Materials for Magnetic Memory: Commentary
Magnetic Materials in Electronic Applications
John M. Parsey, Jr.
Editor's Note: Materials for magnetic memory
is the topic of this month's installment of JOM-e,
the journal's electronic supplement. The articles that are referenced here appear
only on the JOM
web site and include hypertext enhancement. The titles and addresses of the
articles appear in the issue's table
of contents in both the print and on-line versions.
During the past decade there have been many important advances in electronics,
semiconductor-material epitaxy, photonic devices, thin-film metallurgy, and nondestructive
characterization techniques for assessing these new classes of materials. This
issue of JOM-e
looks at an interesting new area of activities undertaken by members of the TMS
Electronic, Magnetic, & Photonic Materials Division: magnetic-film materials
for advanced electronic devices.
Thin films of magnetic materials can be used for high-speed read/write heads in
disk memory devices or as permanent memory for computer applications. Devices
such as these retain the state of the memory cell when power is turned off, in
contrast to the volatile memory in a standard dynamic RAM device. Also, such memory
should consume a negligible amount of power when in operation, in contrast to
the steady drain of semiconductor dynamic random access memory (DRAM).
The utility of this new class of magnetic materials is founded in the giant magnetoresistive
(GMR) effect, identified in the late 1980s by Babich
and coworkers and by Sato
et al. (see the references in both of the articles in this issue). By layering
conductors with domain-oriented layers of magnetic thin films, a large resistance
change can be induced in the conductor. This can be understood from the interaction
of electric and magnetic fields in current transport (the "right-hand rule") and
Maxwell's equations. The GMR effect produces a relatively large signal when correctly
incorporated in a layered device structure, which makes it very attractive for
commercial applications. The understanding of the GMR effect has led to several
device constructs, but absolute signal levels are still quite small, which hinders
processing, testing, and robustness in applications.
Recent findings have shown that by modifying the magnetic materials, layer structure,
and chemical makeup, substantially higher performance structures can be created
that overcome the limitations of simple GMR-based devices. These devices are based
on NiFe thin films and are characterized as tunneling devices, where a current
is forced through a nominally nonconducting material, such as a very thin oxide
layer, while interacting with magnetic fields of varying orientations.
Presented here are two contributions from both industrial and academic viewpoints
that discuss deposition methods for magnetic tunnel junction (MTJ) and GMR films,
the fabrication of MTJ devices, and characterization methods for the extremely
thin nanostructured magnetic film materials. Some of the critical layers in an
MTJ are on the order of 1-2 nm, which requires very precise and powerful characterization
tools for assessment.
The first paper comes from an advanced technology research group at Motorola Physical
Sciences Research Laboratory, investigating MTJ materials for high-speed magnetic
random-access memory (MRAM) devices. Slaughter
et al. present an analysis of a multilayer stack of magnetic thin films coupled
with a dielectric tunneling layer. In their structure, current flow is modulated
through the very precisely controlled dielectric layer by tunneling processes.
They have realized even higher responsivity in their MTJ devices than those reported
from GMR devices. The key feature in their work is the interaction of a fixed-orientation
magnetic layer with a variable-orientation magnetic layer that produces the 0
or 1 nature for the memory cell. The authors describe some of the issues in creating
and fabricating devices and characterizing the thin layers of materials by nondestructive
methods.
One of the interesting points of the article is controlling the deposition of
such thin layers uniformly over large areas (the reader is invited to review some
of the past issues of JOM for
nondestructive thin-film characterization methods and large-scale epitaxy of semiconductor
materials). Another issue is pinning or fixing the orientation of one of the two
magnetic films in the MTJ device. This is an interesting materials science problem
involving domain pinning, long-studied in magnet materials, but an absolute necessity
for the memory cell function. They note that film processing and the pinning effect
are closely coupled, necessitating a low-thermal-budget processing and fabrication
sequence. Slaughter
et al. note that their materials are very well suited for MRAM and magnetic-sensor
applications, but are not yet performing at a level necessary for read/write heads
in disk memory devices. They propose that a very simple MTJ cell connected to
a complementary metal-oxide semiconductor transistor would function as a memory
element for fast MRAM applications. Their initial device results were quite encouraging,
demonstrating an access time of ~14 ns in a 256 ´ 2
memory cell configuration.
The paper by Keavney
and Falco looks at the characterization of GMR-type materials by optical methods.
They note that many standard characterization techniques cannot be applied because
of detection limits or the penetration capability of the probe. However, several
approaches are available that utilize the interaction of light with magnetic materials.
They describe the magneto-optical Kerr effect (polarization-rotation) for polarized
light interacting with a material containing magnetic atoms (e.g., iron, chromium,
or nickel). This approach is phenomenally sensitive, capable of detecting materials
properties in films of 1-2 nm, which is in the range needed for MTJ or other GMR-type
devices. A limitation is noted in the ability to penetrate thick films.
Another approach involves probing the magnetic film, again with polarized light,
for inelastic interactions. This method is known as spin-wave brillouin light
scattering. A high-intensity laser is focused on the material, and the backscattered
light is analyzed for bulk and surface/interface magnetic spin effects. By careful
analysis and modeling, the magnetic properties of the films may be extracted.
X-ray absorption is a third approach for dissecting the magnetic thin-film materials.
Magnetic circular dichroism allows the unique detection of core-related spins
in magnetic atoms. The authors comment on three approaches as they relate to the
characterization of thin magnetic films: photocurrent (absorption); fluorescence
yield (emission); and total transmission intensity (the net signal passed through
the sample).
The use of these new magnetic materials systems is still in the R&D laboratory
environment, but the prospects for commercial development are exciting. Many companies
manufacturing computer disk drives or other mass-storage devices (such as huge,
semipermanent, memory banks) are deeply involved in the development of thin-film
magnetic devices for read/write and storage applications. If the practical aspects
of film deposition, processing, and device fabrication can be resolved, such devices
could be manufactured early in the new millennium.
John M. Parsey, Jr., is the section
manager for advanced materials in the Digital
DNA Laboratory of Motorola SPS. He is the advisor to JOM
from the Electronic Materials Committee of the Electronic,
Magnetic, & Photonic Materials Division of TMS.
Copyright held by The Minerals, Metals & Materials
Society, 2000
Direct
questions about this or any other JOM page to jom@tms.org.
