Evidence for an Anhydrous Carbonaceous Extrasolar Minor Planet
M. Jura, P. Dufour, S. Xu, B. Zuckerman, B. Klein, E. D. Young, C. Melis
EEvidence for an Anhydrous Carbonaceous Extrasolar MinorPlanet
M. Jura a , P. Dufour b , S. Xu a,c ( 许 偲 艺 ), B. Zuckerman a , B. Klein a , E. D. Young d , & C.Melis e ABSTRACT
Using Keck/HIRES, we report abundances of 11 different elements heavierthan helium in the spectrum of Ton 345, a white dwarf that has accreted oneof its own minor planets. This particular extrasolar planetesimal which was atleast 60% as massive as Vesta appears to have been carbon-rich and water-poor;we suggest it was compositionally similar to those Kuiper Belt Objects withrelatively little ice.
Subject headings: planetary systems — white dwarfs
1. INTRODUCTION
Because of the vagaries of gravitational dynamics that occur within a white dwarf’splanetary system, a minor planet’s orbit can be strongly perturbed so that it passes closeenough to the host star to be tidally disrupted (Debes & Sigurdsson 2002; Bonsor et al.2011; Veras & Wyatt 2012; Frewen & Hansen 2014). A circumstellar disk is then formed,and the white dwarf ultimately accretes the resulting debris, thereby imparting a signature inthe stellar spectrum which would otherwise be essentially pure hydrogen or, less often, purehelium (Jura 2003). By determining the abundances of heavy elements in the atmosphere of a Department of Physics and Astronomy, University of California, Los Angeles CA 90095-1562;[email protected], [email protected], [email protected] b D´epartement de Physique, Universit´e de Montr´eal, Montr´eal, Qu´ebec H3C 3J7, Canada; [email protected] c European Southern Observatory (ESO), Garching, Germany; [email protected] d Department of Earth, Planetary, and Space Sciences, University of California, Los Angeles, Los AngelesCA 90095, [email protected] e Center for Astrophysics and Space Sciences, University of California, San Diego, CA 92093-0424;[email protected] a r X i v : . [ a s t r o - ph . E P ] N ov and CO . Consequently, little carbon accumulates into planetesimals unless they form in anextremely cold environment far from the central star. Therefore, within asteroids, carbonis expected to be significantly more depleted than oxygen, as typically found in the innersolar system (Lee et al. 2010) and extrasolar planetesimals (Jura & Young 2014; Xu et al.2014).This simple snow line scenario for carbon/oxygen ratios is not universally valid. Anhy-drous Interplanetary Dust Particles (IDPs), among the most primitive material in the solarsystem, are relatively carbon-rich and are thought to derive from comets despite their lack ofhydrous minerals (Thomas et al. 1993). These anhydrous IDPs might therefore be relatedto those Kuiper Belt Objects (KBOs) such as Haumea (Lacerda & Jewitt 2007; Lockwoodet al. 2014) and Eris (Sicardy et al. 2011) that are sufficiently dense to be no more than 15%ice by mass (Brown 2012) even though they likely contain large amounts of carbon. Here,we suggest that the minor planet being accreted onto Ton 345 is compositionally similar toan ice-poor KBO.
2. TON 345
Originally identified as a faint blue star at high galactic latitude, Ton 345 ( = WD0842+231) with m(g) = 15.73 mag in the Sloan Digital Sky Survey (SDSS) has an atmospherecomposed almost entirely of helium. Ton 345 was singled out to be of special interest becauseit displays broad emission lines characteristic of a circumstellar gaseous disk orbiting withinthe tidal radius of the central white dwarf (Gaensicke et al. 2008). Subsequently, thisstar also has been found to have excess infrared emission produced by an orbiting dust disk(Brinkworth et al. 2012; Farihi et al. 2010; Melis et al. 2010). Experience has shown thatat high spectral resolution, multiple elements can be detected in white dwarfs that display 3 –excess infrared emission, and we therefore obtained spectra at the Keck I telescope.As listed by Xu et al. (2013), there are four well-studied white dwarfs with heliumdominated atmospheres, dust disks, and more than 10 elements heavier than He detectedin their atmospheres: GD 362, GD 40, PG 1225-079 and WD J0738+1835, the only one ofthese four also to display a gaseous component to its circumstellar disk. Here, we presentresults for Ton 345, an additional white dwarf with these distinctive characteristics.
3. OBSERVATIONS
The data reported here were acquired in 2008 at the Keck I telescope with HIRES (Vogtet al. 1994), an echelle spectrograph with a spectral resolution near 40,000. Table 4 of Meliset al. (2010) provides the exact exposure times and dates. In total, we obtained 6600 s and9000 s of exposure time for the blue spectral range between 3130 ˚A and 5960 ˚A and the redspectral range between 4600 ˚A and 9000 ˚A, respectively.The spectra were extracted from the flat-fielded two-dimensional image of each exposureas described in Klein et al. (2010, 2011). The most challenging task was removing the broadundulations in the continuum likely caused by variable vignetting (Suzuki et al. 2003). Also,as described in Klein et al. (2010), an additional re-normalizing processing step was appliedto calibrate and remove second (diffraction) order flux contamination in the region 8200 -9000 ˚A. Wavelength calibration was performed using the standard Th-Ar lamps. FollowingKlein et al. (2010, 2011), we used IRAF to normalize the spectra and combine echelleorders.With our signal to noise ratio which varied between 30 and 45, lines with an equivalentwidth as weak as 20 m˚A or even a little less can be detected. As a result, well over 100 linesfrom 11 elements heavier than helium are seen; there are no unidentified lines.As noted by the referee, archived
Hubble Space Telescope data acquired with the
CosmicOrigins Spectrograph were acquired under program ID
4. ELEMENTAL ABUNDANCES
We first tried to estimate the atmospheric parameters by fitting the
Sloan Digital SkySurvey (SDSS) spectroscopic data with a grid appropriate for DB white dwarf stars. However,given that many of the helium lines are contaminated by metal absorption lines, we donot believe that the standard spectroscopic technique can provide accurate atmosphericparameters, especially for the surface gravity. We thus decided to instead obtain the effectivetemperature by fitting the ugriz photometry and keeping log g fixed at 8.0. We obtain T eff = 19,535 ± ± T eff = 18,700 K, log g = 8.0, log H/He = -5.0 and the approximate amount of heavy elements discussed above, wenext analyze the Keck observations following the procedure described in Dufour et al. (2012).The heavy element abundances found in this way were then used to compute a new modelatmosphere (we keep T eff and log g fixed to the values cited above). We next repeat thefitting procedure using this new atmospheric structure and find that the abundances remainpractically unchanged, indicating that we have converged to a final solution (see Table 1)and that no further iteration is required. Most of the lines are well separated from each otherand therefore we can infer a value of an elemental abundance from an individual line. Thedispersion in the various measurements for a given element can roughly be used as minimumabundance uncertainty, which is typically around 0.1 dex. For some elements, such as O, thedispersion of the abundances determined from individual lines can be as small as 0.02 dex.However, given the uncertainties in continuum placement, atomic parameters and the modelatmosphere, we adopt a minimum error associated with each abundance determination of 0.1dex. There are also uncertainties associated with the effective temperature and gravity, but,fortunately, the relative abundances are insensitive to small variations in these parameters(Klein et al. 2011). We show model fits to the spectral lines listed in Table 2 in Figures 2-9. 5 –Fig. 1.— SDSS optical spectrum of Ton 345. Black denotes the data; the agreement of theHe line profiles with the model denoted in red, supports our inferred atmospheric parameters. 6 –As far as we know, Ton 345 is the only externally-polluted white dwarf to display opticalcarbon lines as seen in Figure 2. The carbon to oxygen ratio in Ton 345 is a factor of 10 -100greater than found in other heavily polluted white dwarfs (Jura & Young 2014). The H α lineis at best only very marginally detected; we can only place an upper bound to the hydrogenabundance. Some of the Si II lines are not well fit; previous studies also have found difficultiesin simultaneously fitting all available silicon lines in the spectra of externally polluted whitedwarfs (Jura et al. 2012; Gaensicke et al. 2012).Table 1 – Abundances in Ton 345Element [log n (Z)/ n (He)]H ≤ - 5.47C -4.63 (0.19)O -4.58 (0.10)Mg -5.02 (0.10)Al -5.96 (0.10)Si -4.91 (0.12)Ca -6.23 (0.10)Ti -7.74(0.10)Cr -6.91 (0.10)Mn -7.54 (0.10)Fe -5.07 (0.10)Ni -6.20 (0.10)In this Table, we follow astronomical convention and report abundances by number. Keylines used in the abundance determinations include: H α ; C II 4267 ˚A, 6578 ˚A; O I 7771 ˚A,7774 ˚A, 7775 ˚A; Mg I 3838 ˚A, 5184 ˚A; Mg II 7877 ˚A, 7896 ˚A; Al II 4663 ˚A, 6243 ˚A, 7042 ˚A,7471 ˚A; Si II 3210 ˚A, 3854 ˚A, 3856 ˚A, 3863 ˚A, 4128 ˚A, 4131 ˚A, 5958 ˚A, Ca II 3159 ˚A, 3179˚A, 3181 ˚A, 3706 ˚A, 3737 ˚A, 3933 ˚A, 3968 ˚A; Ti II, Cr II, Fe II: multiple lines (Klein et al.2010); Mn II 3442 ˚A, 3460 ˚A, 3474 ˚A; Ni II 3514 ˚A.
5. THE ACCRETED PARENT BODY
Because Ton 345 possesses a dust disk, it is likely that the pollution results from onelarge parent body with a well defined angular momentum vector; otherwise grains likelywould be destroyed by mutual collisions (Jura 2008).While dredge-up might enhance the carbon abundance in white dwarfs with effectivetemperatures near 25,000 K, this process is probably negligible for stars cooler than 20,000 7 –Fig. 2.— Spectrum of Ton 345 with lines of C, O, and Fe. Black denotes the data while reddenotes the model with our inferred abundances. The model is wavelength shifted to thephotospheric frame of the star; wavelengths are in air and the heliocentric frame of rest. 8 –Fig. 3.— The same as Figure 2 with lines of Mg, Al, and Si. 9 –Fig. 4.— The same as Figure 2 with lines of Al and Si. 10 –Fig. 5.— The same as Figure 2 with lines of He, Si, Cr and Fe. 11 –Fig. 6.— The same as Figure 2 with lines of Ca, Cr and Fe. 12 –Fig. 7.— The same as Figure 2 with lines of He, Ca, Ti, Cr and Fe. 13 –Fig. 8.— The same as Figure 2 with lines of Si, Mn, Cr, Fe, and Ni. 14 –Fig. 9.— The same as Figure 2 except for selected lines of H, Fe and Ni. The fit to H α represents our upper bound to the abundance of this element which is only very marginallydetected. 15 –K (Koester et al. 2014b). We therefore proceed by assuming that all of the heavy elementsin the atmosphere of Ton 345 are accreted from its circumstellar disk.Because different heavy elements settle at different rates, the abundances within thephotosphere of an externally-polluted white dwarf do not necessarily directly reflect theabundances in the parent body (Koester 2009). For Ton 345, the typical settling time is10 yr, comparable to estimates for the lower bound of a typical dust disk lifetime (Girvenet al. 2012). Possibly, the outer convective zone might be in a steady state where therate of accretion is balanced by the rate of gravitational settling. Alternatively, because itis both observed and theoretically predicted that accretion rates onto externally-pollutedwhite dwarfs can be variable on time scales much shorter than 10 yr (Metzger et al. 2012;Rafikov & Garmilla 2012; Wilson et al. 2014; Xu & Jura 2014), the abundances in theatmosphere of Ton 345 might reflect a recent burst of accretion.Here, we assume the “instantaneous” approximation where the abundances in the pho-tosphere equal the abundances in the parent body. If the system is in a steady state, therelative abundances of the lighter elements, C through Si, would be unchanged because theirrelative settling times differ by less than a factor of 1.1 from their mean value. In contrast,the relative abundances of the heavier elements such as Ca and Fe would increase by a factorof two. However, even though the fraction of the mass of the parent body mass carried inthese heavy elements would be larger, our most important results – that carbon is unusuallyabundant and the material has little water, would be unaltered.There are two arguments that the parent body accreted onto Ton 345 was anhydrous.If the matter accreted onto the white dwarf is carried within familiar minerals, then Mg, Al,Si, and Ca are bonded to oxygen in the proportions matching the oxides MgO, Al O , SiO and CaO. Iron may be found either as an oxide or in metallic form. By this mineralogicalargument, we find that all the oxygen is bound into minerals; none is left to form water.Also, the accreted minor planet was low in water because there is relatively little hydrogen inthe atmosphere of Ton 345; from Table 1 and that most of the oxygen was bound in oxides,we compute that less than 10% of the oxygen was in the form of H O.We now consider the composition of the accreted parent body. Using the abundanceslisted in Table 1, we compute the mass fraction of each element as provided in Table 2.Because of its large carbon abundance, it is possible that the planetesimal accreted ontoTon 345 resembles the most primitive meteorites, the CI chondrites. However, such a fitto our data is not very good, because CI chondrites have relatively more oxygen and lesscarbon then we measure in Ton 345. Among five well studied CI chondrites, the carbon ∼ koester/astrophysics/
16 –mass percentage is 3.5% with a dispersion of 0.48% and a maximum value of 4.4% (Lodders2003), much less than inferred for the material accreted onto Ton 345. Among the same fiveCI chondrites, the average oxygen mass percentage is 46% with a dispersion of 5.8 % and aminimum value of 41%(Lodders 2003) is notably higher than our value of 23% inferred forthe minor planet accreted onto Ton 345.Because of its high carbon to oxygen abundance ratio, the most familiar solar systemmaterial that matches the composition seen in Ton 345 is anhydrous Interplanetary DustParticles (IDPs), primitive matter whose average mass fractions (Thomas et al. 1993) alsoare given in Table 2. Approximately 50% of IDPs are anhydrous (Flynn et al. 2003), andthese are the ones we consider here. The elemental mass fractions for Ton 345’s pollutionand of anhydrous IDPs agree except for Ni.For our model atmosphere, we compute that the He mass in the outer convection zoneis 9.1 × g. Consequently, from the abundances given in Table 1, the total mass of theaccreted parent body must have been at least 1.6 × g, about 60% of the mass of Vesta(Russell et al. 2012) If its density was 3 g cm − , then the parent body diameter was atleast 470 km, well within the range inferred for these parameters for extrasolar planetesimalsaccreted onto heavily polluted white dwarfs (Jura & Young 2014).Table 2 – Percentage of Total MassElement Ton 345 IDP% %)C 15 (5.6) 12.5 (5.7)O 23 (4.0) 32.9 (4.0)Mg 13 (2.5) 10.7 (4.6)Al 1.6 (0.36) 1.3 (1.1)Si 19 (4.2) 14.6 (2.9)Ca 1.3 (0.29) 0.9 (0.3)Cr 0.35 (0.079) 0.2 (0.1)Mn 0.086 (0.020) 0.2 (0.3)Fe 26 (4.4) 17.6 (6.3)Ni 2.0 (0.45) 0.7 (0.4)Following cosmochemical convention, the abundances ratios are expressed by mass ratherthan by element number. The values for Ton 345 are derived from Table 1. Except for Hand Ti whose abundances were not reported, the values for anhydrous IDPs in the thirdcolumn are taken from Table 3 of Thomas et al. (1993) with their 1 σ dispersion given inparentheses. 17 –
6. DISCUSSION
Previous observations have shown extrasolar minor planets accreted by white dwarfstypically have little water (Jura & Xu 2012), although the parent body accreted onto GD61 is an exception in being water-rich (Farihi et al. 2013). In contrast, the object accretedonto Ton 345 is carbon-rich and water-poor indicating that the carbon to water content inextrasolar minor planets varies by more than a factor of 100.A star’s C/O ratio influences the carbon and water content of planetesimals that areformed in its protoplanetary nebulae (Johnson et al. 2012). However, essentially all mainsequence stars near the sun have more oxygen than carbon (Fortney 2012; Nissen 2013;Teske et al. 2014) and the observed range in the carbon to oxygen ratio among externallypolluted white dwarfs must be largely a consequence of substantial separation of these twoelements during formation and evolution.Heavily polluted white dwarfs with dust disks typically have low carbon abundances(Jura & Young 2014). In contrast, Koester et al. (2014b) have found that some whitedwarfs where they only report C and Si abundances with relatively modest levels of pollutionhave notably higher carbon to silicon abundance ratios than do systems where abundancesof multiple elements have been measured. One possible explanation for the pollution withrelatively high carbon abundance is that the parent bodies originate in the outer planetarysystem. However, we do not yet have a full understanding of the cosmochemical evolutionof carbon in extrasolar planetary systems.Because we have measured abundances of 11 heavy elements in the atmosphere of Ton345, we can constrain many potential scenarios for the origin of the accreted minor planetaccreted. Brown (2012) has suggested that high-density KBOs with relatively little waterare the consequence of the collisional erosion of a differentiated parent body which had an iceshell and a rocky core. Because differentiation is widespread among extrasolar planetesimals(Jura et al. 2013), some related model may pertain to the minor planet accreted onto Ton345. We show in Figure 10 a comparison between the mass fractions listed in Table 2 forTon 345 and those for an anhydrous IDP, normalized to the abundances in CI chondrites.The good agreement can be understood as being the consequence of the evolution of adifferentiated minor planet which initially had an iron core, a rocky mantle and an iceexterior. At some later time, all the ice, some of the mantle and none of the core was losteither by a collision or some other process such as sublimation of surface water ice during thestar’s highly luminous red giant phase before it became a white dwarf (Jura 2004) althoughburied water ice could be retained (Jura & Xu 2010). Regardless of how this hypotheticaldifferentiated planetesimal lost its outer ice, ultimately, it would achieve a bulk compositionresembling those solar system KBOs with relatively high density and little water. 18 –
7. CONCLUSIONS
We have obtained optical spectra of Ton 345 and measured the abundances of 11 el-ements heavier than helium. We find that we are observing the disintegration of a minorplanet that likely was carbon-rich and ice-poor; it appears to have been compositionallysimilar to a high density Kuiper Belt Object.This work has been partly supported by the NSF. We thank M. Nabeshima for helpwith the data analysis.
REFERENCES
Bergin, E. A. 2013, to be published, XVII Special Courses at the National Observatory ofRio de Janeiro. AIP Conference Proceedings, astro-ph 1309.4729Bonsor, A., Mustill, A. J., & Wyatt, M. C. 2010, MNRAS, 414, 930Brinkworth, C. S., Gaensicke, B. T., Girven, J. M. et al. 2012, ApJ, 750, 86Brown, M. E. 2012, Ann. Rev. Earth Plan. Sci., 40, 467Debes, J. H., & Sigurdsson, S. 2002, ApJ, 572, 556Dufour, P., Bergeron, P., & Fontaine, G. 2005, ApJ, 627, 404Dufour, P., Kilic, M., Fontaine, G. et al. 2010, ApJ, 719, 803Dufour, P., Kilic, M., Fontaine, G. et al. 2012, ApJ, 749, 6Farihi, J., Jura, M., Lee, J.-E., & Zuckerman, B. 2010, ApJ, 714, 1386Farihi, J., Gaensicke, B. T., & Koester, D. 2013, Science, 342, 218Flynn, G. J., Keller, L. P., Feser, M., Wirick, S., & Jacobsen, C. 2003, Geochim. Cosmochim.Acta, 67, 479Fortney, J. 2012, ApJ, 747, L27Frewen, S. F. N., & Hansen, B. M. S. 2014, MNRAS, 439, 2442Gaensicke, B. T., Koester, D., Marsh, T. R., Rebassa-Mansergas, A. & Southworth, J. 2008,MNRAS, 391, L103 19 –Gaensicke, B. T., Koester, D., Farihi, J. et al. 2012, MNRAS, 424, 333Girven, J., Brinkworth, C. S., Farihi, J. et al. 2012, ApJ, 749, 154Henning, T. & Semenov, D. 2013, Chem. Rev., 113, 9016Johnson, T. V., Mousis O., Lunine, J., & Madhusudhan, N. 2012, ApJ, 757, 192Jura, M. 2003, ApJ, 584, L91Jura, M. 2004, ApJ, 603, 729Jura, M. 2008, AJ, 135, 1785Jura, M. & Xu, S. 2010, AJ, 140, 1129Jura, M. & Xu, S. 2012, AJ, 143, 6Jura, M., Xu, S., Klein, B., Koester, D., & Zuckerman, B. 2012, ApJ, 750, 69Jura, M., Xu, S., & Young, E. D. 2013, ApJ, 775, L41Jura, M. & Young, E. D. 2014, Ann. Rev. Earth Plan. Sci., 42, 45Klein, B., Jura, M., Koester, D., Zuckerman, B., & Melis, C. 2010, ApJ, 709, 950Klein, B., Jura, M., Koester, D., & Zuckerman, B. 2011, ApJ, 741, 64Koester, D. 2009, A&A, 498, 517Koester, D., Gaensicke, B., & Farihi, J. 2014a, A&A, 566, 34Koester, D., Provencal, J., & Gaensicke, B. T. 2014b, A&A, 568, 118Lacerda, P. & Jewitt, D. C. 2007, AJ, 133, 1393Lee, J.-E., Bergin, E. A., & Nomura, H. 2010, ApJ, 710, L21Lockwood, A. C., Brown, M. E., & Stansberry, J. 2014, Earth Moon Planets, 111, 127Lodders, K. 2003, ApJ, 591, 1220Melis, C., Jura, M., Albert, L., Klein, B., & Zuckerman, B. 2010, ApJ, 722, 1078Metzger, B. D., Rafikov, R. R., & Bochkarev, K. V. 2012, MNRAS, 423, 505Nissen, P. E. 2013, A&A, 552, 73 20 –Provencal, J. L., Shipman, H. L., Koester, D., Wesemael, F., & Bergeron, P. 2002, ApJ, 568,324Rafikov, R. R., & Garmilla, J. A. 2012, ApJ, 760, 123Russell, C. T., Raymond, C. A., Coradini, A. et al. 2012, Science, 336, 684Sicardy, B., Assafin, M., Jehin, E. et al. 2011, Nature, 478, 493Suzuki, N., Tytler, D., Kirkman, D., O’Meara, J. M., & Lubin, D. 2003, PASP, 115, 1050Teske, J. K., Cunha, K., Smith, V. V. Schuler, S. C., & Griffith, C. A. 2014, ApJ, 788, 93Thomas, K. L., Blanford, G. E., Keller, L. P., Klock, W. & McKay, D. S. 1993, Geochim.Cosmochim. Acta, 57, 1551Veras, D., & Wyatt, M. C. 2012, MNRAS, 421, 2969Vogt, S. S., Allen, S. L., Bigelow, B. C. et al. 1994, SPIE, 2198, 362Wegner, G., & Koester, D. 1985, ApJ, 288, 746Wilson, D. J.,Gaensciek, B., Koester, D. et al. 2014, MNRAS, in pressXu, S., Jura, M., Klein, B., Koester, D., & Zuckerman, B. 2013, ApJ, 766, 132Xu, S., & Jura, M. 2014, ApJ, 792, L39Xu, S., Jura, M., Koester, D., Klein, B., & Zuckerman, B. 2014, ApJ, 783, 79
This preprint was prepared with the AAS L A TEX macros v5.2.
21 –