Diagnosing Triggered Star Formation in the Galactic H ii region Sh 2-142

PMS stars are known to exhibit the Hα line emission in their spectra as a sign of chromospheric activity or circumstellar accretion, both indicative of stellar youth. We adopted the sample of 14 Hα emission stars compiled by Ogura et al. (2002) that are mostly located outside or near the tip of BRC 43, as illustrated by the blue square in Figure 1 of the spatial coverage of our study. The equivalent width listed in their work allows distinction between classical T Tauri stars (CTTS, with EW > 10 Å) and weak-lined T Tauri stars (with EW < 10 Å), the latter being relatively more evolved so as to have largely cleared their inner disks. This sample is limited in spatial coverage and, being detected in optical wavelengths, is also biased against embedded sources. Relatively evolved stars, such as magnetically active dwarfs (dMe) are potential contaminants.

Excess of infrared emission is the consequence of absorption of starlight in UV and visible by circumstellar grains, which then reradiates at longer wavelengths. The position in a color–color diagram essentially represents different slopes in the spectral energy distributions (SEDs), with $\alpha =d\mathrm{log}(\lambda$ ${F}_{\lambda })/d\mathrm{log}(\lambda )$, where λ is the wavelength, F_λ is the flux density, and the slope α serves as the proxy of the amount of thermalized dust, and hence the evolutionary stages of YSOs. Here we use a J − H versus H − K_S color–color diagram capable of qualifying the level of infrared excess and distinguishing this from interstellar reddening. Of the 18,316 2MASS sources in our field, 6901 have the photometric quality flag "AAA," i.e., with an uncertainty less than 0.1 mag in all of the three bands. These are the stars used in our analysis for the 2MASS sample. Comparison of the observed 2MASS colors with the intrinsic loci of main-sequence stars and of giants, including added effects of interstellar reddening, leads to classification, in progressively increasing infrared excess (proxy of amounts of circumstellar dust), of classical T Tauri (solar-type) candidates, Herbig Ae/Be (intermediate-mass PMS) candidates, or protostars. In Figure 2, the rightmost trapezoid enclosing reddened T Tauri stars, but excluding main-sequence stars experiencing normal interstellar reddening, signifies the region within which young stars are likely located. A total of 29 such near-infrared excess objects, namely classical T Tauri candidates, were identified. These are shown in Figure 2, together with Hα emission sources.

Zoom In Zoom Out Reset image size

Deeply embedded stars may escape 2MASS detection; these are then selected from archival WISE data. Of the 9457 sources in the region, 1846 have photometric uncertainties less than 0.1 mag in 3.4 μm (W1), 4.6 μm (W2), and 12 μm (W3), with the quality flag "AAA." We adopted the scheme of Koenig & Leisawitz (2014), i.e., with color cuts from W1 to W3 for different YSO classes while winnowing out extragalactic contaminants such as active galactic nuclei or star-forming galaxies. Using (W1)−(W2) versus (W2)−(W3), we retrieved eight Class I and 21 Class II objects, as illustrated in Figure 3. Three out of eight Class I objects have reliable detection in every 2MASS band, each with the consistency of near-infrared excess. The remaining five do not have reliable 2MASS measurements, plausibly indicative of elevated extinction. Out of the 21 WISE-selected T Tauri candidates, 13 appear to also exhibit an excess of near-infrared (J − H ≳ 0.75 and H − Ks ≳ 0.4 in Figure 2). Four of the 21 WISE-selected T Tauri candidates are also common to the near-infrared-selected CTTS sample.

Zoom In Zoom Out Reset image size

More embedded sources need to be investigated at longer wavelengths. Figure 4 presents the W1 − W2 versus W3 − W4 color–color diagram, in which transition disk objects stand out, exhibiting excess in 10–20 μm. By contrast, Class II objects with prominent near-infrared excess can no longer be selected distinctly in this diagram as in Figure 2 or Figure 3. Note that in Figure 4, a few Hα stars, candidate CTTSs are also common to transition disk objects, which are a subset of T Tauri stars with the inner disks being cleared; that is, they are in an evolutionary phase between Class II and Class III objects, and the common ones are included in our young stellar sample. The analysis above demonstrates the necessity of using a combination of colors, data quality permitting, with a wider range of wavelength coverage to acquire a more complete YSO sample. The catalog of PMS candidates is presented in the Appendix (Table 2).

Zoom In Zoom Out Reset image size

We matched counterparts within a 3'' radius with the BVI (this work), 2MASS, and WISE sources to construct the SED of each source for further investigation. The SED fitting was done with the tool developed by Robitaille et al. (2007), which uses a precomputed grid of 20,000 young stellar models to constrain the parameters for the central object (e.g., mass, temperature, age) and for the envelope (e.g., disk mass and accretion rate). The number of observational data points, and respective wavelength coverage, dictates the reliability of the constraints on the parameters. We required a minimum of four data points in an SED, and allowed in the fitting the extinction (A_V ) to vary from 1.5 mag (minimum foreground extinction toward the cluster) to 15 mag, and the distance to vary from 2.2 to 3.5 kpc, encompassing the literature values (Georgelin & Georgelin 1976; Moffat 1971; Chavarria-K et al. 1994). Figure 5 illustrates the distinct SED behavior of an example of a Class I with an ascending SED with wavelength versus that of a Class II object. The best-fit parameters such as the disk mass (M_Disk), central object mass (M_*), temperature (T_*), adopting the approach given in Samal et al. (2012) using weighted mean and the standard deviation, are presented in Table 1 for Class I objects. The results suggest that most Class I objects have ages ≲ 1 Myr, consistent with them being young and embedded.

Zoom In Zoom Out Reset image size

Table 1. SED Fitting Parameters for Class I YSOs

ID	α (J2000)	δ (J2000)	[3.4] − [4.6]	[4.6] − [12]	Age	M*	T*	MDisk	$\dot{M}$
	(deg)	(deg)	(mag)	(mag)	(Myr)	(M⊙)	(103 K)	(M⊙)	(10−5 M⊙ yr−1)
MIR-1	341.440025	57.847372	1.590	3.475	0.002	0.623	3.776	0.006	17.637
MIR-2	341.575326	57.989582	1.611	3.221	0.937	0.850	4.842	0.006	10.649
MIR-3	341.642634	57.995816	1.234	3.549	0.021	2.455	4.09	0.033	466.956
MIR-4	341.943429	58.130137	1.274	3.859	0.015	4.73	4.224	0.036	67.637
MIR-5	341.969577	58.131370	0.934	3.256	0.713	3.26	5.864	0.063	16.305
MIR-6	341.981224	58.160661	2.221	4.099	2.470	3.661	12.547	0.025	92.607
MIR-7	341.898887	58.153404	0.960	3.331	0.500	4.891	7.9	0.000	50.080
MIR-8	341.93185	58.328075	0.909	3.153	0.258	0.343	2.961	0.004	0.786

Download table as:  ASCIITypeset image

We cross-matched the list of young objects in the region with Gaia EDR3 sources (Gaia Collaboration 2020) within a coincidence radius of 2'', resulting in two out of eight candidate protostars (Class I), and 46 out of 54 PMS objects having parallax and proper motion measurements. To our knowledge, no radial velocity information is available in the literature for any of the matched sources. The Gaia EDR3 ID and measurements are listed in the YSO catalog presented in the Appendix. The distributions of parallax and proper motion values are shown in Figure 6. The region of our study is nebulous, rendering accurate astrometric estimation, with which parallax and proper motion are derived, difficult. This is evident by a noticeable fraction of negative parallax values, including some of our young candidates. The median motion of the young stellar sample is (${\mu }_{\alpha }\cos \delta ,{\mu }_{\delta })=(-2.52,-2.16)$ mas yr⁻¹ which is also close to the average proper motions of the cluster NGC 7380, (${\mu }_{\alpha }\cos \delta ,{\mu }_{\delta })=(-2.517\pm 0.131,-2.144\pm 0.131)$ mas yr⁻¹ (Cantat-Gaudin et al. 2018). Those objects that deviate more than 4 times the median absolute deviation, therefore being inconsistent with the cluster's motion, are noted in Figure 6(d) and in Table 2 in the Appendix. In Figure 6(c) those young stars having ϖ ≤ 0 or ϖ ≥ 1.2 mas, which would have been excluded of membership however exhibit a grouping in proper motion (Figure 6(d)) consistent with that of the cluster NGC 7380.

Zoom In Zoom Out Reset image size

There are objects with large deviations from the cluster's average proper motion, yet with youth indicators such as Hα or infrared excess. For example, Star 34, a near-infrared candidate exhibiting outlying proper motion of (−10.2, 5.8) mas yr⁻¹, a parallax of 1.06 ± 0.14 mas, and a distance of ${0.94}_{-0.11}^{+0.13}$ kpc, is likely not a member. Other examples include the Hα star 9, and the candidate young star 37, both having motions inconsistent with membership. Four objects, namely stars 6, 23, 26, and 43, have negative parallax values and their proper motions are also not in accord with membership.

To remedy the effects of negative parallax, Bailer-Jones et al. (2021) derived the Bayesian geometric distances by direction-dependent prior from a parallax value and its uncertainties. Additionally these authors also estimate a photogeometric distance based on the color, apparent magnitude, and extinction of a source. We compare the distributions of both distance determinations, shown in Figure 7, for all Gaia sources and for the YSO sample in our field of study. It is evident that the errors in distance computation for the YSOs are substantially large, essentially for those with negative parallax values.

Zoom In Zoom Out Reset image size

Notably, Star 11 ⁶ is plausibly one of the stars whose parallax data are affected by the copious H ii nebulosity. As remarked in Table 2, it is a known emission-line object (Ogura et al. 2002) and exhibits prominent infrared excess. Its parallax 2.42 ±0.26 mas, however, would place it at a heliocentric distance of ∼413 pc and a geometric distance of ${409.1}_{-40.7}^{+73.7}$ pc, making it a puzzling, isolated, foreground young star. Nevertheless, its proper motion remains consistent with being a member of the cluster. The above analysis thus indicates the primacy of proper motions over parallaxes in establishing the association of young objects to Sh 2-142.

Our CO data presented in Figures 8 and 9 evince spatial and kinematic structures of the molecular gas in the region. Spatially averaged and summed spectra, for ¹²CO and isotopic ¹³CO lines, clearly indicate two main lobes in the velocity range of −55 to −25 km s⁻¹, bifurcating around −39 km s⁻¹. The ¹³CO lines are weaker, but consistently validate the kinematic and spatial segregation of the gas. The C¹⁸O emission is much weaker, hence not considered in our analysis.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The radial velocity of DH Cep, −35.4 ± 2.0 km s⁻¹ (Wilson 1953), matches that of the average value of NGC 7380, −34.13 ± 6.09 km s⁻¹ (Kharchenko et al. 2005; Conrad et al. 2017). Jointly these then agree with the line-of-sight velocity of the bulk molecular gas, thus establishing the physical connection of the gas, star cluster, and the main exciting star in the complex.

Morphologically, the CO emission clearly embraces the cluster (Leisawitz et al. 1989; Chen et al. 2011), with a northeast–southwest extension, mainly in the two lobes comparable in spatial extents but distinguished in kinematics. The two velocity peaks are ∼4 km s⁻¹ apart, with each peak corresponding to a lobe, in agreement with what has been reported by Joncas et al. (1988). Similarity in peak velocities of the two lobes with distinctive space volumes suggests that the clouds are not independent. Slicing through the channel maps with a step of 1 km s⁻¹ reveals no systematic trend indicative of an overall cloud rotation. On the other hand, the lobes are projected to have a separation of ∼25 pc at 2.6 kpc, a length scale too long for a protostellar outflow. Typical length scales for molecular outflows are of the order of a parsec (Levreault 1988). The two lobes therefore cannot be due to either cloud rotation or stellar outflows.

The relative symmetry of the lobes, plus the small velocity gradient across the cloud, indicates that the lobes are near the plane of the sky. We propose that the two lobes are parts of the "working surface" of the stellar winds and ionizing shock fronts upon the compressed clouds. There are a couple of known O-type stars in the region (Chen et al. 2011), notably the DH Cep binary system, which emits copious X-rays possibly due to powerful colliding winds (Pittard & Stevens 2002; Bhatt et al. 2010). Such a phenomenon is seen also, for example, in Orion and Monoceros where massive stars, molecular gas, and photoionized gas interplay.

Of the Hα stars (14), near-infrared (29), and mid-infrared excess (29) objects, 38/62 are consistent with being classical T Tauri stars in the NIR color–color diagram (see Figure 10). A few objects displaying less infrared excess yet with Hα emission are likely evolved PMS, namely weak-lined T Tauri stars. To estimate the extinction of each star, we tracked back in the diagram along the reddening vector from the observed colors to the intrinsic classical T Tauri locus (the red line) or, for those outside the T Tauri regime, to the main sequence (the black curve). This analysis obviously does not apply to heavily embedded objects as they are not detected by 2MASS.

Zoom In Zoom Out Reset image size

Once the extinction is estimated, a young object is dereddened and placed in a color–magnitude diagram to infer its age and mass by comparison with theoretical isochrones, e.g., in the J versus J − H diagram as exhibited in Figure 11. While isochrone ages may not be reliable in an absolute sense, it is seen that a few Hα stars are relatively older than the near-infrared excess sources. Most of the young population have masses ≲1 M_⊙. The objects comprising the youngest group, with isochrone ages ≲0.5 Myr, are marked separately for further scrutiny.

Zoom In Zoom Out Reset image size

With the list of the young population at different evolutionary stages, we are ready to delineate their relative locations with respect to parental/remnant gas. Figure 12 displays their positions, now including also those of IRAS point sources as potential protostars. The young stars are numerous in the northeastern part of the cloud and are concentrated around BRC 43. To the northeast, where the molecular gas interacts with the ionization front, there are one IRAS source and three Class I (MIR-4, MIR-5, and MIR-6) sources, together with a young group of PMS objects associated with the cloud, some located near the rim. Outside the cloud, there is only one Class I source (MIR-7), and there are several T Tauri stars, all aligned in the direction of DH Cep. This MIR-7 source, earlier considered as a field variable star (Lata et al. 2016), in fact exhibits a Class I SED and has proper motion (−2.53, −2.12) mas yr⁻¹ and parallax 0.37 ± 0.04 mas (Gaia Collaboration 2020), consistent with being a member of NGC 7380, which has an average proper motion of (−2.51, −2.14) mas yr⁻¹ and a parallax of 0.33 ± 0.05 mas (Cantat-Gaudin et al. 2018).

Zoom In Zoom Out Reset image size

In the BRC 43 extension, five out of 14 Hα objects (limited to this region) are common to the classical T Tauri stars in our sample, whereas the remaining Hα objects are possibly weak-lined T Tauri stars. All these line up roughly toward DH Cep as well. Our analysis of the infrared data shows no Class I or Class II objects inside BRC 43, except a previously known IRAS point source (Sugitani et al. 1991) and a submillimeter source (Morgan et al. 2008). As in the northeast extension, here the youngest objects, those encircled in Figure 11, seem preferentially associated with dense parts of the cloud.

In addition to these two extensions, there are less prominent concentrations revealed by the CO data, notably the one in the southwest that harbors an IRAS source and with MIR-2 and MIR-3 located near its periphery, tantalizingly enough, facing DH Cep. The southernmost reach of the complex, distant from the massive stars and from ionized gas, still shows the presence of a few PMS objects and a protostar (MIR-1). Unlike the northeastern lobe where young stars are present inside and outside of the remnant molecular clouds that must have been carved out, here there is a paucity of relatively evolved young stars, and existing young stars have all been formed in situ within the comparatively unperturbed cloud.