Hot Luminous Star Research Group

Paul Crowther, Saida Caballero-Nieves, Emile Doran, Chris Rosslowe

ESO La Silla (Chile)
The main interests of this research group is to study massive stars in a variety of environments, and so overlaps with other Sheffield astrophysics research groups involving studies of star clusters and starbursts. Joanne Bibby is involved with studying Wolf-Rayet populations in nearby star forming galaxies (now at UCLan), PhD student Emile Doran is involved with studies of individual massive stars within star clusters, and PhD student Chris Rosslowe is studying Wolf-Rayet stars at infrared wavelengths, while PDRA Saida Caballero-Nieves is studying the content of the R136 star cluster using HST/STIS spectroscopy. We also host a Galactic Wolf-Rayet Catalogue.

Massive Stars

Recent Papers:

Light-weight stars, like our Sun, make up the vast majority of stars in our Galaxy, and live a long and quiet life. In contrast, massive stars are very rare, live short, but intensive lives. Wolf-Rayet stars have powerful winds, up to 1,000 million times stronger than the solar wind that is seen during a total eclipse, with speeds up to one percent of the speed of light. They violently end their life as core-collapse supernovae, and most likely Gamma Ray Bursts, amongst the most powerful events in the universe. Heavy-weight stars also burn much hotter than the Sun, so are capable of fusing elements producing carbon, oxygen and iron. Most chemical elements came from the cores of massive stars, producing the very material our bodies are made of - so we are all composed of stardust!

R136a1 We have reanalysed VLT/SINFONI spectroscopy of the brightest members of the NGC 3603 and R136 star clusters, together with high spatial resolution near-IR imaging, revealing high stellar temperatures and luminosities. These have been combined with contemporary evolutionary models for rotating main-sequence to suggest initial masses as high as 320 solar masses (R136a1), and are supported by dynamical mass determinations for the NGC3603 A1 eclipsing binary system. These findings have been reported by ESO (Stars just got bigger). The artists impression to the right compares the relative sizes of young stars, from red dwarfs (0.1 solar masses), yellow dwarfs (1 solar mass), blue dwarfs (8 solar masses) and R136a1 - see FAQ.

Life cycle of a mega star Massive stars play a dominant role in the ecology of their parent galaxies since their stellar winds inject a great deal of material and energy into their environment. They have received renewed attention in recent years since spectra of high redshift galaxies, witnessed at a time when the universe was only a few billion years old, bear a striking resemblance to nearby starburst galaxies, which themselves show the characteristic wind signatures of hot, O-type stars. However, we are poorly equipped to interpret these important data.

Hypernova Solely hydrogen and helium were formed in the Big Bang, so that all heavier elements were subsequently created through nuclear reactions in stars. Consequently, the heavy element content (or metallicity) of a young galaxy will be much lower than that of the current Milky Way. The metallicity of a galaxy plays a crucial role in massive star evolution, since it defines their internal structure, opacities and stellar wind properties. The precise relation between metallicity and mass-loss, a key ingredient for the reliable population synthesis of young galaxies, remains imprecisely known.

NGC300X1 Our surveys of Wolf-Rayet stars within nearby star-forming galaxies have been compared to deep X-ray images to reveal candidates of rare Wolf-Rayet plus black hole or neutron star systems. Prior to our studies only one Milky Way system was known (Cyg X-3), while we now have confirmed three instances of Wolf-Rayet plus black hole systems (IC10 X-1, NGC300 X-1, M101 ULX-1), whose black hole components are the highest known to date amongst stellar-mass systems. These findings have been reported by NASA for IC10 X-1 (Massive Black Hole Smashes Record), ESO for NGC 300 X-1 (Black Hole Hunters Set New Distance Record) and Gemini for M101 ULX-1 (Fast, Furious, Refined: Smaller Black Holes Can Eat Plenty). The image to the left shows an artists impression of these high-mass X-ray binary systems.

OB stars and Luminous Blue Variables

Recent Papers:

The most direct method of deriving empirical mass-loss rates for hot stars is through analysis of the UV resonance transitions from dominant metal ions. However, difficulties in deriving the wind ionization balance using currently available (trace) ions means that mass-loss rates remain uncertain. With FUSE, additional wind lines are being observed, spanning a much wider range of species. From these data, the degree of ionization can be determined accurately, so that physical and wind properties can be measured with confidence. As part of the VLT FLAMES Massive Star Large Program consortium, notably Rohied Mokiem, I have studied massive stars in the Galaxy and Magellanic Clouds - spanning a factor of 5 in metallicity - so that the variation of mass-loss properties with metal content can be measured, of importance in the study of high redshift galaxies. SMC O stars indeed possess weaker winds than those at Solar composition, with higher temperature for individual O subtypes.

With Danny Lennon and Nolan Walborn I have studied the physical properties of Galactic B stars, revealing good agreement with predicted wind strengths for B0-B0.5 Ia supergiants, after allowance for wind clumping, but winds from B0.7-B3 Ia supergiants are overestimated. The bistability jump in B supergiants - signalling a sudden division in wind to escape velocity - was re-investigated but not confirmed. With Laurent Drissen and Veronique Petit I am monitoring an Luminous Blue Variable star in the Magellanic galaxy NGC 2366 10,000,000 light years away. This star has a B supergiant spectral type except that it is currently undergoing a giant eruption, such that one solar earth mass of material is being ejected every day! Such high and constant mass-loss is exceptional even amongst massive stars, although all are irregular and structured at a basic level. It is believed that such intense episodes of mass-loss are responsible for the removal of hydrogen envelopes in massive stars, causing evolution to the Wolf-Rayet phase.


Wolf-Rayet stars

Recent Papers:

It is widely accepted that OB stars are driven by radiation pressure, whilst the situation for the denser winds of Wolf-Rayet stars is less clear. Recent evidence in support of radiation pressure as a source of driving for WR stars has been derived from studies of LMC and Galactic WC stars. These are the immediate precursors of Type Ic SN and possibly `hypernovae' (with a connection to long duration Gamma Ray Bursts). We find that carbon-rich WC stars stars in the LMC (with early spectral types) have weaker winds than Galactic WC stars (later spectral types). This difference naturally explains the shift in spectral class to earlier type at lower metallicity. The situation for nitrogen-rich WN stars is less straightforward since they show a much wider range of wind properties.

NGC300X1 Zoom CFHT and ESO have been used to quantify the properties of Wolf-Rayet stars in the Spiral Galaxies M33 , NGC 300, NGC1313 and M83 to investigate how their properties are affected by metal content. WR stars in M83, located 10 million light years away, are several billion times fainter than the brightest stars in the Night Sky. Nevertheless, we can use images taken in the light of emission lines and nearby continuum to identify individual massive stars in this galaxy. We host a catalogue of Galactic Wolf-Rayet stars. I have also investigated the WR content of metal poor galaxies IC10 and especially IZw18, one of the most metal-poor galaxies known.

Fundamental properties of stars can most readily be obtained from analysis of eclipsing double-lined eclipsing binaries. We have shown that WR20a in the open cluster Westerlund 2 has the highest masses of any star known to date, with each component of the 4 day binary having a mass in excess of 80 times the Solar mass, each shining a million times brighter. Rarely, Wolf-Rayet stars are members of close binary systems with compact objects. We have identified the only two unambiguous cases of Wolf-Rayet plus black hole binaries from our surveys of IC10 and NGC300 (see figure). These possess the highest stellar-mass black holes known to date, and are plausible progenitors of binary black-holes. See 'Farthest' star-mass black hole report from BBC Science News in Jan 2010.


Young massive clusters

Recent Papers: Super star cluster

Westerlund 1, shown here, is a candidate for the most massive young cluster in the Local Group, with a population of WR stars, Red Supergiants and yellow hypergiants with a central density of 100,000 Solar masses per cubic parsec - see the ESO Press Release, and so is the nearest contender for a young Globular Cluster. A neutron star CXO J164710.2-455216 has recently been detected in Chandra X-ray observations of Wd1 -- see the Chandra Press Release.

We have re-evaluated the initial masses of the most massive stars in the youngest, highest mass Milky Way (NGC 3603) and LMC (R136) clusters, from which initial masses of 170 and 320 solar masses were obtained, together with at least 185 solar masses for the Arches cluster. Monte Carlo simulations suggest that each are consistent with an upper mass limit of circa 300 solar masses, a factor of two higher than hitherto believed. We are now investigating all the high mass stars in R136 with an upcoming Hubble Space Telescope STIS programme (see details). Beyond R136, the central cluster of 30 Doradus, we are undertaking the VLT-FLAMES Tarantula Survey, a multi-epoch optical spectroscopic study of 800 massive stars in the 30 Doradus region.

30 Doradus

We have also obtained Gemini GNIRS spectroscopy of massive stars within the Galactic 1806-20 cluster, from which a downward revision to its distance to 8.7 kpc was obtained (its kinematic distance has hitherto been taken as 15 kpc). This revision is significant since it leads to a factor of three reduction in the 2004 December outburst of the magnetar (highly magnetized pulsar) SGR 1806-20 - providing it is associated with the young massive cluster - which in turn indicates a reduction in the contaminaton of BATSE short GRBs from giant outbursts from magnetars. It is significant that both 1806-20 and Westerlund 1 apparently host magnetars, with progenitor masses of approx 50 Msun.

Starbursts in galaxies

Recent Papers:

NGC3125 Clusters of massive stars are born in `Starbursts' when galaxies collide, as seen in the Antennae galaxies. In an Infrared Space Observatory (ISO) programme I have used mid-IR spectra of WR galaxies to determine their physical and chemical properties and the nature of their massive star populations. We have also provided a recent update -- Starburst02 to the Starburst99 population synthesis code, to take into account metallicity dependent line blanketed model atmospheres for O and WR stars. We have also used ESO telescopes to study the massive stellar content of starburst clusters in NGC 3125 (see HST/ACS figure), for which we have resolved previous UV and optical results through use of common interstellar extinction law.

Finally, we are also investigating the chemical properties of high-redshift Lyman break galaxies, such as MS1512-cB58 and Q1307-BM1163, sampling an epoch when the universe was only a few Gyr old, using the stellar wind signatures from massive OB stars in their integrated spectra. Surprisingly, the metal content of such early galaxies as indicated by oxygen, is already within a factor of a few of the present Milky Way galaxy.

Star Formation

Recent Papers:

UCHII 
regions viewed in the mid-IR
GHII
RCW49 seen with Spitzer GLIMPSE survey
The birth of massive stars remains a major puzzle for theorists and observers alike. What we do know is that they are formed within Giant Molecular Clouds (such as Orion). Once the hot star starts to `shine', it illuminates its dense surrounding region - which is called an UltraCompact HII region - the surrounding material prevents direct observation of the star until the natal gas has been blown away, but one can study the birth environment of massive stars at radio and mid-infrared wavelengths - as shown here for two Galactic UCHII regions, based on observations with the Midcourse Space Experiment. Near-IR spectroscopy from VLT/ISAAC has provided robust classification of the ionizing star of the G23.96+0.15 UCHII region, only the second such case due to high interstellar and circumstellar dut extinction. New observations are also being made with the Spitzer mid-IR telescope, primarily throught the GLIMPSE survey, shown on the right for RCW49, plus our own IRAC and IRS study of the W31 star forming region.

Core-collapse Supernovae

Recent Papers: Association of ccSNe
with HII regions in M51
Core-collapse supernovae involve the death of massive stars, whose lifetimes are 5-50 Myr. As such, they might be expected to be associated with star-forming regions. Historical ground-based surveys (approx 0.1 kpc spatial resolution) identified a majority of type II and Ib/c ccSNe with HII regions, while more recent studies have indicated differences amongst the different ccSNe flavours, with an increasing likelyhood of HII region association for type II, Ib and Ic supernovae. We interpret this difference to the longer lifetime of many type II-P progenitors (up to 50 Myr) than the 10-20 Myr duty cycle of giant HII regions. In contrast, if type Ib/c progenitors have shorter lifetimes (up to 20 Myr), most will be associated with giant HII regions. The accompanying figure shows the immediate environment of recent ccSNe in M51, ranging from no association (SN 2011dh, II-P), association with a faint Orion-like nebula (SN 2005cs, II-P) and association with a bright HII region (SN 1994I, Ic). From inspection of high spatial resolution (approx 10 pc) imaging, a very small fraction of ccSNe are associated with compact HII regions whose duty cycles are much shorter (approx 5 Myr).

CSPN

Recent Papers: IC 4663
In addition to studies of massive stars, some low and intermediate mass stars pass through a phase at the end of their (long) life when a `Planetary Nebula' is being formed, in which late pulses remove surface hydrogen layers, revealing a spectrum reminiscent of massive stars - a subset of these Central Stars of Planetary Nebulae (CSPN) exhibit a Wolf-Rayet spectrum. I have studied a variety of these with Orsola De Marco - e.g. SwSt1.

Although many examples of carbon-sequence Wolf-Rayet CSPNe are known, we have identified the first unambiguous example of a nitrogen-sequence CSPNe, IC 4663. The image of IC4663 to the right shows a greyscale Gemini image ([OIII]) plus a colour HST composite (RGB = Halpha, [OIII], V-band).


______________________ 08-Oct-2012

paul.crowther@shef.ac.uk