Are organic semiconductors doomed to remain slow? Transistors built of organic semiconductors, which show promise for applications such as large-area electronics--think wall-size TV screens--perform poorly compared with their silicon counterparts when it comes to speed. The sluggishness arises because organic semiconductors lack the exquisite crystalline order of silicon, so electrons inside the organic material bounce around in it instead of traveling in a relatively straight line.
Now a collaboration of engineers at the University of Illinois at Urbana-Champaign; Columbia University, in New York City; and Dupont, in Wilmington, Del., has found a way to make organic transistors better. The group seeded thin layers of organic semiconductors with conducting carbon nanotubes. The nanotubes make up only about 1 percent of the hybrid material, so it retains the physical robustness of a normal organic semiconductor. But the nanotubes produce crystalline high-conductivity regions distributed throughout the transistor.
Turning on the transistor connects the nanotube regions, through which the electrons can travel with less rebounding off the underlying atomic structure. The electrons, in effect, take a shorter path through the transistor. The decreased distance increases the device's transconductance--its ability to control current with applied gate voltages--which is directly related to the speed of the transistor. The group's published work demonstrates a 60-fold improvement in transconductance in sample transistors. Although the improvement is not yet enough for commercial applications, the group says further experimentation will bring those within reach.
Carbon Nanotubes�Semiconductor Networks for Organic Electronics: The Pickup Stick Transistor , by X.-Zao Bo et al., Applied Physics Letters , Vol. 86, 2 May 2005.An Algorithm You Can Dig
Efficient digging requires people who are experienced with a variety of soil types and tools. Even the most experienced workers get fatigued, however, as most digging jobs are very repetitive. Workers using bulldozers, backhoes, and other mechanical equipment can typically make digging corrections only visually, because they do not receive reliable force feedback through the machines' hydraulics. What's needed is automated machinery that can adjust itself to soil conditions. Recently, a group of engineers at Kings College, in London, developed an algorithm for such machinery.
The algorithm uses force information from a few trial attempts at digging to generate parameters that it can plug into soil models. Because the algorithm is so much faster than previous techniques of similar accuracy, digging can be controlled in real time. The group acknowledges that real soils can exhibit considerable variation in their mechanical parameters, but the method performed well over a range of soils. Next, the researchers hope to improve their soil models and to extend the range of digging angles at which the algorithm works.
Online Soil Parameter Estimation Scheme Based on Newton-Raphson Method for Autonomous Excavation , by Choo Par Tan et al., IEEE/ASME Transactions on Mechatronics , Vol. 10, no. 2, April 2005.Zooming In on Networks
When network researchers and engineers talk about graphs, they don't mean the simple x-y plots of high school algebra. They mean complex abstract structures, where nodes are connected by edges, similar to the connections on a subway map of stations with train lines. Graphs naturally lend themselves to describing network structures, from the simplest local area network to the Internet itself. But drawing large graphs in a way that is meaningful to the viewer is notoriously challenging. Graphs can have millions of nodes and edges, making straightforward display on conventional computer screens impossible.
To alleviate this problem, computer scientists at AT&T Research, in Florham Park, N.J., recently developed a new visualization tool for large graphs. The tool lets users zoom in on a particular subsection of a graph while it preserves a faithful, if coarse, representation of the rest of the graph. The AT&T group achieved this by using a combination of algorithms to first derive a hierarchy of increasingly simplified graphs from the original and then merge various levels of detail, depending on how much magnification the user demands. Animated transitions ease users between different views, allowing people to maintain context as they navigate through the graph.
Topological Fisheye Views for Visualizing Large Graphs , by Emden R. Gansner et al., IEEE Transactions on Visualization and Computer Graphics , Vol. 11, no. 4, July/August 2005.