Students' names link to their final papers.

Kasey Kelly (Kenyon College; advisor Rena Zieve)  studied avalanches when grains composed of welded spherical ball bearings are confined to a single layer in a rotating tumbler. The motivation was to understand how earlier work by other students may have been affected by warping of the container walls. Kasey found two signatures that indicated warping. The first was a decrease in the apparent density of the grains, since they could take up less space within the plane of the tumbler by twisting slightly out of that plane. Since the density varies with the exact arrangement of grains, observing density changes requires many avalanches. Kasey's second signature solved this problem by relying on individual configurations. She calculated histograms of the separations of pairs of ball bearings. One would expect a first peak located at the ball diameter, corresponding to any ball bearings that touch each other. Kasey found that as the tumbler becomes less two-dimensional, this first peak splits into three parts. Her work enables us to test quickly for warping in our further measurements.

At a nanometer length scale, surfaces are far from smooth. Complicated energetics determine whether one atomic species deposited on another material grows in clumps, spreads smoothly, or some combination. Even without clumping, the deposited material may be thicker in certain regions than others and form steps. The steps have their own dynamics of migration and recombination. James Morad (Solano Community College; advisor Shirley Chiang)  used scanning tunneling microscopy (STM) and low-energy electron microscopy (LEEM) to study surface structure when a small amount of gold, corresponding to a few atomic layers, is adsorbed onto a germanium sample. The work had two thrusts: obtaining a clean germanium surface as a starting point, and later observing the structure formed by gold layers. Sample cleaning involved bombarding with argon ions to blast any contaminant atoms off the surface and then annealing at high temperature to remove any craters formed on the surface. Although the procedure was not perfected during the summer and some craters typically remained, James was able to observe a series of atomic steps once gold was deposited. James has transfered to UC Davis and is continuing his work during the school year.

Jason Veatch (University of Evansville; advisor Rena Zieve)  studied the effect of uniaxial pressure on CeCoIn5 at low temperatures, measuring how pressure shifts the superconducting transition temperature. Jason's contribution was to extract useful data from a run where we had no working thermometer at the sample. By comparing signals on warming and cooling, Jason found the thermal time constant linking the sample to the refrigerator. He then wrote a program to compute the sample's temperature based on the refrigerator's temperature history. In a related project, Jason set up CeIrIn5 samples for pressure cell measurements by polishing the crystals and making low-resistance contacts for resistance measurements. This material is unusual in that the transition to superconductivity as measured by resistance appears at a significantly higher temperature than when the transition is detected in heat capacity or magnetic susceptibility, and we plan to monitor the temperature discrepancy under pressure.

Condensed Matter Theory

In a crystal, the electron orbitals of neighboring atoms overlap, leading to interactions among the electrons and new orbitals that extend through the crystal. The electron-electron interactions can give rise to phenomena such as superconductivity or magnetic order. One of the simplest models that captures some of these behaviors is the Hubbard Model, which considers the ability of electrons to ``hop" from one location to a neighboring site and also an interaction energy from having two electrons at the same site. When the two-dimensional Hubbard Model has on average one electron per site, the result is antiferromagnetism: magnetic order where electronic spins alternate direction and produce no bulk magnetic moment. Shauna Story (University of Arizona; advisor Richard Scalettar)  used Quantum Monte Carlo to test how antiferromagnetism disappears as a lattice is diluted by having sites removed. Site removal reduces the number of nearest neighbors and hence the electron hopping; eventually it leads to disconnected sites with no possible interactions.

Computational Complex Systems

Jake Ellowitz (Clark University; advisor Jim Crutchfield)  studied information sharing in simulations of reinforcement-based learning. In the simulations, the problem was for robots to collide with walls as little as possible while traveling through a room. The robots first had to discover where the walls were, through trial-and-error. The robots can shorten the learning process by communicating with each other. Jake accumulated quantitative data on how information sharing affected learning, as a function of both the likelihood of sharing and the density of robots. For high densities, collective behavior of the robots emerges.

Biological Physics

Misfolded proteins are responsible for diseases from cystic fibrosis to Alzheimer's. In many cases, misfolding leads to aggregation of proteins. Diana Qiu (Yale University; advisor Daniel Cox)   examined configurations of the tumor suppressing protein p53. A protein consists of a chain of amino acids, folded into a pattern that often involves common structures such as helices. Computers are not yet fast enough to determine how proteins fold. Instead, Diana ``threaded" the known amino acid sequence for p53 onto a helix, with some flexibility in the spacing between amino acids. Interactions among amino acids determine the energy of a threading; for example, it may be favorable for one amino acid to lie directly above another, exactly one turn of the helix apart. Diana looked for potentially low-energy configurations, which she used as starting points in a molecular dynamics code. The computer simulation let protein molecules evolve in time, and helix structures that persisted for a few nanoseconds were judged stable. Diana did find a stable helix structure. Results for interacting proteins were less conclusive, perhaps because the length of threaded helix was short.

Emily Ricks (University of Virginia; advisor Xiangdong Zhu)   worked on the Oblique-Incidence Reflectivity Difference (OI-RD) microscope that Professor Zhu's group is developing. OI-RD detects chemical interactions through polarization changes in reflected light. The technique has potential for high throughput. It is also non-invasive and cannot itself affect the reactions, unlike the commonly used fluorescent tagging. Emily set up a library of several hundred small molecules in an array of tiny wells, while keeping the molecules ordered in an intuitive way that can reduce errors in positioning the different chemicals or in reading the results. Emily also had to adapt the standard procedures for printing these arrays since some of the equipment used was designed for fluorescence measurements, which have somewhat different requirements than OI-RD.

Elementary Particle Experiment

Kathryn Marable (Brandeis University; advisors Bob Svoboda and Mani Tripathi)  worked on two separate projects. One related to data analysis from the CACTUS telescope, which primarily observed the nearby Draco cluster. Draco is about 95\% dark matter. If dark matter particles coming from Draco interact with the earth's atmosphere, they can create brief pulses of Cerenkov radiation. Having a thorough understanding of the detector is crucial to the experiment. Kathryn installed and debugged a new piece of simulation code for modelling the detector. In separate work, she also designed and prototyped electronics for controlling photomultiplier tubes. This latter work was for the LUX experiment, an attempt to view dark matter through its interactions with an underwater tank of xenon.

The LUX experiment searches for dark matter through its interactions with xenon. To reduce background noise, the xenon is surrounded by a large tank of water. Energetic particles, such as muons, that enter the tank will produce Cerenkov radiation in the water. Matching the time of the Cerenkov burst and a signal from the inner xenon can eliminate certain events as spurious. Carlos Rojo (Rice University; advisors Bob Svoboda and Mani Tripathi)  worked on setting up photomultiplier tubes within the water bath to detect Cerenkov radiation. The difficulty is in sealing the tube casings to prevent water damage. To mimic the lengthy time periods the sealings will have to hold during an actual experiment, Carlos heated the casings over two weeks. The 3M polypropylene adhesive he chose did appear sufficiently water-tight.

UC Davis particle physicists have worked for years on the Compact Muon Solenoid (CMS), a detector at the Large Hadron Collider (LHC). Although the hardware portion was completed several years ago, other work continues: modeling how the detector will behave when LHC is turned on later this year, and developing analysis tools. Matt Sanders (Washington and Lee University; advisor John Conway investigated how a Higgs particle that decayed via four muons might appear at LHC. He wrote macros in ROOT to reconstruct the initial Higgs energy from the muon information. By comparing results from his algorithm and from a different calculation designed by another researcher, Matt was able to verify that both methods functioned as desired.

Nuclear Physics

The huge accelerators for experimental particle physics boost light particles such as electrons, protons, and positrons to almost the speed of light before colliding them together. However, heavy ions such as gold nuclei can produce even more energetic collisions. Their extra mass more than compensates for their slower speeds. Analyzing a gold-gold collision is immensely complicated, in part because it is truly multiple near-simultaneous events involving individual nucleons. Matt Caulfield (Stevens Institute of Technology; advisors Manuel Calderon de la Barca Sanchez and Daniel Cebra)  wrote a Glauber model simulation to extract hidden information about heavy-ion collisions, such as the closest approach of the ion centers during the event and the number of nucleons that actively collide. Assuming a certain distribution for the nucleons within the nucleus, Matt computed a distribution for the number of charged hadrons that would be observed from collisions. The ultimate goal is to be able to reconstruct properties of the nucleon collisions from the data.

Cosmology

Over a period of six years, the Deep Lens Survey gathered optical data from five regions of space. The "depth" of the survey comes because of the long observation time, which averages out noise and allows distant, faint objects to be seen. Although measurements ceased several years ago, analysis of the data continues. Tom Johnson (Colorado College; advisor Dave Wittman)  worked on distinguishing galaxy clusters. As very large-scale structure, these clumps of galaxies can be used to study how the mass distribution in the universe originated and evolved. Tom searched for parts of the images where several neighboring pixels indicated more than average numbers of galaxies. After comparing his list of galaxy cluster candidates to lists generated through other algorithms, he examined remaining candidates individually. In the end he found 58 previously unidentified possible clusters, which will be used in further analysis efforts.