The latest news from academia, regulators
research labs and other things of interest
Posted: Mar 08, 2006
Europe leads the world in spintronic materials research
(Nanowerk News) With CMOS technology likely to reach the end of its development path during the next decade, ‘spintronic' chips incorporating nanoscale magnets could form the memory and logic devices of the future. The FENIKS project has showed how.
The modern microelectronics industry has evolved on the basis of semiconductor devices that function by controlling electrical charge in circuitry of ever-reducing dimensions, now enabling billions of transistors to be packed onto a single integrated circuit. But, as the relentless march of miniaturisation leads toward chip features of just some tens of nanometres in size, technologies such as today's ubiquitous CMOS (complementary metal-oxide semiconductor) are approaching the limits of their potential for further shrinkage.
For some years, researchers have actively been seeking new solutions that harness the quantum mechanical effect of electron spin that reigns in the nano-world. By spinning in one of two directions, electrons act as tiny magnets, the orientation of which can be controlled to represent the 0's and 1's of binary logic – just like the charge states of conventional devices.
Exploiting this property could produce new generations of ‘spintronic' chips such as non-volatile memories, reprogrammable logic circuits and optoelectronic devices, offering much more functionality and lower power consumption than those of today.
The FENIKS project achieved valuable synergy by combining what were initially separate studies into two different approaches to the development of magnetic/semiconductor heterostructures for such applications. Sub-project Fertile focused on the growth and characterisation of III-V ferromagnetic semiconductors (FMS) incorporating manganese (Mn) as the ferromagnetic element, while the Mosaiks project sought to combine so-called 'half-metallic' materials exhibiting very high spin polarisation with narrow-bandgap III-V semiconductors.
In total, the consortium comprised 14 institutes and universities, plus two industrial partners, together representing eight EU countries. Although much of the work of the sub-projects proceeded independently, joint meetings and the involvement of the coordinator, the Interuniversity Microelectronics Centre (IMEC – Belgium) in both research threads meant that interactions were strong and important strides were made in both fields.
To date, spintronic semiconductor activity had been demonstrated only at low sub-zero temperatures. The challenge for Fertile was to progress beyond the long-standing previous best of 110 K (-163°C) towards the goal of delivering devices capable of functioning under ambient conditions. To this end, the team pursued the production of ferromagnetic semiconductors based on GaAs and GaN doped with Mn.
"We set out to fabricate these materials using molecular beam epitaxial growth and to fully characterise and model their magnetic, electrical and optical properties," explains Feniks coordinator Willem Van Roy. "In the event, we advanced the state of the art on several fronts, pioneering a low-temperature annealing technique and showing that the resultant out-diffusion of interstitial Mn was responsible for improved material quality. It enabled us to achieve the world-record Curie temperature (Tc – the operating temperature above which a ferromagnetic material loses its spontaneous magnetism) of 173 K for bulk GaMnAs".
"GaMnN, which we also examined in great detail, proved to be less favorable," continues Van Roy. As a result of our work, GaMnAs is now globally recognised as the leading candidate for continued FMS research. It has the potential to reach even higher Tc levels – although we do not yet know how much higher."
"A number of routes using other materials could also lead to higher Tcs, but these are less well understood. I therefore see two tracks for building on our effort – first, further development and optimisation of device concepts based on GaMnAs (In fact, the majority of the Fertile partners are already collaborating in a follow-up project, Nanospin, launched at the beginning of 2006.); and, second, more materials development and better understanding of room temperature FMS, enabling the device concepts to be transplanted into new areas."
Thin film integration
The initial aims of the Mosaiks sub-project were to integrate highly spin-polarised ferromagnetic metallic thin films (SPFM) of CrO2 and NiMnSb (known as a ‘Heusler alloy'), with quantum wells and epilayer structures of the narrow-gap semiconductors (NGS) InAs and InSb. Recognition of the technical difficulties in achieving high spin injection efficiency between the SPFM and the NGS led subsequently to a revision of these targets.
The growth of pure CrO2 thin films proved to require temperatures above 390°C, at which levels NGS hybrids decompose or melt. The presence of phases such as Cr2O3 was undesirable, because they are not half-metallic or have a much lower effect on spin polarisation. However, significant progress was made with other magnetic oxides such as Fe3O4 and Co-doped TiO2, which could be grown at lower temperatures.
Following the involvement of IMEC at the 18-month point, the remainder of the 42-month programme was modified to concentrate its efforts on producing spintronic structures with Heusler alloys alone. Use of IMEC's GaAs spin LED, plus the development of InAs and InSb LEDs as one-terminal spin injection test structures, also proved particularly useful.
This helped to reduce the overall problem, which embraces electrical spin injection, spin transport and electrical spin detection, into smaller and more manageable pieces. IMEC's spin-LED technology gives direct access to the electrical spin injection, using optical detection to bypass the other two steps.
Another highlight for the original Mosaiks consortium was the elucidation of various spin properties – transport, lifetimes and manipulation mechanisms – of the NGS, again aided by the use of optical techniques.
"Although the replacement of CMOS by spintronics is not an immediate prospect, continuation of the solid work now being carried out in Europe will help to secure a competitive position for EU contenders in future globalised market for such advanced semiconductors," Van Roy concludes.