When chemists and materials scientists talk about "top-down" and "bottom-up" approaches to designing chips and nanostructures, what are they talking about? And what are the advantages and disadvantages of the two techniques?

Xia: Traditionally, engineers fabricate nanostructures through the "top-down" approach by carving them out of a large substrate. The "bottom-up" approach, on the other hand, relies on the assembly of building blocks on a smaller scale to generate nanoscale and larger structures or systems. The top-down approach is good in terms of precision. That means you can generate structures with tightly controlled dimensions and at well-defined locations. One primary concern with this approach is always the capital requirement—i.e., it’s expensive. Another is that the number of structures you can produce in any period of time is very limited. For example, if you count the total number of transistors that Intel has been able to manufacture in the last several decades, it’s a very small fraction of one mole, which is the term that chemists use to count molecules. A mole is 6.02 x 10²³. The point is that chemists can easily make one mole of nanostructures in each run of synthesis. The challenge here is to create the right building blocks from smaller units such as atoms and molecules. In principle, we can tailor the structures and therefore the properties of final products by controlling the shape, size, and chemical composition of the building blocks. The major disadvantage of this bottom-up approach is that we still don’t have precise control over the size and shape of nanostructures, not to mention the know-how to assemble them into functional devices or systems. So there’s a long way to go with this bottom-up approach.

  Can you describe to us how the bottom-up process works, and maybe give us an example from one of your own structures?

Xia: From the perspective of methodology, all bottom-up approaches to nanostructures share the same physics: nucleation and growth. But the details of these two steps can be substantially different depending on the materials and structures involved. This means we have to develop a specific procedure for each type of nanostructure. As an example, take the most extreme structure that we’ve created to date. This is the gold nanocage, which is a little gold nanobox with all eight corners cut off. It’s kind of amazing to realize that this type of structure can be made as small as tens of nanometers in size and with single-crystal walls as thin as 1 nanometer. But we can do it, and it’s robust enough to be dispersed or dried, and even injected into tissues. As for the process of how it’s formed, that’s almost hard to believe. Here’s the picture: when a silver nanocube is dumped into chloroauric acid solution, silver atoms are oxidized and removed from one of the six faces of the cube to form a small pit. The process is just like corrosion. The electrons released from this reaction then migrate to other faces and are captured by the chloroauric acid to generate gold atoms. These gold atoms are immediately deposited on the surface of the silver cube to form a thin shell. In the meantime, some of the silver atoms diffuse into the gold shell to form an alloy. The plating of gold atoms continues until there’s no silver left inside the cube. At this point, the small hole from pitting closes up to form a seamless box. In the next step, we add more chloroauric acid, and the silver atoms in the alloyed shell are then selectively etched out to leave behind many small pinholes in the shell. Finally, these pinholes merge to form a big hole at each corner of the cubic box. This synthesis is simple and elegant. The reaction occurs sequentially to form gold nanocages of increasing porosity as the volume of chloroauric acid is increased.

  What do you consider the most challenging part of this process?

Highly Cited Papers by Younan Xia  and Colleagues, Published Since 1996 (Ranked by total citations) Rank Paper Citations 1 Y.N. Xia, et al., "One-dimensional nanostructures: Synthesis, characterization, and applications," Adv. Materials, 15(5): 353-89, 2003. 916 2 Y.N. Xia, et al., "Unconventional methods for fabricating and patterning nanostructures," Chem. Rev., 99(7): 1823-48, 1999. 507 3 Y.N. Xia, et al., "Monodispersed colloidal spheres: Old materials with new applications," Adv. Materials, 12(10): 693-713, 2000. 499 4 Y.N. Xia, G.M. Whitesides, "Soft lithography," Ann. Rev. Materials Sci., 28: 153-84, 1998. 476 5 Y.G. Sun, Y.N. Xia, "Shape-controlled synthesis of gold and silver nanoparticles," Science, 298(5601): 2176-9, 2002. 450 SOURCE: Thomson Scientific Web of Science

Xia: In a sense it’s just knowing that there’s a reaction that works in this way. I came across this one accidentally when I was teaching introductory chemistry. There was an example in the textbook that described galvanic replacement reaction between zinc plate and a copper sulfate solution. Of course, after the proof-of-concept step, you then need to understand all the details involved in the reaction and try to extend it to other metals or materials.

  How has this nanostructure research changed in the decade since your graduate-school work? And how has your own approach to research changed?

Xia: Well, for starters, the number of people engaged in this research has exponentially increased, as has the number of journals and publications. The research tools have also advanced to the point that we can now do routine observation and manipulation of objects as small as a single atom. For chemical synthesis of nanostructures, we now have an array of methods to choose from for each type of nanostructure. Essentially, we can process any material as nanostructures, although our control over size and shape varies from material to material. For my own research, I’ve shifted the focus towards more fundamental aspects involved in the formation of nanostructures, rather than the demonstration of a synthetic protocol. At the same time, my research group has also started to look at applications enabled by various nanostructures we have synthesized.

  What do you see as the ultimate goal for your nanostructure research?

Xia: I have two goals in mind. First, I want to establish the logic of nanostructure synthesis. If someone needs a particular type of nanostructure made of a specific material, I want to be able to tell that person what’s the best synthetic route and experimental conditions. Ultimately, I hope we’ll have the same level of understanding and control over nanostructure synthesis as we now have for organic molecules. Second, I want to demonstrate one or two killer applications for the nanostructures that we’ve synthesized or will synthesize.

  Your hottest research seems to cover a somewhat dizzying variety of simple nanostructures: one-dimensional wires, cubes, solid spheres, hollow spheres, etc. Why these different structures? And how do you see them ultimately being used?

Xia: The simple answer is that different applications require different types of nanostructures. For example, we need nanowires so that we can place at least two electrodes onto them and then measure how electrons or phonons—i.e., heat—are conducted through nanostructures. Nanowires are also the obvious components of choice for fabricating interconnects between electronic devices, as well as nanoscale waveguides that can direct the flow of light like an optical fiber. Colloidal spheres are essential for applications such as photonic crystals because they have a natural tendency to crystallize into lattices with three-dimensional periodicity. Cubes are unique in that each of their six equivalent faces can be given individual functions, so that we can control their interactions and their self-assembly into different types of structures. The sharp corners and edges on a cube also make it easier to localize surface changes, and we can use the resulting electric fields to enhance the optical signals derived from molecules adsorbed on the surface.

  Why hollow structures? What do those give you?

Xia: As a rule of thumb, any time you make nanostructures with hollow interiors, you’re going to get new functions or new applications. For example, nanotubes, or nanowires with hollow interiors, make it possible to investigate how liquids are transported through these nanoscale structures. Hollow spheres allow us to engineer the optical properties by controlling the shell thickness—to improve the performance of photonic crystals, for example. Hollow cubes, which are just boxes, enable us to tune the optical properties. For example, gold cubes of 50 nanometers in size will strongly absorb light with a wavelength of around 550 nanometers, while gold boxes of similar size can strongly absorb light in the near-infrared region from 800 to 1,200 nanometers, depending on the thickness of the wall. Finally, hollow structures are critical for what we call encapsulation, which is when we put something—a drug, for instance—inside the box. The process has already found widespread use in drug delivery and controlled release. So, the simple answer about structures is that the broader the range of nanostructures, the greater the diversity of applications.

  One of the lines you often hear about nanotechnology is that it’s a vision in search of a reality. Do you think that the technology has arrived?

Xia: The answer really depends on your definition of "nanotechnology." There are plenty of examples of applications where nanostructured materials play important roles. Everything from ancient examples—the manufacturing of church windows, and the late-Roman Lycurgus cup, for instance, where gold nanoparticles were used for coloration—to plenty of modern examples. These include catalysis, where metal nanoparticles are ubiquitous, or microelectronics, where the critical dimension of individual components on a microprocessor is already on the scale of around 70 nanometers. So, in that sense, nanotechnology is a reality. I think the more important point is that it isn’t and won’t be everything. There are many applications where you simply can’t use nanostructures, or it doesn’t bring any additional benefits if you do. For example, when you deal with optical diffraction, the feature size of the grating has to match the wavelength of the light. This means you have to use gratings with pitches on the micro- and millimeter scale to control the diffraction of infrared light and microwaves, respectively. Nanostructures would be too small to have any influence on the propagation of electromagnetic waves with long wavelengths. Nano, of course, will be advantageous for applications where small sizes, high surface areas, and higher reactivities are needed, but it’s still just one of the many parameters you have to keep in mind when you’re developing commercial products. The way I see it, nanotechnology will probably not cause revolutionary changes to our society and daily life, certainly not the way microtechnology has. However, the performance of most devices and the efficiency of most reactions could be significantly enhanced by incorporating nanostructured materials.

  Are there applications that now use your nanostructures?

Xia: There are applications under various stages of development. For example, we’ve synthesized the hollow gold nanostructures—nanoboxes and nanocages—and these are being evaluated for use as imaging contrast agents for optical diagnosis of cancer and as therapeutic agents for the photo-thermal treatment of cancer. We’re also looking at putting antibodies on the surfaces of these gold nanostructures, and these antibodies will then recognize the receptors on tumor cells and attach to the tumors specifically. We’ve also licensed the technology for our silver nanowires to a start-up company, which is looking to develop them for what’s called "flexible electronics." You mix the silver nanowires with polymers to make highly conductive and transparent sheets.

  Where do you see your research going over the next decade? And how do you see nanotechnology itself evolving over that time frame?

Xia: My research has been and will stay focused on the fundamental side of nanomaterial synthesis. We’re trying to understand and control the nucleation and growth steps involved in the formation of nanostructures. At the moment, my group is developing a suite of tools capable of capturing, identifying, and quantifying the nuclei that serve as a bridge between atoms and nanostructures. Down the line, I hope this research will give us a detailed picture of the evolution pathway from atoms to nuclei and nanostructures as well as the design rules for synthesizing metal and semiconductor nanostructures with precisely controlled electronic, magnetic, catalytic, and optical properties. The long-term goal is to build a scientific base for the large-scale production of nanomaterials with whatever specific properties are necessary for applications in areas such as electronics, photonics, catalysis, information storage, optical sensing, biomedical imaging, and drug delivery. In the next five to ten years, I’m sure nanomaterials will continue to find new applications. What’s important is not that we witness some new era built completely on nanotechnology, but that these nanomaterials and nanostructures keep finding new applications and particularly, perhaps, in solving the bigger problems that confront us as a society—in health care, for instance, or the energy crisis.