Chemical interaction and ligand exchange between a [(CH 3 ) 3 Si] 3 Sb precursor and atomic layer deposited Sb 2 Te 3 films Taeyong Eom, a Taehong Gwon, a Sijung Yoo, a Byung Joon Choi, b Moo-Sung Kim, c Sergei Ivanov, d Andrew Adamczyk, d Iain Buchanan, d Manchao Xiao d and Cheol Seong Hwang * a The chemical interaction between the [(CH 3 ) 3 Si] 3 Sb precursor and atomic layer deposited Sb 2 Te 3 thin films was examined at temperatures ranging from 70 to 220  C. The trimethylsilyl group [(CH 3 ) 3 Si] displays greater affinity for Te than for Sb, and this drives replacement of Te in the film with Sb from the [(CH 3 ) 3 Si] 3 Sb precursor, while eliminating volatile [(CH 3 ) 3 Si] 2 Te, especially at elevated temperatures. The compositions of the resulting Sb–Te layers lie on the Sb 2 Te 3 –Sb tie line. The incorporation behavior of [(CH 3 ) 3 Si] 3 Sb was explained in terms of a Lewis acid–base reaction. The exchange reactions occurred to relieve the unfavorable hard–soft Lewis acid–base pair between the trimethylsilyl group and Sb in [(CH 3 ) 3 Si] 3 Sb. Such a reaction could be usefully adopted to control the chemical composition of ternary Ge–Sb–Te thin films. 1. Introduction Phase Change Random Access Memory (PCRAM) is one of the promising next generation memory candidates, having non- volatile data retention properties and rapid writing and reading speeds. The mostly widely adopted phase change materials (PCM) are pseudo-binary compounds lying on the tie-line con- necting Sb 2 Te 3 and GeTe. Among the many compositions, Ge 2 Sb 2 Te 5 (called GST225) comprises the base material for many PCRAMs. Although GST225 has several promising prop- erties as a feasible PCM, its relatively slow switching speed (program and erase time longer than 50 and 300 ns, respec- tively 1 ) has been identied as one of the problems of this material. In contrast, compositions containing only Sb and Te have been reported to have a much faster switching speed, although the stability of amorphous phases has been a concern. 2,3 Among these compositions, Sb 2 Te 3 is the most thermodynamically favored due to the stable +3 and 2 valence states of Sb and Te, respectively. However, it was reported that the Sb 2 Te 3 -based materials show nucleation-dominant crystallization kinetics which are slower than growth-dominant crystallization kinetics. In contrast, Sb rich Sb–Te alloys [Sb x Te (1x) , (0.4 # x # 1)] show growth-dominant crystallization kinetics, making the switching speed faster. Therefore, there have been several attempts to deposit an Sb rich Sb–Te PCM. 2,4,5 Meanwhile, Sb rich Sb–Te alloys have been considered as promising anode materials for high-energy density lithium (Li)-ion batteries (LIBs) due to their higher theoretical charge storage capacity than commercial graphite (372 mA h g 1 ). 6,7 Sb reacts with Li, forming compounds such as Li 2 Sb and Li 3 Sb, for which the theoretical capacity is 660 mA h g 1 . 7 Sb rich Sb–Te alloys are even more useful for sodium (Na)-ion batteries (SIB). Na-ions are barely stored in other anode materials such as graphite and Si, whereas the theoretical Na-ion capacity of Sb is as high as 610 mA h g 1 . 8–11 Both applications for the Sb rich Sb–Te alloys require conformal lm deposition, which is most feasibly accomplished by Atomic Layer Deposition (ALD). This is because the PCM must be placed within a very narrow trench 12 or hole 13 in order to improve the thermal efficiency of the reset current, which must be <60 mA according to the international technology roadmap for semiconductors at the sub-20 nm technology node. 14 For LIBs and SIBs, conformal formation of Sb rich Sb–Te alloys on nanostructured materials, such as porous templates, carbon-based materials, and biological materials, is necessary. 15 Metal-Organic Chemical Vapor Deposition (MOCVD) of Sb 2 Te 3 was reported by Groshens et al., 16 and further developed as an ALD process by Pore et al., 17–20 with introduction of appropriate precursors for ALD-type reactions. Several previous reports adopting the MOCVD process, 5,21–23 generally failed to achieve conformal, uniform, and smooth lms with low a Department of Materials Science and Engineering, Inter-university Semiconductor Research Center, Seoul National University, Seoul 151-744, Republic of Korea. E-mail: cheolsh@snu.ac.kr b Department of Materials Science and Engineering, Seoul National University of Science and Technology, Seoul 139-743, Republic of Korea c Air Products Korea, 15 Nongseo-dong, Giheung-gu, Yongin-si, Gyeonggi-do, 446-920, Republic of Korea d Air Products and Chemicals, Inc., 1969 Palomar Oaks Way, Carlsbad, CA 92011, USA Cite this: J. Mater. Chem. C, 2015, 3, 1365 Received 25th November 2014 Accepted 10th December 2014 DOI: 10.1039/c4tc02688h www.rsc.org/MaterialsC This journal is © The Royal Society of Chemistry 2015 J. Mater. Chem. C, 2015, 3, 1365–1370 | 1365 Journal of Materials Chemistry C PAPER Published on 10 December 2014. Downloaded by INFOTRIEVE on 26/02/2015 17:11:08. View Article Online View Journal | View Issue