CHAMPAIGN, Ill. -- The p-n junction diode is the basic element in nearly all semiconductor devices. Trillions of these diodes -- which permit current to flow only in one direction -- are produced daily. More than 10 million p-n junction diodes can be found in a typical personal computer.
However, since the invention of the transistor 50 years ago, certain characteristics of the p-n junction have been poorly understood and improperly described in textbooks. Now, a new theory of p-n junction performance promises to resolve past misconceptions, says a University of Illinois researcher.
"It may sound strange, but the precise physics of what makes these devices work has not been fully understood," said Karl Hess, a U. of I. Swanlund Professor of electrical and computer engineering and a researcher at the university's Beckman Institute for Advanced Science and Technology. "We found deficiencies in every textbook description of p-n junction diodes. For example, the diffusion capacitance -- the conductance property for alternating current -- was predicted incorrectly in all cases."
Hess and Steven Laux, a researcher at the IBM Thomas J. Watson Research Center who spent a yearlong sabbatical at Illinois, combined computer simulations and numerical analyses to come up with a much more complete understanding of p-n junction performance. They developed a mathematical expression that provides excellent agreement between precise computation and analytical approximation.
"This expression permitted us to identify individual contributions to the diffusion capacitance, and to separate these contributions as they arise from the space-charge or quasi-neutral regions," Laux said. "We therefore have introduced a new set of alternating-current boundary conditions and a more precise treatment of the distribution of the mobile charge."
One surprising result of the new theory is that the diffusion capacitance for long diodes is different from what had been assumed. "Instead of growing exponentially, as taught in all textbooks, the diffusion capacitance actually vanishes in many cases," Hess said.
Why had researchers not identified this before, and how could trillions of well-working p-n junctions have been made without this knowledge?
"Some of the mysteries of p-n junction performance were 'explained' in the past by researchers who inserted erroneous terms into their equations," Hess said. "Fortunately for previous analyses, the mistakes were biggest for long diodes, while most diodes are relatively short. Nevertheless, even very short diodes show deviations from the standard understanding at high forward current densities."
In addition to replacing the previously incomplete and incorrect theories in future textbooks, the new theory should be useful to researchers working with certain types of p-n junctions, such as those used in semiconductor laser diodes.
Laux and Hess describe their new theory in the February issue of IEEE Transactions on Electron Devices.
Cite This Page: