8 Polarization and SED in the far IR of young stars

 

Table 4: Sample of stars with different type of SEDs in near and far IR (1, 2, 3, 4, 5 - according to Fig. 19). The last line in the table indicates the number of stars in each class and an average value of polarization $\bar{p}$

1		2		3				4		5
name	p,%	name	p,%	name	p,%	name	p,%	name	p,%	name	p,%
V 376 Cas	23.4	HD 259431	1.0	HD 98922	0.2	HD 179218	0.4	HD 17081	0.37	Vega	0.005
LkH$\alpha$198	1.9	HD 50138	0.9	HD 104237	0.25	HD 142666	0.5	HD 158643	0.46	HD 216956	0.007
Elias 1	4.7	RR Tau	0.4	V856 Sco	0.5	HD 163296	0.2	HD 41511	0.55	HD 4881	<0.1
R Mon	13	HD 150193	5.2	HD 139614	0.15	HD 190073	0.4	HD 141569	0.10
R CrA	6	MWC 297	2	HD 132947	1.3	AB Aur	0.5	HD 169142	0.3
DG Tau	5	MWC 1080	2.6	HD 34282	0.11	VV Ser	1.0	HD 52721	1.25
LkH$\alpha$233	11	TW Cha	2	HD 37357	0.52	HD 36917	0.9	$\beta$ Pic	0.02
LkH$\alpha$25	3.5	TY CrA	1.0	HD 31648	0.25	HD 35929	0.11	HD 58647	0.2
T CrA	2.9	NX Pup	0.7	HD 36112	0.25	HD 53367	0.7	HD 37490	0.45
RCW 34	5.6	HK Ori	1.8	HD 37258	0.3	LkH$\alpha$208	2.8	+46 $\hbox{$^\circ$ }$3471	0.7
V 1686 Cyg	5.4	LkH$\alpha$234	1.4	HD 37806	0.7	+40 $\hbox{$^\circ$ }$4124	1.2	+ $41\hbox{$^\circ$ }$3731	0.35
KK Oph	4	HD 250550	1.1	HD 144432	0.35	LkH$\alpha$218	1.6	TW Hya	0.15
HL Tau	11	XZ Tau	1.6	HD 97048	2.2	HD 37411	0.25
PV Cep	14	T Tau	1.2	HD 95881	1.5	VZ Cha	1
VV CrA	3.8	MWC 137	6	SU Aur	0.4	+65 $\hbox{$^\circ$ }$1637	1.2
		+61 $\hbox{$^\circ$ }$154	1.8	CS Cha	0.7	HD 100546	0.8
		LkH$\alpha$215	1.1	BF Ori	0.3	HD 76534	0.55
		V380 Ori	1.2	AK Sco	0.5	HD 200775	0.95
		AA Tau	1.5	CQ Tau	1.0	RY Lup	0.5
		CX Tau	0.65	CO Ori	1.6	WW Vul	0.5
		T Ori	1.5	BP Tau	0.3	DF Tau	0.4
		DN Tau	0.7	UZ Tau	0.75	V773 Tau	0.11
		DS Tau	0.8	V819 Tau	1.3	V826 Tau	0.95
		Haro 1-1	2.9	V827 Tau	0.6	V830 Tau	0.6
		UX Ori	2.0
15	7.7%	25	1.7%		48	0.7%		12	0.4%	3	$\leq$0.1%

According to many authors (see for example Adams et al. [1987]) young stars show a random shape of their SEDs in far IR from that with flat or rising IR spectra through two-component IR SEDs to that with decreasing IR spectra. There is no doubt at present that these changes in the shape of the SED are caused by evolution of dust CS shells. Hillenbrand et al. ([1992]) discussed the evolutionary changes of the SEDs for HAEBE stars from their group II (with flat or rising IR spectra) through group I to group III (with small IR excesses). Malfait et al. ([1998]) have studied the more evolved HAEBE stars (45 objects), that arguably are evolving towards Vega-type stars. Most of the objects in their study may be classified as the group I objects of Hillenbrand ([1992]) but 11 stars from their sample lie on the (J-H)/(H-K) diagram closer to the MS region. They discussed a possible evolutionary scenario in the changing of the SEDs from the embedded sources (group II of Hillenbrand ([1992]) through the stars with broad IR excesses and HAEBE stars with a composite excess (group I of Hillenbrand ([1992]) to Vega-type stars. A possible evolutionary sequence among young stars has been proposed recently by Yudin et al. ([1999]) as follows: "Herbig Ae/Be stars (i.e. pre-MS stars) $\Longrightarrow$Vega-like stars with near-IR excess (i.e. young stars near the end of the pre-MS evolutionary phase) $\Longrightarrow$Vega-like stars without near-IR excess but with some far-IR excesses (i.e. young MS stars) $\Longrightarrow$MS stars without near- and far-IR excesses". If the changes in dust structure of CS environment during the evolution are real, the changes in average polarization should be observed simultaneously. Using even small statistics Hillenbrand et al. ([1992]) noted that the objects in their groups I and III generally show smaller net polarizations (p<2%), than the objects in group II (p>5%). This behaviour will be investigated below (using larger statistics).

Figure 19: Different types of SEDs for young stars

Like Malfait et al. ([1998]) we separate young stars from our sample with different types of IR SEDs and plot this SEDs sequence in Fig. 19. The results of our selection are tabulated in Table 4. We have not constructed here SEDs for all stars from our sample but have used SEDs which were available in the literature (Hillenbrand et al. [1992]; Malfait et al. [1998]; Mannings [1994]; Gauvin & Strom [1992]; Grinin et al. [1991]; Miyake & Nakagama [1995]). Thus we compare the values of p and SEDs for about 100 young stars.

As can be seen immediately from Table 4 there is a definite sequence of changing in polarization values from group 1 (with flat or rising IR spectra) - $p\approx 8$% through group 2 (with decreasing IR spectra) - $p\,\approx\,1.7$%, to groups 3 and 4 (with small IR excesses) - $p\,\approx\,0.8$% and 0.4% and further to group 5 which contains the objects which are close to the MS - p<0.1%.

Note that without exception all stars in group 1 are sources of active outflow. In spite of the fact that early stages of evolution are also characterized by high value of accretion rate, the existence of active matter outflow may well counteract the accretion. During further evolution of CS shells (changes in SED from group 1 to group 4) CS disk-like envelopes being compressed along a line-of-force of magnetic field become more flattened and the accretion becomes more important mechanism for perturbation of dust in CS environment. Simple comparison of the sources from Table 4 with the lists of objects from Grady et al. ([1996]) indicates that some of the objects in our group 2 and 3 show evidence of both the matter outflow and accretion. At this stage of evolution the changes in polarization may occur for a few reasons: the decreasing of amounts of hot dust in CS disks, the compression of dust disks into very narrow structures which is unfavorable for polarization and possible increasing of optical depth around the disk's midlplane which results in multiple scattering and depolarization effects. Note also that all stars with Algol-like minima (such as those stars in the HAEBE and TT groups) have their SEDs around class 3 objects which is in good agreement with the scheme of evolution depicted above, particularly with respect to compression of CS disks. All UXOrs show clear evidence of a disk accretion (Nata et al. [1997]) as well as many other stars in groups 3 and 4.

Finally, the evolution of CS disks from group 4 to group 5 and further, leads to a strong decrease in both near and far IR excesses. However the decreasing of near IR excesses occurs more rapidly (see Malfait et al. [1998]) because no matter outflow exists at this stage. All prototypes of Vega-excess stars show evidence of accretion but no matter outflow. At this stage of evolution CS disks have very small optical depth, are very narrow and in addition some fraction of CS dust may coagulate into the large bodies. All of this provides unfavorable conditions for polarization (clearly indicated in Table 4 and as follows from all the above discussion).

We emphasize that the above discussed scheme of CS evolution is only a simple sketch and no numerical calculations have been made here. For detail theoretical discussion of this subject see for example Miyake & Nakagama ([1995]) and Myers et al. ([1998]).

We must also emphasize that the changes in polarization and IR excesses may not be caused directly by the age of a star itself and rather reflect the evolutionary changes in CS disk-like shells. The link between the age of a star and the stage of the evolution of it's CS shell certainly exists but this link can be very complicated and depends on many factors (the mass of a star, the rotational velocity, binarity etc.).