Associate Director for Science
Rowland Institute at Harvard
100 Edwin H. Land Blvd.
Cambridge, MA 02142
Website: Burns Lab
Over the years I've participated in a number of, to me, fascinating projects, most of them at the Rowland Institute. I have found no particular common thread other than simple curiosity coupled with an opportunity to indulge that curiosity. Some of the more interesting ones are described below.
Michael Burns - Associate Director for Science
Over the years I've participated in a number of, to me, fascinating projects, most of them at the Rowland Institute. I have found no particular common thread other than simple curiosity coupled with an opportunity to indulge that curiosity. Some of the more interesting ones are described below. (For a complete list of publications.)
Electromigration in single crystal copper
In a metal conducting a current the electrons, in the process of colliding with the atoms in the lattice, can actually knock the atoms out of their lattice positions. Since the electrons are all flowing in one direction, there can be a net movement of atoms in the wire. At high enough current densities this can lead to a measurable macroscopic movement of the metal along the wire (one mode of failure in modern integrated circuits). We are studying some of the effects of surface electromigration in thin single crystal copper wires in an effort to better understand the electromigration process itself. Copper is a particularly interesting metal in this regards technologically since the most recent semiconductor processes have moved from using aluminum interconnects to using copper. Understanding (and controlling) electromigration is necessary for successful use of copper in this new role.
Single crystal metal whiskers
To observe the binding of cold atoms around a wire (described below), we needed a long, thin wire. One potential method for producing such a wire is to use an old technique of growing metallic single crystal metal whiskers. Although it is surprisingly easy to grow forests of these whiskers in a rather uncontrolled fashion, it is a much more difficult problem to grow them on demand, at a certain location, with a particular crystal orientation. Although the wires needed for the atom-orbits are now being produced by other methods, the technique of metal whisker growth is still being pursued. One reason is that single crystal iron whiskers produced by this technique look like they can be used as the tip in an STM (scanning-tunneling microscope), perhaps as a source and/or detector of spin polarized electrons. This would be of great interest in the emerging field of spintronics. Additionally they are the source of the wires in the electromigration experiments described above.
Atoms orbiting a wire
This project is an exploration of the binding of neutral atoms in stable orbits around a wire charged by a time-varying sinusoidal voltage. Both classical and quantum-mechanical theories for this system have been studied and a unified approach to the Kapitza picture of effective potentials associated with high-frequency fields produced. It appears that cavities and waveguides for neutral atomic matter waves may be constructed using these principles. [Phys. Rev. A, 45, 6468, (1992)]
One may also bind a magnetic atom to a current in a wire through the interaction between the atomic dipole moment and the wire's magnetic field. The theoretical description is based on an extension of the concept of supersymmetry to multi-component wave functions. An analytic solution for spin 1/2 particles can be obtained directly in coordinate space. Experimentally, the system should be realizable for 25 micro-Kelvin sodium atoms around a wire with a diameter of 0.5 microns and a current of 400 micro-amps. [Phys. Rev. Lett. 74,3138 (1995)]
Spin 1 particles present more of a problem theoretically. We have found bound states via numerical and approximate analytic results, and have calculated the decay due to various effects, which bodes well for experimental realizations. Theoretically the supersymmetry method that was so successful in the spin 1/2 state is only approximate for the spin 1 state. The eigenvalues are not degenerate in the angular quantum number, but they almost are. This presents an interesting, unanswered, question: what is the small physical parameter responsible for the small breaking of the symmetry, which we observe in the numerically computed energy eigenvalues? [Phys. Rev. A, 53, 1653, (1996)]
Properly fashioned electromagnetic fields coupled to microscopic dielectric objects can be used to create arrays of extended crystalline and noncrystalline structures. Organization can be achieved in two ways: In the first, dielectric matter moves in direct response to the externally applied standing wave optical fields. In the second, the external optical fields induce interactions between dielectric objects that can also result in the creation of complex structures. In either case, these new ordered structures, whose existence depends on the presence of both light and polarizable matter, are referred to as optical matter. [Science, 249, 749 (1990)]
The interactions in the second group are formed by the significant forces between dielectric objects induced by intense optical fields. These forces are very long range (the forces decay only as 1/r) and oscillate in sign at the optical wavelength. We performed an experiment that demonstrates the simplest case by observing a series of bound states between two 1.43 micron diameter plastic spheres in water. The application to more complex cases (with more spheres) is not currently understood: naive integration of the 1/r forces leads to infinities, which in practice must somehow saturate. It is not known, for example, what the ground-state equilibrium configuration is and hence we do not yet even know what optical matter formed in this fashion would look like. [Phys. Rev. Lett., 63, 1233 (1989)]
Human vision has the remarkable property that, over a wide range, changes in the wavelength composition of the source light illuminating a scene result in very little change in the color of any of the objects. Computations for the color perception of an object depend on knowing more than the amount of light from a point on that object (as in Dr. Land's retinex theory, for example); hence long-range interactions of neural signals is necessary. It was not clear whether these long-range interactions take place right in the retina or further along the pathway in the cortex. We tested the role of the cortex in a human subject in whom the nerve fibers connecting cortical areas subserving two separate parts of the visual field had been severed, and found that the cortex is necessary for long-range color computations. [Nature, 303, 616 (1983)]