The physics and chemistry of star formation is played
out in dense interstellar clouds, which are the nurseries
of new stars, together with their attendant proto-planetary
disks, from which planets condense. The discipline that
considers the formation, interaction, and destruction of
molecules in space is astrochemistry. Its importance has
increased considerably with the explosive growth of
astrobiology and the attendant quest to understand the
molecular origin of life in the universe. The use of
astrochemical models to interpret molecular line
observations is widespread.
A Special Topics examination of astrochemistry research over the past
decade highlights the contributions of Professor Tom Millar, who ranks at
#1 by cites/paper, #3 by cites, and #9 by number of papers, with 14 papers
cited a total of 534 times. Two of these papers rank at #2 and #4
respectively on our list of the 20 most-cited papers in the field.
Professor Millar is Dean of the Faculty of Engineering and Physical
Sciences and Professor of Astrophysics in the School of Mathematics and
Physics at Queen’s University, Belfast, in Northern Ireland.
In the interview below, Dr. Simon
Mitton, ScienceWatch.com's European
correspondent, interviewed Professor Millar to find out
more about his high-ranking papers related to
I picture chemistry as an industrial
enterprise, with its intellectual puzzles normally being pursued in
laboratories. But you are an exception to this simplistic view because
your interests are the chemical reactions that take place in space, in
environments where the physical conditions, particularly densities,
temperatures, and time scales could scarcely be more different than in
the laboratory. How did you get into such an esoteric field?
I attended the University of Manchester Institute of Science and Technology
(UMIST) where I read mathematics, staying on to do a Ph.D. with David
Williams in astrophysics. In practice though, my doctoral research was in
the field we now term astrochemistry; I was probably the first person to
get a Ph.D. in astrochemistry from a UK university.
My thesis looked at molecule formation in interstellar clouds. Back in the
mid 1970s, new interstellar molecules were first being discovered through
techniques of microwave radio astronomy. Everyone got tremendously excited
by the discoveries of ammonia (NH3) and formaldehyde
(H2CO), the latter a central building block in the synthesis of
many other compounds.
I well remember that excitement: I was a graduate
student in radio astronomy at Cambridge, although my interests were
certainly not on the molecular scale because I was observing quasars
and radio galaxies.
"What my research is all
about is trying to use molecules as tracers of the
history of molecular clouds."
It was indeed a most stimulating time, and I started working on models of
how those interstellar molecules could form. I did some work on
investigating the importance of various gas-phase mechanisms on molecule
formation. A big chunk of my thesis was on numerical modeling of cloud
collapse with molecule formation included. By the time I finished I felt I
knew as much as one could then about molecules and gaseous chemistry, but I
knew very little about interstellar dust, which is important in
astrochemistry because molecules freeze onto the dust at very low
I went off to be a postdoc in Toronto with Walt Duley, who had done a lot
of work on interstellar dust; I wanted to learn from him how to study dust,
particularly its optical and physical properties. I was interested in how
dust could protect molecules from strong radiation, which would destroy
them, and I was also interested in surface chemistry. From Toronto I went
to Oxford as a postdoc, thence to an appointment as a lecturer in
mathematics at UMIST, before moving to Queen’s University here in
What are your current interests? I note that you
are currently Chair of the International Board of the James Clerk
Maxwell Telescope (JCMT) in Hawaii and that you are serving as
President of Division VI (Interstellar Matter) of the International
The underlying theme that interests me is the use of chemistry of molecules
as probes of the physical conditions in interstellar space. To me
that’s more than a question of using molecules as proxies to get
temperatures, densities, the strength of magnetic fields and so on. What my
research is all about is trying to use molecules as tracers of the history
of molecular clouds.
Are giant molecular clouds, of a million solar
masses or more, the most important sites for star formation?
They certainly are. An important aim at which I have been working hard to
understand for several years is: what are the best molecular probes of the
star formation process? One of the big problems in star formation is the
transition from an interstellar cloud that has relatively low density and
is cold, to a protostar, in which the nuclear fusion processes can
commence. And if you try to follow the physical conditions theoretically
you find you get to a stage where the gas is dense but still cold—10
At those temperatures everything sticks to the dust on timescales of
hundreds of years, much shorter than the time scale of star formation.
Molecules such as carbon monoxide or ammonia, which astrochemists and radio
astronomers use to probe physics, have frozen out and have ceased to be
applicable as probes. With the molecules frozen out the only components of
the gas are hydrogen, deuterium, and helium, and they form very few
observable species, such as H2D+ and
D2H+, but in surprisingly high abundance.
Are your interests today are broader than just the
theoretical side of the astrochemistry of molecular clouds?
You bet. I apply astrochemistry to a whole variety of environments such as
interstellar clouds, protoplanetary disks, planetary nebulae, late-type
stars, masers, and models for molecular clouds in external galaxies. I have
an observational program as well that uses molecules to trace what’s
actually happening in the process of star formation. I’ve also put in
place an observational program that allows us to test our own models. To do
so we use the JCMT and other radio and submillimeter telescopes around the
world. I also work very closely with scientists who are doing laboratory
studies on reaction rates. Indeed one of the pleasures of working in this
field has been the interdisciplinary nature of the research.
Your two most-cited papers are for the 1995 and
1999 editions of the UMIST databases for astrochemistry (Millar TJ,
et al.,Astron. Astrophys. Suppl. 121:139-85, 1997
and Le Teuff YH, et al.,Astron. Astrophys. Suppl.
146: 157-68, 2000, respectively). The results of those compilations
are clearly much in demand. How did the project start?
When we first started working on the chemistry of interstellar clouds, my
group was doing cutting-edge research, but we were finding that there were
few groups around with whom we could interact. The fundamental data for our
models are chemical reaction rates. We have spent many years building this
from a variety of sources. Importantly for the science, we had a tough time
trying to convince observers to run tests for our models. So we decided to
make all of our computer code and our data available to the community
worldwide, so that the observers could develop models and see that they are
of value in understanding interstellar physics.
The first edition was published in 1991. By going
public your group had, to a large extent, taken a calculated risk for
the public good that you would still have plenty of research to do on
refining your models. How did publication affect the citation
"ALMA will be able to peer in
detail at star-formation regions in molecular
The publication of the UMIST databases and the computing codes to go with
them was not a decision that we took lightly. We had a long argument within
our group about whether it’s best to keep your codes and data
private. The advantage of non-disclosure is that you can keep your research
cutting edge because you possess some tools that none of your competitors
have. You can get your papers published quickly without competition. On the
other hand, we could make everything available to the world and then have
to live with the competition that comes from other people using our codes
and having better ideas.In the end we decided that it would be better for
the project as a whole and better for us—we had to have the
confidence to remain competitive in an even more competitive world—if
we simply made everything public. So that’s the publishing story
underlying our most highly cited papers.
By going that route we have stimulated far more groups, internationally, to
take an interest in astrochemistry. It has meant that observers can do
their own modeling to interpret the observations, and that has to be a good
thing. We published the first version back in 1991, and we release updated
versions about every 4–5 years. The 2006 database was released in
The 2006 database (Woodall J, et al., Astron.
Astrophys. 466:1197-U203, 2007) is already ranked #7 in citations
of your papers. What’s new in this version?
The current version contains some 4,573 binary gas-phase reactions, an
increase of 10% from the previous (1999) version, among 420 species, of
which 23 are new to the database. We have made updates to ion-neutral
reactions, neutral-neutral reactions, particularly at low temperature, and
dissociative recombination reactions. We have included for the first time
the interstellar chemistry of fluorine. Everything is available at
The high citation level suggests that these papers
are appealing to a broad audience.
That’s right. In addition to the astrochemistry, we get a lot of
citations from atmospheric physics. Also we get citations from
astrobiologists looking at the chemistry of the early solar system.
Furthermore, the combustion and fusion communities use our data.
I notice you have a cluster of highly cited papers
on deuterium fractionation in molecular clouds.
The D/H ratio at the end of the Big Bang was about 10-5.
However, since the mid-1970s observers have been finding D/H ratios of 1%
in interstellar molecules, which represents an enhancement by a factor of
1,000. When we started to model this effect we found that because of the
fact that deuterated molecules have lower zero-point energies than their
hydrogenated form, they get formed preferentially at low temperatures. We
found that a D/H ratio of 1% could be reached in an interstellar cloud at a
temperature of 10K, such as the dense cloud in Taurus.
In more recent years when people have been using radio interferometers to
look at star formation regions they started finding out that some of the
D/H ratios were very large, and they found multiply deuterated molecules.
Most spectacularly, triply deuterated ammonia was 1012 times
more common than the raw statistics would suggest.
Our papers look at the physical conditions required to get these enormous
enhancements. What we found in the series of papers is that in regions of
cloud where the matter density is about 1000 times higher than the average,
but still at a temperature of 10 K, everything but hydrogen and deuterium
freezes. This leads to very large enhancements in deuterium fractionation
in the gas phase and, through collisions with the dust grains, to an ice
chemistry in which deuterated molecules are made very efficiently.
Finally, let’s look at where your research
is now headed. The Atacama Large Millimeter Array (ALMA) is one of the
largest ground-based astronomy projects. When completed in 2012, this
major new facility will surely impact on observational
This is a very exciting time for astrochemistry. ALMA will have
unprecedented spatial resolution, frequency coverage, and sensitivity for
the study of molecular emission. We are trying to make predictions for what
ALMA might see in very small, cold, star-forming regions or protoplanetary
disks. We’re also interested in the connection between the early
solar system and the interstellar gas. ALMA will be able to peer in detail
at star-formation regions in molecular clouds. ALMA should get plenty of
data to challenge our models and the laboratory studies we have on
Professor Tom Millar, Bsc, Ph.D., FRAS
Faculty of Engineering and Physical Sciences
Queen's UniversityBelfast, Northern Ireland
Millar's most-cited paper with
152 cites to date: