Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.arcetri.astro.it/~lt/preprints/pianetone/ms.ps.gz Дата изменения: Tue Sep 11 16:56:55 2007 Дата индексирования: Sat Dec 22 07:21:55 2007 Кодировка: Поисковые слова: photosphere

A Young Very Low-Mass Object surrounded by warm dust
L. Testi 1 , A. Natta 1 , E. Oliva 1;2 , F. D'Antona 3 , F. Comeron 4 , C. Baa 1 , G. Comoretto 1
and S. Gennari 1
ABSTRACT
We present a complete low-resolution (R100) near-infrared spectrum of the
substellar object GY 11, member of the -Ophiuchi young association. The ob-
ject is remarkable because of its low estimated mass and age and because it is
associated with a mid-infrared source, an indication of a surrounding dusty disk.
Based on the comparison of our spectrum with similar spectra of eld M-dwarfs
and atmospheric models, we obtain revised estimates of the spectral type, eec-
tive temperature and luminosity of the central object. These parameters are used
to place the object on a Hertzprung-Russell diagram and to compare with the
prediction of pre-main sequence evolutionary models. Our analysis suggests that
the central object has a very low mass, probably below the deuterium burning
limit and in the range 8{12 M Jupiter , and a young age, less than 1 Myr. The
infrared excess is shown to be consistent with the emission of a ared, irradiated
disk similar to those found in more massive brown dwarf and TTauri systems.
This result suggests that substellar objects, even the so-called isolated planetary
mass objects, found in young stellar associations are produced in a similar fashion
as stars, by core contraction and gravitational collapse.
Subject headings: Stars: low-mass, brown dwarfs { Stars: fundamental parame-
ters { Stars: atmospheres { Infrared: stars
1. Introduction
The discovery of Brown Dwarfs (BDs) and objects with masses comparable to those
of giant planets, well below the deuterium burning limit (M<13 M J ), \free- oating" in
1 Osservatorio Astrosico di Arcetri, INAF, Largo E. Fermi 5, I-50125 Firenze, Italy
2 Telescopio Nazionale Galileo and Centro Galileo Galilei, INAF, P.O. Box 565, E-38700, Santa Cruz de
La Palma, Spain
3 Osservatorio Astronomico di Roma, INAF, via Frascati 33, I-00044 Roma, Italy
4 European Southern Observatory, Karl-Schwarzschild-Strasse 2, D-85748 Garching bei Munchen, Ger-
many

{ 2 {
young stellar clusters (Zapatero-Osorio et al. 2000; Lucas & Roche 2000) has opened an
interesting debate on their origin. Do they form like ordinary stars from the collapse of
molecular cores (Shu et al. 1987)? If so, the very existence of very low-mass objects and
their mass function put strong constraints on star formation theories. Alternatively, BDs
and isolated planetary mass objects may be stellar \embryos" ejected from multiple forming
systems before reaching a stellar mass (Reipurth & Clarke 2001), or they may form like
planets by coagulation of dust particles and subsequent gas accretion within circumstellar
protoplanetary disks (Lin et al. 1998). Young (proto)-planets could then be extracted or
ejected by dynamical interactions from the forming planetary systems. In this scenario,
isolated BDs and planetary mass objects are intrinsically dierent from stars; their study
sheds no light on star formation theories, but provides instead a chance of studying the early
evolution of giant planets where they can be observed in isolation, rather than very close to
a much brighter star.
One way to shed light on the origin of very low mass objects is to ascertain their
association with circumstellar disks, which are characteritics of stellar formation from the
contraction of a molecular core. Deep images in the L band Muench et al. (2001) and
in the mid-infrared (Persi et al. 2000; Bontemps et al. 2001) have shown that many very
low luminosity objects have excess emission at these wavelengths, and detailed studies of
some of them have proven that the central objects are bona-de BDs. Their infrared excess
is consistent with the presence of a surrounding disk similar to those found around more
massive pre-main sequence stars (Comeron et al. 1998, 2000; Natta & Testi 2001). These
initial ndings, albeit limited, seem to suggest that indeed BDs form like ordinary stars.
We report here the rst results of a project aimed on one hand to improve our under-
standing of disks around BD, following the approach of in Natta & Testi (2001), and on
the other hand to nd evidence of a circumstellar disk around bona de isolated planetary
mass objects. We have used the Near-Infrared Camera and Spectrograph (NICS) on the
3.56m Telescopio Nazionale Galileo to acquire low-resolution (NIR) spectra of a sample of
mid-infrared sources located within the -Ophiuchi cloud Bontemps et al. (2001). Our tar-
get list included objects detected at 6.7 and 14.3 m with the ISOCAM camera on board
of ISO (Kessler et al. 1996; Cesarsky et al. 1996), having low eective temperature (T eff ),
luminosity (L  ) and extinction (AV ) based on photometric or limited 2.2 m spectroscopic
estimates. The goal of our new observations was to obtain improved determinations of these
parameters and to derive accurate values of masses and ages of the targets by comparison
with theoretical evolutionary tracks. The results for the complete sample will be discussed
elsewhere (Natta et al. 2002). In this Letter, we report on one of the sources, number 33 of
the ISOCAM list of Bontemps et al. (2001), which is associated with the NIR source GY11
(Greene & Young 1992). The substellar nature of this object was already proposed by Rieke

{ 3 {
& Rieke (1990) and conrmed by Comeron et al. (1998); Wilking et al. (1999). We derive a
new, accurate spectral classication of GY11, and of its eective temperature and luminosity.
Our data suggest that GY11 is a planetary mass object, with an infrared excess that can be
roughly modeled as due to a circumstellar disk similar to those associated to T Tauri stars.
These observations provide the rst evidence that objects of such small mass actually form
in a star-like manner, and thus that they are genetically dierent from \planets".
2. Observations
A near-infrared low-resolution spectrum of GY11 was obtained with the Telescopio
Nazionale Galileo on La Palma on July 9, 2001, using a 0: 00 5 slit and the high-throughput
low-resolution prism-based disperser unique to NICS (Baa et al. 2001), the Amici device
(Oliva 2000); this setup oers a complete NIR spectrum, 0.85 to 2.35 m, at R100 accross
the entire range, and it allows an accurate classication of faint and cool dwarfs (Testi et al.
2001). Instrumental and telluric correction was ensured by observations of A0 stars. The
shape of the nal spectrum was checked using near infrared photometric measurements from
the 2MASS second incremental data release; synthetic magnitudes were computed using the
appropriate transmission curves and compared with the source photometry, colors were found
to be consistent with those of 2MASS to within 10%, as expected from typical uncertainties.
To better constrain the values of the extinction, we also obtained optical i-band (0.77 m)
photometry at the ESO-La Silla 1.54m Danish Telescope using the DFOSC instrument.
Photometric calibration was ensured by observations of a set of Landolt (1992) standard
stars, converted into the Gunn system using the transformations given by Fukugita et al.
(1996).
3. Central source parameters
Bontemps et al. (2001) associated the ISOCAM source 33 with a Class II object member
of the -Ophiuchi young stellar cluster known as GY11 (Greene & Young 1992). The brown
dwarf nature of GY11 has been suspected for some time (Rieke & Rieke 1990). However,
there is a large uncertainty in the literature as to the exact value of its photospheric pa-
rameters and mass. Bontemps et al. (2001) estimate bolometric luminosity and extinction
to be L   0:001 L , AV =2.7 mag from NIR photometry. Based on multiband infrared
photometry and a 2.2m low resolution spectrum, Wilking et al. (1999) estimate a spectral
type M6.5, AV 5 mag, L  0.002 L , and T eff 2650 K. These authors noted that due to
veiling caused by dust emission, the spectral type could easily be some 2{3 subclasses later,

{ 4 {
and the extinction signicantly underestimated. In fact, Comeron et al. (1998) derive a
higher value of AV=10 mag from broadband photometric measurements that include optical
and infrared bands.
Our complete near infrared spectrum oers the possibility of a better estimate of the
source parameters, as it allow us to use the global spectrum shape below 2 m, a region
which is less aected by the continuum veiling due to the dust emission. Given the expecta-
tion that the surface gravity of very young BDs shouls be similar to sub-giants, we derive the
photospheric parameters by comparison with eld dwarfs spectra and model atmospheres
with appropriate surface gravity, as suggested by the evolutionary models (Comeron et al.
2000). We rst derive extinction and spectral type by matching the observed GY11 spec-
trum with that of eld dwarfs in the solar neighborhood (Testi et al. 2002), obtained with
the same instrumental set-up and reddened using the Cardelli et al. (1989) extinction law
most appropriate for Ophiucus (R V =4.2; Fig. 1a). We try to provide the best t to the
global shape of the spectrum, with particular attention to the shape of the H-band, the
J-band features and the drop due to water vapor absorption at the red edge of the J-band.
Overall, the best spectral match is found with the eld dwarf with spectral type M8.5 and
extinction AV 7.0 mag. Lower values of AV (by  1 mag) oer a better match of the
spectrum with later dwarf spectra (M9{L0), but are not consistent with the broad band
optical measurements (see inset in Fig. 4). A higher value of the extinction causes a too
steep rise of the spectrum below 1.2 m. Field dwarfs with spectral types earlier than M7.5
show large deviations from the observed shape of the H-band and the drop at 1.3 m. Given
the uncertainties of a classication based on objects with very dierent surface gravity, we
expect our classication to be accurate within one spectral class and the visual extinction
estimate within one magnitude.
As a second step, in order to derive an estimate of the photospheric eective temperature
(Fig. 1a), we compare the observed GY11 spectrum, with reddened, appropriate surface
gravity, log 10 (g)=3.5, model atmospheres (Allard et al. 2000; Chabrier et al. 2000). The
best estimate of the eective temperature is 2400100 K. Higher temperature models oer
a better match of the H-band shape, but underestimate the drop near 1.3 m and the global
shape of the spectrum at J-band. The derivation of the eective temperature of young dwarfs
based on theoretical synthesis of the near infrared spectrum is very uncertain (Lodieu et al.
2002); however, our value of T eff
for an object of spectral type M8.5 is consistent with
the spectral type vs. eective temperature scale discussed by Wilking et al. (1999), and
only marginally higher than the latest eective temperature scales derived for eld dwarfs
(Leggett et al. 2001).
To estimate the luminosity (L  ) of the object, we used the 2MASS J-band magnitude,

{ 5 {
dereddened by AV  7:0 mag, and the bolometric correction derived from the best tting
(2400 K) atmospheric model. The value of this \theoretical" bolometric correction is nearly
identical to the empirical value adopted by Wilking et al. (1999). We derive a value of
L  0.008 L30%.
Using the L  and T eff values derived above, we can compare the position of GY11 in the
Hertzprung-Russell diagram with the predictions of theoretical pre-main sequence evolution-
ary models. In Figure 2 we show this comparison for the latest release of evolutionary tracks
from three leading groups in the theory of pre-main sequence evolution of substellar objects.
In spite of the relatively large uncertainties on L  and T eff , and on the limited accuracy of
pre-main sequence evolutionary tracks at these ages and masses, we conrm that GY11 is
a young ( < 1 Myr), very low mass object, probably below or very close to the deuterium
burning limit, with a best mass estimate in the range 7 to 12 M Jupiter .
4. Infrared Excess and Disk Models
GY11 is the lowest mass object with a clearly detected infrared excess. It is detected by
ISOCAM in the two broad-band lters LW2 and LW3 ( eff 6.7 and 14.3 m, respectively)
used by (Bontemps et al. 2001) in their  Oph survey, as well as in the three intermediate-
band lters, SW1, LW1, LW4 ( eff =3.6, 4.5 and 6.0 m, respectively), used by Comeron et
al. (1998) in their pointed observations of optically identied candidate brown dwarfs. The
ISOCAM measurements in the broad- and narrow-bands are in good agreement, within the
ux calibration uncertainties, which we assume to be  20%.
As usually with ISOCAM, the 6 00 beam includes multiple sources seen in higher reso-
lution near-infrared images, which may contribute to the observed uxes. Figure 3 compares
a K s image of the region around GY11 extracted from the VLT/ESO archive (originally
observed as part of ESO proposal 63.I-0691) and the ISOCAM LW1 ( eff =4.5 m) image
from the ISO archive (Comeron et al. 1998). The mid-infrared emission peaks very close to
the position of GY11. Although a small contamination from the NIR source 8 00 to the east
is possible, we think that most of the mid-infrared observed ux comes from GY11; a similar
conclusion was also reached by Comeron et al. (1998).
In Figure 4, we show the spectral energy distribution (SED) of GY 11 at all wavelengths
and compare it to that of an irradiated, ared disk similar to those that reproduce the
observed characteristics of TTauri systems (Chiang & Goldreich 1997). The disk has a dust
mass of 1 Earth mass (3% of the mass of the central object, for an assumed gas-to-dust
mass ratio of 100) and is heated by a a central source with the GY11 temperature, luminosity

{ 6 {
and mass. We show the predicted SED when the disk extends inward to the stellar surface
(solid line) and when it has an inner hole of 3 stellar radii (dotted line). More details on the
disk models can be found in Natta & Testi (2001) and Natta et al. (2002). The agreement
between observed and predicted uxes is rather good, especially for the disk with the large
inner hole. In particular, both models predict total (star+disk) uxes that in the J, H, K
bands are smaller than the calibrated TNG uxes by 15% at most.
As an independent check, we computed optical and NIR broad-band magnitudes from
the model-predicted SED. They are compared in the inset of Figure 4 with the observed
dereddened magnitudes in i (this paper), R,I,L 0 (Comeron et al. 1998), J,H,K (2MASS). The
agreement is again quite good.
The ISOCAM measurements, especially that at 14.3 m, have large error bars, and one
should not overinterpret them. However, is of some interest to point out that, if indeed the
mid-infrared excess is due to disk emission, the disk must be ared. Therefore, dust and gas
must be well mixed, as in the majority of pre-main{sequence stars, and no major settling
of the dust onto the disk midplane has occurred during the lifetime of GY11. The disk
must be optically thick to mid-infrared radiation; this however sets only a lower limit to the
disk mass of roughly 10 5 -10 6 M depending on the exact value of the dust mid-infrared
opacity and surface density prole. Note that the disk mass has to be very small; if we
assume the ratio of the disk mass to the mass of the central object typical of TTS ( 0.03),
then the disk mass is about 3  10 4 M , and the disk contains only 1 Earth mass of dust.
As a consequence, the accretion rate (if any) is also likely to be low, with an average value
over the lifetime of the object that cannot exceed 310 10 M yr 1 (for an age of 1 Myr).
The accretion luminosity is also small, about 40 times lower than the luminosity of the
photosphere. A direct determination of the disk mass can be derived from (sub)millimeter
wavelength observations. We predict for GY11 a 1.3mm ux of about 0.6 mJy, which is
well below the upper limit set by the survey of Motte et al. (1998), but within the expected
capabilities of the ALMA millimeter array.
5. Conclusions
The results presented in this letter show evidence that a young isolated planetary mass
objects in -Oph, with mass of about 10 M J , is surrounded by warm dust, possibly distributed
on a disk similar in properties to those around young brown dwarfs and T Tauri stars.
The implications of this nding, that should be conrmed by higher spatial resolution mid-
infrared observations, and should be extended to a large sample of similar objects, are
profound, since it gives a clear indication that isolated BDs and even planetary mass objects

{ 7 {
form like stars and are not produced in a planet-like fashion within protoplanetary disks
around more massive objects, and later ejected by dynamical interactions. Isolated BDs and
planetary mass objects are thus an extension of the stellar and substellar sequence to very
low masses and have dierent origins from \planets".
This work is partly based on observations collected at the Italian Telescopio Nazionale
Galileo, Canary Islands, Spain, at the European Southern Observatory telescopes on La
Silla and Paranal observatories, Chile, and on data obtained by the European Space Agency
Infrared Space Observatory. This publication makes use of data products from the Two
Micron All Sky Survey, which is a joint project of the University of Massachusetts and the
Infrared Processing and Analysis Center/California Institute of Technology, funded by the
National Aeronautics and Space Administration and the National Science Foundation. This
work was partly supported by ASI grant ARS 1/R/27/00 to the Osservatorio di Arcetri.
REFERENCES
Allard F., Hauschildt P.H., Schweitzer A. 2000, ApJ, 539, 366
Baa C., et al. 2001, A&A, 378, 722
Bontemps S., et al. 2001, A&A, 372, 173
Burrows A., Marley M., Hubbard W.B., Lunine J.I., Guillot T., Saumon D., Freedman R.,
Sudarsky D., Sharp C. 1997, ApJ, 491, 856
Cardelli J.A., Clayton G.C., Mathis J.S. 1989, ApJ, 345, 245
Cesarsky C., et al. 1996, A&A, 315, L32
Chabrier C., Barae I., Allard F., Hautschildt P. 2000, ApJ, 542, 464
Chiang E.I. & Goldreich P. 1997, ApJ, 490, 368
Comeron F., Rieke G.H., Burrows A., Rieke M.J. 1993, ApJ, 416, 185
Comeron F., Rieke G.H., Claes P., Torra J., Laureijs R.J. 1998, A&A, 335, 522
Comeron F., Neuhauser R., Kaas A.A. 2000, A&A, 335, 522
D'Antona F. & Mazzitelli I. 1997, Mem. Soc. Astron. Italiana, 68, 807

{ 8 {
Fukugita M., Ichikawa T., Gunn J.E., Doi M., Shimasaku K., Schneider D.P. 1996, AJ, 111,
1748
Greene T.P. & Young E.T. 1992, ApJ, 395, 516
Kessler M.F., et al. 1996, A&A, 315, L27
Landolt A.U. 1992, AJ, 104, 340
Leggett S.K., Allard F., Geballe T.R., Hauschildt P.H., Schweitzer A. 2001, ApJ, 558, 908
Lin D.N.C., Laughlin G., Bodenheimer P., Rozyczka M. 1998, Science, 281, 2025
Lodieu N., Caux, E., Monin, J.-L., Klotz, N., 2002, A&A, in press
Lucas P.W. & Roche P.F. 2000, MNRAS, 314, 858
Marcy G.W., Cochran W.D., Mayor M. 2000, in Protostars & Planets IV, V. Mannings,
A.P. Boss & S.S. Russell eds., (Tucson: University of Arizona Press), p. 1285
Muench A.A., Alves J., Lada C.J., Lada E.A. 2001, ApJ, 558, L51
Nakajima T., Oppenheimer B.R., Kulkarni S.R., Golimowski D.A., Matthews K., Durrance
S.T. 1995, Nature, 378, 463
Motte F., Andre Ph., Neri R. 1998, A&A, 336, 150
Natta A. & Testi L. 2001, A&A, 376, L22
Natta A. et al. 2002, A&A, submitted
Oliva E. 2000, Mem. Soc. Astron. Italiana, Vol. 71, p. 861
Persi P., et al. 2000, A&A, 357, 219
Rebolo R., Zapatero-Osorio M.R., Martin E.L. 1995, Nature, 377, 129
Reipurth B. & Clarke C.J. 2001, AJ, 122, 432
Rieke G.H. & Rieke M.J. 1990, ApJ, 362, L21
Shu F.H., Adams F.C., Lizano S. 1987, ARA&A, 25, 23
Testi L., et al. 2001, ApJ, 552, L147
Testi L., et al. 2002, in preparation

{ 9 {
Wilking B.A., Greene T.P., Meyer M.R. 1999, AJ, 117, 469
Williams D.M., Comeron F., Rieke G.H., Rieke M.J. 1995, ApJ, 454, 144
Zapatero-Osorio M.R., Bejar V.J.S., Mart  in E.L., Rebolo R., Barrado y Navascues D., Bailer-
Jones C.A.L., Mundt R. 2000, Science, 290, 103
This preprint was prepared with the AAS L A T E X macros v5.0.

{ 10 {
Fig. 1.| The spectrum of GY11 (thick solid line) is compared with reddened spectra of
eld M-dwarfs (dotted lines) as labelled (from Testi et al. 2002), all spectra are normalized
at the mean ux in the 1.6-1.7 m range and shifted with constant osets for clarity. The
spectrum of GY11 is reproduced at every oset to ease the comparison. b) similar to a), but
the dotted spectra are reddened theoretical atmospheric models (Allard et al. 2000), T eff
as labelled. In both panels, the observed spectra have a lower signal to noise in the region
corresponding to the strong telluric absroption (1.35{1.45 m and 1.82{1.95 m).

{ 11 {
Fig. 2.| Hertzprung-Russell diagram for three sets of evolutionary tracks (D'Antona &
Mazzitelli 1997; Chabrier et al. 2000; Burrows et al. 1997). The position of GY11 is shown
as a blue circle. The tracks are labelled with the appropriate mass, hydrogen burning
stars are shown in red, deuterium burning brown dwarfs in green, and objects below the
deuterium burning limit in cyan. Isochrones are shown as dotted lines and are labelled with
the appropriate ages.

{ 12 {
Fig. 3.| The ISOCAM-LW1 (contours, 4.5m) image of the region surrounding GY11 is
overlaid on the VLT K s (greyscale, 2.2m) image. The ISOCAM image has been aligned
with the VLT image by matching the position of GY10 (2320.8-1708, Comeron et al. (1998))
with the associated mid-infrared source.

{ 13 {
Fig. 4.| Disk and photosphere model of the GY11 SED. The red circles with errorbars show
the mid-infrared uxes from ISOCAM (Comeron et al. 1998; Bontemps et al. 2001), the black
line is our NIR dereddened Amici spectrum. The green jagged line is the photospheric model
for T eff
=2400 K and Log(g)=3.5 (Allard et al. 2000). The blue lines show the combined
photosphere plus disk emission computed as in the text. In one case (solid line) the disk
inner radius is equal to R ? , in the other (dotted line) to 3R ? . In the inset, we show the
comparison between the dereddened broad band photometry (from R to L 0 ; red squares) and
the models (practically coincident), photosphere as a dashed line, photosphere plus disk as
solid line.