Physics News Update, Number 373

Number 373, May 27, 1998 by Phillip F. Schewe and Ben Stein

CAN EPILEPTIC SEIZURES BE PREDICTED? At the Clinic of Epileptology at the University of Bonn (located on Sigmund Freud Strasse), one of the largest centers for epilepsy surgery in the world, doctors must locate the source of seizures as accurately as possible so as to minimize post-operative loss of brain activity once the offending tissue has been removed. To do this Klaus Lehnertz (klaus@jersey.meb.uni-bonn.de) and Christian Elger scan suspected trouble areas in the brain; they monitor EEG patterns over time and look not just at the rapid, violent neuronal firings during an epileptic attack but at the electrical landscape before and after the brainstorm. In particular, they look for suspicious changes in the "correlation dimension," a number which typifies the local complexity of neural activity. In some patients they see a decrease in the correlation dimension which, they believe, corresponds to the sort of increasing synchronization in the pathological firing of neurons that (above a critical level) leads to a seizure. The Bonn researchers argue that this identification and study of the pre-seizure state in the brain should result in a better understanding of how seizures come about and might suggest new ways (chemical, electrical, or psychological) of preventing seizures. (Physical Review Letters, 1 June 1998.)

QUANTUM DOT CELLULAR AUTOMATA (QCA) might make possible a new type of transistor-less computing. A quantum dot is essentially a zero-dimensional artificial atom, isolated on (or in ) a semiconductor substrate. Loosing a pair of electrons within a cell of four closely spaced dots (with an appropriate nudge the electrons can tunnel from dot to dot) creates a binary bit: the configuration of the electrons establishes either a 1 or a 0. Put many of these cells together and you have a programmable cellular automata network. Data input and output occurs at the periphery of the cell ensemble, which acts like a neural network in that the computing is performed by the quantum interactions within the array. Wolfgang Porod at Notre Dame (porod@graz.ee.nd.edu, 219-631-6376) reports this week at the APS atomic/molecular/optical physics meeting in Santa Fe on the modeling and operation of a QCA array, including the demonstration, for the first time, of the manipulation of a single electron by another nearby single electron.

A MOMENTUM MICROSCOPE FOR VIEWING SINGLE-MOLECULE COLLISIONS has been demonstrated, allowing physicists to determine how the alignment of a molecule can affect the final outcome of a collision. At the APS meeting, Michael Prior of Lawrence Berkeley Lab (510-486-7838) will describe (Talk K5.02) how he and his colleagues combined the imaging of molecular fragments with a new application of the technique called COLTRIMS, short for "cold target recoil ion momentum spectroscopy." Originally developed in the early 1990s for studying collisions between ions and atoms, COLTRIMS collects the products of a collision in a weak electric field, which projects them onto position-sensitive detectors. Measuring the particles' positions and the times it takes them to fly to the detector, one can determine the particles' momentum values and thereby reconstruct the collision itself. Colliding a beam of helium hydride (HeH⁺) molecular ions with a cold helium target, Prior and his colleagues have determined that the amount of energy exchanged during the collision depends upon the way the molecular ion approaches the target atom, such as whether the hydrogen in HeH⁺ faces towards or away from the He atom. (Figures at Physics News Graphics; see also W.Wu et al, Phys. Rev. A, January 1998.)