An interview with Guido Garay

Interview by Bo Reipurth, SFN #342 - June 2021

When did you decide to be an astronomer, what was your thesis about, and who was your advisor?

I decided to become a radio-astronomer early in my life, as a high school student, when my math teacher told the class about the inauguration in 1963 of the Arecibo radio-telescope, which opened the observations to a new kind of cosmic objects. I was fascinated with the idea of studying a universe hidden from the eye. Years later, after graduating with a bachelor’s in science at Universidad de Chile, I continued my original path by moving to Harvard to perform doctoral studies. I approached Jim Moran for a thesis project, and he proposed me three topics related to the early stages of star formation: (i) To probe the heart of the Orion Nebula in order to investigate the nature of its embedded IR sources, in particular, the Becklin-Neugebauer object; (ii) To determine the dynamical relationship between OH masers and compact regions of ionized gas; and (iii) To determine the nature of the strong (> 10⁶ Janskys) and variable water-vapor maser emission towards the Orion nebula by monitoring its polarization characteristics. The first two projects required making observations using the Very Large Array, which in the early 80s was the most powerful radio interferometer and had just started operations. Being worried about the possibility of not obtaining observing time, I decided to propose to my thesis committee, composed of several distinguished astronomers such as Alexander Dalgarno, Giovanni Fazio, Paul Ho, and Mark Reid, to carry out one of these three projects. They all agreed that any one of the proposed projects would be good for a thesis. Amazingly, I got observing time to undertake all three projects, which of course was good, but also meant that I had to face a lot of work. Under the great guidance of two of the world's foremost experts in the field of interferometry, Jim Moran and Mark Reid, who acted as co-adviser, I was able to carry out the projects successfully and finish them in a reasonable amount of time.

In a paper from 1985 you, with Mark Reid and Jim Moran, presented a model of OH masers associated with compact HII regions, in which the masers are part of an infalling envelope. How has this concept held up with time?

With the goal of determining the dynamical relationship between compact HII regions and OH maser emission (one of the topics of my thesis), we used the Very Large Array to observe the emission in radio recombination lines from the ionized gas, which allowed us to determine its line center velocity and to map the morphology and velocity field of the OH maser emission. We found that the OH maser emission arises from spots that are projected on the face and located in front of the CHII region and that their radial velocities are redshifted with respect to that of the ionized gas. These results are simply explained if the OH maser spots are part of an in-falling molecular envelope that is still collapsing towards the newly formed star. This scenario cannot, however, be extrapolated to all CHII regions, particularly to those in the later stages in which the expansion of the ionized gas has started and is likely to revert the motions of the surrounding envelope. The concept of an in-falling envelope has been recently revitalized both theoretically, by the global collapse model for massive star formation by Vazquez-Semadeni and collaborators, and by observations of a large number of massive star-forming clumps (e.g. the MALT90 survey) which shows that 31% of the clumps associated with CHII regions exhibit infall motions.

Around the same time, you did a study of compact radio continuum sources in the Orion Nebula and found 21 radio sources. What were your findings, and what is known about these sources today?

In this work, we reported multi-epoch multi-wavelength observations of the Orion Nebula, made with the VLA, in order to characterize the spectral index of the radio emission from the compact radio continuum sources we found within the heart of the Nebula, which could indicate whether they correspond to either ultra-compact HII regions, indicative of embedded OB stars, or winds associated with pre-main sequence, low-mass stars. The initial observation, made as part of my Ph.D. thesis, was designed to search for compact radio sources which would help to establish which of the infrared sources in the molecular cloud might contain primary energy sources. In addition to detecting radio emission from the Becklin- Neugebauer object, we serendipitously discovered 20 other radio sources. I still remember the excitement when after a thorough and detailed process of data reduction at the VLA, guided by Jim Moran, the radio images showed the presence of 21 radio sources clustered within 30'' from Θ¹C. These sources made up what we called the Orion radio zoo, which contained DEERS (Deeply Embedded Energetic Radio Sources), FOXES (Fluctuating Optical and X-ray Emitting Sources) corresponding to pre-main sequence low-mass stars, and PIGS (Partially Ionized Globules), namely neutral condensations surrounded by ionized envelopes that are externally excited by Θ¹C. Among the DEERS is the BN object, for which we determined that its radio emission is free-free emission from a dense, compact, ionized envelope surrounding and internally excited by an early B-type star. About six months later, Ed Churchwell and collaborators published VLA observations of the Orion Nebula, suggesting an alternative model for the PIGS in which the radio emission arises from evaporating protostellar disks around low-mass (~ 1 M_ⵙ) stars. In both cases, the ionization is external, produced by UV radiation from Θ¹C. Later HST observations clearly established that these objects are protoplanetary neutral disks surrounding low mass pre-main sequence stars and externally ionized by Θ¹C.

In 1993 you and your collaborators presented a VLA study of compact HII regions with IRAS counterparts.

As I have mentioned, most of my early research was done using the VLA. Before answering your question, please let me use this forum to thank NRAO for its policy of open access to its facilities to all astronomers in the world. I benefited enormously from this policy. In the work you mention, done with Moran, Rodríguez, and Churchwell, we made multi-frequency radio continuum observations towards 16 luminous IRAS point sources with far-infrared colors indicating that they contained an embedded high-mass star (the famous Wood & Churchwell color criteria). The observations were designed to determine the nature of the exciting source (or sources) of the compact region of ionized gas. We found that in about half of the IRAS point sources, the morphology of the radio emission is simple (typically shell or cometary) consistent with the compact HII region being excited by a single star. The other half shows complex radio morphologies, which can be decomposed into multiple components. We found that the average infrared luminosity of the complex sources is 3 times larger than that of the simple sources, indicating that the complex regions are excited by a cluster of stars. In addition, from the observed infrared and radio properties, we concluded that these clusters contain only massive stars (top-heavy IMF). We also showed that typically ~55% of the ionizing stellar photons are absorbed by dust within the HII region.

In 1996 you observed with the VLA the Cepheus A region, an active region of star formation containing H₂O and OH masers, IR sources, and outflows. What did you find?

In this work, we reported multi-frequency, matching-beam, VLA continuum observations of the Cep A East star-forming region, which contains 16 compact (~ 1'') radio sources clustered within a 25'' radius region, most of which are aligned in string-like structures. We found that the spectral indices of the emission from these compact objects cover a wide range, from -0.6 to 0.7. Four sources exhibit positive spectral indices, elongated morphologies, and angular-size and flux-density dependences with frequency that suggest they correspond to confined jets of ionized gas. Most of the objects that appear in string structures exhibit a mixture of flat and negative spectral indices across their faces, which indicate the presence of both thermal and non-thermal emission. We suggested that the radio emission from the string sources arises in shocks resulting from the interaction of confined stellar winds with the surrounding medium. The duality in emission mechanisms is expected in shock waves where a small fraction of the electrons are accelerated to relativistic velocities, giving rise to non-thermal emission, while most of the electrons produce thermal free-free emission. Our interpretation was radically different from that reached by Hughes & Wouterloot a couple of years earlier, who proposed that the emission from the radio sources in the strings correspond to thermal emission excited by about a dozen of very young B3 type stars. Our new data did not support this thesis.

Later you studied molecular enhancements associated with the famous BHR 71 outflow. What did you learn?

The remarkable BHR 71 outflow was discovered independently by Tyler Bourke and by me. In one of my visits to CfA, I stopped at Tyler's office, a predoctoral CfA fellow, to talk about his research. He mentioned to me, with a lot of excitement, that he was analyzing data taken with the SEST (Swedish-ESO Submillimetre Telescope) of a highly collimated, very extended, bipolar outflow, he discovered (but not published) within the Southern Bok globule BHR71. As soon as he showed me the images, I realized that it was the same outflow I had discovered (but not published) with the Columbia Telescope in La Serena, Chile and that I was also observing with SEST. After the initial shocking surprise for both of us that we were observing the same object with the same telescope, we decided to combine our data and published the discovery paper in 1997. The subsequent step was to investigate the molecular enhancement in the lobes, mapping, using SEST, the line emission in transitions of silicon monoxide, carbon monosulfide, methanol, and formyl cation over several arcminutes. We found that the abundance of methanol and silicon monoxide in the outflow lobes is enhanced with respect to that of the ambient cloud by factors of up to ~40 and 350, respectively. We concluded that the large enhancements are most likely due to the release from grains of ice mantles and Si-bearing species via shocks produced by the interaction between the outflow and dense ambient gas. On the other hand, we found that the abundance of HCO⁺ in the outflowing gas is smaller than that in the ambient gas by about a factor of 20, a decrease consistent with theoretical predictions of shock models.

You and Susana Lizano wrote in 1999 a large review on young massive stars. What were the key questions back then, and how has the field developed since then?

In this paper, we reviewed the physical characteristics and kinematics of objects intimately associated with the formation and early evolution of high-mass stars, such as compact regions of ionized gas and dense and hot molecular cores, that were being gathered at that time with high angular resolution observations. The key question at that time was whether massive stars were formed by an accretion process, similar to that of low mass stars, or by the merger of low mass stars. We compiled the evidence for the existence of hot and very dense molecular structures undergoing large mass accretion rates, attesting that during the process of collapse of high-mass stars massive disks are formed. We also showed that the bipolar outflow phenomenon is commonly seen toward young high-mass stellar objects. These results indicate that the formation of massive stars from massive prestellar cores shares similar characteristics with those of low-mass stars. During the last twenty years, our knowledge of the high-mass star formation process has greatly increased through a combination of studies of the phenomena covering a wide range of physical scales (1 to 0.01 pc). Emphasis has been placed, both theoretically and observationally, on understanding the formation and evolution of cluster of condensations that will eventually give rise to high-mass stars. Observations at scales of pc have shown that the maternities of high-mass stars correspond to distinctive molecular structures with masses of ~10³ M_ⵙ, densities of ~10⁴ cm^-3, and sizes of ~0.6 pc. However, the evolution of starless massive and dense clumps, also known as MDCs, to give rise to massive and dense cores remains debatable. Proposed scenarios are: (i) starless MDCs are supported by a high degree of micro-turbulence leading to the production of cores with a large distribution in mass. The most massive cores (100 M_ⵙ) contract quasi-statically to become high-mass pre-stellar cores before becoming proto-stellar cores. This model referred to as a core-fed scenario, predicts the existence of one, or a few, high-mass pre-stellar cores in the starless MDCs; (ii) starless MDCs fragment into a cluster of low-mass cores with initial masses of the order of the Jeans mass. The low-mass pre-stellar cores accrete gas from their surroundings and those located near the center of the gas reservoir accretes at a higher rate to eventually become high-mass protostars. In this view high-mass pre-stellar cores do not develop; (iii) the global hierarchical collapse scenario. In this view high-mass stars form into infalling massive dense clumps at 1 pc scales, whose global collapse drives inflowing gas streams toward protostars at 0.01-pc scales. During the last decade, several ALMA projects have been performed with the aim of identifying pre-stellar cores and protostars within MDCs. So far very few high-mass pre-stellar core candidates have been reported. The lack of cores with masses greater than 30 M_ⵙ supports clump-fed rather than core-fed scenarios. In addition, the large incidence of blueshifted profiles among MDCs strongly supports the global collapse scenario. At the smallest physical scales (~10³ - 10⁴ AU), ALMA dust polarization observations are adding significant insight for a better understanding of the role of magnetic fields in the process of high-mass star formation, providing unprecedented images of the magnetic field morphology. In a few cases, the magnetic field vectors clearly show the expected hourglass-shaped morphology, while in other cases they closely follow the morphology of infalling and rotating spiral molecular structures.

A glance to your refereed publications shows that you have worked with astronomers from all over the world.

I have been extremely lucky to collaborate, during my carrier, with a long list of astronomers, from whom I have learned a lot in stimulating discussions. Besides Jim Moran and Mark Reid, my mentors, my most appreciated collaborators are Luis Felipe Rodríguez, always a source of wisdom, knowledge, and inspiration in all aspects of life, Susana Lizano, my favorite theoretician, Yolanda Gomez, a loving, intelligent, and generous scientist, and Chema Torrelles, a warm and committed human being. I would also like to mention two researchers from whom I have benefited not only from the exchange of ideas but also from the support they have given me continuously: Ed Churchwell and Karl Menten. Finally, all my students and postdoctoral fellows have been a continuous source of fresh ideas and challenging questions which have made me rethink the problems under analysis more deeply.