Abstract.In the framework of the EU-funded MF-STEP project, autonomous drifting profilers were deployed throughout the Mediterranean Sea to collect temperature and salinity profile data and to measure subsurface currents. The realization of this profiler program in the Mediterranean, referred to as MedArgo, is described and assessed using data More than twenty drifting profilers were deployed from research vessels and ships-of-opportunity in most areas of the Mediterranean. They were all programmed to execute 5-day cycles with a drift at a parking depth of 350 m and CTD profiles from either 700 or 2000 m up to the surface. They stayed at the sea surface for about 6 h to be localised by, and transmit the data to, the Argos satellite system. The temperature and salinity data obtained with pumped Sea-Bird CTD instruments were processed and made available to the scientific community and to operational users in near-real time using standard Argo protocols, and were assimilated into Mediterranean numerical forecasting models.In general, the cycling and sampling characteristics chosen for the MedArgo profilers were found to be adequate for the Mediterranean. However, it is strongly advised to use GPS and global cellular phone telemetry or the future Argos bidirectional satellite system in order to avoid data compression and losses, for the continuation of the Mediterranean drifting profiler program.

Abstract. In the framework of the EU-funded MFSTEP project, autonomous drifting profilers were deployed throughout the Mediterranean Sea to collect temperature and salinity profile data and to measure subsurface currents. The realization of this profiler program in the Mediterranean, referred to as MEDARGO, is described and assessed using data collected between June 2004 and March 2006 (including more than 1500 profiles). Recommendations are provided for the permanent future implementation of MEDARGO in support of operational oceanography in the Mediterranean Sea. More than twenty drifting profilers were deployed from research vessels and ships-of-opportunity in most areas of the Mediterranean. They were all programmed to execute 5-day cycles with drift at a neutral parking depth of 350 m and CTD profiles from either 700 or 2000 m up to the surface. They stayed at the sea surface for about 6 h to be localised by, and transmit the data to, the Argos satellite system. The temperature and salinity data obtained with pumped Sea-Bird CTD instruments were processed and made available to the scientific community and to operational users in near-real time using standard ARGO protocols, and were assimilated into Mediterranean numerical forecasting models. In general, the cycling and sampling characteristics chosen for the MEDARGO profilers were found to be adequate for the Mediterranean. However, it is strongly advised to use GPS and global cellular phone telemetry or the future Argos bi-directional satellite system in order to avoid data compression and losses, for the continuation of the Mediterranean drifting profiler program.

No abstract

Key remaining concerns raised for implementation of Ni FUSI into manufacturing are addressed and solved suggesting that Ni FUSI is worthy for manufacturing. We studied NiSi, Ni 2 Si and Ni 31 Si 12 FUSI gates and their CMOS integration showing 1) Excellent reliability (NBTI, PBTI and TDDB) on HfSiON (EOT=1.1 nm), with lifetimes >10 years at 1.2V for optimized HfSiON (BTI similar/improved compared to reference MG, strong effect of N (DPN HfSiON) finding optimal point in NMOS-PMOS BTI trade-off). 2) No Ni penetration into substrate and no additional reliability degradation with multilevel metallization BEOL thermal budget. 3) Excellent mismatch characteristics and low V t variability down to L G~4 0nm W~130 nm (no FUSI grain orientation effects), 4) Excellent EOT scalability with no PMOS V FB roll-off down to EOT~0.7 nm (Ni 31 Si 12 , WF~4.9 eV); 5) SRAM defectivity analysis finding main type of defects and solutions for their elimination. We also showed 6) phase formation (NiSi, Ni 31 Si 12 ) similar to blanket films at L G =30 nm.Introduction Much progress was made recently in Ni FUSI [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15]. Several methods of WF tuning were demonstrated such as phase control (n: NiSi, p: Ni-rich), implantations (gate: B, P, As, Sb, Al, Yb; channel: F, N) and alloying, which allow to reach adequate WF and V t values [1-10]. In parallel, manufacturability, scalability and reliability have been addressed [2, 3, 8,[11][12][13][14][15], including the demonstration of good process capability and scalability for linewidth independent phase-V t control (e.g. RTP1 PW~20 o C with SiGe cap flow). However, several issues have been raised as potential show-stoppers for implementation of Ni FUSI into manufacturing which have not been clarified. This include impact of BEOL thermal budget (potential Ni penetration, reliability) [14], PBTI on high-k dielectrics [15] and ability to meet reliability specs for scaled dielectrics, V t variability, EOT scalability to ~1nm and beyond (V FB roll-off towards midgap at thin EOTs reported for several p-metal gates [16]), phase formation at short gate lengths (~30 nm), and the key issue of defectivity. Bridging [11], punch-through [13], incomplete silicidation and lateral S/D silicide growth ("pipes") have been identified as potential yield killer defects. All the above issues are addressed in this paper, demonstrating that none of them appears as a showstopper for implementation of Ni FUSI into 45 and 32 nm nodes.Experimental Ni FUSI devices were fabricated using CMP flows previously described [2], and a flow with PMOS poly etch-back (to 50 and 30 nm for Ni 2 Si and Ni 31 Si 12 FUSI resp.) for phase controlled CMOS integration (NMOS: 100 nm poly to obtain NiSi) [11, 12]. Gate stacks on blanket wafers, and with patterned arrays (L G =30 nm) were used for physical characterization. SiON and HfSiON (MOCVD) dielectrics (EOT in 0.7-1.6 nm range) were studied, with NH 3 or DPN nitridation (with varying N content). Optimized 2-step RTP FUSI was used for phase control...

No abstract