Division home page

Technical Highlights

• Nanoscale Physics Facility Comes On-line. A new Nanoscale Physics Facility was commissioned by researchers in the Electron Physics Group to support the development of new measurement methods for future nanotechnologies. This new laboratory will be used to fabricate electronic and magnetic nanostructures and measure their physical properties with atomic scale resolution. Current research is focused on quantum and spin electronics, areas central to the emerging nanotechnology revolution. The laboratory came on-line with the successful operation of the main measurement system, a scanning tunneling microscope (STM) operating at cryogenic temperatures in an intense magnetic field.

  
  
  Figure 1. Interior of the Nanoscale Physics Facility

  The cryogenic STM was designed and built at NIST. Novel features of the microscope include a solid molybdenum body design (made possible by the high-speed machining capabilities of the Fabrication Technology Division), an integral 3-axes positioning system of the sample and tip with picometer precision, optical access to the sample/tip junction, and non-magnetic construction. The microscope is a completely self-contained unit that can be translated between a room temperature system and a liquid He cryostat. The microscope was designed to have very high spatial and energy resolution. It can measure displacements below 1 pm and resolve electron energy levels separated by 600 µV. Operating in cryogenic, high magnetic field, and ultra-high vacuum environments the microscope measures electronic and magnetic properties of nanostructures on an atom-by-atom basis. The microscope has achieved atomic resolution measurements on the surface of Cu(111) at 2.3 K and in magnetic fields up to 10 Tesla. The stability of the microscope allows atomic resolution measurements to remain in registry while the magnetic field is swept.

  The laboratory contains facilities for the fabrication of samples, tips, and nanostructures. The facilities include traditional molecular beam epitaxy of III-V semiconductors, superconductors, and magnetic materials, as well as bottom-up nanofabrication using autonomous atom assembly. This laboratory establishes a new state-of-the-art for nanostructure fabrication and measurement and positions NIST to be a major contributor in this rapidly emerging nanotechnology arena. (J. Stroscio, R. Celotta, and E. Hudson)

• Development of an Efficient Laser-Plasma EUV Source. Extreme-ultraviolet lithography (EUVL) is a promising candidate as the next generation patterning tool for microelectronics. A large effort to develop this tool has been underway for several years by a team at Sandia and Lawrence Livermore National Laboratories, supported by the EUV-Limited Liability Corporation, which consists of Intel, Motorola, and Advanced Micro Devices. Finding a suitable source of EUV radiation for the tool has been one of the major challenges. The source must be compact, bright, and free of contaminating debris. The team has chosen a laser plasma source in which a 10 ns laser pluse irradiates a target of tiny, micron-sized droplets of liquid xenon. The source is compact and the xenon plasma is practically debris-free, but more brightness is desired for higher throughput.

  In collaboration with researchers at the University of Maryland, we have shown that EUV emission from laser-irradiated Kr micron-sized droplets is a strong function of pulse duration. Using a specialized laser that allowed the variation of pulse durations from 100 fs to 10 ns while keeping the pulse energy constant at 50 mJ, we found that the EUV emission of Kr is greatest at 300 ps, where it is five times brighter than the emission created by 10 ns pulses. These measurements will be repeated for Xe, which is being supplied by the Livermore-Sandia team.

  A version of this source has been installed at NIST to calibrate EUV photodiodes for use as dosimeters in EUVL steppers. This source is run in a manner that produces emission with a time structure very similar to the source being used at the EUV-LLC Engineering Test Stand I to produce the most accurate measure of EUV photodiode responsivity under actual operating conditions in the stepper. (S. Grantham, T. Lucatorto, C. Tarrio, and R. Vest).

• NIST SEMPA Facility Upgrade. Our ability to image small magnetic structures has recently been improved by the installation of a new, ultra-high vacuum, field emission scanning electron microscope in the Scanning Electron Microscopy with Polarization Analysis (SEMPA) facility. The new microscope replaces two older units and provides a high intensity field-emission electron source and state-of-the-art electron optics. This new instrument, when coupled to the NIST polarization detection system, will allow faster, higher resolution imaging. The magnetization of magnetic nanostructures will be able to be imaged with 10 nm spatial resolution, the highest of any SEMPA system world-wide. The improved performance is essential to keep pace with the shrinking length scales of magnetic structures in magnetic media, spintronic devices and magnetic sensors. The microscope is also an Auger microprobe with facilities for in situ thin film growth and tools for nanoscale compositional and structural analysis. The microscope's operation is fully automated and under computer control, permitting remote control of the microscope, and immediate electronic transfer of SEMPA images and measurements to NIST customers. (J. Unguris)

Figure 2. Field-emission electron microscope in new SEMPA Facility.

• Large-Chamber EUV Reflectometer Installed on SURF. A specially designed, large (5000 kg) EUV reflectometer has been installed installed on SURF III to meet the demands of the U.S. developers of EUV lithography. In 2001, the EUV-LLC consortium will begin operation of a prototype EUV stepper called the Engineering Test Stand I. This stepper will have mirrors as large as 280 mm in diameter. The new NIST instrument is the only reflectometer in the world capable of making detailed measurements on mirrors this large. Designs for future production steppers incorporate even larger mirrors, which have been anticipated by the NIST planners in the design of the instrument. (L. Deng, S. Grantham, T. Lucatorto, and C. Tarrio)

Figure 3. The large reflectometer chamber is interfaced to SURF III, December 2000.

• Magnetism Beyond 2000. "Magnetism Beyond 2000," a special issue commemorating the publication of the 200th volume of the Journal of Magnetism and Magnetic Materials, features invited papers that address the current status of magnetism research and highlight questions for the future. Two of these papers were solicited from the Division's researchers.

  Figure 4. Spin structure as determined by analysis of the temperature dependence of SEMPA image data.

  Our recent research has focused on the coupling between two magnetic thin films separated by a non-magnetic spacer layer. These thin film structures have been of interest since the discovery of the giant-magnetoresistance (GMR) effect in 1988, and of significant commercial importance since their use in read heads for magnetic disk drives in 1998. One of the invited papers "Interlayer Exchange Coupling," by M. Stiles, reviews these systems, with emphasis on the comparison of predictions of a "quantum-well" model with experimental measurements.

  For most lattice-matched systems, where the physical structure of the multilayers approaches the theoretical ideal, theory and experiment are in good agreement. The best agreement, based largely on work done in the Division, is found for Au/Fe multilayers. An interesting exception to the good agreement is the widely studied Fe/Cr system. While many of the important discoveries in magnetic multilayers, interlayer exchange coupling and giant magnetoresistance for example, have been discovered in experiments on the Fe/Cr system, the many experiments on the Fe/Cr system have also led to conflicting and sometimes puzzling results. The paper, "Effect of Roughness, Frustration, and Antiferromagnetic Order on Magnetic Coupling of Fe/Cr Multilayers", by D. Pierce, J. Unguris, R. Celotta, and M. Stiles, reviews experimental and theoretical work on Fe/Cr multilayers. It explains the complex behavior in Fe/Cr multilayers with a unified picture, based on the delicate interplay between interfacial disorder and the antiferromagnetic order. (R. Celotta, D. Pierce, M. Stiles, and J. Unguris).

• Record Currents, Energy, and Beam Stability Attained in SURF Operations. The SURF III electron storage ring gave its first light in December 1998, and all key operations were brought back on line during 1999. In late 1999 and throughout 2000 we instituted a systematic approach to logging all operational parameters, automating control procedures, increasing the level and improving the quality of RF power, and optimizing the conditions of "fuzzing" the beam. This approach has led to significant qualitative and quantitative improvements in the service offered to SURF users. We have attained a record operating energy of 387 MeV, and now routinely inject initial currents in excess of 500 mA. Thus we can now produce, as a matter of routine, optical power output that is an order of magnitude higher than that ever produced by SURF II even under exceptional conditions. In addition, great progress has been made in understanding the origin of beam noise, and controlling it by modifying the (complex) harmonic spectrum of the RF power system. It is now believed that SURF III is sufficiently quiet to serve as the source for a UV Fourier transform spectrometer, and during 2000 a beamline was designed for this purpose, and its construction was started. This beamline will be dedicated in the first instance to accurate determination of refractive indices of DUV optical materials, which are required by the manufacturers of 157 nm lithography systems.

  The main goal of the SURF III program is the establishment of an absolute primary standard for source-based optical radiometry, based on the calculability of the synchrotron radiation source: if the energy and current of the electron beam are known, the radiated power in any spectral regions can be calculated from first principles. Thus, accurate electromagnetic measurements - of electron current, and of the magnetic field, which determines the electron beam energy - yield correspondingly accurate radiometric determinations. During 2000, we have continued to improve the quality of these two classes of measurements. Beam current is monitored indirectly in our single electromagnet storage ring using the synchrotron radiation emitted by the circulating electrons. The radiation is detected by a silicon photodiode connected to a low-noise operational amplifier whose gain can be varied from 10⁴ to 10¹⁰ using precision feedback resistors. At very low current levels, our determination is absolute, since we are able to count all the electrons in the ring, from the observed quantized decrements of light intensity as individual electrons leave the beam. The challenge is to bridge these results to the high-current regime of practical interest in operations. Towards this end, we have improved shielding, cabling, and filtering techniques to provide additional immunity from noise pickup and ground level variation. Magnetic field mapping performed during the upgrade to SURF III revealed peak to peak variations in magnetic field strengths measured azimuthally at the orbital radius of ± 0.006 % to ± 0.033 % over a range of fields corresponding to 52 MeV to 417 MeV. Using a new calibration and monitoring system we expect to achieve uncertainties under routine conditions of 0.05 % or better in the absolute electron energy value.

  Major improvements have been made in the SURF central control system to take advantage of improved data logging for the automation of basic machine operations. The availability of surplus computer equipment from the Bureau of the Census has given us an opportunity to construct an integrated network that will handle all central control and beamline data-handling procedures. (U. Arp, A. Farrell, E. Fein, M. Furst, R. Graves, E. Hagley, L. Hughey, and G. Mehena).

Collective Excitations of Bose-Einstein Condensates. The Division maintains a theoretical program on the physics of quantum-degenerate gases, which collaborates with experimental efforts in the Atomic Physics and Quantum Physics Divisions. A main focus of this effort is the understanding of the excitation spectra of these systems. In 2000, collaboration with researchers in the Quantum Physics Division led to the first direct experimental observations of vortex rings in Bose-Einstein condensates. A "dark soliton" state was created in a two-component (different spin states) Bose-Einstein condensate of 87Rb atoms: the wavefunction of one component has a nodal plane, which is filled with atoms of the other component. A filled soliton is found to be stable for hundreds of milliseconds, but as the filling is depleted, the soliton becomes susceptible to dynamical instabilities. These lead to its decay into stable vortex rings. Our simulations guided the production and observation of this phenomenon. (D. Feder, B. Schneider, M. Edwards, and C. Clark) Figure 5. Simulations of the decay of a BEC soliton into vortex rings. Time sequence (top to bottom) shows density profile (left), and internal vortex structures from two perspectives.

• Insight into Magnetic Coupling from Scanning Tunneling Microscopy. The study of the magnetic coupling between two ferromagnetic films separated by a non-ferromagnetic metallic spacer is an extremely active area of. The interlayer exchange coupling that causes the magnetization of the two ferromagnetic layers to be parallel or antiparallel, depending on the thickness of the spacer, is now well understood when the spacer is a diamagnetic or paramagnetic metal. However, when the spacer layer is an antiferromagnet, such as Mn, the observed coupling is non-collinear. J. Slonczewski (IBM, Yorktown Heights) proposed the torsion model to describe the more complicated interlayer coupling when exchange coupling within the antiferromagnetic spacer is important.

  The torsion model predicts that the coupling angle between the ferromagnetic layers is 90° for rough interfaces. For sufficiently smooth interfaces, the torsion model predicts that the coupling angle between the ferromagnetic layers will vary around a mean value of 90° by an amount that is very sensitive to the width of the thickness distribution of the spacer layer.

  We have used STM measurements of the growth of Mn on nearly-perfect Fe single crystal whisker surfaces to determine the thickness distribution of the Mn layer for particular growth conditions. The torsion model predictions of the coupling angles for specific thickness distributions were found to be qualitatively consistent with the coupling angles actually measured for Fe/Mn/Fe(001) tri-layers by SEMPA. Scanning tunneling microscopy measurements of the lateral scale of the Mn thickness distribution provided insight into how to go beyond the torsion model to obtain a better explanation of the results. (D. Pierce)