Advancing Basic Science for Humanity
New Tools for Nanoscience
IN NANOSCIENCE, RESEARCHERS ARE TRULY LIMITED BY THE TECHNOLOGY OF THEIR FIELD, needing increasingly more advanced tools for studying, analyzing and manipulating objects and systems at the scale of individual molecules and atoms.
To expand the boundaries of nanoscience, the Kavli Institute at Cornell for Nanoscale Science is now devoted to the development and utilization of next-generation tools for exploring the nanoscale world. The new director of the institute is Cornell University Physics Professor Paul McEuen, widely known for his work with carbon-based systems such as graphene and nanotubes. Serving as co-director is Associate Professor of Applied and Engineering Physics David A. Muller, whose pioneering work includes developing electron energy loss spectroscopy as a tool for predicting materials properties.
Recently, McEuen and Muller discussed the institute’s new mission and the need for advanced technology in nanoscience. In particularly, they described how they had come upon limits of observation and control in their own work, and how they plan to launch “high-risk, high-payoff” projects with the potential of changing the way scientists work worldwide. In McEuen’s words, “we’re looking for projects where you could say, ‘If I succeed, suddenly everybody’s going to want one of these.’"
KIC Director Paul McEuen (left) and Co-Director David A. Muller. Courtesy: Cornell University
As new co-directors of the Kavli Institute for Nanoscience, you’re assuming leadership of an institute that is now six years old. How much are you changing its mission and structure, and how much of the new mission is a continuation of what KIC has been doing?
Paul McEuen (PM): In the first years of the Kavli Institute, the mission was to engage the nanoscale community, both at Cornell and outside, in a discussion about what were the most important issues facing nanoscale science. A series of conferences and workshops addressed this issue in a variety of contexts, all the way from the first Kavli Futures Symposium in Greenland to a series of conferences here on nanoscale imaging that David was very involved in. As a result of those conversations, one thing that emerged very clearly was that it’s our tools that are still an enormous limiting factor in what we can do at the nanoscale world. We don’t have eyes and hands at the nanoscale to see and control things the way we’re used to at the milli-, micro- or macro-scale. As we thought about where we could have the most impact, we thought that trying to push the cutting edge in development of new sets of measurement approaches and tools would be a great way to go. If you make a new tool, you can have an enormous impact in a variety of different areas.
David A. Muller (DM): That about sums it up. I’d say that, in terms of what our strengths were, that this was the most high-impact, most efficient way we could think of going forward. So that’s why we’re re-focused.
Could you give me an example of a case in your work up to this point where you came upon the limit of a tool for measurement or manipulation, where the limitation really made a difference and you felt like you’d hit a wall?
DM: We probably both can give you examples. The one that really drove it home for me was when I was working at Bell Labs and we were trying to understand the fundamental limits to the scaling of computer chips. As you make the transistor smaller, every dimension has to shrink: the smallest part of the transistor is a thin layer, a little insulator called the gate dielectric. This layer was shrinking by about a couple of atoms every year, and we realized that we’d be down to about five atoms thick by about now. When you have a layer of material that’s only five atoms thick, two of those five atoms are at an interface and arranged differently to any bulk material. That’s 40% of your material – something that’s different from what you can just go and look up from a reference library. So in order to understand how thin you could make this layer – the answer turned out to be four atoms – you had to be able to look not just at those four atoms but also zoom in on the one atom at the interface, and work out what was going on there. The success or failure of the device came down to a single layer of atoms. So we needed to develop tools that could look at a single layer of atoms buried inside a material. If you wanted to figure out if a particular device would work, it would be the same challenge as trying to look at a pin on the seat of a car parked somewhere in the United States. The limits that we identified back then, in 1999, are the limits that we hit in production effectively a year ago and required a switch to a new set of materials.
Paul, do you have some case that really made it hit home for you?
PM: The case you just heard was about imaging. For me it was more about measurement and control. New materials had appeared on the scene, like carbon nanotubes, and we wondered if we could make a tiny resonator out of them -- the world’s smallest guitar? In fact, that was easy. We could suspend a carbon nanotube over a little open space, so that it looked like a string on a guitar. But we wanted to play it. We wanted to know how to pluck it. More to the point, we wanted to know how to listen to it. And we didn’t know what technique to use. There was no way of doing a mechanical measurement on a system this small -- to know that the nanotube was vibrating, for example. We tried all kinds of techniques before hitting upon a certain one that involved doing an electromechanical measurement that was adapted basically from the radio industry. This was a case where we had to find a new way of doing a measurement on an ultra-small system that just didn’t exist before. And based on that, we were able to make the world’s smallest electric guitar.
How small is that electric guitar, and how much smaller do you want to make it?
PM: The string is about a nanometer across and a micron long, and it will be pretty hard to make it much smaller. We could in principle make it a little bit smaller, but it’s not clear we need to. What we would like to be able to do is make a lot more of them and measure them en masse, not just one by one. A great challenge going forward is, how do we make lots of little things and get information into and out of them. This means building an interface from our world down to the nanoscale world on all fronts – through electrical measurements, optical measurements, imaging things, pushing on things, pulling on things. You could say that the mission of KIC is to build this tool set that allows us to see and to push and tug at the nanoscale world in every way possible.
What is the big physical barrier that you face? Does it have something to do with the basic qualities of matter? How do you propose overcoming it?
PM: This sounds kind of technical, but the statement for nanoscale electronic devices is that their resistances tend to be high in comparison to things that you can make at a larger scale. It’s easy to do certain kinds of resistance measurement at low frequencies, but if you want to do a fast measurement, it’s very challenging and you have to use very different techniques.
So it’s not just size. It’s also speed.
PM: That’s right.
DM: As a practical matter, we are talking about things that are just a few atoms in size. So the question is, how do I bridge a contact that is a few atoms in size on one side but, on the other side is in an everyday kind of dimension? You want to get things that are a hundred million times different in scale to talk to each other. That’s challenging. Another problem is that if something is only a few atoms across, it’s very easy for just a few atoms to disrupt things. If you have a single monolayer of contamination, in everyday scale you would never know it happened, but now that’s a big barrier to getting contact at the nanoscale. So we’re at a place where a single layer of atomic impurities matters. And that kind of single layer of atomic impurities can get on a clean sample in air almost immediately.
DM: We have a lot of different ways to tackle measurement at the atomic scale, and each has its advantages and its drawbacks. To get specific, with an electron microscope, we can detect a single impurity atom inside a material, image it and tell you its chemical state and what type of atom it was. But we also tend to whack it around a lot, because to do that measurement takes a lot of energy. And so there are other, much more gentle ways you can do it too. For instance, with an atomic force microscope, you could gently tickle it with Van der Waals forces
PM: Or you can just tap the object in question very gently. You can learn certain things that way, but just like with your eyes closed, tapping on an object is not nearly as useful as seeing it with your eyes as well and listening to it and being able to thump it and hear how it sounds. So we have little bits and pieces of the puzzle. As David was saying, we have electron microscopes that in very rarified conditions can see amazing things. But if you want to measure a cell while it’s doing the business of a cell, you can’t use those techniques, at least not yet.
Nanoscience researchers at Cornell University. Courtesy: KIC/Cornell
Because the energy that’s supplied is just too destructive?
DM: With electron microscopy, you’re putting a lot of energy in and only getting a little out, and that can be damaging. So there are much gentler techniques as Paul mentioned, like this gentle tapping. You could imagine just putting in one little tiny photon of light and detecting it coming out – that, too, would be much more gentle. People are trying, for instance, to do MRI (magnetic resonance imaging) with very tiny MRI sensors, and these might someday shrink down to the sensitivity of a single atom.
On Paul’s side of the fence, on what one might call the “manipulation” side, we have the same question: What is the state of the science now?
PM: The mechanisms we want to use to control things at the nanoscale are everything that we can possibly think of. For example, among our members there are people who use very different techniques to manipulate and control things at the nanoscale. I tend to work a lot with electronic forces -- with voltages, electrical measurements, things like that. Dan Ralph, another member of our team, has great expertise in using magnetic forces to control and manipulate things. And Michal Lipson, also a member of our team, has great expertise in using optical forces -- for instance, induced by light on systems -- to control and manipulate systems. For anything that you could use in a macroscopic control environment – such as using a voltage to make a switch throw, or using magnetism to cause something to happen – we want to be able to do the equivalent of that at the nanoscale. We need all possible techniques. It’s just like when you go to the hospital. They don’t just use one imaging technique on you; they have a whole variety of techniques at their disposal, each of which has a different function. We want to develop that same suite of tools that allow you to solve bits and pieces of the problem.
DM: Also, because we’re using very small objects, some of the forces that on an everyday scale are too weak to be noticeable actually can have a very big effect on nanoscale objects. Paul mentioned optical forces, for instance. There’s an optical pressure when you shine light on something. You don’t normally think about it every day, because you can’t feel it; it’s just too small. But on a very small object this becomes quite noticeable. And so we have new opportunities to manipulate things beyond what you would infer from your everyday experience.
DM: We’re looking for new imaging techniques that are going to enable new areas of science and discovery. Ideally, it would be an imaging technique that lets us see what no one has seen before, and lets us learn new information. For instance, a lot of imaging has happened at very low temperatures, and if we want to look at a biological system or a catalyst, we’d like to see this happen at room temperature, when things move around a lot more. The challenge here is how to take a technique that now works only in a very quiet, clean environment, and get it to work in the dirty real-world environment.
PM: I’ll mention one idea that we’re playing with. A graphene membrane is a sheet of carbon atoms that is literally one atom thick, yet can be a completely impenetrable barrier. On one side of it you can have liquid with cells and DNA, and on the other side you can have a vacuum. This would allow you to develop an instrument where you can have a scanning tunneling microscope or a transmission electron microscope on the vacuum side, and yet three angstroms away you can have something living in water. The goal is to make a new kind of chamber, where the sample gets to live in the environment that it wants to while you get to live in the environment that you want to, and yet you’re only separated by one atom.
Is that three-angstrom separation penetrable by a measuring device? In other words, can you detect what is on the other side?
PM: That is an excellent question. It’s a physical barrier, meaning that atoms won’t leak through. But you can send light and electrical signals through it; it’s mostly transparent. That’s the kind of thing we may very well try to do -- to build these new kinds of nano-chambers that allow us to use our advanced imaging techniques in new environments.
You have said that KIC wants to fund “high-risk, high-payoff” projects. How will that work in practice? In particular, what are you looking for in terms of high potential payoff? What would it be about an idea that would make it worth the risk of failure?
PM: For me, one measure of a new technique’s success will be that we find it all over the world inside of five years. The atomic force microscope and scanning tunneling microscope are past examples of that. They were invented and started propagating enormously quickly. That’s one of the great strengths of creative new techniques-- often they can be duplicated with reasonable expense and then can be used for all kinds of different problems. So we’re looking for projects where you could say, “If I succeed, suddenly everybody’s going to want one of these.”
DM: Paul’s statement is a good test. Another is that the new technique will change the way people begin to do a particular branch of science. What we’re looking for is people saying, “If only I could solve this problem; if only I could see what’s happening, things would be different for me and my field.” If someone is able to propose a technique that will clear this roadblock, that’s the kind of thing we want to pursue.
Is there any place or any potential for sub-atomic technology or sub-atomic science here? Is everything at the atomic level?
DM: I guess for a lot of what we’re thinking about, the atom is going to be the smallest building block we’re going to need to deal with. But it is important to remember that there are things like magnetic resonance imaging, where you’re looking at things like a signal from the nucleus of the atom rather than the atom itself. And that’s often a nice signal because it doesn’t disturb any other chemistry if you play with the nucleus and just flip it around, as opposed to smashing it to bits and pieces. So in that case, nuclear magnetic resonance is a technique that could be developed. In electron microscopy there’s a big advantage to making your probe of sub-atomic dimensions, because then you can resolve atoms more clearly.
PM: Just adding to that, I think it’s safe to say that we would like to be able to look at the scale of the individual molecular bond – the connective piece between atoms. But we’re not doing particle physics. Looking at the structure of the nucleus for its own sake is not something that’s in our purview.
What are the major funding sources for these projects that you’re going to be doing?
DM: Funding comes in different flavors. There are some funding pools that let you build very large pieces of equipment that are going to be used by a lot of people. That might be, for instance, the National Science Foundation’s Major Research Instrumentation Project. There are other sources of funding that will support students. Different agencies have different pools of money set aside for these different things. What we tried to do with the Kavli Institute was to identify areas that were falling in the cracks between these different programs
PM: What does not exist in the scientific world to a large extent is the equivalent of venture capital – or something closer to pre-venture capital, such as angel investing – to give you some money to get an idea off the ground. The government doesn’t do this well. It also tends to fund mission-oriented stuff, where you’re going to make a better widget for a certain application. In the biology world it’s the same deal: You’re going to cure a certain malady. It’s very hard to get money to do creative instrumentation development of the sort that we’re talking about here. Inventions like the scanning tunneling microscope, the atomic force microscope, single-molecule fluorescence techniques in biology and, going back further, the invention of scanning electron microscopes and transmission electron microscopes, has really revolutionized what can be done across a vast number of fields. But somehow the existing funding structures don’t focus well on developing such things. We’re really excited about being able to have a potentially huge amount of impact with the resources available to us, because we’re going to be doing something that’s key.
How unusual or how unique is KIC in this focus on frontier work in nanoscale instrumentation and control? Are there other institutes or other academic departments or schools that are kind of focusing on similar things and that you are collaborating with?
PM: I would say that they’re all focused on such work, in that they are all consumers of these tools and very interested in them. In applied physics, in physics, in chemistry, in the engineering disciplines, they’re all big consumers of these tools. And invariably, scientists in these areas are doing tool development because they need it. But they do it sort of on the side, because there’s not a department of instrumentation development. Everyone’s involved in this area, but it’s no one’s prime focus.
DM: Sometimes they are collaborators. But, as Paul said, in many cases their work is done as a bootleg operation under the radar. The projects they’re supported on are very mission-oriented – solving particular problems in materials or medicine – but they need tools to get the work done.
PM: And we expect fully to draw collaborators from a wide variety of departments, because -- as we’ve tried to emphasize often -- the breakthrough technique is likely to come from somewhere you were not expecting. The biologists may know how to do something from their world that nobody in the physics world is aware of; then you put the two pieces together and off you go.
PM: One of the very exciting things on the frontier is the development of increasingly complex nanoscale structures that have lots of electronics in it and other kinds of detectors, to be sent into your bloodstream or wherever to try to interact with the nanoscale world up close and personal. We will have to enable these machines not only to interact with the world around them – by electrical, optical, chemical measurements, for instance – but we’ll also have to send information back and forth between the machines and the outside world. We basically have to build the communication system for little nano-submarines. I want to be able to do the equivalent of calling it on the cell phone and saying, “Hey dude, what’s up down there?” and, “That looks bad, kill it,” and that sort of thing. By the way, you can’t do this with a cell phone. It won’t work, because you need a big antenna. So we have to invent technologies that will allow us to do that kind of interfacing with the nanoscale world. We’re trying to bring to the nanoscale the same kind of control to that world that we’re used to at the human scale.
DM: There’s been a lot of success with nanotechnology in a few specific fields such as, say, electronics, where transistors are now at the nanoscale. But there are a lot of fields out there that have never really thought of themselves as nano, yet what they’re looking at are nano-systems. For instance, say you’re in soil science or geology and you’re worried about carbon sequestration. When you look at the soil, it’s a nano-structured system, and yet there’s very, very little work done on characterization of soil at the nanoscale. People just don’t think that way. The way they think about the problem is through macroscopic, average, homogeneous method – the reason of course being that the tools you would need to look at these materials just haven’t been available. But if we can develop those tools, we ought to be able to change the way people think about those areas of science, such as thinking of soils as a nanoscience rather than as a macroscopic science.