Optical computer chips, which are based on the flow of light, could work faster and use less power than conventional computer chips, which rely on the flow of electrical currents. Optical computer chips require various components including on-chip lasers, waveguides, modulators, and isolators. Isolators act as diodes for light, letting it flow in one direction but blocking its flow in the other direction, and are essential components of an optical circuit. Isolators are made from transparent magnetic materials, but the traditional materials, which consist of magnetic garnets, have proven to be very difficult to grow on common semiconductor substrates from which computer chips are made. There is therefore a great need to develop new magnetic materials which can be conveniently grown on substrates such as silicon because this will enable fully functional optical devices to be made, with the ultimate goal of transforming the field of optical computing and enabling faster, lower-power computers. This project studies new magnetic oxide materials, understanding the relation between the composition, structure, strain state of the materials and their magnetic and optical properties, and prototyping new isolator designs based on these materials. The broader impacts of the work include the training of students, and outreach to the public through the MIT OpenCourseWare initiative, mentoring of high school teachers, and interactions with schools.
TECHNICAL DETAILS Magnetic oxides which are magnetooptically active are an essential component of photonic devices such as optical isolators. Isolators act as optical diodes, protecting lasers from back-reflected light, and are therefore of critical importance in photonic devices. Integrating all the components of an optical circuit onto an optoelectronic chip would enable optical computation to be carried out, with its advantages of high speed and low power consumption. However, the traditional magnetooptical material, garnet, has proved difficult to integrate onto a Si or III-V platform. Therefore, there is considerable interest in developing alternative materials which have good optical transparency and high Faraday rotation, and can also be integrated onto photonic substrates. This project develops thin film magnetooptical materials based primarily on perovskites, and controls the magnetic and optical properties by the substitution of ions onto the A and B sites and the presence of vacancies on the oxygen sites, and by controlling the strain and the magnetoelastic anisotropy. In addition, magnetooptical isolators are designed and modeled based on the nonreciprocal phase shift experienced by light passing through the films. The broader impacts of the work include the development of an optical isolator that, if successful, could transform the field of integrated optics; the training of students; and outreach to the public through the MIT OpenCourseWare initiative, mentoring of high school teachers, and interactions with schools.
