Progress in Surface Science 88 (2013) 39–60
Contents lists available at SciVerse ScienceDirect
Progress in Surface Science journal homepage: www.elsevier.com/locate/progsurf
Review
Oxide-free hybrid silicon nanowires: From fundamentals to applied nanotechnology Muhammad Y. Bashouti a,⇑, Kasra Sardashti a, Sebastian W. Schmitt a, Matthias Pietsch a, Jürgen Ristein b, Hossam Haick c, Silke H. Christiansen a,d a
Max-Planck Institute for the Science of Light, Günther-Scharowsky-Str.-1, Erlangen 91058, Germany Technical Physics, University of Erlangen – Nürnberg, Erlangen 91058, Germany c The Department of Chemical Engineering and Russell Berrie Nanotechnology Institute, Technion – Israel Institute of Technology, Haifa 32000, Israel d Institute of Photonic Technology, Albert-Einstein-Str. 1, 07745 Jena, Germany b
a r t i c l e
i n f o
Keywords: Silicon nanowire Hybrid functionalization Solar cells Heterojunction Field effect transistors
a b s t r a c t The ability to control physical properties of silicon nanowires (Si NWs) by designing their surface bonds is important for their applicability in devices in the areas of nano-electronics, nano-photonics, including photovoltaics and sensing. In principle a wealth of different molecules can be attached to the bare Si NW surface atoms to create e.g. Si–O, Si–C, Si–N, etc. to mention just the most prominent ones. Si–O bond formation, i.e. oxidation usually takes place automatically as soon as Si NWs are exposed to ambient conditions and this is undesired is since a defective oxide layer (i.e. native silicon dioxide – SiO2) can cause uncontrolled trap states in the band gap of silicon. Surface functionalization of Si NW surfaces with the aim to avoid oxidation can be carried out by permitting e.g. Si–C bond formation when alkyl chains are covalently attached to the Si NW surfaces by employing a versatile two-step chlorination/alkylation process that does not affect the original length and diameter of the NWs. Termination of Si NWs with alkyl molecules through covalent Si–C bonds can provide long term stability against oxidation of the Si NW surfaces. The alkyl chain length determines the molecular coverage of Si NW surfaces and thus the surface energy and next to simple Si–C bonds even bond types such as C@C and C„C can be realized. When integrating differently functionalized Si NWs in functional devices such as field effect transistors (FETs) and solar cells, the physical properties of the resultant devices vary. Ó 2013 Elsevier Ltd. All rights reserved.
⇑ Corresponding author. E-mail addresses:
[email protected],
[email protected] (M.Y. Bashouti). 0079-6816/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.progsurf.2012.12.001
40
M.Y. Bashouti et al. / Progress in Surface Science 88 (2013) 39–60
Contents 1. 2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Functionalization of oxide-free Si NW surfaces via the chlorination–alkylation process . . . . . . . . . . . . 2.1. Fabrication of Si NWs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Organic functionalization via chlorination/alkylation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1. Hydrogen-terminated Si NW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2. Chlorination/alkylation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Secondary functionalization: alkenyl and alkynyl as the starting monolayers . . . . . . . . . . . . . . . 2.4. Stability of molecules on the Si NW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1. Effect of molecular coverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2. Effect of molecular backbone bonds: C–C vs. CC vs. CC . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Effect of surface energy: Si NWs vs. planar surfaces. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hybrid Si NW electronic properties and devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Manipulation of the surface fermi level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Controlling the work function, electron affinity, surface dipoles and band bending . . . . . . . . . . 3.3. Hybrid Si NW-based field effect transistor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Hybrid Si NW-based solar cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1. Introduction Much effort has been devoted to developing new nanomaterial-based technological devices with significant performance improvements at a reduced cost [1–4]. In this quest, a substantial number of nanomaterials have been developed, including but not limited to nanoparticles [5–14], nanorods [15–18], and nanowires [18–25]. Nanowires provide additional advantages compared to the other nanostructures in terms of their strong light absorbance, efficient charge separation and direct charge transport path. This makes them a promising candidate for many technological applications. In short, nanowires are important one-dimensional nanostructures that have demonstrated advantageous characteristics for applications in electronics [26–28], photovoltaics [29] and sensing [30–33]. Up till now, nanowires of many different inorganic materials have been synthesized and characterized, including silicon [34], gallium nitride [35] and lead selenide [18,36] among others. The use of silicon nanowires (Si NWs) in particular to further miniaturize devices while avoiding major process changes and their corresponding costs has been a topic of much research in the past decade. Thus, Si NWs are generally considered to be an essential class of nanodevice building blocks [37– 39]. Si NWs have shown significant potential for field effect transistors (FETs). Potential applications of Si NWs in FETs includes aligning on insulating substrate surface, selective deposition of the source and drain contacts on the Si NW edges, and the configuration of either a bottom or top gate electrode [26,40]. Another significant potential application of Si NWs is solar power generation. The indirect optical band gap of 1.12 eV, low absorption co-efficient of 104/cm and high wafer cost of bulk Si reduces the suitability of this material for photovoltaic and optoelectronic devices. Differently engineered band gaps (1.4 eV) and a high absorption co-efficient (at least 105/cm) are required to significantly improve the solar cell efficiency. Si NWs are promising alternatives to bulk Si for use in solar cell applications due to their high light harvest; convenient band gap tuneability from 1.1 to 1.4 eV, which is achievable through decreases in the diameter [41]; and the possibility of fabricating the structures with a axial and lateral junction to decrease the charge bath separation [25,42–44]. Surface treatment is one of the main obstacles to Si NW application. A large body of chemistry has been developed for chemically linking moieties to oxidized Si NW surfaces, generally through –OH
M.Y. Bashouti et al. / Progress in Surface Science 88 (2013) 39–60
41
chemistry. Many pertinent applications have been illustrated [28,45–48]. Although Si NWs offer unique opportunities as building blocks for nanoelectronic devices, the presence of native oxide (SiOx), which instantly forms on the Si NW surfaces upon air exposure, is undesirable [49,50]. This is because of the poor quality of the oxide that forms on the surface, which induces a large number of unwanted and uncontrolled interfacial states in the Si band gap [50,51]. The low quality, highly impure oxides that form in ambient conditions have a tendency to cause uncontrolled oxide/silicon interfaces and surface states. This necessitates the protection of Si NW surfaces against oxidation. Moreover, due to the larger surface to volume ratio, surface properties become more significant in smaller Si NWs and dominate the properties of the device as a whole [26]. Therefore, it is necessary to explore models and methods to predict and control the surface physical characteristics of oxide-free Si NW. Etched Si NWs with surface-bound hydrogen atoms, the simplest termination group, show low charge carrier surface recombination velocities [41,52]. However, these surfaces tend to oxidize within a few minutes of exposure to ambient air, leading to higher surface recombination velocities [40]. Therefore, it is of considerable interest to develop a sound strategy to prevent extensive Si NW surface oxidation while preserving the low surface recombination velocities. A promising method for controlling the surface properties of Si is termination through the use of organic molecules (polar and nonpolar) which form Si–C surface bonds. Functionalization by Si–C bonds causes the stability time to increase to a few hundred hours at room temperature [53–55]. Therefore, termination of the dangling surface bonds with chemical and biochemical moieties [39,56–58] can affect the device reliability over time. This is expected to have extreme significance on the final physical and chemical properties of the Si NWs. Generally speaking; the functionalization of organic molecules introduces a net electrical dipole perpendicular to the surface/interface. The dipole, in turn, modifies the work function and electron affinity, alters the band offset and band bending, and tunes the surface Fermi level [26]. Moreover, many future devices will require their building blocks to be selectively sensitive to the environment. Indeed, the main objectives of molecular functionalization can be outlined as follows: (1) to increase the oxidation resistance of Si NWs; (2) to allow a systematic tuning of the desired physical properties of an electric device by appropriate choice of the functional groups [59,60]; (3) to exhibit selectivity to different environmental species, depending on the type of molecules and their coverage [61]. In order to achieve the above objectives, several techniques that induce a high density of organic molecules attachments to the silicon surface have been suggested. Here we will describe one of these techniques while showing the disadvantages of the others. The most widely reported organic monolayers on the oxide-free Si surface are alkyl chains. As a starting step, the oxide is removed and surfaces are covered with either hydrogen or halogens. The H-terminated monolayer can be converted to alkyl monolayers through the use of alkyl-magnesium reagents [31,53–55,62–64] while the halogenated surfaces can be alkylated using alkyl-magnesium or alkyl-lithium reagents [65]. Alkylation has been successfully carried out on two dimensional surfaces through free radical initiation methods such as irradiation with ultraviolet light [66,67], chemical free-radical activation [68], thermal activation [69], Lewis acid catalyzed hydrosilylation [70,71], and visible-light-initiated modification [72]. However, these methods have as yet not been applied on Si NW. Unless extreme measures are taken, these reactions may result in an incomplete coverage of the organic monolayer (i.e. less than one monolayer of coverage) as well as significant amounts of oxygen on the surface [73,74]. Alternative reactions such as terpyridine termination have been directly demonstrated on hydrogenated Si NW [75]. However, these reactions are limited to terpyridine and cover the surface only partially. The current review describes strategies used to prepare organic monolayers on Si NWs, which achieve both electrical and chemical passivation, and allow for the introduction of productive oxide free components into practical nanodevices. A detailed report on scientific progress in the production and characterization of oxide-free high-density organic monolayers (single and multiple) on Si NW, through a two-step process involving chlorination/alkylation is described in Section 2. The kinetics of the alkylation process and the effect of organic chain length as well as second monolayer functionalization are also discussed. The stability of the hydrogen and organic molecules on the Si NW and the effect of molecular coverage and bonding type are demonstrated. Section 3 presents two examples of functional devices based on hybrid Si NWs: Field Effect Transistors (FETs) and solar cells. The electrical characteristics of molecularly-modified Si NW-based devices are presented, analyzed and compared to pristine Si NWs. Fig. 1 schematically illustrates the organization of the current review.
42
M.Y. Bashouti et al. / Progress in Surface Science 88 (2013) 39–60
Fig. 1. Scheme showing the basic steps from the synthesis to the application of hybrid Si NWs.
2. Functionalization of oxide-free Si NW surfaces via the chlorination–alkylation process 2.1. Fabrication of Si NWs The fabrication of Si NWs with a controlled diameter, length, and electronic properties is essential to technical application [76–78]. Significant progress has been made in recent years in the development of superficially controlled methods of Si NW fabrication [79]. There are two general Si NW fabrication approaches: bottom-up and top-down. The bottom-up approach is an assembly process joining Si atoms to form Si NWs and includes the vapor–liquid solid (VLS) growth technique [80]. The top-down approach prepares Si NWs by etching a bulk Si by a lithography or etching technique such as reactive ion etching (RIE) [81] or metal-catalyzed electroless etching (MCEE) [82]. Each approach has its advantages and disadvantages. For example, the top-down method has difficulty producing Si NWs with diameters below 10 nm. In contrast, the bottom-up approach can readily grow Si NWs with very small diameters
