A Curved Magnetic field in the ring-like shell of bubble N4

TL;DR

This study detects a curved magnetic field in the shell of bubble N4, showing magnetic field strengthening due to compression, which may influence star formation processes in the region.

Contribution

It provides the first measurement of a curved magnetic field in the shell of bubble N4 using near-infrared polarization data.

Findings

Magnetic field strength in the shell is about 120 μG.

The shell's magnetic field is parallel and curved along the ring.

All submillimeter clumps are magnetically subcritical.

Abstract

We report the detection of a curved magnetic field in the ring-like shell of the bubble N4, derived from near-infrared polarization of reddened diskless stars located behind this bubble. The magnetic field in the shell is curved and parallel to the ring-like shell, and its strength is estimated to be $\sim120\,\mu$G in the plane of the sky. The magnetic field strength in the shell is significantly enhanced compared to the local field strength. We calculate the mass-to-flux ratio for the submillimeter clumps in the shell and find that they are all magnetically subcritical. Our results demonstrate that the magnetic field strengthens as the interstellar medium is compressed into a shell, and suggest that the magnetic field has the potential to hinder star formation triggered by \ion{H}{2} region expansion.

Tables

Table 1. Table 1: H 𝐻 H -band Polarimetric Data of Background Stars Behind the Bubble N4

R.A.	Decl.	$P_H$	$\delta P_H$	${\theta }_H$	$\delta {\theta }_H$
(J2000)	(J2000)	%	%	$°$	$°$
272.020601	-18.232673	2.1	0.1	174	1
272.021080	-18.344659	4.5	0.1	69	1
272.023085	-18.243796	4.6	0.6	7	4
272.024903	-18.321352	2.0	0.4	7	6
272.025196	-18.277893	4.6	0.5	172	3
272.025758	-18.262770	5.0	0.2	169	1
272.027667	-18.264612	5.3	0.5	173	3
272.028186	-18.262282	4.8	0.3	168	2
272.028691	-18.295514	3.5	0.2	1	1

Table 2. Table 2: Physical properties of the dust condensations on the shell of N4

ID	R.A.	Decl.	FWHM	Integrated $870\mu$m Flux	Mass	Mass-to-flux Ratio
	(J2000)	(J2000)	(pc)	(Jy)	($M_{☉}$)
G011.8937+0.7780	18 8 46.5	-18 15 24	0.45	1.23	47.6	$\lesssim 0.2$
G011.9116+0.7767	18 8 49.0	-18 14 30	0.41	1.18	45.6	$\lesssim 0.3$
G011.9191+0.7536	18 8 55.0	-18 14 47	0.37	2.42	93.6	$\lesssim 0.7$
G011.9145+0.7384	18 8 57.8	-18 15 28	0.37	0.91	35.2	$\lesssim 0.3$
G011.9261+0.7442	18 8 58.0	-18 14 41	0.37	0.89	34.4	$\lesssim 0.3$
G011.9039+0.7194	18 9 0.7	-18 16 35	0.53	1.98	76.6	$\lesssim 0.3$

Full text

ISM: magnetic fields — ISM: bubbles — polarization — ISM: individual objects (N4)

††journal: ApJ

1 Introduction

Parsec-scale bubbles are created as stellar winds and/or H II region ionization fronts from massive stars expand into the interstellar medium (ISM; Churchwell et al., 2006). The expansion of a bubble sweeps up ISM into a shell of enhanced density, which could result in fragmentation into compact star-forming cores, the so-called “collect and collapse” model of triggered star formation (Elmegreen, 1998). Deharveng et al. (2010) used multi-wavelength public data to resolve the spatial distributions of cold dust condensations in the borders of 65 bubbles and found that 40% of them are good candidates for the “collect and collapse” process. Nevertheless, these good candidates need follow-up studies to determine their physical conditions and to verify if star formation is triggered by the “collect and collapse” process. Beaumont & Williams (2010) mapped seven bubbles with dense gas tracers (CO $J=3-2$ and HCO+ $J=4-3$ lines), and derived column densities which are not high enough for initiating the “collect and collapse” process of induced star formation. They proposed that bubbles expanding into flattened molecular clouds might not be able to collect sufficient gas into shells and would be unable to trigger the formation of new stars. However, the role of the magnetic field is not considered in any of the above studies.

The effects of magnetic fields in the expansion of bubbles have been studied in magnetohydrodynamic simulations. These show that the magnetic field gets stronger in the shell as the ISM is compressed and the field greatly reduces the strength of shocks and the density contrast in directions perpendicular to the magnetic field. The strong, ordered magnetic field in the shell hinders the formation of thin-shell instabilities and reduces the efficiency of star formation triggered by massive stars (Krumholz et al., 2007; van Marle et al., 2015). Another result of magnetic fields influencing the expansion of bubbles is that the elongation direction of an elliptical bubble is parallel to the magnetic field direction. Although several aspects, such as a non-uniform ISM and the inclination angles can affect the projected elongation of a bubble in a realistic ISM, this phenomenon can be investigated observationally. Pavel & Clemens (2012) found that the superthermal H II region–driven bubbles are preferentially aligned with the average orientation of the Galactic magnetic field in the disk, and subthermal H II region–driven bubbles are consistent with random alignments. This observational result implies that magnetic fields play important roles in the early evolution of H II region–driven bubbles. However, a detailed observational study of the influence of magnetic fields on bubbles has not been performed to date.

In this paper we report measurements of the magnetic field for the bubble N4 and the associated star-forming clump N4W. The bubble N4 is a ring-like shell structure enclosing the H II region G11.898+0.747 identified by Lockman (1989). The molecular gas of N4 shows a systematic velocity $V_{\mathrm{LSR}}\sim 25$ km/s (Li et al., 2013), which suggests a most probable distance of $2.80\pm 0.30$ kpc, according to the parallax-based distance estimator (Reid et al., 2016). Several papers have sought evidence for triggered star formation in the bubble N4 (e.g. Deharveng et al., 2010; Watson et al., 2010; Liu et al., 2016; Yan et al., 2016), and all concluded there is no such formation in N4. At the same $V_{\mathrm{LSR}}$ of the bubble N4, N4W is found to harbor several intermediate-mass Class I/II objects and one cold dense molecular core (Chen et al., 2016, hereinafter, Paper I). The current paper makes use of data taken using SIRPOL, mounted on the Infrared Survey Facility 1.4 m telescope at the South African Astronomical Observatory, during the nights of 2013 July 7–9. The instrument performance, observational information, weather conditions, and data reduction process were previously described in Paper I.

2 Polarimetry of point sources

The average full width at half maximum (FWHM) of point sources in the $J$-, $H$-, and $K_{s}$-bands is $2\farcs 0$, $1\farcs 9$, and $1\farcs 8$, respectively. We apply a fixed aperture radius of $1\farcs 35$ (3 pixels) to conduct aperture photometry on the reduced $JHK_{s}$ images at each wave-plate angle. To secure precise photometry, we only consider data from objects fulfilling two conditions: 1) the object’s FWHM is between $1\farcs 4$ and $2\farcs 7$, not largely different from the mean FWHM, and 2) the object is not blended with another at the spatial resolution of the SIRPOL data. The integrated detector counts of these point sources are used to calculate the Stokes parameters $I$, $U$, $Q$ (Equation (1) of Paper I), which together provide the degree of polarization $P$ and the polarization position angle $\theta$ according to Equation (2) of Paper I. The position angle $\theta$ is measured anticlockwise from north to east. The error of $P$ can be obtained by $\delta P=\frac{\sqrt{2}}{2}~{}\delta I/I$, where $\delta I$ is the photometric error. The error of $\theta$ is $\delta\theta=28\fdg 6~{}\delta P/P$; thus larger $P/\delta P$ corresponds to smaller $\delta\theta$. In the following analysis, we only consider point sources satisfying $P\geqslant 3\,\delta P$.

3 Classifying reddened background stars

The SIRPOL $JHK_{s}$ images are too shallow to penetrate through the dense regions in in N4 and N4W. In order to circumvent this problem, we retrieve much deeper $JHK$ photometric data from the DR6 release of the UKIDSS/GPS survey111http://www.ukidss.org/surveys/gps/gps.html. We cross-match the DR6 release of UKIDSS/GPS and the $H$-band polarization data. Applying a maximum separation of $0\farcs 5$ in cross-match, we derive a sample of 538 point sources with both UKIDSS/GPS $JHK_{s}$ photometry and $H$-band polarization. The distribution of these 538 point sources in the $H-K$ versus $J-H$ diagram is shown in Fig. 1. There are 441 point sources located between the two parallel reddening vectors (dotted lines in Fig. 1). These are diskless stars whose extinctions are due to interstellar dust along the line of sight (LOS). Therefore, the polarization of these diskless stars is mostly dominated by interstellar polarization originating from the dichroic extinction of nonspherical dust grains with short axes aligned along the interstellar magnetic field (see the review by Andersson et al., 2015). Thus, the polarization of these diskless stars is expected to be parallel to the interstellar magnetic field. Indeed, the observed polarizations of these stars exhibit the mean direction of the interstellar magnetic field averaged along the LOS. In order to trace the magnetic field in the bubble N4, it is necessary to identify stars whose polarization is caused by dichroic extinction by dust grains in the the bubble.

The obvious gap at around $H-K=0.6$ mag in Fig. 1 naturally splits the 441 diskless stars into low-extinction and high-extinction groups. The extinction of diskless stars is estimated by applying the conversion 222In the conversion from the $H-K$ color to $A_{V}$, a transformation from the UKIRT photometric system to the 2MASS photometric system is applied (Carpenter, 2001). The conversion factor of 14.7 is derived from the recent near-infrared extinction law $A_{\lambda}\propto\lambda^{-1.95}$ (Wang & Jiang, 2014) and a reasonable assumption $A_{J}=0.29$ when $R_{V}=3.1$. equation $A_{V}=14.7~{}([H-K]_{obs}-0.2)$ to their $H-K$ colors. The low-extinction group have less dust extinction and show a smaller extinction range ($A_{V}<6$ mag), indicating they are foreground objects. The high-extinction group show a much larger extinction range ($7<A_{V}<50$ mag), most likely contributed by the ISM in the bubble N4. We regard the reddened diskless stars ($H-K>0.7$ mag) as the background stars. The $H$-band polarization data of these 356 background stars are listed in Table 1.

Although the foreground objects have lower extinctions, a number of them exhibit $H$-band polarization at high significance levels ($S/N>3$). Fig. 2 presents the distributions of the polarization position angles and degrees of polarization for these foreground stars. Most of these foreground stars show a degree of polarization $P_{H}<2\%$, with two exceptions that show extremely large values ($P_{H}>5\%$). On the other hand, the position angle distribution of these foreground stars is roughly flat. A one-sided Kolmogorov–Smirnov test shows that the probability that the foreground $\theta_{H}$ distribution was drawn from a random distribution is 0.14, much higher than the typical significance level of 5%. The result of the Kolmogorov–Smirnov test indicates that the foreground $\theta_{H}$ distribution is very close to a random distribution.

In Fig. 3 we overlaid the $H$-band polarization of the 356 background stars on the extinction map of the bubble N4 333In the transformation from $N(\mathrm{H}_{2})$ to $A_{V}$, the equation $A_{V}=9.0\times 10^{-22}~{}\mathrm{cm}^{2}~{}N(\mathrm{H}_{2})~{}\mathrm{mag}$ from Güver & Özel (2009) is used.. The polarization vectors of the background stars show a certain degree of alignment in smaller areas. For instance, the polarization vectors are oriented at about $90\arcdeg$ in the area just below the bubble N4; in the area between the bubble N4 and the star-forming clump N4W, we find the polarization vectors are roughly parallel to the Galactic plane orientation of $29\arcdeg$. The area below the bubble N4 is a region of low-density molecular gas with $A_{V}\sim 7\,$mag, as inferred from the extinction map shown in Fig. 3. The contribution of the unrelated foreground ISM along the same LOS may be more or less comparable to the local ISM in the bubble N4. Therefore, the polarization direction of background stars in this area is affected by both components. In contrast, the area between N4 and N4W is high-density, so that the reddening of background stars in this area is mostly contributed by the ISM in the bubble N4.

\floattable