Abstract
Nanoplasmonics has recently experienced explosive development with many novel ideas and dramatic achievements in both fundamentals and applications. The spaser has been predicted and observed experimentally as an active element -- generator of coherent local fields. Even greater progress will be achieved if the spaser could function as a ultrafast nanoamplifier -- an optical counterpart of the MOSFET (metal-oxide-semiconductor field-effect transistor). A formidable problem with this is that the spaser has the inherent feedback causing quantum generation of nanolocalized surface plasmons and saturation and consequent elimination of the net gain, making it unsuitable for amplification. We have overcome this inherent problem and shown that the spaser can perform functions of an ultrafast nanoamplifier in two modes: transient and bistable. On the basis of quantum density matrix (optical Bloch) equations we have shown that the spaser amplifies with gain greater than 50, the switching time less or on the order of 100 fs (potentially, 10 fs). This prospective spaser technology will further broaden both fundamental and applied horizons of nanoscience, in particular, enabling ultrafast microprocessors working at 10 to 100 THz clock speed. Other prospective applications are in ultrasensing, ultradense and ultrafast information storage, and biomedicine. The spasers are based on metals and, in contrast to semiconductors, are highly resistive to ionizing radiation, high temperatures, microwave radiation, and other adverse environments.