As teams of researchers begin to detect molecules that could indicate the presence of life in the atmospheres of exoplanets, controversies will emerge. In the early stages, the method will be transmission spectroscopy, in which light from the star passes through the planet’s atmosphere as it transits the host. From the resulting spectra various deductions may be drawn. Thus oxygen (O₂), ozone (O₃), methane (CH₄), or nitrous oxide (N₂O) would be interesting, particularly in out of equilibrium situations where a particular gas would need to be replenished to continue to exist.
While we continue with the painstaking work of identifying potential biological markers — and there will be many — new findings will invariably become provocations to find abiotic explanations for them. Thus the recent flurry over K2-18b, a large (2.6 times Earth’s radius) sub-Neptune that, if not entirely gaseous, may be an example of what we are learning to call ‘hycean’ worlds. The term stands for ‘hydrogen-ocean.’ Think of endless ocean under an atmosphere predominantly composed of hydrogen. Now put it in the habitable zone.
Thus the interest in K2-18b, which appears to orbit within the habitable zone of its red dwarf host. Astronomers have known about water vapor here for some time, while JWST results in 2023 further indicated carbon dioxide and methane. On its 33-day transiting orbit, this is a planet made to order for spectral analysis of its atmosphere. Now we have new work that leans in the direction of a biological explanation for a possible biosignature, one that is tantalizing but clearly demands further investigation.
The biosignature, deduced by researchers at the University of Cambridge led by Nikku Madhusudhan, involves dimethyl sulfide (DMS) and/or dimethyl disulfide (DMDS). These are molecules that, at least on Earth, are produced only by life, most commonly by marine phytoplankton, photosynthetic organisms that play a large role in producing oxygen for our atmosphere.. The detection, say the authors, is at the three-sigma level of statistical significance, which means a 0.3% probability that these results occurred by chance. Bear in mind that five-sigma is considered the threshold for scientific discovery (below a 0.00006% probability that the results are by chance). So as I say, we can call this intriguing but not definitive, a conclusion the authors support.
Image: Nikku Madhusudhan, who led the current work on K2-18b. Credit: University of Cambridge.
What excites the researchers about this work is that they first saw hints of DMS using the James Webb Space Telescope’s NIRISS (Infrared Imager and Slitless Spectrograph) and NIRSpec (Near-Infrared Spectrograph) instruments, but found further evidence using the observatory’s MIRI (Mid-Infrared Instrument) in the mid-infrared (6-12 micron) range. That’s significant because it produces an independent line of evidence using different instruments and different wavelengths. And in the words of Madhusudhan, “The signal came through strong and clear.”
But Madhusudhan said something else that has excited commentators. Noting that the concentrations of DMS and DMDS in K2-18b’s atmosphere are thousands of times stronger than what we see on Earth, there is an implication that K2-18b may be a specific type of living planet:
“Earlier theoretical work had predicted that high levels of sulfur-based gases like DMS and DMDS are possible on Hycean worlds. And now we’ve observed it, in line with what was predicted. Given everything we know about this planet, a Hycean world with an ocean that is teeming with life is the scenario that best fits the data we have.”
Centauri Dreams readers will want to check Dave Moore’s Super-Earths/Hycean Worlds for more on this category. A hycean world is considered to be a water world with habitable surface conditions, and in earlier work, Madhusudhan and colleagues have noted that K2-18b could well maintain a habitable ocean beneath a hydrogen atmosphere. We have no analog to planets like this in our own system, but the category may be emerging as a place where conditions of temperature and atmospheric pressure may allow at least microbial life.
Image: Artist’s conception of the surface of a hycean planet. Credit: Amanda Smith, Nikku Madhusudhan.
So what is producing these chemical signatures? There may be reason for some optimism about a life detection but the possibility of unknown chemical processes remains alive, and thus will spawn further work both theoretical and experimental. And this is the problem for the entire landscape of remote biosignature detection. We’re going to be seeing a lot of interesting results as our instrumentation continues to improve, but at the level of uncertainty that will ensure debate and the need for further taking of data. This is going to be a long and I suspect frustrating process. Astrophysicists are going to be knocking heads at conferences for decades.
So this is an example of how the debate is going to be playing out at many levels. The evidence for biology will be sifted against possible abiotic processes. From the paper:
…both DMS and DMDS are highly reactive and have very short lifetimes in the above experiments (i.e., a few minutes) and in the Earth’s atmosphere (i.e., between a few hours to ∼1 day), due to various photochemical loss mechanisms (e.g. Seager et al. 2013b). Thus, the resulting DMS and DMDS mixing ratios in the current terrestrial atmosphere are quite small (typically ≲1 ppb), despite continual resupply by phytoplankton and other marine organisms…. sustaining DMS and/or DMDS at over 10-1000 ppm concentrations in steady state in the atmosphere of K2-18 b would be implausible without a significant biogenic flux. Moreover, the abiotic photochemical production of DMS in the above experiments requires an even greater abundance of H2S as the ultimate source of sulfur — a molecule that we do not detect — and requires relatively low levels of CO2 to curb DMS destruction (Reed et al. 2024), contrary to the high reported abundance of CO2 on K2-18 b (Madhusudhan et al. 2023b).
Image: The graph shows the observed transmission spectrum of the habitable zone exoplanet K2-18 b using the JWST MIRI spectrograph. The vertical shows the fraction of star light absorbed in the planet’s atmosphere due to molecules in the planet’s atmosphere. The data are shown in the yellow circles with the 1-sigma uncertainties. The curves show the model fits to the data, with the black curve showing the median fit and the cyan curves outlining the 1-sigma intervals of the model fits. The absorption features attributed to dimethyl sulphide and dimethyl disulphide are indicated by the horizontal lines and text. The image behind the graph is an illustration of a hycean planet orbiting a red dwarf star. Credit: A. Smith, N. Madhusudhan (University of Cambridge).
So there are reasons for optimism. We’ll keep taking such results apart, motivated by the unsparing self-criticism of the Cambridge team, which goes out of its way to scrutinize its findings for alternative explanations (a good lesson in scientific methodology here). Case in point: Madhusudhan and colleagues point out evidence for the presence of DMS on comet 67P/Churyumov-Gerasimenko, which could mean an abiotic source delivered by comets into the atmosphere. Because comets contain ices and gases that could be interpreted as biosignatures if found in a planet’s atmosphere, we’re again reminded of the need for caution. Even so, we can deflect this.
For at K2-18b, the atmosphere is massive compared to the trace gases that could be induced by cometary delivery, and the authors doubt that DMS and DMDS would survive in their present form during a hypervelocity comet impact. K2-18b just has too much DMS and DMDS, per these findings, to be accounted for by comets alone.
Detecting a biosignature will require accumulating more and more evidence to demonstrate first the actual presence of the detected molecules and second the possible abiotic photochemical ways of producing DMS and DMDS in an atmosphere like this. Madhusudhan cites this work as an opportunity for pursuing such investigations within a community of continuing research. No one is claiming we have detected life at K2-18b, but we’re getting a nudge in that direction that will be joined by quite a few other nudges as we probe alien atmospheres.
Not all these nudges point to the same things. For among papers discussing K2-18b, another is about to appear that questions whether it and another prominent sub-Neptune (TOI–270 d) are actually hycean worlds at all. This deep dive into sub-Neptune atmospherics, led by Christopher Glein at Southwest Research Institute, will be our subject next time. For before we can make the call on any hycean biosignature, we have to be sure that oceans are possible there in the first place.
The paper is Madhusudhan et al., “New Constraints on DMS and DMDS in the Atmosphere of K2-18 b from JWST MIRI,” accepted at Astrophysical Journal Letters (preprint).
A very level-heading posting compared to what I’m seeing elsewhere on the ‘Net.
Reading the entry from 2022 suggest the ocean temperature on this planet is likely hot, up to the triple-point temperature of 374C. This makes biology unlikely as a source of the DMS and DMDS. There is a photo-chemical reaction that can produce these materials. It is also possible that there could be a biological process, but one that is not at the level of cellular life. If it is cellular life, it would be similar to the extremophiles you see in the Yellowstone hot springs. But my bet is a abiotic photo chemical reaction.
I was wondering the same thing that there can be another way to make DMS and DMDS through natural chemical reactions especially since there is no oxygen. Without oxygen, there is no life. it is interesting that there are other biosignature gases like methane and carbon dioxide in K2-18B. Neptune has some methane. Titan has methane and nitrogen. It is good to see that the JWST can detect complex chemistry there in the spectra. I am waiting for the spectra of some more Earth like planets instead of super Earths and mini Neptunes.
” Without oxygen, there is no life.”
For pity’s sakes, Geoffrey, that’s not true and you know it.
https://en.wikipedia.org/wiki/Anaerobic_organism
While there is a lot of excitement, I hope this doesn’t prove as disappointing as the phosphine in Venus’ atmosphere announcement. There is still controversy over whether phosphine was detected, whether the signal was more likely SO2, and whether phosphine, if present, might be generated abiotically on Venus. ( I am still waiting for the launch of the Morningstar Mission, with an instrument to start the analysis of the droplets in the temperate zone of the atmosphere. c.f. The Great Venusian Bug Hunt )
I have heard that the Cambridge team will release the spectrographic data so that other groups can also analyze it. This is good as it means that more eyeballs and analyses of the data may ensure that the authors are not fooling themselves with the data.
If nothing else, it demonstrates the value of JWST’s NIR and MIR instruments, which cover the wavelengths of chemical bond stretching and biosignature molecule species detection.
As the OP states:
Indeed.
Here is the official Web site of the Hycean Worlds Web site:
https://hycean.group.cam.ac.uk/
This includes getting email updates from them:
https://hycean.group.cam.ac.uk/engage/
You did a great job of highlighting issues with the abiotic explanations. Looking back at their Table 2 and the probability graph, it seems like the most straightforward predictions cluster around maybe 150 to 600 ppm of CH3SCH3 and/or CH3SSCH3. Existing work shows the ability to make 0.5-0.9 and 0.2-0.4 ppm of these compounds out of 1000 ppm CH4 and 20 ppm SH2 in nitrogen, using a UV lamp that includes the 121.6 nm radiation that photolyzes CH4.
At first glance, given a planet with enough CH4 and SH2, you should be able to make more DM(D)S. If the compound constantly breaks down and the rate law is second order in CH4 and SH2 (? is it ?), the amount present should multiply proportional to the concentrations of those two reactants. Madhusudhan previously identified 10000 ppm (1%) CH4 in this planet’s atmosphere. However, the SH2 hasn’t been demonstrated to exist, and DM(D)S appears to break down in the presence of CO2 (and certainly in Earth’s oxidizing atmosphere).
Even so, the abiotic problems seem easier to explain away. I presume 20 ppm or higher of H2S might be invisible to Webb’s limited degree of sensitivity? The red dwarf might flare a lot? The CO2 might be at a different altitude than where the DM(D)S forms? (hmm, can there be H2S ‘clouds’ in the high atmosphere?).
I have a harder time seeing the path to life here. What is life doing in the upper atmosphere pumping out small reactive chemicals that cost energy and can’t persist? If life is deeper, how does it get energy under a thick atmosphere where the photons it absorbs and those it emits are about the same temperature? (rather than 6000 K vs 300 K on a planet with low clouds or clear skies)
I noticed this article on RealClearScience this morning.
https://arstechnica.com/science/2025/04/skepticism-greets-claims-of-a-possible-biosignature-on-a-distant-world/
It seems to echo some of the skepticism I am seeing here and on other science-based websites. Skepticism is good except when dining at your mother-in-law’s house and she asks “how was dinner dear?”
Am I interpreting this correctly:
1. The spectrum in the image and the paper is only from the MIR instrument and not extended to to NIR.
2. NIST reference data extends into the NIR. The NIST IR spectrum for DMS implies that the 3000 cm-1 peak is in the NIR and outside the range of the data. The 250 cm-1 data is also out of the range of the MIR data. (wavenumber cm-1 conversion to wavelength um = 1E4/wavenumber.
3. For absorption, the transit depth INCREASES , i.e. 0.325% implies greater absorption than 0.225%. The higher transit depth should correspond to the absorption peaks of DMS and DMDS. Is this correct? If so, we are really only concerned with the data points above the assumed canonical atmosphere.
4. DMS would be distinctive due to the C-S bond stretching. This appears at ~ 9.7 and 11.4 um. If my point 3 is correctly interpreted, there are no strong increases in transit depth at these values.
Can someone help clarify my current interpretation of the data by eye because this is at odds with the interpretation in the paper? (I will be reading the paper today to understand how the analysis was done).
The data massaging pipelines are quite complex. The 2 pipelines JexoPipe and JExoRES, transit analyses are shown in Figure 4. Concordance seems to decline for wavelengths > 10um. Whether this is important or not, I cannot say.
Figure 2 suggests that DMS and DMSD can be distinguished by the absorption on either side of the 10um wavelength. This must be something that is missed by the NIST IR spectra, which shows that both molecules have very similar IR spectra. The 10um point = wavelength 1000 cm-1. The 2 spectra can be found for DMS and DMDS here.
Figure 6 seems to show that the transit data is best modeled if DMS/DMDS is accepted with IR data ~ 7um (wavenumber 1400) and 8.75 um (wavenumber 1100 cm-1). These are both strong absorption peaks in the NIST spectra. Yet Figure 14 seems to show that at 7um, CS2 might be a confounding molecule. Similarly, at 8.75 um, SO2 might be a confounding molecule. The authors’ Bayesian fit model seems to rule this out, but…
More generally, DMS is a breakdown product of Dimethylsulfoniopropionate (C5H10O2S), which is produced by algae on Earth. This is a more complex molecule, and if the source, it seems less likely to be abiogenic. However, it does suggest that the life in K2-18b hypothetical ocean must have a convergent evolution to that on Earth. This may or may not be very remarkable, but it does have implications, because it suggests that organisms with complex, eukaryote traits have evolved pathways that produce at least one product of terrestrial algae, and that it breaks down (on Earth by bacteria) to DMS. But where does the DMSD come from? If the source organism is like a photosynthesizing alga, where is the O2 in teh atmosphere? Is it present but not measured? Is it consumed before it can accumulate? Or perhaps the organisms fix carbon from CH4 with H2S instead of water, forming CH2S and releasing H2?
Overall, I can see this is going to be a possibly contentious finding. I hope we can get better data, and preferably orthogonal data to support the claim. Could a laboratory experiment find organisms that could produce a similar result under the proposed conditions on K2-18b? If life is eventually confirmed, that would certainly suggest that, at a minimum, life is common in the galaxy, bolstering the likelihood that we will detect more inhabited planets. That would be a momentous finding.
Quote by Alex Tolley: “2. NIST reference data extends into the NIR. The NIST IR spectrum for DMS implies that the 3000 cm-1 peak is in the NIR and outside the range of the data. The 250 cm-1 data is also out of the range of the MIR data. (wavenumber cm-1 conversion to wavelength um = 1E4/wavenumber.” I don’t what you mean by this. Wave number is not used in the spectra here in this paper. It’s in microns. Wave number can be converted into microns. Wave number is a fraction with one on the top or divided into one. For example Carbon dioxide spectral absorption wave number is around 669 cm-1 which is the same as (15 μm). Google AI. Any number divided into one can also be expressed as a number with a minus sign in algebra.
The diagram here of JWST spectra shows the range of gases between 6 and 12 micros which is the range of DMS, DMDS spectra. Carbon Dioxide is at 2.7, 4.3 and absorbs most strongly at 15 um or fifteen microns, and Methane is at 3.3 to 3.7. These are not shown in the diagram since it only shows the spectral range of DMS and DMDS between 6 and 12 microns which match our know absorption model.s of DMS and DMDS which always stays the same since every atom and molecule has distinct energy levels and therefore absorbs EMR at a specific wavelengths. The other biosignature gases Methane and Carbon Dioxide, etc are not included in this diagram. I would also like to see the other spectra.
Wavelength Ranges:
Near-Infrared (NIR): Typically ranges from 0.7 to 2.5 micrometers (μm).
Mid-Infrared (MIR): Generally falls between 3 and 25 μm.
Far-Infrared (FIR): Defined as wavelengths above 25 μm
Note that the MIR is between 3 and 25 microns. DMS and DMDS don’t fall in the NIR range.
Oh, the source is Google AI
I am sorry I did not explain this too well. DMS and DMDS will only absorb electromagnetic radiation between six and twelve microns. All other wavelengths in near infra red and far infra red and even the whole electromagnetic spectrum pass right through those molecules in these two gases and a not absorbed due to the molecular energy levels of those molecules which are distinct. The absorption is at the ground state between zero and one, the first quantum Jump which is at an energy equal to a specific wavelength and if it does not match that wave length energy, it will not be absorbed. Every molecule and atom his its unique absorption spectrum which is why spectroscopy works.
The ground state is the lowest energy level of the electron, and then there is the first, second, third, quantum jumps depending on the atom. It is the electron’s ground state which absorbs the EMR which are called absorption lines and when a molecule or atom gets hot, it re emits the light or EMR at exactly the same wavelength, called emission lines.
My best guess is that the DMS and DMDS are natural atmospheric constituents.
From my article I did on this planet
https://www.centauri-dreams.org/2023/10/20/atmospheric-types-and-the-results-from-k2-18b/
Based on the paper:
P. Woitke, O. Herbort, Ch. Helling, E. Stüeken, M. Dominik, P. Barth and D. Samra, Coexistence of CH4, CO2, and H2O in exoplanet atmospheres, Astronomy & Astrophysics, Vol. 646, A43, Feb 2021
https://doi.org/10.1051/0004-6361/202038870
I conclude that this is a planet with a carbon dominated (surplus) atmosphere, which means that, with the exception of H2O, atmospheric species preferentially combine with carbon as it is the lowest thermodynamic energy state for the atmosphere.
Why we don’t see CS2 is that the carbon to sulfur ratio is 0.5. DMS has a C/S ratio of 2, and in DMDS it’s still 1.
Also, from my article on Hycean planets, I concluded that K2-18b has most likely a steam super fluid atmosphere over high pressure ice. It’s touch and go, but the greenhouse effect and the adiabatic lapse rate keep things just a bit too hot for water to condense even at high pressures where its boiling point could be over 300°C.
Of course, there are a lot of unknowns about this, but this does seem to be the simplest explanation.
Well, at the very least I want to congratulate or thank everyone who has kept us up to date on “Hycean developments”. Whether K2-b “lives” up to expectations or not, this has not been a development that catches Centauri-Dreams readers unaware.
And consequently the critiques of the results are reasonable enough, pro or con.
If nothing else, this ambiguous detection should result in the grinding of more
powerful lenses, mounted here on Earth or in space. And a market for higher resolution spectrometers too…
I found this to be very enlightening:
Laboratory studies on the viability of life in H2-dominated exoplanet atmospheres.
“Theory and observation for the search for life on exoplanets via atmospheric “biosignature gases” is accelerating, motivated by the capabilities of the next generation of space- and ground-based telescopes. The most observationally accessible rocky planet atmospheres are those dominated by molecular hydrogen gas, because the low density of H2-gas leads to an expansive atmosphere. The capability of life to withstand such exotic environments, however, has not been tested in this context. We demonstrate that single-celled microorganisms (E. coli and yeast) that normally do not inhabit H2-dominated environments can survive and grow in a 100% H2 atmosphere. We also describe the astonishing diversity of dozens of different gases produced by E. coli, including many already proposed as potential biosignature gases (e.g., nitrous oxide, ammonia, methanethiol, dimethylsulfide, carbonyl sulfide, and isoprene). This work demonstrates the utility of lab experiments to better identify which kinds of alien environments can host some form of possibly detectable life.”
https://arxiv.org/abs/2005.01668
Here is the version for those not so enlightened:
Hydrogen-breathing aliens? Study suggests new approach to finding extraterrestrial life.
https://theconversation.com/hydrogen-breathing-aliens-study-suggests-new-approach-to-finding-extraterrestrial-life-137630
K2-18b, with it’s density, is more like a mini-neptune than a hycean world. Under earth-like insolation, the ocean would be hot indeed. In https://arxiv.org/pdf/2002.11115, temperature-density profiles were calculated based on different overall compositions. In most likely case (50% ices, 50% silicates/iron, filled up to observable radius with H2/He), conditions on H2/water boundary are 700 bar and 1500 K, give or take an oomag in pressure and half in temperature. Quite difficult to get ocean below critical point, almost impossible below 400 K, which requires almost pure water without any rocks and methane.
So, a very exotic biochemistry, if any. It’s much more likely that we don’t know something important about photochemistry, catalytic processes or else in such worlds than carbon-based life could be pushed to such extremes. And, unlike Venus, no early habitable period on K2-18 b, it always was a pressure-cooker.
Hi Paul
This is one very interesting place, while an interesting planet I tend to agree with some of the above readers comments, and Life on these Hycean/Ocean would could be a tough place to live and survive.
I’ll be looking forward to the next post on the transition to a Sub Neptune world to a hycean world too.
Thanks Edwin