No abstract

No abstract

The effect of Ta and Ta/Cu seed layers, and Ta and Cu cap layers on the effective magnetic thickness of ultrathin permalloy (Ni81Fe19) was investigated for MRAM applications. The films were deposited by Ion Beam Deposition. The magnetic moment of each as-deposited permalloy film was measured using a B-H looper and a SQUID magnetometer. The films were further annealed at either 525 K for 1/2 h or 600 K for 1 h to study the effect of thermally driven interdiffusion on the magnetic moment of the permalloy film. Our theoretical calculations showed that the presence of 12% intermixing at the interface reduces the Ni moments to zero. Experimentally, it was shown that the tantalum rather than the copper interfaces are primarily responsible for the magnetically dead layers. The Ta seed layer interface produces a loss of moment equivalent to a magnetically dead layer of thickness 0.6±0.2 nm. The Ta metal in the cap layer results in a loss of moment equivalent to a dead layer of thickness 1.0±0.2 nm. Upon annealing, thermally driven interdiffusion is concluded to have a strong effect on the Ta(seed)/ Ni81Fe19 as-deposited interface, based on the doubling of the magnetically dead layer to 1.2±0.2 nm. The Ni81Fe19/Ta(cap) as-deposited interface slightly increases its equivalent magnetically dead layer upon annealing to 1.2±0.2 nm. As-deposited interfaces of Ta(seed)/permalloy and permalloy/Ta(cap) are not chemically equivalent and result in different magnetically dead layers, whereas after annealing to 600 K both interfaces attain comparable intermixing and magnetically dead layers. It was also shown that a half-hour anneal at the lower 525 K annealing temperature, which is closer to the actual processing temperature, results in only slight increase of the magnetically dead layer at both interfaces.

No abstract

Future network should be able to eficiently serve Most existing telecommunication wide area networks (WAN) packet-based networks, such as the Internet. In this paper, based and Metropolitan Area Networks (MAN) have an SDH based. on results from COST 266, we explore characteristics of Optical electronically circuit switched transport core. Connection setBurst switching (OBS) and Optical Packet Switching (OPS).up or tear down may require days or weeks, and switching as Both node and Area Network (MAN) are well as inultiplexing/demultiplexiilg always requires complex discussed. A unique joint comparative performance evaluation of optical/electro,optical (oEIo) conversions, tile contention resolution in OBS and OPS are presented, as well as which controls set-up and tear-down of connections. Work on networks, and their performance. automatically switched optical network (ASON) and Xeywords: burst switching, packet switching, contention generalised multi-protocol label switching (GMPLS) takes place within ITU and IETF, respectively. Resulting Optically of Quality of Service (QoS) differentiation in OBS,OPS operators and vendors =e working on an optical Control plane, resolution, node design, Qualie of Service, simuldios -7ih International Conference on Telecommunications -ConTEL 2003 ISBN: 953-184-052-0, June 11-13, 2003, Zagreb, Croatia 775 Circuit Switched (OCS) networks can offer explicit transfer guarantees, since circuit establishments are confirmed. However, this generates a delay at least equal to the round-trip time, typically several ms. Even though OCS networks will offer more flexibility than today's solution, the access to the optical bandwidth is still provided with fihrdwavelength granularity. Future networks should he able to serve a client layer that includes packet-based networks, such as the Internet, which may have a highly dynamic connection pattern with a significant portion of bursty traffic between the communicating pairs. In this case, OCS transport may not be flexible enough. It would require over-dimensioning of the number of connections and of the bandwidth reservation of each connection, to avoid excessive delay and extensive buffering at the ingress routers. Here is when Optical Packet Switching (OPS) and Optical Burst Switching (OBS) come into play, with the goal of reducing delays and improving the utilization of the network's resources through statistical multiplexing. This comes at the expense of not being able to offer explicit transfer guarantees. However, suitable node design and proper dimensioning of network resources may enable support of most services over the same network. Moreover, OPS and OBS may share the WDM layer with an OCS scheme, serving applications with need for explicit transfer guarantees. In the first part of this paper we present a strictly non-blocking node design suitable for OPS/OBS. Then, we show how the node-design can he simplified, with the drawback that it becomes blocking. We then elaborate on how the performance of this blocking node can approximate the ...