Atomic Physics Division

1995/1996 Technical Highlights

• Precise Measurements of Iodine Lines for Laser Wavelength Standards. Doppler-free spectra of molecular iodine provide a particularly convenient source of easily observed and highly reproducible lines for use as wavelength standards in laser spectroscopy. Precise wavelength values, however, have been reported previously for only a small number of lines. We have observed a selected group of 102 lines in a room temperature, 30 cm iodine cell by using Doppler-free, frequency modulation spectroscopy. The lines, which are uniformly distributed over the range 560 nm to 656 nm, were measured with our Fabry-Perot wavemeter with uncertainties of about 1 MHz (about 2 parts in 10⁹). Excellent agreement is obtained for six of the lines that have been measured previously with comparable or better accuracy in other laboratories. Our results will be also useful as reference lines for the calibration of wavemeters and interferometers. (C. Sansonetti)
• Spectroscopy with the Large Fourier Transform Spectrometer. We have implemented a much improved data acquisition system for our high-resolution IR-vis-UV spectrometer obtained in 1994. The new design replaces VME-bus computers with a more flexible PC-based data acquisition system, with much lower noise and less sensitivity to the effects of non-uniform sampling. The figure on the Division cover page compares spectra obtained with the old and new systems. The signal-to-noise ratio in the absorption spectrum has been increased by almost an order of magnitude, and the ghosts in the spectrum of the He-Ne laser have been virtually eliminated. We have installed the spectrometer in a vacuum tank, and expect to have the instrument fully operational by the end of 1996. (G. Nave, C. Sansonetti, and U. Griesmann)
• Laser-Spectroscopic Tests of Wavelength Calibration in Fourier Transform Spectroscopy (FTS). We have undertaken experiments to assess limitations of the wave number accuracy of FTS. In principle, the wave number scale of a FT spectrum can be calibrated by applying a simple multiplicative correction based on a single standard line. Measurements of precisely known lines of Ar II and ¹⁹⁸Hg made in several laboratories, however, have raised doubt about the reliability of this calibration procedure. In our experiments we illuminate a high resolution UV/VIS FTS with red light from a cw dye laser, UV light produced by frequency doubling this laser, and emission from a ¹⁹⁸Hg electrodeless discharge lamp. Light from these sources is combined in an integrating sphere. Regardless of the absolute calibration, the ratio of the wave number of the UV laser light to that of the red light should be exactly 2. We have observed that deviations from 2 as large as several parts in 10⁸ can easily result from minor misalignment of the interferometer that normally would go unnoticed. Using the ratio of the UV and red laser wave numbers as a diagnostic, we are working on alignment procedures that will ensure a calibration error less than one part in 10⁸. We are also remeasuring wavelengths of ¹⁹⁸Hg, which are recommended by the CCDM as an optical realization of the meter, by locking the red laser to an iodine line that has been precisely measured by laser spectroscopy and using the red and UV lasers as calibration standards. (C. Sansonetti and D. Veza with M. Salit and J. Travis of CSTL)
• High Resolution Measurements of Atomic Spectra for Space Astronomy. Spectral analyses have been completed for elements of interest for the interpretation of spectra obtained with the Hubble Space Telescope (HST) of chemically peculiar stars. The over-abundance (factors of 10⁴ or 10⁵) of certain heavy elements in the atmospheres of such stars continues to be a major puzzle of stellar astrophysics. In this connection, we completed a new energy-level analysis of Hg II from observations with our 10.7 m, normal incidence vacuum spectrograph. We also calculated oscillator strengths for this ion. We have provided rapid response to several requests for data from HST investigators. For example, our new measurements of Bi I, II, and III in the far UV will be used to establish the abundance of Bi in the star HR7775. Similarly, we made an accurate measurement of a Pb III line near 155 nm, which will be used to determine the abundance of Pb in the chemically peculiar star chi Lupi.
  
  We have also measured the wavelengths of 473 Fe II lines between 93 nm and 200 nm. Lines were identified in high-resolution grating spectra taken with the 10.7 m, normal incidence grating spectrograph, and precise Ritz wavelengths were derived from energy levels obtained from ultraviolet, visible, and infrared Fourier transform spectra. The uncertainties of the wavelengths span a range from 9 × 10^-6 nm to 7 × 10^-5 nm (0.005 cm^-1 to 0.02 cm^-1). The data will be useful for both calibration of laboratory spectrometers and also for analyses of astrophysical spectra. (J. Reader, C. Sansonetti, and G. Nave)
• Observation of Spectra of Highly Ionized Atoms. We have continued our studies of spectra of highly ionized atoms of importance for tokamak diagnostics. We completed measurements of spectra of highly-ionized rare earth elements from Gd³⁷⁺ to Hf⁴⁶⁺, generated by focusing a single beam from the Glass Development Laser at the University of Rochester onto flat metallic targets. The results will provide improved understanding of the iron, cobalt, and copper isoelectronic sequences and aid in the evaluation of calculations of relativistic effects in heavy ions. (J. Reader)
• New Critically Compiled Atomic Energy Level and Wavelength Data and Enhancement of the Atomic Spectroscopic Database. We have completed a large one-volume compilation of Spectral Data for Highly Ionized Atoms, including wavelengths, energy-level classifications, and transition probabilities for the elements Ti through Cu, and Kr and Mo. Other recently completed or ongoing compilations of energy-level and wavelength data include all spectra of Ar, Zn, and Ga, and of the lighter elements Be, B, F, and Ne. As the new compilations are completed, the data are being added to our Atomic Spectroscopic Database on the Internet. Hypertext links to references and ionization energies for particular spectra have been added, and other improvements of the interactive selection/retrieval procedures have been made. A bibliographic database on atomic energy levels and spectra covering the period since 1988 has also been prepared for the NIST Physics Lab Web site. (W.C. Martin, J. Sugar, A. Musgrove, with G. Dalton of the SRD Program)
• Atomic Transition Probability Data Compilations Published. A new data volume "Atomic Transition Probabilities of Carbon, Nitrogen, and Oxygen, A Critical Data Compilation" has been published in 1996. This 530 page book contains critically evaluated numerical data for about 13,000 allowed and forbidden transitions. Analogous work on He, Li, and Na is nearing completion. An expanded bibliographic database on atomic transition probabilities, which contains over 6600 references, is now accessible on the NIST Physics Lab WWW site. (W.L. Wiese, J.R. Fuhr, D.E. Kelleher, and H. Felrice)
• Collision Rates in Bose-Einstein Condensates. Atomic collisions are crucial in determining the evaporative cooling rates that lead to nanokelvin temperatures and Bose-Einstein condensation of cold atomic gases of ²³Na or ⁸⁷Rb atoms; collisions also control the stability and lifetime of condensates. Photoassociation spectra of trapped atoms permit the most accurate measurement of scattering lengths and other atom interaction parameters that characterize evaporative cooling kinetics Daltonand condensate properties. We have therefore constructed quantitative quantum mechanical computational methods for interpreting such spectra. Using experimental data provided by the Laser Cooling Group, we have obtained a very accurate value for the scattering length of the Na ground state F=1, M=-1 hyperfine component. We have also calculated the collisional relaxation rates for various combinations of hyperfine components of ²³Na or ⁸⁷Rb atoms in the limiting case of weak magnetic trapping fields. Our results are consistent with recent observations on ²³Na and ⁸⁷Rb condensates. (P.S. Julienne, F. Mies, E. Tiesinga, and C.J. Williams)
• Complex Quantum Nanostructures. As semiconductor nanotechnology develops, the nanostructures being fabricated become more complex, with complicated geometries and strong coupling between structures. For example, complex multilayer quantum dot structures, called quantum dot quantum wells in analogy with quantum well structures, can be made by chemical growth techniques. Such structures may be viewed as a generalization of the usual concept of an atom or molecule, whereby the energy states and optical properties are controlled by the design of the structure. We made a special effort to assess the effect of electron-hole correlation in quantum dot quantum wells and T-shaped quantum wires. We find that simple, effective mass models for the single-particle states, combined with a full three-dimensional treatment of pair-correlation, provide accurate results for pair state transition energies in these structures. Proper inclusion of the pair interaction is critical. For T-shaped quantum wires, we find that the confinement is much weaker than previously inferred from the experimental data. The confinement in these structures is not strong enough to quantum confine the electrons and holes. Rather, there is significant pair interaction and correlation in all directions, which must be taken into account in future experimental attempts to fabricate T-shaped structures with enhanced binding. (G.W. Bryant and P.S. Julienne)
• Theory of Near-Field Optical Microscopy. The diffraction limit in optics, once considered the ultimate resolution of any optical system (such as the light microscope), can be overcome by exploiting the properties of the "near" optical field. However, a serious drawback of near-field optics is the complexity of the near-field interaction between probe and sample. We are working in collaboration with the Optical Technology Division to develop near-field scanning optical microscopy (NSOM) as a nanoscale metrology tool. We are developing theoretical models for NSOM that fully account for the far field propagating through the probe that is connected to the light source or detector, the near fields that couple the sample to the probe, and the near fields that couple parts of the sample together or the sample to the substrate. In our first effort we have modeled the NSOM images obtained by probing nanochannel glass arrays. We include a Bethe-Bouwkamp model for the tip, exact calculations for the bulk photonic modes of the nanochannel glass, and full treatment of the tip/sample coupling at the surface, to calculate the optical intensity gathered by the collection optics. With this model we can accurately reproduce the observed NSOM images. Our theory allows us to directly determine the sensitivity of the images to tip/sample separation, probe polarization, aperture size, and other details of the probe fields. This information is critical for developing the understanding of NSOM metrology needed to interpret images and extract information from them. (G.W. Bryant and P.S. Julienne with L. Goldner and E. Shirley of Division 844)
• Fundamental Constants. Work on the new least squares adjustment of the fundamental constants to produce CODATA recommended values is progressing well and is expected to be completed in FY97. As part of this project, we have carried out a thorough review of the theory of the electron anomalous magnetic moment, the muon anomalous magnetic moment, and the hyperfine splitting in muonium. This theory is critical for the determination of the fundamental constants from the corresponding measurements. Also, there is a bibliographic database on fundamental constants available on the Physics Laboratory WWW server. (P. Mohr and B. Taylor of Div. 840)

• Accurate Electron Impact Ionization Cross Sections for Molecules and Radicals. We have developed a new theory for calculating electron-impact total ionization cross sections for atoms and molecules. Our Binary-Encounter Bethe (BEB) theory uses orbital energy (in lieu of the binding energy), kinetic energy, the occupation number of each molecular orbital in the ground state, and an analytic formula to generate total ionization cross sections of molecules. The resulting cross sections are in excellent agreement (within 10%) with known, reliable experimental data from threshold to a few keV in incident energies, for a wide range of molecules, H₂ through SF₆. The BEB theory works as well for radicals. Cross sections for over 50 molecules and radicals have been calculated, particularly those for air pollutants and plasma etching of semiconductors, as well as hydrocarbons in tokamaks. These data will be made available to fusion plasma and plasma etching modelers through the Physics Laboratory WWW server. (Y.-K. Kim, M.A. Ali, W. Hwang, and E. Rudd)
• Corrections to Relativistic Atomic Calculations. We have found that relativistic "multiconfiguration" wave functions sometimes do not have the correct nonrelativistic limit. This results in inaccurate transition probabilities for weak or nonrelativistically forbidden transitions. This incorrect limit has been overlooked in the past, causing misleading published results. We are now developing a general method to correct the situation that will apply to a wide range of configurations. Such a method will lead to more reliable transition probabilities for spin-forbidden transitions, e.g., the 2s^{2 1}S₀ → 2s2p ³P₁ transition of Be-like ions. (Y.-K. Kim and J.-P. Desclaux)
• Nanoscale Surface Modification with the Highly Charged EBIT Ion Beam. Our electron beam ion beam trap (EBIT) has been equipped with an ion beam extractor to provide ions with extremely high charge states (Q>40+) for use in ion-surface interaction studies. The ion beam line delivers continuous and short-pulse beams with orders of magnitude more flux than that of the only other EBIT beam line in the world. We are presently developing novel methods of nanoscale surface modification. A molecular dynamics simulation of the hypothesized ion-induced surface Coulomb explosion has been carried out (Fig. 1), and laboratory data are being collected from a series of experiments (L.P. Ratliff, J.D. Gillaspy, and D. Parks)



Figure 1. Molecular dynamics simulation of a surface Coulomb explosion at the surface of a silicon lattice (shown in cross section as a crater is being formed). The motion is tracked in sub-femtosecond time steps, as nearly 35,000 atoms interact with each other via realistic potentials. An explosion such as this is believed to be initiated by a single incident highly charged ion.

• Metastable Atom Lithography. In a collaboration with researchers from the Electron and Optical Physics Division and Harvard University, we have provided the first demonstration of a new type of microlithography in which excited atoms of noble gases are used to pattern silicon. The technique relies on the local deposition of the atoms' internal energy, stored in an excited metastable state, to expose an ultra-thin, self-assembled monolayer resist. After projecting the metastable atoms through a grid and onto the resist, chemical processing was used to produce a high-resolution image of the grid in silicon. The 100 nm edge roughness of the image (Fig. 2) was limited by the grid, not the metastable exposure process itself. The resist is sensitive to exposure by as little as about one metastable atom per resist molecule. (J.D. Gillaspy, S.L. Rolston and W.D. Phillips with J.J. McClelland of Div. 841)



Figure 2. This grid-pattern in silicon was the first demonstration of metastable atom microlithography.

• Lifetime Measurement of Magnetic Dipole Transitions with EBIT. We have measured the lifetime of a magnetic-dipole transition in Xe⁺³² by using a novel technique, the magnetic trapping mode of EBIT. The measured lifetime is of the order of milliseconds. Such long lifetimes of highly charged ions have been observed previously only by using storage rings or accelerators. (F.G. Serpa and J.D. Gillaspy)
• Characterization of the GEC RF Reference Cell. We have observed the spatially resolved optical emission (OES) and laser induced fluorescence (LIF) from a pure SF₆ discharge. SF₆ and other fluorine containing gasses are used in the etching of silicon and tungsten. Because SF₆ is an electronegative gas, large numbers of negatively charged ions exist within the plasma. The results of this investigation, combined with electrical measurements and ion-energy mass spectrometry, have proven to be a more comprehensive and consistent characterization of 13.56 MHz discharges in SF₆ than has been possible from previous, single-technique measurements. These results should prove useful both in validation of theoretical models for SF₆ discharges and in the interpretation of other experimental results. This is primarily due to the fact that these measurements were performed using the GEC reference cell, which exists in many other laboratories and for which proven procedures have been developed. The OES and LIF spatial intensity distribution profiles exhibit sharp maxima in narrow regions in front of the electrodes, consistent with a constricted sheath. The vertical OES and LIF profiles exhibit secondary maxima, suggestive of a double-layer formation, which has been previously predicted.
  
  We have also developed a new plasma uniformity monitor for axially symmetric plasmas. This device is being tested on the GEC rf reference cell, but is intended for use on commercial etching systems which have limited optical access. (E. Benck, J. Roberts, and A. Schwabedissen)
• Characterization of an Inductively Coupled Plasma (ICP) Cell. Our new inductively-coupled rf-powered version of the GEC rf reference cell is fully operational. This new class of high density, low pressure plasma sources is becoming increasingly important to meet the increasing demands of reducing the dimensions of etched structures. Langmuir probe measurements have demonstrated the effects of a variety of different feed gasses on the electron density and energy distribution function within the inductively coupled plasma (ICP), and time-resolved optical emission spectroscopy has been used to study the sheath region adjacent to an rf-biased electrode in an argon discharge. We have found that, unlike capacitively coupled discharges such as the GEC reference cell, rf biasing of electrodes in an ICP provides control, independent of the plasma production, of the ion energies involved in the etching process. (E. Benck, J. Roberts, and A. Schwabedissen)
• Accurate Atomic Transition Probability Measurements. Atomic transition probabilities and branching fraction data are of importance to industrial applications of plasma physics (e.g., the lighting industry and plasma processing), astrophysics, and basic atomic theory. Atomic structure calculations for light elements have recently achieved an accuracy that approaches that of the best available experimental data. In response to this development, we have improved the uncertainty of our measurements (better than 10%) by the use of modern experimental techniques such as photon counting and Fourier transform (FT) spectrometry. We have measured relative transition probabilities of O I, and branching fractions of weak intersystem lines of N II, Ne II, and Fe I, by observing spectral lines emitted by a wall-stabilized arc and from a hollow cathode lamp. (J.M. Bridges, U. Griesmann, and W.L. Wiese)
• Development of New Infrared Source. The Atomic Physics and Optical Technology Divisions have collaborated on a project resulting in the development of a new, brighter IR source. This new source yields better signal-to-noise ratios, and therefore higher accuracy, in IR measurements. The source is a stabilized argon arc, which has been characterized in the spectral range from 1 µm to 20 µm. Its radiance was calibrated and found to be approximately equal over much of this range to that of a 10,000 K blackbody. A high-resolution spectrum taken with a FTIR instrument shows mostly line emission below 5 µm, and pure continuum between 5 µm and 20 µm. The stability and geometrical properties of the radiance were determined, as well as its dependence on pressure and current. As a result of this project, this source is now being used in calibrating IR detectors, as well as in projects aimed at advancing IR measurements and technology. (J.M. Bridges and A. Migdall of Div. 844)
• Deep-UV Refractive Index Measurements. We have teamed with the Optical Technology Division to make high-accuracy, deep-UV index-of-refraction measurements of materials considered for use in the optical components of photolithography steppers for future-generation IC fabrication. This is part of a collaborative project with the MIT Lincoln Laboratory and SEMATECH. To meet the immediate need for accurate values of the index of refraction of fused silica at 193 nm, we have upgraded a precision refractometer, including precisely characterized UV line sources, to enable minimum-deviation-angle, refractive-index measurements, accurate to 1 part in 10⁵, with a temperature control of 0.1 °C. The system also enables the determination of the temperature coefficient of the index. These measurements are needed to design the transmissive optics for the steppers for 0.18 µm minimum-feature-size IC fabrication (1 Gbit DRAM), which is scheduled by the SIA roadmap for production by the U.S. semiconductor industry beginning in 2001. We have begun these measurements on fused-quartz samples provided via SEMATECH by several of the potential suppliers. We will also make similar measurements on calcium fluoride and other deep-UV optical materials being considered for 0.18 µm and shorter wavelength lithography. (J.H. Burnett and J.R. Roberts)
• Bragg Scattering from Optical Lattices. Atoms laser cooled in intersecting laser beams become trapped in the periodic light-shift potentials created by the interference of the laser beams. This optical lattice holds the atoms at precisely periodic locations, as in a solid crystal, but with a lattice spacing on the order of optical wavelengths.

  Just as x-rays Bragg-reflect from regular crystal planes, light tuned near an atomic resonance can Bragg-reflect from our optical lattices. Among the properties of this Bragg scattering are that it depends sensitively on satisfying the wavelength and angle conditions for coherent addition of the waves reflected from successive atomic planes and that its amplitude depends on how well localized to those ideal planes the atoms are. Using this latter fact, we have studied the time-dependent evolution of atomic motion in the optical lattice. For example, for a given optical lattice, atoms with a higher temperature will be less tightly localized at the precise lattice sites, the periodically located potential minima, and will give less Bragg scattering. By measuring the Bragg scattering as a function of time after the lattice is switched on, we have measured for the first time how fast atoms are laser cooled. We find that the cooling rate is proportional to one parameter, the photon scattering rate, over a wide range of lattice laser detunings, shown in Fig. 3.

  

  Figure 3. The product of the photon scattering rate Γ′ and the cooling time constant τ for various lattice laser detunings in 1-D. The filled squares are experimental results, and the open diamonds are quantum Monte Carlo simulations for the same parameters.

  In both 3-D and 1-D we find this proportionality, although the cooling is about six times slower in 3-D. In 1-D, where calculations are possible, we find essentially perfect agreement with our fully quantum treatment of both internal and center-of-mass atomic motion. Although this is satisfying, the result is in disagreement with the semiclassical picture of laser cooling that has guided the thinking of the community since 1989. How to provide a new physical picture consistent with these as well as earlier results, and how to understand the differences between 1-D and 3-D remain open questions. (G. Birkl, G. Raithel, M. Gatzke, S. Rolston, I. Deutsch, and W. Phillips)
• Coherent Atomic Wave Packet Motion. We use Bragg scattering to study induced and driven oscillations of atoms in an optical lattice. By suddenly increasing the intensity of the lattice laser beams, we compress the trapped atoms toward the minima of the potential wells. The atoms then oscillate about the potential minima, so that the average atom cloud "breathes" at twice the oscillation frequency. We see this as an oscillating Bragg signal that decays due mainly to the anharmonicity of the potential wells, as seen in Fig. 4.

  

  Figure 4. Oscillations of the mean square displacement Δξ² of atoms from the potential minima. The oscillations follow a sudden compression where the intensity of the lattice light is increased by a factor of 4. The oscillations decay quickly due to anharmonicity while Δξ² returns to equilibrium more slowly due to laser cooling.

  In addition to such sudden compression, we have induced oscillations by continuously modulating the intensity of the lattice beams. This parametric driving of the atomic motion produces results similar to quadrature squeezing of light, in that the position and momentum spreads of the atoms are modulated periodically and in quadrature, although we have not yet achieved squeezing below the standard quantum limit. (S. Rolston, G. Raithel, G. Birkl, and W. Phillips)
• Atomic Fountain Clock. In collaboration with the Time and Frequency Division we have built and installed an atomic fountain frequency standard operating with laser-cooled Cs atoms. This standard is just beginning operation and we are using it to study optimal ways of launching, cooling, and probing the atoms. Cooling to a few microkelvin is a pre-requisite for a good fountain clock, but atoms even colder than this would be a significant advantage. Among the strategies being studied for further temperature reduction are adiabatic expansion in optical lattices, and sub-recoil Raman cooling. (C. Ekstrom, W. Klipstein, M. Golding, S. Rolston, and W. Phillips)
• Optically Controlled Biological Collisions. Collision and adhesion of biological particles such as cells and pathogens are a process of fundamental interest in biomedicine. We use optical tweezers to hold and manipulate particles, producing controlled collisions for quantitative study of adhesion. We coat glass spheres with viruses and collide them with red blood cells, measuring the sticking probability in the presence of chemicals that inhibit adhesion. In collaboration with the Whitesides group at Harvard, we measure inhibition constants that were unmeasurable by any other techniques. (K. Helmerson, R. Kishore, and W. Phillips)
• Bose-Einstein Condensation. Building on the success of BEC in the Quantum Physics Division, we have begun an effort to Bose condense Na atoms as a source of coherent atoms for atom optics experiments. Our approach is a hybrid of successful efforts at JILA and MIT. We load cold atoms into a magneto-optical trap (MOT) using the Zeeman slowing technique first developed in our group. Co-located with the MOT is a time-orbiting-potential (TOP) trap, modified from the JILA design to make transfer from the MOT more efficient in phase space. As of this writing, we have trapped about 5 × 10⁸ atoms in a MOT cloud about 1 mm in diameter, a density that should be sufficient for evaporative cooling and condensation of a cloud with close to a million atoms. (R. Thompson, A. Steinberg, M. Gatzke, G. Birkl, S. Rolston, K. Helmerson, and W. Phillips)
• Photoassociative Spectroscopy. When two atoms collide at very low velocity in the presence of a light field, they can absorb a photon during the collision and become bound together as an excited molecule. This photoassociation of ultracold atoms allows the study of molecular states that are otherwise inaccessible. In particular, we can study "purely long-range" molecular states in which both the inner and outer turning point of vibrational motion are larger than tens of atomic units, well beyond the range of exchange and chemical forces. Such molecules are particularly easy to understand because their properties are determined almost entirely by long-range dipole-dipole forces. By studying the spectra of such molecules, we have obtained the best measurements of the ground state scattering length and the atomic radiative lifetime for Na, and we have seen for the first time the effect of radiative retardation on a molecular spectrum. A portion of such a molecular spectrum is shown in Fig 5. Because the excitation probability of the various rotational states is determined by corresponding angular momentum partial-wave ground state scattering wavefunctions, this spectrum allows us to pinpoint the location of the s-wave nodes and therefore the scattering length. This, in turn, determines both the elastic collision cross section at low energy and the mean-field interaction energy, quantities of central importance in the study of Bose-Einstein condensation of atomic gases.

  

  Figure 5. Spectrum of the vibrational ground state of the 0_g^- potential of singly excited Na₂. Detection is by ionization of the excited molecule, as shown in the inset. The relative heights and absolute positions of the peaks give information about the atomic radiative lifetime, scattering length, and about radiative retardation effects.

  The molecular binding force between the atoms depends on the same dipole matrix element that determines the atomic radiative lifetime of the resonance level, so that measurement of the spacing between the vibrational levels allowed us to determine the atomic lifetime to an accuracy of 0.1%, the best Na lifetime determination ever. Because the fields binding the atoms propagate at light speed over distances that are unusually large for molecules, there is a shift in the energy levels from the retardation. We measure the shift of v=0 to be 122(10) MHz, compared to the calculated shift of 121 MHz. (P. Lett, K. Jones, L. Ratliff, and W. Phillips)
• Three Tunable X-Ray Spectrometers Delivered to NASA Programs. NASA missions in x-ray astronomy required widely tunable monochromatic x-radiation for pre-flight calibrations and subsystem development. The AXAF version of one of these instruments, covers the range from 0.3 keV to 12 keV; it was installed at Marshall Space Flight Center (MSFC) in late 1995. This instrument uses a MSFC supplied rotating anode source and provides monochromatized beams at the entrance to an 875 m long vacuum whose exit chamber contains the AXAF telescope. The NIST monochromator may be visited at the URL: http://wwwastro.msfc.nasa.gov/xray/xraycal/xssrr/dcm/. (J.-L. Staudenmann, L.T. Hudson, A. Henins)
• Spectrometric Standardization of Mammographic X-Ray Sources. The NIST curved-crystal x-ray spectrometer has now been tested in a variety of medical research environments. This device (patented and licensed for commercial development) provides high voltage calibration and spectral characterization for mammographic x-ray sources in support of diagnostic radiation quality requirements. A significantly overdetermined calibration is obtained using the well-known location of K absorption edges of a few metal foils. Data fitted to the formal dispersion function give residues below 0.1 kV, i.e., well below currently understood clinical significance. Additionally, the acquisition of the spectral distribution allows improved modelling and refinement of the mammographic paradigm. In the framework of a supporting grant from the Army's Breast Cancer Research Program, cooperative studies have begun at the Center for Devices and Radiological Health (a component of the Food and Drug Administration), the University of California-Davis Medical Center, and Radcal Corporation, the commercial licensee. Through Radcal's efforts, the first commercial model (aimed at the research market) is to be available by the end of 1996. A summary publication has recently appeared in the journal Medical Physics. (L.T. Hudson, R. Deslattes, and A. Henins)
• Lattice Changes in Si Epilayers and Si Substrates. Certain high-performance microprocessors are fabricated using epitaxially deposited, thin Si layers grown on highly doped Si wafers. In at least one case, it was found that material from different vendors gave differing device yields, although all sources met stated electrical criteria and appeared consistent using the manufacturer's current metrology toolbox. We examined substrate lattices using high-resolution lattice parameter comparison techniques developed in the Group. The measurements showed considerable variation (fractional changes of up to 5 × 10^-5) among the sources and even a rather large difference between nominally identically processed samples from 20 cm and 15 cm boules. In a second set of measurements, the lattice constant of the epilayer was measured with respect to that of the substrate using conventional high-resolution double-crystal diffractometry. Fractional differences obtained in these measurements range from 2 × 10^-5 to 1 × 10^-3. The two measurements can be combined to obtain the lattice parameter of the epilayer itself. There is some indication that the lattice parameter differences seen here (somewhat smaller than can be resolved by conventional diffractometry) correlate with device yield although the needed control studies have not yet been undertaken. (R. Deslattes, J.E. Schweppe, L.T. Hudson, and A. Henins)
• A New X-Ray Optics Geometry for Powder Diffraction. The need to recertify powder-diffraction standards required realization of accurate powder-diffraction measurements with a parallel x-ray beam. In addition, accurate determination of the diffraction angle zero demands an instrument operable in mirror symmetric configurations. These geometric constraints and the accuracy needs led us to construct an entirely new apparatus for the measurement. The restriction to parallel beam geometry leads to a significant loss in signal levels in comparison with conventional (focussing) geometries. One solution would be to use a synchrotron radiation source. This clearly addresses the intensity problem but requires separate determination of the input wavelength and entails establishing a good metrological environment on the floor of an accelerator facility. In a recent development, we have been able to obtain incident beam intensity from a conventional (2 kW) diffraction tube comparable to that available at a synchrotron radiation powder-diffraction beamline. The key to this development is a newly realized combination of a graded spacing multilayer paraboloid with a flat multilayer optic having a spacing near the mean value of the graded spacing mirror. The beam from this mirror pair has a divergence of about 107 µrad and provides a photon rate exceeding 1 GHz at the sample. The divergence of this beam in the orthogonal direction is restricted to 10 mrad. Preliminary powder diffraction scans indicate good peak to background ratios, symmetric profiles, and counting rates sufficient to proceed with the needed measurements in the more benign environment of the Gaithersburg laboratories. As reported in a recent publication, we have shown that the lattice spacing for powder Si can be measured with our present apparatus with a relative uncertainty at the 2 × 10^-6 level. This work is supported in part by the NIST Standard Reference Material Program. (J.-L. Staudenmann, L.T. Hudson, A. Henins, and R. Deslattes)
• Field Emitter Arrays for Flat Panel Displays. Field emission displays (FEDs) are being pursued by US companies as a leap-frog flat panel technology that has the potential of increasing the U.S. flat panel display market share from the current 0.4%. The advantages of FEDs over other flat panel display technologies include CRT-like display quality, a wide operating temperature range, and low power consumption. Research on the physics of field emission cathodes in support of FEDs is performed with funding by the ATP program to provide technical support to two ATP-funded companies: FED Corp. and SI Diamond Technologies, Inc. For FED Corp., we provided simulation of electron emission and electron trajectories from gated field emitter arrays and experimental testing of FED Corp's displays. For SI Diamond, we performed electron trajectory simulations and will fabricate and characterize well-controlled diamond-like-carbon thin films. Additional support from DARPA has produced the fabrication of field emitter cathodes with integrated lenses coplanar to the gate electrode for collimating electron beams from tips of gated field emitter arrays. Linear planar lens electrodes on both sides of a line of emitters have been demonstrated to provide focusing by application of appropriate voltages. The chips were tested in a high vacuum chamber where the emitted electrons were accelerated to a phosphor screen anode parallel to the chip. The resulting image was magnified in a telemicroscope and captured by a CCD camera. With focussing, the resulting line image was less than 0.035 mm wide at 10 mm to 20 mm from the anode, while the unfocused image is about 100 times larger. This is the first demonstration of electron beam focusing from field emitter arrays with an integrated planar lens design in a well documented study with calibrated image registration. (C.M. Tang and J. Pedulla)

Most Recent Technical Activities   |   Archive of Technical Activities