VOLUME 21 THE UNIVERSITY OF WISCONSIN-MILWAUKEE | GRADUATE SCHOOL NUMBER 1


Answers Don't Always Lie Beneath The Surface

Physics professor's discoveries are establishing
a new area in surface science


When a metal surface is exposed to air, the oxygen gradually reacts with the metal to form an oxide, what we call rust. Understanding and preventing rust is a major area of study in the field of surface science.


"The implications of growth are huge in all of technology,"Marija Gajdardziska- Josifovska says. "Most of our modern-day technology depends on being able to control a surface and grow something in a controlled way on a surface."

A car's catalytic converter takes harmful exhaust emissions, like carbon monoxide, hydrocarbons, and nitric oxides, and—through reactions with surfaces of substances such as platinum and rhodium—produces less harmful substances: carbon dioxide,water vapor, nitrogen, and oxygen. Promoting reactions such as these is called catalysis, another major area of surface science.

In a CD player, a tiny, infrared laser beam "reads" compact discs. When it bounces off the CD's ridged aluminum surface, the beam is altered. By measuring the changes in the laser beam, the CD player interprets the ridge patterns in the disc and reproduces the music. Protective coatings for surfaces such as eyeglasses are tough, but yet very thin. Achieving the right composition of the thin protective films or the CD player's laser source involves growth on surfaces, another area of surface science.

To help non-specialists understand the important role of surfaces, Marija Gajdardziska-Josifovska compares the surface of a solid to the coasts and borders of a country.

"If you have a big country, the heart of the country is the heartland, the Midwest," the associate professor in UWM's Physics Department explains. "But then a lot of the things that are important to the country happen at the borders: How you interact with the other countries that are around you. Also, many special events happen at the coasts, the interface between water and land.

"And so in the same way you can think about a solid. (Atoms on) the surface of a solid have a different amount of neighbors than on the inside, and a number of important phenomena happen at surfaces."


Growth

The importance of growth goes beyond the earlier examples of lasers and eyeglasses.

"The implications of growth are huge in all of technology," Gajdardziska says. "Most of our modern-day technology depends on being able to control a surface and grow something in a controlled way on a surface." Besides CD players and eyeglasses, Gajdardziska also includes the entire electronics industry as beneficiaries of growth research. "Those are all interface and surface problems. All the action happens at the interface."

Of course, this includes the prospering Silicon Valley industries. "To make your computer work, you start with a big silicon crystal, and you have to grow it as well as possible. Then you create slices out of it. Then you take the surface of that slice and you want to make an entire chip on it, which has many layers."

In growth and catalysis, knowing the atomic arrangement, composition, and properties of the surfaces involved is of fundamental importance. Slices created from a block of pure silicon must be made at just the right angle.

"Different orientations of silicon have different structures at their surfaces, so on some surfaces you can grow things better than on others," Gajdardziska says. "On some surfaces you can grow a layer which arranges itself so that it mimics the underlying crystal, while on others it's harder to do that."

Surface scientists also must answer many questions about how a substance will grow on the underlying surface, called the substrate. Where will a molecule attach itself? Will it move along the surface? How many molecules are needed create a patch large enough to expand? Will large patches absorb the small? If patches form, how thick will they be? When making computer chips, layers grown on silicon must be even and uniform, so current can flow smoothly through them.

"How atoms will assemble depends on the combinations of surface and interface energies of the substrate and the film you're growing."

"You have a lot of everyday experience with these types of phenomena, like rain on your car," Gajdardziska points out. "The raindrops grow very differently, depending on whether you've waxed your car. The interface energy between water and wax is large, so water droplets ball up to minimize the area of contact, and they grow large and far apart. When you have a dirty, unwaxed car, the dirt particles act as very nice nucleation sites for growth of many smaller and flatter water drops.

"A soapy surface reacts very differently with the water, and you do not see drops. In scientific terms, the water wets the soapy surface because the interface energy between water and soap is smaller than the surface energy of water.

Scanning electron micrograph of a polar MgO(111) surface showing triangular pyramids (the image was recorded at the UWM Advanced Analysis Facility by Dr.Plass abd Dr.Sklyarov). The surface properties of magnesium oxide are important for catalysis and for growth of thin films. Prior to the work of Gajdardziska and her collaborators it was believe that the faces of these pyramids were the neutral MgO(100) type surfaces, and these observaion were the basis for the general idea that clean polar surfaces must facet into neutral surfaces. Using a combination of transmission and scanning electron microscopy, along with atomic force microscopy, Gajdardzisk's group recently demonstrated that the faceting is introduced by acid etching and that the faces of the pyramids are not the neutral (100) type crystallographic planes (for example, in the figure above one of the three sides should have vanished if the faces were of (100) type). these pyramids can be erased by high temperature annealing when the polar(111) surface becomes stabilized by a surface reconstruction. Image and text provided by Marija Gajdardziska-Josifovska.


"We are trying to use this and understand these mechanisms at the atomistic level."

The atomistic events described here are taking place on a very small scale. In her research, Gajdardziska routinely makes measurements in nanometers— billionths of meters. Why must the films be so thin? With very thin layers, surface scientists can often force materials into unique structures, with beneficial properties that do not exist in larger crystals.

Gajdardziska assisted Carolyn Aita, materials professor in UWM's College of Engineering and Applied Science, on a U.S. Army-funded project to develop protective thin films for glass surfaces. Aita received a patent for the results. Gajdardziska cites this project, which used alternating layers of zirconia (zirconium oxide) and alumina (aluminum oxide), as an example of controlling thickness of layers.

"If you make the layers only a few atomic layers thin, the zirconia, rather than being monoclinic, which is what it likes to be at normal temperatures and pressures, you can force it into a new crystalline structure, a tetragonal structure. That structure has advantageous mechanical properties. This is a clear example of nanotechnology, controlling the size and going to the nanometer size.

"I always think in terms of social sciences," the well-traveled scientist says, providing another geographical analogy. "My country of origin is Macedonia, which is tiny. When you have a huge country like the United States, the borders with Canada and Mexico are less important, because there is so much area inside that can more or less define the nature of the country. But when you have a tiny little country like Macedonia or Switzerland, the length of the borders compared to the area inside starts getting substantial, and then that interface with your neighbors ends up really affecting the nature of the country and the future of the country, much more so than with a big one.

"When the crystallites become so small the surfaces become a non-negligible portion compared to their volume, you start seeing new structures, new nano-scale phenomena that don't exist for large crystals. This is one project that we've worked on which was interesting from a fundamental point of view, and also has a direct application."


Catalysis

In a catalytic reaction, a surface promotes reactions between atoms or molecules that would not normally react.

"That molecule, when it is sitting in air, has a binding energy, which is altered when it attaches itself to a surface; that small modification may be sufficient to allow a certain reaction to happen," Gajdardziska explains.

"If you want to generate a particular molecule to burn in a cleaner engine, you have to find a surface that will allow you to do that type of a reaction, and you have to understand how the reaction happens, how the products fly away. Sometimes you have things that stay on the surface and eventually clutter the surface, so that you kind of poison the surface. Eventually your catalytic converter stops working. Either you have to do something to regenerate it or you toss it away and start with a new one."


Electron microscopy

Gajdardziska uses a high-end transmission electron microscope to study objects with nanometer sizes, which cannot be seen with optical microscopy. She established UWM's Laboratory for High Resolution Electron Microscopy with start up funding she received from the university upon her arrival on campus in 1993. To clarify a basic principle of microscopy—wavelength—she employs her "Take Your Daughter to Work Day" explanation.

"You can think about the wavelength of your probe as a size of a pencil," she explains. "The thing you do when you're a kid is to trace your hand. In order to see all your fingers you need a sharp pencil. If you use classroom chalk, you may not be able to go in all the spaces between your fingers, but you still see that there is a hand, although you don't see all the details. If you use a fat sidewalk chalk, you may not even be able to recognize that the object is a hand.

"In order to see something, the tool that you're using to see it has to have a finer dimension than the dimension of the thing you want to see. In microscopy, that is the wavelength. Both light and electrons have wave properties, and their wavelength is that critical dimension. Optical microscopy is not suitable for atomic resolution imaging because the wavelength of visible light is of the order of hundreds of nanometers, while the interatomic distances are of the order of tenths of nanometers. To continue with the hand-tracing analogy, trying to see atoms with visible light is like laying a chalk the size of the Sears Tower on its side to try to trace your hand."

Transmission electron diffraction patterns recorded by Gajdardzisk's group show that polar oxide surfaces can be stabilized by surface reconstructions which persist in air. The experimental TED pattern in figure(a), obtained by Dr.Plass from an anealed MgO crystal which is terminated by (111) type surfaces, consists of diffracted beams due to the bulk crystal periodicity (i.e. spots of 220 type ona hexangon) and of additional fainter diffracted beams which are due to a surface reconstruction with(|3X|3) R30 periodicity. Analysis of the data with a novel, direct-methods approach reveals that the surface is stabilized by oxygen trimers(as shown by the direct methods map in figure(b) and the atomic model in figure(c). This is the first experimental observaion of the unstable cycle from of ozone i which the three oxygen atoms from an equilateral triangle, as opposed to the open form of ozone which is present in the atmosphere. Images and text provided by Marija Gajdardziska-Josifovska.


The structures Gajdardziska studies are too small to be seen with visible light, so she uses electrons, which have a much smaller wavelength. High-energy electrons are also better suited for her current research area, oxide surfaces, because they are less disturbed by charges carried by oxide surfaces.

Gajdardziska also uses spectroscopy and diffraction to analyze surfaces. In x-ray spectroscopy, she directs electrons at the sample, which in turn emits x-rays of different energy. From the x-ray spectrum, she can identify what type of atoms are in the sample. From electron diffraction, which is the interference between waves scattered by the sample, she can tell how atoms are distributed in the sample—whether they're in an ordered, crystalline structure, or a random structure.

"Those are the kinds of information that we obtain from electron microscopy, using a combination of imaging, diffraction, and spectroscopy," Gajdardziska says. "And I do all three usually to completely attack a problem."


Stable polar oxide surfaces

Metal oxides are commonly used as substrates to promote catalytic reactions or surface growth. When oxides form, the metal atoms lose electrons, and the oxygen atoms gain those electrons, so the oxide is composed of charged particles, or ions.

The surfaces containing an equal number of positive and negative ions are neutral surfaces, and as such were believed to be the only possible terminating surfaces for metal oxide crystals. A basic, widely accepted principle in surface science has been that surfaces with a net charge— polar surfaces—cannot exist, and must rearrange themselves into neutral facets. It was thought that the surface could not withstand a net charge, so it would always make itself neutral.

But Gajdardziska has found that the surface properties that scientists thought were natural for polar surfaces actually were the result of acid etching—a step in preparing the surface after it's created—and not a basic property of the surface. With her research group, she discovered that under the right conditions at very high temperatures, polar surfaces can be created and stabilized.

"The polar surface can stabilize itself by a surface reconstruction, where it rearranges its atoms in a special way that it can still stay flat and solve its charge problem at the atomistic level," Gajdardziska says.

"The really surprising thing about this surface structure is that it's even stable in air. Usually with all the metal and semiconductor reconstructions, they're only stable in what we call an ultra-high vacuum. You have to spend a lot of time and money on very good pumps to achieve that. And we have vacuum equipment like that in the Laboratory for Surface Studies."

The unique atomic structure of the new polar surfaces may produce different catalytic reactions or allow the growth of new particles that are more active.

"If you can just simply promote a different morphology of the metal nanoparticle, you can increase the number of atoms which will be active even though you're working with the exact same size of a particle," Gajdardziska says.

"If you can grow those small platinum nanocrystals that are used in catalytic converters in the cars, and you can grow them the exact same size so you're using the exact same amount of platinum, but you grow them with a different shape so they end up having more active sites, then that would be the gain to be obtained."

Although Gajdardziska can cite some possible uses for her discoveries, her polar oxide research is so new that its future applications are unknown." That is one of the problems that is so new and so borderline at the frontier that I cannot say what will be the applications of it right now," she says.

"Very often when you're carrying on a more basic research you really cannot predict exactly what will be the application, because you're trying to open up a new area."


New atomic-level imaging techniques

Until recently, imaging of growth on surfaces was done after the growth was finished. But to learn about certain growth properties, measurements must be made while growth is happening. As a post-doctoral research associate at Arizona State University, Gajdardziska recorded the first reflection electron microscopy movies of cluster growth during non-congruent evaporation of compound semiconductor surfaces. These semiconductor surfaces consist of group III and group V atoms. In non-congruent evaporation, the group III atoms leave the surface, while the group V atoms stay. This usually occurs at much lower temperatures than the expected melting point of the substance.

"It's really interesting when you have a movie," Gajdardziska says. "You get a very different story than when you have a snapshot at the very end of the event."

Gajdardziska and Margaret Malay, one of her doctoral research assistants, are studying non-congruent evaporation of indium phosphate, an important component in many semiconductor lasers. "The growth really is critical and depends on the surface structure and the temperature," Gajdardziska says. "Having something like this happen to your surface is very important to know about and control, so that you can grow the desired structure for a laser."

Another limitation of conventional imaging is its two-dimensional nature. With Gajdardziska's use of electron holography at the atomic level, more detailed understanding of surfaces is possible.

A wave has two components: amplitude, which you can think of as its energy, or in the case of light waves, how bright something looks to you; and phase, which tells you what distance and through what media the wave has traveled.
Post-doctoral student Richard Plass and doctoral student Margaret Malay work with Gajdardziska at the high resoultion electron microscope.


In conventional imaging techniques, only the amplitude, or brightness is recorded, which is why photographs are two-dimensional. The same phenomenon happens with electrons or any kind of a wave.

Holography, which literally means "to write everything," is a modern, three-dimensional imaging technique that allows both amplitude and phase information to be recorded.

The only way to "write everything" is to have two waves that are coherent with each other, which means they have the exact same wavelength and they were generated from the same original source. A beam splitter is used to create the two coherent beams.

In a basic holographic scenario, one wave, the object wave, is sent to the object you're trying to learn about and bounces to a receptor, while the other one, the reference wave, is sent directly to the receptor. The interference pattern created when the two waves meet contains information about the phase.

"If I can measure the phase, then I can tell you what is the thickness of something, because the phase is directly proportional to the thickness," Gajdardziska explains. "If I'm trying to look at this nano-particle, which I cannot see with anything else but electrons, and I want to find out what is its shape, what is its thickness, if I were to record just one conventional image, then I have a projected image and I don't know what is the thickness. If I do a holographic image then I have all the information about these three-dimensional particles stored into that hologram."

Gajdardziska, who started using holography while at Arizona State, now travels to Oak Ridge National Laboratory in Oak Ridge, Tennessee for holography work.

She has developed the most accurate holographic technique for measuring the index of refraction for electrons. "Just like there is refraction for visible light, there actually is refraction for electrons. But to measure that, you need the holographic techniques."


Life in Academia

Gajdardziska's research often enhances her teaching. Besides an electron microscopy course she has established at UWM, she teaches optics to physics majors and general physics courses for scientists and engineers.

"You can really use your research results," she says. "I use my electron microscopy lab and all the computer software that I have for processing digital images. I use that in my optics class for some of the demonstrations. I can show them things that they wouldn't ordinarily see if they were taking the course with somebody who isn't doing research. It just adds one other dimension to the course and also makes it more connected. They can see how the general concepts fit into the different applications."

Teaching the optics course actually reminded Gajdardziska of one principle that helped her in her polar oxide surface research.

"Part of the misinterpretation of what people were doing before was based on actually not remembering one fact from optics," she says. "There are some intangible ways in which different ideas connect."

Teaching is an important and satisfying part of Gajdardziska's life in academia.


Schematic of the high electron microscope(and peripherals) found in Gajdardziska's laboratory. Electrons are generated at the top of the microscope column and are focused onto the sample by a series of elecetronmagnetic lenses. An image or diffraction pattern from the sample is formed by the objective lens which is magnified and projected onto a phosphorous screen for viewing. Images can be recorded conventionally on electron-senstive film and later developed for analysis, or they can be recorded digitally by CCD camera and viewed immediately on the computer. Chemical analysis of samples can be performed with the attached energy dispersive x-ray(EDX) spectrometer which collects x-rays generated from regions of the sample that are exposed to the electron beam of the microscope. The computer attached to the EDX detector allows the user to record spectra and identify chemical species present in samples. Image and text provided by Marija Gajdardziska-Josifovska.

"I was attracted to being able to work with people, and teach people how to do research as well—help them make that step from just learning existing knowledge to actually making the transition to generating new knowledge. That's the wonderful thing that happens at graduate school. In that sense its very different from the other stages of education."

Among her numerous honors, awards, and grants, Gajdardziska received a 1995 Presidential Faculty Fellows Award, a five-year, $500,000 federal grant recognizing her exceptional work as a young scientist. She has used part of the award to fund her student researchers. Along with graduate students, Gajdardziska has also included select undergraduates in her research program, and even found funding for some of them.

"That's something that gives them a very different flavor, different experience." Most of those students are now either in graduate school or are definitely trying to find or have found positions associated with doing physics after undergraduate studies."

Ken Egan graduated in May and began Ph.D. studies at Arizona State University this fall. Last year, one of her students, Amalia Hicks, who was accepted to the Massachusetts Institute of Technology's Ph.D. program, was involved in research as a summer intern at the Graduate School's Advanced Analysis Facility.

"It's nice to have some of your undergraduates be accepted at very good schools. I think the fact that in addition to the coursework, she had this direct research experience, looked good in her entire application package."

With all the positive aspects of working in academia, there are some things that have been challenging for Gajdardziska.

"It's like running an enterprise," she says. Challenges go beyond establishing and funding a laboratory and funding research assistants. Maybe the water chiller needs to be fixed. Or the air conditioning broke down. Professors have to find funding to fix those problems, too.

"You always think of the academics as spending their days only worrying about fundamental things. Running an experimental lab, one oscillates between all the extremes from the most mundane, everyday problems to inventing good scientific problems and getting results and getting published.

"Once you figure out how to keep a lab running at a university, there are some advantages to it because you do have the freedom of selecting the problems, as opposed to working on someone else's problems, and you also do have the interactions with students and young people."

Gajdardziska says her research group, with its members each working on different but related projects, increases the amount of knowledge generated.

"You increase your power by having a research group. You can exchange ideas and try ideas that you come up with. Your power is in a sense multiplied by how big your group becomes."

Gajdardziska is the author of over 65 refereed publications and 55 conference proceedings and presentations. She also is featured in the book Who's Who in Science and Engineering in America.