We see images of Tokamak plasma with all sorts of colours from red to purple. Why do we see any light at all, since the plasma should be so hot to have dissociated all its electrons? It is all from contamination or unwanted cooling?

  • $\begingroup$ You might be interested to know that questions of this type (determining the wavelengths given off most intensely for materials heated to various temperatures) is what led to the foundation of quantum mechanics... $\endgroup$ Commented Dec 16, 2018 at 17:03

4 Answers 4


The sun is mostly plasma, and it almost fits well the black body radiation curve, which is the way we gauge its average temperature: almost 6000K which radiates away. The center is much hotter ( see below).


It is reported that the ITER plasma will be ten times hotter than the sun at the center. This will squeeze the black body type curve towards high frequencies, but there will still be a tail in the visible at the outer cooler layers, to give this type of visible radiation:

In the depths of the Sun where fusion reactions produce the energy that we perceive as light and heat, temperatures reach 15 million °C. In the centre of the ITER plasma temperatures will soar to between 150 and 300 million °C.


In the heart of the Korean tokamak KSTAR, in operation since 2008, a plasma pulse burns brightly. But don't be fooled—the brightest areas of the photo are in fact the coolest. At 150 million °C (the temperature in the centre), the plasma doesn't emit in the spectrum of visible light. © National Fusion Research Institute, Korea.

So the black body curve is really squeezed to the left.

In general from a temperature and over there will be radiation in the visible from black body radiation, but for a plasma there are differences..

The theory of black-body radiation in thermal equilibrium with a homogeneous and isotropic plasma is presented. The relevant thermodynamic quantities of the radiation are obtained. The presence of the plasma changes qualitatively and often quantitatively the concept of black-body radiation, essentially through the effect of plasma density.

Edit after comments.

I used this black body calculator for 150.000.000 K


and this for comparison with sun temperatures

bb her

This is what I mean by "squeezed to the left" beyond the visible.

  • $\begingroup$ (1/2) The analogy with the Sun is inaccurate, for two reasons. First, the optical thickness is very, very different (tokamak plasmas are optically thin, while the solar photosphere is optically thick), so in tokamak plasmas, there simply isn't enough radiating material in a column to see the low-intensity blackbody tail. Second, the degree of ionization is very, very different (the degree of ionization of the photosphere is less than 1 part per thousand, see astronomy.stackexchange.com/questions/7883/…) $\endgroup$ Commented Dec 6, 2018 at 16:57
  • $\begingroup$ (2/2) while the degree of ionization in the core of a tokamak plasma is quite close to 1. If your explanation were correct, the edge of the plasma would either be blue, or there would be a rainbow-like effect, getting redder towards the edge and bluer towards the core. We don't see this; instead, the optical signal is usually pretty monochromatic ("pinkish" is often reported), which is consistent with recombination on the edge of the plasma. $\endgroup$ Commented Dec 6, 2018 at 17:05
  • $\begingroup$ @probably_someone I am sorry, but if you look at the black body radiation, the tail is low as the curve shifts to the left, (assuming a black body curve which will be distorted for such high temperature plasma). It will be in intensity at the far right as the infrared and microwave from the sun, very weak. The edge in the center would not be blue, but black. It is blue at the edges as the image shows, where the temperatures are lower. $\endgroup$
    – anna v
    Commented Dec 6, 2018 at 17:42
  • $\begingroup$ The black body formula in princi[ppe takes into account all the various electromagnetic quantized modes in an overall estimate. That is why I gave a link for plasma and black body, as there are changes. $\endgroup$
    – anna v
    Commented Dec 6, 2018 at 17:51
  • 1
    $\begingroup$ Black body radiation has nothing to do with what determines the color of the edge of the plasma. It exists, of course, but it is not a significant contributor compared to transitions and recombinations from partially-ionized gas. Besides, the fact that the glow is strongest at the edge is totally inconsistent with how blackbody radiation behaves. Decreasing the temperature decreases the intensity at all wavelengths. $\endgroup$ Commented Dec 6, 2018 at 18:02

As Maury Markowitz mentioned, the color of the edge of the plasma (the core is colorless and transparent) is determined by the composition of the partially-ionized gas that is recombining with plasma electrons and undergoing transitions at the edge of the plasma. Sometimes this gas is purposely injected (as was the case with DCX and xenon gas), and sometimes it is simply residual gas in the chamber. For the tokamak MAST, it is the latter; since the plasma is made of deuterium, the optical radiation consists of the spectral lines of the dueterium atom, which gives a pinkish glow to the edges of the plasma.

This is consistent with the explanation given by the ITER tokamak's public pages here: https://www.iter.org/newsline/258/1512

  • $\begingroup$ dang how did I miss this page on my searches! Thanks that answers my question. Always wondered how they get these images from such a dangerous area. I appreciate there's barely any mass involved but there must be some serious ionising radiation in there that cameras need to be hardened to? $\endgroup$
    – Gonzonator
    Commented Dec 16, 2018 at 15:23

In most cases it's a fake: they often use test gasses with the explicit goal of being partially ionized so they give off light and thus allows us to take fancy photographs. The ultimate example of this was the DCX at the 1958 Atoms for Peace where they injected xenon (IIRC) to make a visible glowing ring. Its also possible to take pictures during startup, when it isn't entirely thermalized. There are some youtube videos showing the MAST startup and you can see this happen.


As you correctly state, a plasma is composed by a certain density of charged particles (ions, electrons,...). Due to many different reasons, such as the presence of external and internal (self consistent) electromagnetic fields, these charged particles are moving under the action of various forces. It is possible to show that a charged particle that is accelerated will be emitting electromagnetic radiation (not only in the visible spectrum).

Emission of electromagnetic radiation due to the accelerated motion of charged particles in a plasma is often an important part in the energy balance of a plasma system (as the Tokamak configuration you are talking about).

  • $\begingroup$ I don't think this has to do with the color in the visible spectrum. ITER seems to think that the color is from recombination at the edges of the plasma: iter.org/newsline/258/1512. In fact, the core of the plasma is basically colorless. $\endgroup$ Commented Dec 6, 2018 at 15:47
  • $\begingroup$ @probably_someone Actually, emission of em radiation comes from many different factors. The inner core of the plasma is colorless because the radiation that is emitted with the mechanism I am considering can also be reabsorbed by other regions of the plasma. Recombination at the boundaries is just another mechanism, in my answer I was assumed a more idealized version of a plasma, neglecting boundary effects. $\endgroup$
    – JackI
    Commented Dec 6, 2018 at 15:53
  • $\begingroup$ The issue is that the core of the plasma is not only colorless, but also transparent. If there was significant absorption as you suggest, then external light would also be scattered, so the plasma would be translucent. $\endgroup$ Commented Dec 6, 2018 at 16:07

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.