By Michael K. Miller, Philippe J. Pareige, and Kaye F. Russell

The concept of the atom has existed since the 5th century B.C., but it was not until the 19th century that this concept was developed into a scientific theory. Even then, the ability to see individual atoms was considered an impossible task because they must be magnified a million times to be visible. However, in 1951 Erwin W. Müller, a professor at Pennsylvania State University, achieved this magnification after constructing a fundamentally new type of microscope that he called a field ion microscope. The device enabled him to see atoms on a routine basis. In those pioneering days of the technique, the scientist was forced to sit in a completely dark room for 20 minutes to let his eyes become accustomed to the dark so that he could see the faint glow of the atoms on a phosphor screen. Fortunately, after the introduction of image intensifiers similar to those used in night vision binoculars, sitting in the dark was no longer necessary. In 1967, Müller and his colleagues further advanced the development of the field ion microscope by adding a mass spectrometer. The new instrument, called the atom probe field ion microscope (or atom probe, for short), made it possible not only to see but also to identify individual atoms. The properties of materials can be dramatically changed by the presence of small quantities of certain elements, especially in the boundaries between a material's crystalline grains, called grain boundaries. Scientists have found that changes in properties can be related to changes in structure and composition at the microscopic level. Thus, metallurgists at ORNL and elsewhere were excited about the potential of this new instrument in characterizing the structure and composition of metals at the highest possible spatial resolution.

The ORNL atom probe includes a three-dimensional atom probe (3DAP) developed here.

In 1981, Jim Bentley and Everett Bloom, materials scientists with the Microscopy and Microanalysis Sciences Group in ORNL's Metals and Ceramics Division, recognized that the atom probe technique was developing into a powerful tool for characterizing metals and alloys. They decided to add the capability to their group to complement the other powerful microanalytical techniques at their disposal. As a result of their initiative, the ORNL atom probe facility began operation almost 10 years ago.

Today, this atom probe is used routinely to examine materials literally on an atom-by-atom basis. This facility has been used in a large number of collaborations with other government agencies, industry, and universities both in the United States and around the world. During discussions in 1983 on setting up this facility, we decided to take the evolution of the atom probe technique another major step forward by developing an instrument to determine the position and identity of all atoms in a small volume. This concept has been adopted by other atom probe groups around the world. As a result, a new type of atom probe, known as a three-dimensional atom probe (3DAP), is now available as a commercial instrument.

A Look under the Microscope's Hood

The design of the field ion microscope is remarkably simple despite its atomic resolution. The basic instrument consists of a vacuum system in which an electrically insulated specimen is placed about 50 millimeters (2 inches) in front of a phosphor screen. The specimen is in the form of a needle that is over 1000 times sharper than an ordinary household sewing needle. In fact, the end of the needle is so sharp that it cannot be seen by the naked eye or even a standard optical microscope; not surprisingly, fabricating these needles is an art itself.

In this field ion micrograph of a nickel-molybdenum (Ni4Mo) intermetallic compound, each dot is a single atom.

Once the system has been evacuated to a very low pressure and the specimen is cooled to about 60 K (-213°C), a small trace of neon gas is let back into the vacuum system, and a positive voltage is slowly applied to the specimen. Because of the specimen's needle geometry, the neon atoms around the end of the needle are attracted to the positively charged specimen. If the voltage on the specimen is high enough, an electron may be removed from the neon atom, leaving a positively charged neon ion on the surface of the specimen. However, because the surface of the specimen also is positively charged, the neon ion is repelled rapidly from the specimen toward the phosphor screen. When the neon ion strikes the phosphor screen, it produces a spot of light. This process occurs continuously all over the specimen surface, and the resulting picture on the phosphor screen is called the field ion image. Consider a field ion image of a nickel-molybdenum (Ni4Mo) intermetallic compound, shown above. Each of the individual dots in this image is an individual atom. One characteristic feature of the field ion image is the array of concentric rings. These rings arise from the intersection of the atomic planes in the crystal and hemispherical nature of the specimen surface, as seen in the computer simulation of the end of the needle, shown at left.

In this computer reconstruction of the sharp end of a needlelike field ion specimen, concentric rings appear because of the intersection of the atomic terraces with the surface of the specimen.

The technique is not restricted to examination of the surface of the specimen. By increasing the voltage on the specimen slightly higher than that needed to ionize the neon atoms or by applying high-voltage pulses of short duration, the surface atoms themselves can be ionized and removed, thereby revealing the interior of the specimen. This process, known as field evaporation, is the cornerstone of the atom probe technique. This process can be very carefully controlled, as shown in the following micrograph. In this nickel-zirconium (Ni7Zr2) intermetallic catalyst, the specimen was field evaporated until nine atoms were left on the central atomic terrace. A series of short, high-voltage pulses were then applied to the specimen until one of the nine atoms (shown at the top left corner) was removed, leaving eight atoms on the terrace. This sequence of field evaporation was then repeated for the remaining atoms until only one atom was left on the atom terrace, as seen in the lower right frame. Field evaporation is also used to clean the specimen and remove any irregularities immediately after it is made.

The only limitation on the type of material that can be examined with the field ion microscope is that it must be capable of conducting electricity to some extent. Thus, the technique may be applied to almost all metals and alloys, semiconductors, and a few special ceramic materials.

This field evaporation sequence shows the gradual removal of eight of the last nine atoms on the central atomic terrace of a nickel-zirconium (Ni7Zr2) intermetallic catalyst. One atom is evaporated from the central terrace between each pair of frames.

Although some information about the identity of the atoms can be deduced from the brightness of the atom in the image, a more precise method is to use the mass spectrometer section of the atom probe. By rotating the specimen, the image of the atom of interest is positioned over a small aperture in the phosphorus screen that acts as the entrance to the mass spectrometer. The atom can be removed by applying a 10-nanosecond, high-voltage pulse to the specimen to ionize all the atoms in a thin surface layer. Although all the atoms in this surface layer are removed and repelled toward the phosphor screen, only those atoms that pass through the small aperture and enter the mass spectrometer are analyzed. The effective size of this aperture can be selected to allow passage of just one atom or the atoms originating from an area a few atoms wide.

The identity of each atom is determined based on the time it takes the atom to travel from the specimen to the detector—a distance of a little more than 2 meters (7 feet). Because all the different types of atoms in the specimen have the same kinetic energy (0.5 mv2) when they leave the specimen, the elements of low mass (m) travel faster than the heavier elements and reach the detector first. However, this measurement requires sophisticated high-speed electronics because the typical velocity (v) of these atoms is approximately 300 miles per second. To perform this measurement with sufficient resolution to be able to identify all isotopes of each element in the periodic table, we use a clock that ticks 1 billion times a second and count the number of ticks. By repeating this process, the composition of small volumes may also be determined by simply counting the number of atoms of each element in that volume. A series of these small volumes can be collected to estimate the variation in composition in different regions of the specimen.

The Atom Probe's Applications

In today's rapidly changing world, there is a continuing need for new materials with improved properties. For example, materials that can be fabricated into various shapes and that can withstand high temperatures are needed in highly efficient engines being designed to reduce fuel use and emissions of pollutants. One of the standard approaches to achieving this goal is to design the alloy by adding different elements to refine a particular property. For example, in the commercial nickel-based superalloys used in turbine blades in modern jet engines, this refinement may include the addition of 10 or more elements to achieve the desired combination of properties, including strength, toughness, creep resistance, and oxidation resistance. The number of permutations of the possible interactions between this number of elements and other microstructural features makes it difficult to predict properties accurately. Therefore, it is desirable to determine the role of each of these additions experimentally. The atom probe is particularly suitable for determining the location and distribution of all these elements within the structure of the alloy.

The atom probe is particularly suitable for determining the location and distribution of elements within the structure
of the alloy.

A relatively simple example of this approach is the addition of boron to nickel aluminide (Ni3Al). Pure Ni3Al is extremely brittle at room temperature, limiting its technological application as a potential new lightweight high-temperature material. However, if as little as 200 parts per million of boron is added to the alloy, Ni3Al becomes ductile, permitting it to be shaped into useful components without cracking (see related article "Nickel Aluminides: Breaking into the Marketplace"). This beneficial addition of boron may be understood by characterizing the distribution of boron throughout the alloy. The field ion micrograph below reveals a region in a boron-doped Ni3Al specimen that contains part of a grain boundary. The grain boundary is clearly seen in this micrograph because of the presence of a series of bright dots and the abrupt disruption of the concentric rings at the boundary where the orientation of the two crystalline grains changes. The identity of these bright dots may be uniquely determined by positioning the probe aperture over an individual bright spot in this image. Analysis of the atoms forming these dots in the mass spectrometer section of the atom probe revealed that they were individual boron atoms that had segregated to the grain boundary. This type of individual atom identification is possible only with this instrument. Atom probe analysis also revealed that boron was present at all the other disturbances in the crystal structure, including dislocations, low-angle boundaries, stacking faults, and antiphase boundaries. From these observations, we concluded that a fraction of the added boron atoms segregated to the grain boundary, where they acted as a type glue at these otherwise weak points, thus preventing the nickel aluminide alloy from failing preferentially at these locations.

In this field ion micrograph of boron-doped nickel aluminide (Ni3Al), the bright dots are individual boron atoms that have segregated to a grain boundary (arrowed).

Unfortunately, not all elements are beneficial to the properties of a material. Some combinations can be potentially deleterious if the levels present are too high. For example, most steels used to fabricate the pressure vessel of a nuclear reactor contain small amounts of phosphorus and copper. Copper is not a typical impurity in commercial steels but can be present because of the use of copper-coated welding rods in the production of the welds in older reactors. Although these elements do not normally have a significant influence on the mechanical properties of the new pressure vessel, the distribution of these elements may change during the service lifetime of the vessel as a result of its exposure to a temperature of approximately 290°C for many years.

Added boron atoms segregated to the grain boundary, where
they acted as a type of glue, preventing the nickel aluminide
alloy from failing.

Two possible changes in the distribution of elements that could affect the long-term stability of the welds in these steels are the formation of extremely small copper-enriched precipitates and the segregation of phosphorus to the boundaries. The small copper precipitates strengthen the steel in much the same way as steel reinforcing rods strengthen concrete, but the precipitates also have the undesired side effect of making the steel brittle. Because embrittlement is not desirable in nuclear reactor vessels, it is important to determine whether these features are present, and if so, whether they change over the lifetime of the reactor. It is also economically important to know whether it is possible to extend the operating lifetime of a reactor pressure vessel without compromising safety. Because of the low service temperature, the size of early changes in the features approaches atomic dimensions. Therefore, the atom probe is an ideal technique for detecting and characterizing early changes in the microstructure of these materials.

This image is a computer reconstruction of an atom probe analysis of a small copper-enriched precipitate in a neutron-irradiated steel used in reactor pressure vessels. The precipitate contains not only copper (Cu) but also nickel (Ni), manganese (Mn), and silicon (Si) atoms. The iron atoms are not shown for clarity. The precipitate is approximately 10 atom diameters across.

Atom probe characterizations of a number of different pressure vessel steels reveal that they have extremely complex microstructures that involve a variety of small precipitates. Through use of the atom probe at U.S. Steel, Mike Miller and S. S. Brenner were the first to demonstrate that, in neutron-irradiated materials, the solubility of copper in the alloy was extremely low and the excess copper formed copper-enriched precipitates. Most recent research at ORNL demonstrated that these copper-enriched precipitates contained not only copper but also some iron, manganese, nickel, and silicon. This finding can be clearly seen in the computer reconstruction of the atom probe data of a volume of material that encompasses a precipitate. In this volume, the iron atoms (which are the majority of atoms in steel) have been omitted for clarity, and the extent of the entire volume analyzed is shown by the gray cylinder. The diameter of this precipitate was determined to be only approximately 10 atoms across. Atom probe analysis has also revealed that the phosphorus had segregated to the grain boundaries but that its deleterious influence there was minimized because the grain boundaries were also coated with a thin film of molybdenum carbide and nitride precipitates, which prevents failure from occurring prematurely at these locations. This type of atomic-level information has produced a new level of understanding of control parameters required to design alloys that have properties needed for these types of applications.

This type of atomic-level information has produced understanding of control parameters required to design alloys for special uses.

More recently, Raman Jayaram and Miller have used atom probe field ion microscopy to demonstrate an important link between ultrafine-scale chemistry and a critical bulk mechanical property. If small amounts of molybdenum and zirconium are added to nickel aluminide in the proper proportions, the modified alloy is more ductile at room temperature, stronger and more resistant to deformation at high temperature, and less likely to react with oxygen. However, as the temperature reaches 777°C, the modified alloy is transformed from a ductile to a brittle material. This transition occurs at a lower temperature, 277°C, for undoped nickel aluminide. Using the atom probe, the ORNL researchers showed that the nickel aluminide matrix was severely depleted of zirconium and molybdenum. They found that the molybdenum precipitated out as spherical particles, and they obtained the first direct experimental evidence of zirconium segregation to dislocations in nickel aluminide when the material is annealed at elevated temperatures, 400 to 800°C. Identification of elements segregated to dislocations is nearly impossible to do by other techniques such as analytical electron microscopy. These changes in dopant positions and concentrations are responsible for property changes, such as loss of ductility.


These examples show it is possible to see and catch individual atoms routinely using the ORNL atom probe field ion microscope. Our technique enables complex materials to be characterized at the atomic level. Scientists from industry and universities have access to this state-of-the-art instrument through the Shared Research Equipment (SHaRE) program. This atom probe facility is the only one in the U.S. Department of Energy's national laboratory system that is available to outside users. Our ability to see and catch atoms is being shared with others.

Our atom probe facility is the only one in DOE's national laboratory system that is available to outside users.

Sponsors and Sources

The SHaRE program, including atom probe R&D, is supported by the DOE Office of Energy Research, Office of Basic Energy Sciences, Division of Materials Sciences.

Suggested reading

E. W. Müller, Z. Physik, 31 (1951), 136.

E. W. Müller, J. A. Panitz, and S. B. McLane, Rev.Sci. Instrum., 39 (1968), 83.

M. K. Miller and G. D. W. Smith, Atom Probe Microanalysis: Principles and Applications to Materials Problems, published by Materials Research Society, Pittsburgh, Pennsylvania, 1989.

M. K. Miller and M. G. Burke, "Atom Probe Field-Ion Microscopy: Imaging at theAtomicLevel," Imaging of Materials, eds. D.B. Williams, R.Gronsky, and A. R. Pelton, Oxford University Press, Oxford, England, 1990.

J. A. Horton and M. K. Miller, Acta Metall., 35(1987), 133.

M. K. Miller, M. G. Hetherington, and M.G.Burke, Metall. Trans., 20A (1989) 2651.

B I O G R A P H I C A L Sketches

Michael K. Miller, a native of Beaconsfield, Bucks, England, is a senior research staff member in the Microscopy and Microanalytical Sciences Group in ORNL's Metals and Ceramics (M&C) Division. He has a bachelor of science and technology degree in materials science from the University of Bradford, England, and a D. Phil. degree in metallurgy from Wolfson College, University of Oxford, England. He was a Science and Engineering Research Council Fellow at Oxford's Department of Materials Science for two years. He came to the United States in 1979 to work at U.S. Steel Research Laboratories in Monroeville, Pennsylvania, and at the University of Pittsburgh. In 1983, he began atom probe research at ORNL. In 1986 he developed the innovative concept of the three-dimensional atom probe. His research interests include atom probe field ion microscopy, phase transformations, radiation damage, intermetallics, and nickel-base superalloys. He is president of the International Field Emission Symposium. Since 1985 (except for two years), he has been editor of the annual conference proceedings of the International Field Emission Society. He is the coauthor of two books: Atom Probe Microanalysis: Principles and Applications to Materials Problems (with G. D. W. Smith), published in 1989, and Atom Probe Field Ion Microscopy (with A. Cerezo, M. G. Hetherington, and G. D. W. Smith), published in 1996.

Kaye F. Russell, a native of Clinton, Tennessee, is the principal technologist whose primary assignment is technical support and atom probe field ion microscopy research for the Microscopy and Microanalysis Group. She attended the University of Tennessee as a chemistry major. She came to ORNL in June 1967 as a Youth Opportunity Program student in the M&C Division's Welding and Brazing Laboratory. She transferred to the Metallography Group in the fall of 1967 and worked as a photographer's aide and then a metallographic technician. In 1982, she moved to the division's Radiation Effects and Microanalysis Group as a technician to prepare specimens for transmission electron microscopy. She began atom probe research in 1986. In 1993, she received a Martin Marietta Energy Systems Award for Technical Achievement.

Philippe Jean Pareige of Le Havre, France, is a postdoctoral researcher in the M&C Division's Microscopy and Microanalysis Group through the Oak Ridge Institute of Science and Education. He has a Ph.D. degree in physics from Rouen University in France. While working on his doctoral degree, he conducted research at the Laboratoire de Microscopie Ionique et Electronique (Ion and Electron Microscope Laboratory) of Rouen University. In 1991 he received the "young metallurgist" first prize from the Societe Francaise de Metallurgie et des Materiaux (French Society of Metallurgy and Materials Sciences) in Paris. His research interests include physical metallurgy, atom probe field ion microscopy, and nuclear pressure vessel steel surveillance.

Where to?

[ Next article | Search | Mail | Contents Review Home Page | ORNL Home Page ]