Recent advances in cantilever-based force detection have allowed detection of forces below an attonewton. We have applied this capability to the study of dissipation and fluctuations in nanometer-scale systems. Our work is largely motivated by our effort to extend magnetic resonance force microscopy (MRFM) to single-spin sensitivity, where we have been confronted with a variety of unanticipated noise issues. Phenomena that we have studied include magnetic moment fluctuations in nanoscale ferromagnets, non-contact friction and force fluctuations near surfaces, and increased electron spin relaxation rates observed when closely monitoring electron spins by MRFM. The enhanced spin relaxation rate is believed to be caused by Rabi frequency magnetic noise that is generated by thermal vibrations in high order cantilever modes. Overcoming these various noise issues will be key to achieving single-spin quantum readout.
This work was performed in collaboration with H. J. Mamin, R. Budakian, B. Chui, B. Stipe and C. S. Yannoni. We thank ONR and the DARPA Mosaic program for financial support.
High rate of areal density growth rate enables the reduction in cost per MB and increasing demand for data storage. Micromechanics is soon likely to be needed to accommodate the increased mechanical position precision needed for reading and writing data. Two examples will be described. A microactuator can be used as the fine actuator of a two-stage actuator servo system for very high bandwidth magnetic head slider positioning. This device is batch-fabricated metal structure using high-aspect-ratio lithography and stencil plating. The fabrication process and characteristics will be described. As a second example, micromachined atomic force microscope (AFM) probes are used to generate and detect fine pits on the surface of a polymer disk. An areal density of 25 Gb/in<SUP>2</SUP> is achieved with a data reading rate of above 1 Mb/s. A very low mass (0.3 ng) silicon nitride AFM probe has been used in this study, and the fabrication and performance are described.
Charge trapping in thin films of silicon nitride has long been studied for use as a non-volatile semiconductor memory. Recently, this technology has been combined with scanned probe technologies with the sharp probe tip serving as the upper electrode in a Si<SUB>3</SUB>N<SUB>4</SUB>- SiO<SUB>2</SUB>Si (NOS) structure. By applying a voltage pulse between the tip and silicon substrate, charge carriers can be made to tunnel through the oxide and be trapped in the nitride. This trapped charge causes a shift in the capacitance-voltage curve along the voltage axis; the voltage at which depletion occurs is increased. It has been proposed that such a system could be used as a high density data storage device. We have begun to explore some of the issues related to such an application, including data lifetime and data rates. In thermally accelerated life tests, no sign of charge spreading was seen after 100 days at 150 degree(s)C and from the rate of charge decay we would predict room temperature lifetimes in excess of 1 million years. We have also used an air-bearing spindle to conduct high speed measurements on a spinning NOS sample and obtained data rates as high as 500 kHz with carrier-to-noise ratios of approximately 60 dB in a 3 kHz bandwidth.