Identification of Methyl Isocyanate and Other Complex Organic Molecules in a Hot Molecular Core, G31.41+0.31

The continuum image of G31 at 94.3 GHz is shown in Figure 1. The synthesized beam size of the continuum is 119 × 098 with a PA of ∼ 76°. Using a two-dimensional Gaussian fitting over the dust continuum emission, we obtained the deconvolved size of G31 ∼ 090 × 068.

Zoom In Zoom Out Reset image size

Assuming a distance to the source of 7.9 kpc, we obtained a source size of ∼7110 × 5372 au. This source size changes to 3330 × 2516 au when a distance of ∼3.7 kpc is used.

Beltrán et al. (2018) observed G31 with an angular resolution of 022 and found a continuum brightness temperature of ∼132 K, which is comparable to the dust temperature (∼150 K). This suggests that the dust continuum emission observed at 217 GHz is optically thick. Rivilla et al. (2017) also observed that dust opacity is high (τ = 2.6) at 220 GHz.

We have observed that the peak intensity of the continuum is ∼158 mJy/beam which corresponds to a brightness temperature of 18 K (using Rayleigh–Jeans approximation, 1 Jy/beam ≡ 118 K). This estimation is much smaller than the assumed dust temperature of 150 K, which indicates a small optical depth (τ_ν) of ∼ 0.12 (using T_mb = T_d(1−exp(−τ_ν)).

It should be noted that in our observation, the source was not resolved or at best marginally resolved; therefore, the estimated brightness temperature could be underestimated. After adding the beam dilution correction for the unresolved source, the brightness temperature is estimated to be ∼36 K. Considering this value of brightness temperature, dust emission remains optically thin (τ_ν ∼ 0.24 ). Due to the collapsing envelope, we can expect a higher density and temperature toward the inner region than the outer envelope of this source. Since dust density is higher toward the inner region, opacity would be higher, as observed by Beltrán et al. (2018). It may reduce the true line intensity of the observed transitions. However, due to the lower angular resolution, our present data would not provide further details about the dust opacity of the whole region of G31.

Assuming the dust continuum to be optically thin, flux density can be written as,

$$\begin{eqnarray}&&{S}_{\nu }^{\mathrm{beam}}={{\rm{\Omega }}}_{\mathrm{beam}}{\tau }_{\nu }{B}_{\nu }({T}_{d}),\end{eqnarray} \tag{ 1 }$$

where Ω_beam = $\tfrac{\pi }{4{ln}2}$ × θ_major × θ_minor is the solid angle of the synthesized beam, S${}_{\nu }^{\mathrm{beam}}$ is the peak flux density measured in units of mJy/beam, τ_ν is optical depth, T_d is the dust temperature, and B_ν(T_d) is the Planck function (Whittet 1992). Optical depth can be expressed as

$$\begin{eqnarray}&&{\tau }_{\nu }={\rho }_{d}{\kappa }_{\nu }L,\end{eqnarray} \tag{ 2 }$$

where ρ_d is the mass density of dust, κ_ν is the mass absorption coefficient, and L is the path length. Using the dust-to-gas-mass ratio (Z), the mass density of the dust can be written as

$$\begin{eqnarray}&&{\rho }_{d}=Z{\mu }_{{\rm{H}}}{\rho }_{{{\rm{H}}}_{2}}=2Z{\mu }_{{\rm{H}}}{N}_{{{\rm{H}}}_{2}}{m}_{{\rm{H}}}/L,\end{eqnarray} \tag{ 3 }$$

where ${\rho }_{{{\rm{H}}}_{2}}$ is the mass density of the hydrogen molecule, ${N}_{{{\rm{H}}}_{2}}$ is the column density of hydrogen, m_H is the hydrogen mass, and μ_H is the mean atomic mass per hydrogen atom. Here, we used Z = 0.01, μ_H = 1.41 (Cox 2000), and a dust temperature of 150 K. A two-dimensional Gaussian fitting of the continuum yields a peak intensity (S${}_{\nu }^{\mathrm{beam}}$) of dust continuum of 158.4 mJy/beam and an integrated flux (F_ν) of 242.4 mJy. The rms noise of the continuum map was found to be 5 mJy/beam. From the above equations, the column density of molecular hydrogen can be written as

$$\begin{eqnarray}&&{N}_{{{\rm{H}}}_{2}}=\displaystyle \frac{{S}_{\nu }^{\mathrm{beam}}/{{\rm{\Omega }}}_{\mathrm{beam}}}{2{\kappa }_{\nu }{B}_{\nu }({T}_{d})Z{\mu }_{{\rm{H}}}{m}_{{\rm{H}}}}.\end{eqnarray} \tag{ 4 }$$

According to the interpolation of the data presented in Ossenkopf & Henning (1994), we estimated the mass absorption coefficient. By considering a thin ice condition here, we obtained the mass absorption coefficient at 94 GHz (3187.90 μm) ∼0.176 cm² g⁻¹. By using Equation (4), the total hydrogen column density of the G31 source is ∼1.53 × 10²⁵ cm⁻².

Observed spectra were obtained within the circular region having a diameter of 11 centered at R.A. (J2000) = 18^h47^m34309, decl. (J2000) = −1°12'4600. Line parameters of all the observed transitions were obtained using a single Gaussian fit. From fitting, we have estimated the line width (FWHM), LSR velocity, and the integrated intensity. All the line parameters of observed molecules such as molecular transitions (quantum numbers) along with their rest frequency (ν), line width (FWHM, ΔV), integrated intensity (∫T_mbdV), upper state energy (E_u), V_LSR, and transition line intensity (Sμ², where S is the transition line strength and μ is the dipole moment) are presented in Table 1. Overall, we have observed a broad line width (≥5 km s⁻¹) and all lines are at the center around the systematic velocity of the source (∼97 km s⁻¹; see Table 1). In the Appendix, we have presented observed and Gaussian-fitted spectra (see Figures A2–A6).

Table 1.  Summary of the Line Parameters of Observed Molecules toward G31.41+0.31

Species	($J{{\prime} }_{{K}_{{\rm{a}}}^{{\prime} }{K}_{{\rm{c}}}^{{\prime} }}$-$J{{\prime\prime} }_{{K}_{{\rm{a}}}^{{\prime\prime} }{K}_{{\rm{c}}}^{{\prime\prime} }}$)	Frequency (GHz)	Eu/kB (K)	FWHM (km s−1)	Peak Intensity, Iν (K)	VLSR (km s−1)	Sμ2(Debye2)	∫Tmbdv (K km s−1)
SO2	83,5 − 92,8	86.63908	55.20	8.46	14.63	97.27	3.01	131.9 ± 3.6
H2CO	61,5 − 61,6	101.33299	87.57	10.20	75.90	97.26	5.05	824.4 ± 22.4
CH3SH	40,4 − 30,3 (A+)	101.13915a	12.14	8.70	6.00	97.00	6.61	55.2 ± 17.7
	40,4 − 30,3 (E+)	101.13965a	13.56	6.10	5.90	95.45b	6.61	38.5 ± 14.8
	42,3 − 32,2 (A−)	101.15933c	31.26	6.75	3.50	97.00	4.96	26.1 ± 4.7
	43,2 − 33,1 (E−)	101.15999c	52.39	7.09	1.56	95.03d	2.90	13.1 ± 2.9
	43,1 − 33,0 (A+)	101.16066c	52.55	7.09	1.56	93.05d	2.90	13.1 ± 2.9
	43,2 − 33,1 (A−)	101.16069c	52.55	7.09	1.56	92.94d	2.90	13.1 ± 2.9
	42,3 − 32,2 (E−)	101.16715e	29.62	6.94	8.20	97.00	4.96	60.6 ± 5.2
	42,2 − 32,1 (E+)	101.16830e	30.27	6.20	7.70	93.58f	4.96	50.8 ± 4.5
CH3OH	${7}_{2,{5}^{-}}-{6}_{3,{3}^{-}},{{\rm{v}}}_{{\rm{t}}}=0$	86.61557	102.70	9.06	84.55	97.02	1.35	815.3 ± 26.8
	${7}_{2,{5}^{+}}-{6}_{3,{4}^{+}},{{\rm{v}}}_{{\rm{t}}}=0$	86.90291	102.72	9.13	85.16	97.00	1.35	827.8 ± 24.9
	${15}_{3,{13}^{+}}-{14}_{4,{10}^{+}},{{\rm{v}}}_{{\rm{t}}}=0$	88.59478	328.26	8.98	82.26	97.10	4.20	786.4 ± 14.6
	${15}_{3,{12}^{+}}-{14}_{4,{11}^{-}},{{\rm{v}}}_{{\rm{t}}}=0$	88.93997	328.28	8.55	79.78	97.16	4.20	726.5 ± 9.9
	5−2,3 − 51,4, E2, vt = 0	101.12685	60.73	8.37	23.87	98.88	0.01	208.4 ± 18.8
	7−2,5 − 71,6, E2, vt = 0	101.29341	90.91	7.72	45.71	97.38	0.05	373.2 ± 18.0
	8−2,6 − 81,7, E2, vt = 0	101.46980	109.49	7.25	55.31	97.98	0.09	426.9 ± 16.5
CH3NCO	100,0 − 90,0	86.68655	34.97	7.70	14.63	97.83	83.11	120.0 ± 6.8
	100,10 − 90,9	86.68019	22.88	8.28	13.50	97.90	83.06	119.1 ± 11.5
	102,9 − 92,8	86.78077	46.73	6.80	10.45	97.32	79.73	75.8 ± 4.5
	102,8 − 92,7	86.80502	46.74	5.82	11.16	97.10	79.74	69.2 ± 3.8
	10−3,0 − 9−3,0	86.86674	88.58	10.14	15.80	97.00	75.32	170.6 ± 16.6
	102,0 − 92,0	87.01607	58.88	6.22	8.91	97.12	79.18	59.0 ± 6.2
	100,10 − 90,9	86.67492	285.03	⋯	⋯	⋯	83.07	⋯
CH3OCHO	173,14 − 173,15	86.91761	99.72	6.41	7.09	98.03	1.87	48.5 ± 7.6
	113,9 − 112,10 E	88.68688	44.97	5.41	12.90	97.86	2.44	74.3 ± 9.2
	113,9 − 112,10 A	88.72326	44.96	7.84	17.11	98.32	2.43	142.8 ± 29.0
	71,6 − 61,5 E	88.84318	17.96	8.07	45.87	98.18	18.04	394.3 ± 58.9
	71,6 − 61,5 A	88.85160	17.94	6.64	49.17	97.50	18.04	348.0 ± 22.4
	81,8 − 71,7 E	88.86241	207.10	5.83	23.49	97.64	20.89	145.9 ± 10.1
	133,11 − 132,12 E	101.37050	59.64	5.16	15.90	97.93	2.61	87.4 ± 6.7
	133,11 − 132,12 A	101.41474	59.63	5.93	14.56	97.81	2.60	92.0 ± 13.5
SiO	(2-1)	86.84696	6.25	⋯	⋯	94.1	2.82	⋯

Notes.

^aSpectral profile 1. ^bVelocity with respect to 101.13915 GHz transition. ^cSpectral profile 2. ^dVelocity w.r.t 101.15933 GHz transition. ^eSpectral profile 3. ^fVelocity w.r.t 101.16715 GHz transition.

Download table as:  ASCIITypeset image

3.2.1. Simple Molecules

We have observed SO₂ (86.63908 GHz) and H₂CO (101.33299 GHz) in G31. We have also identified some simple molecules such as HCN, H¹³CO⁺, and SiO, which are widely used to trace various physical processes of star-forming regions. The observed spectral profile of H¹³CO⁺ shows redshifted absorption, which is shown in Figure 2(a). This is a signature of an inverse P-Cygni profile. This suggests that the source has an infall nature. Observed emission of SiO and HCN traces the outflows in this source.

Zoom In Zoom Out Reset image size

Figure 2(b) shows the observed spectral profile of HCN. Here, we have identified three hyperfine transitions of HCN (F = 1 → 1, 2 → 1, and 1 → 0 hyperfine components of the J = 1 →  0 transition, see Table 2). Three hyperfine transitions of HCN appear in the absorption and are blended together. These three absorption lines together result in a broad feature. HCN traces the outflow associated with the source. The observed transitions of both HCN and H¹³CO⁺ are found to be optically thick.

Table 2.  Observed Molecular Transitions in Absorption

Species	($J{{\prime} }_{{K}_{{\rm{a}}}^{{\prime} }{K}_{{\rm{c}}}^{{\prime} }}$-$J{{\prime\prime} }_{{K}_{{\rm{a}}}^{{\prime\prime} }{K}_{{\rm{c}}}^{{\prime\prime} }}$)	Frequency in GHz	Eu (K)
H13CO+	1-0	86.75430	4.16
HCN	1-0(F = 1-1)	88.63041	4.25
HCN	1-0(F = 2-1)	88.63184	4.25
HCN	1-0(F = 0-1)	88.63393	4.25

Download table as:  ASCIITypeset image

Here, we have also detected one transition of SiO at 86.84696 GHz (J = 2 → 1). SiO shows the absorption feature toward the center of the dust continuum and the emission profile at all other positions. Beltrán et al. (2018) also observed a similar spectral feature of SiO (5-4, 217.10498 GHz). SiO traces outflows associated with this source.

3.2.2. Complex Organic Molecules

We have observed numerous COMs that were previously reported in G31 (e.g., Rivilla et al. 2017, and references therein). Additionally, for the first time in this source, we are reporting the detection of two new molecules, methyl isocyanate (CH₃NCO) and methanethiol (CH₃SH). Altogether, we have identified six transitions of CH₃NCO with the ground vibrational state (v_b = 0) and one transition with v_b = 1. Among the six transitions of CH₃NCO, one transition (86.86674 GHz) is blended with ethylene glycol (confirmed from LTE modeling).

We have identified three different peaks of CH₃SH, where each peak contains multiple transitions. These transitions are reported in Table 1. We are unable to separate these observed transitions. Because of closely spaced transitions, it is difficult to distinguish two or more different transitions from a single observed peak of CH₃SH. Thus, to obtain the observed line parameters of CH₃SH transitions, we have fitted multiple Gaussian components to the observed spectra. For multiple Gaussian fittings, we have used fixed values of velocity separation and the expected line intensity ratio. Only the FWHM is kept as a free parameter. This method works well in the observed spectral profiles 1 and 3 around 101.139 GHz and 101.167 GHz of CH₃SH, where there are two transitions around each profile (see Table 1). Two-component Gaussian fitting corresponding to the two transitions is used to fit those spectral profiles (see Table 1 and Figure A5). However, this method does not work well for spectral profile 2, around 101.159 GHz, where four transitions are present. The four transitions can be classified into two classes based on their expected peak intensity (equivalent to Sμ²) and E_up. The expected intensity ratio between these four transitions is 1:0.58:0.58:0.58. We have used two-component Gaussian fitting corresponding to the two classes of transitions, one for the 101.15933 GHz (Sμ² = 4.96, E_up = 31.26 K) transition and another component for the other three transitions, 101.15999 GHz, 101.16066 GHz, 101.16069 GHz (Sμ² = 2.90, E_up = 52.55 K) around the 101.159 GHz line profile. For the fitting, we have used a 1:1.74 (1.74 is the sum of the intensity of three transitions) intensity ratio for these two components. Therefore, the resultant Gaussian corresponding to the three transitions can be obtained via this method; as these transitions are equally intense and have a similar excitation temperature, the area under the Gaussian can be divided by three to get the contribution from each transition multiplet. We have used only one transition from these multiplets for the rotation diagram of CH₃SH since the integrated area of all three transitions is the same and has similar upper state energy (E_up). Therefore, these fitting results are used meaningfully to estimate rotational temperature using the rotation diagram method. Obtained line parameters are further used in the rotation diagram of CH₃SH. For the rotation diagram of CH₃SH, we use 101.13915, 101.13965, 101.15933, 101.1599, 101.16715, and 101.16830 GHz transitions. Gaussian fits of CH₃SH observed spectra are provided in Figure A5.

We also have identified the presence of two abundant COMs, methanol and methyl formate, in this source. Methanol is one of the most abundant COMs in star-forming regions. Here, we have identified seven transitions of CH₃OH with different upper state energy in the ALMA band 3 regions. We have also identified CH₃OCHO transitions with a wide range of upper-level energies (17–207 K), which are listed in Table 1. The integrated intensity maps of some unblended transitions of SO₂, H₂CO, CH₃SH, CH₃OH, CH₃OHCO, and CH₃NCO are shown in Figures 3(a)–(f) where the continuum is overlaid on the integrated intensity map, and all other maps corresponding to different frequencies are given in the Appendix (see Figures A9–A12). Using local thermodynamic equilibrium modeling and comparing that with the integrated intensities of observed molecular transitions, we found that various transitions of CH₃OH, CH₃NCO, and CH₃OCHO are optically thin.

Zoom In Zoom Out Reset image size

For example, in the LTE and optically thin condition, the expected line ratio between the 87.01608 GHz and 86.68019 GHz transitions of CH₃NCO is 1.6 while the observed intensity (brightness temperature) ratio is 1.44. This suggests that these two lines are optically thin. Similarly, for 86.78078 GHz and 86.68019 GHz transitions of CH₃NCO, the observed intensity ratio between these two lines is 1.35, and the expected line ratio is 1.30. For the 101.12685 GHz and 101.46980 GHz transitions of CH₃OH, the observed intensity ratio is 2.3, which is consistent with the expected LTE line ratio of 2.2. For CH₃OCHO, the observed line ratio between 88.68688 GHz and 88.85160 GHz transition is 4, which is similar to the observed integrated intensity ratio 3.7. The similarity between the expected and observed line ratios of these transitions suggest that they are optically thin.

We have detected multiple transitions of CH₃OH, CH₃OCHO, CH₃SH, and CH₃NCO, with different E_u. Thus, we employed the rotation diagram analysis to obtain the rotational temperatures. We performed the rotation diagram analysis by assuming the observed transitions are optically thin and in LTE. For optically thin lines, column density can be expressed as (Goldsmith & Langer 1999),

$$\begin{eqnarray}&&\displaystyle \frac{{N}_{{\rm{u}}}^{\mathrm{thin}}}{{g}_{{\rm{u}}}}=\displaystyle \frac{3{k}_{B}\int {T}_{{mb}}{dV}}{8{\pi }^{3}\nu S{\mu }^{2}},\end{eqnarray} \tag{ 5 }$$

where g_u is the degeneracy of the upper state, k_B is the Boltzmann constant, ∫T_mbdV is the integrated intensity, ν is the rest frequency, μ is the electric dipole moment, and S is the transition line strength. Under LTE conditions, the total column density can be written as,

$$\begin{eqnarray}&&\displaystyle \frac{{N}_{{\rm{u}}}^{\mathrm{thin}}}{{g}_{{\rm{u}}}}=\displaystyle \frac{{N}_{\mathrm{total}}}{Q({T}_{\mathrm{rot}})}{\exp }^{-{E}_{{\rm{u}}}/{k}_{B}{T}_{\mathrm{rot}}},\end{eqnarray} \tag{ 6 }$$

where T_rot is the rotational temperature, E_u is the upper state energy, Q(T_rot) is the partition function at rotational temperature. The Equation (6) can be rearranged as,

$$\begin{eqnarray}&&\mathrm{log}\left(\displaystyle \frac{{N}_{{\rm{u}}}^{\mathrm{thin}}}{{g}_{{\rm{u}}}}\right)=-\left(\displaystyle \frac{\mathrm{log}\ e}{{T}_{\mathrm{rot}}}\right)\left(\displaystyle \frac{{E}_{{\rm{u}}}}{{k}_{B}}\right)+\mathrm{log}\left(\displaystyle \frac{{N}_{\mathrm{total}}}{Q({T}_{\mathrm{rot}})}\right).\end{eqnarray} \tag{ 7 }$$

Equation (7) shows that there is a linear relationship between E_u and $\mathrm{log}({N}_{{\rm{u}}}/{g}_{{\rm{u}}})$. Two parameters, column density and rotational temperature, can be obtained by fitting a straight line to the values of $\mathrm{log}({N}_{{\rm{u}}}/{g}_{{\rm{u}}})$ plotted as a function of E_u, where N_u/g_u is obtained from the observations through Equation (5).

Rotational diagrams of methyl isocyanate, methanol, methanethiol, and methyl formate are presented in Figure 4. For CH₃OCHO, we have several transitions, but we do not see two components. Hence we fit all the data points with a line, which are not aligned. Though we did not find a very good fitting, the obtained rotational temperature of methyl formate ∼162 K is consistent with the previously measured value in this source for the same molecule (Beltrán et al. 2005; Rivilla et al. 2017). We have obtained a rotational temperature of 76 K for CH₃SH and 48 K for CH₃NCO. The rotational diagram analysis suggests two temperature components of methanol; one is a cold component and the other is a hot component. The cold component (solid black line, see Figure 4(b)) temperature is obtained 33 K. In contrast, the hot component (solid blue line, see Figure 4(b)) temperature is 181 K. The rotational diagram of CH₃OH suggests that the transitions associated with the higher E_u transitions (86.61557 GHz, 86.90291 GHz, 88.59478 GHz, 88.93997 GHz; hot component) are mainly arising from the hot gas surrounding the G31. In contrast, the emissions associated with the lower E_u (101.12685 GHz, 101.29341 GHz,101.46980 GHz; cold component) transitions are from the cold region of the HMC. However, due to our observation's lower angular and spatial resolution, we marginally resolved these transitions. Therefore, it is not possible to provide a clear insight into the spatial distribution of methanol in this source.

Zoom In Zoom Out Reset image size

The error bars (vertical red bars) in Figure 4 are the absolute uncertainty in a log of (N_u/g_u), and this arises from the error of the observed integrated intensity that we measured using a single Gaussian fitting to the observed profile of each transition. The integrated intensities with their uncertainties in all observed transitions are provided in Table 1. Our estimated rotational temperatures for methyl formate and the hot component of methanol is similar to the typical hot core temperature. On the other hand, methyl isocyanate and the cold component of methanol have lower rotational temperatures (33–48 K). However, due to poor angular resolution, we could not provide further details about these two temperature components' spatial distribution.

To obtain the molecular column density of the observed chemical species, we used LTE modeling with CASSIS. For this modeling, we used the estimated rotational temperature discussed above. We calculated the beam average column density (N_b), where the effect of beam dilution was not taken into consideration. For the model, we used five parameters: (i) column density, (ii) excitation temperature (T_ex), (iii) line width (ΔV), (iv) velocity (V_LSR), and (v) source size (θ_s). To obtain the best fit with the observation, only column density was varied by keeping all other parameters (Tex, source size, FWHM, and hydrogen column density) fixed and we chose the best-fit results when LTE model spectra are the same as the observed spectra. For LTE fitting, we used source sizes for molecular emission similar to the deconvolved size we obtained from a 2D Gaussian fitting. Line width (FWHM) for molecular transitions is estimated from the Gaussian fitting of spectral profiles, and excitation temperature (Tex) is assumed to be ∼150 K, a typical hot core temperature. Fixed parameters are noted in Table 3. We did not have a rotation diagram for all the species (e.g., SO₂, H₂CO) presented here. Thus, we followed the LTE model to derive the column density of all the species. We have also compared the column densities of CH₃OH, CH₃SH, CH₃NCO, and CH₃OCHO between the rotation diagram analysis and the LTE model (see Table 4 and Section 3.6). Observed column densities and fractional abundances of all the species are presented in Table 4. Abundances of all species are noted with regard to H₂.

Table 4.  Comparison of Observed Molecular Column Density, Fractional Abundances, and Molecular Ratios between G31 and Other Sources

Species	G31.41+0.31,		Sgr B2 (N)	Orion KL	G10.47+0.03	Chemical Model
	Column Density		Column Density	Column Density	Column Density	Column Density
	[Abundance]		[Abundance]	[Abundance]	[Abundance]	[Abundance]
	LTE Rotation Diagram
CH3OH	1.84(19)[1.2(−6)]a	2.94(19)[1.92(−6)]	4.00(19)[1.14(−5)]b	3.24(18)[1.62(−6]c	9.0(18)[1.8(−6)]d	7.56(19)[2.10(−5)]e
CH3OCHO	6.50(18)[4.25(−7)]	4.30(18)[2.81(−7)]	1.20(18)[3.33(−7)]f	1.04(17)[5.2(−8)]g	7.0(17)[1.4(−7)]d	3.31(17)[9.20(−8)]e
CH3NCO	7.22(17)[4.72(−8)]a	1.58(16)[1.04(−9)]	2.20(17)[6.11(−8)]h	7.0(15)[3.5(−9)]i	1.2(17)[8.88(−9)]j	4.68(16)[1.30(−8)]f
CH3SH	4.13(17)[2.70(−8)]	2.85(16)[1.86(−9)]	3.40(17)[9.44(−8)]b	2.0(15)[1.0(−9)]k	⋯	1.33(16)[3.70(−9)]b
SO2	2.88(18)[1.88(−7)]		8.28(16)[2.30(−8)]l	1.0(17)[5.0(−8)]m	3.0(17)[6.0(−8)]d
H2CO	2.90(18)[1.90(−7)]		1.82(17)[5.06(−8)]n	1.5(17)[7.5(−8)]o	3.0(18)[6.0(−7)]d
SiO	1.46(15)[9.54(−11)]p		4.32(15)[1.20(−9)]l	5.4(14)[2.7(−10)]q	⋯
CH3SH/CH3OH	0.0225	0.0010	0.0082	0.0006		0.0002
CH3OCHO/CH3OH	0.3541	0.1463	0.0292	0.0321	0.0778	0.0044
CH3NCO/CH3OH	0.0393	0.0005	0.0054	0.0022	0.0049	0.0006

Notes. Orion KL (${N}_{{{\rm{H}}}_{2}}$) = 2.0 × 10²⁴ cm⁻², Sgr B2(N) (${N}_{{{\rm{H}}}_{2}}$) = 3.6 × 10²⁴ cm⁻², G10.47+0.03 (${N}_{{{\rm{H}}}_{2}}$) = 5.0 × 10²⁴ cm⁻², G31.41+0.31 (${N}_{{{\rm{H}}}_{2}}$) = 1.53 × 10²⁵ cm⁻².

^aFor the hot component of methanol, the column density is derived using a higher value of excitation temperature T = 300 K. Using LTE, we have found opacities ≥1 for a few transitions of CH₃OH (86.68655, 86.68019, 86.78077, and 86.80502 GHz) and CH₃NCO (86.68019 and 86.68655 GHz). Thus, the column density estimated for CH₃OH and CH₃NCO by using LTE should be considered a lower limit. ^bMüller et al. (2016). ^cFavre et al. (2015). ^dRolffs et al. (2011). ^eGarrod (2013). ^fBelloche et al. (2016). ^gFavre et al. (2011). ^hBelloche et al. (2017). ⁱCernicharo et al. (2016). ^jGorai et al. (2020). ^kKolesniková et al. (2014). ^lHiguchi et al. (2015). ^mTercero et al. (2018). ⁿBelloche et al. (2013). ^oEsplugues et al. (2013). ^pSiO column density and fractional abundance calculated in the outflowing gas measured at V_LSR = 94.1 km s⁻¹ by averaging over the box (7'' × 7'') centered at (α(J2000), δ (J2000)) = [18^h47^m34121, −1°12'48975]. ^qSchilke et al. (2001).

Download table as:  ASCIITypeset image

Absorption studies are best suited to trace the physical conditions (e.g., density, kinetic temperature) of the ISM because it is not induced by beam dilution as long as the absorbing layer is larger than the continuum source. Here, we used non-LTE modeling to derive excitation conditions and physical properties from these absorption features. For the non-LTE modeling, we use the RADEX program (Van der Tak et al. 2007). The collisional rates were taken from the LAMDA database and the spectroscopic constants were taken from the JPL/CDMS catalog. In RADEX, we used a continuum background temperature (T_c) of 36 K, which corresponds to the peak brightness temperature of the 3.2 mm continuum, considering the beam dilution. For this modeling, we have used a colum density of 3.55 × 10¹⁵ cm⁻², 8.92 × 10¹⁵ cm⁻² and 1.18 × 10¹⁵ × cm⁻² for H¹³CO⁺, HCN, and SiO respectively. Figures 5(a)–(c) show the variation of radiation temperature of HCN (1-0), H¹³CO⁺ (1-0), and SiO (2-1), respectively, for a wide range of parameter space. Our parameter space consists of the number density of hydrogen (varies in between 10³ and 10⁷ cm⁻³) and kinetic temperature (varies between 1 and 100 K). Figures 5(a)–(c) show the parameter space for the absorption features of HCN, H¹³CO⁺, and SiO, respectively. We have estimated the critical density of HCN (1-0), H¹³CO⁺ (1-0), and SiO (2-1), which is 1.51 × 10⁶ cm⁻³, 1.83 × 10⁵ cm⁻³, and 2.26 × 10⁵ cm⁻³, respectively. We have shown that for a hydrogen density ≥1.0 × 10⁵ cm⁻³, all these transitions are showing absorption features. If H₂ density is below the critical density, level populations are mostly governed by collisional excitation and de-excitation rather than a Boltzmann distribution. Thus, these molecules might show emission even at T ≤ T_c. The observed spectral profile of SiO is in line with these results. We found SiO emission at all positions except the source's central region (where density is high, which is probably greater than the critical density).

Zoom In Zoom Out Reset image size

In this source, we obtained a temperature range 33–180 K. The hydrogen column density of this source is estimated as ∼ 1.53 × 10²⁵ cm⁻². Various observed molecular emissions are discussed below.

3.5.1. Infalling Envelope

3.5.2. Molecular Outflow

Many authors previously studied molecular outflows associated with G31 (e.g., Olmi et al. 1996; Cesaroni et al. 2011; Beltrán et al. 2018). SiO and ¹³CO revealed the presence of two molecular outflows: One in the east–west (E-W) direction and another one in the north–south (N-S) direction. SiO is an excellent tracer of shock and molecular outflow associated with star-forming regions (e.g., Zapata et al. 2009; Leurini et al. 2014). Beltrán et al. (2018) showed average redshifted and blueshifted wings of SiO emission which reveal three outflow directions associated with this source: E-W, N-S, and NE-SW. Among these three outflow directions, the N-S and E-W components are clearly present in our observation. However, the NE-SW component is not very evident (see Figure 6, left panel, of Beltrán et al. (2018) and Figure A1). Channel maps of SiO emission are shown in Figure 6. The blueshifted and redshifted integrated emission of SiO along with the continuum is depicted in Figure 7. The presence of HCN is widespread in diverse interstellar environments ranging from the dark cloud to the star-forming region and circumstellar envelope of a carbon-rich star (e.g., Snyder & Buhl 1971; Ziurys & Turner 1986). HCN also traces molecular outflow in G31. The observed spectrum profile of HCN (see Figure 2(b)) is symmetric with respect to the systematic velocity of the source. However, the intensity of emission wings associated with the absorption profile is asymmetric. This may be due to the asymmetry in molecular distributions. Figure 8 shows the redshifted and blueshifted integrated emissions of HCN overlaid on the continuum.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

To obtain the outflow parameters, we estimated the outflow velocity as the maximum velocity of outflow emission with respect to the systematic LSR velocity (97 km s⁻¹). For the blue lobe and red lobe of SiO outflow, velocities are found to be 23 km s⁻¹ and 33 km s⁻¹. We estimated the dynamical timescale of outflow by using t_dyn = $\tfrac{l(b,r)}{{V}_{\max }(b,r)}$. For this, we calculate two different lengths (lb1, lb2) of the blue lobe and two lengths (lr1, lr2) for red lobes (see Figures 7–8). We obtained an average dynamical timescale of the outflow of ∼ 4 × 10³ years by considering the maximum velocity of SiO of the blue lobe and red lobe. The dynamical timescale of HCN outflow is found to be ∼3 × 10³ yr. These values are in good agreement with the dynamical timescale of outflow of ¹²CO and SiO, as previously reported in G31 by Cesaroni et al. (2011) and Beltrán et al. (2018). For all individual positions, the different dynamical timescales for both blue and red lobes are presented in Table 5. For HCN, we calculate the outflow parameters for one blue lobe (lb1) and one red lobe (lr1). To estimate the other outflow parameters such as mass, momentum, energy, mass loss, momentum loss, and energy loss per year, we used the formula mentioned in Cabrit & Bertout (1992). To estimate the outflow parameters, we assumed the same abundance (1.0 × 10⁻⁸) and excitation temperature (20 K) of SiO (Codella et al. 2013) and HCN. All the outflow parameters are summarized in Table 5. All these outflow parameters are in good agreement with the observed values of Araya et al. (2008).

Table 5.  Physical Parameters of the Outflow

Parameters	SiO	HCN
Distance from continuum position to outer contour (pc)
Red lobe (lr1)	0.117 (0.055)	0.118 (0.055)
Red lobe (lr2)	0.087 (0.041)	⋯
Blue lobe1 (lb1)	0.108 (0.051)	0.09 (0.042)
Blue lobe2 (lb2)	0.135 (0.063)	⋯
Typical outflow velocity: Voutflow (km s−1)
Red lobe	33.0	40.0
Blue lobe	23.0	30.0
Dynamical time: tdyn (yr)
Blue lobe	4.59 × 103 (1.62 × 103)	2.87 × 103 (1.35 × 103)
	5.74 × 103 (1.21 × 103)	⋯
Red lobe	3.47 × 103 (2.16 × 103)	2.90 × 103 (1.36 × 103)
	2.58 × 103 (2.70 × 103)	⋯
Average	4.10 × 103 (1.92 × 103)	2.88 × 103 (1.36 × 103)
Outflow mass (M⊙)
Blue lobe	27 (6)	12 (2.6)
Red lobe	24 (5)	55 (12)
Total	51 (11)	67 (14.6)
Momentum (M⊙ km s−1)
Blue lobe	619 (136)	352 (77)
Red lobe	790 (173)	2213 (485)
Total	1409 (309)	2565 (562)
Kinetic energy (L⊙ yr)
Blue lobe	7.1 × 103 (1.6 × 103)	5.3 × 103 (1.2 × 103)
Red lobe	1.3 × 104 (8.8 × 103)	4.4 × 104 (9.7 × 103)
Total	2.0 × 104 (1.0 × 104)	4.9 × 104 (1.1 × 104)
Mass loss (M⊙ yr−1)
Blue lobe	5.2 × 10−3 (4.1 × 10−3)	4.1 × 10−3 (1.9 × 10−3)
Red lobe	7.9 × 10−3 (2.1 × 10−3)	1.9 × 10−2 (8.9 × 10−3)
Total	1.3 × 10−2 (6.2 × 10−3)	2.3 × 10−2 (10.8 × 10−3)
Momentum loss (M⊙ km s−1 yr−1)
Blue lobe	0.12 (0.09)	0.12 (0.06)
Red lobe	0.26 (0.07)	0.76 (0.35)
Total	0.38 (0.16)	0.88 (0.41)
Energy loss (L⊙)
Blue lobe	1.38 (1.10)	1.84 (0.85)
Red lobe	4.31 (3.63)	15.26 (7.14)
Total	5.69 (4.73)	17.10 (7.99)

Note. To estimate the SiO outflow physical parameters, we considered an average value of the red lobe $\left({{r}}_{r}=\tfrac{{lr}1+{lr}2}{2}\right)$ and blue lobe $\left({{r}}_{{b}}=\tfrac{{lb}1+{lb}2}{2}\right)$ and the corresponding average timescales of the blue lobe and red lobe. We have calculated the lobe distance in arcseconds and then converted into parsecs by considering a source distance of 7.9 kpc. In parentheses, we have noted the outflow parameters by considering the new distance of 3.7 kpc for G31.

Download table as:  ASCIITypeset image

In Figure 6, we have shown the channel maps of SiO emission associated with the HMC. This figure depicts that blueshifted channels are distributed in the southwest region and redshifted channels toward the southeast and northeast regions. Bipolar lobe morphology observed from 93.2 km s⁻¹ and 103.1 km s⁻¹ channels suggests wide-angle bipolar outflow. Integrated emission of SiO and HCN transition (Figures 7 and 8) shows that there is a clear outflow direction along the east–west direction. In the integrated emission, the red lobe of the outflow is slightly tilted to the southeast. This may occur due to another outflow direction along the southwest-northeast direction. There is also an outflow in the N-S direction. Outflow of HCN along the N-S direction is wider (see Figure 8) as compared to the SiO outflow (see Figure 7). This may be attributed to the excitation conditions and chemical origin of HCN. The estimated outflow is massive, 50 M_⊙ for SiO and 67 M_⊙ for HCN and the outflow is characterized as young (3–4 × 10³ yr). However, by considering the new distance measurement of ∼3.7 kpc of this source, we obtained a dynamical timescale of outflow around 1–2 × 10³ yr. The estimated outflow mass is ∼11 M_⊙ for SiO and ∼14.6 M_⊙ for HCN.

In this section, we present the molecular abundance of all the observed species. We compare the abundance of all detected species between G31 and other sources, which is presented in Table 4. All the molecular abundances are reported with regard to H₂. Figure A1 in the Appendix shows the schematic diagram of kinematics and molecular distributions of G31 based on the present observation. There are three outflow directions; one is along the E-W direction, one is along the N-S direction, and another is along the NE-SW direction (only for SiO). Outflow components are traced by HCN and SiO emissions. Here, we have obtained different molecular distributions of COMs. The emitting regions of different transitions of COMs are different. We have separated those into three regions such as the emitting region is smaller than the beam size (θ ∼ 08; see the purple circle in Figure A1), comparable to the beam size (θ ∼ 11; see the green circle outer edge in Figure A1), and greater than the beam size (θ ∼ 14; see the cyan circle in the Figure A1). The emitting regiona of all observed species are provided in Table B1 (see Appendix B). The uncertainties of emitting regions listed in Table B1 are the lower limit of true values as estimated emitting regions might be affected by several issues such as poor angular resolution, image noise, and the non-Gaussian shape of the source. Emitting regions of CH₃OH and CH₃OCHO are comparable or slightly greater as compared to the beam size. Molecular distributions show that CH₃NCO and CH₃SH have a slightly lower-emitting region as compared to beam size (marginally resolved). The data presented in this paper suggest possibly different emitting regions for different COMs, but these are not conclusive because of comparable beam and source size. High sensitivity data with high angular and spatial resolution is required to understand the spatial distributions of these molecules.

3.6.1. Sulfur Dioxide, SO₂

3.6.2. Formaldehyde, H₂CO

3.6.3. Methanol, CH₃OH

3.6.4. Methanethiol, CH₃SH

3.6.5. Methyl Formate, CH₃OCHO

3.6.6. Methyl Isocyanate, CH₃NCO

CH₃OH and CH₃SH molecules are the chemical analog of each other and have a similar structure; the only difference is that one has an oxygen atom and the other has a sulfur atom. The emitting regions of CH₃SH transitions are compact and marginally resolved, whereas the emitting region of methanol is slightly extended compared to CH₃SH. A large number of previous experimental, theoretical, and observed results suggest that these two species are mainly produced on the grain surface via successive hydrogen addition reactions (e.g., Das et al. 2008, 2010, 2016; Das & Chakrabarti 2011; Fedoseev et al. 2015; Müller et al. 2016; Gorai et al. 2017; Vidal et al. 2017) with CO and CS, respectively. The main route of formation of methanol and methanethiol is given below,

$$\begin{eqnarray*}&&\mathrm{CO}\mathop{\to }\limits^{{\rm{H}}}\mathrm{HCO}\mathop{\to }\limits^{{\rm{H}}}{{\rm{H}}}_{2}\mathrm{CO}\mathop{\to }\limits^{{\rm{H}}}{\mathrm{CH}}_{3}{\rm{O}}\mathop{\to }\limits^{{\rm{H}}}{\mathrm{CH}}_{3}\mathrm{OH}\end{eqnarray*}$$

and

$$\begin{eqnarray*}&&\mathrm{CS}\mathop{\to }\limits^{{\rm{H}}}\mathrm{HCS}\mathop{\to }\limits^{{\rm{H}}}{{\rm{H}}}_{2}\mathrm{CS}\mathop{\to }\limits^{{\rm{H}}}{\mathrm{CH}}_{3}{\rm{S}}\mathop{\to }\limits^{{\rm{H}}}{\mathrm{CH}}_{3}\mathrm{SH}.\end{eqnarray*}$$

For the above reaction schemes, the first and third steps possess an activation barrier, whereas the second and fourth steps are barrier-less (e.g., Hasegawa et al. 1992; Gorai et al. 2017). At low-temperature regions, methanol and methanethiol are efficiently produced on the grain surface and released to the gas phase via non-thermal desorption in the cold phase and during the warm-up stage via a thermal desorption process. From the astrochemical simulation, Müller et al. (2016) found that the peak abundance of methanol occurs around 130 K. Here, our observed results of methanol predict a gas temperature of ∼180 K, which is similar to the typical hot core temperature. In addition to grain surface reactions, H₂CS can also be produced in the gas phase. In the gas phase, H₂CS is mainly produced from H₃CS⁺, where H₃CS⁺ is formed via the ion-neutral reaction between S⁺ and CH₄. Below ∼20 K with a density (n_H) around 10⁴–10⁵ cm⁻³, H₂CS is accreted onto the grain surface directly from the gas phase, which may also contribute to the CH₃SH formation (Bonfand et al. 2019).

The observed abundance of CH₃SH in G31 is 2.7 × 10⁻⁰⁸ and 1.0 × 10⁻⁰⁹ in Orion KL (Kolesniková et al. 2014, see Table 4). The observed abundance of methanethiol in Sgr B2 is ∼3.5 times higher than the G31. Methanol abundance is almost similar in G31, Orion KL, G10.47+0.03, and 10 times higher in Sgr B2 (see Figure 9). The observed CH₃SH/CH₃OH ratio in G31 is similar to Orion KL and it is eight times higher in Sgr B2. Modeling results of these two species also have comparable abundances and molecular ratios (see Table 4).

Methyl formate can efficiently be produced on the surface of dust grains through the reaction between methoxy radical (CH₃O) and formyl radical (HCO; Garrod et al. 2008; Garrod 2013). Garrod et al's simulation shows that, around 30–40 K, these radicals are mobile, and the reaction is efficient. It is clear from their simulation that gas phase CH₃OCHO is mainly coming from the ice phase (see Figure 1 of Garrod 2013). Increased UV photodissociation of CH₃OH leads to the formation of various radicals CH₃, CH₃O, and CH₂O, around 40 K temperature; these are mobile and form the COMs (e.g., CH₃OCHO, CH₃OCH₃). In the gas phase, when T ∼ 40 K, protonated methanol and H₂CO (when it is abundantly released to the gas phase) react and form HC(OH)OCH${}_{3}^{+}$. Then electron recombination of HC(OH)OCH${}_{3}^{+}$ produces CH₃OCHO (Bonfand et al. 2019). An efficient gas phase reaction of methyl formate production was proposed by Balucani et al. (2015) for the cold environment, where dimethyl ether is the precursor of methyl formate. Dimethyl ether is also present in this source (Isokoski et al. 2013). Therefore methyl formate can be produced through both grain surface and gas phase reactions. The observed abundance of CH₃OCHO in G31 is 4.25 × 10⁻⁷, which is similar to the observed abundance in Sgr B2, G10.47+0.03, and 10 times higher as compared to the Orion KL and modeling results. In G31, we have observed a CH₃OCHO/CH₃OH abundance ratio of ∼0.14–0.35. For other high mass star-forming regions, this ratio is found to be 0.03, 0.03, and 0.08 in Sgr B2, Orion KL, and G10.47+0.03, respectively (see Table 4).

In this work, we find the emitting regions of CH₃NCO are compact. In this source, Rivilla et al. (2017) also observed a similar emitting region of ethylene glycol. It is primarily produced on the grain surface via recombination of two HCO radicals followed by successive hydrogen addition reactions (Coutens et al. 2017). In the case of CH₃NCO formation, Halfen et al. (2015) proposed two routes: (i) a reaction between HNCO(HOCN) and CH₃ and (ii) a reaction between HNCO(HOCN) and ${\mathrm{CH}}_{5}^{+}$. Cernicharo et al. (2016) proposed that it could be produced on the grain surface. The experiment result suggests that CH₃NCO can efficiently be produced on the solid phase through the reaction between CH₃ and (H)NCO (Ligterink et al. 2017). The hot core model suggests that it could be produced via a radical–radical reaction between CH₃ and OCN (Belloche et al. 2017). Since the reaction mentioned above is exothermic and barrier-less, it can efficiently be produced on the grain surface at a little warmer temperature (>30 K), when radicals get enough mobility. Though CH₃NCO is efficiently produced on the grain surface, it could also be destructed by hydrogen addition reactions and produce n-methyl formamide (Belloche et al. 2017). Quénard et al. (2018) also considered similar radical–radical reactions in their chemical model to study the evolution of CH₃NCO under hot core conditions. In the hot core, when the temperature reaches around 100 K, CH₃ and HNCO can thermally desorb from the grain surface and increase the formation of CH₃NCO. Our estimated CH₃NCO abundance in G31 is ∼4.72 × 10⁻⁸. For other hot molecular cores (e.g., Sgr B2, Orion KL), we have found a comparable abundance of CH₃NCO. The observed fractional abundance of CH₃NCO in this work is well within the modeled results (model 5) of Belloche et al. (2017).

The identification of CH₃OH, CH₃OCHO, CH₃NCO, and CH₃SH in G31 suggests both grain surface and gas phase chemistry are efficient for the formation of COMs in this source. However, we cannot distinguish the contribution from the gas and grain phase separately due to the observational limit. To explain the various observed line profiles in this work requires a combined chemical and radiative transfer model, which will be carried out in our follow-up study.