Spatial Distribution of AlO in a High-mass Protostar Candidate Orion Source I

High-temperature molecular gas containing metallic elements is potentially a good probe to trace the kinematics/dynamics of circumstellar disks, and its presence in circumstellar disks around young stellar objects (YSOs) may also give some insights into formation processes of high-temperature meteoritic components formed in the Sun’s protoplanetary disk. The Orion Kleimann–Low (KL) region is the most famous and nearest massive star formation site, and has been extensively studied since the 1970s. The KL region harbors a candidate high-mass YSO, Source I, which has a hot circumstellar rotating gas disk emanating a magnetocentrifugal wind of SiO. In this study, we report spatially resolved distributions of aluminum monoxide (AlO) emission lines at 497 and 650 GHz in the rotating outflow of Orion Source I based on subarcsecond observations obtained by the Atacama Large Millimeter/Submillimeter Array for the first time in star-forming regions. These AlO emissions are detected only at the base of the outflow as the high excitation line of H2O in spite of their low excitation temperatures. The limited distribution of AlO to the launching point of the outflow indicates that AlO is not in the gas phase in the outer part of the outflow lobes away from the disk surface, which could be attributed to recondensation of AlO as dust due to its refractory nature.


Introduction
High-temperature molecular gas containing metallic elements can be used as a probe to trace the kinematics/dynamics of circumstellar disks (e.g., Plambeck & Wright 2016;Hirota et al. 2017;Ginsburg et al. 2018Ginsburg et al. , 2019Kim et al. 2019). The Orion Kleimann-Low (KL) region is the nearest active massive star formation site (e.g., 418 pc, Kim et al. 2008) and has been extensively studied for high-mass star formation (e.g., Zuckerman et al. 1976;Genzel et al. 1978;Hirota et al. 2012Hirota et al. , 2015Hirota et al. , 2016Hirota et al. , 2017Plambeck & Wright 2016;Ginsburg et al. 2018Ginsburg et al. , 2019. Hirota et al. (2017) observed a rotational transition of silicon monoxide (Si 18 O, J=12-11) around Orion Source I, which is a young stellar object (YSO) candidate (5-7 M e : Plambeck & Wright 2016; 15 M e : Ginsburg et al. 2018) located in the Orion KL region, using the Atacama Large Millimeter/submillimeter Array (ALMA). They found that Si 18 O molecules are ejected as rotating outflow (10 km s −1 ) from >10 au from the central object, which should be driven by the magnetocentrifugal disk wind launched by the high-mass YSO candidate and should contribute to the angular momentum transfer of the system. Ginsburg et al. (2018) observed SiO v=0 J=5-4 and 29 SiO J=5-4 emissions with similar spatial distributions to that of Si 18 O J=12-11, which also indicates the presence of a rotating, expanding outflow. Kim et al. (2019) detected the emission lines of SiO in Orion Source I ( 29 SiO v=2 J=11-10, SiO v=2 J=10-9, and SiO v=4 J=11-10), among which the detection of the SiO v=4 line was the first in star-forming regions. The SiO v=2 J=10-9 emission was detected in the rotating outflow as the Si 18 O J=12-11 line (Hirota et al. 2017), while the distributions of 29 SiO v=2 J=11-10 and SiO v=4 J=11-10 lines are limited to the base of the outflow. Ginsburg et al. (2019) recently made the first detection of NaCl, KCl, and their isotopologues with 37 Cl and 41 K in the base of the outflow of Orion Source I, which could be useful molecules to study the kinematics and physical conditions of the disk surrounding Orion Source I.
Aluminum monoxide (AlO) is a molecule that can be present in the gas phase at high temperature (e.g., >1650 K at total pressure of 10 −5 bar in a gas of solar composition ;Ebel 2006), and is also potentially a good tracer of the kinematics of hightemperature region around YSOs. There has been no definitive detection of AlO in the star-forming regions although tentative detection of AlO emissions at 229.7 and 344.4 GHz is claimed in Orion Source I (Ginsburg et al. 2019). The molecule has been observed around oxygen-rich evolved stars (Tenenbaum & Ziurys 2009;Kamiński et al. 2016;De Beck et al. 2017;Decin et al. 2017;Takigawa et al. 2017). The spatially resolved observation of AlO around W Hya with ALMA showed that AlO is present only at <3 stellar radii with a consistent distribution with dust, while 29 SiO is present in the accelerated stellar outflow beyond 5 stellar radii without severe depletion inside <3 stellar radii (Takigawa et al. 2017). This indicates that AlO form aluminum oxide dust around W Hya and that the aluminum oxide dust contributes to the wind acceleration, which prevents the efficient formation of silicate dust around W Hya due to abrupt decrease of the gas density and explains the aluminum oxide dust-enriched nature of W Hya (Takigawa et al. 2017).
Aluminum monoxide is a key molecule in the formation of most refractory solids in a gas of solar composition. In outflows of oxygen-rich AGB stars, AlO form the first dust as aluminum oxide including corundum (α-Al 2 O 3 ; Takigawa et al. 2015Takigawa et al. , 2017, and such circumstellar corundum grains formed prior to the Sun's birth are found as a rare but important component (presolar grains) in pristine undifferentiated chondrites (e.g., Takigawa et al. 2014Takigawa et al. , 2018. Chondrites also contain inclusions consisting of minerals enriched in refractory elements (e.g., Ca, Al, and Ti). Such Ca-and Al-rich inclusions (CAIs) are the oldest solid objects formed in the solar system (e.g., Connelly et al. 2012). The chemistry, mineralogy, and petrology of CAIs strongly suggest their high-temperature origin in the very early stage of disk evolution, but it has been under debate where and how CAIs formed in the early solar system. The observation of AlO around YSOs may provide a new insight into the formation of the oldest high-temperature solid objects in the early solar system.
In this study we analyzed the ALMA observation data for Orion Source I in Orion KL to search for AlO molecules and report the spatially resolved detection of AlO around Orion Source I for the first time.

Data Reduction
We used the ALMA observation data of Orion Source I obtained in Cycle 1 (2012.1.00123.S) and Cycle 2 projects (2013.1.00048.S). The Band 8 data in the Cycle 2 observation and the Band 9 data in the Cycle 1 observation were used for investigating AlO emissions at 497 and 650 GHz, respectively. The data were reduced following the procedure described in Hirota et al. (2017). The reduction was processed with the Common Astronomy Software Applications software (https:// casa.nrao.edu/). We first applied the self-calibration process to obtain better quality data. Both phase and amplitude solutions were derived from the continuum image of the target by using "gaincal" and "clean" tasks. The channels to make the continuum image were defined by checking by eye the spectrum derived from the dirty image. After applying the phase and amplitude solutions, continuum emission was subtracted with the "uvcontsub" task. The line images were synthesized from these continuum-subtracted data through the "clean" process with uniform weighting. The synthesized beam was 122 mas × 97 mas (50 au × 40 au) with a position angle of 85°for the 497 GHz data, and that for the 650 GHz data was 180 mas × 145 mas (75 au × 60 au) with a position angle of −34°.
The integrated flux (moment 0) maps of the two emission features show that they are localized near the central object,  (Figures 2(c) and (d)), which were derived by applying the same fitting analysis as shown in Figure 1 to the spectrum of each pixel, are also consistent with the maps of intensity weighted velocity fields of gas species observed nearby the launching point of the outflow (Hirota et al. 2017;Ginsburg et al. 2019;Kim et al. 2019).
The velocity widths of 20-24 km s −1 (Figure 1) represent the velocity variation of rotating gas along the line-of-sight direction, and both lines show a line-of-sight velocity of around 5.5 km s −1 in the central region, where the systemic velocity of the object should be observed without a significant effect from the rotating motion. The derived velocities of ∼5.5 km s −1 are close to the systemic velocity of Orion Source I (Plambeck et al. 1990). This result consolidates our conclusion that the 497 and 650 GHz emissions originate from AlO molecules. We note that the 497 and 650 GHz emissions are not contaminated from complex organic molecules (COMs) observed in the Orion-KL region (Pagani et al. 2017). The 497 and 650 GHz emissions have broader velocity widths (Figure 1) and are present only in the compact region around Orion Source I (Figure 2), which are not the characteristics of COMs (velocity widths ∼3-5 km s −1 ) in this region.
These lines of evidence strongly suggest that AlO is the molecule emitting at 497 and 650 GHz and that AlO is present in the launching point of outflow from the circumstellar rotating disk of Orion Source I. The report of AlO emission lines at 229.7 and 344.4 GHz in Orion Source I (Ginsburg et al. 2019) also corroborates the conclusion in this study.

Discussion
The distribution of AlO is limited to the base of the rotating outflow from the disk as H 2 O at 463 GHz (v 2 =1, J Ka,Kc =4 2,2 -3 3,1 ) and the rotating outflow nearby the launching region (Figure 2), while the Si 18 O J=12-11 emission line (Hirota et al. 2017) are observed in an expanding rotating outflow (10 km s −1 ). The lower energy state of H 2 O at 463 GHz is 2744 K, while those of Si 18 O J=12-11 is 128 K. The high excitation line of H 2 O is therefore observed only at the base of the outflow.
The upper state energies of AlO lines at 497 and 650 GHz are ∼167 and ∼281 K, respectively (Pickett et al. 1998), and are much lower than that of H 2 O at 463 GHz. The AlO emissions could thus be detected in the wide range of the outflow from the disk as Si 18 O J=12-11 if AlO molecules were present in the gas phase with the solar Al/Si ratio of ∼0.08 (Lodders 2003). The limited distribution of AlO emissions to the base of the outflow (Figure 2) indicates that AlO molecules are not present in the gas phase, which is likely due to its refractory nature. Aluminum monoxide molecules condense as refractory dust such as corundum (Al 2 O 3 ) and hibonite (CaAl 12 O 19 ) that are more refractory than silicate dust. Aluminum monoxide may thus not be present in the gas phase in the expanding outflow due to recondensation of refractory dust as discussed for alkali halides (Ginsburg et al. 2019), while SiO molecules sublimated from silicate dust are present in the gas phase in the outflow.
We note that the AlO emissions distribute to a greater extent in the outflow than alkali halides, which are observed only just above the continuum disk (Ginsburg et al. 2019), not significantly exceeding the contour of 0.3 Jy beam −1 for the 497 GHz continuum in Figure 2(a). This implies that AlO and alkali halides behave differently in the outflow, and that alkali halides may condense as solid more effectively than AlO. Gas phase reactions that consume alkali halide molecules or dissociation of alkali halide molecules could also be responsible for their limited distributions (Ginsburg et al. 2019). Asymmetric distribution of AlO is also a notable feature, where the emissions are stronger in the northeast than in the southwest (Figure 2). The asymmetric distribution is suggested for the SiO v=4 line (Kim et al. 2019), but the degree of asymmetry of AlO distribution appears to be larger than SiO, H 2 O, and alkali halides observed in the outflow (e.g., Hirota et al. 2017;Ginsburg et al. 2019). Further investigation, including observation of multiple molecules with different chemical characteristics, is required to understand the chemical processes in the outflow and the origin of the asymmetric distribution of AlO.
The detection of AlO around the central object of Orion Source I will provide further opportunities to study the structure, kinematics, and dynamics of the high-temperature inner region of the circumstellar disk of Orion Source I. The hot nature of the disk around the embedded massive protostar suggests that solid objects that are enriched in refractory elements, similar to the oldest solar system solid, could form in the disk. Further observations of high-temperature metalbearing gas molecules in the disk around YSOs could link the high-temperature CAI formation event occurred in the early solar system to the star formation and subsequent planetary formation processes, which is a key research topic in meteoritics. We note that Orion Source I could be an evolved star in the very early stage of the protoplanetary nebula phase (Báez-Rubio et al. 2018), peculiarly located in the active massive-star formation site. If this is the case, the detection of circumstellar AlO molecules would provide insights into dust formation and possibly the kinematics of mass loss as in the case of W Hya (Takigawa et al. 2017).