Abundant Methanol Ice toward a Massive Young Stellar Object in the Central Molecular Zone^*

Methanol (CH₃OH) is a key species in the formation of complex organic molecules. CH₃OH is observed in gas phase throughout the Central Molecular Zone (CMZ), the innermost ∼400 pc region of the Milky Way (see Morris & Serabyn 1996), with a wide variation in abundance relative to H₂ (∼10⁻⁸–${10}^{-6};$ Requena-Torres et al. 2006; Yusef-Zadeh et al. 2013). Complex organic molecules are also observed throughout the CMZ, with high and constant gas-phase abundances, relative to gas-phase CH₃OH, that are matched by hot cores in the Galactic disk (Requena-Torres et al. 2006).

Gas-phase production of CH₃OH is inefficient, but formation by the hydrogenation of CO ices in dense (∼10⁵ cm⁻³) and cold (∼15 K) environments, such as dense molecular cores, is very active as shown by laboratory experiments and Monte Carlo simulations (e.g., Watanabe et al. 2003; Cuppen et al. 2009; Coutens et al. 2017). A number of non-thermal desorption mechanisms, including shocks (Requena-Torres et al. 2008), cosmic rays (Yusef-Zadeh et al. 2013), and episodic explosions (Coutens et al. 2017) have been proposed to explain the widespread gas-phase CH₃OH (possibly along with other complex molecules; see Rawlings et al. 2013) in the CMZ. However, previous searches for CH₃OH ice in the CMZ have been unsuccessful; only upper limits on the abundance of solid CH₃OH relative to H₂O ice, ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(H₂O), have been established from Sgr A* ($\lt 0.04$) and the Quintuplet cluster star GCS 3-I ($\lt 0.27$) (Chiar et al. 2000; Gibb et al. 2004; Moultaka et al. 2015).

Indirect evidence of the presence of solid CH₃OH came from mid-infrared spectra of massive young stellar objects (MYSOs) in the Galactic center (GC) region (An et al. 2009, 2011). We identified these GC MYSOs by their wide absorption profile of 15 μm CO₂ ice, with a strong "shoulder" absorption component centered at 15.4 μm. To date, this 15.4 μm shoulder has only been observed toward disk MYSOs (Boogert et al. 2015 and references therein). Laboratory studies attribute the 15.4 μm shoulder absorption to Lewis interaction of CO₂ in icy grains with other molecules such as methanol, ethanol, butanol, or diethylether (Dartois et al. 1999a). CH₃OH has the highest abundance of these toward Galactic disk MYSOs.

The question emerges as to whether the observed 15.4 μm shoulder CO₂ ice absorption in the CMZ is produced by Lewis-base molecules other than CH₃OH because of the unusual CMZ conditions. CMZ molecular clouds are warmer, denser, and more turbulent than molecular clouds in the disk. Stronger tidal shear forces and magnetic fields pervade the CMZ (see Morris & Serabyn 1996 and references therein). The cosmic ray ionization rate is higher than in the disk (Goto et al. 2008). All these effects may complicate and lead to chemical networks that are distinct from those under local cloud conditions.

In this letter, we report the first direct detection of solid CH₃OH in the CMZ by searching for the 3.535 μm ${\nu }_{3}$ C–H stretching mode of CH₃OH ice in the GC MYSO SSTGC 726327. The 3.535 μm absorption is almost independent of ice mantle composition (Hudgins et al. 1993; Ehrenfreund et al. 1999), and therefore can be used to reliably measure a column density of solid CH₃OH. SSTGC 726327 is one of 35 MYSOs identified in the GC using 5–35 μm Spitzer/IRS spectra (An et al. 2009, 2011), with both the 15.4 μm shoulder CO₂ ice and 14.97 μm gas-phase CO₂ absorption. Its silicate feature implies an integrated (both foreground and internal) visual extinction ${A}_{V}=46\pm 3$ mag, while the foreground extinction from the Galactic disk is 30 ± 4 mag (Schultheis et al. 2009).

2.1. IRTF

2.2. Subaru

Figure 1 compares our $L^{\prime}$ image (middle) with the K-band image from the UKIDSS Galactic Plane Survey (Lucas et al. 2008; top) and the Spitzer/IRAC [4.5] channel image (Ramírez et al. 2008; bottom). UKIDSS photometry plus our point-spread function fitting photometry shows that $L^{\prime} =9.3$ for SSTGC 726327E, with $H-K=2.7$ and $K-L^{\prime} =1.6$. The $L^{\prime}$ (3.8 μm) photometry is in good agreement with [3.6] = 9.3 for SSTGC 726327 (Ramírez et al. 2008). The extended emission we detect at $L^{\prime}$, but not at K, roughly coincides with the 3'' (0.12 pc) diameter H ii region Sgr B1 A (Mehringer et al. 1992, dotted circle), also known as GPSR5 0.488–0.028 (Becker et al. 1994).

Zoom In Zoom Out Reset image size

The IRAC source SSTGC 726327 is a composite of flux from SSTGC 726327E, SSTGC 726327W, the H ii region, and the MYSO. The IRAC position, relative to SSTGC 726327E, is $0\buildrel{\prime\prime}\over{.} 9$ west at 4.5 μm, but is located further away ($1\buildrel{\prime\prime}\over{.} 1$) at 8.0 μm (Ramírez et al. 2008). Morales & Robitaille (2017) find that, when a candidate YSO selected based on Spitzer/IRAC photometry is matched to a UKIDSS source by their spectral energy distributions, 94% of the Spitzer and UKIDSS sources are separated by $\leqslant$057. This suggests SSTGC 726327 and SSTGC 726327E are distinct sources with a projected separation, for a GC distance of 8 kpc, of 6000 to ${\rm{10,000}}\,\mathrm{au}$.

The top panel in Figure 2 shows our 2–4 μm IRTF/SpeX spectrum of SSTGC 726327E (blue line). The spectrum shows photospheric 2.3 μm CO band-head absorption, demonstrating that SSTGC 726327E is probably a background giant star rather than a MYSO. The observed Br γ (2.166 μm) and Br α (4.051 μm) emission likely arise in the Sgr B1 A H ii region.

Zoom In Zoom Out Reset image size

We searched for the best-fitting spectral type, the peak optical depth of the 3.0 μm H₂O ice band $[{\tau }_{\mathrm{solid}}$(H₂O)], and foreground dust extinction of SSTGC 726327E, by utilizing the IRTF/SpeX spectral library (Rayner et al. 2009), the optical constants of 10 K amorphous H₂O ice on 0.4 μm grains, and the extinction curve derived from the line of sight to a dense core (Boogert et al. 2011). We minimized the reduced ${\chi }^{2}$ (${\chi }_{\nu }^{2}$) of the fit to the SpeX spectra and J and H UKIDSS photometry ($J=18.391\pm 0.068$ and $H=13.571\pm 0.003$). We masked Br α and Br γ emission lines, and excluded data points at $3.1\,\mu {\rm{m}}\leqslant \lambda \leqslant 3.7\,\mu {\rm{m}}$ to avoid the 3.4 μm absorption band discussed later. We multiplied the K and L spectra by a constant factor (1.16) to match the UKIDSS K-band measurement ($K=10.915\pm 0.003$).

We found that SSTGC 726327E is likely a M giant or K/M supergiant, along with ${\tau }_{\mathrm{solid}}$(H₂O) = 1.7 ± 0.3 and ${A}_{K}=4.2\pm 0.5.$ The errors represent a range of values within ${\rm{\Delta }}{\chi }_{\nu }^{2}\lt 2$ from the minimum ${\chi }_{\nu }^{2}$ (0.83). We found a 30% smaller A_K if we used the GC extinction curve from Fritz et al. (2011) instead; spectral types and ${\tau }_{\mathrm{solid}}$(H₂O) essentially remain unchanged within the errors. The green line in the top panel of Figure 2 shows our best-fitting spectrum (M6.5 III; HD 142143) with added reddening (A_K = 4.15). The red line represents the best-fitting spectrum to our IRTF/SpeX data, with the 3 μm H₂O ice absorption $[{\tau }_{\mathrm{solid}}$(H₂O) = 1.51] added to the reddened M6.5 III spectrum. The optical depth spectrum with respect to the best-fitting model is shown in the bottom panel of Figure 2. The strong, broad 3.4 μm absorption band in SSTGC 726327E is a blend of two distinct absorption features from icy molecular cloud grains (Brooke et al. 1999) and foreground diffuse cloud dust (Sandford et al. 1991; Pendleton et al. 1994).

In the left panel of Figure 3, our 3.4–4.0 μm IRCS spectra of SSTGC 726327E are shown before ("original") and after scaling the flux to match the IRTF data ("corrected"). Methanol ice absorption is seen at 3.535 μm in our IRTF spectrum (Figure 2), but it is more clearly observed in the IRCS spectra. We do not detect the weaker CH₃OH features at 3.84 μm and 3.94 μm (Dartois et al. 1999b).

Zoom In Zoom Out Reset image size

We derived a local continuum across the 3.535 μm band using a first order polynomial fitted at $3.448\leqslant \lambda \,\leqslant 3.497\ \mu {\rm{m}}$ (2860–2900 cm⁻¹) and $3.597\leqslant \lambda \leqslant 3.650\ \mu {\rm{m}}$ (2740–2780 cm⁻¹) in wavenumber space (dotted line). The right panel in Figure 3 shows the optical depth derived by dividing the "corrected" IRCS spectra by the local continuum. We employed a laboratory transmission spectrum of 10 K pure CH₃OH ice (Hudgins et al. 1993), in which the 3.535 μm band overlaps with another, wide band, centered at 3.39 μm (2960 cm⁻¹) from the ${\nu }_{9}$ C–H stretching mode. We fitted a first order polynomial to the laboratory spectrum over the same wavelength intervals adopted for the continuum construction, and isolated the 3.535 μm component by subtracting the local baseline. The solid lines in Figure 3 show our best-fitting model spectrum for CH₃OH ice, which has a peak optical depth at 3.535 μm, ${\tau }_{\mathrm{solid}}$(CH₃OH), of 0.10 ± 0.01. This agrees with a result from an alternative approach, in which we restricted our fit to 3.45–3.64 μm and simultaneously searched for the flux scaling factors of individual IRCS orders and the best-fitting laboratory CH₃OH ice spectrum.

Table 1 summarizes the 3.535 μm CH₃OH ice optical depth and column density for SSTGC 726327E. The shape and the intrinsic integrated band strength (A) for the 3.535 μm C–H stretch mode have a weak dependence on temperature and abundance of ice mantles (e.g., Ehrenfreund et al. 1999; Kerkhof et al. 1999). We averaged values from Hudgins et al. (1993) and Schutte et al. (1996) to adopt $A=5.95\pm 0.65\times {10}^{-18}$ cm molecule⁻¹ for CH₃OH ice, then derived its column density. We also list ${\tau }_{\mathrm{solid}}$(H₂O) for SSTGC 726327E. We computed ${N}_{\mathrm{solid}}({{\rm{H}}}_{2}{\rm{O}})$ by multiplying ${\tau }_{\mathrm{solid}}$(H₂O) by FWHM ≈ 330 cm⁻¹, and assuming $A=2.0\times {10}^{-16}$ cm molecule⁻¹ (Hagen et al. 1981).

Table 1.  Peak Optical Depth and Column Density of Solid H₂O and CH₃OH

Object(s)	AV	${\tau }_{\mathrm{solid}}({{\rm{H}}}_{2}{\rm{O}})$	${N}_{\mathrm{solid}}({{\rm{H}}}_{2}{\rm{O}})$	${\tau }_{\mathrm{solid}}({\mathrm{CH}}_{3}\mathrm{OH})$	${N}_{\mathrm{solid}}({\mathrm{CH}}_{3}\mathrm{OH})$	${N}_{\mathrm{solid}}({\mathrm{CH}}_{3}\mathrm{OH})/{N}_{\mathrm{solid}}({{\rm{H}}}_{2}{\rm{O}})$
	(mag)		(1018 cm−2)		$({10}^{18}\,{\mathrm{cm}}^{-2})$
SSTGC 726327E	24–43a	1.7 ± 0.3	2.8 ± 0.5	0.10 ± 0.01	0.47 ± 0.05	0.17 ± 0.03
Sgr A*	30	0.50 ± 0.01	1.24 ± 0.25	<0.01	<0.05	<0.04
GCS 3 I	29	0.23 ± 0.02	0.47 ± 0.04	<0.03	<0.13	<0.27
Massive YSOs						<0.03–0.31
Low-mass YSOs						<0.01–0.25
Quiescent clouds						<0.01–0.12

Note. References for the Galactic center sources: SSTGC 726327E (this paper); Sgr A* and GCS 3-I (Chiar et al. 2000; Gibb et al. 2004). Minimum and maximum values for source classes from Boogert et al. (2015).

^aA range bracketed by values from our model fitting, either ${A}_{K}=4.2\pm 0.5$ with a dense core extinction curve (Boogert et al. 2011) or a 30% lower A_K from model fitting with a GC extinction law (Fritz et al. 2011). We assumed ${A}_{K}/{A}_{V}=0.11$.

Download table as:  ASCIITypeset image

In this letter, we present the first detection of CH₃OH ice absorption toward the CMZ, in the background star SSTGC 726327E. Geballe & Oka (2010) identified 2.27 μm absorption in a background GC star as possibly due to CH₃OH ice, but the absence of 3.53 μm CH₃OH ice absorption rules out this identification (T. R. Geballe, 2016, private communication). Table 1 compares ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(H₂O) = 0.17 ± 0.03 toward SSTGC 726327E to previous upper limits from a 14'' × 20'' aperture placed on Sgr A* and from the Quintuplet cluster star GCS 3-I (Chiar et al. 2000; Gibb et al. 2004). ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(H₂O) is ≳4 times higher toward SSTGC 726327E than toward Sgr A*.

SSTGC 726327E is ∼8000 au in projection from the MYSO SSTGC 726327 in Sgr B1. The CH₃OH ice absorption we detect toward SSTGC 726327E likely arises from the extended envelope of the MYSO, which can extend to ∼30,000–90,000 au (e.g., van der Tak et al. 2000b). Pontoppidan et al. (2004) have measured column densities of CH₃OH ice from an extended envelope of a low-mass Class 0 protostar, Serpens SMM 4, based on the L-band spectra of 10 pre-main sequence stars. These "background" stars show a constant ${N}_{\mathrm{solid}}$ $({\mathrm{CH}}_{3}\mathrm{OH})$/${N}_{\mathrm{solid}}$ $({{\rm{H}}}_{2}{\rm{O}})\sim 0.3$ at projected distances of 4000–10,000 au from Serpens SMM 4, then no CH₃OH at a projected distance of 19,000 au. They conclude, as we do, that some or all CH₃OH ice absorption arises from the extended envelope of the YSO.

While we have discovered CH₃OH ice in one line of sight to CMZ, CH₃OH ice is found in abundance not only in YSOs, but also toward stars behind cold, quiescent molecular clouds and cores in the Galactic disk at ${A}_{V}\gtrsim 9$ mag (Table 1; Boogert et al. 2015). For two disk MYSOs (RAFGL 7009S and W33 A; Brooke et al. 1999; Dartois et al. 1999b), ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(H₂O) is higher than the maximum value observed toward stars behind quiescent cores in the disk (0.12; Boogert et al. 2011). This empirical division also makes the case for SSTGC 726327E being behind the envelope of the MYSO SSTGC 726327.

If SSTGC 726327 has physical and chemical structures similar to those of W33 A (van der Tak et al. 2000a, 2000b), the maximum gas density encountered along the line of sight at the projected distance of $8000\,\mathrm{au}$ is ∼6 × 10⁵ cm⁻³, which is sufficiently large enough to maintain CH₃OH formation in its envelope. However, the gas temperature within the MYSO envelope likely exceeds the sublimation temperature of CO (∼20 K) all the way to the edge of the MYSO, under which additional growth of CH₃OH on ice mantles should be suppressed (Watanabe et al. 2003; Cuppen et al. 2009). CMZ clouds are on average warmer than in the disk, but there are dense shielded regions with cold (∼15 K) dust grains (Rodríguez-Fernández et al. 2004; Molinari et al. 2011), where CH₃OH can form. Then it is possible that CH₃OH ice grains arise in dense molecular clouds within the CMZ, in which case the star SSTGC 726327E is projected by chance next to the MYSO.

The fractional abundance of CH₃OH ice with respect to H₂ is $\sim {10}^{-5}$ toward SSTGC 726327E, if our best-fitting A_K is taken with ${A}_{K}/{A}_{V}=0.11$ and $N({{\rm{H}}}_{2})/{A}_{V}\sim {10}^{21}$ cm⁻² mag⁻¹ (see Hasenberger et al. 2016, and references therein). This order of magnitude estimate is ∼10¹–10³ times larger than the gas-phase CH₃OH abundance in the CMZ (Requena-Torres et al. 2006; Yusef-Zadeh et al. 2013). While systematic searches for CH₃OH ices should be proceeded before making any firm conclusions, our estimate suggests that gas-phase CH₃OH in the CMZ can be largely produced by desorption of CH₃OH from icy grains (e.g., Requena-Torres et al. 2008; Yusef-Zadeh et al. 2013; Coutens et al. 2017).

In spite of abundant CH₃OH ices in various sight lines, broad CO₂ ice bands with the 15.4 μm shoulder absorption have, to date, only been observed toward YSOs (Boogert et al. 2015). About 19% of the CO₂ absorption in SSTGC 726327 is attributed to the shoulder CO₂ (An et al. 2011). With our measurement of CH₃OH abundance, we found ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(CO₂ shoulder) = 3.9 ± 0.4. We applied the same spectral decomposition procedure to the observed CO₂ ice profile of the ISO spectrum of W33 A (Gerakines et al. 1999), and found 10% for the fraction of the shoulder component, which results in ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(CO₂ shoulder) = 11.2 ± 2.4 based on ${N}_{\mathrm{solid}}$(CH₃OH) in Brooke et al. (1999). Although there exists a factor of three difference in this comparison, systematic errors likely dominate, since our measurement only traces CH₃OH ice in the outer part of the MYSO. The similar ${N}_{\mathrm{solid}}$(CH₃OH)/${N}_{\mathrm{solid}}$(CO₂ shoulder) implies that CH₃OH is indeed the best candidate for interacting molecules in the observed 15.4 μm shoulder CO₂ toward GC MYSOs (An et al. 2009, 2011), as it is in disk MYSOs. Additional observations are clearly needed to extend a source list, establish their properties, and estimate the variance of CH₃OH ice in the CMZ.

D.A. acknowledges support provided by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2015R1D1A1A09058700) and by the Korean NRF to the Center for Galaxy Evolution Research (No. 2010-0027910).

Facilities: IRTF - Infrared Telescope Facility, Subaru - Subaru Telescope.