Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/multidisciplinary-digital-publishing-institute/towards-reconfigurable-electronics-silicidation-of-top-down-fabricated-yMVDun2LNd
- Publisher
- Multidisciplinary Digital Publishing Institute
- Copyright
- © 1996-2019 MDPI (Basel, Switzerland) unless otherwise stated
- ISSN
- 2076-3417
- DOI
- 10.3390/app9173462
- Publisher site
- See Article on Publisher Site
Abstract
applied sciences Article Towards Reconfigurable Electronics: Silicidation of Top-Down Fabricated Silicon Nanowires 1 , 2 , 3 , 1 , 2 , 3 1 , 2 1 , 3 , 4 , 5 Muhammad Bilal Khan *, Dipjyoti Deb , Jochen Kerbusch , Florian Fuchs , 3 , 6 2 , 3 , 6 7 2 , 3 , 8 Markus Löer , Sayanti Banerjee , Uwe Mühle , Walter M. Weber , 1 , 2 , 3 , 4 , 9 2 , 3 , 4 , 5 , 9 1 , 2 , 3 1 , 3 , 10 , Sibylle Gemming , Jörg Schuster , Artur Erbe and Yordan M. Georgiev * Institute of Ion Beam Physics and Materials Research, Helmholtz-Zentrum Dresden-Rossendorf (HZDR), 01328 Dresden, Germany International Helmholtz Research School for Nanoelectronic Network, HZDR, 01328 Dresden, Germany Center for Advancing Electronics Dresden, Dresden University of Technology, 01062 Dresden, Germany Institute of Physics, Chemnitz University of Technology, 09126 Chemitz, Germany Fraunhofer Institute for Electronic Nano Systems, 09126 Chemnitz, Germany Dresden Center for Nano-Analysis, Dresden University of Technology, 01062 Dresden, Germany Fraunhofer Institute for Ceramic Technologies and Systems IKTS, Dresden, 01277 Dresden, Germany Namlab gGmbH, Nöthnitzer Strasse 64, 01187 Dresden, Germany Dresden Center for Computational Materials Science, Dresden University of Technology, 01062 Dresden, Germany On leave of absence from the Institute of Electronics at Bulgarian Academy of Sciences, 72, Tsarigradsko Chausse Blvd., 1784 Sofia, Bulgaria * Correspondence: m.khan@hzdr.de (M.B.K.); y.georgiev@hzdr.de (Y.M.G.) Received: 30 June 2019; Accepted: 14 August 2019; Published: 22 August 2019 Featured Application: This work has implications for the fabrication of nickel silicide–silicon Schottky junction-based devices such as reconfigurable field-e ect transistors. Abstract: We present results of our investigations on nickel silicidation of top-down fabricated silicon nanowires (SiNWs). Control over the silicidation process is important for the application of SiNWs in reconfigurable field-e ect transistors. Silicidation is performed using a rapid thermal annealing process on the SiNWs fabricated by electron beam lithography and inductively-coupled plasma etching. The e ects of variations in crystallographic orientations of SiNWs and di erent NW designs on the silicidation process are studied. Scanning electron microscopy and transmission electron microscopy are performed to study Ni di usion, silicide phases, and silicide–silicon interfaces. Control over the silicide phase is achieved together with atomically sharp silicide–silicon interfaces. We find that {111} interfaces are predominantly formed, which are energetically most favorable according to density functional theory calculations. However, control over the silicide length remains a challenge. Keywords: Schottky junction; field-e ect transistors; nickel silicide; annealing 1. Introduction In the last few decades, the conventional downscaling of complementary metal-oxide- semiconductor (CMOS) transistors dominantly relied on the reduction of size to improve the performance as well as to reduce the costs and the power consumption of devices. With the end of physical scaling of field-e ect transistors (FETs), further device performance enhancement is expected to be based on new concepts. These concepts include new materials (e.g., high mobility channel materials [1–3] such as 2D materials), metal gates with high-k gate dielectrics [4], new device Appl. Sci. 2019, 9, 3462; doi:10.3390/app9173462 www.mdpi.com/journal/applsci Appl. Sci. 2019, 9, 3462 2 of 13 architectures (e.g., 3D integration) [5], new computation principles (e.g., spintronics) [6], and new functionality (e.g., reconfigurability) [7]. Reconfigurability, in particular, can be achieved by selectively injecting a chosen type of charge carrier in the channel and controlling the concentration of these charge carriers at Schottky junctions [8]. This concept is typically implemented by fabricating Schottky junction-based FETs, also known as reconfigurable FETs (RFETs) [9]. In RFETs, undoped silicon nanowires (SiNWs) are nickel (Ni) silicided from both ends using annealing, and the Ni di usion process creates a silicide–silicon–silicide structure with two Schottky junctions. The observed silicide phases vary in terms of the sheet resistance, Schottky barrier height, built-in strain etc. [10,11]. Hence, the phase of the silicide influences the performance of devices as the injection and the extraction of charge carriers in the silicide–silicon junction depend on the properties of the phase [12,13]. Since the performance of RFETs strongly depends on the quality of the Schottky junctions, proper control over the silicidation process (silicide phase, sharpness of the silicide–Si interfaces, silicide length along the nanowire) in these devices is very important but remains a challenge [14]. This work aims at understanding the Ni-silicide di usion mechanism and attaining specifically the NiSi phase in top-down fabricated SiNWs by studying the silicide–Si interfaces. The formation of NiSi –Si contacts is attractive because NiSi exhibits similar electron and hole barriers [10]. It also has 2 2 a low lattice mismatch of 0.4% with respect to Si at room temperature, which allows the formation of single-crystalline structures [10,15]. NiSi forms an abrupt heterojunction to Si, thus enabling its use in unipolar devices such as RFETs [16]. To understand the sequence of phase formation within the desired temperature range, various studies have been reported for bottom-up grown nanowires [17–31]. However, fewer studies are available for top-down fabricated nanowires [32,33]. In addition to the formation of the suitable silicide phase, control over the sharpness of silicide–Si interfaces, and the silicide length is important for applications such as RFETs and Schottky barrier FETs [34]. Silicides are mostly formed by a temperature-induced interfacial reaction of a transition metal with Si. This process is called the reactive phase formation, which indicates di usion between two adjacent phases leading to the formation of a single or multiple products along the chemical gradient between those phases [35,36]. Silicidation is governed by di usion [37,38] and nucleation [35,39] and can be controlled by steps involved in the reaction phase formation. Ni-silicide formation on bulk Si substrates has been extensively studied in the past [39–43]. Moreover, Ni-silicided nanowires can be produced by direct synthesis or by solid-state reaction of Ni with SiNWs. Wu et al. [44] reported single-crystalline NiSi nanowires produced by radial silicidation of vapor–liquid–solid grown SiNWs, which were coated with thermally-evaporated Ni and annealed at 550 C in forming a gas atmosphere to produce single-crystalline NiSi NWs. The first report of the longitudinal growth of metal silicides in nanowires and the formation of flat and abrupt interfaces to silicon was given by Weber et al. [16] by the example of NiSi –Si nanowire hetero-structures. Schmitt et al. [45] published a detailed review of several processes to produce transition metal silicide nanowires. The desired silicide phase formation is facilitated by proper control of annealing parameters [46] in an axial hetero-structure. These hetero-structures have been used for FET applications and exploration of innovative device concepts [8,16,45,47,48]. The silicide di usion and phase formation in SiNWs has been described in various publications. Appenzeller et al. [29] demonstrated that the longitudinal growth of silicide in SiNWs is a function of the radius or the cross-sectional area of the nanowire and the produced silicide phase. The nature of the Si–silicide interface also influences the growth and the phase of silicide. The most common Ni-silicide phases in SiNWs are Ni Si, NiSi, and NiSi . The phase of silicide can be controlled by a proper choice 2 2 of Ni quantity and quality [27,39], conditions of the reaction [30], strain in the nanowire [28,49], and crystal orientation of nanowires [26]. In <112> nanowires, the hexagonal -Ni Si is formed at 300 C, whereas this phase is not observed till 800 C in bulk structures [47]. In <111> nanowires, first the epitaxial NiSi is formed and it withstands temperatures upto 700 C. The low-resistive NiSi is formed at 700 C [26]. The thermal history of the SiNW influences the silicidation rate. Yamashita et al. [50] presented a comparative illustration of silicidation rate for SiNWs fabricated by two di erent processes: Appl. Sci. 2019, 9, 3462 3 of 13 The “doping first” process, where the annealing for dopant activation is carried out before the NW patterning; and the “patterning first” process, where NWs are patterned first and then subjected to further heat treatment for oxidation and dopant activation. They concluded that the silicidation rate in “patterning first” approach nanowires is higher than that for the “doping first” approach, as the oxidation induced strain in the “patterning first” approach is altered and distributed by dopant activation annealing in the later process step. Here we present the results of our investigations on the Ni silicidation of top-down fabricated SiNWs. In the next section, details of the SiNW fabrication and the analysis techniques are given. The initial investigations were carried out on Si thin film structures to identify the process window for SiNW silicidation. In the third section, the results obtained from scanning electron microscopy (SEM) and transmission electron microscopy (TEM) are presented and compared to simulations. Results of electrical characterization of back-gated RFETs are also presented. 2. Materials and Methods The devices were fabricated on (100) surface-oriented silicon-on-insulator (SOI) substrates. One set of substrates had a 20 nm top Si layer and a 102 nm buried SiO (BOX) layer, while the second set of substrates had a 50 nm top Si layer with a 110 nm BOX layer. SiNWs were fabricated using electron beam lithography (EBL) and inductively-coupled plasma (ICP) etching. EBL was performed with a Raith e-Line Plus system at an acceleration voltage of 10 kV, 1200 C/cm dose, 30 m aperture size, and with 2 nm area beam step size using the negative tone EBL resist, hydrogen silsesquioxane (HSQ) from Dow Corning (X-1541). HSQ is known to have sub 5 nm resolution and high etch resistance [51]. The original 6% HSQ solution was diluted to 2% concentration in methyl isobutyl ketone (MIBK) and spin coated at 2000 rpm for 30 s to yield a 40 nm thick HSQ layer. Then the samples were baked at 90 C for 2 min and exposed using the EBL tool. After EBL, the samples were developed with a high-contrast tetramethylammonium hydroxide (TMAH) based development process [52] and dried with a N gun. The HSQ patterns were transferred into the top SOI layer using a SENTECH’s ICP-Reactive Ion Etcher SI 500. The parameters of the etching process were: ten standard cubic centimeters per minute (sccm) SF , 20 sccm C F , 5 sccm O , 0.9 Pa chamber pressure, 400 W ICP power, and 12 W RF power. After 6 4 8 2 the SiNW fabrication, a second EBL-based patterning was performed for Ni contact formation using a positive tone resist, ZEP520A (ZEON Corporation, Tokyo, Japan). The exposure parameters were: A total of 10 kV acceleration voltage, 10 m aperture size and area dose of 40 C/cm . The development was performed in n-amyl acetate solution (ZED-N50, ZEON Corp.). This was followed by baking at 200 C for 5 min to smoothen the side walls of the patterned structures. Prior to Ni sputtering, the Si native oxide was etched away from NWs using 1% bu ered hydrofluoric acid (BHF) solution and then the samples were transferred into the sputtering chamber. The sputtering was performed at 10 keV with a Gatan 681 High-Resolution Ion Beam Sputter Coater. The sample holder was rotated at a speed of 40 rpm for uniform deposition of Ni at a rate of 1.3 0.2 Å/s. The thickness of Ni was varied between 35 and 50 nm according to the requirements of the experiments. After the sputtering step, lift-o was carried out in N-methyl-2-pyrrolidone (NMP) at 50 C for 10 min. The samples were then rapid thermally annealed (RTA) at 450 C in forming gas atmosphere (a mixture of 10% H and 90% N ) for silicidation of the NWs. The annealing time was varied in order to control the axial di usion length of Ni into the SiNW. The role of the Ni pads was two-fold: Firstly, they acted as a Ni reservoirs for silicidation of NWs, and secondly, they served the purpose of a contact electrode. To investigate silicidation of nanowires covered with an oxide shell, the samples were thermally oxidized. The oxidation process started with removal of the native oxide from the samples by dipping them in 1% BHF for 15 s. This was followed by an immediate transfer of samples in a pre-heated (875 C) rapid thermal oxidation chamber. The samples were kept in the chamber for 10 min in the presence of an oxygen flow of 10 standard liters per minute (slm). Subsequently, the samples were annealed at 875 C in a nitrogen environment for 5 min. This was followed by annealing in forming gas atmosphere. The oxidation process generates heavy stress in the nanowire due to the volume Appl. Sci. 2019, 9, 3462 4 of 13 Appl. Sci. 2019, 9, x FOR PEER REVIEW 4 of 13 expansion. This impacts the oxygen di usion in Si and the surface reaction rate at the Si–SiO interface the Si–SiO2 interface drops. Therefore, the oxidation process is self-limited [53,54] (i.e., the stress drops. Therefore, the oxidation process is self-limited [53,54] (i.e., the stress generated by the oxidation generated by the oxidation process prevents further oxidation of the nanowires). After the oxidation process prevents further oxidation of the nanowires). After the oxidation of nanowires, a second EBL of nanowires, a second EBL step for Ni contacts formation and subsequent annealing for silicidation step for Ni contacts formation and subsequent annealing for silicidation was performed according to was performed according to the aforementioned conditions. the aforementioned conditions. To analyze the silicidation process, SEM and TEM investigations were performed. The systems To analyze the silicidation process, SEM and TEM investigations were performed. The systems used for SEM and TEM were Raith e-Line Plus and Zeiss Libra 200 TEM, respectively. TEM samples used for SEM and TEM were Raith e-Line Plus and Zeiss Libra 200 TEM, respectively. TEM samples were prepared using focused ion beam (FIB) in a FEI Helios Nanolab 660 machine applying were prepared using focused ion beam (FIB) in a FEI Helios Nanolab 660 machine applying low-damage low-damage recipes to preserve the native crystal structure [55]. recipes to preserve the native crystal structure [55]. 3. Results and Discussions 3. Results and Discussion 3.1. Silicidation in Si Thin Film Structures 3.1. Silicidation in Si Thin Film Structures To study silicidation in thin films structures, Si pads were patterned in Si substrates in <110> To study silicidation in thin films structures, Si pads were patterned in Si substrates in <110> and and <100> orientations using EBL and dry etching processes. Ni sputtering and subsequent <100> orientations using EBL and dry etching processes. Ni sputtering and subsequent annealing annealing for silicidation was carried out according to conditions mentioned in Section 2. The for silicidation was carried out according to conditions mentioned in Section 2. The purpose of this purpose of this first study was to determine an initial process window for the nanowire silicidation. first study was to determine an initial process window for the nanowire silicidation. Figure 1 shows Figure 1 shows top-view SEM images of a sample after silicidation. Annealing was performed at top-view SEM images of a sample after silicidation. Annealing was performed at 450 C in forming 450 °C in forming gas atmosphere for 10 min. It is evident from the bright regions of the Si pads in gas atmosphere for 10 min. It is evident from the bright regions of the Si pads in Figure 1 that the Ni Figure 1 that the Ni diffusion was not fast enough to reach the nanowires and it was confined within di usion was not fast enough to reach the nanowires and it was confined within the Si pads. Moreover, the Si pads. Moreover, the silicide–Si interface has different shapes in Si pads with <110> and <100> the silicide–Si interface has di erent shapes in Si pads with <110> and <100> orientations. The silicide orientations. The silicide makes a 90° angle with Si in <110> orientation, whereas in the <100> makes a 90 angle with Si in <110> orientation, whereas in the <100> orientation, the silicide has a orientation, the silicide has a step-like interface with Si. step-like interface with Si. Figure 1. Top-view scanning electron microscopy (SEM) images of silicon nanowires (SiNWs) with Figure 1. Top-view scanning electron microscopy (SEM) images of silicon nanowires (SiNWs) with Si Si pads in (a) <110> and (b,c) <100> orientations. The silicided regions of the Si pads are brighter. pads in (a) <110> and (b,c) <100> orientations. The silicided regions of the Si pads are brighter. Silicide–Si interface appears to be flat in <110> pads while in the case of <100> it has a step-like shape, Silicide–Si interface appears to be flat in <110> pads while in the case of <100> it has a step-like shape, as indicated by yellow arrows. as indicated by yellow arrows. TEM analysis was performed on lamellas, prepared by FIB, from the selected samples. The TEM analysis was performed on lamellas, prepared by FIB, from the selected samples. The cross-section was inspected by high-resolution transmission electron microscopy (HREM) and electron cross-section was inspected by high-resolution transmission electron microscopy (HREM) and energy loss spectrum (EELS). Figure 2 shows the overview of the cross-section with the di erent layers electron energy loss spectrum (EELS). Figure 2 shows the overview of the cross-section with the included in it. The top most sample layer is covered by a carbon protection layer, which is part of the different layers included in it. The top most sample layer is covered by a carbon protection layer, which is part of the FIB milling process to mitigate the damage incurred by FIB. Below the carbon layer, there are two Ni silicide layers and the Si substrate. The top Ni-containing layer depicts a Appl. Sci. 2019, 9, x FOR PEER REVIEW 5 of 13 Ni-rich silicide phase (region I). The following layer has a Si-rich silicide phase (region II). The interface between region I and II is smooth, whereas the interface between regions II and the Si Appl. Sci. 2019, 9, 3462 5 of 13 substrate (region III) is rough and jagged. It is evident from Figure 2b–d that the silicide in region II has followed preferential crystal directions. The HRTEM image in Figure 3e shows an abrupt change in the silicide phase from NiS2 to Ni-rich phase. FIB milling process to mitigate the damage incurred by FIB. Below the carbon layer, there are two Appl. Sci. 2019, 9, x FOR PEER REVIEW 5 of 13 Ni silicide layers and the Si substrate. The top Ni-containing layer depicts a Ni-rich silicide phase (region I). The following layer has a Si-rich silicide phase (region II). The interface between region I Ni-rich silicide phase (region I). The following layer has a Si-rich silicide phase (region II). The and II is smooth, whereas the interface between regions II and the Si substrate (region III) is rough and interface between region I and II is smooth, whereas the interface between regions II and the Si jagged. It is evident from Figure 2b–d that the silicide in region II has followed preferential crystal substrate (region III) is rough and jagged. It is evident from Figure 2b–d that the silicide in region II directions. The HRTEM image in Figure 3e shows an abrupt change in the silicide phase from NiS to has followed preferential crystal directions. The HRTEM image in Figure 3e shows an abrupt change Ni-rich phase. in the silicide phase from NiS2 to Ni-rich phase. Figure 2. Cross-sectional transmission electron microscopy (TEM) analyses of a lamella taken from the sample shown in Figure 1. (a) SEM top-view image of a SiNW with a Si pad. Focused ion beam (FIB) section was made along the white line to get the cross-sectional image of silicidation in the Si pad. (b) TEM image of the cross-section showing three different regions. Region I: Ni-rich Ni-silicide phase; Region II: Si-rich Ni-silicide phase; and Region III: pure Si from the Si pads. (c) Energy-filtered TEM (EFTEM) image of Ni to confirm its concentration in the three different regions. Higher brightness corresponds to higher concentration of Ni. (d) EFTEM image of Si of the same Figure 2. Cross-sectional transmission electron microscopy (TEM) analyses of a lamella taken from Figure 2. Cross-sectional transmission electron microscopy (TEM) analyses of a lamella taken from regions. Higher brightness corresponds to higher concentration of Si. (e) High resolution the sample shown in Figure 1. (a) SEM top-view image of a SiNW with a Si pad. Focused ion beam the sample shown in Figure 1. (a) SEM top-view image of a SiNW with a Si pad. Focused ion transmission electron microscopy (HRTEM) image of the interface between Region I and Region II, (FIB) section was made along the white line to get the cross-sectional image of silicidation in the Si beam (FIB) section was made along the white line to get the cross-sectional image of silicidation in showing an abrupt change in the silicide phase. pad. (b) TEM image of the cross-section showing three di erent regions. Region I: Ni-rich Ni-silicide the Si pad. (b) TEM image of the cross-section showing three different regions. Region I: Ni-rich phase; Region II: Si-rich Ni-silicide phase; and Region III: pure Si from the Si pads. (c) Energy-filtered Ni-silicide phase; Region II: Si-rich Ni-silicide phase; and Region III: pure Si from the Si pads. (c) To further investigate the silicidation process, zero-loss TEM was performed [56]. The images TEM (EFTEM) image of Ni to confirm its concentration in the three di erent regions. Higher brightness Energy-filtered TEM (EFTEM) image of Ni to confirm its concentration in the three different regions. are presented in Figure 3. The Ni map shows higher concentration of Ni in region I compared to corresponds to higher concentration of Ni. (d) EFTEM image of Si of the same regions. Higher Higher brightness corresponds to higher concentration of Ni. (d) EFTEM image of Si of the same region II. Fast Fourier transform (FFT) confirms that the silicide phase in region II has a cubic lattice brightness corresponds to higher concentration of Si. (e) High resolution transmission electron regions. Higher brightness corresponds to higher concentration of Si. (e) High resolution structure in accordance to NiSi2. The silicide in region I is Ni-rich and it has a non-cubic structure. microscopy (HRTEM) image of the interface between Region I and Region II, showing an abrupt change transmission electron microscopy (HRTEM) image of the interface between Region I and Region II, The NiSi2–Si interface is atomically sharp (Figure 3c). in the silicide phase. showing an abrupt change in the silicide phase. To further investigate the silicidation process, zero-loss TEM was performed [56]. The images are presented in Figure 3. The Ni map shows higher concentration of Ni in region I compared to region II. Fast Fourier transform (FFT) confirms that the silicide phase in region II has a cubic lattice structure in accordance to NiSi2. The silicide in region I is Ni-rich and it has a non-cubic structure. The NiSi2–Si interface is atomically sharp (Figure 3c). Figure 3. TEM analysis of silicide junctions. (a) Overview of inter-phases between Ni-rich phase and Figure 3. TEM analysis of silicide junctions. (a) Overview of inter-phases between Ni-rich phase and Si-rich phase. (b,c) Magnified images show atomically sharp NiSi –Si junction. Si-rich phase. (b,c) Magnified images show atomically sharp NiSi2–Si junction. To further investigate the silicidation process, zero-loss TEM was performed [56]. The images are 3.2. Silicidation in Nanowires presented in Figure 3. The Ni map shows higher concentration of Ni in region I compared to region II. Fast Fourier transform (FFT) confirms that the silicide phase in region II has a cubic lattice structure in accordance to NiSi . The silicide in region I is Ni-rich and it has a non-cubic structure. The NiSi –Si 2 2 interface is atomically sharp (Figure 3c). Figure 3. TEM analysis of silicide junctions. (a) Overview of inter-phases between Ni-rich phase and Si-rich phase. (b,c) Magnified images show atomically sharp NiSi2–Si junction. 3.2. Silicidation in Nanowires Appl. Sci. 2019, 9, 3462 6 of 13 Appl. Sci. 2019, 9, x FOR PEER REVIEW 6 of 13 3.2. Silicidation in Nanowires Appl. Sci. 2019, 9, x FOR PEER REVIEW 6 of 13 After determining an initial silicidation process window from the experiments for silicidation After determining an initial silicidation process window from the experiments for silicidation in After determining an initial silicidation process window from the experiments for silicidation in in thin film structures presented in the previous section, experiments for nanowire silicidation were thin film structures presented in the previous section, experiments for nanowire silicidation were thin film structures presented in the previous section, experiments for nanowire silicidation were performed. For these experiments, nanowires were fabricated in <100> and <110> crystallographic performed. For these experiments, nanowires were fabricated in <100> and <110> crystallographic performed. For these experiments, nanowires were fabricated in <100> and <110> crystallographic orientations using the top-down approach. Ni was sputtered according to the conditions described orientations using the top-down approach. Ni was sputtered according to the conditions described orientations using the top-down approach. Ni was sputtered according to the conditions described in the Materials and Methods section and, subsequently, rapid thermal annealing was performed at in the Materials and Methods section and, subsequently, rapid thermal annealing was performed at in the Materials and Methods section and, subsequently, rapid thermal annealing was performed at 450 C in forming gas atmosphere for 80 s. The resulting structures are shown in Figure 4. 450 °C in forming gas atmosphere for 80 s. The resulting structures are shown in Figure 4. 450 °C in forming gas atmosphere for 80 s. The resulting structures are shown in Figure 4. Figure 4. SiNWs in <110> and <100> crystallographic orientations: the brighter segments of NWs Figure 4. SiNWs in <110> and <100> crystallographic orientations: the brighter segments of NWs indicate Figure 4. SiNWs in <110> and <100> crystallographic orientations: the brighter segments of NWs indicate indicate silicide formation. Silicidation is faster in <100> NWs. Annealing is performed at 450 C for silicide formation. Silicidation is faster in <100> NWs. Annealing is performed at 450 °C for 80 s. silicide formation. Silicidation is faster in <100> NWs. Annealing is performed at 450 °C for 80 s. 80 s. As illustrated in Figure 4, the silicide diffusion length is larger in <100> NWs compared to <110> As illustrated in Figure 4, the silicide di usion length is larger in <100> NWs compared to <110> As illustrated in Figure 4, the silicide diffusion length is larger in <100> NWs compared to <110> NWs. Appenzeller et al. [29] have shown that silicidation diffusion length is inversely proportional NWs. Appenzeller et al. [29] have shown that silicidation di usion length is inversely proportional to NWs. Appenzeller et al. [29] have shown that silicidation diffusion length is inversely proportional to square of the diameter of the NW. The widths of <100> and <110> NWs in Figure 4a are 28 and 24 square of the diameter of the NW. The widths of <100> and <110> NWs in Figure 4a are 28 and 24 nm, to square of the diameter of the NW. The widths of <100> and <110> NWs in Figure 4a are 28 and 24 nm, respectively, while the silicide diffusion lengths are 745 and 628 nm, respectively. This shows a respectively, while the silicide di usion lengths are 745 and 628 nm, respectively. This shows a strong nm, respectively, while the silicide diffusion lengths are 745 and 628 nm, respectively. This shows a strong dependence of silicide length on the nanowire orientation. To investigate in detail the dependence of silicide length on the nanowire orientation. To investigate in detail the dependence strong dependence of silicide length on the nanowire orientation. To investigate in detail the dependence of Ni diffusion and silicide formation on crystallographic orientations of NWs, a of Ni di usion and silicide formation on crystallographic orientations of NWs, a circular array of dependence of Ni diffusion and silicide formation on crystallographic orientations of NWs, a circular array of NWs with five-degree separation was fabricated. It was followed by the NWs with five-degree separation was fabricated. It was followed by the aforementioned steps for circular array of NWs with five-degree separation was fabricated. It was followed by the aforementioned steps for silicidation. The results are shown in Figure 5. silicidation. The results are shown in Figure 5. aforementioned steps for silicidation. The results are shown in Figure 5. Figure 5. SEM images showing (a) nanowires fabricated in circular array with a five-degree Figure 5. SEM images showing (a) nanowires fabricated in circular array with a five-degree separation Figure 5. SEM images showing (a) nanowires fabricated in circular array with a five-degree separation to study the dependence of silicide diffusion length on the nanowire orientation; and (b) a to study the dependence of silicide di usion length on the nanowire orientation; and (b) a higher separation to study the dependence of silicide diffusion length on the nanowire orientation; and (b) a higher magnification image showing clearly different diffusion lengths in different NWs. Annealing magnification image showing clearly di erent di usion lengths in di erent NWs. Annealing is higher magnification image showing clearly different diffusion lengths in different NWs. Annealing is performed at 450 °C for 80 s. performed at 450 C for 80 s. is performed at 450 °C for 80 s. Although the silicide diffusion length varies in different crystallographic orientations of the Although the silicide di usion length varies in di erent crystallographic orientations of the NWs, it Although the silicide diffusion length varies in different crystallographic orientations of the NWs, it was not possible to extract a clear relation since the results vary in different arrays. To was not possible to extract a clear relation since the results vary in di erent arrays. To further investigate NWs, it was not possible to extract a clear relation since the results vary in different arrays. To further investigate the orientation dependence of the silicide formation, silicidation of nanowire the orientation dependence of the silicide formation, silicidation of nanowire arrays fabricated in <100> further investigate the orientation dependence of the silicide formation, silicidation of nanowire arrays fabricated in <100> and <110> crystallographic orientations was carried out and the results are and <110> crystallographic orientations was carried out and the results are presented in Figure 6a–d. arrays fabricated in <100> and <110> crystallographic orientations was carried out and the results are presented in Figure 6a–d. The width of the NWs is 20 nm. The average silicide diffusion length in The width of the NWs is 20 nm. The average silicide di usion length in <100>- and <110>-oriented presented in Figure 6a–d. The width of the NWs is 20 nm. The average silicide diffusion length in <100>- and <110>-oriented SiNWs is 812 and 744 nm, respectively (see Figure 6e). The images SiNWs is 812 and 744 nm, respectively (see Figure 6e). The images distinctly show that the di usion of <100>- and <110>-oriented SiNWs is 812 and 744 nm, respectively (see Figure 6e). The images distinctly show that the diffusion of Ni and the silicide progression into the NWs of the same array is Ni and the silicide progression into the NWs of the same array is not homogeneous and on average has distinctly show that the diffusion of Ni and the silicide progression into the NWs of the same array is not homogeneous and on average has scattering of up to 300 nm in <100> and 400 nm in <110> NWs. scattering of up to 300 nm in <100> and 400 nm in <110> NWs. Silicidation rates are known to vary not homogeneous and on average has scattering of up to 300 nm in <100> and 400 nm in <110> NWs. Silicidation rates are known to vary based on different factors, including the quality of interface Silicidation rates are known to vary based on different factors, including the quality of interface between the Ni reservoir and the NW [21]. However, our investigation shows this variation also between the Ni reservoir and the NW [21]. However, our investigation shows this variation also within an array of NWs, which is processed under the same conditions. We attribute this variation in within an array of NWs, which is processed under the same conditions. We attribute this variation in silicide length to different surfaces of NWs, as these NWs are top-down fabricated and their surfaces silicide length to different surfaces of NWs, as these NWs are top-down fabricated and their surfaces Appl. Sci. 2019, 9, 3462 7 of 13 based on di erent factors, including the quality of interface between the Ni reservoir and the NW [21]. However, our investigation shows this variation also within an array of NWs, which is processed Appl. Sci. 2019, 9, x FOR PEER REVIEW 7 of 13 under the same conditions. We attribute this variation in silicide length to di erent surfaces of NWs, as these NWs are top-down fabricated and their surfaces can be di erent due to the subtractive nature can be different due to the subtractive nature of the etching process. Because of this uncontrolled of the etching process. Because of this uncontrolled silicide formation and scattered positioning of the silicide formation and scattered positioning of the Schottky junctions, large top gates have to be Schottky junctions, large top gates have to be placed to ensure that they properly cover the junctions at placed to ensure that they properly cover the junctions at the two sides of the NWs [14,34]. This the two sides of the NWs [14,34]. This limits the possibilities for downscaling the size of devices and limits the possibilities for downscaling the size of devices and compromises their large-scale compromises their large-scale fabricability. fabricability. <100> <110> Crystal Orientation of Nanowires (e) Figure 6. SEM images showing silicidation flow in arrays of NWs fabricated in (a,b) <100> and (c,d) Figure 6. SEM images showing silicidation flow in arrays of NWs fabricated in (a,b) <100> and (c,d) <110> crystallographic orientations. Large scattering of silicidation length is evident in NWs within <110> crystallographic orientations. Large scattering of silicidation length is evident in NWs within each array. (e) Graph showing average silicide diffusion length (rectangular bars) and its scattering each array. (e) Graph showing average silicide di usion length (rectangular bars) and its scattering (linear “error bars”) in <100> and <110> SiNWs: silicide diffusion length is higher for <100> NWs, (linear “error bars”) in <100> and <110> SiNWs: silicide di usion length is higher for <100> NWs, while large scattering is observed in both types of NWs. Annealing is performed at 450 °C for 80 s. while large scattering is observed in both types of NWs. Annealing is performed at 450 C for 80 s. To have better control over the silicidation length, further investigations were carried out with To have better control over the silicidation length, further investigations were carried out with sets sets consisting consisting of of two two and and thr thre ee e NWs. NWs. The The r re esults sults ar are e shown shown in in Figur Figure e 7 7.. It It was was expected expected that that the the outer two NWs might consume excess Ni, giving a better control over the silicide progression in the outer two NWs might consume excess Ni, giving a better control over the silicide progression in the central central NW NW. . However However ,,it it was was still stillnot not possible possible to to obtain obtain a a pr proper oper contr contro oll over over the the silicide silicide pr progr ogre ession ssion with this technique. The silicide di usion length varied in di erent groups of NWs. Interestingly, the with this technique. The silicide diffusion length varied in different groups of NWs. Interestingly, outer the onanowir uter nan es owi appear res appe to exhibit ar to ex a distinct hibit a silicide distinct phase silicid compar e phase edco tom the painner red to ones, the as inn indicated er ones, by as their higher brightness. While the inner NWs appear to have a NiSi phase with a comparable lattice indicated by their higher brightness. While the inner NWs appear to have a NiSi 2 phase with a constant comparato ble that latti of ceSi, co the nsta outer nt toones that have of Si,Ni-rich the outer phases ones with hava e lar Ni ger -rich lattice phases constant. with aThis large might r latti be ce constant. This might be an effect of Ni source “competition” between neighboring nanowires. Apparently, the outer nanowires receive their Ni supply from a longer extent, thereby limiting that supply to the inner nanowires. Silicide Length (nm) Appl. Sci. 2019, 9, 3462 8 of 13 an e ect of Ni source “competition” between neighboring nanowires. Apparently, the outer nanowires A receive ppl. Sci. their 2019, 9Ni , x Fsupply OR PEER fr R om EVIEW a longer extent, thereby limiting that supply to the inner nanowires. 8 of 13 Appl. Sci. 2019, 9, x FOR PEER REVIEW 8 of 13 Figure 7. SEM images of (a) single, (b) double, and (c,d) triple nanowires, showing uncontrolled Figure 7. SEM images of (a) single, (b) double, and (c,d) triple nanowires, showing uncontrolled Figure 7. SEM images of (a) single, (b) double, and (c,d) triple nanowires, showing uncontrolled silicidation. The outer nanowires in images (c) and (d) have Ni-rich silicide phase while the inner silicidation. The outer nanowires in images (c,d) have Ni-rich silicide phase while the inner ones appear silicidation. The outer nanowires in images (c) and (d) have Ni-rich silicide phase while the inner ones appear to have NiSi2. Annealing is performed at 450 °C for 10 min. to have NiSi . Annealing is performed at 450 C for 10 min. ones appear2 to have NiSi2. Annealing is performed at 450 °C for 10 min. Dellas et al. [27] and Lin et al. [28] claimed that an oxide shell around the NWs hinders di usion Dellas et al. [27] and Lin et al. [28] claimed that an oxide shell around the NWs hinders diffusion Dellas et al. [27] and Lin et al. [28] claimed that an oxide shell around the NWs hinders diffusion of silicide into the NWs. Therefore, expecting a better control, further investigations were made with of silicide into the NWs. Therefore, expecting a better control, further investigations were made with of silicide into the NWs. Therefore, expecting a better control, further investigations were made with oxidized NWs. However, it also resulted in non-uniform di usion of silicide into the NW, as shown in oxidized NWs. However, it also resulted in non-uniform diffusion of silicide into the NW, as shown oxidized NWs. However, it also resulted in non-uniform diffusion of silicide into the NW, as shown Figure 8. in Figure 8. in Figure 8. Figure 8. SEM micrograph of silicidation of oxidized nanowires. Annealing is performed at 450 C for Figure 8. SEM micrograph of silicidation of oxidized nanowires. Annealing is performed at 450 °C for Figure 8. SEM micrograph of silicidation of oxidized nanowires. Annealing is performed at 450 °C for 80 s. 80 s. 80 s. 3.3. Properties of Silicide–Si Interface 3.3. Properties of Silicide–Si Interface 3.3. Properties of Silicide–Si Interface To investigate the phase of the silicide and the quality of the silicide–Si interface, an exemplary To investigate the phase of the silicide and the quality of the silicide–Si interface, an exemplary To investigate the phase of the silicide and the quality of the silicide–Si interface, an exemplary NW was sectioned parallel to its length. Subsequently, HRTEM was performed. The resulting images NW was sectioned parallel to its length. Subsequently, HRTEM was performed. The resulting NW was sectioned parallel to its length. Subsequently, HRTEM was performed. The resulting are shown in Figure 9. Starting with Ni-rich phases near the Ni-reservoir, the Si fraction increases images are shown in Figure 9. Starting with Ni-rich phases near the Ni-reservoir, the Si fraction images are shown in Figure 9. Starting with Ni-rich phases near the Ni-reservoir, the Si fraction towards the silicide–Si interface. increases towards the silicide–Si interface. increases towards the silicide–Si interface. Appl. Sci. 2019, 9, 3462 9 of 13 Appl. Sci. 2019, 9, x FOR PEER REVIEW 9 of 13 Figure 9. HRTEM images of silicide–Si Schottky junction: (a) NW sectioned along the NW length, Figure 9. HRTEM images of silicide–Si Schottky junction: (a) NW sectioned along the NW length, (b) (b) high magnification image of the interface showing an atomically-sharp Schottky junction, and (c) high magnification image of the interface showing an atomically-sharp Schottky junction, and (c) Fast fourier transform (FFT) studies confirm formation of NiSi phase near the silicide–Si interface. Fast fourier transform (FFT) studies confirm formation of NiSi2 phase near the silicide–Si interface. Annealing is performed at 450 C for 80 s. Annealing is performed at 450 °C for 80 s. HRTEM images show the formation of an atomically sharp Schottky junction, which is required HRTEM images show the formation of an atomically sharp Schottky junction, which is required for size downscaling and enhanced performance of the devices. FFT studies confirmed the formation for size downscaling and enhanced performance of the devices. FFT studies confirmed the formation of the desired NiSi near the silicide–Si interface. The interface orientation is {111} in accordance of the desired NiSi2 near the silicide–Si interface. The interface orientation is {111} in accordance with with [16]. [16]. In order to better understand our experimental observations, we calculated interface energies In order to better understand our experimental observations, we calculated interface energies using density functional theory (DFT) as implemented in Atomistix ToolKit 15.1 [57]. The using density functional theory (DFT) as implemented in Atomistix ToolKit 15.1 [57]. The Perdew– Perdew–Burke–Ernzerhof exchange-correlation function from [58] was used and the reciprocal space Burke–Ernzerhof exchange-correlation function from [58] was used and the reciprocal space was was sampled with a spacing of around 0.25 nm . −1 sampled with a spacing of around 0.25 nm . A periodic structure consisting of alternating silicon and NiSi was studied. The length of the A periodic structure consisting of alternating silicon and NiSi 2 was studied. The length of the resulting supercell perpendicular to the interface was at least 12 nm. Periodic boundary conditions resulting supercell perpendicular to the interface was at least 12 nm. Periodic boundary conditions were employed in the parallel directions. The atomic arrangements of the studied interfaces are were employed in the parallel directions. The atomic arrangements of the studied interfaces are described in [59]. In order to mimic our experimental setup of a nanowire lying on a substrate, the described in [59]. In order to mimic our experimental setup of a nanowire lying on a substrate, the lattice constants in the silicon and NiSi regions were both fixed to the silicon lattice constant of the lattice constants in the silicon and NiSi2 regions were both fixed to the silicon lattice constant of the undisturbed silicon crystal. The atomic coordinates of all atoms within a 10 nm neighborhood around undisturbed silicon crystal. The atomic coordinates of all atoms within a 10 nm neighborhood the interface were relaxed. around the interface were relaxed. We defined the interface energy density as: We defined the interface energy density as: 𝜖 == (E𝐸 −N𝑁 E𝐸 −N𝑁 E𝐸 )//22A𝐴 .. (1) (1) interface interface super super cell cell NiSi NiSi NiSi NiSi SiSi SiSi 2 2 2 2 𝐸 is the total energy of the calculated supercell, 𝐸 are the total energies of bulk supercell NiSi /Si E is the total energy of the calculated supercell, E 2are the total energies of bulk unit supercell NiSi /Si unit cells, and 𝑁 follows from the number of unit cells in the supercell. 𝐴 is the cross-section NiSi /Si cells, and N follows 2 from the number of unit cells in the supercell. A is the cross-section of the NiSi /Si of the system and the factor 2 arises because two interfaces are included in a single supercell. system and the factor 2 arises because two interfaces are included in a single supercell. The calculation resulted in interface energy densities of around 2.2 eV/nm for the {111} The calculation resulted in interface energy densities of around 2.2 eV/nm for the {111} interface interface (an A-type interface was assumed, see [59] for details) and around 4.2 eV/nm for the {110} (an A-type interface was assumed, see [59] for details) and around 4.2 eV/nm for the {110} interface. interface. This effect is partially compensated by the higher area of a {111} interface in a <110> This e ect is partially compensated by the higher area of a {111} interface in a <110> nanowire, because nanowire, because an angle of 35° between the normal vector of the interface and nanowire an angle of 35 between the normal vector of the interface and nanowire orientation occurs. For orientation occurs. For geometrical considerations, the interface area is increased by a factor of geometrical considerations, the interface area is increased by a factor of around 1.2 compared to a {110} around 1.2 compared to a {110} interface. Hence, the total energy contribution of the interface is interface. Hence, the total energy contribution of the interface is smaller for a tilted {111} interface, smaller for a tilted {111} interface, which makes such an interface energetically more favorable. The which makes such an interface energetically more favorable. The experimental observations of the {111} experi interface mental o ar be; ser ther vati efor ons e, of explained the {111} by inter the facalculated ce are; ther interface efore, expl ener ain gy eddensities. by the calculated interface energy densities. 3.4. Electrical Characterization of Fabricated Devices To test the outcome of the silicidation process, electrical characterization of devices based on single unoxidized NWs patterned in both <100> and <110> orientations was carried out by Appl. Sci. 2019, 9, 3462 10 of 13 3.4. Electrical Characterization of Fabricated Devices Appl. T So ci.test 2019the , 9, xoutcome FOR PEERof RE the VIEW silicidati on process, electrical characterization of devices based on single 10 of 13 unoxidized NWs patterned in both <100> and <110> orientations was carried out by back-gating. back-gating. The back-gate voltage (Vbg) was swept between −40 to 40 V in a butterfly loop (0 to 40 V, The back-gate voltage (V ) was swept between 40 to 40 V in a butterfly loop (0 to 40 V, 40 to bg 40 to −40 V, and −40 to 0 V). The drain to source voltage (Vds) was varied from 0.25 to 1 V. The 40 V, and 40 to 0 V). The drain to source voltage (V ) was varied from 0.25 to 1 V. The devices ds devices exhibit large hysteresis as the nanowires are not passivated. We extracted a single sweep exhibit large hysteresis as the nanowires are not passivated. We extracted a single sweep from the from the transfer characteristics, as illustrated in Figure 10. The minimum of the curves was shifted transfer characteristics, as illustrated in Figure 10. The minimum of the curves was shifted to left by to left by 17 V to center the curves around 0 V. An ambipolar behavior of devices was found. The 17 V to center the curves around 0 V. An ambipolar behavior of devices was found. The currents in currents in <110>-oriented devices are higher compared to the currents in <100> devices. At Vds = 1 V, <110>-oriented devices are higher compared to the currents in <100> devices. At V = 1 V, the values ds the values of n- and p- currents in <110> devices are 40.7 and 450.0 nA, while in <100> the values are of n- and p- currents in <110> devices are 40.7 and 450.0 nA, while in <100> the values are 16.8 and 16.8 and 207.0 nA, respectively. The shift of original curves away from 0 V is attributed to a built-in 207.0 nA, respectively. The shift of original curves away from 0 V is attributed to a built-in potential on potential on the NW surface in ambient conditions in the absence of passivation with, for example, the NW surface in ambient conditions in the absence of passivation with, for example, an oxidation an oxidation layer [34]. Unipolar behavior in these devices can be attained by using multiple gate layer [34]. Unipolar behavior in these devices can be attained by using multiple gate electrodes [9], electrodes [9], while p/n current symmetry, which is required for energy efficient functioning of while p/n current symmetry, which is required for energy ecient functioning of circuits based on circuits based on RFETs, can be tuned by oxidation induced stress [9,14]. RFETs, can be tuned by oxidation induced stress [9,14]. Figure 10. Electrical characterization with back-gate sweeping of single NWs-based devices with Figure 10. Electrical characterization with back-gate sweeping of single NWs-based devices with (a) (a) <110> and (b) <100> orientations. The devices show ambipolar behavior. To facilitate a p- and <110> and (b) <100> orientations. The devices show ambipolar behavior. To facilitate a p- and n-current comparison, curves were shifted left by 17 V to center. Currents in <110>-oriented devices n-current comparison, curves were shifted left by 17 V to center. Currents in <110>-oriented devices are higher than those in <100>. Channel width and length for both types of nanowires are 20 nm and are higher than those in <100>. Channel width and length for both types of nanowires are 20 nm and 3 m, respectively. 3 µ m, respectively. 4. Conclusions 4. Conclusions In this work, we studied Ni silicidation of silicon thin films and top-down fabricated SiNWs by In this work, we studied Ni silicidation of silicon thin films and top-down fabricated SiNWs by using the RTA technique. We paid special attention to the formation of the required silicide phase using the RTA technique. We paid special attention to the formation of the required silicide phase (NiSi ), the quality of the Schottky junctions, and the reproducibility of the silicide length along (NiSi2), the quality of the Schottky junctions, and the reproducibility of the silicide length along the the nanowires, which are very important for device performance and scalability. To investigate the nanowires, which are very important for device performance and scalability. To investigate the influence of the NW orientations on the silicidation process, SiNWs with di erent crystallographic influence of the NW orientations on the silicidation process, SiNWs with different crystallographic orientations were silicided. Although large scattering in the silicidation lengths of nanowires was orientations were silicided. Although large scattering in the silicidation lengths of nanowires was observed even for the nanowires of the same orientation, the silicidation in <100> nanowires is observed even for the nanowires of the same orientation, the silicidation in <100> nanowires is found found to be faster than that in <110> nanowires. Furthermore, TEM and FFT analyses revealed the to be faster than that in <110> nanowires. Furthermore, TEM and FFT analyses revealed the formation of sharp Schottky junctions with {111} interfaces and the NiSi phase of silicide required formation of sharp Schottky junctions with {111} interfaces and the NiSi2 phase of silicide required for the fabrication of RFETs. Density functional theory calculations showed that {111} interfaces are for the fabrication of RFETs. Density functional theory calculations showed that {111} interfaces are energetically favorable. Control over the di usion length of silicide into the NW was not achieved. energetically favorable. Control over the diffusion length of silicide into the NW was not achieved. This makes placing of top gates on the Schottky junctions and downscaling of devices a challenge. This makes placing of top gates on the Schottky junctions and downscaling of devices a challenge. Alternative annealing techniques with shorter annealing times, like, for example, flash lamp annealing, Alternative annealing techniques with shorter annealing times, like, for example, flash lamp may be employed to have better control over the silicidation process. Transfer characteristics of the annealing, may be employed to have better control over the silicidation process. Transfer back-gated devices with single unoxidized NWs patterned in <100> and <110> orientations illustrate characteristics of the back-gated devices with single unoxidized NWs patterned in <100> and <110> orientations illustrate ambipolar behavior, with currents in <110>-oriented devices being higher compared to the currents in <100> devices. Appl. Sci. 2019, 9, 3462 11 of 13 ambipolar behavior, with currents in <110>-oriented devices being higher compared to the currents in <100> devices. Author Contributions: Conceptualization, D.D. and M.B.K.; methodology, D.D., M.B.K. and J.K. (device fabrication), F.F., S.G. and J.S. (simulations), M.L., S.B. and U.M. (lamella preparation and TEM); writing—original draft preparation, M.B.K., D.D. and F.F.; writing—review and editing, Y.M.G., M.L., F.F., S.G., J.S., A.E. and W.M.W.; supervision, A.E. and Y.M.G. All authors have read and approved this paper for submission. Funding: We acknowledge funding by the Helmholtz Initiative and Networking Funds for support through the International Helmholtz Research School NanoNet via grant VH-KO-606 (M.B.K, D.D., F.F., S.B., W.M.W, S.G., J.S., A.E.) and the W2/W3 Programme for the first-time appointment of excellent female scientists via grant W2/3-026 (S.G., F.F.). Acknowledgments: Authors thank Claudia Neisser and Tommy Schönherr for their help in conducting experiments. We are thankful to Habil Peter Zahn for his support in administrative a airs. We also thank Phanish Chava for his help in data compilation. Conflicts of Interest: The authors declare no conflicts of interest. References 1. Yeo, Y.C.; Gong, X.; Van Dal, M.J.H.; Vellianitis, G.; Passlack, M. Germanium-based transistors for future high performance and low power logic applications. In Proceedings of the 2015 IEEE International Electron Devices Meeting (IEDM), Washington, DC, USA, 7–9 December 2015; p. 2. 2. Dong, R.; Han, P.; Arora, H.; Ballabio, M.; Karakus, M.; Zhang, Z.; Shekhar, C.; Adler, P.; Petkov, P.S.; Erbe, A.; et al. High-mobility band-like charge transport in a semiconducting two-dimensional metal–organic framework. Nat. Mater. 2018, 17, 1027. [CrossRef] 3. Wang, J.; Yao, Q.; Huang, C.-W.; Zou, X.; Liao, L.; Chen, S.; Fan, Z.; Zhang, K.; Wu, W.; Xiao, X.; et al. High Mobility MoS2 Transistor with Low Schottky Barrier Contact by Using Atomic Thick h-BN as a Tunneling Layer. Adv. Mater. 2016, 28, 8302–8308. [CrossRef] [PubMed] 4. Chau, R.; Brask, J.; Datta, S.; Dewey, G.; Doczy, M.; Doyle, B.; Kavalieros, J.; Jin, B.; Metz, M.; Majumdar, A.; et al. Application of high- gate dielectrics and metal gate electrodes to enable silicon and non-silicon logic nanotechnology. Microelectron. Eng. 2005, 80, 1–6. [CrossRef] 5. Arabi, K.; Samadi, K.; Du, Y. 3D VLSI: A Scalable Integration beyond 2D. In Proceedings of the 2015 Symposium on International Symposium on Physical Design, Monterey, CA, USA, 29 March–1 April 2015; pp. 1–7. 6. Das Sarma, S.; Fabian, J.; Hu, X.; Žutic, ´ I. Spin electronics and spin computation. Solid State Commun. 2001, 119, 207–215. [CrossRef] 7. Risch, L. Pushing CMOS beyond the roadmap. Solid-State Electron. 2006, 50, 527–535. [CrossRef] 8. Heinzig, A.; Slesazeck, S.; Kreupl, F.; Mikolajick, T.; Weber, W.M. Reconfigurable Silicon Nanowire Transistors. Nano Lett. 2011, 12, 119–124. [CrossRef] 9. Heinzig, A.; Mikolajick, T.; Trommer, J.; Grimm, D.; Weber, W.M. Dually Active Silicon Nanowire Transistors and Circuits with Equal Electron and Hole Transport. Nano Lett. 2013, 13, 4176–4181. [CrossRef] 10. Chang, Y.-J.; Erskine, J.L. Di usion layers and the Schottky-barrier height in nickel silicide—silicon interfaces. Phys. Rev. B 1983, 28, 5766–5773. [CrossRef] 11. Ottaviani, G.; Tu, K.N.; Mayer, J.W. Barrier heights and silicide formation for Ni, Pd, and Pt on silicon. Phys. Rev. B 1981, 24, 3354–3359. [CrossRef] 12. Léonard, F.; Talin, A.A. Electrical contacts to one- and two-dimensional nanomaterials. Nat. Nanotechnol. 2011, 6, 773–783. [CrossRef] 13. Larson, J.; Snyder, J. Overview and status of metal S/D Schottky-barrier MOSFET technology. IEEE Trans. Electron Devices 2006, 53, 1048–1058. [CrossRef] 14. Simon, M.; Heinzig, A.; Trommer, J.; Baldauf, T.; Mikolajick, T.; Weber, W.M. Top-down technology for reconfigurable nanowire FETs with symmetric on-currents. IEEE Trans. Nanotechnol. 2017, 16, 1. [CrossRef] 15. Bagdasarov, K.S.; Lube, É.L. Growth of Crystals; Springer: Boston, MA, USA, 1991; Volume 16, p. 278. 16. Weber, W.M.; Geelhaar, L.; Graham, A.P.; Unger, E.; Duesberg, G.S.; Liebau, M.; Pamler, W.; Cheze, C.; Riechert, H.; Lugli, P.; et al. Silicon-Nanowire Transistors with Intruded Nickel-Silicide Contacts. Nano Lett. 2006, 6, 2660–2666. [CrossRef] [PubMed] Appl. Sci. 2019, 9, 3462 12 of 13 17. Tang, W.; Dayeh, S.A.; Picraux, S.T.; Huang, J.Y.; Tu, K.-N. Ultrashort Channel Silicon Nanowire Transistors with Nickel Silicide Source/Drain Contacts. Nano Lett. 2012, 12, 3979–3985. [CrossRef] [PubMed] 18. Tang, W.; Nguyen, B.-M.; Chen, R.; Dayeh, S.A. Solid-state reaction of nickel silicide and germanide contacts to semiconductor nanochannels. Semicond. Sci. Technol. 2014, 29, 054004. [CrossRef] 19. Beregovsky, M.; Katsman, A.; Hajaj, E.; Yaish, Y. Di usion formation of nickel silicide contacts in SiNWs. Solid-State Electron. 2013, 80, 110–117. [CrossRef] 20. Chen, Y.; Huang, Y. Phase control in solid state silicide nanowire formation. Phys. Status Solidi 2013, 10, 1666–1669. [CrossRef] 21. Yaish, Y.E.; Katsman, A.; Cohen, G.M.; Beregovsky, M. Kinetics of nickel silicide growth in silicon nanowires: From linear to square root growth. J. Appl. Phys. 2011, 109, 94303. [CrossRef] 22. Ogata, K.; Sutter, E.; Zhu, X.; Hofmann, S. Ni-silicide growth kinetics in Si and Si/SiO2 core/shell nanowires. Nanotechnology 2011, 22, 365305. [CrossRef] [PubMed] 23. Katsman, A.; Beregovsky, M.; Yaish, Y.E. Evolution of nickel silicide intrusions in silicon nanowires during thermal cycling. J. Appl. Phys. 2013, 113, 084305. [CrossRef] 24. Dellas, N.S.; Abraham, M.; Minassian, S.; Kendrick, C.; Mohney, S.E. Kinetics of reactions of Ni contact pads with Si nanowires. J. Mater. Res. 2011, 26, 2282–2285. [CrossRef] 25. Lu, K.-C.; Tu, K.N.; Wu, W.W.; Chen, L.J.; Yoo, B.-Y.; Myung, N.V. Point contact reactions between Ni and Si nanowires and reactive epitaxial growth of axial nano-NiSi/Si. Appl. Phys. Lett. 2007, 90, 253111. [CrossRef] 26. Dellas, N.S.; Liu, B.Z.; Eichfeld, S.M.; Eichfeld, C.M.; Mayer, T.S.; Mohney, S.E. Orientation dependence of nickel silicide formation in contacts to silicon nanowires. J. Appl. Phys. 2009, 105, 94309. [CrossRef] 27. Dellas, N.S.; Schuh, C.J.; Mohney, S.E. Silicide formation in contacts to Si nanowires. J. Mater. Sci. 2012, 47, 6189–6205. [CrossRef] 28. Lin, Y.-C.; Chen, Y.; Xu, D.; Huang, Y. Growth of Nickel Silicides in Si and Si/SiOx Core/Shell Nanowires. Nano Lett. 2010, 10, 4721–4726. [CrossRef] [PubMed] 29. Appenzeller, J.; Knoch, J.; Tutuc, E.; Reuter, M.; Guha, S. Dual-gate silicon nanowire transistors with nickel silicide contacts. In Proceedings of the 2006 International Electron Devices Meeting, San Francisco, CA, USA, 11–13 December 2006. 30. Katsman, A.; Beregovsky, M.; Yaish, Y.E. Formation and Evolution of Nickel Silicide in Silicon Nanowires. IEEE Trans. Electron Devices 2014, 61, 3363–3371. [CrossRef] 31. Chou, Y.-C.; Tang, W.; Chiou, C.-J.; Chen, K.; Minor, A.M.; Tu, K.N. E ect of Elastic Strain Fluctuation on Atomic Layer Growth of Epitaxial Silicide in Si Nanowires by Point Contact Reactions. Nano Lett. 2015, 15, 4121–4128. [CrossRef] [PubMed] 32. Habicht, S.; Zhao, Q.-T.; Feste, S.F.; Knoll, L.; Trellenkamp, S.; Ghyselen, B.; Mantl, S. Electrical characterization of strained and unstrained silicon nanowires with nickel silicide contacts. Nanotechnology 2010, 21, 105701. [CrossRef] 33. Habicht, S.; Zhao, Q.; Feste, S.; Knoll, L.; Trellenkamp, S.; Bourdelle, K.K.; Mantl, S. NiSi nano-contacts to strained and unstrained silicon nanowires. In Proceedings of the 2011 Materials for Advanced Metallization (MAM), Dresden, Germany, 8–12 May 2011. 34. Pregl, S.; Weber, W.M.; Nozaki, D.; Kunstmann, J.; Baraban, L.; Opitz, J.; Mikolajick, T.; Cuniberti, G. Parallel arrays of Schottky barrier nanowire field e ect transistors: Nanoscopic e ects for macroscopic current output. Nano Res. 2013, 6, 381–388. [CrossRef] 35. D’Heurie, F.M.; Lavoie, C.; Gas, P.; Philibert, J. Reactive Phase Formation: Some Theory and Applications. In Di usion Processes in Advanced Technological Materials; Springer Science and Business Media LLC: Berlin, Germany, 2005; pp. 283–332. 36. Laurila, T.; Molarius, J. Reactive Phase Formation in Thin Film Metal/Metal and Metal/Silicon Di usion Couples. Crit. Rev. Solid State Mater. Sci. 2003, 28, 185–230. [CrossRef] 37. Ohring, M. Chapter 1-A Review of Materials Science. In Materials Science of Thin Films, 2nd ed.; Academic Press: San Diego, CA, USA, 2002; pp. 1–56. 38. Tu, K.-N. Silicon and Silicide Nanowires: Applications, Fabrication, and Properties; Pan Stanford Publishing: Redwood City, CA, USA, 2013. Appl. Sci. 2019, 9, 3462 13 of 13 39. Koyama, M.; Shigemori, N.; Ozawa, K.; Tachi, K.; Kakushima, K.; Nakatsuka, O.; Ohmori, K.; Tsutsui, K.; Nishiyama, A.; Sugii, N.; et al. Si/Ni-Silicide Schottky junctions with atomically flat interfaces using NiSi source. In Proceedings of the European Solid-State Device Research Conference (ESSDERC), Helsinki, Finland, 12–16 September 2011; pp. 231–234. 40. Iwai, H.; Ohguro, T.; Ohmi, S.-I. NiSi salicide technology for scaled CMOS. Microelectron. Eng. 2002, 60, 157–169. [CrossRef] 41. Baglin, J.; Atwater, H.; Gupta, D.; D’Heurle, F. Radioactive Ni tracer study of the nickel silicide growth mechanism. Thin Solid Films 1982, 93, 255–264. [CrossRef] 42. Julies, B.A.; Knoesen, D.; Pretorius, R.; Adams, D. A study of the NiSi to NiSi transition in the Ni–Si binary system1. Thin Solid Films 1999, 347, 201–207. [CrossRef] 43. D’Heurle, F. Di usion in intermetallic compounds with the CaF structure: A marker study of the formation of NiSi thin films. J. Appl. Phys. 1982, 53, 5678. [CrossRef] 44. Wu, Y.; Xiang, J.; Yang, C.; Lu, W.; Lieber, C.M. Single-crystal metallic nanowires and metal/semiconductor nanowire heterostructures. Nature 2004, 430, 61–65. [CrossRef] [PubMed] 45. Schmitt, A.L.; Higgins, J.M.; Szczech, J.R.; Jin, S. Synthesis and applications of metal silicide nanowires. J. Mater. Chem. 2010, 20, 223–235. [CrossRef] 46. Murarka, S.P. Chapter IV–Formation. In Silicides for Vlsi Applications; Murarka, S.P., Ed.; Academic Press: San Diego, CA, USA, 1983; pp. 99–131. 47. Lavoie, C.; D’Heurle, F.; Detavernier, C.; Cabral, C. Towards implementation of a nickel silicide process for CMOS technologies. Microelectron. Eng. 2003, 70, 144–157. [CrossRef] 48. De Marchi, M.; Sacchetto, D.; Frache, S.; Zhang, J.; Gaillardon, P.-E.; Leblebici, Y.; De Micheli, G. Polarity control in double-gate, gate-all-around vertically stacked silicon nanowire FETs. In Proceedings of the 2012 International Electron Devices Meeting, San Francisco, CA, USA, 10–13 December 2012; p. 8. 49. Chen, Y.; Lin, Y.-C.; Zhong, X.; Cheng, H.-C.; Duan, X.; Huang, Y. Kinetic Manipulation of Silicide Phase Formation in Si Nanowire Templates. Nano Lett. 2013, 13, 3703–3708. [CrossRef] 50. Yamashita, H.; Kosugiyama, H.; Shikahama, Y.; Hashimoto, S.; Takei, K.; Sun, J.; Matsukawa, T.; Masahara, M.; Watanabe, T. Impact of thermal history of Si nanowire fabrication process on Ni silicidation rate. Jpn. J. Appl. Phys. 2014, 53, 085201. [CrossRef] 51. Mirza, M.M.; Velha, P.; Docherty, K.E.; Samarelli, A.; Ternent, G.; Zhou, H.; Li, X.; Paul, D.J. Nanofabrication of high aspect ratio (50:1) sub-10 nm silicon nanowires using inductively coupled plasma etching. J. Vac. Sci. Technol. B 2012, 30, 6. [CrossRef] 52. Henschel, W.; Georgiev, Y.M.; Kurz, H. Study of a high contrast process for hydrogen silsesquioxane as a negative tone electron beam resist. J. Vac. Sci. Technol. B 2003, 21, 2018. [CrossRef] 53. Liu, H.I. Self-limiting oxidation of Si nanowires. J. Vac. Sci. Technol. B 1993, 11, 2532. [CrossRef] 54. Liu, H.I.; Biegelsen, D.K.; Ponce, F.; Johnson, N.M.; Pease, R.F.W. Self-limiting oxidation for fabricating sub-5 nm silicon nanowires. Appl. Phys. Lett. 1994, 64, 1383–1385. [CrossRef] 55. Huang, J.; Loeer, M.; Muehle, U.; Moeller, W.; Mulders, J.; Kwakman, L.; Van Dorp, W.; Zschech, E. Si amorphization by focused ion beam milling: Point defect model with dynamic BCA simulation and experimental validation. Ultramicroscopy 2018, 184, 52–56. [CrossRef] 56. Reimer, L. Energy-Filtered Transmission Electron Microscopy; Springer: Berlin/Heidelberg, Germany, 1995. 57. Atomistix ToolKit. 2015. Available online: https://www.synopsys.com/silicon/quantumatk.html (accessed on 16 August 2019). 58. Perdew, J.P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865–3868. [CrossRef] [PubMed] 59. Fuchs, F.; Gemming, S.; Schuster, J. Electron transport through NiSi –Si contacts and their role in reconfigurable field-e ect transistors. J. Phys. Condens. Matter 2019, 31, 355002. [CrossRef] [PubMed] © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Journal
Applied Sciences – Multidisciplinary Digital Publishing Institute
Published: Aug 22, 2019