Tungsten ͑W͒ thin film as a nucleation layer for a W plug-fill process was deposited using a modified chemical vapor deposition ͑CVD͒ called pulsed CVD, and properties of the films were investigated. Basically, the deposition stage was composed of four separate steps for one deposition cycle: (i) reaction of WF 6 with SiH 4 ; (ii) inert gas purge; (iii) SiH 4 exposure without WF 6 ; and (iv) inert gas purge. A higher deposition rate, ϳ1.5 nm/cycle, was obtained as compared to that of atomic layer deposition ͑ALD͒ (ϳ0.25 nm/cycle). The film deposited by pulsed CVD showed a much lower root mean square ͑rms͒ roughness ͑0.87 nm͒ and better conformality at the contact holes with an aspect ratio of 14 ͑contact height: 3.51 m and top diameter: 240 nm͒ as compared to the layer deposited by conventional CVD using WF 6 and SiH 4 .Chemical vapor deposited ͑CVD͒ tungsten ͑W͒ has been used as a plug material in etched oxide contacts and vias at high aspect ratio in multilevel metallization of dynamic random access memory ͑DRAM͒. 1-5 The conventional W plug-fill process is composed of three steps: 1 (i) silane (SiH 4 ) pretreatment, (ii) W nucleation, and (iii) W plug-fill. Prior to W deposition, Ti/TiN is typically deposited as a glue/barrier layer because of poor adhesion of W on SiO 2 . The first step, a SiH 4 pretreatment induces a homogeneous nucleation of W film on TiN, resulting in a continuous and uniform W film less than 20 nm thick. 5 In the nucleation step, WF 6 is reduced with SiH 4 for fast deposition on TiN with little incubation time while H 2 reduction is used in the plug-fill step with an excellent conformality ͑Fig. 1͒.Generally, the SiH 4 reduction in the W nucleation step produces poor step coverage of W films in deep contacts compared to H 2 reduction. Thus, a thin nucleation layer is favored to minimize the effect on the W plug-fill. However, a thin nucleation layer is subject to be attacked by WF 6 during the plug-fill step, leading to volcano formation. 6 Moreover, as the contact and via sizes continuously shrink with increased aspect ratio, the poor step coverage of the W nucleation layer becomes more problematic. In this respect, a deposition process with better conformality is required. Atomic layer deposition ͑ALD͒ of W is considered as a most promising solution with a superior conformality at the aggressive topographic structures. 7-9 Using Si 2 H 6 , 7 SiH 4 , 8-9 and B 2 H 6 9 as reducing agents of WF 6 , ALD-W films were deposited with very smooth surface and excellent step coverage as compared to those of CVD-W. However, an ideally very low deposition rate of 1 monolayer of W ͑0.25-0.3 nm͒ per cycle as compared to conventional CVD-W process 7,9 may limit its applications in the semiconductor processing.In this article, we report the modified CVD-W nucleation process, called pulsed CVD. We introduced an additional SiH 4 exposure without WF 6 after a short CVD reaction of WF 6 with SiH 4 . And, between the two steps, a purge step using inert gas was carried out to prevent a possible gas-phase rea...
In order to resolve the scalability and handover performance issues of existing approaches, we propose a scalable mobility platform which is based on the hierarchical Distributed Hash Table (HDHT), referred to here as MoHiD, to provide host mobility through a DHT. In MoHiD, the location information of the hosts is stored in the HDHT running on the MoHiD Access Router to provide mobility. In the HDHT, the storage level of each entry can be specified, which drastically reduces the handover delay by limiting the number of overlay hops occurring during the query processing steps. In order to demonstrate the feasibility of MoHiD, we used a commercial cloud to measure the entry query time of the HDHT on the global scale and constructed a testbed to measure the handover performance. The experimental results show that the HDHT query delay and the total handover delay are 16.7 ms and 115.9 ms, respectively, this providing evidence of rapid handovers.
With the aim of accumulating data that can be used to infer the basis for the acculturation of trees called 'Goe,' this study aims to identify the survival status of the pagoda and zelkova trees known as 'Goe' trees by comparing and analyzing the status of pagoda trees (Sophora japonica) and zelkova tree (Zelkova serrata) designated as a protected tree in Korea. The results of this study are as follows; Zelkova serrata designated as a protected tree grows the most, with 2,147 trees (29.4%) in Cheonnam, followed by Chungnam (16.5%) and Gyeongbuk (14.4%). However, Sophora japonica showed a different result from zelkova Serrata as the total number of 210 Sophora japonica (55.7%) in Gyeongbuk and Daegu is much larger than that of zelkova Serrata. As a result, in the Yeongnam region, where the Confucianism of Yeongnam was actively practiced, the existence of Sophora japonica is much larger than that of the Zelkova Serrata, which is not a coincidence, and it is difficult to determine it only based on their flora and planting distribution. Results of comparing protected trees of Sophora japonica and Zelkova Serrata showed that the average age of Zelkova Serrata wass 289 years, while that of Sophora japonica was 302 years, and that the average height of Zelkova Serrata wass 18 m, which is higher than the height of 16 m of Sophora japonica. The average diameter at breast height of Zelkova Serrata was 398 cm and that of Sophora japonica was 314 cm, which indicates that Zelkova Serrata is relatively big. Therefore, it can be assumed that Zelkova Serrata has a larger growth potential than Sophora japonica, and the possibility of growth as a big tree is also high, but it seems that the explanation that "they are relatively long-lived" is not clearly determined.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.