Spin Transistors and OLEDs Draw Closer
University of Utah physicists have controlled an electrical current using electron spin, pointing toward an organic spin transistor. The researchers also demonstrated that fabricating efficient OLEDs may be more difficult than expected.
Alexander E. Braun, Senior Editor -- Semiconductor International, 9/15/2008 9:22:00 AM
 |
OLEDs are easier and cheaper to make than inorganic LEDs. Although inorganic LEDs convert 47-64% of electricity into light instead of heat, they remain too expensive to replace incandescent or compact fluorescent lighting. However, the researchers demonstrated that fabricating efficient OLEDs may be more difficult than expected. It was hoped that OLEDs would provide a 63% conversion efficiency; instead, it appears that they may be only ~25% efficient. This is derived from a study of an organic polymer, MEH-PPV, but probably holds true for others as well, they said.
 |
| Schematic of the device used in the electrically detected pulsed-ESR experiment. Electrodes fabricated from thin (~70 nm) aluminum films connect the contact pads and the MEH-PPV/PEDOT stack to prevent alteration of the eigenstates of the low-Q dielectric resonator into which the device is placed. (Source: University of Utah) |
LEDs emit light when negative and positive charges (electrons and holes) combine. The spin of each electron-hole pair can combine in four quantum states, which may combine in two different ways to form a singlet with a 0 net spin, or a triplet with a net spin of 1. In some organic materials, singlets emit light when they decay, while triplets do not. Thus, OLED efficiency is linked to singlet and triplet production, and because a singlet is only one of the four quantum states, this curtails efficiency.
An aim of spintronics is more efficient computers that use electron spin and charge to store and transmit information, with the up and down spins of electrons representing 0s and 1s. The Utah study shows that spins can carry information in an organic polymer, and demonstrates that a spin transistor is possible because information can be converted into a current, manipulated, and changed. Boehme thinks spin transistors and other spin electronics could enable smaller chips, raising computing capabilities by orders of magnitude. “The smaller transistor today is made up of hundreds of thousands of atoms,” he said, “and the ultimate miniaturization goal is electronics on the scale of atoms and electrons.”
Lupton, an expert in optical spectroscopy, developed a new spectroscopic technique for the research. “Surprisingly, little work has been done on electron spin resonance spectroscopy of organic semiconductors,” he said. “We can now manipulate using spin resonance, and spectroscopically directly locate transitions in an organic semiconductor.”
Lupton explained that in a classical situation, spin resonance is the spin’s magnetic moment — either parallel or anti-parallel — to the external magnetic field. “When you shine microwave radiation onto the sample, you flip the spin from parallel (lower energy) to anti-parallel (higher energy),” he said. “You’re controlling spin with the electron spin resonance condition, and we measure this in an organic electronic device by detecting the photocurrent.”
This reveals much about intrinsic spin lifetime in an organic semiconductor. In spin resonance, one distinguishes between the two types of lifetimes — the population lifetime (T1 lifetime), which is how long the population of a particular spin is maintained, and an often-ignored parameter: spin coherence. Both amplitude and phase are necessary to describe wave propagation as well as spin population. The researchers measured both the T1 time (population decay), and defined the boundaries of spin coherence lifetime. “We’re really measuring the phase of spin oscillations in an organic device,” Lupton said.
The spectroscopic technique provides a way to quantify and classify spin lifetimes in an organic semiconductor. In terms of material development, this could enable a semiconductor engineer to ask a chemist to change a particular atom or shape in a molecule, and then directly measure the impact on spin lifetime, allowing the engineering of very long spin lifetimes for computation and storage purposes.
“This has immediate applications,” Lupton said. “Even with something as mundane as an OLED, you must understand what the spin lifetime is, because you have many carbon-based molecules that help low-spin orbit coupling, which means that spin is maintained. If you think about an OLED, where you inject positive and negative charges, each of these charges — plus and minus — carries spin either up, down, plus a half or minus a half, so you get four different combinations of spin when these electrons and holes recombine. In organic semiconductors, only the singlet configuration — one in four — that corresponds to the anti-parallel spin configuration decays radiatively. If the spin is long-lived, then you’ll get more triplets than singlets. Thus, the technique provides a direct spectroscopic window to understanding spin excitations in organic semiconductors. “All our present device concepts are based on the discovery of new degrees of freedom. We’ve demonstrated that the spin degree of freedom in organic semiconductors is well-worth exploiting.”
So far, the research group has not investigated OLED electroluminescence. The next step is to not just read out these spin manipulations electrically, but optically — to look at the electroluminescence change as a function of spin manipulation. “When you change an excitation’s spin configuration by going from the singlet anti-parallel to the triplet parallel configuration, luminescence properties change,” Lupton said. “Usually, you only get singlets emission, but you can make it produce triplet luminescence — that is, phosphorescence. We want to monitor electroluminescence as a function of this coherent spin manipulation, and detect the change in singlet to triplet emission — the fluorescence-to-phosphorescence ratio. This will open a unique window to spin dynamics for all semiconductors.”
