Fashioning themselves "latter-day Edisons," researchers at the University of Nebraska contend that their architecture for quantum-dot development is 500 percent better than its nearest competition. Quantum-dot devices, which use the quantum nature of electrons to switch between binary states, could be a solution to problems encountered by ever-shrinking conventional transistors.

"We set a world record by demonstrating the largest nonlinear coefficient for a semiconductor quantum dot," said Supriyo Bandyopadhyay, the lead researcher. "Previous architectures have been highly praised for achieving a tiny percent increase, but we got a 500 percent increase with our design."

Band yopadhyay's work has led to a University of Nebraska patent issued on the 500 percent-better quantum dot. Besides Bandyopadhyay, five other researchers assisted in the work. They are Rod Dillon, Ned Ianno, Latika Menon, Paul Snyder and Frazer Williams.

The patent covers an inexpensive construction method for vast arrays of quantum dots, involving an easy-to-perform electrochemical process on an aluminum substrate. Quantum dots are described as "spontaneously" forming atop the aluminum substrate in a regular array suitable for processing information.

"We've been working on this process for five or six years now, and it's far from perfect — the chemical solution has to be just right, the power level of the electrical current has to be just right and used for exactly the right length of time," said Bandyopadhyay.

Without the precise controls of their patented process, Bandyopadhyay explained, the quantum dots won't arrange themselves in the regular arrays needed to create computing machinery. Bandyopadhyay estimated that it will take his team five more years to perfect his quantum-dot manufacturing process.

Binary replacements

Quantum dots, in theory, can produce tiny computing architectures that fit as many as 10,000 devices in a space with the thickness of a human hair. According to Bandyopadhyay, quantum computers later in the decade will start to replace binary computers, and will begin to make military satellite circuitry immune from electronic-warfare attacks even sooner.

"You could build a quantum computer that has 2 to the 1,000th bits of data, which you could never do with a binary computer, because the number 21,000 is larger than the number of atoms in the known universe. But a quantum computer could store that many bits with just 1,000 atoms," said Bandyopadhyay.

The trick that quantum dots perform to encode so much information is called "superposition," which allows a quantum bit, or "qbit," to keep its state nebulous until data is read out at the end of a computation. In this manner, a qbit can simultaneously represent all the possible states a normal bit could be in, which is imagined as having all states superimposed on top of each other.

This superposition of states enables, for instance, an 8-qbit addition to simultaneously represent all possible 8-bit values added to all possible 8-bit values simultaneously. Bandyopadhyay calls the superposition phenomenon in qbits its "quantum parallelism." In his view, the hardest part of building a quantum computer comes from arranging atomic states of qbits in such a way that quantum mechanical manipulations will result in the desired calculations.

"We have been working on quantum computer research for about three years and have succeeded in demonstrating computer memories," he said. "But we are at least five years away from having a small-scale quantum computer in the lab, and commercial versions won't be available for at least 20 years."

Much sooner, however, the Nebraska team hopes to demonstrate specific improved optical devices based on quantum dots that should shield military satellites from laser attacks. Such devices could also improve infrared imaging, night vision, surveillance and other electronic-warfare attributes. Since quantum dots operate on principles immune from most external tampering techniques, highly secure communications systems can be built, according to Bandyopadhyay.