≡ Menu

Two Unusual Brown Dwarfs

I track brown dwarfs closely because they have so much to teach us about the boundary between planet and star. I’m also intrigued by what might be found on a planet orbiting one of these objects, though life seems unlikely. Brown dwarfs begin losing their thermal energy after formation and continue cooling the rest of their lives, a period I’ve seen estimated at only about 10 million years. We know nothing about how long abiogenesis takes — not to mention how common it is — but the outlook for brown dwarf planets and astrobiology seems bleak.

It’s intriguing, though, that we’ve identified a number of brown dwarfs with planetary systems, including 2M1207b, MOA-2007-BLG-192Lb, and 2MASS J044144b, and in the latest news from the NEOWISE mission, we have two brown dwarfs that stand out for other reasons. What used to be the Wide-Field Infrared Survey Explorer would become a tool for the detection of near-Earth objects, but data from the earlier WISE incarnation is still turning up red and brown dwarfs. Along the way, the Backyard Worlds: Planet 9 project, armed with the talents of 150,000 citizen scientists, has more than proven its worth, but more on that in a moment.

Image: This artist’s concept shows a brown dwarf, a ball of gas not massive enough to power itself the way stars do. Despite their name, brown dwarfs would appear magenta or orange-red to the human eye if seen close up. Credit: William Pendrill (CC BY).

The brown dwarfs in question are notable because of their extreme lack of iron. In fact, a typical brown dwarf would have up to 30 times more iron and other metals than we find in either. Low levels of metallicity are something we would expect in ancient exoplanet systems, where supernovae spewing metals were scarce in nearby space, but the mechanisms involved remain problematic. Marc Kuchner is principal innvstigator for Backyard Worlds: Planet 9:

“A central question in the study of brown dwarfs and exoplanets is how much does planet formation depend on the presence of metals like iron and other elements formed by multiple earlier generations of stars. The fact that these brown dwarfs seem to have formed with such low metal abundances suggests that maybe we should be searching harder for ancient, metal-poor exoplanets, or exoplanets orbiting ancient, metal-poor stars.”

Adam Schneider (Arizona State) found one of the brown dwarfs, WISE 1810, back in 2016. Enter Backyard Worlds: Planet 9, which counts among its discoveries more than 1,600 brown dwarfs as well as the oldest known white dwarf, surrounded by a debris disk. Schneider was able to use WiseView, a tool created by Backyard Worlds citizen scientist Dan Caselden, to track the object’s motion, an indication of its relative proximity. The second brown dwarf is WISE 0414, this one discovered by a group of citizen scientists whose names deserve recognition: Paul Beaulieu, Sam Goodman, William Pendrill, Austin Rothermich, and Arttu Sainio.

Image: These images show the newly discovered brown dwarf WISE 1810 as seen with the WiseView tool. The object has an orange hue in these false-color images. In both images, a gray arrow on the left indicates the object’s position in 2010; the black arrow on the right indicates its position in 2016. Credit: Schneider et al. 2020.

The needed follow-up by astronomers using near-infrared spectroscopy confirmed that what the citizen scientists had found by working through hundreds of WISE images were indeed brown dwarfs. At the Jet Propulsion Laboratory, Federico Marocco and Eric Mamajek confirmed WISE 0414 using the Hale Telescope at Palomar Observatory in California. The unusual composition of both objects was apparent, as noted in the paper:

We find that the high proper motion objects WISEA 0414−5854 and WISEA 1810−1010 have exceptionally unusual spectroscopic and photometric properties that likely reflect significantly subsolar metallicities. The best-fitting models for these objects suggest very low metallicities ([Fe/H] ≤ −1), though no single model provides a satisfactory fit across all wavelengths. Further astrometric and spectroscopic observations are warranted to better characterize these enigmatic systems.

What caught my eye in this paper is its discussion of what the authors call a “stellar-substellar transition phase.” Metal-poor brown dwarfs — and these are extreme examples — show higher rates of cooling than other dwarfs. Moreover, hydrogen fusion does not stop abruptly as we scale down from star to brown dwarf, but rather, can continue in an interesting way:

…covering a narrow range of masses and large range of effective temperatures over which “unstable fusion” can occur. The masses encompassed by this transition, and the range of temperatures it spans depend on the metallicity and age of the population, with a more pronounced spread expected for metal-poor halo subdwarfs.

Some seven objects meeting this description have previously been found. WISE 1810 and 0414 have the lowest masses and effective temperatures yet discovered in this grouping. The authors point out as well that using proper motion instead of colors to identify cold, low luminosity objects has its advantages, including detecting dwarfs with spectra as unusual as these.

The paper is Schneider et al, “WISEA J041451.67-585456.7 and WISEA J181006.18-101000.5: The First Extreme T-type Subdwarfs?” accepted at the Astrophysical Journal (preprint).

{ 15 comments… add one }
  • Alex Tolley July 13, 2020, 10:39

    WISE 1810 and 0414 have the lowest masses and effective temperatures yet discovered in this grouping.

    knowing next to nothing about star formation, is there any chance what is happening is that the metals are fractionating by gravity? IOW, are the heavier elements falling into the core of the BD, and therefore not appearing in the spectra that are reflecting the composition of the surface layers? That these 2 BDs are smaller and colder than other BDs naively suggest that they may have more planet-type formation features, one of which is the separation of heavier elements to form metallic cores. An obvious test is to look at super Jupiters and their spectra for comparison.

    • Patient Observer July 13, 2020, 21:33

      …the heavier elements falling into the core of the BD…
      That conjecture reminded me of a stunning (to me) confirmation that there is significant nuclear fission occurring at the earth’s core. The more than smoking evidence was the unambiguous detection of antineutrinos consistent with nuclear fission. The resulting heat generation accounts for an estimated 50% of the total heat generation within the earth. The rest is from radioactive decay and other, less exotic, sources.


      Other articles on the topic suggest that the reactor periodically shuts down due to a build up of fission products that absorb neutrons. These products, being lighter (of course), diffuse outwards. At some point, the fission reaction resumes therey repeating the cycle.

      One assertion is that the flipping of earth’s magnetic field is due to the on-off nuclear reaction and the associated convection currents waxing and waning affecting the strength and orientation of the geomagnetic field.

      “Georeactors” have been proposed to exist in gas giants as well as very small stars that may help initiate nuclear fusion or, at least, produce enough thermal energy to keep the BD or whatever hotter than otherwise expected.

      Wonders will never cease.

  • wdk July 13, 2020, 12:07

    When I first heard some of the arguments about whether Pluto was a planet or not, I wanted to put my hands over my ears. But I guess there is kind of whirlpool effect about these arguments.

    If I understand the pre-print correctly, I believe that these examples of brown dwarfs or sub-dwarfs do not have planets – or satellites. And that designation issue was what led me to look at this dividing line for this
    boundary case. If the objects in the report are primaries, then secondary objects are either planets or moons. When Jupiter and the other gas giants were discovered to have satellites, no one was thinking about the possibility that these bodies were steps away from being suns themselves. So, if you have a brown dwarf in orbit around a star, as we frequently find, are its satellites moons or planets? Same issue, free floating.

    The other thing to note was that many brown dwarfs were iron deficient. Elsewhere we had some side discussion about transition through gas giant to red dwarf, finally noting that the smallest red dwarfs had had convection enough enough to keep the pile up of
    fusion “ashes” from shutting the hydrogen fusion process down for
    many solar life times ( it feels unnatural to say “trillions of years”).
    What with short deuterium (or other fusion processes), just wondering if the shutdown would tend to trap iron in non-convective zones?
    Or, is the iron representative of low metallicity or a particular
    “vitamin” deficiency?

    • Ron S. July 13, 2020, 13:58

      When I was a teenager I had a friend with an odd relationship with music. He liked rock and hated jazz. When a song came out that was somewhere in between he couldn’t decide whether or not he liked it because it was very important to him that he first decide whether it was rock or jazz.

      We tried to convince him to focus on whether he liked the song and not on its (arbitrary) classification. He couldn’t and didn’t.

  • Bruce D Mayfield July 13, 2020, 14:58

    I had two questions re these BD that shed light on these questions about their low Fe contents; what are their masses and ages. I found this on the Nasa Neowise site reporting this discovery:
    “Scientists call the newly discovered objects “the first extreme T-type subdwarfs.” They weigh about 75 times the mass of Jupiter and clock in at roughly 10 billion years old. These two objects are the most planet-like brown dwarfs yet seen among the Milky Way’s oldest population of stars.”

    If memory serves 80 Jupiters is the mass needed to have the start of main sequence stellar proton-proton fusion, so these BD are close to being massive enough to be RD stars, which are fully convective. So these almost stars should have had much convection to for a long fraction of their lives, but once their initial supplies of Deuterium and Li ran out the energy source driving convection would end. So it shouldn’t be surprising that, as Alex Tolley suggested, Fe would tend to settle out. But also, their would have been less Fe to start with 10BYA, so again not surprising to see low Fe for these objects.

    • Andrei July 16, 2020, 12:16

      That Sir is a very good and convincing line of thought.
      Patient Observer also added some interesting points that added to my knowledge on these objects, that beside the “unstable fusion” which I already suspected would occur – yet very nice to have confirmed here.
      So thank you again Mr Gilster for an interesting entry!

  • Geoffrey Hillend July 13, 2020, 17:20

    Quote by Alex Toley: “knowing next to nothing about star formation, is there any chance what is happening is that the metals are fractionating by gravity? IOW, are the heavier elements falling into the core of the BD, and therefore not appearing in the spectra that are reflecting the composition of the surface layers? That these 2 BDs are smaller and colder than other BDs naively suggest that they may have more planet-type formation features, one of which is the separation of heavier elements to form metallic cores. An obvious test is to look at super Jupiters and their spectra for comparison.”

    I like your idea which seems very plausible. If true, then being metal poor has nothing to do with a brown dwarfs age.

    • Henry Cordova July 13, 2020, 20:18

      Energy is transferred from a star’s core to surface by three mechanisms, conduction, radiation and convection. The relative importance of these varies during a star’s lifetime, its evolutionary history and its mass. I do know that smaller main-sequence stars have an extensive convection zone in their upper layers, and massive stars transfer energy by conduction in their cores, but that most of the time radiative transfer is the mechanism by which energy escapes the core and moves to the surface. Which mechanism dominates at what level in each type of star during its history is the stuff that stellar models are made of. It has something to do with temperature, pressure, intensity and opacity gradients within the star itself and how they interact.

      Its been a long time since I’ve studied these issues, so I imagine there has been some progress in stellar structure since I took those courses…

      • Ron S. July 14, 2020, 7:54

        When the extinction distance is so short within the ultra-dense stellar interior I find it difficult to think of radiation heat transfer being distinguishable from conduction. The mechanisms differ but perhaps not in their effects.

        • Henry Cordova July 14, 2020, 11:54

          Perhaps, but convection allows allows fusion products to be transported to upper layers where they are detectable by spectroscopy. It also means that the chemistry of the deeper levels evolves as some nucleides are removed and replaced by surface material.

          I also learned that conduction plays a role in the cores of very massive stars because quantum effects in electron degenerate plasmas tend to modify the gas laws. At a certain point higher temperatures and pressures can no longer push back against the weight of the upper layers, and the core collapses. This is one of the causes of supernovae.

          I’m certainly no expert, I took only one course in stellar structure and evolution a long time ago, but the material stayed with me because of how, even back then, human beings had the audacity to seriously consider what is going on inside the stars. And I think that is wonderful.

          • Ron S. July 14, 2020, 15:06

            You may have misread since I didn’t mention convection and I certainly agree with you on that point. All you need is an energy gradient and enough freedom of motion for mass transport. Non-degenerate stars and BD, etc. have that. That may not apply for certain more compact objects, but that’s outside the scope of the article.

            If by radiation we mean EM the object must be less compact than a white dwarf. Then we are left with QM to address energy transfer and a more entertaining equation of state. I really don’t think we disagree.

            Fusion interruption (fuel exhaustion), and the consequent equilibrium change, followed by collapse and rebound (nova) is another matter. When/if it occurs in a BD it is a “gentler” process.

  • Henry Cordiva July 13, 2020, 19:51

    Does the Mass Function for stars have a discontinuity somewhere between red dwarfs and gas giant planets? I.e., do the numbers of these objects gradually increase in the Brown Dwarf zone between these two extremes, or we are talking about three fundamentally different object types here, with different origins and histories? I know “stars” have internal thermonuclear reactions that provide the heat, and this is not possible below about 0.1 solar masses, and brown dwarfs glow from the gradual release of gravitational energy. Gas planets still “glow” at certain wavelengths, but are cold at the tops of their atmospheres. But do their numbers increase as mass decreases?

    • Ivar Brouwer July 14, 2020, 5:38

      Isn’t it that a Brown Dwarf fuses Deuterium (or has fused Deuterium) that makes them a brown dwarf? As such I think that would make them stand out from regular planets that can’t fuse but only mix and decay. (Planets can’t make heavier elements)

      It’s origins, I assume are not in play: heavy planets & stars seem to form preferable in pairs/groups. But can form alone. There is a cut of from which a cloud of matter can collapse in on itself but that has no direct baring on capability Of the object to fuse later in live.
      (I hope I got this right, by no means an expert)

  • Mike Serfas July 14, 2020, 6:08

    According to https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3612282/ there is a potentially habitable zone (a very small one, naturally) around brown dwarfs even after billions of years. They seem very pessimistic on the chance the planets retain water through the bright early phase, and note the possibility that tides could either provide too much heat or push the planet out of the habitable zone. Still, I’m not sure they really considered all those possibilities, and their most basic finding is indeed that a planet could be warmed by a failed star.

  • wdk July 14, 2020, 17:55

    Had a response on the underlying question, but I think my server went down before submission…
    The brown dwarf range of 13 to 80 Jupiters has been based on the modeling notion that that range is a region where deuterium fusion ( H with a neutron and proton) would generate helium 3. This is exothermic and would be an episode in early stellar or substellar life. An early report I remember from those stellar structure course days too, was a 300,000 year period of production. Looking on line I see estimates that range around 40 to 100 million years. Improved models, probably. But if deuterium burning imitates hydrogen burning, one would expect more massive brown dwarfs burning brighter and faster much as main sequence stars would with hydrogen. Brown dwarfs
    and gas giants with lesser masses ( 1 to 13 Jupiters) appear not to vary much in radius, but their thermal reservoirs could be significantly different due to considerable deuterium burning.

    It is also noted:

    1. That deuterium burning occurs in both brown dwarfs and stars and tends to expand outward in a shell.
    2. That deuterium burning is not as intense as hydrogen burning. And since it ignites earlier, the environment it creates allows continued mass
    accretion building toward main sequence stars. Were there no deuterium fusion, but just hydrogen, MS star mass build up would have a hard time getting past 3 solar masses. Now that might create a universal shortage of iron, no?

    Since the deuterium burning is a spherical shell heading outward in both brown dwarfs and early MS stars, conditions in its wake might depend on heat sources and heat transport. If there is no ignition of hydrogen, then things boil (?) down to the question of which heat transport will remain after deuterium fusion shutdown: convective, radiant or conductive. At the low end of mass, I would suspect conduction due to high pressure and low temperatures. At the end close to red dwarf mass, it might depend on how much red dwarf convection is dependent on hydrogen ignition, a chicken or the egg proposition.

Leave a Comment