Recent Projects (as of March 2017)

I have been working on several projects to study star formation in the Central Molecular Zone (CMZ) of the Galaxy. The most recent one is to use interferometer (SMA/JVLA) + single-dish (CSO/APEX/GBT) spectral line observations to characterize the molecular gas environment of star formation in the 20 km s-1 cloud (for the star formation in this cloud, see below). We focus on two impacts that star formation may have on the environment: chemistry, and temperature. In the end we find that at 0.1-pc scales in this cloud, star formation and shocks may together change the chemical compositions of molecular gas as well as heat the gas. Note that here the shocks are not related to star formation, but induced by pc-scale dynamic processes (orbital motions around the Galactic center, cloud-cloud collision, etc.). The paper has been submitted to AAS journals.

The plots below show two aspects of the impact on the environment: chemistry (enhancement of shock tracers), and temperature (signatures of shock/protostellar heating).

Shock tracers in the 20 km/s cloud
Above: enhancement of 'slow' shock tracers relative to the 'fast' shock trace SiO toward the massive clumps. This may suggest slow shocks or hot moelcular cores in the massive clumps.
Temperatures in the 20 km/s cloud
Above: ratios of kinetic temperatures derived from NH3 (2,2)/(4,4) lines and H2CO 303-202/321-220 lines. The large ratios may suggest two gas components, while in several dense cores signatures of internal heating are found.

The other project I am working on is a JVLA C-band survey of the CMZ. The 6.7 GHz (class II) CH3OH maser, a radio recombination line (RRL), several H2CO lines, as well as continuum were observed. The objective is to search for high-mass star formation using the class II CH3OH maser and the RRL. The plot below shows the coverage of the observation (large blue circles). We have obtained the data, and the analysis is in progress.

JVLA Cband Survey of the CMZ


Research in PhD

High-mass stars (M ≥ 8 M) dominate the energy output in star clusters through energetic stellar winds, and produce heavy elements that are essential for life when dying as supernovae. They often form in a clustered environment in dense molecular clumps, which are often opaque in optical, sometimes even in infrared bands. To study their physical properties, we can make use of radio and millimeter/submillimeter spectral lines and continuum, which can penetrate the dense gas around the massive protostars. Such observations have proved that high-mass stars could form in a similar way with their low-mass counterparts, through accretion disks and bipolar outflows.

In spite of these advances, there are still many unsettled questions. I am particularly interested in three of them:

  • the relation between high-mass star formation and filamentary structures in molecular clouds;
  • the initial fragmentation in massive molecular clouds that leads to high-mass protostellar cores;
  • and star formation in the extreme physical conditions in the central molecular zone (CMZ), the inner 500 pc of the Galactic Center where massive, dense molecular gas resides.

To address these questions, I used observations from e.g., the Very Large Array (VLA), the Submillimeter Array (SMA), and the Green Bank Telescope (GBT), to study high-mass star formation in dense cores embedded in Galactic molecular clouds, as well as in massive clouds in the CMZ. The goal is to find out how dense gas in filaments is accreted into cores thus fuel star formation, how the fragmentation shapes dense core population within IRDCs, and how the extreme physical conditions in the GC affect star formation. The results have been published in three papers (Lu et al. 2014, 2015a,b)

I. Multi-Line Study of Filaments in High-mass Star Forming Regions

SMA observations of Filaments

In my earlier work (Lu et al. 2014), I compiled a sample of 62 high-mass star formation regions, and obtained high angular resolution NH3 (J,K)=(1,1), (2,2) spectral lines data from the VLA, which were then used to derive rotation temperatures, column densities, and masses of dense cores. When I analyzed the data, I noticed ubiquitous filamentary structures in the NH3 emission. Some are spatially coincident with infrared extinction features in Spitzer images, suggesting their nature as IRDCs, while others are associated with star formation tracers such as masers or Hii regions. These sources thus cover a wide range of evolutionary phases. Similar filamentary structures have been found in both nearby clouds (Bally et al. 1987; Hacar et al. 2013) and distant massive clouds (Pillai et al. 2006; Jackson et al. 2010).

To investigate the relation between high-mass star formation and filaments, I selected 9 filamentary sources, including both IRDCs and more evolved clouds, and made follow-up studies in two steps. First, I used the SMA sub-millimeter continuum and spectral line observations, to confirm the high-mass star formation in these sources. The continuum emission was used to locate dense cores and the molecular lines were used to assess protostellar nature of the cores. Second, I conducted 3 mm spectral line observations using the IRAM 30m telescope to derive the gas infall rates along filaments, to understand whether and how filaments promote high-mass star formation.

As an example, a figure above presents preliminary results of the SMA observations toward a filamentary IRDC, IRAS 18308-0841. It presents three filaments that are all coincident with infrared extinction features. The velocities of the NH3 line clearly reveal gradients along the filaments, probably tracing convergent motions into the center of the source, or the 'hub'. The SMA 1.3 mm continuum observation resolved three dense cores in the hub, and one in the filament. The most massive one of the three cores in the hub is >100 M. Collimated 12CO and SiO bipolar outflows, as well as molecular lines such as CH3CN, DCN, and OCS that are usually found in hot molecular cores, were detected in this core. Therefore, protostars have emerged in the hub, and they are likely high-mass given the large mass of the core and the energetic outflows.

For the IRAM 30m telescope observation, I obtained HNC/HCO+ 1-0 lines, both of which are optically thick and usually present blue-asymmetry features, which could trace infall motions (Chira et al. 2014). I also included an optically thin line, N2H+ 1-0, to define the cloud velocities. Signatures of infall (blue-asymmetries) have been found in the raw data. Once I finish further reduction of the data, I will estimate the infall velocities based on analytical models (e.g., Myers et al. 1996) and derive the infall rates.


II. Initial Fragmentation in Massive Clumps in G28.53

Overview of G28.53-0.25
Above: An overview of the IRDC G28.53. Left: Spitzer and Herschel composite image overlaid with 1.2 mm continuum emission taken from Rathborne et al. (2006). The colors represent emission at 8.0 μm (blue), 24 μm (green), and 70 μm (red). The contours levels are 30 (3σ), 60, 90, 120, 180, 240 mJy/beam. The four clumps, MM1, MM5, MM7, and MM8 are marked. The green circles mark the most massive cores in each clump. The four crosses mark the H2O masers. Right: Rotation temperature map using the combined GBT and VLA NH3 data, overlaid with contours of the integrated intensities of the NH3 (1,1) main hyperfine line. The contours are in steps of 5 mJy beam-1 km s-1×[3, 6, 9, 12, 15, 20, 25]. The dash circles represent the 2' primary beams of the VLA observations. ‘Core 2’ in MM1 (Swift 2009) is marked with a star. The four H2O masers are marked with crosses and labelled as W1-W4.

To study the impact of turbulence in the initial fragmentation and core growth in IRDCs, I selected four clumps in an IRDC G28.53-0.25 (G28.53 hereafter). G28.53 has a kinematic distance of 5.4 kpc (Rathborne et al. 2006) and a luminosity of ~3500 L (Rathborne et al. 2010). It has been mapped in 1.2 mm continuum with the IRAM 30m (figure below; Rathborne et al. 2006), which reveals a total mass of ~104 M. Ten continuum peaks are identified, each of which is a few hundreds of M and of ~1 pc scales, therefore they are typical clumps that form massive stars.

In order to reveal what controls the fragmentation in these clumps, I used high angular resolution NH3 line observations from VLA and GBT to trace kinematics and determine gas temperatures (see figure above). NH3 itself is also useful for tracing dense gas with a critical density of ~104 cm-3 (Swade 1989). I also used 1.3 mm continuum and spectral lines from the SMA to resolve 0.1 pc scale cores in the four clumps and to trace star formation. The results were published in Lu et al. (2015).

The SMA observations of the four clumps revealed twelve 0.1 pc scale cores, as shown in the figure below. I derived core properties, including NH3 rotation temperatures, core masses based on mm (dust) emission, Jeans masses of the clumps, and virial masses of the cores.

SMA observations of G28.53-0.25
Above: The SMA 1.3 mm continuum emission of MM1, MM5 and MM7/8. The dashed circles in each panel mark the 56" primary beams of the SMA. The ellipses mark FWHM of the 2D gaussians fitted to each core. (a) The contours are in steps of 1.1 mJy beam-1×[-5, -3, 3, 5, 7, 9, 15, 25, 35, 45]. The two H2O masers, W1 and W4, are marked with crosses. (b) The contours are in steps of 0.9 mJy beam-1×[-3, 3, 5, 7]. (c) The contours are in steps of 0.81 mJy beam-1×[-3, 3, 5, 7].

I found that the MM1 clump is likely forming several high-mass stars. The massive, gravitationally bound cores and the energetic outflows suggest that MM1 might be a progenitor of high-mass star cluster. In contrast, MM5 clump is intermediate and MM7/8 clumps are quiescent in term of evolutionary phases. However they will continue fragmenting and collapsing until star formation starts.

After comparing the Jeans masses (using either the turbulent line widths or the thermal line widths based on NH3 temperatures) and the molecular masses of the cores, I found that when these 1 pc scale clumps fragment into 0.1 pc scale cores, turbulent pressure is more important than thermal pressure in supporting the gas, which makes it possible to form super-thermal-Jeans-mass cores that are massive enough to produce high-mass stars.


III. Deeply Embedded Protostellar Population in the CMZ

Overview of CMZ clouds
Above: An overview of the SMA 1.3 mm continuum and CH3CN emission of the six CMZ clouds. The upper panel shows a Spitzer three-color image of the CMZ (red:24 μm; green:8 μm; blue:3.6 μm). The first row of the lower panels shows the 1.3 mm continuum (dust) emission. Color scale starts at 3σ=10 mJy/beam level and is unit of mJy/beam. The second row shows the integrated CH3CN emission. Color scale starts at 3σ=0.4 Jy/beam level and is unit of Jy/beam×kms−1. The dotted regions are the FWHP field of the SMA. The protostellar core candidates are marked by red dashed circles.

Unlike molecular clouds in the Galactic disk, clouds in the CMZ are subject to strong radiation and gravitational perturbation from the supermassive black hole and violent dynamics. The present-day star formation rate per unit dense gas is at least an order of magnitude lower than what the Kennicutt-Schmidt star formation relation predicts (Kauffmann et al. 2013; Longmore et al. 2013). Is the star formation in the CMZ working in a periodical way, regulated by the strength of turbulence, as modeled by Kruijssen et al. (2014, 2015)? Could the apparently quiescent clouds form high-mass star clusters similar to the Arches or the Quintuplet in the future?

To investigate these possibilities, I worked with collaborators to conduct a survey of six massive CMZ clouds (see figure above) using the SMA in the 230 GHz band, to obtain the 1.3 mm continuum and molecular lines such as CH3CN, SiO, 12CO, 13CO, C18O, and H2CO, all of which will be used to constrain physical properties of dense cores. In addition, we surveyed the six clouds using the JVLA, in order to obtain NH3 (1,1) to (5,5) transitions, H2O maser and 1.3 cm continuum simultaneously, all of which will be used to locate embedded protostars.

Surprisingly, we found a large population of H2O masers in these six clouds, most of which are coincident with the SMA dust peaks. We also detected CH3CN emission towards some dust peaks, which usually traces hot molecular cores. These cores are the first protostellar candidates found in the CMZ outside of Sgr B2/C/D and the circumnuclear disk.

Overview of CMZ clouds
Above: Left: H2O masers in the 20 km s-1 cloud. The red crosses (W6, W18) are known OH/IR stars. The blue crosses show H2O masers that have no known OH/IR star counterparts. The contours (ranging from 5σ to 55σ in steps of 10σ) and color scale show the SMA 1.3 mm continuum and integrated CH3CN emission, respectively. The dotted loop shows FWHM of the SMA primary beams and the dashed loop shows that of the VLA primary beams. Right: the spectra of the 18 H2O masers. Most have fluxes <0.5 Jy thus are undetectable by previous surveys, e.g., HOPS.

As an example, the figure above presents preliminary results of VLA and SMA observations toward the 20 km s-1 cloud. We detected 18 H2O masers, among which two are known OH/IR stars, while velocities of the other 16 ones are within -40 to 50 km s-1, consistent with the cloud velocity in general. The SMA dust emission reveals five aligned massive clumps, labelled as C1 to C5. Compact CH3CN emission is detected toward C1/C4/C5. The H2O maser and CH3CN emission likely trace a population of protostellar cores at very earlier evolutionary phases. This newly found protostellar population might suggest increasing level of star formation in the last ~104 years in this cloud. A paper on these results has been submitted to ApJ Letters.

The SMA observations toward the six massive CMZ clouds have become the basis of an SMA large scale project The SMA Legacy Survey of the CMZ (PIs: C. Battersby & E. Keto), in which we requested to survey a 240 arcmin2 area in the CMZ above a column density threshold of 1023 cm-2. As a co-I of the project, I helped prepare observing scripts, reduced data, and helped improve the calibration and imaging strategies.


Papers submitted

  1. "The Molecular Gas Environment in the 20 km s-1 Cloud in the Central Molecular Zone",
    Lu, X. et al., submitted to AAS journals.

Publications in Refereed Journals

  1. "Deeply Embedded Protostellar Population in the 20 km s-1 Cloud of the Central Molecular Zone",
    Lu, X., Zhang, Q., Kauffmann, J., Pillai, T., Longmore, S. N., Kruijssen, J. M. D., Battersby, C., & Gu, Q. 2015, ApJL, 814, L18. IOP link
  2. "Initial Fragmentation in the Infrared Dark Cloud G28.53-0.25",
    Lu, X., Zhang, Q., Wang, K., & Gu, Q. 2015, ApJ, 805, 171. ADS link
  3. "Fragmentation of Molecular Clumps and Formation of Protocluster",
    Zhang, Q., Wang, K., Lu, X., & Jiménez-Serra, I. 2015, ApJ, 804, 141. ADS link
  4. "Very Large Array Observations of Ammonia in High-mass Star Formation Regions",
    Lu, X., Zhang, Q., Liu, H. B., Wang, J., & Gu, Q. 2014, ApJ, 790, 84. ADS link

Conference Contributions

  1. Intro talk: "An SMA/VLA Mini-survey of Six Massive CMZ Clouds: Searching for 'Hidden' Protostellar Population",
    Lu, X., Zhang, Q., Kauffmann, J., & Pillai, T. 2015, Harvard-Heidelberg Star Formation Workshops. Web link
  2. Talk: "SMA and VLA Observations of Dense Cores at Different Evolutionary Phases in Filamentary IRDCs",
    Lu, X. & Zhang, Q. 2014, Workshop on Dense Cores: Origin, Evolution, and Collapse, AAS Topical Conference Series. ADS link
  3. Talk: "Molecular Spectral Lines in Filamentary Infrared Dark Clouds",
    Lu, X., Zhang, Q., & Liu, H. B. 2014, 69th International Symposium on Molecular Spectroscopy. ADS link
  4. Poster: "Revealing Initial Conditions of High-mass Star Formation in IRDCs with the SMA",
    Lu, X., Zhang, Q., & Liu, H. B. 2014, The Submillimeter Array: First Decade of Discovery. Web link
  5. Talk: "Gas Kinematics in Filamentary Infrared Dark Clouds",
    Lu, X., Zhang, Q., & Liu, H. B. 2014, AAS 224. ADS link
  6. Poster: "SMA Observations towards Massive Clouds in the Central Molecular Zone",
    Lu, X., Zhang, Q., Kauffmann, J., & Pillai, T. 2013, IAU Symposium 303. ADS link

Successful PI Regular Proposals

  • SMA, 2015B, "High-mass Star Formation in Dense Cores Embedded in Filaments"
  • SMA, 2015A, "Deeply Embedded Protostars in the Central Molecular Zone"
  • IRAM 30m, 2014 winter, "Filamentary Structure, Infall Convergent Flow and Massive Star Formation"
  • SMA, 2014B, "Massive Star Formation in Progress in Filamentary Clouds"
  • SMA, 2013B, "Sgr B2: A Star-forming Cloud in the Central Molecular Zone"
  • SMA, 2013B, "Gas Kinematics in Filamentary Infrared Dark Clouds"
  • SMA, 2013A, "High-mass Clouds in the Central Molecular Zone"
  • SMA, 2012B, "Gas Kinematics and Condensations in Filamentary Infrared Dark Clouds"

Participated Large Scale Proposals

Observing Experience

  • SMT, remote observing, Nanjing, China, November 2015
  • SMA, on-site observing, Mauna Kea, HI, USA, September 2014
  • CSO, remote observing, Cambridge, MA, USA, April 2014
  • CARMA, on-site observing, Big Pine, CA, USA, July 2012 (during CARMA summer school)
  • SMA, on-site observing, Mauna Kea, HI, USA, June 2012
  • DLH 13.7m telescope, on-site observing, Delingha, Qinghai, China, January 2012