Articles you may be interested inRemote plasma enhanced atomic layer deposition of TiN thin films using metalorganic precursor J. Vac. Sci. Technol. A 22, 8 (2004); 10.1116/1.1624285 Plasma induced microstructural, compositional, and resistivity changes in ultrathin chemical vapor deposited titanium nitride filmsIn situ physical vapor deposition of ionized Ti and TiN thin films using hollow cathode magnetron plasma source Hydrogen plasma pretreatment effect on the deposition of aluminum thin films from metalorganic chemical vapor deposition using dimethylethylamine alane Titanium nitride ͑TiN͒ thin films deposited by metalorganic chemical vapor deposition, using tetrakis-dimethyl-amino-titanium as a precursor, are known to have high sheet resistivity. A plasma treatment in forming gas ambient is one way of reducing the high sheet resistivity down to that of the films deposited by physical vapor deposition. This plasma treatment is normally done in situ ͑without breaking the vacuum͒ immediately after deposition. However, an ex situ ͑breaking vacuum after deposition͒ plasma treatment will provide a more economical way of manufacturing in which less time is required in the chemical vapor deposition chamber, hence allowing high deposition throughput. Furthermore, the plasma treatment can be done in a high-pressure chamber, which translates into low cost. In this study, we compared film properties of TiN with in situ and ex situ plasma treatments. Analyses were performed on chemical composition, sheet resistance, amount of Ti atoms, via resistance and reliability using x-ray photoelectron spectroscopy, four-point probe, x-ray fluorescence, and a current-voltage multiprobe. The ex situ plasma treatment was capable of removing carbon from the TiN film, but not without a sacrifice of the reliability and electrical performance of the TiN film. Further development on the ex situ plasma treatment must be made to improve the reliability and electrical performance of the TiN film.
Oxide semiconductor field-effect transistors (OSFETs) have been actively developed for display applications. The off-state current in an OSFET is more than ten orders of magnitude lower than that in a Si-FET [1]. Displays using OSFET backplanes achieve low power consumption through idling-stop (IDS) driving [2], which dramatically decreases the refresh rate when the flame image is static. The extremely low off-state current of the OSFET is applicable not only to displays but to various other applications. Recently, OSFET technology has been applied to ULSI memory [3][4][5][6]. When used as a pass transistor connected to a storage node (SN), an OSFET allows the long-time retention of electric charge in the SN. Unlike conventional DRAM, the memory with OSFET eliminates the need for frequent refresh operations and reduces the refresh power. Furthermore, the memory can perform power gating (PG) in the idle state, solving the problem of leakage power observed in conventional SRAM. Thus, OSFETs enable low-power ULSI. Figure 1 shows three types of OS memory with different features. DOSRAM (Dynamic Oxide Semiconductor RAM) [3] has a small number of elements (1T1C) and can achieve a small cell area. OSSRAM (Oxide Semiconductor Static RAM) [4] is a combined circuit that plays two roles. The normal-operation portion of the circuit operates as fast as conventional SRAM; however, it is volatile. The backup portion can retain data over a long period. In the idle state, the memory cell transfers data from the normal-operation portion to the backup portion and cuts the leakage power by performing PG. In the recovery operation, the memory cell restores the data from the backup portion to the normal-operation portion and immediately resumes its operation. The cell area and operation speed of NOSRAM (Nonvolatile Oxide Semiconductor RAM) [5] are between those of DOSRAM and OSSRAM. Multilevel NOSRAM [5] has also been proposed with the aim of reducing cost per bit. All kinds of OS memory can be simultaneously fabricated by the same process on the same substrate. When embedded in computer architecture, OS memory enables normally-off (N-off) operation, wherein CPU processing and idle state are operated in the power-on and power-gate modes, respectively. The N-off operation effectively reduces the power consumption by reducing the leakage power to substantially zero. For the N-off operation, the embedded memory should operate at a high speed (or backup and restore data at a high speed) and has to retain data over a long period in PG. These requirements are met by DOSRAM, NOSRAM, and OSSRAM. Figure 2 shows a die photograph and specifications of a prototype microprocessor with embedded DOSRAM. The performance of LSI devices will be further improved in the future by miniaturization, which increases the integration and leakage current of FETs. Therefore, lower power consumption will be particularly demanded. OS-LSI is a prospective candidate that achieves both higher performance and lower power consumption. In addition to memory devices, th...
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.