≡ Menu

Low Metallicity in Compact Multi-Planet Systems

When astronomers talk about metals, they’re using the term in a specific sense. A metal in stellar terms is any element heavier than helium. Thus iron, silicon, magnesium and carbon qualify, all elements that are components of small, rocky planets. It was iron that John Michael Brewer (Yale University), Debra Fischer and colleagues singled out as a proxy in their recent work on the metal content of exoplanet systems. The work focuses specifically on compact, multi-planet systems as one of several system architectures found in close orbit of a host star.

What’s interesting here is that these domains seem mutually exclusive, or almost so. Unlike our Solar System, a system with multiple planets on tight orbits can squeeze its worlds into a region as close as Mercury. Likewise near the host star, we sometimes find massive planets in close orbits, known as ‘hot Jupiters.’ Few of these have close planetary neighbors, and few compact multi-planet systems have massive planets.

And there is another distinguishing factor. Where the relatively uncommon hot Jupiters appear, they are most numerous around stars with high metallicity, whereas small worlds appear around stars with a wide range of metallicity. Do systems lighter in metals have problems creating larger planets? The jury is still out, but some studies have made an association between low metallicity and small planet formation that Brewer and Fischer now strengthen in this new work.

Image: Artist’s conception of a compact multi-planet system. Credit: Michael S. Helfenbein.

The researchers compared systems of a particular architecture to all known planet hosts as a function of metallicity, adding in a third architecture, systems with cool Jupiters (orbiting further than 0.3 AU from the host star). They identified in their sample 104 hot Jupiter systems, 87 cool Jupiter systems and 105 compact multi-planet systems. The latter were defined as systems with three or more planets orbiting closer than 1 AU. Note that only one hot Jupiter system is found in a compact multi-planet system, while nine cool Jupiters are found in compact systems.

The abundance of compact, multi-planet systems around stars of low metallicity that emerges is clear and points in interesting directions. Many more of these systems may exist than we have assumed. Bear in mind that such worlds are a tricky catch for radial velocity methods, but because they tend to be co-planar, they can be spotted in transit searches, which can observe transits of multiple worlds. We seem to be in a very early phase of compact system detection.

It was Fischer who demonstrated over a decade ago that higher metallicity stars were the most likely to form large gas giants, in work that supported the core accretion model of gas giant planet formation. But smaller systems have been more difficult.

“Our surprising result, that compact systems of multiple, small planets are more likely around lower metallicity stars suggests a new, important clue in understanding the most common type of planetary system in our galaxy,” said co-author Songhu Wang, a 51 Pegasi b Fellow at Yale.

Standing out in this work is the apparent connection between iron and silicon. The researchers found a high silicon-to-iron ratio in stars with lower metallicity. Moreover, low-metallicity stars are long-lived, offering older, potentially habitable worlds whose longevity may make them prime targets for astrobiology. We are likely to learn that such systems are common. From the paper:

Stars of lower metallicity and higher Si/Fe [silicon to iron] ratios are generally older or members of the galactic thick-disk population (Kordopatis et al. 2015). This could point to a changing mix of planet architectures based on formation time and location. In fact, one of the oldest verified and low-metallicity planet hosts, Kepler-444, is home to a compact multi-planet system (Campante et al. 2015). New high-precision radial velocity surveys looking for Earth-massed planets (Jurgenson et al. 2016; González Hernández et al. 2017) may find a much larger population of small planets around these lower-metallicity stars.

Fischer, in this Yale University news release, likens the ratio of silicon to iron to a “thermostat for planet formation. As the ratio increases, nature is dialing up the formation of small, rocky planets.”

Image: This is Figure 2 from the paper. Caption: Stars with low metallicity or a high ratio of Si/Fe do not seem to form hot Jupiters, and are increasingly likely to host compact multi-planet systems. Plotted here are planet-hosting main sequence stars from our sample (green points), comparing the Si/Fe ratio to the log solar relative iron abundance, [Fe/H], with hot Jupiters (orange stars) confined to the lower right-hand side of the plot and compact multi-planet systems (blue triangles) making up a large fraction of the upper-left region. The Sun (yellow circle) is plotted for reference and the dashed lines are drawn to highlight the different populations. Stars with both high Si/Fe and high [Fe/H] are thought to be from the galactic thick-disk and are poorly represented in most planet search samples due largely to their greater distance. Credit: Brewer et al.

Exciting stuff, because we are moving into an era when new instruments like the Extreme Precision Spectrometer (EXPRES) developed by Fischer’s team at Yale, not to mention next-generation ground- and space-based resources, will make it more likely that we can find compact multi-planet systems where no transits occur.

The paper is Brewer et al., “Compact Multi-planet Systems are more Common around Metal-poor Hosts,” The Astrophysical Journal Letters Vol. 867, No. 1 (24 October 2018). Full text. Also note Zhu, “Influence of Stellar Metallicity on Occurrence Rates of Planets and Planetary Systems,” submitted to The Astrophysical Journal Letters, which likewise finds multi-planet systems more common around lower-metallicity stars (preprint).


Comments on this entry are closed.

  • Michael Fidler November 13, 2018, 14:52

    A hidden reservoir of multi-planet systems?
    by Eckhart Spalding | Nov 13, 2018

    “The fact that compact multi-planet systems can form around stars which have a range of metallicities also stands, with an interesting twist– at the lowest tested metallicities, -0.5 < Fe/H < -0.3, there appears to be a rising increase in the likelihood. Is there a glut of undiscovered multi-planet systems that lie undiscovered at still lower metallicities? There are ancient stars in the Milky Way with metallicities of Fe/H=-3.5 and even less! Kepler didn’t have a chance to say the final word on this, and what we need now are next-generation, high-precision RV measurements of those erstwhile underappreciated metal-poor stars."


  • Robin Datta November 13, 2018, 18:29

    Nucleosynthesis is stellar cores, supernova explosions and neutron star mergers produce metals, but those from the latter two include heavier elements and would be incorporated into newer generation stars. Could a paucity of the heavier elements that are needed in trace amounts by life on earth have a constraining influence on the emergence of life on older stars?

    • kzb November 14, 2018, 8:28

      The heaviest element needed for life on Earth is iodine. The only other 5th row element needed is molybdenum. Life uses these for somewhat niche uses and if they were not available I am sure life would find a workaround for it. Nothing is needed for life after the 5th row down on the periodic table.

  • Bruce D. Mayfield November 13, 2018, 19:43

    This finding improves the odds that old stars in the galaxy’s halo would also host rocky planets. Nice!

  • Rob Flores November 13, 2018, 21:42

    Just a moment; before declaring low metalicity systems
    complex life potentials;

    Are we throwing away the idea that the recycling of a planets crust is
    vital to continued biology? Because with lower iron content does that not imply a less dense core and therefor much faster cooling;
    Also is iron not the ideal liquid for creating a strong magnetic field on earth sized planets? With a Weaker magnetic field how is a planet’s atmosphere to survive long term, especially since the compact systems
    we are talking about are mostly red dwarfs (flaring)

    • Bruce D. Mayfield November 14, 2018, 14:40

      Agreed. And, while it’s exciting to find another area or population of stars that might have many planets, it may well be that even the unstable heavy elements like uranium are needed to keep the interior of a planet molten over the long term.

    • Ashley Baldwin November 14, 2018, 14:44

      Not necessarily. There are plenty of tightly packed planetary systems as metallicity increases . The correlation is between the SI:Fe ratio rather than the overall amount of either element per se . Such that although silicon increases from baseline as a greater function than iron, iron still increases with metallicity too and being heavier would migrate in larger quantities to an erstwhile terrestrial planetary core to form part of magnetic Dynamo .

      Bigger threats to the Dynamo would come from the tidally locked reduced rotation rate of close in planets , that contribute to the planetary magnetic moment especially in the convective liquid outer core.

      Also impacted by be the equally important transfer of heat across the core/mantle boundary due to reduced convection in an increasingly silicon dominated mantle.

  • Alex Tolley November 13, 2018, 21:59

    I wonder why the x-axis is Fe/H rather than H/Fe. The latter ratio, like Si/Fe, would mean low metallicity and the graph would have the expected positive correlation. Rather than a visual sense of ratios, the data could have been binned to show that the frequency of hot Jupiters increased as the metallicity increased, while compact rocky planet systems relatively increase as metallicity decreases.

    The other question I have is why a fixed distance is used for compactness, rather than some other factor like the nebula disk temperature as a way to normalize the data across the range of stellar types. After all, from an astrobiology POV, don’t we really care about the frequency of rocky worlds in the HZ capable of supporting [carbon based] life?

    • Ron S. November 14, 2018, 12:01

      “I wonder why the x-axis is Fe/H rather than H/Fe.”

      Does it really matter? The approximately center of the x-axis is zero, a ratio of 1, so the reciprocal would almost be a left-right mirror image. Perhaps their choice improves the visual presentation.

    • kzb November 14, 2018, 13:37

      I think it’s just the accepted way of presenting metallicity. [Fe/H] is the traditional way of presenting it. [H/Fe] might be read wrongly, simply because people are so used to seeing [Fe/H].

      • Alex Tolley November 14, 2018, 19:09

        And the same for Si/Fe rather than Fe/Si?

        • kzb November 15, 2018, 12:08

          Don’t know to be honest Alex! Also notice that is not log scale (apparently, because no square brackets on Si/Fe) but [Fe/H] is apparently log.

  • Michael November 14, 2018, 5:22

    Could the increased metal content of the later generation stars have more powerful magnetic fields. These fields could interact more with gas and dust nearing the proto star causing the observed distribution of planets.

  • kzb November 14, 2018, 8:20

    There was a paper a few weeks ago which argued small planet frequency DOES increase with metallicity. But small planets were defined as about Saturn mass and under if I remember correctly.
    Anyhow this paper is yet another one which supports terrestrial planets being common around low-metallicity stars. This is important for evaluating the extent of the GHZ (Galactic Habitable Zone). Studies like this support the idea that low-metallicity regions of the galaxy should be included, with the result that the Fermi paradox is intensified still further !

  • Harry R Ray November 14, 2018, 11:34

    Avery OBVIOUS exception to the above “rule” is TRAPPIST-1, which jhas a relatively high metalicity. Two new ArXiv preprints have just come out. ArXiv: 1811.04149; Magnetic Fields on the Flare Star TRAPPIST-1: Consequences for Radius Inflation and Planetary Habitability.” by D. J. Mullan, J. Macdonald, S. Dietrich, H Fausey – if TRAPPIST-1 has a poloidial field, it could prevent coronal mass ejections from hitting the planets, rendering them more habitable. ArXix: 1811.04877; “Disentangling the planet from the star in the late type M dwarfs: A case study for TRAPPIST-1g.” by Hannah R. Wakefield et al – no dense H/He atmosphere for TRAPPIST-1g to a 3 sigma confidence level, leaving ONLY TRAPPIST-1h with a H/HE atmosphere(or lack thereof)yet to be determined.

    • Harry R Ray November 15, 2018, 13:05

      After yesterday’s excitement, I finally had time to read the PDF’S. The Wakefield et al paper states that the Teff for TRAPPIST-1 is 111K LESS than the STANDARD value(i.e. 2400K as opposed to 2511+/- 37k. This has HUGE implications for TRAPPIST-d.

  • Harry R Ray November 14, 2018, 12:18

    Three tweets posted on PALE RED DOT: RedDots@REDDotsSpace: Major update tomorrow. Stay Tuned. PaleRedDot@Pale_red_dot: 23h uuuh! RedDots@RedDotsSpace: 1h Twitter feed starting at 17:50 GMT. See you in a bit!

    • Harry R Ray November 14, 2018, 14:12

      Barnard;s Star b: 3.5 Earth mass(minimum) planet with an orbital period of 233 days. “Nature” paper out tomorrow.

    • Ashley Baldwin November 14, 2018, 14:30

      The campaign is exploring for hab zone terrestrial planets around nearby red dwarfs . Barnard’s star at last perhaps ? ( if so, would have been music -and vindication- to dear old Peter Van de Kamp’s ears . God rest his soul)

      • ljk November 15, 2018, 10:51

        Although PVDK thought he had found two Jupiter type exoworlds circling Barnard’s Star (and later many others around other stars) and never changed his mind despite later astronomers finding errors in his work. Though I agree I think he would be happy just the same.

        The really sad thing is, he died just months before the announcement of the discovery of 51 Pegasi b in 1995. Although he must have known about the first pulsar exoplanets found in 1992.

  • Ronald November 21, 2018, 12:53

    Exciting stuff indeed! We are gradually learning more about the different possible configurations of planetary systems and their relative occurrence.

    Btw, I found fig. 1 in the Brewer paper at least as telling, maybe idea, Paul, to show that one as well?

    Pity that ‘compact systems’ was not very well defined with regard to planet size, but I assume here (from experience) that most of those planets are medium-sized: gas dwarfs to Neptunes.

    Some interesting quotes from the paper:
    “hot Jupiters and compact systems of multiple planets are nearly mutually exclusive.”
    “We show that compact multi-planet systems occur more frequently around stars of increasingly lower metallicities”.
    “(…) the steep rise in the fraction of hosts containing hot and cool Jupiters (with higher metallicity)”.
    “the occurrence rate of compact multis is higher than previously reported”.

    In the Zhu paper, see Fig. 2 (Cumulative distribution functions (CDF) of metallicities of stars in different categories): Above Fe of -0.1 there is a steep increase in giant planets.

    Also from another recent paper, Exploring the realm of scaled Solar System analogs with HARPS, Barbato et al. (2018):
    “The lack of planets found on inner orbits in our sample could be interpreted as evidence for the main observational prediction of the inward migration model, namely the anti-correlation between planetary systems hosting close-in super-Earths and those hosting long-period gas giants.”

    Combining all three (and prior knowledge), we can then conclude that compact systems of medium-sized planets are anti-correlated with *any* gas giants, hot, cool or (to a lesser extent?) even cold.

    So, tentatively, there arise the picture of (at least?) 3 different predominant types of planetary systems, as I have suggested before:
    1) Low-metallicity compact systems consisting of medium-sized planets (gas dwarfs, Neptunes), no gas giants.
    2) (Very) high metallicity systems with exclusively hot/warm gas giants, and (almost?) nothing else.
    3) Intermediate metallicity systems with cool/cold gas giants and small (terrestrial) planets on the inside.

    There may be narrow transitional zones (i.e. combinations) between those three types of systems of course.