DOW-UAP-D129: Metallic Spintronics - Giant Magnetoresistance, Spin-Transfer Torque and the Promise of Radiation-Resistant Electronics

Source file: DOW-UAP-D129_AAWSAP-DIRD-Metallic-Spintronics-March-23-2010.pdf Originating agency: Defense Intelligence Agency (DIA), Defense Warning Office, Directorate for Analysis; prepared by the Acquisition Support Division (DWO-3) under the AAWSA Program Document type: Defense Intelligence Reference Document (DIRD), "Acquisition Threat Support" series, control number DIA-08-1003-011 Date: 23 March 2010 (information cutoff date, ICOD: 1 December 2009) Classification: UNCLASSIFIED//FOR OFFICIAL USE ONLY, with the FOUO caveat struck through on every page; released in 2026 Page count: 27 VIRIN: 260918-D-D0360-1118 PURSUE Release: 6

Summary

This is one of the most down-to-earth documents in the AAWSAP DIRD series: an orderly, conservative review of physics that in 2010 was part of mainstream research and, in part, already part of the computer industry. The author, whose name is redacted as "AAP Person 71," explains spintronics (a contraction of "spin-based electronics"): electronics in which information is carried not only by the electron's charge but also by its spin, a quantum property that behaves like a tiny magnet.

The paper runs to 22 numbered pages and 97 references, in five chapters: an introduction; giant magnetoresistance (GMR); spin-transfer torque (STT), including experiments, applications and the motion of magnetic domain walls; spintronics in antiferromagnetic metals; and a summary. It carries 12 figures, most of them schematics and measurements drawn from the scientific literature.

The core message is that GMR provides a way to detect spin-encoded signals and STT a way to manipulate them, and that together they could lead to a "universal" memory and to low-power, radiation-resistant electronics. The paper does not mention UFOs, unidentified aerial phenomena or advanced propulsion; its only link to aerospace is an argument about radiation hardness for electronics in space.

Research Article

Why spin: the introduction and the heat wall

The introduction starts from a problem in the computer industry: forty years of progress rested on miniaturization, and further shrinking runs into rising power dissipation. In the author's words, "Changing the spin of an electron is faster and requires less power than moving it" (page 4). The paper divides the field into three subfields: materials that are both magnetic and semiconducting, where the author says most devices "are still theoretical concepts"; new magnetotransport effects in ferromagnetic metals; and techniques for manipulating individual spins. The paper concentrates on the last two and observes that "The confluence of intense basic science and industry interest in ferromagnetic metal spintronics has not occurred on this scale in physics in a long time" (page 5).

The concluding chapter puts numbers on the problem. Under Moore's law, transistor density would reach about 10^13 per cm2 by 2035, and at clock speeds of about 10 GHz a chip would dissipate about 40 MW/cm2, a heat load greater than that of a rocket nozzle (pages 21-22). The International Technology Roadmap for Semiconductors (ITRS) calls the looming collapse the "Red Brick Wall," a point where no "known manufacturable solutions" exist.

GMR: reading magnetism through resistance

Chapter 2 explains GMR, the research on which received the 2007 Nobel Prize in Physics. In multilayers, where a few atomic layers of a ferromagnetic metal alternate with non-magnetic layers, the electrical resistance drops sharply when an external magnetic field is applied. The original observation was made on iron-chromium (Fe/Cr) multilayers grown by molecular beam epitaxy (MBE); later work on sputtered cobalt-copper (Co/Cu) multilayers found room-temperature magnetoresistance 3 to 4 times larger than Fe/Cr and 13 times larger than the permalloy films then used in read heads (pages 6-7).

The explanation rests on the "two-current model" based on a suggestion by Mott: spin-up and spin-down electrons carry current in two parallel channels and scatter more strongly when their spin opposes a layer's magnetization. When the layers are aligned, one channel passes almost unhindered and "shorts" the current; when they are antiparallel, both channels are equally blocked. The paper gives the formulas (RAF = (R+r)/2 versus RF = 2Rr/(R+r)) and distinguishes current in the plane of the layers (CIP) from current perpendicular to them (CPP), where a sample 1 mm2 wide and 1 µm thick has a resistance of only about 10^-7 to 10^-8 ohm (page 8). Applications already in use included hard-drive read heads based on the "spin valve" and "exchange bias," gear-rotation sensors, current monitors and galvanic isolators.

STT: writing magnetism with current

Chapter 3, the longest, covers the inverse effect. STT, predicted in 1996, means that a spin-polarized current transfers angular momentum to a magnet and can reverse its magnetization or set it into precession. The paper's simple picture rests on conservation of angular momentum and Newton's third law: an electron entering a ferromagnetic layer has its spin realigned with the magnetization, and the magnetization feels a reaction torque. Because the torque from one electron is negligible, enormous current densities are needed, above 10^12 A/m2, so every experiment forces the current through a tiny constriction: a mechanical point contact smaller than 10 nm, a lithographic contact, a nanowire or a nanopillar (pages 9-10).

The author describes the first observation (1998, Reference 5) using a point contact to a Co/Cu multilayer; the Landau-Lifshitz-Gilbert equation (Equation 1) and the "critical current" above which STT overcomes damping; and switching in a Py/Cu/Py nanopillar with a 40 x 120 nm cross-section (Figure 7). High-frequency experiments showed precession at tens of GHz, and synchrotron x-ray microscopy can follow the magnetization with about 100-picosecond resolution and a spot of about 30 nm. Even so, the author writes, "the detailed understanding of STT is still the subject of debate" (page 14).

The applications (pages 15-18) are nanoscale oscillators tunable from a few GHz to more than 100 GHz, perhaps into the terahertz range, with linewidths of 2 MHz and quality factors above 18,000; STT-MRAM, in which bits are written by a current passed directly through the memory cell rather than by a slowly decaying magnetic field that disturbs neighbouring cells; and "racetrack" memory, in which current pulses move magnetic domain walls along a nanowire, an effect first predicted by Luc Berger in 1978.

Antiferromagnets: a research direction not yet proven

Chapter 4 is the newest and most cautious. Theoretical predictions (MacDonald and coworkers, 2006) held that antiferromagnetic metals, whose magnetic moments alternate and cancel, should also show GMR-like and STT-like effects, at current densities of only about 10^9 A/m2 compared with about 10^11 A/m2 in ferromagnetic multilayers. The author immediately qualifies this: "all calculations to date are for perfect samples and depend on quantum coherence" (page 18).

The experiments the author reviews confirm only half of the picture. The one experimental search for antiferromagnetic GMR (2009, Reference 87) found none in sputtered multilayers, possibly because of scattering in imperfect films. By contrast, four separate experiments found that a strong current changes the "pinning" of a ferromagnetic layer attached to an antiferromagnet. In the detailed example (Figure 12), across 29 contacts on three samples, current in one direction strengthened the pinning and current in the other direction weakened it; because the shifts reverse with the current, they cannot be due to Joule heating alone (pages 20-21).

What the paper does not say

The paper makes no reference to UFOs, unidentified craft, propulsion or weapons. It offers no timelines apart from the Moore's-law projection to 2035, and no data on radiation hardness; the claim that metallic spintronics resists radiation is a qualitative argument in the conclusion. In the author's own words, "much fundamental work remains to be done before we see commercial applications of these devices" (page 22). There are also small editorial slips: "SST" for STT in several places, and in the description of Figure 12 the "free" and "pinned" layer labels (F1 and F2) are swapped relative to the figure caption.

Among the twelve technical areas of the program's Statement of Objectives (DOW-UAP-D110), the paper belongs mainly to "supporting topics," with a possible link to control (onboard processing and sensors) and to materials (nanoscale magnetic multilayers). As general context: STT-MRAM, which the paper anticipated without dating it, has since reached commercial production.

Significance

The document's value lies less in scientific novelty than in what it shows about AAWSAP. Some of the program's reference documents were conventional reviews of materials science and electronics, written at the level of an academic review article and drawn almost entirely from peer-reviewed literature (Physical Review Letters, Nature, Science). The aerospace connection is stated cautiously, at the very end: "The radiation resistance would be of particular interest for aerospace applications because the radiation in space is known to severely damage conventional electronics by building up a destructive charge in transistors" (page 22).

Readers looking in this series for evidence of anomalous technology will find none here. Readers who want to understand the breadth of the program will find a clear example of its "supporting topics": electronic infrastructure that any future aircraft or spacecraft, conventional or not, would need.

Key People

Locations

The paper names no other places; it deals with laboratory experiments reported in the literature without giving their locations.

Key Concepts

Notable Quotes

"In spintronic devices, information is carried not by the electron's charge, as in conventional microchips, but by the electron's intrinsic spin." -- page 4

"Changing the spin of an electron is faster and requires less power than moving it." -- page 4

"The confluence of intense basic science and industry interest in ferromagnetic metal spintronics has not occurred on this scale in physics in a long time." -- page 5

"Linewidths as narrow as 2 megahertz were demonstrated (Reference 55), leading to quality factors over 18,000." -- page 16

"Note, however, that all calculations to date are for perfect samples and depend on quantum coherence. It is known that disorder can reduce the predicted effects." -- page 18

"Unless the energy dissipation from transistor switching can be reduced dramatically, the thermal load associated with 40 MW/cm2 will exceed that in a rocket nozzle." -- pages 21-22

"The radiation resistance would be of particular interest for aerospace applications because the radiation in space is known to severely damage conventional electronics by building up a destructive charge in transistors." -- page 22

"As the spintronics field is still in a relative state of infancy, new and more exciting phenomena are likely to be uncovered in the future." -- page 23