The ‘perfume’ of 67P/C-G

With inputs from Kathrin Altwegg, ROSINA science team, University of Bern.

Since early August, the Rosetta Orbiter Sensor for Ion and Neutral Analysis (ROSINA) has been ‘sniffing the fumes’ of 67P/C-G with its two mass spectrometers.

As reported previously in this blog, even though the comet is still more than 400 million kilometres from the Sun, the mixture of molecules detected in the comet’s coma is surprisingly rich already. Before arriving at 67P/C-G, the ROSINA team thought that at these vast distances from the Sun, its relatively low intensity would only release the most volatile molecules via sublimation, namely carbon dioxide and carbon monoxide.

However, ROSINA has detected many more molecules. Indeed, as of our 11 September report, ROSINA’s inventory of detected gases 67P/C-G looked like this:

Water (H2O)
Carbon monoxide (CO)
Carbon dioxide (CO2)
Ammonia (NH3)
Methane (CH4)
Methanol (CH3OH)

But today we can report that the following have also been detected:

Formaldehyde (CH2O)
Hydrogen sulphide (H2S)
Hydrogen cyanide (HCN)
Sulphur dioxide (SO2)
Carbon disulphide (CS2)

High resolution mass spectrum of ROSINA DFMS, taken on October 10 at a distance of 10 km from the comet centre. Hydrogen sulphide is easily seen. The second peak is due to the heavier isotope of sulphur 34S, which is a fragment of all sulphur bearing species. Image courtesy K. Altwegg, University of Bern

High resolution mass spectrum from ROSINA’s Double Focusing Mass Spectrometer (DFMS), taken on 10 October at a distance of 10 km from the comet centre. The plot shows the detection of hydrogen sulphide and the heavier isotope of sulphur, 34S, which is a fragment of all sulphur bearing species. The plot shows intensity vs. the mass-to-charge ratio*.  Image courtesy K. Altwegg, University of Bern

If you could smell the comet, you would probably wish that you hadn’t 🙂

As the Kathrin Altwegg, principal investigator for ROSINA, put it: “The perfume of 67P/C-G is quite strong, with the odour of rotten eggs (hydrogen sulphide), horse stable (ammonia), and the pungent, suffocating odour of formaldehyde. This is mixed with the faint, bitter, almond-like aroma of hydrogen cyanide. Add some whiff of alcohol (methanol) to this mixture, paired with the vinegar-like aroma of sulphur dioxide and a hint of the sweet aromatic scent of carbon disulphide, and you arrive at the ‘perfume’ of our comet.”

While this is unlikely to be a particularly attractive perfume, remember that the density of these molecules is very low, and that the main part of the coma is made up of water and carbon dioxide, mixed with carbon monoxide.

The key point, however, is that a detailed analysis of this mixture and how it varies as 67P/C-G grows more active will allow scientists to determine the comet’s composition. Further work will show how 67P/C-G compares with other comets, for example by revealing differences between comets originating from the Kuiper Belt (like 67P/C-G) and comets that hail from the distant Oort cloud (like Comet Siding Spring, which recently flew past Mars). The goal is to gain insights into the fundamental chemical make-up of the solar nebula from which our Solar System and, ultimately, life itself emerged.

 *For a simple explanation of mass spectra and associated units, see Wikipedia’s entry: 
For more detailed information on ROSINA, see our description here.




  • morganism says:

    Is this actually water spectra, or hydroxyls ?

    Are they looking for a certain water stretching band?

    Hope you got to enjoy the AGU field trips too !
    the tweets were sweet, but no links to posters…

    • Cometstalker says:

      Its dihydrogen monoxide not -OH the difference of one missing H atom is clearly noticeable in the Rosina. The resolution in the chart is better then 0.005 m/z . At large H2O sums to about 18 and -OH makes it to about 17 thats a huge difference also the odd -OD could be ruled out but it is to rare to make a fuzz about.

    • Alembe says:

      I believe the water was detected by mass spec, in which case it would be the whole moleceule (H20).

  • greg_coyote says:

    A rich chemical mix- this mission continues to be completely fascinating to this armchair astronomer. Am I right in saying that the presence of chemicals such as these certainly does no harm to the theory of comets as a possible distribution system for simple life/organic molecules around the solar system?

    • logan says:

      If time gives reason to this argument, then expecting findings to include the ‘simplest’ forms of life. So simple as to redefine the term.

    • Cometstalker says:

      No harm at all is done if the comet is of moderate size relative to the planet and the planet is without life.
      The chemistry of this comet presented so far is only a fraction of the whole spectra available. This does not mean that without comets no life is possible as these chemicals are easy to create. To create life out of planetary soup spiced with comet dust or not took quite some time. Then this first life developed in an accelerated speed to the extent that we try our best to create artificial intelligence with silicon, graphen, polymer and some other stuff also in a very steep accelerating curve. Where and why all of this started is of minor interest, the major concern is if it will continue for a while or not.

    • Marco says:

      Hi greg_coyote, I certainly need no convincing. Comets are extremely complex and unpredictable just like living things. The inside of the comet even more so to explain what and when it exhales these chemicals. Life within the comet would explain a lot.

    • James says:

      It is a good question. I wonder what kind of structures can emerge when these chemically diverse “perfumes” get cooked and ionized in a close approach to the sun before being carried off into space by solar wind and eventually funneled into a magnetically active planet.

    • Kamal Lodaya says:

      Greg: Phosphorus still to be seen.

  • Pat says:

    Isn’t hydrogen sulphide H2S rather than H2S2?…

  • Gabriel says:

    I think there may be a typo among the newly reported gases : You wrote “Hydrogen sulphide (H2S2)” when it should instead be H2S.

    Very interesting info anyway. I wonder what complex organic compounds the dust may contain. There sure seems to be the ingredients for a lot of things!

  • Martin Stoermer says:

    A small point but hydrogen sulphide is listed incorrectly in the text as H2S2 but is correct in the graphics as H2S

    • emily says:

      Thanks to all who noticed the typo – corrected! 🙂

      • Cometstalker says:

        The odd thing about di-hydrogen sulphide is that if there is a lot of it around it is not smelly at all as the sensors in the nose are getting paralysed, the bad news is that at this concentration level its a huge risk that the rest that is atatched to this nose gets paralysed as well. Most of the other stuff is quite unhealthy too if in a bit higher concentration.

        Interesting would be to get a report if the comet has a magnetic field indicating an iron-nickel core, alternatively the gravity vectors indicating that the comet is uniform in density.

      • Cometstalker says:

        H2S2 actually does exist and it is called Disulfan its a yellow liquid at NTP to some extent related to H2O2.
        That H2S is present on the comet is not a surprice as this place is bombarded with protons, C S2 is a bit less common. On the other side the solar wind contains quite a bit more then just protons and @particles plus the electromagnetic spectrum.

  • Jacob nielsen says:

    “Please Sir, I want some more”
    Very smooothed out sectrum, no need to do that! Oh! Lets add H2S to the list. I was planning to keep my spacesuit on anyway. The faint smell of hot springs would come second to the cold and vaccum. H, C, N, O, S, ?

    • Jacob nielsen says:


    • Kathrin Altwegg says:

      This spectrum is not smoothed, that’s raw data coming directly from the sensor, just the offset has been deducted.

      • Jacob Nielsen says:

        Ok! Then it is just clean. And narrow too, I guess actually not much else to expect within 0,1u.

    • Robin Sherman says:

      Lots of Chlorine compounds have been found on Mars. I would expect both Fluorine and Chlorine gas or a simple Chlorinated Hydrocarbon like CH3Cl to be found also at some point in 67P’s coma.

  • Cometstalker says:

    Oh dear,

    i did say that comets are the leftovers of the creation of our solar system. Trash cans are smelly and if left in the sun for too long the really start to become abusive.

    So, the material in comets have a lower variety then what can be found on planet earth. The atmosphere of Titan is quite well analysed qualitative and to some extent quantitative an so far nothing on 67/pcg is presented that is not already found in Titans atmosphere. The ESA scientist might take a look in that list to get a hint what else is to be expected on the comet. Also in Nebulas a lot is to be found including tera tons of ethanol. Something for the next pipe line project?

  • 342rsj says:

    The spectrum shows 2 nice signals attributed to sulfur, molecular and isotopic, nicely resolved!
    There are also 2 small peaks on the right, one is located at 34.03 m/z, the other at 34.005, is there any molecular/isotopic interpretation for those signals?
    My understanding is that the signal at 34.03 is most probably a protonated organic molecule (13CMethanol?)
    But what about the the peak at 34.005 (SiH4, H2O2)?

    • logan says:

      Give us some playground. H202 🙂

    • Cometstalker says:

      Im surprised why a log y-axis scale is not used and why this almost 2 weeks delay, the curve is ready almost instantly and to make it a bit more presentable takes halve an hour. All kind of data is of interest but the filtering time through the censor committee or whatever is creating the delay is frustrating.

    • Cometstalker says:

      SiH4 is a bit more stable then the very reactive H2O2 also its boiling point is a lot lower. The propellant is at least not H2O2 so it is not polution. Mono-methyl-hydrazine plus nitrogen-tetroxide is used as a bi-propellant.

    • Jacob Nielsen says:

      Hi 342rsj!
      I can’t really make your suggestions for the minor peaks fit… Am I not doing my math right?

    • Kathrin Altwegg says:

      Good spotting! These are actually fragments of heavier organic molecules, fragmented in our ion source. We are analyzing them.

  • Ross says:

    “Before arriving at 67P/C-G, the ROSINA team thought that at these vast distances from the Sun, its relatively low intensity would only release the most volatile molecules via sublimation, namely carbon dioxide and carbon monoxide.”

    Is this statement implying that sublimation is not responsible for the existence of the various compounds found in the coma? Or is it simply saying that the sublimation process is “weirder” than originally thought?

    • Cometstalker says:

      The streamers visible on several pictures clearly shows that they are not due to solar radiation. Another hint is it random outbursts. As the surface is supposed to have a relative high thermal insulation the only alternative is that the inner parts of the comet in the neck region are exhaled due to gravitation induced force and tidal grinding converting mechanical energy into heat sublimating part of its inside substance and make it to diffuse into its coma.
      Analysing the contents of these streamers is the best opportunity ever to get samples from deep inside of the comet and at 8 km from the comet surface the sample material is not yet poluted with the solar wind.

    • Robin Sherman says:

      Ross that statement raised a question for me too. Does this indirectly imply that some process other than near surface sublimation is the source and that the other source we have seen, the streamers, is thought to be a process not related to solar radiation and surface sublimation? We know ROSINA can measure the velocities of gases, so if the gases detected did come from the higher velocity streamers, they would know.

      Just another little teaser. The comet produces a witches brew of noxious, smelly chemicals, from somewhere, somehow. What about other NON smelly molecules, N2, Argon and Hydrogen Peroxide or nice smelling compounds like Ketones, the simplest being Acetone. I know the notion of foul smelling chemicals and the comets “Perfume” is designed to have a broader public appeal and thus generate some headlines and interest, but a simple add on at the end the post, of which other molecules have been detected, would have kept us science types satisfied as well.

      As has been said by others in previous posts, some of us are never satisfied even when we do get some science results. So many thanks to Kathrin and Emily for at least sharing something with us.

    • logan says:

      Gas in ‘high speed’ jets trawls other materials. Once outside, well, it’s another history.

    • Jon says:

      @ roughly 300 million miles from the sun, I doubt in any theorists dream would they detect much of anything but the carbon. Surely, this isn’t sublimation….right?

  • logan says:

    Slightly trashy 😉

    • logan says:

      Very grateful to you and ROSINA team, Emily. Discourse changing info. Specially ‘sensorial’ your document. Like it a lot. (About to ask for a shower for 67P).

  • tntbaba says:

    Is the angle of entry into the inner regions of the solar system that help Astronomers know whether a comet originated from the Kuiper belt or the Ort cloud?

  • Hi Emily–

    Good job getting news from some of the other science teams! Can Dr Altwegg comment on the relative abundances of the different species? If we normalize to H2O=1e10^6, what’s the abuncance of H2S, for example?


    • Kathrin Altwegg says:

      Roughly 1 %. However, we didn’t yet have time to do a thorough analysis. This implies a few factors like sensitivity of the Instrument to the different species, how the ratio changes with the nucleus rotation, what the fragmentation patterns of the different species are, etc. This takes time. But we will tell once we have it.

    • Jacob nielsen says:

      Hi Richard, – oh, no! we would choke in here on so much information 😉

  • Jacob nielsen says:

    Phosphine PH3 is around 34 g/mol too. Just dreaming 🙂

    • Jacob Nielsen says:

      Am I imagining, or would it be possible, that PH3 is hiding in the right portion of the positive skew of the major peak? Any other observations support the presence of phosphine?

  • Robin Sherman says:

    I know that evidence of amino acids would be a big prize, but aromatic hydrocarbons and pyridine would also be significant in terms of life enabling molecules. Any sign that they and the expected Tholins are present in the coma yet?

  • Jacob nielsen says:

    Phosphate chemistry is as essential to life (as we know it) as carbon itself…

  • A.Cooper says:

    In the light of the fact that so many sulphur compounds have been found I’d like to reproduce here an email I sent to a colleague recently which is largely concerned with sulphur compounds being a possible signature of 67/C-G flying through Jupiter’s ring system in the distant past. 

     To give some context, I posted several comments on the August 25th Rosetta blog post, “Landing Site Search Narrows” regarding the hypothesis that 67/C-G underwent a close approach to Jupiter in the distant past which stretched a single body into the two visible lobes:

    I estimated that close approach to be between 150,000 km and 220,000 km from Jupiter’s centre. I now think it more likely to be between 120,000 km and 135,000 km which would send it through the densest part of the rings.

    The two main pieces of evidence are 1) that the pole of rotation of the comet is oriented in the right way for such a close approach to have occurred and 2) that numerous features on the head and body line up, suggesting a single body that was stretched.

     Here’s the reproduced email.

     Dear C

    I have a hypothesis regarding 67P/Churyumov-Gerasimenko which might be proven with data from instruments on Rosetta. I gather you have also researched the Jovian system so I wondered if you might entertain the theory which is as follows:

    1) I believe 67P/C-G was once a single body that has been stretched to its present form and is therefore not a contact binary. 

    2) This stretching could have occurred if 67P/C-G made a close approach to Jupiter, under the Roche limit and through its inner ring system in the last 10,000 years or so.

     There is some evidence for this:

    1) the pole of rotation of the comet is roughly at right angles to its Jovicentric radiant meaning the end-over-end rotation would be prograde along that radiant on approach. F, the modeller I collaborate with found that the rotation vector induced in geocentric close approaches is always prograde, that is, along the radiant. 

    2) using the rubber duck analogy, the stretching of the head away from the body is a classic Roche pass stretch when compared with our geocentric Roche modelling. 

    3) I’ve analysed all photos of 67P/C-G and have found two dozen close matches between the head and the body. I’m quite certain they fit together and were once one solid, single entity, not a binary. 

    4) In addition, there was a possible shearing away of three vast slabs on the side of the body. They would have sheared along the rift caused by the uplifting head and been dislodged further down by the upward leverage. Those slabs are no longer present and would have drifted beyond 67P’s hill sphere due to differential g forces in the Roche pass. However, a pile of slab-like detritus remains wedged against the cliff in landing site A (site A is an exposed area formerly covered by a supposed slab). The detritus is wedged along what would have been the delta g force vector for much of the Roche pass. 

    My calculations suggest the Roche pass was between 120,000 km and 135,000 km above jovicentre. This means 67P would have gone through Jupiter’s rings at relative velocities above 30km/sec. I worked out that if it went through the Amalthea ring it would exit with between 0.3 and 30 micrograms of dust per square metre of surface area. If it was the main ring (more likely) it could be up to 2 milligrams dust and 2kg of meteoroid material per square metre. 

    Since the Jovian ring dust is thought to be short chain sulphur from the inner moons’ surfaces I wondered if it would be possible for the ALICE team to see a signature of this dust in their spectrographic results. I realise it may not be a straightforward search for the actual dust itself but I make the following observations.

    [POINTS 3 and 8, asterisked, are especially pertinent to the new ROSINA data in this Rosetta blog post].

    1) the dust should be detectable only from one side of the comet due to the very small amount of rotation during its ~5-minute ring traverse. The angle of attack in the traverse would have dictated the deposition areas. This hemispheric anomaly of ring dust would be strong evidence for the Roche pass. However, dust resettlement over time might have bled that signature round to the other hemisphere somewhat.

    2) If 67P/C-G went through the Jovian main ring it might have been hit by small meteoroids well as dust. This would betray the attitude of attack in the form of craters. There does seem to be an larger number of craters on the back of the duck and none on the sides. This is to be expected in a Roche pass as the sides would remain protected from the prograde rotation through the rings. (I realise some are probably not impact craters). 

    *3) the dust might have formed sulphides depending on where on the comet it hit. 

    4) the Roche pass velocity at over 30km/sec would mean the impacting dust would contain more than enough energy to satisfy all of the first six ionisation energy levels for sulphur (8600Kj/mol of impacting dust). So perhaps exotic sulphate derivatives were formed on impact such as salts, acids or peroxides.

    5) the ring dust could have tunnelled into the bare, rock-like areas and still be intact a few mm below the surface. Perhaps ALICE could detect the UV spectral lines for sulphur just below the surface. VIRTIS might detect the red spectral lines. CIVA and APXS might see sulphur in the dust around the lander. 

    6) the ring dust could have spallated while tunnelling- perhaps more likely in the cometary dust areas- allowing for a predictable suite of ions at predictable depths, derived from native cometary compounds. ROSINA might see these. 

    7) although the comet has an albedo of 0.04, it may still be red in the redder wavelengths (betraying the collected ring dust). I read that Amalthea and Thebe have albedos of 0.09 and 0.047 but are described as dark red, not black. Of course, this red colour is because they are the same colour as the ring dust and are the sources of that dust. (However dust accumulation on the moons, ultimately from Io, would have been a whole lot gentler than 30km/sec so it stayed chemically intact). 

    8) If the dust didn’t ionise, there may be enough of it to give an absorption line for the sulphur itself or lines for its different-length catenated polymers. I understand the dust is derived from cooled liquid sulphur- the pi allotrope which has quite a large proportion of S-6 and S-8. [The graph from Rosina in this “perfume” post shows absorption lines for the heavier

    9) Moreover, it may be possible to predict sulphides for a specific location on the comet (depending on surface composition) and look for the expected absorption line for those accordingly. 

    *10) Trace amounts of sulphur dioxide might be streaming out into the coma. MIRO might detect that. [ROSINA DETECTED IT AS PER THIS ROSETTA BLOG POST].

    In conclusion, for all my speculation over the ring dust, one thing is abundantly clear to me: the comet has been stretched and is not a contact binary. It follows from this that the Roche pass is at least a reasonable conjecture and the ring dust, somewhat more conjectural. 

    Kind Regards 


    I think the ring dust theory (that would betray a Roche pass) is still conjectural but somewhat less so in the light of these recent data finding sulphur compounds in the coma.

    • Robin Sherman says:

      Very interested to read your post. I too noticed the distribution of impact sites. I have speculated on the causes of the clean fractures, two on the body and one on the head, as being from impacts. The chances of the right velocity and size to actually do this are pretty remote for one such event, but three? I find your proposed idea a lot more credible.

      The evidence for molten material flowing on the comet is compelling to me, but a large source of energy would be required to achieve this. I suggested earlier that tidal heating from proximity to Jupiter might be one source. How significant might tidal heating of the comet’s interior during this encounter have been?. Significant heat would be created in the neck region and for sure would lead to the expulsion of large amounts volatiles and the production possibly of molten material, but I get the sense that material has mainly flowed towards the neck from the lobes. Tunnelling by impactors does not seem to be the only source of energy for “ice lava”. Should we restrict such a close encounter with Jupiter to a single occurrence? Maybe only one with such severe consequences as “stretching” the comet.

      There does seem to be an awful lot of surface dust/powder/ejecta with visually different albedos in different areas of the comet. Its distribution is quite specific suggesting a large amount is from a source other than sublimation residue and coma fallout, both of which, one would expect to give a more even, comet wide coating.

      As an amateur observer and speculator in broad terms, it is really nice to read the speculations of those who know a lot more about the details of these processes. Unfortunately something that has been non existent from the Rosetta science team so far.

      • A.Cooper says:


        Thanks for your reply. I’m glad someone else has noticed those areas that look like clean breaks. I believe that if there were large slabs that went adrift, the one on landing site A actually extended round as far as the last obvious ridge near to the ‘edge’ where the biggest craters are. This extended area is largely free of craters and has large amounts of scree and boulders scattered in the low parts (see Cometwatch 20th October photo). It also appears to be more pitted with ragged ridges and less of a blanket of dust  (19th Oct- best in hi res).  This seems to be consistent with the damage that would occur if a large slab were ripped almost vertically from the surface. 

        I think there would have been substantial heating during the hypothesised Roche pass due to the stretching effect so I think it could have caused ice lava to flow, as you say (please also see Marco’s interesting comment on ‘spin-up’, below and on the ‘landing site search narrows’ blog entry). 

          I think it could all occur in one pass and, as you say, “Maybe only one with such severe consequences as “stretching” the comet”. The window or keyhole through which the comet has to fly for those severe consequences is small, circa 20,000 km so I doubt it happened even twice.

         Since Jupiter is big, the pass under the Roche limit would’ve taken several hours at 120,000 km close approach altitude and 30km/sec. For much of that time it would have been stretching and shearing in the neck portion. When you consider the large forces required to move the head 1 km from the body (even overcoming the tiny, sub 1mm/sec^2 acceleration due to gravity), it means much energy was expended. 

        The acceleration needed is small but when multiplied by the mass of the head to get the total force, that force is big. It follows that the energy expended just to lift the head from the body as if it’s a boulder with no cohesive attachment would be a lot (mass x g force x 1km height).  So any surplus acceleration in the Roche pass would likely have been commensurately large and available to overcome the cohesive forces of the cometary material. Again, when multiplied by the mass, each cubic metre is trying to pull apart from the two cubic metres either end of it with a force equal to the mass (of 1 cu m) x surplus acceleration. This is because the entire comet is in tension across its longest length as it flies through closest approach, pointing vertically down towards Jupiter (although that assumption is notwithstanding any previous rotational behaviour that might smudge and reduce the effect). That tension is equal to the the same surplus acceleration x mass. Furthermore, if the head and body refused to stretch, as they evidently did, that force would have been focussed on the weakest point, causing a catastrophic shear- that’s what I think I can see along the bottom of the neck i.e. the top of the body. After that shearing, the two lobes would work together to extrude the neck and make quite short work of it. 

        The energy in the extrusion, as it overcame the cohesive forces, would be dissipated as heat and probably melt the ice. That would reduce the cohesive forces greatly and cause a runaway breakdown of the cohesive integrity in the neck. If you bend a thickish metal bar back and forth a dozen times it can be too hot to touch. That’s a similar energy dissipation to extruding the bar across that same small length, if you had the strength to do it. So I think you’re right- what we see in the neck is a combination of the main stretch and some ice lava flow, especially in the middle. 

        The ‘neck’ photo from Cometwatch-18th October shows what looks like a crumbly concrete mix complete with a transverse fracture that you would expect if the head was being twisted and tipped down (this mirrors another characteristic of our modelled Roche passes, as the lobe passes beyond the vertical position but is still pulled back awkwardly on a different vector from the linear stretch vector that occurred in the vertical position). The head of 67P is indeed twisted by just a few degrees anticlockwise as you look down on it and possibly tipped forward too. 

        The Roche limit for Jupiter is 226,000km. Under that limit, Jupiter’s differential g forces would have been trying to ‘lift’ the head from the body but the comet’s own gravity would have resisted those small forces until a much lower altitude. I calculated that altitude to be around 135,000km above jovicentre (for a 3km long axis before stretching occurred). So 135,000km would be the altitude at which loose rocks would drift away from both ends of the comet and the head would come adrift if it had been a loose boulder. The comet would have to go lower if cohesive forces had to be overcome and clearly this was the case. My lower limit of 120,000 km is based on the fact that comet Shoemaker-Levy 9 shredded as it passed not far below this altitude (at 110,000km). But I suppose 67P/C-G could have gone lower than 120,000km if it had good integrity and then stretched at that lower altitude. The fact that SL-9 was shredded into 21 pieces at around this altitude means that any comet, 67P included, would undergo immense stresses. Also, the fact that SL-9’s 21 pieces exited the gravity well in a straight line betrays the stretching effect across the longest axis at or near closest approach. 

        Where did you mention tidal forces from Jupiter having a heating effect? I haven’t read the comments in some of the posts so I must’ve missed it. I would have replied if I’d seen it!

        • Robin Sherman says:

          A, Cooper

          It was clear to me from the images after the middle of September that there had been some sort of plastic or liquid flow of the surface. At first I thought this may be overflow from impact craters, where impact melted enough ice quickly enough for it to flow over the surface.

          Then the streamers or jets appeared and it seemed suggestions of tidal heating caused by the shape of the comet was leading to small scale cryovolcanism and the jets. In a post on 10th October I suggested more large scale cryovolcanism might have taken place to explain the apparent ice lava flows. Searching for a large enough energy source to enable this, it was an obvious association to make with the gas giants where volcanism and cryovolcanism has been explained by tidal heating. I wondered if tidal heating by Jupiter, as seen on Io most famously, might have been the energy source. I asked Claudia in one post if there was any evidence that 67P was once in orbit around Jupiter and thus could have undergone heating like Io. She seemed to think this unlikely, as once in orbit the chances of escaping were vanishingly small. I then returned to impact scenarios and tunnelling to generate the ice lava flows.

          It is striking how these molten surface features appear related on the head and body lobe and then some large energy event has occurred to gouge out the neck region. I’m glad you mentioned the twist, as that is something the contact binary scenario could find very difficult to explain unless angular momentum continued to affect the merger after impact.

          We have not seen the back of the neck clearly, but ESA’s shape model suggests the deep trench of the neck area does not continue right around the comet, this may be the small piece of stronger material that stopped the comet flying apart. It may not have melted but softened enough for some plasticity to allow the twisting.

          The frozen ice flows on the tops of the bodies also appear to be more compact and solid as if made of denser ice. A temporary melting or even just softening of the comets interior would have allowed separation of the material by density, the centripetal forces of the comets rotation forcing heavier material to the top of the head lobe and the bottom of the body lobe. An additional level of stress stretching the neck area, preventing gravity from closing the gap. Given the porous nature of the comet this separation may have occurred anyway without extensive melting, part of Marco’s ideas I was taken with. Every little nugget of insight helps.

          As I said I can only deal in ideas, I neither have the mathematical tools or data to calculate and add numbers to these scenarios, I’ve just watched lots of science documentaries and with a degree in Physics and Chemistry, have enough knowledge of the fundamentals to understand the basic science of possible scenarios.

  • Emanuele Andreola says:

    As far as i konw all aminoacids are made of carbon, oxygen, hydrogen an nitrogen. Moreover two aminoacids out of 20 also include sulfur.
    All of these compound are found in the comet.

    • Jacob nielsen says:

      C, O, H, N and S, as mentioned, are the elements that our 20 proteinogenic amino acids are made of. Many more amino acids can exist, and if Philae finds amino acids on 67p it may well be amino acids unfamiliar to our biology. One characteristic ‘feature’ held by ‘the 20’, is that they are all L- stereoisomeric, a property attributed to their biological synthesis. Finding an overweight of L-stereoisomers, will be somehow linked to biology.

  • Pamela welch says:

    It is an erroneous assumption (and one of my pet peeves) that rotten eggs smell like sulfur. Rotten eggs smell like foul, rotten meat. However, hard boiled eggs often have a sulfur odor.

    • logan says:

      You are right Pamela. Over boiled eggs, at yolk surface.

      • Robin Sherman says:

        Elemental Sulphur has virtually no odour. I have occasion to test powdered Sulphur in my job so I can tell you it hardly smells at all. The unpleasant odours associated with sulphur are Hydrogen Sulphide and the “bad eggs” smell of Sulphur Dioxide. Sodium Sulphide smells pretty bad too. Eggs are rich in Sulphur containing proteins. These are particularly important to make Keratin and hence feathers. When the proteins start to break down the Sulphur is released as Sulphur Dioxide. Over boiling an egg does the same thing.

        A duck should therefore be expected to be a source of these Sulphur compounds. 🙂

  • silver says:

    lots of sulfur, but how much silica? if this is an oort cloud object why should it have anything above carbon nitrogen or oxygen?

    what mechanism instilled sulfur into ‘oort’ cloud “lumps”, and if sulfur is there ( in proportion ) where the hell are the silicates( in proportion) ?

    • Jacob Nielsen says:

      I think we have to consider whether compounds are found in the gas-phase or in dust particles.. Sublimation points have to be considered when estimating relative abundances in the nucleus, based on what is currently detected by ROSINA.

    • logan says:

      Why not? Why not ‘unused’ dust? Why not supernova dust? Why not ‘dissolved’ nebulas? In the incommensurable void particle streams mix their ‘waves’ as if ‘phantoms’ where.

  • Marco says:

    @ A. Cooper. Thank you for putting the question here. I believe you are really on to something, and want to encourage you to keep asking and thinking along these lines. I think the matching head and body features that line up is the crucial eureka evidence at this stage. Please keep a record of images which show this piece of evidence, or put them on a website if you want an open forum on that evidence itself.
    Note that the rotation rate has sped up from 12.7 hours to 12.4 hours – see Wikipedia entry. If verified by peer review, this is also a pertinent piece of evidence of stretching possibilities. That is a very significant angular momentum shift which indicates large internal inward movement converted to angular momentum, or differential jets imparting the needed angular momentum. By my reckoning, a Roche pass would be more sudden and extreme force that would either break or leave unchanged the comet rather than stretch it. Jet actions would be gradual enough to stretch the comet without cleaving it apart in the same impulse.

    • A.Cooper says:

      Thanks for the reply, Marco. 

      Remember you said in the “landing site search narrows” blog entry that the comet could have spun up and then spun back down again. That hadn’t occurred to me at the time and it seems plausible especially if it’s now changed according to Wikipedia.  

      So there are two possible ways the comet could have stretched. If you recall, I worked out on that other blog post that the rotational period would have to be 1.5 to 2 hours to overcome the comet’s own gravity. I’d like to know the amounts of mass ejection and velocity during a 67P/C-G orbit in order to understand spin-up/spin-down better. A paper that was kindly linked below said 1 metre deep averaged over the comet surface disappears per orbit- and it was referencing 67P a lot though not relating it specifically to that 1 metre figure. It was an estimated average for comets if I recall. It must vary greatly according to perihelion. Anyway, 1 metre depth per orbit is quite a lot.

      As for the photos, I haven’t got a site and it’s difficult to point out correlations in writing. It’s much easier to sit with someone and point them out. This is partly because one can be fooled by shadows, foreshortening and perspective until the same feature has been seen from several different angles. I’ve studied every photo published. I sometimes see a stark, snow white area and know that it’s whited out from the sun angle and that there are stratified ridges in that same place which were apparent at lower sun angles. However, I think I’ll link the most obvious ones in due course, citing the blog post date for that photo. You really need to download all the photos to a file, and toggle between ones which look as if they match. 

      Please see my reply to Robin above too!

      • Marco says:

        @A Cooper. I think the mass loss over a full orbit (the change in rotational period is based on very accurate measurements made in 2009 and then recently when Rosetta came close enough in July) can account for the change in angular momentum, but only if a large percentage of it emanates from the lobes rather than the neck, and pushing jets counter to the rotation preferentially. This may be the jet state closer to perihelion. Other than that, internal movement of mass will affect rotation, or counter the effects of the jets. Eg if the jets are spinning up the comet, but mass is moving radially away from the neck at the same time rotation may stay static. Alternatively, a partial collapse of the neck region would speed up the rotation due to the conservation of angular momentum.

  • Cesar says:

    Sulfur dioxide don’t smell like vinegar. It is absolutely different smell, sulfurous and pungent.

  • A. Cooper says:


    I’ve been researching some bookmarks, hence the delayed reply. There are some puzzling aspects to those rotation rates you mentioned, although I don’t think it detracts from your theory of spin-up/spin-down. I had Lowry et al (2012) bookmarked, a paper which shows a rate of 12.76137 hours +/- 0.00006 hrs which appears to be very accurate as you noted. The JPL small body browser lists exactly the same value, citing that paper and “Rev March 2014” as an annotation. This implies it was revised in March 2014 and hadn’t budged a second since 2012. To confuse matters further, it says:

    “Result based on less than full coverage, so that the period may be wrong by 30 percent or so.”

    We also know for sure that it is 12.4 hours (from OSIRIS) so JPL is now out of date. So why is it stated with such exquisite precision in Lowry et al? I didn’t think Earth-based light curve data could resolve to better than a few percent anyway. I doubt if it stayed rock solid at 12.76 hrs for two years and then suddenly sped up since March. However, if it did, your theory is very strong indeed. If so it would be changing by the day and would have to be as a result of inward-moving material because the mass ejected at the moment isn’t enough for that sort of spin-up.

    My feeling is that the light curve data was a few percent off making it ~0.3 hours off the definitive OSIRIS value of 12.4 hrs. I still think the rotation period would be evolving slowly and has had plenty of time to spin up to 1.5 hrs and down again over thousands of years. It’s been a Jupiter family comet since at least 1599 (JPL ephemeris) and, although constantly perturbed, has not diverged more than about 1AU at aphelion or perihelion, except for the 7 or 8 somewhat closer perihelion passages since its 1959 close approach to Jupiter (not a Roche pass). In other words, it’s probably been outgassing for aeons, as it goes through the 2-4 AU mark. That’s plenty of opportunity to vary its spin rate as jets come and go according to erosion. There’s also the YORP effect that operates spin-up on very long time frames and is conjectured to be the reason for some asteroid shreddings when they reach the two-hour-rotation mark and can’t hold together any more.

    Anyway, I’m sure you’ll be eager to see how the spin rate evolves as the comet approaches perihelion. Of course, the most effective way to get a quick spin-up is for the ends to commence outgassing (in favourable directions) and I’m sure that’s most likely to happen when the comet is rotating head-on to the Sun. With that in mind, I worked out when that will be by matching the spin axis to the sun-centred ephemeris and came up with April 30th 2015. It would be about 12 degrees off head-on on March 30th and May 30th. So you could say it’s as good as head-on for those two months.

    Robin (above) said he thought a slab had cleaved from the head. I think he must mean the obvious end crater. I thought it was a giant, sunken/depleted vent and was looking forward to it gearing up to outgas when spinning head-on. I see Robin’s point especially as it’s at the end and subject to the most spin (your theory) or Roche delta g (my theory).

    • Marco says:

      Thanks A. Cooper. I have been checking back for a reply. This is my take on the spin. Back in 2009, C-P had a favourable perihelion with the earth fairly close. Thus, at a time of greatest jet activity, instruments from Earth could detect the spin rate with very good accuracy, which was 12.7 hours. No accurate measurements could be made from Earth since, as both activity reduced and distance increased. In March 2014, Rosetta was still many millions of kilometres away, but Osiris could make an approximation of the spin rate to within 30% accuracy. The 12.4 hours is the current, accurately determined spin rate from Rosetta’s instruments.

    • Marco says:

      For what it’s worth, I predict a further decrease in the orbital period, and a slight elongation of the comet also, over the coming perihelion.

    • Marco says:

      When the comet is spinning head on to the sun, tidal force from the sun would also be pulling the lobes apart.

  • Marco says:

    That should be rotational period. The lobes are not orbiting each other yet 🙂

  • A. Cooper says:


    I checked the ephemeris for this orbit and the next, perihelion to perihelion. Roughly speaking, this one is 6 years 5.5 months and the next one is 6 years 2.5 months so you are right, there will be a reduced orbital period next time.

    • Marco says:

      That is interesting in itself, but a further rotational acceleration, or a gradual process of elongation of neck, separation of lobes is what my hunch is.

  • davew says:

    Question for Prof. Altweg but also a general question for all; We now know that the D-H Ratio for 67P is higher than Earth’s Oceans and so perhaps Comets are NOT the source; But why does that point us back to Asteroids instead ?

Comments are closed.