[Paper Review] The IMF in Starbursts

Bruce G. Elmegreen|arXiv (Cornell University)|Nov 8, 2004

Stellar, planetary, and galactic studies4 citations

TL;DR

This paper reviews the evolution of understanding regarding the initial mass function (IMF) in starburst regions, concluding that while early studies suggested top-heavy IMFs due to high extinction and luminosity deficits, later observations with improved extinction corrections and multiwavelength modeling show the IMF in starburst galaxies is consistent with the Salpeter slope (Γ = −1.35) and a lower mass cutoff between 0.5 and 1 M⊙. The IMF appears slightly shallower in denser environments, but no strong evidence supports a universal top-heavy IMF in starbursts.

ABSTRACT

The history of the IMF in starburst regions is reviewed. The IMFs are no longer believed to be top-heavy, although some superstar clusters, whether in starburst regions or not, could be. General observations of the IMF are discussed to put the starburst results in perspective. Observed IMF variations seem to suggest that the IMF varies a little with environment in the sense that denser and more massive clusters produce more massive stars, and perhaps more brown dwarfs too, compared to intermediate mass stars.

Motivation & Objective

• Reconcile conflicting early observations suggesting top-heavy IMFs in starburst galaxies with later, more accurate data.
• Assess whether the IMF in starburst regions truly differs from the canonical Salpeter IMF due to extreme conditions.
• Evaluate the role of extinction, dynamical mass, and luminosity measurements in shaping IMF interpretations.
• Investigate whether super star clusters (SSCs) exhibit systematically top-heavy IMFs compared to normal clusters.
• Determine if IMF variations correlate with star formation density or environmental conditions in dense stellar systems.

Proposed method

• Reviewed observational data from multiple starburst galaxies (e.g., M82, NGC 3256, UGC 8387) using multiwavelength photometry and spectroscopy (K-band, Brα/Brγ, FIR, radio, CO, HeI lines).
• Applied evolutionary population synthesis models to match observed emission line ratios, luminosities, and spectral energy distributions to different IMF assumptions.
• Used dynamical mass estimates from rotation curves and velocity dispersions to constrain total stellar mass and compare with luminous mass.
• Evaluated extinction corrections using Brγ/Brα and Paβ/Brγ ratios, challenging earlier high extinction values (e.g., AV = 25 mag) in M82.
• Analyzed IMF slopes (Γ) and lower mass cutoffs (ML) by fitting observed L/M ratios, Lyman continuum fluxes, and [NeIII]/[NeII] line ratios.
• Compared results across diverse environments—from local Galactic clusters to high-redshift starbursts—using a consistent IMF framework.

Experimental results

Research questions

• RQ1 Did early observations of starburst galaxies like M82 truly indicate a top-heavy IMF, or were high extinction estimates and luminosity biases responsible?
• RQ2 To what extent do dynamical mass, luminosity, and extinction measurements affect IMF inference in starburst regions?
• RQ3 Is there a systematic environmental dependence of the IMF, particularly in terms of IMF slope or lower mass cutoff in dense star-forming regions?
• RQ4 Do super star clusters (SSCs) exhibit systematically different IMFs compared to normal clusters, and if so, what physical mechanisms could explain this?
• RQ5 Can the observed lack of red giant populations and elevated oxygen abundances in starburst regions rule out top-heavy IMFs with inner truncations?

Key findings

• Revised extinction estimates in M82 (AV ≈ 2–12 mag) from Paβ/Brγ ratios contradict earlier high values (AV = 25 mag), reducing the need for IMF truncation.
• The K-band luminosity in M82 is ~3 times lower than previously estimated, further weakening the case for a truncated IMF.
• Multiwavelength modeling of 19 starburst galaxies supports a Salpeter IMF (Γ = −1.35) from 0.1 to 100 M⊙, with no evidence for top-heaviness.
• High-resolution spectroscopy of M82 at 25 pc resolution shows no need for an upper mass limit >50 M⊙, and suggests an IMF turnover below 1 M⊙.
• Observations of super star clusters (e.g., MGG-11 in M82, NGC 1705-1) show some evidence for top-heavy IMFs, but others (e.g., NGC 6946, M82: MGG-9) have normal Salpeter-like IMFs.
• A trend is observed where denser star-forming regions (e.g., starbursts) have slightly shallower IMF slopes (Γ ≈ −1.35) compared to less dense regions (Γ ≈ −1.8), suggesting environmental dependence.