VLBI observation of OH masers in G 45.07+0.13

VLBI observation of OH masers in G 45.07+0.13

Chin. Astron. Astrophys. Vol. 21, No. 2, pp. 182-190, 1997 A translation of Acta Astron. Sin. Vol. 37, No. 4, pp. 387-395. 1996 0 1997 Elsevier Scienc...

620KB Sizes 0 Downloads 49 Views

Chin. Astron. Astrophys. Vol. 21, No. 2, pp. 182-190, 1997 A translation of Acta Astron. Sin. Vol. 37, No. 4, pp. 387-395. 1996 0 1997 Elsevier Science B.V. All rights reserved Printed in Great Britain PII: SO2751062(97)00025-S 0275-1062/97 $32.00 + 0.00

VLBI

observation of OH masers in G 45.07+0.13t ZHENG Xing-wu

Department

of Astronomy,

Nanjing

University,

Nanjing

210008

Abstract

Both the left-and right-handed circularly polarized radiations of the masers in the ultra-compact hydrogen II region G45.07+0.13, due to the transition from the OH ground state 211~,2,J = 3/2, F = 1 --t 1, have been observed with the US VLBI array. using the multiplepoint fringe rate method, we have constructed the map of the maser structure with a relative position accuracy of 20mas. With the fringe rate of the calibration source and that of the maser component with a velocity of 56.7 km s-l, taken to be the phase reference point, the absolute position of the phase reference point has been calculated as given by (Y (1950.0) = 19hllm00.4G”, b(1950.0) = 10’45’ 43.2”. We have found that most maser components are located in the front of the comet-shaped HI1 region and that their distance to the center of hydrogen II region is approximately 0.4”. This agrees well with the distance from the vertex of the parabolic ionization wavefront to the young celestial body, as estimated with the bow shock model. By use of the tube-shaped model and under the assumption of fully saturated radiation, the number density of the hydrogen molecules in maser regions has been derived to be 4.5 x lo8 cm- 3. Such high density regions are very possibly the thick clouds of neutral molecules swept up by shock waves and are located in the front side of ionization wavefront. In this maser active region a pair of Zeemansplit lines have been found and from the difference of their radial velocities a magnetic field of 3mG has been inferred. Its direction is away from the Earth. Similar to the observation of G34.3&0.2, the diameter of the maser cluster near the compact hydrogen II region has the order of magnitude of 10” cm. This is 50 times larger than the thickness of a neutral gas layer possibly swept up by shock waves. It is still not clear how this can be explained. Key words:

Hydrogen

II region -

maser -

magnetic

field

1. INTRODUCTION

High resolution observations of ultra-compact H-II regions have shown many of them to have a comet-like structure, composed of a dense “head” of ionized hydrogen and a thin “tail”l’1. t A National Received

Scaling-the-Heights 1995-04-10;

revised

Program w&on

1996-04-05

182

OH Maser in G45.07+0.13

Theoretical

astrophysicists

believe

such a structure

183

may have two generating

1) The bow-shock model. A young object moves supersonically cloud, the fast winds it emits interact with the ambient neutral

mechanisms.

in its parent molecular molecular medium and

form the structurel?]. The ionized “head” is the shock region driven by the wind. In front of the ionized shock is a compressed layer of neutral gas, whose sufficiently large neutral molecular column density provides the physical flow” model. It was observed that the majority

condition for masing. 2) The “champagne of young stars are formed on the edge of

molecular clouds, where there are large outward gradients of density and pressure. When strong winds are generated on the young stars, those flowing outward puncture the outer confines of the cloud and ionized gas flows out as from an opened champagne bottle, while inwardly, the wind is compressed due to obstruction by dense molecular gas, forming the comet head of H-II. In the forward direction of the head there is again a compressed layer of neut,ral gas131. The debate between these two models has become one of the most interesting topics in the physical study of the formation of large mass stars. The structure and velocity field of the OH masers in the compressed of the maser active region, obtained from important evidence as to which of the two G 45.07+0.13 is an ultra-compact H-II

layer of neutral gas and the physical properties high resolution VLBI observations, can provide models is correct. region 141. In its NE, four dense condensations of

ionized hydrogen are arranged along an arc, with apparently an empty hole swept clean at the centre. In the SW direction is a “gap” with a very weak continuum emission, rather than a tenuous “tail”. Two research teams have independently observed and studied OH maser emission around this H-II region using VLA and MNRLN15v61. They have given the spatial distribution and velocity structure of the maser sources and the physical properties of the maser active region. For the OH ground state, F = 1 + 1 and F = 2 + 2, 40 maser sources were discovered, grouped into four clusters. Because of errors in the measured absolute positions, they did not give the relation between the masers and the H-II region. Using the method of VLBI multiple point fringe mapping this paper presents the distribution of the maser sources of the region and gives an accurate determination of the absolute position of the maser source with velocity 56.76 km/s, used as the phase reference point. We found the masers to be located near the “comet’s head” of the H-II region. From accurate VLBI positions the structure of the distribution of the masers and their sizes are derived in this paper. Under the assumption of a tube-like geometry, the number density of the active region is found. These observed physical data provide constraints on the setting up of theoretical models.

2. OBSERVATION

AND

DATA

TREATMENT

Using the US VLBI network, I observed the OH maser emission in the young star formation region G 45.07+0.13 at laboratory rest frequency 1665.4018 MHz for the ground state transition *II3/2, J = 3/2, F = 1 -+ 1. The antenna array included the VLA, NRAO, New Mexico, the 43m antenna at Green Bank, NRAO, West Virginia and the 25m antenna of HRAS, Texas. For the VLA an array composed of 21 antennas was chosen with an equivalent aperture of 115 m. All three systems use hydrogen maser clocks as frequency standards. Their parameters are given in Table 1.

184

ZHENG Xing-wu

Table 1 Antenna station

lhquency

Standard

Parameters

System Temperature

(K)

Sensitivity (jk-‘)

Aperture (m)

VLA

Hydrogen Clock

60

0.67

HRAS

Hydrogen Clock

100

10

26

NRA0

Hydrogen Clock

55

36

43

115

With a view to studying the magnetic field surrounding the protostars, both rightcircular polarization (LCP) were obhanded circular polarization (RCP) and left-handed served. The obsERVING BANDWIDTH WAS 125 KhZ, CORRESPONDING TO A Doppler width of 22.5 km/s, which included all the maser emission of G 45.07+0.13. We chose the quasars 3C84, 3C 73, 3C 345 and NRA0 530 as phase and passband calibrators. Because these sources have rather weak signals, their observing bandwidth was set at 2 MHz. The receiver recording system is the MK III magnetic tape system, working at recording mode B and C in order to have the spectra of both polarizations. The recordiug tape was then treated at the Haystack Observatory, using 96 lag channels in the cross and auto-correlations, corresponding to a frequency resolution of 0.17 km/s. The average integration time was 10 seconds. Post-correlation treatment was completed at the US Center for Astrophysics, using a software package jointly compiled by the Center and NRAO. A brief description now follows. (1) The quasar autocorrelation function is Fourier transformed, giving the chamrel amplitude response curve, which is then used to calibrate the channel response in the observatious. (2) For the maser emission of the G 45.07+0.13 source, the autocorrelation function is timeaveraged and Fourier transformed, giving its power spectrum for the given antenna. The one for the NRA0 43 m antenna near transit was taken as the reference spectrum for calibrating the variation in the gain of the antennas. (3) The interference fringes and fringe rates for the continuum observation of the quasars provide the information for calibrating the clocks. (4) Corrections are made for antenna tracking. (5) The maser source with LCP velocity 56.76 km/s was selected as the phase reference point, and its signal was subtracted from the signals of the other sources at both polarizations. Then, using the multiple point fringe rate n~etllod(7~81, fringe rate maps were obtained having a relative positional accuracy of 0.02 arcsec and a 3a sensitivity of 1.0 Jy. A least square reduction of the observed VLA-NRA0 and VLA-HRAS fringe rates then gave the absolute position of the reference maser source to be a( 1050) = 19”11’“00.4G” f O.lO”, 6( 1050) = lO”45’ 43.0” f 0.2”.

3. OBSERVATIONAL 3.1

The Cross

Correlation

RESULTS

AND

ANALYSIS

Spectra

Fig. gives the cross correlation spectrum for G45.07+0.13, observed by the VLANRA0 baseline. The largest peak falls at RCP velocity 56.76 km/s. with intensity 62.0 Jy. The spectrum is composed of four peaks at about 54, 57, GOand 62 km/s, with a total width of 8 km/s centered about 58 km/s. The LCP peak is at 54 km/s with an intensity 3 times smaller than the RCP peak. If we consider both polarizations, then the average velocity is 57.2 km/s and the velocity dispersion is 9 km/s. Each subspectrum has a linewidth of

OH Maser

Part of the velocity

0.3 km/s.

From

the observation

about

2 km/s.

Virial

The remainder

Theorem,

cr = 5 km/s, 30 M,

then

The velocity

mass

the central

dispersion

is

observing

frequency

spectrum

cannot

body,

is probably

field.

a split of

by motions

(such

for according

R =

to the

5x 1016 cm and

which would contradict

the mass of

for an 0 6.6 type star[‘l.

caused by interaction

between

medium.

agrees with that of Baart

distributions.

causing

be caused

line observations

masers

of the magnetic

and if we take

star and the ambient

and intensity

splitting

to be 3mG,

to the central

5u2R/G,

recombination

among individual

Our cross correlation velocity

bound

M =

mass would be 650 Ma,

from the central

as regards

the field is found

7km/s in the dispersion

from the hydrogen

coming

185

is caused by Zeeman

pair,

gravitationally

the central

deduced

shocks

dispersion

of a Zeeman

as free fall and rotation)

in G45.07+0.13

et a1.L61using MERLIN

Some differences

are probably

data

caused by our

being 4 times higher than theirs.

-2 z,x

40-

‘Z 5 30-

Q !! s cr,

20

:: :: .* A ; ::' . 3: : '; :.

10 JL

0

0

.... 1,

1.

1

46

50

52

1

f

54

56

1

““““I

56

60

s

-0.5 62

64

!

,

66

,

00

Cross-power

spectra

of OH masers

Fig. 2

in

circularly

G45.07-tO.13

3.2

Number

Density

Considering OH molecules number

of Hydrogen

density

can be expressed

of hydrogen

molecules

0

right-handed

16G5 MHz radiations

of

in G45.07+0.13

Molecules

that in the environment state

-I -1

,

-08

-0.6

Map of left-and polarized

OH masers

are in the ground

,

-04

relative right ascension (arcsec)

LSR velocity (km/s) Fig. 1

(

-0.2

of protostellar

and that

matter,

the great

majority

also we haveLgl n(OH)/n(Hz)

in a tube-shaped

masing region, assuming

-

lo-‘,

of the the

saturation,

as n( Hz) = 3.2 x 108q-‘D2AvL-3S

where q is the pumping

rate in units of lo- 4,

D

maser line width in km/s, L is the characteristic

cmm3

is the distance length

(I)

in kpc, Av is the individual

of the masing

tube in 1015 cm, and

ZHENG Xing-wu

186

of a single maser

S is the flux density

lo3 and L = 5~10’~

q =

in Jy.

Au = 0.3 km/s and S =

the measured

We take,

62 Jy,

cm, and we obtain

for G 45.07+0.13,

and for the masing a number

density

D = 9.7kpcl”l,

region,

typical

of hydrogen

values

molecules

of

4.5 x lo8 cmS3. In giant molecular is 103-10’.

Inside

clouds such as Orion,

dense cores of clouds,

OMC 1, W 33, the average

typical

value is lo’-106.

higher density local regions may be formed by interaction shocks from young stars and the ambient

protostellar

between

medium.

molecular

It is possible molecular

A detailed

density that

outflows, discussion

still

winds, on this

point will be given below. Structure

3.3

Table

of the

Maser

2 lists the physical

point fringe rate method. Fig. 2. arcsec

We see that across

cluster

appears

to consist Table

LSR velocity

majority

size 0.02~~

as for usual maser clusters

of the individual

The listed relative

the great

(linear

Sources

parameters

positions

of the sources

or 5.8~10~~cm).

(10 16-10’7cm)1111. of two sub-clusters

This

sources

found from the multiple

are used in constructing

the map of

are concentrated

an area 0.4

within

is of the sarne order of magnitude

In the rnap given by Baart of size 0.1 arcsec

separated

and Cohen,

by 0.25 arcsec.

Flux Density and Relative Position of Maser Components

2

Relative

Flux

RA

EITOF

Relative

Dee

(arcsec)

(arcsec)

EWX

(l=d)

CJY)

(arcsec)

54.3OR'

14.6

-0.08

0.01

0.01

0.01

544lR

9.1

-0.02

0.02

0.04

0.02

56.06R

5.9

-0.12

0.01

-0.04

0.01

56.76R

61.9

0.00

0.01

0.00

0.01

57.llR

6.7

0.07

0.02

0.02

0.03

57.46R

3.5

-0.05

0.01

-0.32

0.02

-0.14

0.03

(arcsec)

6o.IOR

8.5

-0.24

0.01

60.46R

4.9

-0.23

0.02

0.08

0.02

61.86R

17 ~

-0.17

0.02

0.14

0.01

62.39R

3.6

-0.11

0.03

I.48

0.02

53.c4L

9.1

-0.12

0.01

-0.13

0.02

53.59L

5.2

-0.07

0.01

0.10

0.02

53.94L

20.0

-0.11

0.02

-0.11

0.01

54.&m

4.9

-0.02

0.01

-0.09

0.03

61.69L

2.9

-0.84

0.03

-0.27

0.02

There

are also two weaker sources,

Our observed we observed

structure

the

is almost

3 fewer RCP

sources,

one to the SE of the concentration,

the same as that observed and 4 fewer LCP sources

by Baart

one to its N.

and Cohen.

However,

than they did. These

7 sources

OH Maser in G45.07+0.13

187

are all weak ones with flux densities close to 1 Jy or less, so the difference is probably due

to our antenna 3.4

array having a smaller

The Maser

Sources

The absolute

positioning

on the radio contour resolution

VLA

respectively.

effective

in relation

of our reference Filled

close to the compact

The

eastern

of G45.07+0.13

and

the two weak components

“champagne

region

source enables

us to superpose

circles

represent

can be seen

LCP

our maser map

distance

as required

sources

in the eastern

part

of # 0.4”(6 x 1016 cm.

to have a bow-shaped

all the OH masers

hydrogen,

and RCP

is located

core at a projected

excepted,

of ionized

Region

source of G 45.07+0.13

of the H-II region

of the compressed

than their MERLIN.

See Fig. 3, where the H-II map is from the high

and unfilled

We see that the reference part

to the H-II

map of the H-II region.

observations.

aperture

H-II

are located

by both

the

concentration on the outside

“bow shock”

and

flow” models.

19"11"0.045'

0.042'

0.038"

R.A. (1950.0) Fig. 3 OH masers superposed on the radio continuum map of G45.07+0.13 The position in the figure. a position, masers.

it is difficult

The relative

similar 3.5

results

A Zeeman

Zeeman

pair

velocities

1~10’~

to establish

configuration

corresponds

any intrinsic

relation

between

by the large cross

of 0.8 arcsec.

For such

the H-II region and the

to be more reasonable

and NGC 6443 F114y151and the theoretical

in the light of expectations.

Pair to the criterion

in the

proposed

G45.07+0.13

of 56.76km/s

cm) and a velocity

3/2, F + 1 +

by Ho et al.B31 is marked

H-II region, at a separation

we found appears

found in G 34.3+0.2

According LSR

of the OH maser determined

It is to the N of the compact

1 transition,

by Reid

region.

(RCP) difference

The

and Silversteinl’61,

two component

and 54.82 km/s (LCP), of -

2 km/s.

a separation

For the OH ground

the Zeeman split is 1.69 mG/(km/s),

to a longitudinal

other maser active regionsl’71,

field of 3mG.

This

we can identify

masers

of the

one have

of 0.07arcsec state

‘I13,s,

hence the velocity

value is similar

pair

(J

=

difference

to the values found

and agrees closely with the value 3.1 mG found by Baart

for and

Cohen. From the relative frequency field of G45.07+0.13

shift between

is away from the Earth.

the two polarizations This agrees

we infer the longitudinal

with the direction

of the Milky

188

ZHENG Xing-wu

Way fieldI161 and suggests and the medium

a close relation

between

the medium

in the star-forming

region

at large.

4. THE

THEORETICAL

MODELS

The accurate VLBI determination of the absolute positions of the maSer components provides an important means for understanding certain physical relations between the maser forming region, the H-II region, the infrared aourcea and the young objects. In G 45.07+0.13 the majority of the OH maser components are located near the “head” of the comet-like structure of the H-II region. While incapable of definitely confirming one model and rejecti11g the other, our observations do provide certain constraints on any model. 4.1

The Bow Shock Model In the bow shock model, the distance star is 11’1 l=5.5

of the apex of the parabolic

x lOi

IX

Cm,

shock from tl1e young

(2)

where M, is the mass loss rate of the young star in units of 10d6 Ma/yr, VW is the wind velocity in 103km/s, p(H) is the 111ea11mass of the hydrogen particle in units of gram, n(H) is the 11umber density of hydrogen particles in lo5 cmm3, and V* is the velocity of the young object inside the cloud in units of 10 km/s. The observations of hydrogen recombination line showed G 45.07+0.13 to be an 0 6.5 type starl’l, implying M, = 1, VW = 3, n(h) = 3 and p(H)= 1.4. If we now assume that the V, is of the Same order of mag11itude as tl1e velocity dispersion insider the cloud (4-10 km/a), then the theoretical formula (2) gives a star-shock front distance of 4.6x 1016 cm. This is very close to our observed distance betwee the 1naser cluster and the core of the H-II region (6x 1016 cm). In the same model, the number density of hydrogen molecules is[‘91 n (H,) = I.4 x lO’n,V,Y;’ where n, is the number

density

in the hydrogen

cm ,

(3)

core in front of the shock in u11its of lo5 cmm3,

V, is the shock velocity in 100 km/s, and V_ is the Alfvk11 velocity in km/s. The observed z’A in molecular cloud&sol is about 2 km/s. If we assume Vs = 100 km/s and n, = lo7 cn~-~[~], then (3) gives a hydrogen density of 7 x 10’ cm -3. This is very close to our observed value of 4.5x 10s cm-3. According to the shock model, masers occur in the shock-compressed neutral layer. This requires the size of the maser cluster to be less than the thickness of the compressed layer. If L is the coherence length required for 1nasing, then the maximum projected size of the maser cluster predicted by the model is

= lS+L’

D m

___ 3L2

(5,

(4)

wl1ere S is the thickness of the neutral, compressed layer. For the shock model, we have S = (9/8)M*1, M being tl1e Mach number. Taking M = 10, we find 6 z 7x 10’” cm. If the

OH Maser in G45.07+0.13

compressed

layer of neutral

then the maximum value is smaller composed

molecules

theoretical

than our observed

of two clusters,

thin neutral

Our observed

each of diameter

and so poses a difficulty

4.2

“Champagne

High resolution

observation

maybe

5. We have used the US VLBI point fringe rate method a 20 mas accuracy

the relevant

and observed relative

positions

1 transition.

of

of the star

Using the multiple

of the maser components

of the reference

observations

a magnetic

component.

with After

of this region we have come to

field of about

the observed

3mG,

field.

are located

region.

pointing

is possible

intensity through

From

away from

For a tube-like

maser

near

from the compact

pair in the maser

of the galactic

saturation

components

of 0.4arcsec

model

requires interaction

the

“head”

of the

core. their

velocity

the Earth,

and under

a hydrogen between

difference

in agreement

the condition

number young

density

stars

of of

and the

gas.

3) Similar

to the case of G 34.3+0.2,

of the order of 1016 cm. This provides

the maser cluster

an observational

of G 45.07+0.13

constraint

Wood

D. 0.

S., Churchwell E., ApJS,

[21

Van Buren D..

[31

Yorke H. W.,

[41

Turner B. E., Mathews

MecLow

M. M., Wood

Tenorio-Tagle

1989, 69, 831 D. 0. S. et al., ApJ, 1990, 353, 570

G., bodenheimer

P., A&A,

1983, 127, 313

H. E., ApJ, 1984, 277, 164

[51

Gamy

Is1

Baart E. E., Cohen R. J., MNRAS,

[71

Walker R. C., ApJ, 1981,

[a1

Walker R. C., Mats&is

G., Reid M. J., Moran J. M., ApJ, 1985, 289, 681 1985, 213, 641

86, 1323

D. N., Garcia-Bameto

has a diameter

on theoretical

References

ill

gradient

data are not available.

the OH maser emission

position

continuum

of the maser

4.5 x 10’ cmm3, which

ambient

from the observa-

helpful in this connection.

the absolute

at a distance

one Zeeman

with the direction about

it appears

there should be a density

At present

we have obtained

majority

H-II region,

2) We found

complete

this model,

CONCLUSIONS

network

and determined

great

we derived

may

conclusions:

1) The comet-like

is of the same order of magnitude

at the 2116,z, J = 312, F = 1 +

with the high resolution

the following

why in a very

size can be produced.

Th is shows that such a maser structure

near the H-II region.

forming region G 45.07+0.13

explain

to be

Model

that in order to confirm

HCO+

the cluster

for the bow shock model.

Flow”

x lo3 cm3/pc

comparing

we still cannot

this model has not the difficulty just mentioned,

tions of G 34.3+0.2 (0.5-1.0)

0.1 arcsec,

of a much greater

in G 34.3+0.2[‘*].

is 2x 1015 cm. This

of 30. Even if we regard

size of the maser cluster in G 45.07+0.13

be general

Although

of the same order as the ionized gasI”l,

for the size of the maser cluster

size by a factor

gas layer maser clusters

(1 x 1016 cm) as was observed

The

has a thickness

estimate

189

J. A., ApJ, 1982, 255, 128

models.

190

[91

ZHENG Xing-wu

Elizur M., In: Astronomical CAA

Masers, Kluwer Academic

1994,18,443

= CJSS 1994,14,

Publishers,

1992, 232

PO1

Zheug Xing-wu,

PI

Matthews

WI

Moran J. M., Burke B. F., Barrett A. H. et al., ApJ, 1968, 152, L97

1131

Ho P. T. P., Haschick A. D., Vogel S. N. et al., ApJ, 1983, 265, 295

H. E., Goss W. M., Winberg

P41

Zheng X. W.,

1151

Zheng Xing-wu,

177

A. et al., A&A,

1977, 61, 261

Moran J. M., Reid M. J., ApJ, 1996, in preparation CJSS 1987, 9, 87

PI

Reid M. J., Silverstein E. M., ApJ, 1990, 361, 483

[I71

Reid M. J., Moran M. J., ARA&A,

PI

Churchwell E., In: C. J. Late, N. D. Kylaf%, eds., The Physics of Star Formation Stellar Evolution,

D91

Hollenback Virginia,

Kluwer Academic

D., Elizur M., McKee

USA

1981, 19, 231

(Springer Verlog)

Publishers,

Do&e&t,

C. F., In: Astrophysical

and Early

1990, 221 Masers,

Proceedings,

Arlington,