DOI: 10.1126/science.1243823, 1346 (2013);342 Science

et al.Bon-Chul KooPhosphorus in the Young Supernova Remnant Cassiopeia A

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): December 15, 2013 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/342/6164/1346.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/suppl/2013/12/11/342.6164.1346.DC1.html can be found at: Supporting Online Material

http://www.sciencemag.org/content/342/6164/1346.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/342/6164/1346.full.html#ref-list-1, 2 of which can be accessed free:cites 43 articlesThis article

registered trademark of AAAS. is aScience2013 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

Dec

embe

r 16

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

o

n D

ecem

ber

16, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

on

Dec

embe

r 16

, 201

3w

ww

.sci

ence

mag

.org

Dow

nloa

ded

from

o

n D

ecem

ber

16, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

Phosphorus in the Young SupernovaRemnant Cassiopeia ABon-Chul Koo,1* Yong-Hyun Lee,1 Dae-Sik Moon,2,3,4 Sung-Chul Yoon,1 John C. Raymond5

Phosphorus (31P), which is essential for life, is thought to be synthesized in massive stars anddispersed into interstellar space when these stars explode as supernovae (SNe). Here, wereport on near-infrared spectroscopic observations of the young SN remnant Cassiopeia A, whichshow that the abundance ratio of phosphorus to the major nucleosynthetic product iron (56Fe) inSN material is up to 100 times the average ratio of the Milky Way, confirming that phosphorusis produced in SNe. The observed range is compatible with predictions from SN nucleosyntheticmodels but not with the scenario in which the chemical elements in the inner SN layers arecompletely mixed by hydrodynamic instabilities during the explosion.

Phosphorus (P) is an indispensable ingredi-ent for life together with carbon, hydro-gen, nitrogen, oxygen, and sulfur (S). In

our solar system, its abundance relative to hy-drogen is 2.8 × 10−7 by number, which is 50 to1900 times less than those of the other life-keeping a elements (1). The abundance of P inthe diffuse interstellar medium and stars in ourgalaxy’s disk is comparable with that of the solarsystem or the cosmic abundance, with some de-pendence on metallicity (2, 3). P is believed tobe mainly formed in massive [≳8 solar mass (M⊙)]stars by neutron capture on silicon (Si) in hydro-static neon-burning shells in the pre-SN stage andalso in explosive carbon- and neon-burning layersduring SN explosion (4, 5).

Freshly synthesized P should thus be found inyoung core-collapse SN remnants (SNRs) result-ing from the explosion of massive stars. CassiopeiaA (Cas A) is the youngest confirmed core-collapseSNR in our galaxy; it has been extensively studiedin all wavebands (Fig. 1). It is located at a dis-tance of 3.4 kpc (6) and thought to be the rem-nant of a SN event in C.E. 1681 T 19 (7). Thespectrum of the light echo from the SN event in-dicated a SN of Type IIb that had a small hy-drogen envelope at the time of explosion (8).The total estimated mass of the SN ejecta is 2 to4 M⊙, and the inferred initial mass of the pro-genitor star ranges from 15 to 25 M⊙ (9, 10).Emission from P II was previously detected fromCas A (11), but no analysis of the emission linehas been done.

We conducted near-infrared (NIR) spectro-scopic observations of the main SN ejecta shellusing the TripleSpec spectrograph mounted onthe Palomar 5-m Hale telescope in 2008 (Fig. 1)(12). TripleSpec provides continuous wavelength

coverage from 0.94 to 2.46 mm at medium spec-tral resolving power of ∼2700. By analyzingthe spectra, we identified 63 knots of emissionand, for each knot, measured the radial veloci-ties and fluxes of emission lines, including [P II]lines (12).

The knots show distinct spectroscopic andkinematic properties depending on their origins(Fig. 2). The knots with strong [S II] lines (Fig. 2,red symbols) have high (≳100 km s−1) radialvelocities, and most of them have strong [P II]lines too, suggesting that the production of P istightly correlated with the production of S. Someknots without apparent [S II] lines have lowradial velocities (≲100 km s−1) and strong HeI lines (Fig. 2, green symbols). Their proper-ties match those of “quasi-stationary flocculi,”the material lost from the hydrogen envelopeof the progenitor during its red-supergiant phase

as slow wind (13, 14). The rest of the knots with-out [S II] lines (Fig. 2, blue symbols) have littleHe I emission, and most of them have high ra-dial velocities. They have [Fe II] lines as strongas those from the other knots. These knots areprobably pure iron (Fe) material from completeSi-burning in the innermost region of the SN,corresponding to the pure Fe ejecta detected inx-rays (9). They are found mainly in the south-ern bright [Fe II] filament, where the x-ray–emitting Fe ejecta is not prominent (Fig. 1).

The P/Fe abundance ratio of the knots canbe derived from the flux ratio of [P II] 1.189 mmand [Fe II] 1.257 mm lines (F[P II]1.189/F[Fe II]1.257).These two lines have comparable excitation en-ergies and critical densities, which together withthe fact that the ionization energies of P and Feare comparable (10.49 and 7.90 eV) greatly sim-plifies the abundance analysis (15). (This is notthe case for [S II] lines that have considerablyhigher excitation energies.) We assume that theline-emitting region is at Te = 10,000 K and hasa uniform density, with equal fractions of singlyionized ions. This simple model yields P/Fe abun-dances accurate to within ~30% mostly for bothSN and circumstellar knots (12). The observedF[P II]1.189/F[Fe II]1.257 ratios (Fig. 3) imply thatthe P-to-Fe abundance ratio by number, X(P/Fe),of the SN ejecta knots is up to 100 times higherthan the cosmic abundance X⊙(P/Fe) = 8.1 × 10−3

(1), whereas the knots of the circumstellar me-dium have ratios close to the cosmic abundance.The enhanced P abundance over Fe in these SNejecta knots is direct evidence for the in situ pro-duction of P in Cas A.

The observed range of X(P/Fe) is compatiblewith the local nucleosynthetic yield of P in SN

1Department of Physics and Astronomy, Seoul National Uni-versity, Seoul 151-747, Korea. 2Department of Astronomy andAstrophysics, University of Toronto, Toronto, Ontario M5S 3H4,Canada. 3Space Radiation Laboratory, California Institute ofTechnology, Pasadena, CA 91125, USA. 4Visiting Brain PoolScholar, Korea Astronomy and Space Science Institute, Daejeon305-348, Korea. 5Harvard-Smithsonian Center for Astrophysics,60 Garden Street, Cambridge, MA 02138, USA.

*Corresponding author. E-mail: koo@astro.snu.ac.kr

Fig. 1. Three-color com-positeimageoftheyoungSN remnant Cas A. Red isa Palomar [Fe II] 1.644-mmimage representingSNma-terial at ∼104 K, green is aChandra x-ray continuum(4.20 to 6.40 keV) imagerepresenting hot gas andrelativistic particles thatare heated by SNR blastwave, andblue is aChandraFe K (6.52 to 6.94 keV)image representing SNmaterial at ∼2 × 107 K,respectively. The slit posi-tions in our NIR spectro-scopic observations aremarked by thin white bars.The centralwhite plus sym-bol represents the positionof the SN explosion centerdetermined from propermotions (28). The scalebar in the lower left rep-resents an angular scaleof 1′, which corresponds to ~1 pc at the distance (3.4 kpc) of the SNR.

13 DECEMBER 2013 VOL 342 SCIENCE www.sciencemag.org1346

REPORTS

models. The internal chemical composition andthe resulting X(P/Fe) profile of a spherically sym-metric SN model for progenitor mass of 15 M⊙is shown in Fig. 4 (16). In the oxygen-rich layer,P is enhanced, whereas Fe is slightly depletedbecause of neutron capture. (56Fe and 58Fe arethe two major Fe isotopes in the oxygen-richlayer.) This makes X(P/Fe) higher by two ordersof magnitude than that of the outer He-rich layer,which is essentially equal to the cosmic abun-dance. The extended P enhancement inMr/Mcore =0.55 to 0.73 is the result of hydrostatic burningin pre-SN, whereas the bump at Mr/Mcore ∼ 0.5is due to explosive burning during the explosion.The knots that we observed are probably fromsomewhere in the oxygen-rich layer. It is, how-ever, difficult to precisely determine the knots’original positions in the progenitor star from thecomparison of the measured P/Fe ratio with themodel prediction because the local X(P/Fe) arevery much model-dependent (Fig. 4B) and sub-ject to further modifications owing to the Rayleigh-Taylor instability during the explosion (17). Weinstead assumed that the knots have been expand-ing freely, with an expansion rate of 2800 km s−1

pc−1 (18). We then converted the obtained spacevelocities to mass coordinates using the ejecta

Fig. 2. NIR line fluxes of the Cas A's knots as afunction of their radial velocities. NIR line fluxesare in erg per square centimeter per second. Shown areHe I 1.083 mm (3P→ 3S1), [P II] 1.189 mm (1D2→

3P2),[S II] 1.032 mm (2P3/2→

2D5/2), and [Fe II] 1.257 mm(a4D7/2→ a6D9/2) lines, from top to bottom. The knotsnot detected in the line emission are marked byopen, upside-down triangles representing 3s upperlimits. The fluxes are corrected for extinction. The dif-ferent colors represent knots of different character-istics: red symbols, knots with strong [S II] lines; greensymbols, knots without [S II] lines but with strong He Ilines at low radial velocities (<~ 100 km s−1); bluesymbols, knots without [S II] lines and with little He Iemission.

10-15

10-14

10-13

10-12

10-11

He

I 1.0

83 µ

m

10 -14

10-13

10-12

[P II

] 1.1

89 µ

m

10 -14

10-13

10-12

10-11[S

II] 1

.032

µm

100 100010-14

10-13

10-12

[Fe

II] 1

.257

µm

|v | (km s-1)

0.00 0.05 0.10 0.15 0.20 0.25 0.30[Fe II]1.677/[Fe II]1.644

0.01

0.10

1.00

10.00

[P II

]1.1

89/[F

e II]

1.25

7

100 X (P/Fe)

10 X (P/Fe)

X (P/Fe)

Fig. 3. [P II] 1.189/[Fe II] 1.257 versus density-sensitive [Fe II] line ratios of the Cas A's knots.The knots are marked in the same colors as in Fig. 2. The open, upside-down triangles represent theknots undetected in the [P II] line emission, whereas the open squares represent the knots undetectedin [Fe II] 1.677 mm line emission, including the one undetected in either line emission at (0.020, 0.17).The dotted lines show theoretical line ratios for X(P/Fe) = 1, 10, and 100 times the cosmic abundance.The solid squares along the lines represent the ratios at ne = 103, 104, 105, and 106 cm−3, from left toright. The [Fe II] line ratio in the abscissa is essentially given by a function of ne, and the observed ratiosindicate that ne = 3 × 103 to 2 × 105 cm−3.

www.sciencemag.org SCIENCE VOL 342 13 DECEMBER 2013 1347

REPORTS

velocity profile of a 15 M⊙ SN IIb model (19)scaled to the chemical structure of the model inFig. 4 (16). The resulting mass coordinates haveconsiderable uncertainties because of model-dependency but provide rough locations of theknots (Fig. 4B).

X(P/Fe) of the [P II] line–emitting ejectaknots fall into the X(P/Fe) range of the oxygen-rich layer in the model, whereas the X(P/Fe) ofthe “pure” Fe knots are often less than the cosmicabundance (Fig. 4B). The fact that the circum-stellar knots have X(P/Fe) close to the cosmicabundance—X(P/Fe) ∼ 2 X⊙(P/Fe)—gives con-fidence in the derived X(P/Fe) in the core. Theextended X(P/Fe) range over nearly two ordersof magnitude may be explained by the hydro-dynamic chemical mixing during the SN explo-sion. But, our result does not support a completemixing of the core below the He-rich layer be-

cause in such a case, X(P/Fe) will be much loweras marked by the brown arrow in Fig. 4B; theavailable yields in the literature imply X(P/Fe) =0.01 to 0.05 for a SN of 15 to 25M⊙ (4, 16, 20, 21),which is represented by the solid part of the ar-row, or 0.01 to 0.15 when the unusually high Pyield of the 20M⊙ model of (16) is included. Thedetection of P-depleted, pure Fe material probablyproduced in the innermost, complete Si-burninglayer also indicates that these dense SN ejecta ma-terials largely retain their original abundance.

The high X(P/Fe) ratio, in principle, could alsoresult from the depletion of Fe in the gas phaseif Fe atoms are locked in dust grains. In Cas A,a substantial amount (∼0.1 M⊙) of newly formeddust grains in the SN ejecta has been indeed de-tected (22–24). But, they are most likely silicatedust, with only 1% of Fe in the form of FeS (25).Also, the fact that [P II] lines in the SN ejecta

are much stronger than in the circumstellar knots(Fig. 2) is best explained by the enhanced abun-dance of P rather than by the depletion of Feonto dust.

References and Notes1. M. Asplund, N. Grevesse, A. J. Sauval, P. Scott, Annu. Rev.

Astron. Astrophys. 47, 481–522 (2009).2. V. Lebouteiller, Astron. Astrophys. 443, 509

(2005).3. E. Caffau, P. Bonifacio, R. Faraggiana, M. Steffen,

Astron. Astrophys. 532, A98 (2011).4. S. E. Woosley, T. A. Weaver, Astrophys. J. Suppl. Ser. 101,

181 (1995).5. Another type of stellar explosions (Type Ia supernovae)

are also thought to contribute, but to a much lesserextent (26, 27).

6. J. E. Reed, J. J. Hester, A. C. Fabian, P. F. Winkler,Astrophys. J. 440, 706 (1995).

7. R. A. Fesen et al., Astrophys. J. 645, 283–292(2006).

8. O. Krause et al., Science 320, 1195–1197 (2008).9. U. Hwang, J. M. Laming, Astrophys. J. 746, 130

(2012).10. P. A. Young et al., Astrophys. J. 640, 891–900

(2006).11. C. L. Gerardy, R. A. Fesen, Astron. J. 121, 2781–2791

(2001).12. Materials and methods are available as supplementary

materials on Science Online.13. S. van den Bergh, Astrophys. J. 165, 457 (1971).14. J.-J. Lee, S. Park, J. P. Hughes, P. O. Slane,

arXiv:1304.3973 (2013)15. E. Oliva et al., Astron. Astrophys. 369, L5–L8

(2001).16. T. Rauscher, A. Heger, R. D. Hoffman, S. E. Woosley,

Astrophys. J. 576, 323–348 (2002).17. K. Kifonidis, T. Plewa, H.-T. Janka, E. Müller, Astron.

Astrophys. 408, 621 (2003).18. T. DeLaney et al., Astrophys. J. 725, 2038–2058

(2010).19. S. E. Woosley, R. G. Eastman, T. A. Weaver, P. A. Pinto,

Astrophys. J. 429, 300 (1994).20. C. Kobayashi, H. Umeda, K. Nomoto, N. Tominaga, T. Ohkubo,

Astrophys. J. 653, 1145–1171 (2006).21. A. Chieffi, M. Limongi, Astrophys. J. 764, 21

(2013).22. J. Rho et al., Astrophys. J. 673, 271–282 (2008).23. B. Sibthorpe et al., Astrophys. J. 719, 1553–1564

(2010).24. M. J. Barlow et al., Astron. Astrophys. 518, L138

(2010).25. T. Nozawa et al., Astrophys. J. 713, 356–373 (2010).26. G. Cescutti, F. Matteucci, E. Caffau, P. François, Astron.

Astrophys. 540, A33 (2012).27. C. West, A. Heger, Astrophys. J. 774, 75 (2013).28. J. R. Thorstensen, R. A. Fesen, S. van den Bergh, Astron. J.

122, 297–307 (2001).

Acknowledgments: This work was supported by Basic ScienceResearch (NRF-2011-0007223) and International Cooperationin Science and Technology (NRF-2010-616-C00020) programsthrough the National Research Foundation of Korea (NRF)funded by the Ministry of Education, Science and Technology,and also by the Korean Federation of Science and TechnologySocieties (KOFST). D.-S.M. acknowledges support from the theNatural Science and Engineering Research Council of Canada.We thank M. Muno for his help in observations and J.-J. Leefor providing the Chandra 1Ms x-ray images.

Supplementary Materialswww.sciencemag.org/content/342/6164/1346/suppl/DC1Materials and MethodsFigs. S1 to S4Tables S1 and S2References (29–49)

26 July 2013; accepted 7 November 201310.1126/science.1243823

0.4 0.6 0.8 1.010 -3

10 -2

10 -1

10 0

10 1

Mr / Mcore

X (

P/F

e)

15 M

20 MB

10 -6

10 -5

10 -4

10 -3

10 -2

10 -1

10 0

10 1

Mas

s fr

actio

n4He16O56Fe

58Fe 31P

32S

A

Fig. 4. Comparison of the abundance ratio X(P/Fe) of the Cas A's knots with predictions fromSN nucleosynthetic models. (A) Internal, post-SN chemical composition of a 15 M⊙ SN model (16).The abscissa is the mass inside a radius r from the explosion center normalized by the mass of the SNcore, which is 4.163 M⊙. The mass inside the “mass cut” (1.683 M⊙ or Mr/Mcore = 0.40) becomes a stellarremnant, whereas the rest is ejected with the explosion. Most of the mass outside the core would havebeen lost as a stellar wind for SN Type IIb, such as Cas A. In a real SN, the profile will be smootherbecause of the hydrodynamic mixing of material. (B) Profile of the abundance ratio X(P/Fe) of the 15 M⊙SN (solid line). The Fe abundance includes all stable Fe isotopes. The observed X(P/Fe) ratios of the Cas ASN ejecta knots are marked in the same colors and symbols as in Fig. 3. The knots of circumstellarmedium (green symbols) are marked at Mr /Mcore > 1. For comparison, the dotted line shows X(P/Fe)profile of a 20 M⊙ SN (16) with a very high P yield because of unusual convection mixing the Si andoxygen layers before the explosion. The other typical SN models of 15 to 25 M⊙ in the literature havechemical structures similar to that of the 15 M⊙ SN model here. The brown arrow to the left representsthe expected range of X(P/Fe) for a complete mixing of the core below the He-rich layer.

13 DECEMBER 2013 VOL 342 SCIENCE www.sciencemag.org1348

REPORTS