reason to expect that chemical reactions are 100% complete, there could be incomplete reaction products as well, even traces of unreacted material. Also, things don't stay "hot" as they expand and radiate in the vacuum of space for \very long. $\endgroup$
    – uhoh
    Jun 5, 2018 at 1:40
  • 4
    $\begingroup$ 1) Who said those are ice crystals? 2) Thruster firings, especially short burps like that, often leave a fair amount of unburned propellant that can contaminate ISS surfaces, especially when it gets baked on by UV light from the sun. The restriction on engine firings near ISS are also due to plume loads and mechanical damage from propellant droplets. $\endgroup$
    – Tristan
    Jun 5, 2018 at 16:42
  • 1
    $\begingroup$ Something to keep in mind: Think the rocket burn is perfect? Perfect mixing so every bit of both chemicals is reacted? $\endgroup$ Jun 6, 2018 at 2:47
  • 2
    $\begingroup$ See this NASA paper about bipropellant thrusters with high speed droplets of unburned and partially burned propellant. $\endgroup$
    – Uwe
    Jun 7, 2018 at 17:14

3 Answers 3


Well, after useful comments and some research I can summarise:

The chemical reaction of bipropellant hypergolic fuel is more complex:

1) There are more products of the reaction

2) Some part of the fuel remains unburned

In this link by @Uwe is stated that exhaust plume contains microdroplets of unburned propellant, with size up to several micrometers. Quote:

Contaminant mass is predicted to be less than 2% of total propellant

In this paper a formula of hypergolic bipropellant burning products is given as:

4a*CH3N2H3 + 5*N2O4 -> np1*CO2 + np2*H2O + np3*O2 + np4*N2 + np5*NO + np6*CO + np7*OH + np8*H2 + np9*O + np10*H + np11*N + np12*NO2 + np13*H2O2 + np14*HO2 + np15*HNO

where a, np1-np15 are coefficients to be found. But unburned propellant is not taken into account there.

In this pdf the basics of plume impingement are given. It has a picture of unburnt propellant droplets inside engine's nossle:enter image description here Also interesting detail - exhaust plume in vacuum has "back flow", so small part of it spreads in opposite direction and it should be taken to account as contaminant:enter image description here

In recent (Jun 8 2018) video of Soyuz MS-09 docking to ISS not only plumes could be seen, but also low-speed ice particles (start at 6.15)

I suppose that particles are frosen droplets fron engine's nozzle.

I would say the question is partly answered. But there is one thing that I have completely forget:

The first stage of Falcon-9 rocket has nitrogen thrusters. So, the plumes of them should contain only nitrogen. But on videos we see the plumes of the nitrogen thrusters! So, my initial suggestion that gases like nitrogen form ice crystals in exhaust plume is correct. But I wonder - how it happens?

This question deals with it. Links there are interesting, but not completely relevant, I think. The nitrogen is in gaseous phase when it exits a nozzle (of nitrogen thruster or hypergolic thruster). So it's rather "conterintuitive" that the gas spreading in vacuum forms ice crystals. The gas "should" only spread ever more in vacuum, not condense..

At this moment I could not find any reference on physics of this process. If anybody could explain it, the question would be completely answered.

  • 2
    $\begingroup$ On Earth compressed nitrogen may be liquified by expansion. This should be possible in space too. If the temperature is very low, you may get solid nitrogen. The LOX for rocket engines is made by compressing air, liquifying it by expansion and separating it by destillation. Without that process, no cryogenic propellants would exist. $\endgroup$
    – Uwe
    Jun 9, 2018 at 16:19

The bright particles you see flying from the Souyz nozzle are the pieces of the "nozzle cowling". It is kinda ceramic material coating of the nozzle. The amount of coating is calclulated for the nozzle operation. It wears off gradually during the nozzle serving time.

  • 2
    $\begingroup$ Welcome to space! Do you have a reference to back up your assertion? $\endgroup$ Jan 6, 2020 at 20:38

In-short, the answer is that at the beginning and especially end of an RCS thruster firing, there are clouds of unburned reactants. Source (for the shuttle, which has a very similar chemistry):

In a discussion with a NASA aerospace engineer familiar with the space shuttle reaction control system, I learned that the thrusters never generate any light while operating, but they always emit a small cloud of unburned propellant just before the thruster fires and a much larger cloud immediately after the thruster shuts down.


When the valves are closed to shut the thruster down, small amounts of propellant are trapped in the tubes between the valves and the combustion chamber. [... In] the vacuum of space, the "microscopic snow" also forms after shutoff just as it does at startup. But[, in this case,] the dribble volume is large enough that the snow generated can be seen as a white plume in reflected sunlight. It is totally invisible without some external source of illumination.

Note: there are a few alternate theories I came across which, while apparently less-plausible, may still at least contribute to the effect. I have seen (less-credible) sources suggesting that either the fuel or the oxidizer is injected first (for safety purposes, presumably), leading to unconsumed propellant/oxidizer, and that in longer OMS burns the plume is invisible. Also, CO₂ ice would be present in addition to water ice.

  • 2
    $\begingroup$ Um, your source is a website called "the living moon" which claims to have pictures of buildings on the moon, etc. It is a fringe website and an unreliable source. $\endgroup$ Jun 9, 2018 at 2:28
  • 1
    $\begingroup$ @OrganicMarble Okay fair enough, that was the only article I skimmed from the website, and the root domain indeed looks pretty bonkers (I saw this explanation a couple other places too, but this seemed to be perhaps the original source). I contend that the explanation still makes sense, though. $\endgroup$
    – imallett
    Jun 9, 2018 at 5:12

Your Answer

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

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