Tải bản đầy đủ - 0trang
C+O star models for SNe Ic
SKe 1998bw and 1997ef,respectively. The former is the explosion of a C+O
ergs, while the
star of MC+O 13.8 Ma (MmS 40 Ma) with 5 x
latter is that of a C+O star of MC+O 10 MQ (Mms 35 M a ) with
Figure 1 . Density profiles of models COMDH (solid line; for SN 2003dh, ),
C0138ES0 (dashed line; for SN 19!38bw, ), and C0100E18 (dottedline; for SN 1997ef,
a Values required to reproduce the peak brightness of SN 2003dh
Model COMDH was constructed for SK 2003dh. We note that in early
phases the spectra of SK 2003dh are similar to those of SK 1998bw, while
in late phases they are similar to those of SK 1997ef. Therefore, we used
C0100E18/2 for o < 15000 km s-', C0138E50 for o > 25000 km s - l ,
and merged the two models linearly in between . Kote C0100E18/2
is a model where the density of COlOOEl8 is reduced by a factor of two
In the following sections, we will show that the model COMDH can well
reproduce both the spectra of SIX 2003dh and its LC.
3. Spectrum Synthesis
We have compared the synthetic spectra of model COMDH with observations a t three epochs .
The first spectrum was obtained a t the M M T on 2003 April 10 . This
12 days after the GRB, i.e.
10 rest-frame days into the life of the
SIX, assuming that the SIX and the GRB coincided in time. The spectrum
is characterized by very broad absorption lines, and it is similar to those
of SIX 1998bw a t a comparable epoch. A good match can be obtained for
logL[erg] = 42.83 and v(ph) = 28000 km s-' (Figure 2). A t this epoch,
the photosphere is located far outside, and this SK 1998bw-likespectrum
is formed in the C0138E50-like part of COMDH.
Figure 2. T h e observed 2003 April 10
spectrum (top, black line), subtracted afterglow spectrum (observed on 2003 April
1, middle, thin gray line), 'net' SN spectrum (bottom, thick gray line), and our
synthetic spectrum (bottom, black line).
Figure 3. T h e observed 2003 April 24
spectrum (top, black line), subtracted afterglow spectrum (observed on 2003 April
1, bottom, thin gray line), 'net' SN spectrum (middle, thickgray line), and our synthetic spectrum (middle, black line).
The next spectrum was obtained a t the M M T on 2003 April 24 [l'i].
Its rest-frame epoch is 23 days. At this epoch, the photosphere falls in the
joining region. So this spectrum is a good test for the merged model. We
obtained a good match with logL[erg] = 42.79 and v(ph) = 18000 km s-'
Figure 4. The observed 2003 May 10 spectrum (black line), spectrum of SN 1997ef
obtained on 1998 January 1 (thin gray line), and our synthetic spectrum (thick gray
Our third spectrum was obtained with Subaru on 2003 May 10 1121,
a rest-frame epoch
36 days, and it resembles that of SK 1997ef a t a
comparable epoch. At this epoch, model COMDH gives a good fit using
logL[erg] = 42.43 and v(ph) = 7500 km s-' (Figure 4). It means that
the photosphere has receded deep inside, and that this SN 1997bw-like
spectrum is determined by the COlOOE18-like part of COMDH.
4. Light Curve Construction
We extracted the SIX component from the observed spectra of [l'i],[la],
. The spectrum taken on April 4 was adopted as the fiducial nonSIX afterglow (AG) spectrum (following Matheson17 et al. 2003), scaled in
flux, and then subtracted from the observations together with a host galaxy
component . We adjusted the scaling factor to make the flux between
3200 A and 3600 A in the rest frame roughly disappear from the results.
This is justified by the fact that! type Ib/c SNe, like SNe Ia, always show
flux deficiency to the blue of
3600 8, due to line-blanketing (e.g. [MI).
The AG flux deduced this way can be fitted using a power-law function of
0.08. The extrapolated AG flux was
time, f M t P a ,where a = -2.18
subtracted from the spectrum taken on May 10, which does not cover the
wavelength range mentioned above.
We derived the absolute V-band LC of SK 2003dh from the decomposed
spectra. We then transformed it to the bolometric LC by using the bolonoting their spectral similarity. The
metric corrections of SK 1998bw
bolometric light curves are ploted in the rest frame in Figure 5. Ref. 13
have estimated the SP; photometry on .June 22 from a spectrum taken with
the Subaru telescope. They assumed a negligible AG contribution, while
our fitting AG evolution predicts about 27% the total flux at that epoch.
Our LC differs from those reported by refs. 8 and 17, which are also
different from one another. It roughly follows that of SK 1998bw before
the peak, while it is fainter than SK 1998bw by about 0.6 mag after the
peak. Though the likely time of peak is not covered, it is possibly 2 - 5
days earlier than SK 1998bw and the peak brightness may be a bit fainter.
The light curve in , however, rises much more rapidly, reaches a brighter
peak, and then drops much faster, while the one in [l7] appears to resemble
that of SK 1998bw. It is likely we subtracted more A C flux than ref. 17 did
because we intended to avoid the un-physical ”blue excess” in the decomposed spectra. Refs. 8 and 17 used a ?I-square fitting method t o decompose
the spectra. We tried that, method, only t o get a LC similar to the one
We also constructed another bol LC using BVRI photometry 1141. The
LC from the spectra is similar to that from the photometry (figure 5).
5 . Light Curve Fitting
Synthetic bolonietric LCs were computed using an LTE radiation t,ransfer
code and a gray 7-ray transfer code [lo]. Generally speaking, y-rays created
in radioactive decays of 561’;i and %o can be either absorhed by SK ejecta
or escape, as determined by transfer calculations. The heated ejecta material then emits optical photons, which escape from optically thin regions.
A constant line opacity of 0.03 om3 g-’ was assumed in our computation.
The whole optical opacity was found t o be dominated by electron scattering. The balance between photoionizations and recombinations has been
taken into account.
First, we discuss the general dependance of the synthetic light curve on
various model parameters. The width of the LC peak, n c , can be roughly
i4 x I E , ~ ~ ~ / ~ A ~ ,
~ , where
~ / ~ IEE K
, ~ is
expressed as mc
opacity and A is a function of the density structure and 56Ki distribution.
The rising time of the LC is strongly affected by the 56Ki distribution. If
56Ki is absent from the outer region, -/-rays and optical photons are long
trapped in optically thick regions and hence the LC rises slow-ly. After the
peak and in later phases, the whole ejecta becomes transparent to optical
photons. So the decline rate of the LC is mainly determined by the optical
depth of the ejecta to y-rays, which roughly scales as K , ~ cc
where K? is y-ray opacity.
Lipkin el al. 2004
Figure 5. The observed bolonietric LC of SN 2003dh (filled suqares; this paper), synthetic LC of the merged model COMDH (solid line), and that of model C0138E50
We tested two models to fit the LC of SIX 2003dh, C0138E50 and
COMDH. 56Ki distribution and the total 561Xi mass were treated as free
parameters. As shown in Figure 5, both niodels can reproduce the observed LC peak. This is consistent with the above parameter analysis. The
best-fit 56Ki mass for each model is listed in Table 1. The synthetic LC
of C0138E50, however, declines faster than observations in later phases.
Model COMDH, on the other hand, has a denser core compared with
C0138E50 (see Figure 1). The denser core increases the efficiency of yray trappings and results in a LC being brighter and slower in later phases
than that of C0138E50.
We have studied the properties of SK 2003dh and constructed model
COMDH for it. We summarize in Figure 6 the model parameters of core-
collapse SSe and hypernovae for comparison . It shows that normal
core-collapse SKe occur for stars having main-sequence mass less than 20
M,, which probably explode with the canonical kinetic energy of 1x 1051
ergs and eject 0.1 M, 56Ki , [Z]
. originated from more massive
stars can be divided into two branches. One is the hypernova branch, which
features high kinetic energy and large ejected 56Ki mass . The other
is the faint SK branch, showing low kinetic energy and small ejected 56Si
mass. It is clear that S S 2003dh belongs to the hypernova branch.
Our spectra and LC fitting require a dense core for SK 2003dh. Such
a dense core is unlikely the outcome of any realistic 1-D hydrodynamical calculations. Actually, all spherically symmetric explosion models, e.g.
C0100E18 and C0138E50 in Figure 1, are very flat inside, and they all
have an inner mass-cut defining a ‘hole’ in the density profile. In 2-D jetinduced explosions, however, a dense core can be formed through accretion
along the equatorial direction and the LC declines slowly according to the
2-D hydrodynamical and light curve calculations .
1. Arnett, W.D., A p J 253 785 (1982).
2. Bloom, J.S., et al., A p J 5 0 8 L21 (1998).
3. Bloom, J.S., et al., 4 p J 572 L45 (2002).
4. Della Valle, M., et al., AtYA 406 L33 (2003).
5. Galama, T.J., et al., Nature 395 670 (1998).
6. Garnavich, P.M., et al., ApJ582 924 (2003).
7 . Hatano, K., et al., BAAS 198 3902 (2001).
8. Hjorth, J., et al., Nature 423 847 (2003).
9. Iwamoto, K., et al., ApJ437 L115 (1994).
10. Iwamoto, K., et al., Nature 395 672 (1998).
11. Iwamoto, K., et al., 4 p J 534 660 (2000).
12. Kawabata, K.S., et al., ApJ593 L19 (2003).
13. Kosugi, G., et al., P,4SJin press (2004).
14. Lipkin, Y.M., et al., ApJ submitted (2004) (astro-ph/0312594).
15. Maeda, K., et al., ApJ593 931 (2003).
16. Maeda, K., Phl) theszs, Univ. of Tokyo (2004).
17. Matheson, T., et al., A p J 599 394 (2003).
18. Mazzali, P.A., Iwamoto, K., & Nomoto, K., ApJ545 407 (2000).
19. Mazzali, P.A., et al., A p J 572 L61 (2002).
20. Mazzali, P.A., et al.,ApJ599 L9.5 (2003).
21. Nakamura, T., Mazzali, P.A., Nomoto, K.,& Iwamoto, K., ApJ 550 991
22. Nomoto, K., et al., Nature 364 507 (1993).
23. Nomoto, K., et al., Nature371 227 (1994).
Nomoto, K., Iwamoto, K., Suzuki, T., Physics Reports 256 173 (1995).
Nomoto, I<., et d.,Foyer book ostroin (2004).
Patat, F., ApJ 555 900 (2001).
Piran, T., Physics Reports 314 575 (1999) (astro-ph/9S10256).
Ricker, G.R., et al., /,4UC8101 (2003).
Shigeyama, T., & Nomoto, K., ApJ360 242 (1990).
Shigeyama, T., et al., A p J 420 341 (1994).
Stanek, K.Z., et al., A p J 591 L17 (2003) (astro-ph/O3041173).
Turatto, M., et al., 4 p J 4 9 8 L129 (1998).
Waxman, E., & Draine, B.T., A p J 537 796 (2000).
Yamazaki, R., Yonetoku, D., & Nakamura, T., ,4pJ 594 L79 (1998).
Zampieri, L., et al., MNRAS 338 711 (2003).
Main Sequence Mass, (Mo)
Figure 6. Top panel: main-sequence m a s of the progeniton for core-collapse
SNe/hypernovae V . S . kinetic energies. Bottorn panel: main-sequence m a 5 V.S. total 56 Ni
mass. The parametersof these SNe were estirnatedby previous papers (SN I1 1987A, ;
SN IIb 19935, [3O]; SN Ic 19941, ; SN I1 1997D, ; SN I1 1999br, ; SN Ic 1997ef,
; SN Ic 1998bw, ; SN Ic 1999as, ; SN Ic 2002ap, ; SN Ic 2003dk1,[ZO]).
THE “DARK SIDE” OF GAMMA-RAY BURSTS AND
IMPLICATIONS FOR NUCLEOSYNTHESIS OF LIGHT AND
Mas-Planck-Institut b r Astrophysik, Garchi.q, Germany
Dept. of Astronomy, University of Tokyo, Tokyo, Japan
Research Lab. for Nuclear Reacto,rs, Tokyo finst. of Tech., Toleyo, Japan
Center for Nuclear Study, Uwiversity of Tokyo, Wako, Saitama, Japan
Recent observational and theoretical advances in studies of gamma-ray bursts
(GRBs) indicate t h a t there can be a significant “dark side” to the GRB energy budget: besides the ultrarelativistic jet giving rise t o the bright gamma-ray emission,
there can be an even greater amount of “dark” energy contained in an accompanying, mildly relativistic outflow. T h e occurrence of such outflows are strongly
supported in theoretical models, particularly for collapsar-type GRB progenitors,
and also has been directly confirmed in observations of GRB030329/SN2003dh.
This outflow component, which should be more baryon-rich than the G R B jet,
can have interesting implications for nucleosynthesis. Inside the outflow, light elements may be produced through reactions similar t o big bang nucleosynthesis.
Heavy element synthesis by neutron capture can also take place, sometimes by a
moderately rapid “n-process” rather than an r-process. The resulting nucleosynthetic products may be observationally relevant for the most metal-poor stars, as
well as the companion stars of black hole binary systems.
Successful generation of GRBs requires the formation of a narrowlycollimated, ultrarelativistic outflow with bulk Lorentz factor r 2 100, implying very low baryon-loading (e.g.). Since the temperature at the
base of the outflow should be of order MeV and the baryons are likely to
contain a high fraction of neutron^,^^^^ some production of nuclei starting
from free protons and neutrons is expected to occur in the expanding flow.
However, as shown in several recent papers,1r10i18in the very high entropy
and extremely rapid expansion characteristic of GR.B jets, nucleosynthesis
is limited to only small production of D and 4He, which are difficult to
Nevertheless, most GRB progenitors should also give rise to associated outflows with higher baryon-loads and lower velocities (baryon-rich
outflows, or BROs), as discussed in the next section. The lower entropy,
slower expansion and higher ejecta mass of BROs compared to GRB jets
make them much more interesting from a nucleosynthesis viewpoint. As
a first study of nucleosynthesis in BROs associated with GRBs, we have
investigated this problem utilizing the simple dynamical framework of the
basic fireball model, but incorporating detailed iiuclear reaction networks
including both light and heavy elements. More details can be found in .
2. Evidence for BROs: the “Dark Side” of GRBs
Besides the ultrarelativistic jet giving rise to the conspicuous GRB emission, there is growing evidence, both theoretical and observational, for the
existence of different types of BROs which constitute a “dark” component
to the GRB energy budget. Theoretically, it is highly probable that different types of “circum-jet winds” of baryon-loaded material surround and
coexist with the narrowly collimated GR.Bjet. In general, the GRB jet production mechanism is likely to act not only on the baryon-poor zones near
the jet axis, but also on the more baryon-contaminated, peripheral regions,
leading to BROs as a natural byproduct. Models involving core collapse
and jet penetration in massive stars”,27 have recently gained strong support through observations of SN2003dh associated with GRB030329.5i13 In
such models, BROs should be induced through entrainment and mixing
with the stellar material. High-resolution numerical simulations of jet-star
~ ~ demonstrate the inevitable generation of
interaction in c ~ l l a p s a r sindeed
a peripheral, low r outflow containing a significant fraction of the total
energy. In generic black hole accretion disk models, baryon-rich winds