An Excited-state OH Maser Survey toward WISE Point Sources

Since the 6.7 GHz CH₃OH masers are considered to be a definite tracer of MSFRs (e.g., Minier et al. 2003; Xu et al. 2008), we selected the sources from the WISE point catalog, which satisfies the color–color or color–magnitude criteria of sources having the 6.7 GHz CH₃OH masers as the MFSR candidates to be observed. The MMB survey is an unbiased survey of the 6.7 GHz CH₃OH maser conducted with the Parkes telescope in a relatively wide region of the Galactic plane (186° ≤ ℓ ≤ 60°, $\left|b\right|\leqslant {2}^{^\circ }$; Green et al. 2009, 2010, 2012; Caswell et al. 2010, 2011; Breen et al. 2015). We used the 6.7 GHz CH₃OH maser sources detected from the MMB survey as initial samples to identify candidate MSFRs from the WISE point-source catalog (for details, see Yang et al. 2017). We cross-matched the 6.7 GHz CH₃OH masers from the MMB survey and the WISE point sources within an angular separation of 7'' (e.g., Caswell 2009). We found that most of the WISE sources associated with the 6.7 GHz CH₃OH masers (in total, 473 sources having the four WISE band data) satisfy the criteria of magnitudes and colors.

• 1.  
  Magnitude criteria: [3.4] <14, [4.6] <12, [12] <11, and [22] <5.5 mag.
• 2.  
  Color criteria: [3.4]–[4.6] >2 and [12]–[22] >2.

We used these criteria to match the sources in the all-WISE point-source catalog and selected ∼13,000 candidates that are possibly associated with the CH₃OH masers. Considering the resolution (∼3') of the Shanghai Tianma Radio Telescope (TMRT) at 6.7 GHz and excluding the known 6.7 GHz CH₃OH maser sources from the MMB survey and the sources with decl. below −30°, we finally obtained 3348 sources as our targeted sample of MSFR candidates to search the 6.7 GHz methanol and the ex-OH masers. The source samples and related source parameters are given in Table 1 of Yang et al. (2017, 2019).

The survey was conducted by the TMRT from 2015 September to 2018 January (Yang et al. 2017, 2019; Chen et al. 2020). Table 1 summarizes the related observation parameters for the ex-OH maser lines in this work. A cryogenically cooled C-band receiver covering the frequency range of 4–8 GHz and the Digital Backend System were used to receive and record signals in LHCP and RHCP polarization simultaneously. The four ex-OH maser transitions at the frequencies of 4750.656, 4765.562, 6030.747, and 6035.092 MHz were covered with four individual spectral windows, each having a bandwidth of 23.4 MHz with 16,384 channels corresponding to the channel width of 1.431 kHz (or a velocity spacing of 0.07–0.09 km s⁻¹ for these ex-OH transitions). The system temperature of the receiver is about 20–30 K, the aperture efficiency of the TMRT is ∼55%, and the corresponding sensitivity is 1.5 Jy K⁻¹. The half-power beamwidth (HPBW) of these observed lines is about $2\buildrel{\,\prime}\over{.} 6$ at 6.0 GHz.

We used the position-switching mode to observe each source with two ON–OFF cycles. The ON and OFF positions have an integration time of 2 minutes for each cycle. The noise diode was used to carry out the prior flux calibration. As part of the 6.7 GHz methanol maser survey (Yang et al. 2017, 2019), W3(OH) and NGC 7538 were observed as further flux density calibrators of the 6.7 GHz methanol maser line during the observations in this project. A flux density uncertainty of the 6.7 GHz line is less than 20% even though the elevation effect was not calibrated during the observation time, which is also deemed the overall uncertainty for all spectral windows in the C-band receiver. Object 3C 147 is used to calibrate polarization because it is known to have low polarization below 10 GHz (Perley & Butler 2017). The LHCP and RHCP spectra were calibrated separately, with flux density uncertainties in both polarizations of less than 20%.

The GILDAS/CLASS ⁹ package was used to process the observed spectral data. In this work, the maser line is considered to be detected only when the peak of the flux density is stronger than the 3σ rms noise level (the typical rms noise level was approximately ∼60 mJy) and the emissions appear from more than two spectral channels and can be inspected simultaneously in both ON–OFF observation cycles. We directly determined the peaks of the maser flux density from their spectra, and the integrated flux densities of the masers were determined by integrating the maser emission over the velocity range of the emission.

Compared with the ex-OH maser catalog compiled by Qiao et al. (2014), the survey of the 6035 MHz ex-OH maser (e.g., Avison et al. 2016; Szymczak et al. 2020), and those referred to in the SIMBAD Astronomical Database, ¹⁰ we identified one, one, and five new ex-OH masers at the transitions of 4765.56, 6030.75, and 6035.09 MHz, respectively. They are labeled in Table 3 and Figures 1–3 with "new." It should be noted that two new 6035 MHz ex-OH masers are detected at higher latitudes ($\left|b\right|$ >3^°): G97.527+3.184 and G211.567−19.295. The WISE IR three-color images of these newly detected ex-OH maser sources are shown in Figure 4. The figure shows that these sources are located in relatively simple IR environments, suggesting that they do not belong to complicated clusters. Remarks for these new ex-OH maser sources are as follows.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

3.1.1. 4766 MHz Ex-OH Maser

3.1.2. 6031 MHz Ex-OH Maser

3.1.3. 6035 MHz Ex-OH Maser

3.2.1. 4766 MHz Ex-OH Maser

3.2.2. 6031 MHz Ex-OH Maser

3.2.3. 6035 MHz Ex-OH Maser

To investigate the variabilities of the detected ex-OH masers more accurately, we compared our ex-OH maser detections with the 6035 MHz ex-OH maser catalog, which was observed at a high position accuracy with the ATCA observations by Avison et al. (2016). We matched the sources detected in our observations to their catalog using a criterion of angular separation of less than 26 (corresponding to the TMRT beam size). A total of 12 maser sources were matched (see Table 3, marked with superscript 2). In Figure 5, we compared the changes in the peak flux densities of these 12 sources between our observations and the Avison et al. (2016) observations. The observations of these masers were made in 2009 and 2013 by Avison et al. (2016), while ours were made between 2016 and 2017. It was found that the majority of them show significant flux density changes in both the LHCP and RHCP, even with different variability behaviors in the two circular polarization spectra. The 6035 MHz ex-OH maser emissions are relatively stable (with flux changes of less than 30%) in only three sources (G30.770−0.804, G32.744−0.075, and G43.148+0.013) during the two observations.

Zoom In Zoom Out Reset image size

We also collected the observational data of Caswell (2001) and Caswell & Vaile (1995) using ATCA, and there are three sources that have 6035 MHz ex-OH maser data reported over these years. Figure 6 shows the variability of the 6035 MHz OH maser during the past 23 yr for the three sources. We found that the flux density of ex-OH masers changed significantly but did not show a periodic trend.

Zoom In Zoom Out Reset image size

In general, the detection rates of the ex-OH masers in our observed samples are very low, with an overall detection rate of less than 1%. In order to check the properties of the ex-OH masers in a limited number of maser detections (less than or equal to 30), we divided the observed data into three different subsamples for statistical analysis of the ex-OH maser detections. These subsamples are as follows: those associated with the 6.7 GHz CH₃OH masers only (137), those associated with RRLs only (414), and those associated with both the 6.7 GHz CH₃OH masers and RRLs (103; see Table 4 of Chen et al.2020 for more detailed information). The RRL data we used were the emission from MSFRs observed with TMRT by Chen et al. (2020). As shown in Table 4, all of the ex-OH maser transitions have the highest detection rates in the subsamples associated with both the 6.7 GHz CH₃OH masers and RRLs (4.85%, 3.88%, and 11.65% for 4766, 6031, and 6035 MHz, respectively). It is a possibility that sources associated with both the CH₃OH maser and RRL emission are more likely to be at the relatively late evolutionary stage of high-mass star formation, when the ex-OH maser is easily pumped. These results are consistent with the interstellar maser evolutionary sequence proposed by Breen et al. (2010) that OH masers are generally excited at more evolved stages of high-mass star formation. We will further discuss the evolutionary stage of the ex-OH masers in Section 4.1.

In Figure 7 (left panel), we plotted the [3.4]–[4.6] versus [12]–[22] color–color diagram of the observed WISE sources with and without ex-OH maser detections. Because the sources we selected were themselves redder in the IR wavelength ([3.4]–[4.6] > 2 and [12]–[22] > 2), we did not find a significant distribution difference in the color–color diagram between the sources associated with and without the ex-OH masers. Qiao et al. (2014) reported that the colors of the WISE sources associated with the ground-state OH masers are much redder than those associated with the 6.7 GHz CH₃OH masers. In the right panel of Figure 7, we present the [3.4]–[4.6] versus [12]–[22] color–color diagram of WISE point sources associated with the CH₃OH masers, ex-OH masers, and RRLs. However, we did not find obvious distribution differences between the WISE point sources associated with the ex-OH maser and those associated with the CH₃OH maser or RRLs.

Zoom In Zoom Out Reset image size

Figure 8 shows the [3.4]–[4.6] versus 22 μm color–magnitude diagram for sources associated with CH₃OH masers, ex-OH masers, and RRLs. Overall, no significant differences can be clearly seen in the color–magnitude distributions of the WISE point sources associated with CH₃OH masers, ex-OH masers, and RRLs. In the top panel of Figure 8, we plot the 22 μm magnitude distributions and find that the 22 μm mean magnitude of the WISE point sources associated with the ex-OH masers is generally smaller than that of those associated with the CH₃OH masers. This may suggest that the sources associated with ex-OH masers are more evolved or have thicker dust envelopes compared to those associated with CH₃OH masers. Meanwhile, the 22 μm magnitudes of WISE point sources only associated with RRLs are located toward a larger magnitude direction. It suggests that the majority of these sources compared to those associated with ex-OH and CH₃OH masers may be produced at a more evolved stage of high-mass star formation, on the basis that the MIR emission will be decreased as the young stellar objects evolve. In the right panel of Figure 8, the distributions of WISE [3.4]–[4.6] color associated with CH₃OH masers, ex-OH masers, and RRLs are shown as comparisons. It can be seen that the majority of the RRL sources have smaller [3.4]–[4.6] color than the CH₃OH and OH maser sources, further supporting that the RRL sources may be at more evolved stages.

Zoom In Zoom Out Reset image size

We plot the RRL luminosity distributions for the sources associated with RRLs only, both RRLs and CH₃OH masers, and both RRLs and ex-OH masers in Figure 9. It can be seen that the emission of CH₃OH and ex-OH masers is likely to be produced toward the sources with higher RRL luminosity. In addition, we also found that among the detected 30 6035 MHz ex-OH masers, the average peak luminosity of the 13 ex-OH masers associated with RRL emission was about 63 Jy kpc², which is 1 order of magnitude stronger than that of the remaining 17 ex-OH masers without RRL emission (∼6 Jy kpc²). This further supports that the luminous ex-OH masers are usually produced from sources with luminous RRL emission. Ouyang et al. (2019) pointed out that the 6.7 GHz CH₃OH maser is more likely produced in the regions with brighter RRLs and higher radio continuum emission. Similar to the CH₃OH masers, from the above results, we can conclude that the regions with stronger RRL emission are conducive to the production of ex-OH masers.

Zoom In Zoom Out Reset image size

It is also worth noting that, as shown in Table 3, among the 33 sources with the ex-OH maser detected (including the 4766, 6031, and 6035 MHz transitions), 28 sources were also detected with the 6.7 GHz CH₃OH masers, and 16 were also detected with RRLs. This indicates that the ex-OH maser is more likely related to the CH₃OH maser compared to the RRL. This also suggests that the ex-OH masers may be produced at evolutionary stages associated with CH₃OH masers, and both are earlier than those only associated with RRLs.

The RRL line FWHM distributions for the sources associated with RRLs only, CH₃OH masers, and ex-OH masers are given in Figure 10. Only the brightest RRL component is considered if there are multiple RRL velocity components detected in the same source. It was found that the median FWHM line widths of RRLs are similar for these sources, with 23.5, 25.2, and 25.1 km s⁻¹ for the sources associated with RRLs only, CH₃OH masers, and ex-OH masers, respectively. A large sample study showed that the typical RRL line width is ∼24 km s⁻¹ (Wenger et al. 2021), which is overall similar to our results. However, it can be seen from Figure 10 that the RRL line widths for the sources associated with the ex-OH masers are usually smaller than 40 km s⁻¹. In addition, the RRL line widths have a wider range for sources associated with RRLs only. This also supports that the RRL-only sources may be at a more evolved stage on the assumption that, with the expansion of the H ii region, stronger turbulence will lead to an increase in the RLL line widths. The Cragg et al. (2002) model suggested that the 6035 MHz ex-OH maser is more likely to appear in the region with a low temperature and high density. With the evolution toward a more expanded H ii region, it is possible that the gas temperature would increase and gas density decrease in this region, so the ex-OH maser would gradually disappear in such an environment.

Zoom In Zoom Out Reset image size

In this work, we calculate the distances of the ex-OH maser sites and their distributions in the Milky Way using the Parallax-Based Distance Calculator V2 (PBDCV2 ¹¹ ; Reid et al. 2019), which is an updated version from the old version presented by Reid et al. (2016). The median velocities derived from the 6035, 6030, or 4765 MHz ex-OH maser spectra were used to calculate the distances and spiral arm assignments of the sources in the Milky Way. We adopted the distances and spiral arm assignments with the best probability from the PBDCV2. The distance parameters and their associated spiral arms are presented in columns (3)–(6) of Table 5. Notably, for six sources (G32.744−0.075, G43.148+0.013, G45.454+0.060, G69.543−0.973, G111.526+0.803, and G111.532+0.759) with distances determined through maser trigonometric parallax measurements, we adopted their trigonometric parallax distances. These sources are marked with a numerical superscript in column (1) of Table 5.

In addition to the ex-OH maser sources detected in our survey, we also collected the known ex-OH maser sources from the catalogs of Qiao et al. (2014), Avison et al. (2016), and Szymczak et al. (2020). We recalculated the distances and spiral arm assignments of these maser sources using PBDCV2.

Figure 11 shows the Galactic distributions of the 6031 and 6035 MHz ex-OH masers detected in our survey and collected from previous works. As shown in the top panel, most of the ex-OH maser sites were distributed at lower latitudes with $\left|b\right|\leqslant {2}^{^\circ }$, while at higher latitudes with $\left|b\right|\gt {2}^{^\circ }$, only five ex-OH maser sites were detected. From the overall Galactic plane distributions shown in the bottom panel of Figure 11, the majority of the ex-OH maser sources are distributed on the spiral arms of the Milky Way. Notably, the 6031 MHz OH masers (blue plus signs) are mostly distributed in the regions close to us, which may be due to the relatively low luminosities of the 6031 MHz OH masers themselves. Most of the sources we detected in this work (blue plus signs and circles) are located in the first and second quadrants of the Milky Way and associated with the Perseus and Sagittarius arms.

Zoom In Zoom Out Reset image size

Based on the distance calculated by PBDCV2 described in Section 4.2, we can derive the peak and integrated luminosities of the ex-OH masers detected in our survey (calculated by averaging the peaks and integrated flux densities of the LHCP and RHCP), and the results are given in columns (7) and (8) of Table 5. We compared the luminosity relations among those of the 4766, 6031, and 6035 MHz ex-OH masers; the 6.7 GHz CH₃OH masers; and the RRLs. Figure 12(a) shows a strong luminosity correlation between the 6035 and 6031 MHz ex-OH masers, with a correlation coefficient of 0.98 in both the peak and integrated luminosity relations. The fitting results of the luminosity relationships are summarized in Table 6. Such a luminosity correlation between the two transitions is theoretically reasonable because they have a similar excitation environment as their pumping models (Cragg et al. 2002). In this work, the 6031 and 6035 MHz OH masers were simultaneously observed at eight sources, and the peak flux density (the mean of RHCP and LHCP) of the 6035 MHz maser was higher by a factor of ∼2 than that of the 6031 MHz maser in all of these sources. This result supports the ex-OH maser model results (Cragg et al. 2002) and is also consistent with the majority of the previous observation results (Gray et al. 1992; Baudry et al. 1997; Desmurs & Baudry 1998; Caswell 2003).

Zoom In Zoom Out Reset image size

It can be seen from Figure 12(b) that there is no significant luminosity correlation between the 6035 MHz OH maser and the 6.7 GHz CH₃OH maser. Avison et al. (2016) compared spatial separations between CH₃OH and ex-OH maser pairs and found that the majority of their separations are larger than 0.03 pc. Due to a typical size of the UC H ii region of 0.03 pc (Kurtz 2005), this may indicate that the majority of the CH₃OH and ex-OH maser pairs we matched do not originate from the same UC H ii regions. Therefore, it is not surprising to find no luminosity correlation between them in our work.

We plotted the luminosity dependence between the 6035 MHz ex-OH masers and the RRLs in Figure 12(c). It can be seen that there is also a strong luminosity correlation between the 6035 MHz OH masers and RRLs, with peak and integrated luminosity correlation coefficients of 0.72 and 0.80, respectively. This result is consistent with our previous work (Ouyang et al. 2019), further supporting that such a correlation is due to the fact that both the ex-OH masers and RRLs are excited at a more evolved stage of the high-mass star formation.

The polarimetric observations of the ground-state OH masers have been proven as the potential tracer for the large-scale interstellar magnetic field (Green et al. 2012). The ex-OH masers are expected to be closer to the massive young star than the ground-state OH masers, so they may be more sensitive to tracing the magnetic field environments closer to the star than the ground-state OH masers. We can use the following relation to estimate the magnetic field strength derived from the ex-OH maser velocity separation of each Zeeman pair: ΔV km s⁻¹/H (mG) = 0.056 and 0.079 at 6035 and 6031 MHz, respectively (Baudry et al. 1997). In this relation, a magnetic field directed away from the observer is defined as positive; it is also expressed as that the velocity of the RHCP component is toward the redshifted side relative to that of the LHCP. For some masers with multiple features, the Zeeman pair we identified is only for the prominent and easily distinguishable ex-OH maser spectral features. For the identification of the Zeeman pair, we directly use the CLASS command to accurately obtain the peak value and its corresponding center velocity for both the RHCP and LHCP. In this work, we have identified 52 Zeeman pair features in 31 sources (see Table 7), among which two maser sources are fully polarized (G24.790+0.084 and G90.921+1.487). The directions and strengths of the magnetic fields were determined for all of the identified Zeeman pairs and are listed in column (7) of Table 7. For the cases with multiple Zeeman pairs identified in the same source, we present multiple magnetic measurements that are generally coincident in their directions, even with different strengths, except for G45.454+0.060. The magnetic fields have been detected in both positive and negative directions simultaneously in G45.454+0.060 from our observations and Szymczak et al. (2020). We derived the strength of the magnetic field in the range from −6.4 to +6.3 mG and did not find any cases where the strength of the magnetic field is beyond ±10 mG, which is consistent with the results of Caswell (1997).

Compared to the magnetic field strengths measured in previous works from the same sources (see Table 7), our results are within the range of previous measurements but with some changes that are likely caused by the varied velocities in the LHCP and RHCP components. In particular, for source G80.865+0.420, the feature near −1 km s⁻¹ was used to derive the positive magnetic field in the Szymczak et al. (2020) observation but was not detected in our and Baudry et al.'s (1997) observations.

When there are multiple magnetic field strengths determined by multiple Zeeman pairs in the same source, all of the absolute magnetic field strengths are taken as for the subsequent analysis and discussion. Figure 13 shows the Galactic plane distributions of the 48 OH masers with the magnetic field direction measurements from our observations and those extracted directly from other published works (Green et al. 2012, 2015) for the OH masers. Notably, the sources shown in this figure are not a subset of those presented in Figure 11 because it also consists of six additional ground-state OH masers from the published samples. It can be clearly seen from this figure that the sources with negative directions are preferably distributed closer to us (most of them coming from the Local Arm and the tail of the Perseus arm). On the Local Arm and the tail of the Perseus arm, 11 out of 13 Zeeman pairs are negative magnetic fields, while on the other arms, many more sources with positive magnetic fields are presented. Therefore, this suggests that they may be somehow coincident in the direction of the magnetic fields in MSFRs near the solar circle. In contrast, in the tangent of the Sagittarius–Carina spiral arm, Green et al. (2012) found a consistency in the magnetic field directions on the spiral arm (the majority directed away from us, i.e., positive direction; see Figure 6 in their work) using the ground-state OH masers. Linked to our results, we found that the magnetic fields may have a uniform "positive" direction along the Sagittarius–Carina spiral arm from the current measurement data, whereas the Scutum–Centaurus arm seems likely to show a random behavior of the magnetic field directions, seen from combining this work and the Green et al. (2015) results (see Figure 9 in their work) using the ex-OH maser Zeeman measurements. These results suggest that magnetic fields have different properties on each Galactic arm from large-scale views, and the ex-OH masers might be an important supplement to the ground-state OH masers for exploring magnetic field information on the Galactic large scale. Nevertheless, only 48 OH maser sources with magnetic field measurements are included, meaning that the above statistics may be biased. Many more locations with magnetic field measurements using ex-OH or ground-OH masers are required to confirm the large-scale property of the Galaxy magnetic fields.

Zoom In Zoom Out Reset image size

We also compared the absolute magnetic field strengths of the ex-OH maser sources associated with or without the 6.7 GHz CH₃OH masers. The relevant statistical results are given in Table 8. It shows that among 30 Zeeman pairs associated with CH₃OH masers, the magnetic field strengths are in a range from 0.4 to 6.4 mG, with an average of 2.6 mG. Among 10 Zeeman pairs without CH₃OH masers, the magnetic field strengths are in a range from 0.2 to 3.8 mG, with an average of 1.9 mG. Therefore, it can be derived that the sources with CH₃OH masers generally have stronger magnetic field strengths than those without CH₃OH masers. All of these sources are associated with the ex-OH masers, so they as a whole are at a late stage of high-mass star formation. However, the sources associated with CH₃OH masers may be at an earlier stage compared to those without them. With the evolution of star formation, the source will expand (such as an expanding H ii region), and the magnetic field strength will decrease due to the conservation of magnetic flux; in the meantime, the CH₃OH maser will gradually disappear. Therefore, it is reasonable to expect that the magnetic field strength of the ex-OH sources without CH₃OH masers is weaker.