close

Вход

Забыли?

вход по аккаунту

Керівництво з експлуатації Руководство по эксплуатации;pdf

код для вставкиСкачать
High resolution spetra of bright entral stars of
bipolar planetary nebulae, and the question of
magneti shaping
Ting-Hui Lee1 , Letizia Stanghellini1 , Lilia Ferrario2 , and Dayal Wikramasinghe2
1
2
National Optial Astronomy Observatory, U.S.A. thleenoao.edu;
lstanghellininoao.edu
The Australian National University, Australia liliamaths.anu.edu.au;
dayalmaths.anu.edu.au
Summary. We present ESO NTT high resolution ehelle spetrosopy of the entral
stars (CSs) of eight southern bipolar planetary nebulae (PNe) seleted for their
asymmetry. Our aim was to determine or plae limits on the magneti elds of the
CSs of these nebulae, and hene to explore the role played by magneti elds in
nebular morphology and PN shaping. If magneti elds do play a role, we expet
these CSs to have elds in the range 102 − 107 G from magneti ux onservation on
the reasonable assumption that they must evolve into the high eld magneti white
dwarfs. We were able to plae an upper limit of ≈ 20, 000 G to the magneti elds of
the entral stars of He 2-64 and MyCn 18. The spetrum of He 2-64 also shows a PCygni prole in He I λ5876 and λ6678, orresponding to an expanding photosphere
with veloity ∼ 100 km s−1 . The detetion of helium absorption lines in the spetrum
of He 2-36 onrms the existene of a hot stellar omponent. We did not reah
the neessary line detetion for magneti eld analysis in the remaining objets.
Overall, our results indiate that if magneti elds are responsible for shaping bipolar
planetary nebulae, these are not required to be greater than a few tens of kilogauss.
1 Introdution
Planetary Nebulae (PNe) are found in a wide range of morphology, with the majority of shapes being round, elliptial, and bipolar. Several mehanisms have been
proposed for shaping the bipolar strutures, inluding a binary system, magneti
elds, and stellar rotation. Among them, magneti elds have been an attrative
hypothesis for shaping PNe into bipolar morphology for single stars.
Interestingly, magneti elds have been deteted in the progenitors and the
progeny of PNe, namely, around proto-planetary nebulae (pPNe), and in white
dwarfs (WDs). Sine PNe and their entral stars (CSs) are the evolutionary link
between pPNe and WDs, at least a fration of them should also ontain magneti elds. However, suh observational evidene has been sare. Very reently,
some hints of the existene of magneti elds in PNe have been given by means of
666
Lee, Stanghellini, Ferrario & Wikramasinghe
spetropolarimetry [1℄. Among the four CSs of PNe observed, they found possible
signature of magneti elds in two CSs.
Isolated magneti WDs (MWDs) with elds greater than 106 G show a mean
mass of ∼ 0.93 M⊙ , ompared to the main peak of the mass distribution of ∼ 0.57
M⊙ for the nonmagneti white dwarfs [2℄. This is interesting, sine the statistial
studies of PNe have shown that bipolar PNe may have more massive progenitors. The
high mass distribution of isolated MWDs, together with more massive progenitors
of bipolar PNe, seem to indiate that isolated MWDs ould be the evolutionary
produt of magneti CSs. If this hypothesis is orret, then, under the ondition of
magneti ux onservation, we should expet a CS to MWD eld ratio of
(BCS /BMWD ) ∼ (RMWD /RCS )2 ,
(1)
where B is the magneti strength and R is the radius of the star. Depending on the
mass of the CS and how evolved it is, RMWD /RCS ould have a range of ∼ 0.01−0.1.
Therefore with a range of magneti elds of ∼ 106 −109 G found in MWDs, we should
expet CS magneti elds in the range of ∼ 102 − 107 G. Thus our aim is to searh
for magneti elds in CSs of bipolar PNe in this range in order to investigated if
CSs of bipolar PNe an be progenitors of MWDs.
We have obtained high resolution ehelle spetrosopy of entral stars (CSs) of
eight southern bipolar planetary nebulae, to look for Zeeman splitting of the stellar
lines aused by possible magneti elds assoiated with the CSs. We seleted bipolar
nebulae whose entral stars are bright and well separated from the nebulae, and most
of them do not have diretly evidene of being a binary system. The spetra were
aquired with the ESO Multi-Mode Instrument (EMMI) on the 3.58 m NTT. Table
1 lists the properties of the entral stars and the observing log.
Table 1.
Objet
PN G
HDW 5
He 2-25
He 2-36
He 2-64
He 2-123
He 2-186
MyCn 18
NGC 2818
218.9-10.7
275.2-03.7
279.6-03.1
291.7+03.7
323.9+02.4
336.3-05.6
307.5-04.9
261.9+08.5
Target List
V (CS) Tz
Ob. Date Exposures (se)
(×103 K)
16.29
Feb 4, 2003 4 × 1800
16.96
< 61
Feb 4, 2003 4 × 1800
11.37
Feb 4, 2003 2 × 900
48.3
Feb 5, 2003 4 × 1800
16.84
< 60
Feb 5, 2003 1 × 1200 + 3 × 1800
16.62
95.5
Feb 5, 2003 2 × 1800
14.5 (B) 51.6
Feb 4, 2003 4 × 900
19.58
215.0
Feb 5, 2003 2 × 1800
2 Results
Overall the spetra of seven objets, exept HDW 5, show many nebular features.
He 2-25, NGC 2818, and He 2-186 also show high expansion veloities in the hydrogen
emission lines. Here we fous our attention to the stellar spetra. We found no
Central Stars of Bipolar PNe
667
evidene of stellar emission or absorption in three objets, He 2-25, He2-186, and
NGC 2818. HDW 5 has one stellar absorption line deteted, and He 2-123 has two
stellar C IV emission line deteted. Only in He 2-36, He 2-64, and MyCn 18 we
deteted several stellar lines. We show the images of these three objets in Figure
1. For full desription of our results, please see [3℄.
Fig. 1.
R-band images of He 2-36, He 2-64, and MyCn 18
3 Disussion
If a Zeeman split of magneti origin is deteted in the spetrum, the eld strength
ould be inferred with:
∆λ = 4.67 × 10−13 geff λ20 B
(2)
where ∆λ is the split separation, λ0 is the wavelength of the spetral line, geff is the
eetive Landé fator and B is the magneti eld strength in Gauss [4℄. Equation
(2) an be also used to estimate upper limits to the magneti eld.
Among the 8 CSs observed, only the spetra of He 2-36, He 2-64, and MyCn 18
have high enough S/N to determine whether their stellar lines might be Zeeman
split. No evidene of Zeeman splitting was found in the spetra of He 2-64 and
MyCn 18. We ould plae an upper limit to the magneti eld assoiated with these
two CSs using Equation (2). Due to the nature of our data faint entral stars thus
low S/N for stellar ontinuum (∼ 10 to 20 for He 2-64 and MyCn 18), we believe
that we are unable to detet a split feature with a separation of less than 0.4 Å. This
translates to a eld strength of ∼ 20,000 G. Therefore no magneti eld stronger
than 20,000 G was found in the entral stars of He 2-64 and MyCn 18. The spetrum
of He 2-64 also shows P-Cygni proles of the He I lines, most learly shown at λ5785
and λ6678 (Figure 2, left). The veloity range of the wind is of the order ∼ 100 km/s
from the emission maximum. This nding indiates that the star has an extensive
expanding atmosphere.
The spetrum of He 2-36 is more diult to interpret. We know that this star has
two omponents, the ool ompanion being an A2 III star, and the hot ompanion
deteted to date only in the ultraviolet spetra [5℄. The prominent features of our
He 2-36 spetrum are those metal lines expeted in a A2 III stellar spetrum. We
have suessfully tted the Balmer lines in our spetrum by using a model with Teff
= 8500 K and log g = 2. There is no need to invoke a hot omponent to justify
668
Lee, Stanghellini, Ferrario & Wikramasinghe
these lines. On the other hand, when omparing He 2-36 with the spetrum of
HD 210111, we were also able to identify two absorption lines at λ4686 and λ5876,
where no metal lines typial of an A2 III star are expeted. We believe that these
two lines orrespond to He II λ4686 and He I λ5876, and that we have onrmed the
presene of the hot ompanion of He2-36 in the optial spetrum with the present
observations.
The spetrum of He 2-36 also shows a feature at λ4780 (Figure 2, right). Sine
the absorption line looks split in the middle, we foused our attention to it, in order
to determine its nature. First, we heked against the possibility of nebular emission
ontamination and onluded that this feature is of stellar origin. Seond, we need to
establish whether the ool or the hot ompanion is responsible for the split feature
λ4780. If absorbed by the ool omponent, this feature should be interpreted as Ti II
λ4779.99. In this ase, though, this would be the only split line among the several
Ti II absorption observed. If, on the other hand, this transition originates in the hot
stellar photosphere, it ould be identied as O IV λ4779.10. But even this possibility
is puzzling, sine in the present spetrum we ould not detet the other omponents
orresponding to the 2s2p(3P0)3p - 2s2p(3P0)3d transition of O IV. If the feature
was indeed O IV, and the observed splitting was produed by the magneti eld,
we ould use equation (2) to estimate the eld strength. This O IV transition has
geff =4/3, thus the observed 0.7 Å split ould imply a eld strength of the order of
∼ 25,000 G if we assume that the line splits into a doublet in the weak eld regime.
If we use the sequene of spetra of hot sub-dwarfs published by [6℄, our dedued
relative strengths of He II λ4686 and He I λ5876 would appear to indiate a star
with log g ∼ 4.5 − 5.0 and Teff ∼ 50, 000 − 70, 000 K. Suh a star would evolve into
a magneti white dwarf with a surfae dipole eld strength of ∼ 8 × 107 − 2.5 × 108
G if we adopt a log g ∼ 8.5 as being typial of magneti white dwarfs. This eld is
omfortably within the range observed for the eld distribution of isolated magneti
white dwarfs.
We realize that our proposal that we are seeing absorption lines from the hot star
in He 2-36 superposed on the A star spetrum is not without diulty, sine suh
absorption features would require a omparable ontribution by the two stars in the
optial. A possible resolution to this dilemma is that the A star is not physially
linked to the nebula and the hot omponent that we have found is indeed the entral
ionizing star of the PN. Whether or not we have measured a magneti eld in He2-36
hinges on the identiation of the split absorption feature at 4780 whih we have
tentatively attributed to O IV from the hot star. This remains to be onrmed
through the detetion of other omponents of the multiplet with their assoiated
Zeeman splitting. Higher resolution data should also reveal Zeeman splitting in the
helium lines.
Central Stars of Bipolar PNe
669
Left: Spetrum of He 2-64: P-Cygni proles. The upper panel shows He
I λ5876 Å, the lower panel shows He I λ6678 Å. The expansion veloity inferred
from these proles is in the order of ∼ 100 km s−1 . The veloity sale at the top
of eah panel is set to 0 at the peak of the emission. Right: Spetrum of He 2-36.
λ4780 absorption line showing splitting into two omponents. The bottom panel is
the smoothed spetrum, whih shows the splitting more learly.
Fig. 2.
Referenes
1. S. Jordan, K. Werner, & S.J. O'Toole: A&A, 432, 273 (2005)
2. D.T. Wikramasinghe & L. Ferrario: PASP, 112, 873 (2000)
3. T.-H. Lee, L. Stanghellini, L. Ferrario, & D. Wikramasinghe: AJ, 133, 987
(2007)
4. F. Leone, W.D. Vaa, & M.J. Stift: A&A, 409, 1055 (2003)
5. W.A. Feibelman, & Y. Kondo: ApJS 136, 735 (2001)
6. S.J. O'Toole, S. Jordan, S. Friedrih, & U. Heber: A&A 437, 227 (2005)
1/--страниц
Пожаловаться на содержимое документа