An entirely new type of semiconductor laser modulation process will be explored in this industry-university collaborative project. A novel Field-Induced Charge Separation Laser (FICSL) structure will provide direct modulation of the gain to enable much higher modulation speeds relative to conventional diode lasers in which current modulation is employed. Preliminary modeling suggests modulation speeds in the 100 GHz range. Collaborations with Ziva Corp., our GOALI partner, will help guide these activities.
Intellectual Merit: The FICSL involves new physics, that of separating holes and electrons with an applied field via a gate structure placed above the active gain region in order to directly modulate the gain. An enhanced understanding of laser dynamics in multi-terminal configurations should result. The team brings high levels of expertise in MBE growth (Palmstrøm), high-speed transistors (Rodwell) and efficient, high-speed vertical-cavity lasers (Coldren) to uniquely address this new problem area. The UCSB labs in MBE growth, III-V nanofabrication, and materials and device characterization, are second to none for the VCSEL-like laser studies to be doneS. Complementary skills exist at Ziva.
Broader impact: Such more-efficient, higher-speed devices may revolutionize optical interconnect approaches and enable more efficient computers and data centers. Interaction with Ziva greatly benefits the students involved. New processes developed within NSF-NNIN facility, will be available to future generations of students. Physics will be integrated into the ECE227 series, and it will be disseminated in journal and conference publications. The co-PIs and graduate students will continue to participate in one or more of the ten internship programs for under-represented minorities, high school, and undergrad students that are sponsored by NSF-MRL, NNIN, & COE.
A new three-terminal Vertical-Cavity Surface-Emitting Laser (VCSEL) has been demonstrated that uses a novel Field-Induced Charge-Separation Laser (FICSL) mechanism to modulate the light output. Instead of modulating the current to moduate the light, one modulates the voltage on a third terminal--the gate--as in an FET in this device. This voltage applies an electric field across the active region, which separates the holes and electrons, thereby frustrating their normal recombination to create light. In theory, this more direct gain-modulation mechanism should enable much higher speed operation. Researchers at UCSB, led by Prof. Larry Coldren and PhD student Chinhan Lin have recently completed an NSF GOALI project that experimentally demonstrated this new FICSL in VCSEL form. Fig. 1 shows a schematic. At DC with a current flowing from the p-injector to the n-channel (which contains the active region), the light power out characteristic is similar to a very good two-terminal VCSEL--See Fig. 2 (2nd gen). Figure 3 now shows the effect of applying a bias voltage to the gate. Part (a) illustrates the reduction of the light power out as the gate voltage is increased, and part (b) shows that the gate current also increases at the same time. This is just as expected, since the gate voltage causes a separation of holes an electrons and thus a reduction in their ability to recombine to create light. Figure 4 gives initial RF data. As expected from the theory, the small-signal modulation response shown in part (a) has an added zero, so it begins to rise rapidly in the 400MHz frequency range; however, due to an unexpected added pole near 10 GHz, this rise is prematurely terminated and the response drops rapidly (~40dB/decade). The theory predicts that it should continue to rise and then roll-off much more gradually (20dB/decade) as compared to a standard two-terminal laser. It is believed that this deviation from ideal might be remedied by improved bandgap engineering just above the active region. Figure 4 (b) indicates that despite less that idea RF behavior in this initial device, it still outperforms a conventional two-terminal diode VCSEL by about 25%. That is, the 3-dB bandwidth of the FICSL-VCSEL is measured to be that much above that projected for a comparable diode VCSEL. In summary, a new 3-terminal laser that uses a new modulation mechanism has been proposed, modeled, designed, fabricated and characterized. The low frequency results are in very good agreement with the theory; the ultimate performace of high frequency RF characteristics appear to be limited by a glitch in the experimentally grown band structure, but even for this first device, at moderate frequencies (up to the microwave region), the RF results fits the modeling well, and the final 3 dB bandwidth exceeds a projection for a comparable 2-terminal structure. This project benefitted from contributions from summer interns both at the undergraduate and high school level. Two graduate students obtained their PhDs as a result of this program.
