A Roundtable Discussion
Science at the interface. Discoveries of new
phenomena and development of new technologies at the intersections of
various disciplines, such as biology, physics, chemistry, engineering,
and the measurement and computational sciences. Itís the ORNL vision.
To Bill Appleton, ORNL's former deputy director for science and technology,
it's a national craze. "Think of advances in medicine that involved
contributions from physics," he said, while moderating a roundtable
discussion on "Science at the Interface" on December 17, 1999,
|"Perhaps the greatest discovery of all this research is that we
can no longer separate basic from applied science. The disciplines
are connected in ways they have never been before."Vice President
Al Gore (1/1999)
The discussion participants talked about the past accomplishments and future of multidisciplinary research and collaborations involving ORNL, especially when the Spallation Neutron Source (SNS) begins operating here in the next decade. They addressed issues such as extracting meaningful information from large amounts of data, the changing nature of collaborations and experimentation (e.g., science at a distance); attracting, training, and retaining highly qualified researchers; applying ORNL strengths to shifting trends in science and technology; and accommodating changes in the research agenda that may come with having the SNS.
Roundtable discussion participants, from left, included Eli Greenbaum, Jacob Barhen, Ed Uberbacher, Michelle Buchanan, Herb Mook, Thomas Zacharia, Tony Palumbo, Jim Roberto (standing), Bill Appleton, Sheldon Datz (standing), Linda Horton, Mike Kuliasha (standing), Frank Plasil, Doug Lowndes, Mike Simpson, Don Batchelor, Reinhold Mann, and John Sheffield. Not pictured are Marilyn Brown and then ORNL Director Al Trivelpiece, who joined the group later. Photographs in this article were taken by Curtis Boles and enhanced by Gail Sweeden.
The participants in the roundtable discussion and their interests were as follows: Al Trivelpiece, then ORNL director; Jacob Barhen, head of the Center for Engineering Science Advanced Research (CESAR), computational aspects of nanostructures; Don Batchelor, Fusion Energy Division, theory and computational simulation; Marilyn Brown, deputy director of the Energy Efficiency and Renewable Energy Program in the Energy Division, impact of advanced technologies on U.S. energy systems and greenhouse gas emissions; Michelle Buchanan, associate director of the Life Sciences Division (LSD), analytical chemistry and structural biology; Sheldon Datz, Physics Division, atomic and collisional physics; Eli Greenbaum, Chemical Technology Division, chemistry and nanotechnology; Linda Horton, deputy director of the Metals and Ceramics Division, who oversees ORNL metal and ceramic sciences research funded by DOE's Office of Basic Energy Sciences; Mike Kuliasha, director of the Computational Physics and Engineering Division; Doug Lowndes, Solid State Division (SSD), thin-film growth, nanostructured materials, and nanotechnology; Reinhold Mann, LSD director, collaborations involving the life sciences and computational sciences; Herb Mook, SSD, neutron scattering; Tony Palumbo, Environmental Sciences Division, bioremediation and genetic engineering; Frank Plasil, Physics Division, nuclear and high-energy physics research; Jim Roberto, ORNL associate director for Physical and Computational Sciences; John Sheffield, director of the Joint Institute for Energy and Environment at the University of Tennessee at Knoxville, all aspects of energy and the environment; Mike Simpson, Instrumentation and Controls Division, nanoelectronics; Ed Uberbacher, head of LSD's Computational Biosciences Section, computational biology and bioinformatics; Tuan Vo-Dinh, LSD, medical applications of physics; and Thomas Zacharia, director of the Computer Science and Mathematics Division, the impact of computer technology on science.
Barhen of ORNL's Computer Science and Mathematics Division defined science at the interface as "a coherent integration and fusion of concepts, methods, and devices with the goal of achieving revolutionary advances that would not be attainable within the constraints of a single discipline." Plasil observed that scientific research needs technical solutions and that most technical solutions need scientific research. For example, teasing out information about whether free quarks were present during the earliest stage of the Big Bang requires sophisticated electronics to sort through a large volume of data coming from the Relativistic Heavy Ion Collider (RHIC) at Brookhaven National Laboratory (BNL). On the other hand, a fusion energy device that works cannot be built without the valuable information provided by materials research.
Simpson, an electrical engineer, gave an example of how two different disciplines can help each other. Gary Saylor, a microbiologist at the University of Tennessee, came to him and asked, "How can we instrument what's going on in living cells?" "We then came up with the idea that we should think of cells as fundamentally new electronic components that could be engineered to do signal processing and other electronic functions." As a result, the two collaborated on the development of the "critters on a chip" technology in which living bacteria on an electronic chip signal the presence of nearby chemicals.
ORNL offers an intellectual interface among the traditional
disciplines of physics, chemistry, biology, and engineering, Greenbaum
said. "ORNL's environment encourages the convergence of ideas and concepts
of separate disciplines. In our work, chemistry and physics are brought
to bear on the understanding of biological systems."
Greenbaum and Simpson are involved in building
an electronic logic device that uses tiny metal electrodes and photodiodes
made of proteins from spinach leaves. The nature of the device is a
metaphor of the meeting of disciplinesscience at the interfacerequired
to make it possible. "These new biosystems have soft condensed
matter and hard condensed matter," Greenbaum said. "Only because
we know the functions of individual biomolecules can we interact with
each of them after they are incorporated into designed nanostructures."
Trivelpiece made an observation that triggered
a discussion about the effects on various sciences of advances in measurement
capabilities and data acquisition and analysis. "Iíve been fascinated
by improvements in measurement sciences," he said. "Data collection
is automated so you can now do in 10 minutes what used to take 10 weeks.
If the Superconducting Supercollider had been constructed, the first
shot would have produced more data than all the high-energy physics
researchers accumulated in their preceding history. The data glut will
be so bad that you won't take data and archive it in the same way you
did in the past. Is there a future out there for doing experiments so
that you don't have to collect data?"
Trivelpiece suggested that computer models
of the future may be able to make instant sense out of mounds of data.
He noted that there is evidence that adrenalinethe fight or flight
hormonecan produce different effects in the brain. "The same chemical
in microscopic quantities can trigger the rage to kill or a catatonic
state depending on the circumstances," he said. He observed that,
as the measurements improve and the computer models become more realistic,
the ORNL-led project on the Virtual Human
might be able to shed light on the actions and interactions of not only
the heart, lungs, and kidneys but also the brain. The following are
edited excerpts from the roundtable discussion:
APPLETON: Haven't improvements
in instrumentation and computation and other advances aided the study
of what different genes do, making functional genomics a successful
example of science at the interface?
BUCHANAN: Bringing together physics, materials
science, computational science, and chemistry with biology enhances
what the biology community does. Advances in measurement science will
some day allow us to measure chemical concentrations in the living cell
in real time.
MANN: You can't do functional genomics without
informatics capabilities and resources. Biology drives other sciences
and these sciences also drive biology.
UBERBACHER: About 10 years ago Reinhold
Mann and I developed computer technology to find genes in DNA sequences.
We found the first few genes that way. By early December the entire
human chromosome 22 with its 5000 genes had been sequenced. We are in
the midst of analyzing the entire human chromosome 22 for genes, and
we will be analyzing human chromosome 19 for
the DOE complex in 2000. We've improved our rate of analysis by orders
of magnitude by using computers in biology. We are now using a gene
chip, a revolutionary biological measurement technology that allows
people to look at an entire biological system at one time by making
thousands of measurements across the genes being expressed in a microbial
or human genome.
PLASIL: With engineers in the Instrumentation
and Controls Division, we physicists have developed application-specific
integrated circuits for the quark-gluon experiments we will be doing
at RHIC. This collaboration is necessary because the amount of data
we will get will be so huge that we will have trouble storing it. With
high-energy collisions, we get thousands of particles, but most of them
aren't very interesting. So we have to dig around for the three or four
that are. This data selection must be done very fast online with specific
instrumentation that really can be developed only at a national lab.
MOOK: In the early days of neutron
scattering at Oak Ridge, scientists counted neutrons at a rate of 1
pixel per second. With the SNS we hope to get 5 million pixels per second.
The question is, how does the user come in and get what he wants out
of these 5 million pixels? The answer is he can't without the help of
high-speed computing. The user may only want to measure the distance
between atom A and atom Ba bond length. The way we do that now is to
take a lot of data, look at the peaks, and then do modeling. What we
really have to do at major facilities now is give the user a computer
model into which to quickly fit the data.
ZACHARIA: We should do more to integrate
computing into other disciplines. Computer tools are being developed
at ORNL to analyze large sets of data. Someday you will be able to build
into your model a simulation tool so that the computer generates the
results or information you are seeking. The computer science challenges
of the future are likely to be networking and data storage.
APPLETON: Are we using the unique aspects of a national lab for our research or
could this work just as easily be done at a university or some industry?
GREENBAUM: Even today university research is driven by traditional departments
and disciplines. Promotions within the department of chemistry are based
on what is done in chemistry. National labs are driven by national missions
and problems. We get beyond the disciplinary aspect of our individual
training by focusing on the problem. We can move more quickly because
we are not bound by traditional departments.
DATZ: A multitude of tools
for analyzing and altering surfaces of materials emerged at ORNL and
other national laboratories because of a marriage of vastly diverse
disciplines that could not happen at universities.
BROWN: By meshing
their strengths in basic research, applied and engineering sciences,
and behavioral studies, national labs have a unique ability to move
revolutionary ideas into the marketplace.
TRIVELPIECE: Academic institutions
have more difficulty with interdisciplinary activities than national
labs. This means that at national labs you have a better ability to
interact with other scientists and engineers in other fields.
SHEFFIELD: ORNL's work in materials
is outstanding, partly because it is coupled with other programs such
as energy and fusion. We couldn't have an effective energy research
program without a materials research program (see the article on gas
turbine power plants and ORNL's materials research). Universities
don't tend to have that kind of extent of work.
LOWNDES: At a university, a professor
may be torn in several directions and forced to work in a narrow area.
Many national labs are the caretakers of quite remarkable national and
international user facilities so we have full-time staff working there.
Despite our costs, I think we're learning how to do a better job of
using postdoctoral researchers and graduate students at these facilities.
The national labs are powerful places for doing big-facility research.
APPLETON: How important will collaborations be to national labs in the
KULIASHA: Multidisciplinary collaborations at national labs
result in a synergy, or team learning, because of the data produced
and the interaction of the participants.
LOWNDES: My own feeling is that we
have to collaborate with universities and use postdocs and graduate
students as efficiently as we can to be highly productive. We are learning
how to bring costs down while increasing our research productivity.
When I came to the Solid State Division in the 1970s, we had 120 staff
members and 20 postdocs. We now have an equal number of staff members
UBERBACHER: In the biology and genomics
domain, the importance of collaborations is exploding. In functional
genomics, if we're doing things with genes that have medical significance
and we want to respond to National Institute of Health calls for proposals,
we may need to team with medical schools or pharmaceutical and biotech
companies. It's more than getting postdocs; we have to partner with
the scientific leadership on the outside that can access different funding
agencies and bring different perspectives.
ZACHARIA: In an age where
information is so readily available and knowledge and advances are coming
at a much faster rate, if we don't have a strategy of partnering with
other universities and national labs as part of our planning process,
I think we will be left behind.
APPLETON: If this is such a good place
to do science at the interface, what projects should we be looking at?
What should ORNL be doing in the future that will make a difference
VO-DINH: At the national labs, I see a fusion of biological
and physical sciences and informatics. It used to be that basic science
was very important. Recently, we see more of the dual role of technology
and science. For example, scientists need a tool such as the atomic
force microscope to analyze the components of a single living cell.
Today we are taking a systems approach to scientific problems. That's
why we need multidisciplinary research.
MANN: Systems thinking will
be increasing in scope over the next 10 years in biology. In fact, systems
thinking will be needed to understand the whole genome.
TRIVELPIECE: In strategic planning, researchers defend their turf, talk to people in other disciplines, and then try to find an empty circle and figure out how to combine strengths to plug the hole. As we plan, we should
focus on our outrageous, unfair advantage and go to our strengths.
PALUMBO: One of our strengths is that
we can use mass spectrometry to get high throughput in DNA sequencing
and expression. The problem is that we can't do that yet for the task
of identifying proteins and discovering their functions and interactions.
We might need a brute force strategy to get high throughput. Shotgun
cloning is not an elegant strategy for getting sequence data, but that's
the one that has characterized a lot of microbial genomes.
BUCHANAN: Fifteen years ago, mass spectrometry
was not generally applied to biology. But after the development of desorption
processes, including refinements made through interactions with the
Solid State Division, we have techniques that allow us to use mass spectrometry
to characterize DNA and proteins. Beyond using mass spectrometry to
identify biomolecules, neutrons from ORNL's High Flux Isotope Reactor
will soon be used to study protein-protein interactions and protein signaling
processes. The SNS and computational tools will become important for
structural biology studies.
HORTON: Right now the Metals and Ceramics
Division focuses on structural materials. The strength we have is in
our partnerships and in applying those partnerships to merging basic,
applied, and industrial science. We need to strengthen what we have
been doing in the life sciences and instrumentation areas to develop
functional materials, such as biomaterials, sensors, and other smart
APPLETON: What are the most important scientific and technology
trends of the future?
BATCHELOR: Physics World recently
listed the ten great unsolved problems in physics. They are quantum
gravity, understanding the nucleus, fusion energy, climate change, turbulence,
glassy materials, high-temperature superconductivity, solar magnetism,
complexity, and consciousness. We have a lot of expertise with respect
to many of these problems. It seems like we should come together to
try to solve these problems using our strengths in computer modeling
I donít think itís a very good list.
SHEFFIELD: One of the strategic
technologies forecast for 2020 by Battelle Memorial Institute is one
area we actually work in, which is the issue of clean water. Water is
getting more and more polluted. We are working on ways to purify water
cheaply, stop pollution at its source, measure pollutant levels in water,
and map water flows and pollution sources.
ROBERTO: The use of neutron
scattering to study soft materials such as polymer blends and proteins
will be an exciting area in ORNL's future.
GREENBAUM: We're learning
what happens when soft condensed matter meets hard condensed matter,
as in logic devices combining spinach proteins with metal electrodes.
KULIASHA: I heard an interesting quote. The social cry of the 18th century
was "No taxation without representation." The social cry of the 21st
century will be "No experimentation without simulation."
MOOK: I was
surprised when I heard Vice President Gore say at the SNS groundbreaking
ceremony that the greatest things we will
do at the SNS are the things we haven't thought of yet because we have
a tool to do it. I once asked John Bardeen when he was here for a superconductivity
conference why the transistor was invented, how was that made possible.
That was certainly a major accomplishment for any society. He said it
was the people that came together at that time at Bell Labs that made
that possible. They had the freedom to interact in a special way. We
have to find ways to bring the best people together and enable them
to interact effectively.
APPLETON: National labs offer a unique meeting
place. They have major research facilities, computing tools, and a multidisciplinary
mode of operation, so that's the place where exciting research is going
to happen. But you have to get the people you need and you have to pick
KULIASHA: I disagree that we have to
bring them here. People feel a closer affinity to their coworkers at
other institutions than they do to people in the office next door. The
reason people had to be close together was because the experiments were
done at a few facilities. Now we have science at a distance. People
are communicating by e-mail and video conference calls. I would argue
that we shouldn't think about organizing around the institution but
instead around the scientific problem and the multiple researchers addressing
APPLETON: I don't
think any researchers are going to make a major discovery by talking
to each other over the Internet. They are going to have to be in the
MANN: What do we have to do to get them here? The field
of genomics is moving so rapidly. We must create the right environment
and have the right facilities and provide the resources. That varies
from opportunity to opportunity.
DATZ: There are many tasks that require
the use of many different techniques at large facilities (as in international
collaborations involving ORNL researchers at CERN near Geneva, Switzerland).
So you have to be able to pack a suitcase and go to a facility and do
the experiment and then go home.
LOWNDES: There is no substitute for
big-facility experimental research. National laboratories should be
sure that facilities are being exploited to the fullest extent. We need
to have interdisciplinary teams of people here putting together a new
facility like the SNS. We should think carefully about the organization
of research teams on site.
KULIASHA: Eighty percent of the people working
at the SNS will be outside users. If you've got an exciting team, access
to the exciting facility, and plenty of exciting work, people will join
the team where the action is.
APPLETON: What is your personal perspective
on the future of the Lab? I personally think that the national labs
are in the best position they have ever been in. There is going to be
more and more basic research done in this country. More of it will be
done at the national labs because basic research is getting more and
more complicated. You have to go to a major facility to study complex
BARHEN: We live in an exciting time and enormous possibilities
lie ahead. National laboratories should be national instead of DOE labs.
It would be nice if we could become a resource across all science agencies.
What will it take for us to be a NASA center as well as a DOE center?
Collaboration across centers and institutions will be driven by how
exciting projects are. We are working in areas of interest to NASA such
as quantum dot computing and mobile robots for space exploration.
BROWN: Our position in the scientific
world is important because we are the custodian of unique scientific
equipment and facilities. We're an aging staff of scientists. Our best
staff scientists are not that young. You can count our 35-year-old scientists
on one hand. We have to take advantage of what I see as a trend toward
distributed science. We have got to reach out and make our equipment
accessible to our young collaborators all across the world so we can
stay fresh. Vice President Gore said the impacts and products of the
SNS will be things we have not yet identified. There had better be some
advances and there had better be some good positive press because there
is an underlying public pessimism about the value of science. I don't
think we should focus on bringing people here as much as we should focus
on getting out our ideas, accomplishments, and data using the Internet.
We need to get our name out to people like graduate students so they
know where there is a precious resource. Our lab should be one that
can be accessed on the Web virtually.
PLASIL: We are thinking that once our
experiment is running at Brookhaven, we can control it from here. I
don't know if this is the way that things will go; this is perhaps the
way we would like to see them go.
We should do fundamental physics using neutrinos
that will be produced at the SNS as well as muons that could also
be produced there.
LOWNDES: In the
first stage of the experiment, you will have to go to the facility.
You need creative people there to help you design peripheral equipment
and processing equipment. You need auxiliary equipment for biological
molecules and thin films of interest. As far as the Internet is concerned,
if I can sit home in my sweatsuit with my Nordic exercise machine nearby
and run my experiment from my computer at home, that's great.
When I do an experiment, I may or may not get good results, but I almost
always learn something from the person next door or down the hall. I
might learn about a neat experiment someone is doing using DNA. I don't
see that kind of information exchange happening over the Internet.
If you're going to run the SNS as a distributed experimental facility,
you will do things very differently than if you're going to run it as
the best neutron center in the world for everyone to come to. That's
the strategic decision that must be made.
HORTON: I see ORNL and the
other national labs in a time of real change. ORNL historically has
not been an institution focused on a single facility. We have the Holifield
Radioactive Ion Beam Facility, the High Temperature Materials Laboratory,
the ShaRE program, and the Surface Modification and Characterization
Laboratory, and concentrations of expertise in chemistry and computing
science. We have been a lot of different things to a lot of different
people here. We are now becoming an institution that will largely focus,
at least on the surface, on neutron science. That is a tremendous change
for this institution and surviving that is something we will have to
enthusiastically embrace. Or we won't make it. We can't have a neutron
scattering facility for the entire world and have just a few of our
researchers using it. We have to embrace the Laboratory's biology, physics,
and materials science communities in neutron science projects. If you
look at Argonne, you see what's happening to them. The Advanced Photon
Source is becoming Argonne National Laboratory. Their basic science
programs are swinging toward incorporating APS in their research in
response to sponsors. It will be a dramatic and challenging time for
LOWNDES: We must find a way to retain some of the best
and brightest scientists.
DATZ: A good source of researchers
for us is the Wigner Fellows program. Our retention rate for these Wigner
Fellow postdocs is over 70%. This is an altruistic way of treating young
men and women and increasing the quality of our staff. But we should
be concerned that there is a paucity of young scientists coming from
the U.S. and Europe. Enrollment in hard sciences and engineering is
going down in this country. We should try to address recruiting good
people early on so we are able to steer them into good university programs.
LOWNDES: Through the National Science Foundation we have opportunities
to introduce interdisciplinary graduate training in national labs to
get over institutional barriers at universities. A proposal will be
made to the NSF to build an interdisciplinary graduate education and
research training program at the University of Tennessee in connection
with the national nanotechnology initiative. We are seeking support
for the active recruiting of good graduate students for our nanotechnology
projects at ORNL.
MANN: We need to take steps like that
of the old Biology Division, which was linked to the UT/ORNL Graduate
School of Biomedical Sciences, which turned out 150 top graduates who
are very successful around the country. A lab program connected to a
graduate school has the potential of providing a pipeline for giving
us staff and putting us on the map.
SIMPSON: I've worked a lot with University
of Tennessee students in the past five years. As costs rise, you have
to sell your time to more and more projects. Once upon a time $400,000
of internal funding could support work in several divisions. Now it
covers only equipment and one full-time employee. You run out of time
to work with students or junior staff members, so they go someplace
PALUMBO: For some non-DOE sponsors we will need not only the SNS
but also a good mass spectrometer facility to develop commercial technologies.
What brings people in is not just a big facility but pockets of facilities.
APPLETON: I thank all of you for participating
in this roundtable discussion. I come away from this discussion even
more convinced that multidisciplinary R&D is not only productive and
rewarding, but also is a mode of research that thrives in the ORNL environment.
It is not just our past success that convinces me of this. It's the
promising opportunities I see for the future such as the nanoscale science,
engineering, and technology initiative. Additionally, I think this kind
of open forum helps all of us hone our thinking. I hope ORNL can have
more of such gatherings in the future.