OHP, june 2006

1
OHP, june 2006
From Symbiotic stars to Quasars Radoslav K.
Zamanov
2
in collaboration with M.F. Bode (Liverpool,
UK) P. Marziani (Padova, Italy) C.H.F. Melo
(ESO Chile) R. Bachev (Bulgaria) R.Konstantino
va-Antova (Bulgaria) J.W. Sulentic (Alabama,
USA) M. Calvani (Padova, Italy) A. Gomboc
(Ljubliana, Slovenia)
3
Symbiotic stars are interacting binaries
consisting of a red giant transferring mass onto
a white dwarf.
4

• We are investigating the projected rotational
  velocities of the mass donors (red giants). Our
  aims are
• To check theoretical prediction that the red
  giants in these binaries are co-rotating (for
  objects with known periods).
• To perform comparative analysis and to check if
  they are faster rotators (comparing with isolated
  giants and those in wide binary systems).
• To give clues for binary periods, individual mass
  loss rates, select candidates for X-ray
  observations.

5

• Observations
• 40 symbiotic stars have been observed with the
  2.2m telescope (ESO, La Silla) FEROS
  spectrograph at resolution 48000.
• Our sample
• All objects from the Symbiotic star catalogue
  with 12hltR.A.lt24h, Dec. lt 20, and brighter
  than Vlt 12.5 mag.
• From literature -12 northern symbiotic

6
To measure the projected rotational velocity (v
sin i) we used the CCF method and numerical
template. The width of the CCF is connected
(calibrated for FEROS) with the v sin i.
Examples of the CCF and the fitting gaussian.
AS 316 v sin i 9.8 1.5 km/s Rapid rotator
V417 Cen v sin i 75 7.5 km/s R Aqr only
upper limit can be given (v sin i lt 3.5 km/s)
7
The rotational period of the red giant versus the
orbit period for 16 symbiotic stars in our sample
with known orbital periods (all they are
S-type). The solid line represents
synchronization (PorbProt). Among these 16
objects there are two, which deviate
considerably from co-rotation CD-43 14304 and RS
Oph. There are doubts that v sin i of CD-43
14304 could be wrong. RS Oph seems to be the only
symbiotic system which is not synchronized.

8
Isolated giants spectral classes K2-K5 III

K2-K5 III giants in symbiotic stars
(238 objects from catalogues of v sin i) (7
objects, our measurements)
Results 238 K2III-K5III stars have vsini in
the interval from less than 1.0 km/s up to 6.7
km/s with mean vsini 1.70 km/s (median 1.50
km/s) and standartd deviation of the sample
0.90 km/s. The K giants of 8 S-type symbiotic
stars with mass donors K2III-K5III have vsini in
the interval from 4.5 up to 8.9 km/s with mean
vsini7.42 km/s (median7.15 km/s) and stddev
1.54 km/s. The Koslmogorov-Smirnov test gives a
probability of 10-6 (K-S statistics 0.60) that
both distributions are coming from the same
parent population. This means that the K-giant
mass donors of symbiotic stars rotate faster than
isolated K-giants.
9
Isolated giants spectral classes M0-M7 III
M0-M7
III giants in symbiotic stars (12 objects from
catalogues of v sin i) (28 objects, our
measurements mostly)
Isolated M giants 1.8 km/s lt v sin i lt 18.1
km/s, mean vsini 5.54 km/s (median3.10
km/s), stddev5.22 km/s. M-giants in
symbiotics 3.0 lt v sin i lt 52 km/s, mean
vsin i9.07 km/s (median7.72 km/s), stddev8.81
km/s. The Kolmogorov-Smirnov test gives a
probability of 0.0074 (K-S statistic 0.54) that
both distributions are coming from the same
parent population. This means that from
statistical point of view the M giants mass
donors of symbiotic stars rotate faster than
isolated M-giants (at confidence level 99).
10

• Reasons for faster rotation of the giants
  in symbiotic systems
• - synchronization, if the time spent by the
  mass-losing star on the giant branch is longer
  than the synchronization time. In all symbiotic
  systems with orbital period Porb 100 years
  tidal interaction overcomes the angular momentum
  loss by the wind (Soker 2002).
• - accretion during the MS phase of the
  present red giant the more massive star in the
  system, the present WD, had transferred material
  at the stage when it had been red giant.
• - backflowing material hot component
  prevents part of the mass blown by the giant from
  acquiring the escape velocity for the binary
  system. This fraction of mass may acquire angular
  momentum, and if it is accreted back by the
  giant, it spins-up its envelope.
• - angular momentum dredge-up when convective
  envelope approaches the core region of the
  giant.
• - planet engulfment during the giant phase.

11
FUTURE WORK To strengthen our results, more
data on M type isolated giants and more v sin i
measurements of K type mass donors in symbiotics
are desirable. 1. 17 more symbiotics have been
observed with the same FEROS2.2m 2. We intend
to expand our sample with northern and fainter
symbiotic stars. ELODIE at Bulgarian Observatory
Rozhen can play important role.
12
From White dwarfs to Quasars
13
The main constuituents of an Active Galactic
Nucleus Central massive Black Hole (MBH106-109
M?) Geometrically Thin Accretion Disk (d? 3 Rg
? 105Rg) Thick molecular torus (d ? 1 pc) Line
emitting gas (clouds?) (d?0.1 pc in low
luminosity AGN d? 104 Rg) Radio Jet along Disk
Axis (from Padovani Urry 1992)
Relativistic Jet
Massive Black Hole
Molecular Torus
Accretion Disk
14
Figure UV region spectral similarity between
CH Cyg and I Zw. The middle spectrum is produced
by scaling and broadening of the CH Cyg spectrum
to imitate the emission lines widths of IZw1.
CH Cyg
CH Cyg symbiotic with 1M? WD
CH Cyg (broad.)
I Zw 1
I Zw 1 narrow line Seyfert 1 galaxy, widely
used as template for all quasars. Mass of the
black hole 107 M? .
From Zamanov Marziani, 2002, ApJ 571, 77
15

• Comparison between the optical spectra in the H?
  - H? region of the interacting binaries CH Cyg,
  MWC 560 and the low redshift quasar I Zw 1. A
  clear similarity between the emission lines is
  visible. Practically every emission feature
  visible in the spectrum of IZw1 has corresponding
  emission line in the spectra of CH Cyg and MWC
  560.

16
The optical emission line spectra of CH Cyg and
MWC 560 are subtracted, broadened and scaled to
imitate I Zw 1. This standard procedure is
widely used for the emission line measurements of
AGN, using I Zw 1 itself as a template (Boroson
Green 1992, Marziani et al. 1996). After this
processing, good identity is achieved with the
spectrum of I Zw 1. Our best fit corresponds to a
width FWHM(FeII) 970?90 km s-1 (Zamanov
Marziani, 2002, ApJ 571, 77)
17
The HI and FeII lines of AGNs are coming from the
so-called broad-line region. This poorly
understood region is thought to be within ? 1 pc
from the central (supermassive) black hole. The
clear spectral similarity means that in objects
like MWC 560 and CH Cyg we are observing a
scaled down version of the famous broad line
region of quasars.
18
JETS Jet velocity 1000-1500 km s-1 in CH
Cyg (Taylor et al. 1986, Crocker et al. 2001) and
1000-6000 km s-1 in MWC 560 (Tomov et al. 1992)
Galactic microquasars (accreting stellar mass
black holes) 0.26c SS 433 (Margon
1984) 0.5c Cyg X-3 (Marti et al. 2001)
0.9c GRS 1915105 (Mirabel Rodriguez
1999) Consistent with an overall picture in
which the jet velocity is of the same order of
the escape velocity (Livio 2001) Vesc(WD)
0.02 c .
19

• JET ENERGY
• MWC 560 and CH Cyg the jets are probably result
  of the propeller action of a magnetic white dwarf
  (Mikolajewski et al. 1996) extraction of
  rotational energy from the compact object.
• Quasars the jet energy is coming from
  extraction of energy and angular momentum from a
  rotating black hole via the Blandford Znajek
  (1977) mechanism.
• Microquasars black hole - Blandford Znajek
  (1977) mechanism
• neutron star - ??? (The jets of Crab are the
  most pure case of extraction of rotational
  energy, even without accretion).
• The jets of CH Cyg and MWC 560 represent probably
  a low energy (non-relativistic) analog of the
  jets of quasars and microquasars, having a
  similar energy source the extraction of
  rotational energy from the central compact
  object.

20
Optical spectra demonstrating the spectral
similarity and the changes of the FWHM(H?). The
filled circles refer to NLSy 1 galaxies, which
are supposed to have systematically lower black
hole masses. The two triangles indicate the
symbiotic stars (CH Cyg and MWC 560). As it could
be expected they are located outside of the AGN
population but from the side of NLSy1.
from Zamanov Marziani, 2003, ASP Conf.Ser,
303, 308
21
The position of two symbiotic stars on the
diagram reinforces the interpretation of Boroson
Green Eigenvector-1 as a mainly result of L/M
ratio. The efficiency of accretion along with
some other factors could play some minor role.
The FeII-H? (Eigenvector-1) diagram. The lines
are plotted (from top to bottom) for MBH1.109
M?, MBH5.107 M?, and white dwarf mass MWD1.4
M?. The L/M ratio was running in the limits
2.5-4.6 for MBH1.109 M? 2.5-5.1 for MBH5.107
M? 3.0-3.9 for white dwarf mass MWD1.4 M?.
The ratio (L/M) is in solar units with the solar
value (L/M)?1.92 ergs s-1 g-1.
from Zamanov Marziani, 2002, ApJ 571, L77
22
The high mass X-ray binaries Can we observe a
scaled down version of the quasar broad line
region in wind-fed X-ray binaries ? In the most
cases they have additional source of ionization
a hot primary OB star, i.e the ionization
conditions are quite different from symbiotics
and AGNs. It will be extremely interesting to
detect a stellar mass black hole accreting from
the wind of red giant (although very difficult
from the evolutionary point of view). A black
hole accreting from wind of red giant will
(probably) represent good imitation of quasar !
23
OIII blue outliers among the AGNs
24
Composite Quasar Spectra from the Sloan Digital
Sky Survey (Van den Berk et al., 2001 AJ 122,
549).
However the quasar spectra are not similar!
25
Examples of subtraction of FeII complex around H?
and OIII lines. Left panels represent the
continuum subtracted spectra and best FeII fit.
Left panels represent fit to the H? broad
component. The difficulties of FeII subtraction
are coming from S/N ratio, wavelength coverage,
presence/absence of HeII4686, HeI 4471, etc.
from Marziani, Sulentic, Zamanov, et al., 2003,
ApJS, 145, 199
26
H? spectral region of the blue outliers after
the deredshift and subtraction of the FeII
template. Spectra are normalized with respect to
the normal continuum and arbitrary constant
added. Solid curves correspond to the subtraction
of IZw1-based empirical template, and dot-dashed
curves to the subtraction of a theoretical
template (Sigut Pradham 2003, ApJ 145, 15).
Vertical lines indicate the position of H?,
OIII?4959 and OIII?5007. The difference in
radial velocities between OIII lines and H? is
obvious.
from Zamanov, Marziani, Sulentic, et al., 2002,
ApJ 576, L9
27

• Forbidden OIII emission arises in the NLR of
  AGNs. This emission has now been partly resolved
  in the nearest AGN, where the geometry of the
  line-emitting gas has been found to be far from
  spherically symmetric. This suggest that
  measures of the integrated OIII emission may
  correlate with source orientation to the line of
  sight. At the same time it is generally believed
  that radial velocity measures of the narrow
  emission lines (e.g narrow component of H? and
  OIII ?4959, ?5007) provide a reliable measure
  of the systemic, or rest-frame, velocity.
• Several observations, however, indicate that the
  NLSy1 prototype I Zw1 shows an blue shift of the
  OIII lines ?V? -500 km/s relatively to other
  rest frame indicators (HI 21cm, molecular CO
  emission).
• We measured the radial velocity difference
  (?V) between the H? and OIII ?4959, ?5007
  lines in 187 objects (our sample 215 objects, 7
  with no detectable OIII emission, 16 with
  poorly defined H? peak).

28
Histogram showing the distribution of the radial
velocity difference between OIII?5007 and top
of H?. As it is visible in most of the objects
?V lt 300 km s-1. However there are some
objects, with ?V down to -1000 km s-1. The
values range from 950 to 280 km/s with average
lt?Vgt -30 km/s and sample standard deviation ?135
km/s. Typical measurement error is ?50 km/s.
from Zamanov, Marziani, Sulentic, et al., 2002,
ApJ 576, L9
29
Radial velocity difference between H? and
OIII?5007 versus the FWHM(H? BC). Vertical
dotted line marks the boundary of the NLSy1
galaxies. Vertical dashed line separates
population A and B sources. In our sample of
215 objects we detected 7 objects with ?V ? -300
km s-1.
from Zamanov, Marziani, Sulentic, et al., 2002,
ApJ 576, L9
30
Location of the outliers in the FWHM(H?BC) versus
W(FeII)/W (H?BC) diagram (the optical E1
diagram). They are not randomly distributed
(2D_KS test gives probability 0.990 0.999).
from Zamanov, Marziani, Sulentic, et al., 2002,
ApJ 576, L9
31
A sketch representing blue outlier. The OIII
lines originate from the wind, the disk is
visible face-on, and the receding part of the
wind is obscured from the disk. In calculations
we adopted cone half-opening angle 850, with the
line of sight oriented at 150, with respect to
the cone axis. The receding part of the flow is
assumed to be fully obscured by an optically
thick disk.
32
Upper panels CIV ?1549 and OIII?5007 profiles
of Ton 28. Lower panels CIV ?1549 and
OIII?5007 outflow model profiles, for optically
thin gas moving at approximately the local escape
velocity.
from Zamanov, Marziani, Sulentic, et al., 2002,
ApJ 576, L9
33
L/LEdd of blue outliers

• For our sample, we calculated the masses using
  reverberation mapping studies(Kaspi et al. 2000).
  Our sample covers
• magnitude range 20 lt MB lt27
• BH mass 7 lt log(M/M?) lt10
• a well defined strip of
• L/LEdd 0.02 - 1.00.
• The blue outliers are located between objects
  with highest L/LEdd ratio.

34
The luminosity-to-mass ratio versus the mass of
the BH. If the blue outliers are oriented
nearly pole-on the effect of orientation could
play a role. It could be as high as ?M?0.4. Even
in these case the blue outliers remain between
objects accreting at higher Eddington ratio.
(Bear in mind that a lot of other objects also
have to be moved in the same way).
35
The blue outliers among AGNs seems to represent
a special case of high L/M ratio, face-on view,
and very compact NLR. They seems to be radio
quiet analog of the core dominated radio loud
quasars. (!) Not all radio quiet AGNs visible
pole-on are blue outliers.
36
In future

• Search for blue-shifted OIII among higher
  redshift quasars (IR data).
• Searching for orientation indicators.
• Exploration of the connection between the L/M
  ratio and blue shifts.

37
THE END