Atomic Physics Division

Technical Highlights

• Observation of Spectra of Importance for Tokamak Diagnostics. In a continuing collaboration with the University of Texas, we have excited spectra of highly ionized atoms needed for tokamak diagnostics by laser ablation of metallic samples into the TEXT tokamak plasma. Analysis of our spectra of highly ionized tungsten obtained in this way has proved unambiguously the correctness of our earlier interpretation of spectra of W from the ORMAK tokamak at Oak Ridge National Laboratory as almost entirely attributable to about 10 strong lines of W²⁷⁺, W²⁸⁺, and W²⁹⁺, rather than to an unresolved array of several thousand lines of W³¹⁺ to W³⁴⁺, as had originally been hypothesized (Figure 1). These results can be used to recognize impurity ions of W in tokamak plasmas and to calculate radiation losses in tokamak plasmas due to the presence of W, a possible material for use in the next generation ITER tokamak.

  

  Figure 1: Tungsten radiation at approximately 50 Å observed with: A -- the ORMAK; B -- the PTL; and (top) the TEXT tokamak.

  Our tokamak observations of Fe were combined with observations of laser-produced plasmas obtained with the Glass Development Laser (GDL) at the University of Rochester to determine an improved system of energy levels and ionization energy for Li-like iron, Fe²³⁺. An extensive spectral analysis was completed for Kr-like niobium, Nb⁶⁺, also of importance for use in future tokamaks. (J. Reader, J. Sugar)
• High-Resolution Measurement of Heavy-Element Spectra for Industrial and Astronomical Applications. We have observed and measured the spectra of zirconium and mercury atoms and ions in various wavelength regions over the range 50 nm to 630 nm. All prominent Hg I lines between 180 nm and 630 nm were observed with a Fourier-transform spectrometer. The data are needed for characterization of pencil-type mercury discharge lamps that are widely used in industry and science for wavelength calibration of spectrometers. We will also evaluate the suitability of these lamps for calibration of the sensitivity of spectrometers as a function of wavelength. We are working on computer software to model the complex Hg I lines and extract hyperfine constants and isotope shifts.
  
  Other measurements and analysis of the spectra of Zr and Hg ions are concentrated on Zr III and Hg III; data on these spectra are especially needed for interpretation of spectra of chemically peculiar stars obtained with the Goddard High-Resolution Spectrometer on the Hubble Space Telescope. (J. Reader and C.J. Sansonetti)
• Precise Tests of Atomic Theory in Be III. Accurate measurements in simple atomic systems provide the most sensitive tests of relativistic and quantum electrodynamic (QED) theory in atomic physics. An excellent new test for two-electron systems is provided by a recent measurement of the 1s2s³3S - 1s2p³P transitions in doubly ionized beryllium. We collaborated in this work with Richard Holt and his co-workers at the University of Western Ontario. Be III spectra were recorded by fast-ion-beam laser spectroscopy. Absolute calibration of the spectra was obtained from uranium lines observed by laser optogalvanic spectroscopy in a U-Ne hollow cathode lamp. We measured the uranium lines with an uncertainty of 4 parts in 10⁹ in our laboratory at NIST.
  
  The wave numbers for the Be III 1s2s³3S - 1s2p³P transitions were determined with an accuracy of 1 part in 10⁸, nearly three orders of magnitude more accurate than previously reported. The results confirm that relativistic terms of order α⁴Z⁴, which have not yet been calculated, make a significant contribution to the energies of these Be states. The size of these contributions can be estimated from the data to be about 100 times the experimental uncertainty. Once these relativistic terms have been computed, the Be III results will provide a test of QED calculations in this two-electron system at the level of a part in 10⁴. (C.J. Sansonetti and J. Gillaspy)
• Acquisition of High-Resolution Fourier Transform Spectrometer. NIST has concluded arrangements to acquire from the Los Alamos National Laboratory one of the world's most powerful FTS instruments. This spectrometer covers the range 200 nm - 18.5 µm with a resolution of 0.0025 cm^-1 and has demonstrated a signal-to-noise ratio of 10⁵. It will complement our existing high-resolution grating and laser-spectroscopic instrumentation, giving NIST an unparalled spectroscopic capability from the extreme UV to the IR. Design work is in progress to modify our facilities to accommodate this instrument at NIST. The spectrometer will be moved to NIST in 1994, and will be fully operational early 1995.
  
  Acquisition of the FTS will permit expansion of Atomic Physics Division programs in a number of areas including analysis of complex spectra for applications in high-efficiency lighting, determination of precise atomic benchmark data, and more rapid and efficient collection of data needed for plasma processing, laboratory, and space astronomy applications. Other groups at NIST have expressed interest in using the FTS to develop new methods of precise elemental analysis by optical isotope dilution spectroscopy, to refine procedures for metals analysis in glow discharges, and to observe molecular spectra at very high resolution. We have also had a number of inquiries from scientists in universities and industry who are eager to propose collaborative projects using the instrument. (C.J. Sansonetti and J. Reader)
• Atomic Spectroscopic Databases. The Data Centers on Atomic Energy Levels and Atomic Transition Probabilities have built two databases in response to our current perceptions of the needs of a broad community of users.

  One database uses the ORACLE relational database management system and now includes essentially all data on energy levels, wavelengths, and transition probabilities produced at NIST during the past 25 years. The main components of the database are wavelengths, energy levels, and transition probabilities; the system allows retrieval, for example, by spectrum, wavelength, multiplet, accuracy of transition probabilities, etc. All newly compiled data are added to the database as completed, and some unpublished extensions and updating are also incorporated. We expect to complete fairly soon the interfacing and other work needed to begin operating this database as a server node in the NASA sponsored Astrophysics Data System.

  A second database has been separately designed and written, complete with user interface. The current prototype contains lines, levels, and transition probabilities for all ionization stages of four elements. The data reside in redundancy-free tables and are integrated: for example, a search for wavelengths automatically brings up all the level and transition probability data for those wavelengths. This self-contained database, designed particularly with industrial users in mind, requires a modest amount of disc space; it can be distributed on floppy discs and is highly portable to different computer environments. We plan to include the database in a group of Physics Laboratory databases to be made widely available on the Internet (see below). (A. Robey, D. Kelleher, W. Martin, G. Dalton)
• World Wide Web Server for Physics Lab. A World Wide Web server for the Physics Laboratory has been installed. This server provides information over Internet to the public through client software such as Mosaic which is freely distributed by NCSA (the National Center for Supercomputer Applications). Documents are being prepared for access over the Web to provide both critically evaluated data and general information about the Physics Laboratory. The documents in preparation include databases that provide atomic wavelength and intensity information in both tabular and graphical form, listings of references to articles on critically evaluated data, descriptions of programs in the Physics Laboratory, and published Physics Laboratory articles of wide interest.
  
  Also, work is being done to make the atomic data that are critically evaluated in the Atomic Physics Division widely available to the public over Internet. A database containing energy level and transition probability data for a number of elements is now available for use by remote log-in. Users can log in, search by a menu driven set of commands, and save sets of data and retrieve them through an anonymous ftp server. (P. Mohr, D. Kelleher)
• Atomic Transition Probability Tables for Carbon, Nitrogen and Oxygen and Other Critical Compilations of Spectroscopic Data. We completed the evaluation and compilation of new atomic transition probability data for the elements carbon, nitrogen, and oxygen in all stages of ionization, are now going through the final editing process, and expect to publish these data, covering about 20,000 spectral lines -- roughly 1000 lines per spectrum -- in late 1994. The new data tables will be about 10 times as large as the earlier ones, and the data quality is typically improved by factors of two to five.
  
  We also completed new compilations of the energy levels for all the zinc spectra (Zn I - Zn XXX) and compilations of wavelengths with energy-level classifications for the high-ionization spectra of manganese (Mn VII - Mn XXV). Similar new wavelength compilations for the high-ionization spectra of iron and krypton are almost ready for publication. We are also preparing for publication the results of a very complete compilation of the wavelength and energy level data for Si I. Work on the other spectra of silicon and on chlorine spectra is underway. (W. Wiese, J. Fuhr, J. Sugar, W. Martin, A. Robey)
• Major Improvement of Relativistic Atomic Structure Code. A major improvement was put in the existing relativistic atomic structure code in collaboration with J.P. Desclaux and P. Indelicato (France). This improvement accounts for the fact that the one-electron orbitals in the initial-state and the final-state wave functions are not orthonormal in high-precision calculations because the wave functions are optimized separately to gain a better accuracy. This new capability was used to calculate lifetimes of the xenon atom measured by the laser trapping group. At present, the theory predicts 50 s to 70 s while the experiment yields 43 s for the lifetime of the weakest transition. Better agreement is found on stronger transitions. Work is in progress to resolve this difference. (Y.-K. Kim)
• New Theory to Calculate Reliable Electron-Impact Ionization Cross Sections. A new theoretical method to predict electron-impact ionization cross sections for neutral and ionized atoms was developed in collaboration with M.E. Rudd (University of Nebraska). This method combines the binary-encounter theory and the dipole interaction, does not contain any adjustable parameters, and has been verified to provide very reliable (10 percent or better) cross sections for light atoms and molecules for incident energies of 10 eV to 10 keV. This new method uses atomic data that can easily be generated with the Hartree-Fock method and photoionization data, which can either be experimental or theoretical. The new method will be used to provide ionization cross sections for atoms and ions needed in plasma processing and magnetic fusion modeling. For instance, this new method will be used to generate ionization cross sections mentioned in the proposed CRADA with GE (see below). (Y.-K. Kim)
• Precise Corrections for Finite Nuclear Size. Work on precise calculation of the finite nuclear size correction to the self energy has been successfully carried out in collaboration with G. Soff (Germany). New algorithms were developed for this calculation. Previous estimates of the effect were significantly modified by our results, particularly at high Z (fermium), where our results are accurate to about 0.003 percent and differ from previous predictions by about 25 percent. A calculation of energy level corrections due to the two-photon Feynman diagrams for the ground state of high-Z, two-electron atoms has been completed in collaboration with W. Johnson, J. Sapirstein (University of Notre Dame), and S. Blundell (France). These corrections are important in high-Z ions, and our work provides the first complete calculation. (P. Mohr)
• Characterization of the GEC RF Reference Cell. We have observed for the first time some unique effects on the shape of the H_α spectral line in an admixture of Ar in hydrogen for an rf plasma discharge. The plasma source was the Gaseous Electronic Conference (GEC) RF Reference Cell. The introduction of Ar in a pure H₂ plasma increases the number of fast neutral H atoms as is evidenced from the increase in the intensity of the broad component of a two-component Doppler broadened H_α line profile.

  In the case of a pure hydrogen plasma many processes have been discussed to explain the characteristics of the profile of hydrogen spectral lines. In order to explain the enhanced Doppler component of the broadened hydrogen Balmer lines in our experiments, the additional process of charge-exchange is introduced. This is due the charge-transfer between metastable argon ions and hydrogen molecules to form the hydrogen molecular ion and neutral argon. Our results indicate that Ar plays a significant role in the increase of the number of excited hydrogen atoms (n = 3) for an rf discharge.

  The spectral line profiles emitted from the plasma were observed in the direction parallel to the electrode surface, i.e., in a direction normal to the electric field. The ratio of intensities between the wide and narrow components varied as a function of the position between the two electrodes. The analysis yields an average temperature of 23.8 eV for H atoms associated with the wide profile component and 0.22 eV for atoms identified with the narrow component. Because of the Doppler shapes of the H_α profiles and because both the narrow and wide profiles were centered at the same wavelength, a random velocity distribution for both velocity components is indicated. Since the observations were made perpendicular to the applied electric field, these observations indicate there were no directional velocity effects from the applied electric field. However, the data obtained in this experiment cannot quantify the contributions among the many possible ways to produce fast hydrogen atoms. (J. Roberts)
• Electron Beam Ion Trap (EBIT). A new facility to create, trap, and study highly charged ions was brought into operation. This EBIT facility will allow us to greatly expand the range of our in-house atomic data production capabilities as well as open up possibilities to explore technological applications in the area of ion-beam lithography. By focusing an intense electron beam with superconducting magnets and high voltage electrodes, the EBIT can strip most or all of the electrons from a collection of atoms and confine the remaining ions in an electromagnetic bottle. The nearly monochromatic electron beam can then be tuned to probe the trapped ions in situ. The capabilities of the EBIT were demonstrated by stripping 46 electrons from Barium atoms to create ions with a neon-like electronic configuration. The strong nuclear charge of the neon-like barium results in a highly compact ion, only about 0.02 Å in diameter in its ground state. Consequently, the spectral resonance lines are located in the x-ray regime. Figure 2 shows the spectrum of x-ray photons emitted from EBIT and observed with a high efficiency solid state detector. (J.D. Gillaspy)



Figure 2: Photon spectrum from Ba⁴⁶⁺ in EBIT.

• UV Lithography for Semiconductor Manufacturing. With the Radiometric Physics Division, we have been working on a project funded by SEMATECH to improve radiometric measurements in semiconductor manufacturing. Etching of fine patterns in semiconductor wafers is carried out by photolithography, using bright ultraviolet (UV) light sources. High quality reproducible results are obtained by irradiating the wafers with the correct radiation dose. This is, however, a difficult procedure due to the unstable output of these intense light sources. Present techniques rely on a time-consuming trial and error method of getting the correct dose. Attempts are now being made to rely on accurate radiometric measurements as a better method. For this purpose, we have designed and furnished to SEMATECH two calibrated UV spectroradiometers. These instruments were made to be used with a new "deep UV" manufacturing tool which operates at 250 nm in wavelength. No commercial instruments were available for this wavelength. The spectroradiometers replace a previously furnished prototype instrument, and have improvements in sensitivity, accuracy and wavelength coverage. We are also furnishing a calibration device for the spectroradiometers. This consists of an additional spectroradiometer connected to an arc lamp. Using this device, the spectroradiometers used with the manufacturing tools can be calibrated periodically to insure that accurate measurements are being made even if changes occur in the measuring instruments. The control of UV exposure in semiconductor manufacturing will help to make U.S. Semiconductor companies more competitive. (J. Roberts and J.M. Bridges)
• Development of new Infrared Source. Many companies have programs in infrared (IR) technology, including development and testing of IR detectors. Measurements in the IR spectral region are usually hindered by signal-to-noise limitations. A stronger IR source than is presently available would be useful in developing, testing, and calibrating detectors, as well as in other applications where a powerful source of IR is needed. A stabilized argon arc plasma source, developed and used by us as an ultraviolet source, has been found to have a high radiance in the IR region, to. We have begun a project to characterize this arc as an IR source, and to optimize it for this spectral range. Measurements made so far have shown that the argon arc can be from six to ten times brighter than a typical source used for the near to middle IR region (1 µm - 20 µm). The investigation of the arc source in the IR will continue with measurement of the radiance as a function of parameters such as arc current and pressure, as well as stability and geometrical and spectral properties of the radiation. (J.M. Bridges)
• Branching Ratios and Transition Probabilities. An ongoing project in the Atomic Physics Division, supported by NASA, is the critical compilation of atomic transition probability data, which continue to be valuable for numerous applications in technology and science. These data are mainly furnished by calculations. Although generally quite good, in some cases large disagreements exist between independent calculations of the same quantities. Determination of the correct values for some typical atomic transitions can have a great leverage in allowing the selection of correct calculated results for many more transitions. In order to do this, we have made accurate measurements of branching ratios and relative transition probabilities for several hundred lines of neutral and singly ionized oxygen, nitrogen, and carbon. The measurements were made on lines emitted from a highly stable arc source, operating in helium with admixtures of either oxygen, nitrogen, or carbon dioxide. Such measurements were previously made with arc plasmas of pure gases, but we found that plasmas operating essentially in helium feature lower electron densities so that the emitted lines are narrow and less blended with other lines, and exhibit much less underlying continuum radiation. This allowed radiance measurements with accuracies of better than 5 %, a significant improvement over previous results. The resulting measurements are leading to greatly improved data compilations. (J.M. Bridges and W.L. Wiese)
• Coherent Momentum Transfer-Pushing Atoms with Darkness. A key process in atom optics is imparting momentum to atoms without any randomness. When light is used to push the atoms, it is essential to avoid any spontaneous emission from the atoms since its random character destroys the atomic coherence that is needed for interferometric applications. In collaboration with a theoretical group from the University of Colorado we have transfered the momentum of eight laser photons using a process that avoids spontaneous emission. In fact, the transfer is done without the atom ever being in the excited state. The key to the process is that the atom stays in a coherent superposition of groundstate sublevels that cannot absorb the laser light. The composition of that "dark" state depends on the polarization of the light. We slowly change the polarization of the light and the state of the atom (both internal and linear momentum) adiabatically follows, always staying in the dark, ground state, and never absorbing any photons. The process coherently changes both the internal state and the momentum of the atom. The absence of excitation, and therefore of spontaneous emission, preserves the coherence of the atomic motion and allows the atoms to interfere. We hope to use this process to perform the function of a mirror or beamsplitter for atoms in an atom interferometer. (L. Goldner)
• Microwave Trap for Neutral Atoms. In 1985 we first demonstrated the confinement of neutral atoms in a trap made with an inhomogeneous, static magnetic field. Since then magnetic traps have been often used to store spin-polarized hydrogen in an attempt to achieve the high density and low temperature needed for Bose-Einstein condensation. Unfortunately atoms in a magnetic trap cannot be trapped if they are in their lowest internal energy state. This means, when in a collision between trapped atoms a transition to this state occurs, the atom leaves the trap. Collisions therefore limit the density that can be achieved in a magnetic trap. In collaboration with a group from Harvard we have therefore worked on a different kind of trap, and we have now demonstrated the first microwave atom trap. Such a trap is capable of trapping the atomic ground state and we hope it will lead to Bose condensation and other experimental results requiring higher atomic density. (S. Rolston, W. Phillips)
• Atoms in Optical Lattices. Laser cooling of neutral atoms is usually accomplished with intersecting laser beams directed at the atoms from different directions. The interference of these laser beams produces a pattern of varying intensity and polarization that creates a lattice of potential wells for the cooled atoms. The atoms can be cooled enough that they become trapped in the sub-optical-wavelength-sized potential wells. Last year we observed such trapping and the resulting quantized motion of the atoms in a one-dimensional situation. Now we have produced a 3-D lattice, using a special configuration of four linearly polarized laser beams. The chosen lattice maximizes the time the atoms remain trapped and allows the atoms to be cooled to just over one microkelvin, about a factor of two lower than usually obtained with laser cooling. By measuring the spectrum of light emitted by the atoms we observe the 3-D quantized motion of the atoms and determine that they are localized to within 1/20 of an optical wavelength. Control of the optical lattice should allow us to reach even lower temperatures. Such optical lattices may be useful in producing permanent structures having small feature size and high periodicity. (S. Rolston, W. Phillips)
• Lifetime of Metastable Xenon. Using laser cooling and trapping techniques, which we demonstrated on metastable xenon for the first time last year, we have measured the radiative decay rate for the 5p⁵6s ³P₂ level. Trapping and cooling and the development of a new method for determining extremely long lifetime allowed us to make this measurement. The lifetime, predicted to be on the order of 100 s, had never been determined experimentally. Vacuum UV photons from the metastable trapped atoms are observed to determine the rate of detected decay photons. Then a separate laser dumps the atoms into another state that rapidly decays by emission of a different vacuum UV photon, observed by the same detection apparatus. The ratio of the observed decay rate from the metastable state to the number of observed photons after the "dump" gives the metastable lifetime. Careful attention to a number of systematic errors allowed the lifetime to be determined as 42.9(9) s for ¹³²Xe. The discrepancy with the predicted value allowed the Atomic Physics Division theory group to discover a previously unknown difficulty in the calculations of atomic structure. (S. Rolston)
• Photoassociative Molecular Spectroscopy. When laser-cooled atoms collide in the presence of light, an excited molecule can be formed in a rotational-vibrational state that depends on the laser frequency. The resolution can be very high because the initial spread of kinetic energies of the cold, free atoms is so low. We demonstrated this photoassociative spectroscopy last year. Now we have identified and assigned several rotational-vibrational series, and extracted interatomic potential parameters. Because the initial kinetic energy of the colliding atoms is so low (less than 1 mK), the excitation can be to molecular states very close to dissociation threshold. Such high lying levels involve vibrations whose outer turning points are hundreds of atomic units and are known as long-range molecules. Besides exciting such states, we have for the first time observed the spectroscopy of "pure long range" states whose inner turning points are at 60 atomic units or more. These states, predicted some years ago, have not been directly observed until now because of the difficulty in producing them from usual ground state molecules or from thermal-energy free atoms. (P. Lett, L. Ratliff, M. Wagshul, K. Helmerson)

Most Recent Technical Activities   |   Archive of Technical Activities