LOW THRESHOLD MULTIPLE QUANTUM WELL LASER FOR HIGH DENSITY ELECTRONIC PACKING APPLICATIONS

OPTICAL INTERCONNECT TECHNOLOGY PROMISES TO PROVIDE SIGNIFICANT RELIEF FROM THE PINOUT PROBLEM, DECREASE IMPLEMENTATION COMPLEXITY, AND PROVIDE IMPROVEMENTS TO THE FLEXIBILITY AND REAL-TIME RECONFIGURATION FOR HIGH DENSITY ELECTRONIC PACKING APPLICATIONS SUCH AS VLSI/VHSIC. OPTICAL INTERCONNECT REQUIREMENTS, IN GENERAL, CAN BE DIVIDED INTO FOUR AREAS: DEVICES, ROUTING MECHANISMS, PACKING TECHNIQUES, AND ARCHITECTURE DEVELOPMENTS. IN PARTICULAR, DEVICES WHICH CAN TRANSFORM ELECTRICAL SIGNALS TO TOPICAL SIGNALS AND VICE VERSA (SEMICONDUCTOR DIODE LASERS, LIGHT EMITTING DIODES, DETECTORS) ARE REQUIRED TO IMPLEMENT OPTICAL INTERCONNECTS. LOW THRESHOLD HIGH SPEED LASERS BEYOND THE CURRENT STATE-OF-THE-ART ARE REQUIRED TO ADDRESS SPEED, AND LARGE FANOUT PROBLEMS. IN PHASE I OF THE PROPOSED EFFORT WE PROPOSE THE DEVELOPMENT OF A HIGH SPEED DIRECT CURRENT MODULATION MQW LASER BASED UPON OUR UNIQUE APPROACH OF REDUCING THE DEVICE CAPACITANCE BY USING SEMI-INSULATING SUBSTRATES AND ETCHING THE BOTTOM N + CONDUCTION LAYER FROM UNDER THE TOP CONTACT PAD. TO THE BEST OF OUR KNOWLEDGE, SUCH AN APPROACH HAS NEVER BEEN TRIED BEFORE. WE FEEL THIS DECREASE OF CAPACITANCE WILL RESULT IN AN INCREASE OF THE MAXIMUM RELAXATION OSCILLATION FREQUENCY. THE FIRST PHASE WILL CONSIST OF PERFORMING REQUIREMENTS AND DESIGN ANALYSIS, FABRICATING LASERS OF SI AND N + SUBSTRATES AND COMPARING THEIR RELAXATION OSCILLATION FREQUENCIES TO PROVE THE VALIDITY OF OUR TECHNICAL APPROACH. THE SECOND PHASE WILL BE AIMED AT DEVICE OPTIMIZATION, PACKAGING ISSUES, BUILDING AND TESTING A DEMO MODEL FOR DEMONSTRATING THE HIGH DENSITY ELECTRONIC PACKING APPLICATION.