So far I had no luck trying to find the visible absorption spectrum of $\rm CO_2$ anywhere, all I get is the far infrared absorption spectrum and stuff like that.

If you just search "what color is liquid $\rm CO_2$" it simply says everywhere that "it's colorless" but... That's also said a lot about other liquids such as water, but we all know that's not true: https://en.wikipedia.org/wiki/Color_of_water#:~:text=While%20relatively%20small%20quantities%20of,and%20scattering%20of%20white%20light.

Water is not colorless at all, it just appears colorless in small quantities since it's a very transparent liquid and one can only appreciate its slightly blue coloration in very large quantities such like in pools, lakes, oceans, etc. Water is not just blue because of Rayleigh scattering and the sky's reflection, it is also and mainly blue because it does in fact absorb more red and green light than it absorbs blue light, just like any other blue thing.

That being said, I don't think liquid CO2 and other liquids such as liquid SO2, NH3, CH4, etc. are colorless as it's said and as they are described everywhere.

However, going back to just liquid $\rm CO_2$. I believe it just appears colorless in small quantities, but in very large quantities some color is perceivable exactly just like it happens with water. But I might be totally wrong since that belief is based on absolutely nothing, that's simply how I guess it is since I have absolutely no idea and it's simply hard to imagine a liquid that totally ignores visible light without absorbing even the slightest bit of it at all.

But If that is the reality and it doesn't absorb visible light at all, then that would mean deep liquid $\rm CO_2$ would just end up appearing slightly blue. Much less than water does though, it would be a much clearer liquid and you would need way more deepness for the blue coloration to appear because this coloration would only come from rayleight scattering. I think this is the same that happens with liquid N2 based on previous research... But there is pretty much almost no information at all about liquid N2's color too.

CO2 is such a common thing in this universe that a lot of exoplanets out there might have vast liquid $\rm CO_2$ ocean, seas or some lakes since the conditions needed for a planet to allow it are not that crazy. That's why Im interested about knowing the answer to this question. I'm simply very curious about it and I want to have an slight idea of what would an hypotetic exoplanet with $\rm CO_2$ oceans look like. Before anyone mentions that I should ask this question somewhere else like in"wordbuilding"... I already tried there without luck. I was told to ask the question here too to see if some of you have an answer to this simple yet so seemingly hard to answer questions due to how impossible it is to find anything about liquid's $\rm CO_2$ visible absorption spectrum over the internet. Nothing in Reddit, Wikipedia, Google, Quora... There's absolutely nothing, and if there is, I wonder where because I searched non stop everywhere.

Sorry if there's any grammar mistake, tried to avoid making one and I believe it's all right.

Had this question unanswered for a long time, would love to finally get an answer.

  • 1
    $\begingroup$ If nothing else: liquid CO$_2$ only forms at very high pressures, so you would need some fairly special conditions for the atmosphere to be negligible. That said, such conditions are definitely possible, so the question is perfectly valid. $\endgroup$ Commented Jul 13, 2021 at 17:15
  • $\begingroup$ @EmilioPisanty Exactly! The right conditions for liquid CO2 are much harder to get when compared to the conditions you need for liquid water, but it's really not that bad. You would need at least something like a planet with a fairly thick 20 bars atmosphere and an average temperature somewhere in between -55°C and -20°C, or if the atmosphere is twice as thick, the temperature can be as high as 4°C still allowing stable liquid CO2 on the planet's surface. $\endgroup$
    – DeMooniC
    Commented Jul 13, 2021 at 17:36
  • $\begingroup$ aip.scitation.org/doi/10.1063/1.3549593 is the closest I can find, with some suggestion that the reflectivity in the blue is higher than at lower energies. But they are mostly focused on very extreme conditions (up to 1 TPa). $\endgroup$
    – Jon Custer
    Commented Jul 13, 2021 at 17:40
  • $\begingroup$ The only thing that makes it conciderably harder for a planet to achive this conditions is that a CO2 rich planet with a thick atmosphere would most likely mean an atmosphere also mainly made of CO2 which as we all know means a lot of greenhouse effect. I guess it's not that big of a deal if the planet is orbiting a cold star, something like a red dwarf, orange dwarf... But liquid CO2 oceans around hotter stars like our sun would be much rarer, as far as we are not concidering supercritical CO2 as valid of course. $\endgroup$
    – DeMooniC
    Commented Jul 13, 2021 at 17:45
  • $\begingroup$ @JonCuster Oh yeah, that might be a bit way too extreme. Im actually a bit surprised tho. Wouldn't CO2 at such crazy temepratures and pressures be a supercritical fluid instead of a liquid? $\endgroup$
    – DeMooniC
    Commented Jul 13, 2021 at 17:50

1 Answer 1


What I think makes sense

I'm ignoring for now the effects of the liquid phase, let's just go for the UV-vis spectrum of CO$_2$. For me this means electronic excitations. Somebody will have measured/calculated those, so let's look for papers. When I search for "electronic excitations CO2" in google scholar, I find this paper: "Vibronic effects on the low-lying electronic excitations in CO2 induced by electron impact" where the introduction states:

The low-energy part of the UV absorption spectrum of the molecule consists of two diffuse bands with maxima at excitation energies of E ∼ 8.4 and 9.3 eV.

and I think that's all we need to know. We can work with the idea that the molecule is so small and simple, and its bonds so strong, that the lack of absortion in the visible spectrum is not an approximation and instead the result of a lack of energy levels in that region of the spectrum. No levels at the right energy above the ground state means no resonance condition means no absortion of visible light.

But, there is this UV absortion. So what color does that leave us with? 9 eV, per the NIST converter are 7.26 x 10$^6$ m$^{-1}$. The inverse of 7.26 x 10$^6$ m$^{-1}$ are 1.38 x 10$^{-7}$ m, or 138 nm. This is, as indicated in the text, in the UV region (<380 nm) but not too far from visible. Since they are diffuse bands, we can assume they are rather broad and could contribute very weakly to some residual absortion in the visible region. The title of the paper also gives us an indication on how this energy region can be reached: through vibronic interactions, i.e. mixing between electronic and vibrational transitions. Above certain temperature, excited vibrational states will be populated, and a (quite forbidden but not zero-probability) transition will happen between the ground electronic state with high vibrational excitation and an excited electronic state with lower vibrational excitation.

The short answer, as for anything that absorbs very faintly in the UV, is that the color would be very faintly yellow.

A revised version of this answer would take into account the effects of aggregation. Surely more interesting things (a richer spectrum of electronic transitions) can happen when CO$_2$ molecules are bunched at high pressure. However here the temperature/pressure regime that resulted in an ocean of CO$_2$ would probably be critical.

Also, what color is this ocean's floor? If the absortion is weak enough, this may be the dominant color as seen from space anyway.

If one absolutely wants to find a spectrum

If hard-pressed to find a spectrum, one could resort to this theoretical simulation: "An instantaneous normal mode theory of condensed phase absorption: The collision-induced absorption spectra of liquid CO2" (Moore, J. Chem. Phys. 107, 5635 (1997)). The positive side here is that collision (and thus condensed matter, in this case liquid phase) effects are taken into account and indeed the topic of the paper, with "liquid CO$_2$" being in the title. The negative side is that I'm way less sure of what is happening here and how much we can trust what we're seeing. The authors do acknowledge that

"Unfortunately, experimental data for the fundamental absorption line-shapes at the state point studied here are not, to our knowledge, available. In fact, we were unable to find band shapes for fundamental absorptions in neat liquid CO2 at any state point, although the corresponding data for CS2 is available."

That having been said, there goes a spectrum together with the corresponding caption:

The bend absorption spectrum is presented. The Fourier transform of the MD dipole-dipole correlation function is presented for comparison with the INM spectrum.

"The bend absorption spectrum is presented. The Fourier transform of the MD dipole-dipole correlation function is presented for comparison with the INM spectrum."

  • $\begingroup$ That's an amazing answer! Couldn't have expected more, thank you so much! I'll wait just a little bit to give you the bounty JUST IN CASE someone else gives a good answer but... I see that very unlikely, you are probably the winner here by far. $\endgroup$
    – DeMooniC
    Commented Jul 20, 2021 at 17:26
  • $\begingroup$ Now, you aked me. What color is this ocean's floor? Since we are talking about liquid CO2 ,unlike water ice, CO2 ice would simply sink to the bottom. So I guess the sea floor would simply be a bright white because of it, which would just reflect back up all the light that reached the bottom and didn't get absorbed, in this case, yellow ligh acording to you, right? Ok so the conclusion here ended up being that "for anything that absorbs very faintly in the UV, is that the color would be very faintly yellow." (1/2) $\endgroup$
    – DeMooniC
    Commented Jul 20, 2021 at 17:29
  • $\begingroup$ So this means that, the deeper the ocean, the yellower it would appear, right? Also acording to your answer (correct me if im wrong), the denser the atmosphere and the colder the ocean, the "yellower" it gets. So that basically means that, for example, in a planet where the atmospheric pressure is around 100 bars and the temp is around -40°C, the oceans would appear way more saturated in color than in a 30 bars -13°C planet with CO2 oceans just as deep. Is that right? And last thing I have doubts about. Would a deep CO2 get as yellow as a deep H2O ocean gets blue? (2/3) $\endgroup$
    – DeMooniC
    Commented Jul 20, 2021 at 17:41
  • $\begingroup$ Oh and almost forgot, what about rayleight scattering? Wouldn't the effects of Rayleight scattering be stronger than the slightly yellowish coloration on liquid CO2 making it look more of a blue/blue-green? (3/3) $\endgroup$
    – DeMooniC
    Commented Jul 20, 2021 at 17:42

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.