Quantum Physics Division

Technical Highlights

• Technique for Measuring Sticking Probabilities as a Function of Energy Developed. A recent program targets the microscopic basis of silicon semiconductor etching. In a new type of measurement, data has been recorded for the sticking and scattering of molecular chlorine with Si(100) over a wide range of kinetic energy, such as occurs in plasma etching reactors. The results are derived from a direct comparison of the signals for the scattering of a species from silicon versus quartz, similar to the method of King and Wells, except that the scattered flux is now time-resolved in a mass spectrometer. The results show, for example, that at intermediate kinetic energies of Cl₂ (1 eV to 2 eV), a large fraction of the molecules are trapped or reacted by the silicon at low coverage. However, at higher coverage, the sticking probability decreases significantly at this kinetic energy. For lower kinetic energy species there is a substantial probability of sticking or reacting at all energies. The data in these types of measurements are correlated with the onset of SiCl_x etching products and are presently being used to assemble a complete picture of the sticking and etching dynamics.



Figure 1: Time-of-flight data for CL₂ scattered by quartz (upper traces) and silicon (lower traces) with increasing coverage.

• Progress Made in Drive to Achieve Bose Condensation. Several JILA scientists are undertaking an experimental and theoretical effort to observe and understand Bose-Einstein condensation (BEC) in a dilute gas. The basic experimental approach is to load optically trapped and cooled atoms into purely magnetic traps where these atoms can be studied and further cooled evaporatively in the absence of light. This will avoid temperature and density limits associated with optical processes. If the atoms can be loaded into the magnetic trap with sufficiently high densities and long lifetimes, cooling can proceed efficiently via selective removal of the high-energy tail of the distribution. The ratio of elastic collision rates to inelastic collision rates is much higher for Rubidium and for Cesium than it is for Hydrogen, so the alkali experiments will not face the severe density limitations to evaporation that the hydrogen trapping groups have encountered.
  
  We have been achieving encouragingly high densities in a Magneto-optical trap (MOT) by increasing the gradient of the MOT to unusually high levels [25 mT/cm (250 G/cm)]. When the laser light is turned off, about a quarter of the atoms remain magnetically trapped in the quadrupole trap formed by the same coils used to generate the MOT fields. We have also been developing a better vacuum chamber, one that can reach 10^-9 Pa while still providing an adequate flux of alkali atoms for trapping.
• Metal Nanocolumns Constructed with an STM. A scanning tunneling microscope (STM) has previously been used to draw silicon atoms across a crystal surface and into tall, narrow columns of nm scale dimensions. The large height-to-width ratio and known shape of these columns provide a basis for studies of nanodevice electronics and optics. We have now used the STM to produce similar nanocolumns of metals, including platinum, which does not deteriorate in air, and tungsten, which makes exceptionally robust columns. The physical processes and STM procedures involved in constructing metal columns are quite different than for silicon, due to the vast difference in conductivities. In addition to their usefulness for studying nanocircuits, these metal columns are valuable in scanned-probe microscopy, as they provide access to the shapes of steep-sided structures. Since the STM image in the figure is a "convolution" of the probe and sample shapes, this image requires a matching column on the probe tip, and the actual columns are considerably narrower than seen in the image.



Figure 2: STM scan of an array of platinum nanocolumns on a platinum substrate.

• Probing Light Propagation and Interactions in a Nonlinear Media. Soliton pulse propagation in an optical fiber is the highest rate method available for information communication. This phenomenon is based on the nonlinear response of a Kerr medium, which produces pulse self-focusing in longitudinal and transverse dimensions. This nonlinear phenomenon is also a potential basis for all-optical switching for optical computing and optical communications routing. While this technology is progressing by leaps and bounds, many critical mechanisms and loss processes are not controlled or well understood, with the resultant slowing of progress.

  

  Figure 3: Intensity contours of the angular distribution, as a function of frequency, of near-resonant light emitted from self-focused beam in strontium vapor.

  We are studying these nonlinear mechanisms in an atomic vapor, a nonlinear medium that is well understood. Using a YAG-pumped dye laser tuned near an atomic resonance, we have studied the self-focusing propagation as well as the generation of new frequencies, which are major loss mechanisms in such media. Our studies to date have found that the theories currently used to describe single-beam frequency generation need major corrections. We are developing a combined theoretical and experimental attack to understand why, for example, for a resonance at ω₀₁ and a visible cone that peaks at ω_cone, a predicted feature at ω_blue is entirely absent. Interactions between pairs of pulses, as proposed for all-optical control mechanisms, can also be studied in this uniquely well characterized system.
• Beams of Aligned Molecules Elucidate Orientation Effects in Collisions. One of the long-standing dreams in chemical physics has been to bring aligned or oriented molecules into collision with one another, and in essence probe the steric effect of chemical reactions. Sensitive process control, such as control of molecular beam epitaxial growth, may depend on this understanding. Unfortunately, typical experimental study of such transformations, even in the rarified environment of crossed molecular beams, still entails a painful averaging over impact parameters, intermolecular geometries, collisional energies, etc. However, some progress towards this goal has been made. We have developed sensitive methods based on fast polarization modulation of high resolution IR diode lasers to probe M_J distributions in rotationally excited molecules, and have applied these methods to demonstrate very high levels of alignment of the rotational angular momentum vector that occurs naturally due to rapid collisions in supersonic jet expansions. Results for CO₂ molecules indicate as much as a 50 % preferential alignment of J perpendicular to the expansion axis, which results from fast He atoms bombarding the slower CO₂ species "from behind."



Figure 4: Mechanism for preferentail rotational alignment of CO₂ molecules in a supersonic jet.

• "Clocking" Hot Radical-Molecule Collisions on Reactive Potentials. Although the Doppler effect can be a cause of substantial headache for the laser metrologist, the Doppler method is being put to good use to probe the final velocity distributions in quantum state resolved collisional reaction dynamics. Specifically, translationally "hot" Cl atoms prepared by UV excimer laser photolysis of Cl₂ molecules are hurled against species such as HCl, H₂O, and CH₄. The quantum state product distributions of these collisions are then observed with time resolved, direct absorption of a high resolution IR laser source, which can be tuned over the full Doppler profile of final state velocities. Analysis of the IR Doppler profiles permits one to extract differential scattering cross sections with final quantum state resolution. These studies probe the key collisional dynamics which occur in the entrance and exit channels to fundamental radical Cl atom chemical reactions, information that may be essential for better characterization of plasma etching processes.
• New Method to Detect Rydberg States Demonstrated. The creation and detection of very highly excited atomic states is an important tool for precision spectroscopists and those studying the behavior of atoms in both strong and weak fields. We have demonstrated a new detection technique that uses an optical "dump" pulse from a visible dye laser to stimulate emission from the Rydberg state of interest to a lower valence level. Fluorescence from this lower level is then detected. This new technique has many advantages compared with earlier methods. It permits the selective interrogation of individual Rydberg states, and the ease of using optical light pulses. The Rydberg state being stimulated to the lower level is easily identified since the stimulated emission process is governed by the normal atomic selection rules and the wavelengths are well-known. The highest principal quantum number that can be unambiguously observed is limited by the "dump" laser's linewidth, not by the wavelength or energy resolution of the detection system. While direct ionization of the Rydberg state occurs, this pathway is relatively limited in the new technique because of the strength of the stimulated emission transition.

  
  
  Figure 5: Schematic of excitation and dump pulse detection scheme.

  This technique has been first tested on Ca atom Rydberg states. Pulsed dye lasers were used to excite a series of ¹D states with principal quantum numbers n = 11 to 38 by a two step process. States with n = 12, 18 and 25-38 were detected with a third pulsed dye laser tuned to the "dump" transition 4s nd ¹D₂ → 4s 5p ¹P₁. Fluorescence from the ¹P state at 671 nm was observed through a monochromator. The first collisional alignment effect of a state-to-state Rydberg system has recently been detected.
• "Doppler-less" Spectroscopy Used for Precision Optical Frequency Interval Determination. With the aid of Magneto-Optic trapping, one can collect a few million atoms from a low density ambient (approximately 10^-5 Pa) and cool and concentrate them into a ball of sub-millimeter dimension. Switching off the magnetic fields and increasing the laser frequency detuning allows cooling of the trapped atomic cloud to rms velocities approximately 30 cm/s. This gas of noninteracting atoms forms an ideal environment for precision spectroscopic studies, at least for the few dozens of milliseconds before the atomic ball falls out of our view. By running the trapping, cooling, and uv measurement steps in a repetitive cycle, many of the used atoms can be recycled, leading to efficient data collection and enabling one to imagine using rare isotopes for these kinds of measurements.
  
  Using the precise methods of dye laser frequency control developed at JILA, we have scanned over the UV 3S-5P third resonance line in these cold sodium atoms, and used the resulting frequency intervals to determine the excited state hyperfine coupling constants with about 50-fold improvement in precision. Current efforts aim to measure the optical decay lifetime with unprecedented accuracy (approximately 0.1 %) for stringent tests of atomic calculations. These experiments are useful and significant for the physical data they produce, but are of fundamental interest for us as prototype experiments for a future generation of high performance optical atomic frequency standards based on super-stable lasers, cold atom technology, and probably Ramsey fringes in an "atomic fountain" geometry.
• Unprecedented Frequency Stability and Reproducibility of a Visible Laser Achieved. We have recently made use of the commercial availability of an efficient diode-laser-pumped Nd:YAG laser emitting at 1.06 µm wavelength, combined with efficient frequency-doubling into the green using modern nonlinear optical methods, to produce about 75 mW of high quality, frequency-stable green light. It has been possible to detect six families of narrow sub-Doppler iodine resonance lines. The resonances are reminiscent of those obtained with the ubiquitous I₂-stabilized HeNe laser but have several distinct advantages, namely 20-fold narrower resonance linewidth, 100-fold better signal/noise ratio, and the fundamental metrological advantage of separating the laser source and frequency reference functions. These resonances, having nearly ideal symmetry, could support stabilization of the laser to sub-100 Hz linewidth if optimally used. The reproducibility appears to be approximately 300 Hz.

  It is instructive to compare these first results to those obtained with the 633 nm iodine-stabilized HeNe laser after 24 years development work in countless national standards labs: for this HeNe system the soon-to-be-published experts' report from CCDM '92 will recommend that 12 kHz be formally accepted as the reproducibility of this system. The already-demonstrated reproducibility in these first experiments with the doubled-Nd 532 nm green is some 40-fold better!

• Isolation Systems for Ground-based Gravitational Wave Antennas and Other Applications. The performance of ground-based gravitational wave antennas could be extended down to frequencies near 1 Hz through improved seismic isolation. Such an isolation system has many practical applications. We are working to demonstrate that a three-stage isolation system can be designed that will provide the necessary isolation without introducing too much thermal noise. During the past year a preliminary six-degree-of-freedom isolation stage with an isolation factor of 100 from 1 Hz to 100 Hz has been constructed and is now being tested. Each isolation stage needs to attenuate oscillations in all six degrees of freedom in order to avoid cross-coupling between the different types of motion. The preliminary stage is large enough to support a vacuum system which can contain the two planned main isolation stages. The preliminary stage also can be used later as a shake table to test the performance of the two main stages. Although scientifically driven by the demanding requirements of ground-based gravitational wave detectors, the techniques developed and the understanding achieved have direct application to industrial vibration isolation systems.

• Program to Trap Ions for High Accuracy Atomic Mass Measurement Advances. A Penning ion trap has been constructed which employs a highly uniform and stable magnetic field and accurate electric fields so that one should in principle be able to make atomic mass measurements to parts in 10¹⁰, and perhaps eventually to parts in 10¹². Masses with such accuracies play a role in possible determination of the neutrino mass, more accurate determination of Avogadro's constant, physics and chemistry of isomers, and the possible adoption of a non-artifact mass standard. The method involves cyclotron resonance of single ions in a cold (4K) ion trap.

  We have built and successfully tested an external ion source and beam line. This allows introduction of arbitrary species of ions and control of the number of ions introduced. Mass comparisons with carbon will be readily accomplished on a short time scale, for example. We will now aggressively proceed to make mass measurements and comparisons using singly-charged ions, with special attention to silicon with an eye to a possible mass standard. A host of other ions will be considered, and work will also be undertaken to introduce multiply-charged ions into the trap in order to analyze comparable Q/M of different nuclei, since ultimately all masses must be related to carbon (¹²C).

• Evidence for Magnetic Field Annihilation on the Surface of Stars. It has long been suspected that magnetic energy is converted into motion and heat on a grand scale on stellar surfaces. We have now discovered a stellar magnetohydrodynamic laboratory for studying this process in detail. Understanding the physical processes that result in efficient energy conversion is critical for developing controlled fusion as a commercial energy source, and we are developing some of the atomic physics tools needed to infer plasma properties from the emitted radiation. In this case we have analyzed the ultraviolet emission lines of three-times ionized carbon and silicon from the star AU Microscopii observed by the Hubble Space Telescope. We find that these emission lines, which are formed in gas at about 10⁵ K, are very broad with wings extending to 200 km/s from line center. Line profiles this broad are seen over small portions of the solar surface where new magnetic flux is emerging from the interior, producing interactions with the pre-existing field that convert magnetic energy into high speed motions and heat. It is believed that the new Hubble data provide the first clear evidence for magnetic field emergence and annihilation on a star, but in the case of this very active star, the phenomenon occurs on a far grander scale than in the Sun.

• Groundbreaking Measurements in Electron-Ion Excitation. The NIST merged beams technique has made possible the first successful measurements of electron-impact excitation of an intercombination level of a multiply-charged ion. We studied the 4s^2 1S → 4s4p ³P transition of Kr⁶⁺, a species introduced into tokamak plasmas for diagnostic purposes. Not only were absolute cross sections measured, but definitive dielectronic resonances were observed, also a first for such an ion. Serious disagreement of the measured cross sections with quite sophisticated calculations indicates a need for further theoretical development, and additional experimental work to guide that effort.

• Atomic Collisions Data Center. The Atomic Collisions Data Center is a part of the Standard Reference Data program. Current activity combines the development of electronic databases and the generation of critical reviews in the areas of optical physics and atomic scattering physics. New thrusts will involve databases and critical reviews in the areas of diode lasers and nonlinear optics.

  A database of electron and atom collision data for atoms and molecules, with 20 000 bibliographic entries and 20 Mbytes of collision data, has been built up over several years in support of the critical reviews. The new version is managed under INGRES with an X-windows interface developed for data entry and retrieval, and for production of publication quality graphical arrays. A PC database, written in CLARION, of gas lasers from the Handbook of Laser Science and Technology, Vol. II and Suppl. I, Marvin Weber, ed. (CRC Press) was delivered to Standard Reference Data in July 1993. Three critical reviews are nearing completion: Argon collision data for the Reference Discharge Cell with Zoran Petrovi; Collisional Alignment and Orientation of Atomic Outer Shells, II. Quasi-molecular Excitation and Beyond, with N.O. Andersen, E. Campbell, J.W. Gallagher, and I.V. Hertel; Collisional Alignment and Orientation of Atomic Outer Shells, III.Spin-dependent Effects, with N.O. Andersen and K. Bartschat. New projects in progress are: Redbook on Electron-Impact Excitation and Ionization, with D. Schultz and J. W. Gallagher; development of a database at Oak Ridge National Laboratory under INGRES with on-line service; and a pilot project on data on Chlorine, with J. N. Bardsley and W. L. Morgan at Livermore National Laboratory to support modelling of inductively coupled discharges used for plasma processing.

Most Recent Technical Activities   |   Archive of Technical Activities