X-ray Spectroscopy of Young SNR
Nearly all of the elements in the Universe except Hydrogen and Helium were
generated in stars. Lighter elements are
generated by nuclear fusion in a star's core during its lifetime. Our Sun, for
example, is converting Hydrogen to Helium in its center now, and will
eventually form Carbon and Oxygen, among other things. Elements up to
Iron on the periodic table are created during the main sequence part of
a star's life. Elements that are
heavier than Iron, such as
gold and silver, are created during supernova explosions, when the
tremendous energies released lead to a rash of nucleosynthesis.
These elements do not stay locked up in the centers of stars (otherwise
people and the Earth would not exist), but are dispersed into the ISM
where they eventually coalesce to become nebulae, planets and new stars.
The dispersal is accomplished by the remnants
that expand into the space around an exploded star. In this
way, supernova explosions are the means by which the elements are
redistributed throughout the Galaxy.
Most known supernova remnants are hundreds or even thousands of years old,
so naturally we cannot have observed the stars that blew up to make them.
the relationship between particular types of stars (massive vs. low
mass; these different types of stars have different evolutionary tracks) and particular types of
supernova remnants (SNRs), therefore, as much of astrophysics, is like putting together pieces
of a puzzle until the picture finally emerges. It may be, for example,
that there is a stellar cinder (a neutron star or a pulsar) in the middle
of the remnant, which, we might assume, is the remains of a massive
progenitor which exploded as a Type II SNe. Or the SNR could be found in the middle of an association of
young, massive stars, in which case the progenitor star was very likely
itself a massive star. But this evidence is circumstantial, or indirect,
and it would be nice to have a more direct idea of the composition or
nature of the star that exploded and gave rise to the
remnant that we observe today.
Why Do We Want to Know?
There are billions and billions of stars in the universe, and a
significant portion of these will become supernovae. By studying the
spectra of the remnants the supernovae leave behind we can gain a better
understanding not only of these most powerful events but also of the
nature, makeup and composition of the stars that exploded to form
them. We can learn how SNR interact with the rest of the Galaxy, forming
new nebulae, associations, clouds of enriched gas, and even planets like
The Mighty Spectrum
One tool which can provide such direct evidence is a spectrum (http://heasarc.gsfc.nasa.gov/docs/objects/snrs/puppisa_spectra.html). A spectrum is a measure of the
light emitted by an object (such as a SNR) as a function of wavelength
(or energy - the shorter a light wave's wavelength, the more energy it
It is useful because particular atoms, when excited, emit and/or absorb
particular, fixed wavelengths or energies. The wavelength corresponds to an electron
transition in the atom between distinct energy states. Since the electrons
of atoms can exist at specific energies depending on the type of atom, the
change in energy resulting from an electron moving between two such states is unique
for each element and each compound element, or molecule.
Material heated and compressed by a SNe radiates strongly in the X-rays
because the gas is very energetic and the possible transitions can be
The various gases in the remnant emit light at specific energies;
this leads to narrow features in the spectrum called emission lines. By
identifying what emission "lines" are present in a spectrum, it is possible
to tell what elements are present in the object observed. So by looking at
a SNR's spectrum, it is possible to tell what elements are in the
Spectrum of supernova remnant Cas A (http://heasarc.gsfc.nasa.gov/docs/objects/snrs/casa_spectra.html)
What Was In that Star??
Studying a SNR's spectrum
is the most direct way of determining the composition of the
progenitor star. In young remnants that have not yet swept up a
significant amount of interstellar gas, the imprint of the stellar ejecta
on the remnant's spectrum is strong, so the spectrum can tell us a lot
about the material that made up the progenitor star. The observed elements
and abundances are
very different for different type of SN explosions (massive star vs. white
dwarf). Type Ia remnants (from white dwarfs)
should have relatively
strong Si, S, Ar, Ca, and Fe, and weak O, Ne, and Mg; Type II (from
massive stars) generally
have the reverse pattern.
But wait, there's More...
But it is possible to glean even more information from the mighty spectrum
by fitting the data to
stellar models or to patterns observed in known systems. For example, what
elements and how much of each element are produced by nuclear fusion in a
star's interior is a strong function of mass in Type II remnants. This
fact can be used to try to estimate more closely the mass of the progenitor based on
relative element abundances. In this case, we match the observed abundances
to the relative abundances in models of stars with different progenitor
masses to give an estimate of the
For Type Ia remnants, the nucleosynthesis yield
can depend on the assumed initial composition and on details of the
explosion (e.g., supersonic detonations versus subsonic deflagrations), but
should be quite uniform between different remnants of this class.
Mixing: Another Piece of the Puzzle
In addition to the question of what exactly the ejecta were composed of, spectra
can give us another puzzle piece; that is, how much the stellar
material was mixed up when the star exploded. During its normal lifetime and
during the explosive nucleosynthesis during a SNe, a
star produces heavier and heavier elements in layers, like an onion skin,
with the more processed, heavier, elements at the very center of the star,
and the lighter elements on the outside. If this is the whole story, then
the elements of the remnant should also show an onion-skin like
distribution, just blown outwards.
But other factors may be at work. Evidence
from (spectra of) SN 1987a indicates that convection effectively
mixed the stellar material in the onion-skin layers. This kind of mixing may be quite
common in Type II SNe. On the other hand, it is expected that
in Type Ia remnants, some of the layers will stay stratified, remaining like the distinct
layers of an onion. The outer layers of Si and other
intermediate mass elements are expected to be convectively mixed, while the
innermost layers of Fe are expected to remain separate; this is observed in
the early optical spectra of supernovae of Type Ia.
How Do We Know?
How is it possible to know if the layers in a SNR are mixed or
stratified, since a spectrum gives frequency, rather than spatial
information? An important advance here is in the development of spatially
resolved spectroscopy. For example, satellites such as ASCA can produce
spectra in a wide range of energy for patches of the remnant arcminutes in size. With this information it is
possible to identify the specific locations of features and elements,
rather than being limited to having only one spectrum for the object as a whole.
As well as giving information on the presence of elements, spectra can be
used to calculate the density and the temperature of the gas that generates
a spectral line. In
Tycho's SNR, Iron is seen to be at very different densities and
temperatures than Sulphur and Silicon. This suggests that the
Iron is in one place while the other elements are in a different place: the
material is stratified.
All of the information derived from SNR spectra would be great if it
were clear cut (but it wouldn't be nearly as challenging or as fun to
figure it out!). However, real data is never completely clear cut and
drawing conclusions is never simple. In the case of SNR there is a
problem because the observed spectra are composed of emission lines
plus some broad, usually smoothly varying underlying emission.
This underlying emission, or continuum, could be
thermal emission from gas at a certain temperature. Or it could be
synchrotron emission from energetic cosmic rays
accelerated in magnetic fields. Whatever the underlying continuum
emission is, it must be subtracted from the
observed line emission to get the proper relative strengths of the lines.
Using a thermal continuum model gives different line strengths than
using a synchrotron emission continuum and this can lead to different
conclusions about the composition of the remnant. It is important to
resolve this issue of what processes are responsible for the emission
added to the line emission in order to get the most accurate information
possible from the spectra of SNRs.
Observational Clues: Current and Future Work
In our work, we are studying the composition of the ejecta from its X-ray
emission and looking for elements which show different emitting conditions
such as temperature, pressure or density, suggesting that the element has
been mixed; the same element is found in very different parts of the
remnant. We are primarily using spectral and imaging data from the ASCA
X-ray observatory. ASCA provides high quality X-ray data with good
spectral resolution over a broad energy range from 0.5 to 10 keV, covering
emission lines of O, Ne, Mg, Si, S, Ar, Ca, and Fe. It also provides
spatial information on arcminute scales (Galactic SNR range in size, from a
couple arcminutes to a couple hundred arcminutes, by comparison, so with ASCA it is possible to isolate features such as
separate clumps of stellar gas). We study these sources by fitting models
to the spectrum over a wide range of energies, by comparing the strengths of different emission
lines, and studying images of the remnant in narrow energy bands.
Additionally, new instruments called "micro calorimeters" are being
developed which will be able to resolve individual lines which make up
a single peak with today's spectrometers. With this higher spectral
resolution, the strengths of lines that are much closer together in
wavelength will be able to be used to compute conditions in the remnant
such as temperature and density. It will not matter what the underlying
continuum is since its contribution will be essentially the same for the
lines being considered (because the lines are so close
together). With the help of these new instruments we will be able to say
with greater confidence what the progenitor stars were made of and what
the conditions of the remnant are like now.
Thank you to Una Hwang for contributing to this article.