Nanomagnet Logic (NML) - Field-Coupled Nanocomputing: Paradigms, Progress, and Perspectives - page 21

One of the driving forces behind this research is the semiconductor industry itself
in the form of the Nanoelectronics Research Initiative (NRI) of the Semiconductor
Research Corporation (SRC). The NRI was formed in response to the impending ''red
brick wall'' in the industry's road map, which is primarily the result of the inability to
manage dissipation associated with computation with field-effect transistors. In an
effort to find alternative, lower-power device technologies, the NRI is searching for
switches based on state variables other than charge. One possibility is the electron's
spin, and associated magnetic phenomena. There now are several research efforts
underway to explore switches where the logic state is represented by the magneti-
zation of a nanomagnet. These various approaches have been summarized and
reviewed in [ 3 ], and our work on nanomagnet logic is one of these efforts.

2 Single-Domain Magnets for NML

Nanomagnet Logic (NML) is based on patterned arrays of elongated nanomagnets that
are sufficiently small to contain only a single magnetic domain. The magnetization
state of a device - i.e. whether it is magnetized along one direction or another,
commonly referred to as ''up or down'' - can be used to represent binary information
in the same way that magnetic islands are used to store information in magnetore-
sistive random access memory (MRAM). Elongated single-domain magnets are
essentially tiny bar magnets with poles on each end, that generate strong stray fields
that can be used to couple to other nearby magnets. While such magnetic interactions
between neighboring nanomagnets are undesirable for data-storage applications, we
have demonstrated that these interactions can be exploited to perform logic

operations.
It should be emphasized here that such single-domain behavior is rather special
and specific to magnets with certain sizes and shapes. For our work with patterned
ferromagnetic thin-film permalloy, these sizes are on the order of hundreds down to
tens of nanometers (nm). If the magnet is too large, its magnetization state breaks up
into multiple internal domains, and the poles at the end - along with their strong
fringing fields - disappear. If the magnet is too small, its magnetization state can be
switched by random thermal fluctuations, and it no longer has a stable magnetization
state; this is the so-called superparamagnetic limit. As a fascinating side comment,
Nature has learned to exploit the stray fields associated with single domain magnets
for navigation in the Earth's magnetic field. Specialized, so-called magnetotactic
bacteria grow perfectly single-domain nanomagnets, that are specific to a particular
animal species [ 4 ]. By the way, much of our work on NML uses nanomagnets with
sizes and shapes between those characteristic for the pigeon and the tuna.
Figure 1 shows a magnetic force microscope (MFM) image of an array of nano-
scale magnets with varying sizes and aspect ratios. The coloring represents magnetic
contrast, and dark and light spots indicate magnetic poles. It can be seen that these
magnets display single-domain behavior if they are sufficiently small and narrow (left
side of image), and these poles generate magnetic flux lines that can interact with
external magnetic fields, or couple to neighboring magnets. Otherwise, their magne-
tization state breaks up into several internal domains (right side of image), and there is