Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2019 Apr 17;13(5):5572–5582. doi: 10.1021/acsnano.9b00538

In-Plane Nanowires with Arbitrary Shapes on Amorphous Substrates by Artificial Epitaxy

Regev Ben-Zvi †, Hadassah Burrows †, Mark Schvartzman †, Ora Bitton §, Iddo Pinkas §, Ifat Kaplan-Ashiri §, Olga Brontvein §, Ernesto Joselevich †,*
PMCID: PMC6994061  PMID: 30995393

Abstract

graphic file with name nn9b00538_0007.jpg

The challenge of nanowire assembly is still one of the major obstacles toward their efficient integration into functional systems. One strategy to overcome this obstacle is the guided growth approach, in which the growth of in-plane nanowires is guided by epitaxial and graphoepitaxial relations with the substrate to yield dense arrays of aligned nanowires. This method relies on crystalline substrates which are generally expensive and incompatible with silicon-based technologies. In this work, we expand the guided growth approach into noncrystalline substrates and demonstrate the guided growth of horizontal nanowires along straight and arbitrarily shaped amorphous nanolithographic open guides on silicon wafers. Nanoimprint lithography is used as a high-throughput method for the fabrication of the high-resolution guiding features. We first grow five different semiconductor materials (GaN, ZnSe, CdS, ZnTe, and ZnO) along straight ridges and trenches, demonstrating the generality of this method. Through crystallographic analysis we find that despite the absence of any epitaxial relations with the substrate, the nanowires grow as single crystals in preferred crystallographic orientations. To further expand the guided growth approach beyond straight nanowires, GaN and ZnSe were grown also along curved and kinked configurations to form different shapes, including sinusoidal and zigzag-shaped nanowires. Photoluminescence and cathodoluminescence were used as noninvasive tools to characterize the sine wave-shaped nanowires. We discuss the similarities and differences between in-plane nanowires grown by epitaxy/graphoepitaxy and artificial epitaxy in terms of generality, morphology, crystallinity, and optical properties.

Keywords: artificial epitaxy, nanoimprint lithography, guided nanowires, ZnSe, GaN, CdS, ZnTe, ZnO

The growth of straight and aligned horizontal nanowires by the guided-growth approach has been widely demonstrated over the past few years,1−8 mainly due to its potential as a way to fabricate and study nanowire-based planar devices.9−16 A vapor–liquid–solid (VLS) process,17 guided by a crystalline surface, yields horizontal nanowires with controlled directions and crystallographic orientations, determined by the epitaxial and graphoepitaxial relations with the substrate. In epitaxial growth, the guidance occurs along specific lattice directions of a flat surface according to the atomic registry of the nanowire and the underlying substrate. In graphoepitaxy, the nanowires grow along nanosteps or nanogrooves and must satisfy both geometrical constraints and lattice constraints with the exposed facets (Figure 1a). This guided growth approach, by either epitaxy or graphoepitaxy, in which both assembly and alignment of nanowires are achieved during growth, eliminates the need for postgrowth processes and enables deterministic control over the position, direction, and length of the nanowires. Although in the past few years it was shown to be useful and general to a large variety of materials and substrates,18−23 it is limited to crystalline substrates and yields only straight or randomly kinked nanowires, but no arbitrary shapes designed at will.

Figure 1.

Figure 1

Open in a new tab

Schematic representation of the three guiding modes: epitaxy, graphoepitaxy, and artificial epitaxy. (a) Straight nanowires are the common outcome of epitaxy (guided growth along flat crystal planes) and graphoepitaxy (guided growth along faceted crystal planes). In artificial epitaxy, nanowires grow along a noncrystalline template and (b) can therefore yield any predesigned shape.

The opportunity to expand the guided growth approach to nanowires with arbitrary shapes is clearly attractive from a technological point of view, enabling the creation of specialized devices with configurations suitable for optical waveguides and electric circuits. Recently, tremendous efforts have been invested in creating in-plane geometries of nanowire networks in pursuit of quantum-computing devices based on Majorana fermions,24−26 using in-plane selective area growth (SAG), in which nanowires are created in a patterned mask on a single-crystal substrate with clear epitaxial relations.27−29 The ability to eliminate the use of a crystalline substrate offers the possibility of using a much larger variety of substrates, such as flexible substrates and commonly used oxidized silicon,30 for CMOS- and MEMS-compatible optoelectronics. The extension of the guided growth approach beyond crystallographic guidance of straight nanowires is not trivial. The question is whether geometry alone is enough to guide the growth of horizontal nanowires, and if so, what would be the effect on their morphology and crystallinity. In addition, we wish to explore the possibility to guide nanowires along arbitrary shapes. Growing nanowires with predesigned curvature can be used for the study of strain-related effects on crystallinity and properties of nanowires from different materials. It was widely demonstrated that when vertically grown nanowires are bent or curved after growth, a red-shift of the near band edge (NBE) emission is observed in points with higher curvature along the nanowire, indicating a strain-related decrease of the band gap.31−35 However, nanowires growing along curved features might differ from nanowires under postgrowth straining, and to the best of our knowledge, the effect on the optical properties in this case has yet to be studied.

The concept of using amorphous lines as nucleation sites and growth guides was termed “artificial epitaxy” in Givargizov’s pioneering work.36 This type of oriented crystal growth is sometimes also referred to as graphoepitaxy, but we use the term artificial epitaxy to specify that the substrate is amorphous, whereas in graphoepitaxy the substrate can also be crystalline (even though the growth is guided by relief features larger than the lattice parameter). The idea of artificial epitaxy is very similar to the scratching of a glass beaker in order to induce and guide recrystallization processes. In artificial epitaxy, growth occurs along relief features on an amorphous substrate by geometric guidance alone. In the past, we found such growth to be rather challenging. In fact, our first attempts to guide GaN nanowires along templates patterned by photolithography failed, primarily due to the limitations of the lithographic technique. The microscale dimensions of the templates were too large, their features too rough, and their density too sparse for successful guidance of nanowires.37 We have overcome these limitations by using higher resolution lithographic techniques, such as electron-beam lithography (EBL), which can produce any arbitrary pattern (Figure 1b). A nonepitaxial, in-plane guided shaping of nanowires is possible by several approaches. For example, the VLS growth of Si and Ge nanowires in predefined shapes was demonstrated by confining the growth in closed channels created by EBL followed by multiple fabrication steps.38 The growth in this case is limited by the diffusion into the closed channel, and the size and shape of the nanowire is determined by the dimensions and quality of the channel. Alternatively, nanowires can be guided along the edges of a shaped open trench, as demonstrated by Xue etal., and presents a monolike high crystallinity.39 However, this process was specifically developed for the solid–liquid–solid growth of Si nanowires40−42 and not for the general growth of nanowires from different materials.

A different approach for creating nanowires with different geometries is based on postgrowth shaping and usually involves placement of vertically grown nanowires along lithographic patterns, such as anchors that result in U-shaped nanowires,43 or scaffolding that results in periodically strained nanowires.44 These techniques offer only partial control over the geometry of nanowires and lack the above-mentioned advantages of the guided growth method. More specifically, since postgrowth manipulations and transfer of nanowires is required in these methods, the nanowires are more prone to fracture and contamination. 3D and in-plane buckled nanowires can be achieved by transferring them onto a prestrained elastomer and releasing the tensile strain.34,35,45−47 In principle, this method can be applied to any nanowire material but is limited to a specific “wavy” geometry. To truly expand the guided growth approach of nanowires, a high-throughput method that is not limited to a specific material and geometry is desired.

Herein we demonstrate the growth of semiconductor nanowires along open nanolithographic ridges and trenches on the amorphous thermal oxide layer of a silicon wafer by artificial epitaxy. We first demonstrate guided growth along straight lines, as a proof of concept that our method can indeed be applied to the growth of nanowires in open amorphous features. Nanowires of several material systems (GaN, ZnSe, CdS, ZnTe, and ZnO) are successfully grown within these ridges and trenches. These artificial guiding features were initially patterned by EBL (a serial process) for prototyping and then by nanoimprint lithography (NIL), demonstrating a fully parallel (i.e., high-throughput) process. We examine two different cross-sections for the guiding features: smooth curved trenches and a 90° profile ridges. We characterize the morphology of ZnSe and GaN nanowires growing in the two different templates by cutting thin, electron-transparent slices across the nanowires with a focused-ion beam (FIB) and observing them under a transmission electron microscope (TEM). We further characterize the quality and crystallographic orientation of the nanowires by using high-resolution TEM. Surprisingly, despite the lack of epitaxial relations, we find preferred crystallographic orientations for both ZnSe and GaN nanowires. Using NIL, we further introduce various curved and kinked guiding features for synthesizing planar nanowires of arbitrary shapes (spiral, zigzag, and sinusoidal shapes). Specifically, sine wave-shaped ZnSe and GaN nanowires are compared and show pronounced vapor–solid (VS) growth and no VS growth, respectively. Photoluminescence (PL) and cathodoluminescence (CL) mapping were performed as noninvasive characterization techniques and show shifts in the NBE emission along the nanowires. While in ZnSe these changes do not correlate with the sinusoidal geometry of the shaped nanowire and are likely due to VS growth, in GaN, a red-shift in PL emission is correlated with higher curvature areas, suggesting strain-induced reduction of the band gap. This combination of top-down and bottom-up approaches can be applied for the large-scale fabrication and study of predesigned shapes of nanowires from different materials.

Results and Discussion

Demonstration of Guided Growth along Lithographic Ridges and Trenches

In order to study the guided growth of nanowires on amorphous substrates by artificial epitaxy, we started by patterning nanometer-scale straight open lithographic guides on a Si wafer covered with a 300 nm oxide layer. The first attempts were done by EBL. After writing and developing the e-beam resist (PMMA 950 A3, see Methods), the lithographic guides are created by either wet-etch of the silica layer using a buffered oxide etch (BOE) or by depositing alumina using electron-beam evaporation and liftoff. These two methods yield either isotropic, curved trenches, or ridges with a 90° angle between surface and ridge walls, respectively (Figure 2a). (More details regarding the fabrication of lithographic guides can be found in the Methods section.) From this point on, the two different cross-section morphologies of the lithographic guides will be referred to as trenches and ridges. Dimensions of lithographic guides range from 10 to 20 nm in height and 80–160 nm in pitch, with various widths. We note that the guides are not designed to confine the horizontal nanowires but only to guide their growth along the artificial features. Figure 2b shows a scanning electron microscope (SEM) image of straight ridges created by EBL followed by alumina evaporation and liftoff. Both ridge and trench configurations (discussed below) show similar quality and uniformity and are both used for the growth of horizontal nanowires by artificial epitaxy.

Figure 2.

Figure 2

Open in a new tab

Guided growth of nanowires by artificial epitaxy in straight lithographic guides patterned by electron-beam lithography (EBL). (a) Schematic of the EBL fabrication process of lithographic guides for artificial epitaxy. Patterns are written by EBL in a positive-tone polymer; the exposed areas are washed away upon development. Two processes are employed to create the lithographic guides: (1) Isotropic grooves are wet-etched by BOE at the exposed areas to create trenches. (2) Stripes of amorphous alumina are placed within the exposed areas by electron-beam evaporation to create ridges. In both cases, after liftoff of unexposed polymer, catalyst patterning is performed, followed by CVD growth of nanowires. (b) Scanning electron micrograph of a ridge configuration before growth. (c) Straight ZnSe nanowires growing in a ridge configuration.

Microscale thin-film islands of metal catalyst are patterned perpendicular to the lithographic guides, using photolithography, electron-beam evaporation (nominal thickness 0.5 nm), and lift-off. Upon heating the sample at 550 °C, the metallic film breaks into nanodroplets in a dewetting process, which serve as the catalyst droplets for nanowire growth in the VLS mechanism. Nanowires of different materials are then grown by chemical vapor deposition (CVD), at similar conditions formerly found by our group for the epitaxial and graphoepitaxial growth of horizontal nanowires on sapphire.5,6,9,11,48 The general growth scheme can be found in the Supporting Information (Figure S1). In Figure 2c we present horizontal ZnSe nanowires guided along straight ridges prepared by EBL. Nanowire lengths and diameters were comparable to those grown by epitaxy and graphoepitaxy on sapphire,9 with lengths surpassing 20 μm. The catalyst droplet is apparent at the edge of the nanowires, indicating the expected VLS growth mechanism.

Parallel Process by Nanoimprint Lithography

Although EBL is a broadly used method for the fabrication of high-quality nanoscale features that can be used for the guided growth of horizontal nanowires by artificial epitaxy, it is a serial, low-throughput process. Therefore, once successful growth was achieved along the lithographic guides, EBL was replaced by thermal NIL. In NIL,49 the same mold is used to pattern a large number of samples, upgrading guided growth by artificial epitaxy into a parallel process. The mold itself is created by EBL, using hydrogen silsesquioxane (HSQ) as the electron-beam resist. HSQ is used to achieve high-resolution features due to its low line-edge roughness and low molecular weight.50 Upon development and thermal treatment, it hardens into porous silica and can withstand multiple uses as a hard mold. In thermal NIL, the hard mold is pressed at high pressure into a thermoplastic polymer (the imprint resist) at a temperature in the viscous phase and then rapidly cooled below the polymer’s glass transition temperature (Tg) before separation51 (technical details regarding the imprint process can be found in the Methods section). Using the same mold to pattern a large number of samples drastically improves the process throughput (Figure 3a). After molding the pattern into the imprint resist, different methods can be used to transfer the patterned features from the imprint resist to the silica layer. All methods include removal of any polymer remaining in the depressed areas by a gentle reactive ion etching (RIE) as one of the first steps (see Methods). We use two pattern-transfer methods based on either wet-etch or deposition of amorphous alumina, which are similar to the final steps in the EBL process. Therefore, the NIL process results in the two same final configurations of trenches and ridges. A schematic of the complete NIL process can be found in the Supporting Information including both pattern-transfer methods for the two configurations of trenches and ridges in Figures S2 and S3, respectively Straight trenches, with a 120 nm pitch, fabricated by NIL, are shown in Figure 3b. The inset shows a cross-section prepared by FIB of similar trenches with an 80 nm pitch. Nanowires of different materials are grown on both trench and ridge configurations. The quality and crystallinity of nanowires grown in the two configurations are discussed below. In Figure 3c, GaN nanowires grown along ridges prepared by NIL are depicted. As can be seen, the quality of lithographic guides produced by NIL (Figure 3b), realized mainly by their uniformity and low roughness, is no less than those produced by EBL (Figure 2b). More importantly, the high quality of the NIL guides is manifested in the high yield and alignment of nanowires growing along them (Figure 3c). As shown in Figure 3d, the nanowire diameter is not determined by the distance between the two ridges, but is rather affected by the size of the catalyst droplet, in accordance with the VLS mechanism. The nanowire is attached to one ridge, where the alumina wall, only 10 nm in height, guides and aligns the nanowire. This configuration of open lithographic guides allows a VLS growth with minimal confinement, where the nanowires are at least partially free to expose their most stable facets under the relevant growth conditions. This issue will be further discussed below.

Figure 3.

Figure 3

Open in a new tab

Guided growth of nanowires by artificial epitaxy in straight lithographic guides patterned by nanoimprint lithography. (a) Schematic of the NIL process. A hard mold with the desired pattern is pressed into a thermoplastic polymer coated on the target substrate and heated above the glass transition temperature (Tg). Pneumatic pressure is applied, followed by rapid cooling of the system and separation of the mold and the now plastic, patterned substrate. Throughput is greatly increased with the repeated use of a single master mold. (b) Scanning electron micrograph of a trench configuration before growth with 120 nm pitch. In the inset, a cross-section made by FIB milling of such trenches before growth with 80 nm pitch. Inset schematic describes the process of transferring the imprinted pattern to the substrate: first, any polymer remaining in the depressed areas is removed by a gentle RIE. Lithographic guides are prepared by the same two methods described in the EBL process. (c) Straight GaN nanowires growing in a ridge configuration. (d) High-magnification image of a single GaN nanowire attached to the ridge’s wall. The diameter of the nanowire, estimated by the SEM image, is 35 nm.

Morphology and Crystallinity

One important issue, arising from the fact that the substrate is amorphous and not a single crystal (as in previous cases of guided growth and SAG), is the crystallinity of the nanowires. In the past few years, we established that guided horizontal nanowires growing by epitaxy and graphoepitaxy on crystalline substrates not only grow as a single crystal but also show relatively low density of defects.5,6,22 Moreover, their high crystal quality is manifested in their optical and optoelectronic properties.9,13,48 A few works on horizontal nanowires grown on amorphous substrates demonstrated that the nanowires in this case also grow as single crystals.38,39 However, the question of preferred crystallographic orientation of these nanowires remains open. One of the main advantages of the guided growth of nanowires on crystalline substrates is the control over the crystallographic orientation of the nanowires. For example, ZnO nanowires guided on R-plane sapphire grow with extremely high yield in a polar orientation (where the [0001] direction aligns with the long axis of the nanowire).5,52 This is due to the strong epitaxial relations between the nanowire and the substrate. Since on amorphous substrates epitaxial relations are completely absent, we ask if the nanowires will exhibit a preferred crystallographic orientation or, with no epitaxy to constrain them, will exhibit completely random behavior. In addition, we wish to characterize the morphology of nanowires guided by amorphous guides, since it can reflect their quality and uniformity. In order to characterize both the morphology and crystallographic orientations of nanowires guided by artificial epitaxy, cross-sectional electron-transparent lamellae were cut across the nanowires by FIB and observed under a TEM.

We start by studying the effect of the amorphous templates on the morphology of the horizontal nanowires. More specifically, we are interested in studying the structure of nanowires, manifested in their shape, diameter, quality, and faceting. Clearly, a separate consideration is required for trench and ridge configurations. A low-magnification TEM image (Figure 4Aa) presents the cross-section of ZnSe nanowires guided by artificial epitaxy along ridges. As can be seen, the ZnSe nanowire grows directly on the amorphous silica surface and is guided by a 10 nm alumina wall. The open configuration of the lithographic guide does not confine the growth and allows the nanowire to accommodate a diameter larger than the dimension of the wall and to display well-defined facets. The wires are observed to grow as high-quality single crystals, and no structural defects are observed. More examples can be found in the Supporting Information (Figure S4). Additionally, a longitudinal lamella was cut and observed by TEM (Figure S5). The images, at both low magnification and atomic resolution including their fast Fourier transform (FFT), reveal high uniformity and crystal quality along the ZnSe nanowire.

Figure 4.

Figure 4

Open in a new tab

Morphology and crystallinity studied by cross-sectional TEM: (A) (a) Low-magnification TEM of a cross-section of a ZnSe nanowire guided along a ridge configuration. The nanowire grows on the amorphous silica layer, where only a small part is attached to the alumina wall. (b) Low-magnification TEM of a cross-section of a GaN nanowire guided in a trench configuration. The bottom part, which is in contact with the silica layer, is rounded, while the upper part exposes facets. (B) (a) Fringes of the ZnSe crystal structure observed by HRTEM. Inset: FFT of a ZnSe crystal, from which crystallographic orientation is determined. (b) Fringes of the GaN crystal structure observed by HRTEM. Inset: FFT of GaN crystal, from which crystallographic orientation is determined.

A second lamella was cut across GaN nanowires guided by artificial epitaxy in a trench configuration. Low-magnification image is presented in Figure 4Ab. As can be seen, the nanowire cross-section appears round at the interface with the substrate and faceted at the upper, exposed part. This cross-section resembles GaN nanowires grown horizontally on quartz, in which the nanowires grow embedded in the quartz substrate.20 The silica layer seems to be enveloping the GaN nanowires, distorting the cross-section of a typical trench profile in comparison to that observed before growth (inset of Figure 3b). In fact, we observe the reconstruction of the amorphous silica at the conditions of the synthesis, resulting in a half-closed channel around the lower part of the nanowire. In its upper part, which is not in contact with the surface, the nanowire is faceted (more examples are available in Figure S6). Although the nanowires grow as single crystals (see below), we observe a relatively high concentration of plane defects in comparison to GaN nanowires grown by epitaxy and graphoepitaxy modes (Figure S6). We suggest that the reconstruction of the silica during the growth of the nanowire leads to a more constrained growth and results in a higher density of plane defects. This observation suggests that the trench configuration is less preferable for the guided growth of materials with relatively high CVD temperature (950 °C for the guided growth of GaN nanowires). In general this observation supports the importance of an open configuration of the lithographic guides for the growth of nanowires with high crystal quality, in which the trench only guides but does not confine the growth. Next, we proceed to the crystallographic characterization of the nanowires under consideration.

Crystallographic Orientation

In order to characterize the crystallographic orientations of nanowires grown by artificial epitaxy, we use high-resolution TEM (HRTEM). Higher magnification images display clear fringes (Figure 4Ba and Bb for ZnSe and GaN, respectively) and enable the determination of the crystallographic orientations by using FFT. The FFT peaks are identified with known crystallographic data and fitted to atomic models of ZnSe and GaN. From HRTEM and crystallographic analysis, we find ZnSe nanowires to be high-quality single crystals. No evidence for structural defects is observed. We attribute this observation to the growth along spaced ridges, allowing the crystal to accommodate its stable structure and expose its stable facets under the relevant growth conditions. More interestingly and to our surprise, we find a well-defined crystallographic orientation. Despite the fact that these nanowires are guided by amorphous features and no epitaxial relations with the substrate play a role in this guiding mechanism, we find a preferred growth orientation. We were able to determine the crystallographic orientation of seven nanowires, five of which grew along the [1213] direction in a wurtzite (WZ) structure (see table in Figure S4). This specific direction was not observed in epitaxial ZnSe nanowires, guided by either epitaxy or graphoepitaxy on different planes of sapphire.9,53

The same methodology is used for the crystallographic analysis of GaN nanowires. We find a relatively large variety of crystallographic orientations (see table in Figure S6). Nevertheless, among the different orientations, we find [1210] to be the most common growth direction. We note that even nanowires with the same growth direction can be found rotated in different angles with respect to the growth axis. These observations all resemble our past findings regarding the guided growth of GaN nanowires on quartz.20 This similarity is explained by the observed amorphization of the quartz at the interface with the GaN nanowires, which in fact obscures the epitaxial relations, very much like the horizontal growth of nanowires on the amorphous oxide layer of the silicon wafer.

The nonpolar orientations, [1213] and [1100], also known as the a and m directions, respectively, are the most frequently observed crystallographic orientation for GaN nanowires.54−62 Kuykendall etal. demonstrated that the catalyst composition in VLS growth plays a central role in directing the crystallographic orientation of GaN nanowires.58 More specifically, they showed that Ni-rich catalyst leads to growth along the a-axis ([1210]) on two different substrates. In analogy, Ni is used as the catalyst for the growth of GaN nanowires in this work, which are indeed found to grow most commonly along the a direction. It is suggested that the catalyst composition either stabilizes the (1210) planes at the growth front (energetically controlled growth) or changes the barrier between the catalyst and the (1210) planes of GaN (kinetically controlled growth). In principle, a third, noncatalytic possible effect can be considered: minute amounts of catalyst can migrate to the sides of the wire and stabilize specific facets, thus dictating the growth direction. Since the nanowires in this work present a round cross-section at the interface with the silica and only a small part is faceted, we consider the last suggestion to be less likely in this case. Based on the absence of epitaxial relations and former observations regarding the preferred growth directions in vertically grown GaN nanowires, the catalyst seems to play a major role in guiding the growth axis of these artificial-epitaxy-guided nanowires.

Generality for Different Materials

To further test the generality of guided growth by artificial epitaxy, in addition to ZnSe and GaN, we grew CdS, ZnTe, and ZnO nanowires along lithographic guides, as depicted in Figure 5Aa, Ab, and Ac, respectively. Nanowires from the different materials were grown using either EBL or NIL with both trench and ridge configurations. As can be seen, all nanowires demonstrate guided growth along the patterned guides, displaying the adaptability and generality of the method. However, nanowire growth from different materials differs in both yield and nanowire morphology. The observed differences between all five materials are also observed in the guided growth by epitaxy and graphoepitaxy on sapphire5,6,9,11,48 and are discussed below. SEM images reveal that after dewetting the density of catalyst droplets is comparable to the density of lithographic guides. Therefore, every guide could potentially support the growth of a nanowire (Figure S7). However, the density of nanowires is clearly lower than that of lithographic trenches, leading to the conclusion that the yield of nanowire growth is affected by the parameters of the CVD process and is not limited by the catalyst density. This issue is material-specific, and synthesis parameters could be adjusted to increase the yield for each material. In general, the CVD process on a patterned silicon substrate is very similar to that on sapphire, and in most cases guided growth by epitaxy, graphoepitaxy, and artificial epitaxy can be achieved under the exact same conditions in the same synthesis. In some cases, adjustment of the sample temperature is required to improve the yield of nanowire growth by artificial epitaxy, mainly due to the different thermal conductivity of silicon and sapphire.

Figure 5.

Figure 5

Open in a new tab

Guided growth of arbitrarily shaped horizontal nanowires of different materials by artificial epitaxy: (A) Guided growth of straight (a) CdS, (b) ZnTe, and (c) ZnO. Lithographic guides were made by EBL in a ridge configuration. (B) Guided growth of GaN in (a) sine wave and (b) right angle kinked (zigzag) shapes. Guided growth of ZnSe in (c) spiral, (d) zigzag, and (e) sine wave shapes. Lithographic guides were made by NIL in a trench configuration.

Guided chalcogenide nanowires often show, in addition to the VLS mechanism that leads to the elongation of the nanowires, a significant extent of VS growth, in which material from the gas phase is directly deposited on the growing nanowire. The VS/VLS growth rate can be controlled to some level by adjusting the parameters of the synthesis. A prominent VS growth can be manifested as nanowalls or very “bulky” nanowires, as was observed in guided chalcogenide nanowires on sapphire.11,48 The VS growth is present in chalcogenide nanowires grown by artificial epitaxy as well. This phenomenon is most pronounced in the growth of ZnTe in our working conditions (Figure 5Ab), but even very thick bulky nanowires are well aligned by the patterned guides. We note that the dimensions of the patterned guides were optimized to the growth of GaN nanowires, which do not show any VS growth. More specifically, the pitch of the array, together with the width and height of the guides, was adjusted to fit the observed yield and typical diameters of GaN nanowires on sapphire, respectively.6 The growth yield of chalcogenide nanowires with typically larger dimensions could be improved by adjusting the spacing, width, and height of the guides.

Similarly to all other materials, the CVD growth of ZnO by artificial epitaxy resolved in elongated nanostructures along the guides, with a droplet at the edge that indicates a VLS growth, as can be seen in Figure 5Ac. However, when performing AFM, we observed that where a nanowire seemed to grow from the SEM top view, there was instead a trench wall that was a few nanometers higher than expected. To better understand the structure, we cut a thin electron-transparent lamella using FIB and examined it by TEM. We found that instead of ZnO nanowires, a thin layer of ZnO was covering the wall of the trench, creating some sort of cupping (Figure S8). By adjusting the synthesis parameters, we did succeed in growing ZnO nanowires inside the trenches, but they were short (∼1 micron) (further information can be found in the Supporting Information). This is not completely surprising. Although the yield and alignment of ZnO nanowires on sapphire guided by epitaxy are extremely high,5,52 from our experience the guidance by graphoepitaxy is usually more challenging. Since guided growth by artificial epitaxy is only topographic, it is expected to be less appropriate for the guidance of long ZnO nanowires.

Arbitrary Shapes

Growing nanowires on patterned amorphous substrates offers the possibility of growing them in arbitrary shapes and expand the guided growth approach beyond the realization of straight nanowires. To test the possibility of growing arbitrarily shaped nanowires, we patterned zigzag, sinusoidal, and spiral features by NIL in both trench and ridge configurations. Mold patterning in HSQ with arbitrary shapes is presented in Figure S9. The growth of GaN and ZnSe within the shaped guides is presented in Figure 5B. As shown in Figure 5Ba, GaN nanowires follow the sine wave-shaped guides in a trench configuration. The nanowires do not show any VS growth, and the diameter stays constant along the wire. In a few cases the nanowires manage to jump between adjacent trenches, indicating the need for higher walls. While the height of the walls proved to be fit for the growth of aligned straight nanowires, it seems that the trenches are a bit too shallow for the growth of nonstraight nanowires. Zigzag-shaped GaN nanowires grown in a ridge configuration are presented in Figure 5Bb. The nanowires nicely follow the sharp 90° turns with a yield that is comparable to that of straight nanowires guided by artificial epitaxy. In general, the GaN nanowires seem to better follow the shaped guides in the ridge configuration in comparison to the trench one. This is the case even when the nanowires are forced to take sharper turns during growth. We suggest that the sharper profile of the ridge is preferred for the growth of nanowires in arbitrary shapes, which grow attached to the ridge walls. GaN nanowires growing in the curved smooth profile of the trenches are more prone to escape from them.

Figure 5Bc, Bd, and Be show the growth of ZnSe along spiral, zigzag, and a sine wave shapes, respectively. The shaped nanowires all show a high level of tapering due to pronounced VS growth. According to the VLS mechanism, the tip of the nanowire, close to the catalyst droplet, is the most recently crystallized segment. This segment has the same dimensions of the patterned trenches. During the synthesis, material from the vapor phase nucleates on top of the nanowire, leading to a thicker and bulkier morphology toward its other end. This logic suggests that the nanowire follows the shaped groove while it is still in the appropriate dimensions, and VS growth occurs on top of the shaped wire. Although VS growth is sometimes apparent in epitaxial- and graphoepitaxial -guided ZnSe nanowires,9,53 it is much more pronounced in the shaped nanowires. We suggest that the growth in a curved or kinked configuration induces a higher density of defects in and on the surface of the nanowires, and thus results in more nucleation sites for VS growth. This is significant for the growth of shaped chalcogenide nanowires, which tend to suffer from VS growth, as manifested by thicker and chunkier morphology.

After growing sine-wave-shaped GaN (Figure 5Ba) and ZnSe (Figure 5Be) nanowires, we proceeded to characterize their optical properties. As mentioned above, wavy configurations of nanowires can be achieved also in postgrowth methods by transferring vertically grown nanowires onto a prestrained elastomer and releasing the tensile strain.34,35 This approach results in a periodic modulation in the NBE energy, where the buckled segments show a red-shifted PL emission with respect to the straight segments. This effect is attributed to a strain-induced band gap shrinking. In general, bent nanowires (and other nanostructures) show a red-shift in their PL33 and CL31,32 emission. However, the nanowires in this work were not bent after growth but rather formed along a curved feature. The optical properties of nanowires grown by this mode, especially the impact it may have on their NBE emission, has yet to be studied. In order to characterize the sine wave-shaped GaN and ZnSe nanowires guided by artificial epitaxy, we map the PL emission along the curved nanowire. We use a micro PL system (Horiba LabRAM HR Evolution) with a He–Cd laser (325 nm) for excitation. Mapping is done by scanning a predetermined area with a motorized stage where at each position a full spectrum is acquired. The NBE peak of the spectrum at each location can then be fitted to a Gaussian, and its height (above the baseline) and center, which corresponds to the emission intensity and wavelength, can be plotted in a color-scale to produce an intensity and a spectral map, respectively. This enables us to follow the NBE emission along the shaped structure.

Figure 6Aa shows an SEM image of a sine wave ZnSe nanowire guided by artificial epitaxy. Significant VS growth is observed, typical of the shaped chalcogenide nanowires, as mentioned above. Figure 6Ab and Ac show the complementary PL intensity map and spectral map, respectively. Variation of 8 nm in the band edge emission is observed in the range of 459.5–467.5 nm, where the maximum error of the fitting is 0.6 nm. However, no correlation between the NBE energy and the periodic shape of the nanowire is found. We attribute the relatively wide range of NBE energies to the significant extent of VS growth. To follow the NBE with a better spatial resolution, we further perform CL on an additional sine wave-shaped ZnSe nanowire (Figure 6Ad). Each point along the nanowire represents the location in which a spectrum was taken. The color represents the wavelength extracted from the center of the Gaussian fitted to the NBE peak, as described above for PL. A representative CL spectrum is shown in the Supporting Information (Figure S10). Similar to PL, a variation in the NBE energy from different points along the nanowire is observed, but with no correlation to the periodic curved geometry of the nanowire. CL spectra also reveal defect emissions at wavelengths larger than 480 nm (Figure S10). This observation is in good agreement with the large extent of VS growth observed by SEM imaging.

Figure 6.

Figure 6

Open in a new tab

Optical properties of shaped nanowires. (A) Photoluminescence and cathodoluminescence mapping of a sine wave-shaped ZnSe nanowire: (a) SEM micrograph of a ZnSe nanowire grown in a sinusoidal open trench prepared by NIL; (b) PL intensity map and (c) PL spectral map of the nanowire in (a); the color change corresponds to variations in wavelength. (d) CL spectral data taken at specific points along the wire. The color corresponds to the wavelength extracted from the center of the NBE peak. (B) Photoluminescence mapping of a sine wave-shaped GaN nanowire: (a) SEM micrograph of a GaN nanowire grown in a sinusoidal open trench prepared by NIL. (b) PL intensity map and (c) PL spectral map of the nanowire in (a); the color change corresponds to variations in wavelength.

Unlike the sine wave-shaped ZnSe nanowires, sine wave-shaped GaN nanowires are much more uniform and exhibit no VS growth, as presented in Figure 6Ba. Figure 6Bb presents the PL intensity map of the area marked in a white rectangle in Figure 6Ba. No significant variation in intensity is observed along the nanowire itself, indicating the uniformity of the structure and diameter of the GaN nanowire that grows in a pure VLS mechanism. In Figure 6Bc we plot the PL spectral map of the NBE emission. In order to exclude signal from adjacent nanowires, we focus on the half-sine wave at the far edge of the nanowire. A clear red-shift of the NBE emission is observed with correlation to the curved geometry of the nanowire. The wavelengths of NBE emission in the curved regions differ from the wavelengths in the straight regions by 4 nm (358 and 362 nm), where the maximum error of the Gaussian fit is 0.2 nm. This red-shift indicates strain-induced band gap reduction, as found in postgrowth bent nanowires. Nevertheless, we note that no yellow luminescence, which is correlated with defects in GaN, is observed either in straight or in curved regions, as was found in GaN nanowires guided by epitaxy and graphoepitaxy on sapphire.6 In other words, except for the strain-related red-shift due to the curved template, the growth by artificial epitaxy on an amorphous substrate does not affect the optical properties of guided GaN nanowires in comparison to epitaxy and graphoepitaxy. The full spectrum can be found in the Supporting Information (Figure S11).

Conclusions

In summary, artificial epitaxy was used to produce in-plane nanowires of five different materials guided along straight and arbitrarily shaped ridges and trenches. Using the high-throughput process of NIL, nanolithographic guides were fabricated with comparable yield and quality to those fabricated by EBL. Two different configurations of lithographic guides were examined and were found to be useful for the guided growth of nanowires by artificial epitaxy. Although the substrate is amorphous, straight nanowires were found to grow as single crystals along preferred crystallographic orientations. We suggest that the catalyst plays an important role in promoting specific orientations in the absence of any epitaxial relations between the nanowire and the substrate. PL and CL were used as noninvasive tools to characterize nanowires with arbitrary shapes containing curved regions. With the absence of VS growth, a red-shift in the NBE emission was observed in higher curvature segments of GaN, suggesting a strain-related band gap decrease, as in the case of postgrowth bending of nanowires. Guided growth by artificial epitaxy proves to be general and adaptable to the growth of different materials and in principle can be used for the growth of any other material. The precise dimensions of the trenches as well as the growth parameters can be optimized for each material to improve the yield and morphology of the nanowires. This combination of top-down and bottom-up approaches expands the guided growth approach and offers the possibility of controlling and manipulating semiconductor nanosystems for optoelectronic and quantum computing, compatible with silicon-based CMOS and MEMS technologies.

Methods

EBL Patterning

Poly(methyl methacrylate) (PMMA 950 A3) was spin-coated (5000 rpm) on a Si/SiO2 (300 nm, thermal oxide) wafer (SVM) and baked on a hot plate for 2 min at 180 °C. Different designs were patterned using a Raith e-Line Plus electron-beam lithography system. Cold developing was done in standard MIBK/IPA (1:3) at 5 °C for 40 s. After development two methods are used to create either trenches or ridges:

(1) A trench configuration is created by wet-etch: BOE (6:1 with surfactant, J.T. Baker) was diluted in a 5:1 ratio in distilled water. First, a sacrificial sample of Si/SiO2 is immersed for 1 min to calibrate the exact etching rate. Using an ellipsometer (Rudolph Auto EL) the thickness of the thermal oxide layer is measured before and after etching to calculate the etching rate. After EBL and development, the sample is etched for the time required to remove 20 nm of the oxide layer, according to the calculated rate, to form 20 nm isotropic trenches in the SiO2 layer.

(2) A ridge configuration is created by evaporation of amorphous alumina: Evaporation of alumina was done in an electron-beam evaporation chamber (PVD, Telemark) at a rate of 1 Å/s.

The two specified methods are followed by liftoff in acetone.

Nanoimprint Lithography

Mold Writing

Hard molds were prepared from 325 μm thick Si wafers (SVM) cut into 1.6 cm2 squares by etching the native oxide in a buffered oxide etch of hydrofluoric acid (BOE 6:1 with surfactant, J. T. Baker), spin-coating the wafers with a 35 nm thickness of 2% HSQ e-beam resist (XR-1451, Dow Corning). The designs were patterned using a Raith e-line Plus electron-beam lithography system at a 30 kV accelerating voltage, developing the mold in AZ 726 (Clariant GmBH) for 60 s followed by a 30 s water rinse. The design elements are one-dimensional pixel-lines exposed with doses ranging from 1800 to 3000 pC/cm depending on the pattern and the line pitch, which ranges from 120 to 180 nm. The dosage value directly determines the line thickness, which ranged from 20 to 40 nm. Arrays of straight lines, zigzags (with acute, right (90°), and obtuse angles), sine waves, and spirals were designed and patterned in this fashion. Each array was 30 × 70 or 30 × 100 μm in size. Each mold had four duplicate areas of 100 arrays. Plasma ashing (1 min, 1 sccm O2, 150 W) and thermal annealing (60 min, 600 sccm Ar, 900 °C) hardened and developed HSQ into porous silica. Finally, the fabricated mold was covered with a mold release agent (Nanonex NXT-100). SEM images of arbitrary shapes patterned in HSQ can be found in the Supporting Information (Figure S9).

Imprint

PMMA resist designed specifically for NIL, PMMA 35K (EM Resist Ltd., ∼13% PMMA in anisole), was diluted with anisole (anhydrous, 99.7%, Sigma-Aldrich) at different ratios to obtain mixtures that could produce thicknesses ranging from 40 to 105 nm. An ellipsometer (Rudolph Auto EL) and an optical profiler (Zeta-20) were both used to measure resist thickness. The resist was spin-coated onto 18 × 18 mm2 squares of 500 μm silicon with a 300 nm thermal oxide layer (SVM) and baked on a hot plate at 180 °C for 2 min.

A homemade thermal imprinting system (Figure S12) was used to perform imprinting. The sample was placed on the imprint chuck. The mold was placed face-down on top of the sample. A 6 × 6 cm2 double layer of silica elastomer was placed on top of the mold, and a vacuum pump was used in order to ensure a good seal. The chamber was lowered onto this sandwich and then heated to 200 °C, which is above the Tg of PMMA (105 °C). High pressure (17 bar N2) was then introduced to the system for a period of 5 min. The system was then rapidly cooled using a flow of chilled water. After cooling to 40 °C, the pressure was vented. Upon removal from the chuck the mold and sample were separated using a razor blade.

Pattern transfer was accomplished using either wet etch to create trenches or alumina evaporation to create ridges, as described below. The two methods were used on samples with ∼40 and ∼70 nm PMMA thickness, respectively.

Pattern Transfer after NIL

Two pattern-transfer processes were employed to create trenches and ridges: (1) For trenches, the residual resist layer was first removed by plasma etching (STS ASE ICP, 30 mTorr, 30 sccm O2, no coil, 20 W platen power), and the exposed silicon dioxide was etched in BOE to form 20 nm isotropic trenches. (2) For ridges, we first deposited a metal mask of 15 nm Ti, using electron beam evaporation onto the imprinted samples tilted by 30°.63,64 Then, residual resist was etched through the deposited mask using oxygen plasma (STS ASE ICP, 30 mTorr, 30 sccm O2, no coil, 20 W platen power, followed by electron-beam evaporation of 10 nm amorphous alumina. The last stage in both pattern transfer methods is liftoff in acetone. Schematics of the two pattern-transfer methods can be found in the Supporting Information (Figures S2 and S3).

Catalyst Patterning and Nanowire Growth

Catalyst islands were patterned perpendicularly to the guiding features by photolithography, using negative-tone resist NR-9 1000PY (developed with RD-6) and a mask aligner (MA/BA6 Karl Suss), followed by electron-beam evaporation (PVD, Telemark) of 5 Å Ni catalyst (for the growth of GaN) or Au (for the growth of all other materials). Dewetting of the catalyst was performed at 550 °C. Growth was performed in a CVD system according to protocols published by our group for various materials. The growth scheme can be found in the Supporting Information (Figure S1).

Structural Characterization

Imaging of nanowires was done by scanning electron microscope (Supra 55VP FEG LEO Zeiss). For the characterization of morphology and crystallinity, a focused ion beam (FEI Helios 600 dual beam microscope) was used to cut thin (50–100 nm) lamellae across the nanowires, which were later inspected under a high-resolution transmission electron microscope (FEI Tecnai F30). To study the crystallographic orientations, HRTEM images were analyzed using FFT from selected areas, and the FFT peaks were fitted to the crystallographic tables of bulk ZnSe and GaN.

Photoluminescence

PL measurements were done using a micro-Raman/micro-PL system (Horiba LabRAM HR Evolution). A He–Cd 325 nm laser was focused on the nanowire through a reflective objective lens, and PL was collected using the same objective and sent to a 300 lines/mm grating and a front-illuminated cooled open-electrode CCD camera.

Cathodoluminescence

CL measurements were done with SEM cathodoluminescence using a Gatan MonoCL Elite system that is installed in a Zeiss GeminiSEM 500, high-resolution SEM. CL signal was excited by a 5 kV electron beam and collected by a diamond-turned mirror. The spectra were acquired by directing the light to a monochromator and a CCD for parallel spectroscopy either as spectral map or point by point on specific locations. The spectral resolution was set to 2.7 nm by choosing a slit size of 1 mm and grating of 1200 lines/mm. SEM and CL images were recorded simultaneously.

Acknowledgments

This research was supported by the Israel Science Foundation 1493/10, Minerva Stiftung (Project No. 711138), European Research Council (ERC) Advanced Grant (No. 338849), ERC Proof of Concept (PoC) Grant No. 838702, Helena and Martin Kimmel Center for Nanoscale Science, Moskowitz Center for Nano and Bio-Nano Imaging, and the Carolito Stiftung. E.J. holds the Drake Family Professorial Chair of Nanotechnology. R.B.Z. acknowledges funding from the Clore Foundation.

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsnano.9b00538.

  • Additional information (PDF)

Author Present Address

‡ Present address of M. Schvartzman: Department of Materials Engineering and Ilse Katz Institute for Nanoscale Science and Technology, Ben Gurion University of the Negev, Beer Sheva, 8410501, Israel.

The authors declare no competing financial interest.

Supplementary Material

nn9b00538_si_001.pdf (2.3MB, pdf)

References

  1. Fortuna S. A.; Wen J.; Chun I. S.; Li X. Planar GaAs Nanowires on GaAs (100) Substrates: Self-Aligned, Nearly Twin-Defect Free, and Transfer-Printable. Nano Lett. 2008, 8, 4421–4427. 10.1021/nl802331m. [DOI] [PubMed] [Google Scholar]
  2. Hsin C.; He J.; Lee C.; Wu W.; Yeh P.; Chen L.; Wang Z. Lateral Self-Aligned p-Type In2O3 Nanowire Arrays Epitaxially Grown on Si Substrates. Nano Lett. 2007, 7, 1799–1803. 10.1021/nl0707914. [DOI] [PubMed] [Google Scholar]
  3. Nikoobakht B.; Michaels C. A.; Stranick S. J.; Vaudin M. D. Horizontal Growth and in situ Assembly of Oriented Zinc Oxide Nanowires. Appl. Phys. Lett. 2004, 85, 3244–3246. 10.1063/1.1803951. [DOI] [Google Scholar]
  4. Shen Y.; Turner S.; Yang P.; Van Tendeloo G.; Lebedev O. I.; Wu T. Epitaxy-Enabled Vapor-Liquid-Solid Growth of Tin-Doped Indium Oxide Nanowires with Controlled Orientations. Nano Lett. 2014, 14, 4342–4351. 10.1021/nl501163n. [DOI] [PubMed] [Google Scholar]
  5. Tsivion D.; Schvartzman M.; Popovitz-Biro R.; Joselevich E. Guided Growth of Horizontal ZnO Nanowires with Controlled Orientations on Flat and Faceted Sapphire Surfaces. ACS Nano 2012, 6, 6433–6445. 10.1021/nn3020695. [DOI] [PubMed] [Google Scholar]
  6. Tsivion D.; Schvartzman M.; Popovitz-Biro R.; von Huth P.; Joselevich E. Guided Growth of Millimeter-Long Horizontal Nanowires with Controlled Orientations. Science 2011, 333, 1003–1007. 10.1126/science.1208455. [DOI] [PubMed] [Google Scholar]
  7. Zi Y.; Jung K.; Zakharov D.; Yang C. Understanding Self-Aligned Planar Growth of InAs Nanowires. Nano Lett. 2013, 13, 2786–2791. 10.1021/nl4010332. [DOI] [PubMed] [Google Scholar]
  8. Rothman A.; Forsht T.; Danieli Y.; Popovitz-Biro R.; Rechav K.; Houben L.; Joselevich E. Guided Growth of Horizontal ZnS Nanowires on Flat and Faceted Sapphire Surfaces. J. Phys. Chem. C 2018, 122, 12413–12420. 10.1021/acs.jpcc.8b04063. [DOI] [Google Scholar]
  9. Oksenberg E.; Popovitz-Biro R.; Rechav K.; Joselevich E. Guided Growth of Horizontal ZnSe Nanowires and Their Integration into High-Performance Blue-UV Photodetectors. Adv. Mater. 2015, 27, 3999–4005. 10.1002/adma.201500736. [DOI] [PubMed] [Google Scholar]
  10. Oksenberg E.; Sanders E.; Popovitz-Biro R.; Houben L.; Joselevich E. Surface-Guided CsPbBr3 Perovskite Nanowires on Flat and Faceted Sapphire with Size-Dependent Photoluminescence and Fast Photoconductive Response. Nano Lett. 2018, 18, 424–433. 10.1021/acs.nanolett.7b04310. [DOI] [PubMed] [Google Scholar]
  11. Reut G.; Oksenberg E.; Popovitz-Biro R.; Rechav K.; Joselevich E. Guided Growth of Horizontal p-Type ZnTe Nanowires. J. Phys. Chem. C 2016, 120, 17087–17100. 10.1021/acs.jpcc.6b05191. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Schvartzman M.; Tsivion D.; Mahalu D.; Raslin O.; Joselevich E. Self-Integration of Nanowires into Circuits Via Guided Growth. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 15195–200. 10.1073/pnas.1306426110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Shalev E.; Oksenberg E.; Rechav K.; Popovitz-Biro R.; Joselevich E. Guided CdSe Nanowires Parallelly Integrated into Fast Visible-Range Photodetectors. ACS Nano 2017, 11, 213–220. 10.1021/acsnano.6b04469. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Wang Q.; Li J.; Lei Y.; Wen Y.; Wang Z.; Zhan X.; Wang F.; Wang F.; Huang Y.; Xu K. Oriented Growth of Pb1-XSnxTe Nanowire Arrays for Integration of Flexible Infrared Detectors. Adv. Mater. 2016, 28, 3596–3601. 10.1002/adma.201506338. [DOI] [PubMed] [Google Scholar]
  15. Wang X.; Aroonyadet N.; Zhang Y.; Mecklenburg M.; Fang X.; Chen H.; Goo E.; Zhou C. Aligned Epitaxial SnO2 Nanowires on Sapphire: Growth and Device Applications. Nano Lett. 2014, 14, 3014–3022. 10.1021/nl404289z. [DOI] [PubMed] [Google Scholar]
  16. Xu J. Y.; Rechav K.; Popovitz-Biro R.; Nevo I.; Feldman Y.; Joselevich E. High-Gain 200 ns Photodetectors from Self-Aligned CdS-CdSe Core-Shell Nanowalls. Adv. Mater. 2018, 30, 1800413. 10.1002/adma.201800413. [DOI] [PubMed] [Google Scholar]
  17. Wagner R.; Ellis W. Vapor-Liquid-Solid Mechanism of Single Crystal Growth. Appl. Phys. Lett. 1964, 4, 89–90. 10.1063/1.1753975. [DOI] [Google Scholar]
  18. Cheng C.; Guo H.; Amini A.; Liu K.; Fu D.; Zou J.; Song H. Self-Assembly and Horizontal Orientation Growth of VO2 Nanowires. Sci. Rep. 2015, 4, 5456. 10.1038/srep05456. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Du W.; Yang X.; Pan H.; Ji X.; Ji H.; Luo S.; Zhang X.; Wang Z.; Yang T. Controlled-Direction Growth of Planar InAsSb Nanowires on Si Substrates without Foreign Catalysts. Nano Lett. 2016, 16, 877–882. 10.1021/acs.nanolett.5b03587. [DOI] [PubMed] [Google Scholar]
  20. Goren-Ruck L.; Tsivion D.; Schvartzman M.; Popovitz-Biro R.; Joselevich E. Guided Growth of Horizontal GaN Nanowires on Quartz and Their Transfer to Other Substrates. ACS Nano 2014, 8, 2838–2847. 10.1021/nn4066523. [DOI] [PubMed] [Google Scholar]
  21. Nikoobakht B.; Herzing A. Where Is the Required Lattice Match in Horizontal Growth of Nanowires?. Nanoscale 2014, 6, 12814–12821. 10.1039/C4NR04537H. [DOI] [PubMed] [Google Scholar]
  22. Tsivion D.; Joselevich E. Guided Growth of Epitaxially Coherent GaN Nanowires on SiC. Nano Lett. 2013, 13, 5491–5496. 10.1021/nl4030769. [DOI] [PubMed] [Google Scholar]
  23. Tsivion D.; Joselevich E. Guided Growth of Horizontal GaN Nanowires on Spinel with Orientation-Controlled Morphologies. J. Phys. Chem. C 2014, 118, 19158–19164. 10.1021/jp504785v. [DOI] [Google Scholar]
  24. Aseev P.; Fursina A.; Boekhout F.; Krizek F.; Sestoft J. E.; Borsoi F.; Heedt S.; Wang G.; Binci L.; Martí-Sánchez S.; Swoboda T.; Koops R.; Uccelli E.; Arbiol J.; Krogstrup P.; Kouwenhoven L. P.; Caroff P. Selectivity Map for Molecular Beam Epitaxy of Advanced Iii-V Quantum Nanowire Networks. Nano Lett. 2019, 19, 218–227. 10.1021/acs.nanolett.8b03733. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lee J. S.; Choi S.; Pendharkar M.; Pennachio D. J.; Markman B.; Rodwell M. J.; Krogstrup P.; Palmstrøm C. J.. Selective-Area Chemical Beam Epitaxy of in-Plane InAs One-Dimensional Channels Grown on InP (001) and InP (111) B Surfaces. ArXiv:1808.04563v1 Cond-Mater., 2018. [Google Scholar]
  26. Friedl M.; Cerveny K.; Weigele P.; Tütüncüoglu G.; Martí-Sánchez S.; Huang C.; Patlatiuk T.; Potts H.; Sun Z.; Hill M. O.; Güniat L.; Kim W.; Zamani M.; Dubrovskii V. G.; Arbiol J.; Lauhon L. J.; Zumbühl D. M.; Fontcuberta i Morral A. Template-Assisted Scalable Nanowire Networks. Nano Lett. 2018, 18, 2666–2671. 10.1021/acs.nanolett.8b00554. [DOI] [PubMed] [Google Scholar]
  27. Okamoto A.; Ohata K. Selective Epitaxial Growth of InAs on GaAs by Molecular Beam Epitaxy. Jpn. J. Appl. Phys. 1987, 26, L1174. 10.1143/JJAP.26.L1174. [DOI] [Google Scholar]
  28. Okamoto A.; Ohata K. Selective Epitaxial Growth of Gallium Arsenide by Molecular Beam Epitaxy. Appl. Phys. Lett. 1987, 51, 1512–1514. 10.1063/1.98619. [DOI] [Google Scholar]
  29. Cho A. Recent Developments in Molecular Beam Epitaxy (MBE). J. Vac. Sci. Technol. 1979, 16, 275–284. 10.1116/1.569926. [DOI] [Google Scholar]
  30. Schmid H.; Borg M.; Moselund K.; Gignac L.; Breslin C.; Bruley J.; Cutaia D.; Riel H. Template-Assisted Selective Epitaxy of Iii-V Nanoscale Devices for Co-Planar Heterogeneous Integration with Si. Appl. Phys. Lett. 2015, 106, 233101. 10.1063/1.4921962. [DOI] [Google Scholar]
  31. Fu Q.; Zhang Z. Y.; Kou L.; Wu P.; Han X.; Zhu X.; Gao J.; Xu J.; Zhao Q.; Guo W. Linear Strain-Gradient Effect on the Energy Bandgap in Bent CdS Nanowires. Nano Res. 2011, 4, 308–314. 10.1007/s12274-010-0085-6. [DOI] [Google Scholar]
  32. Han X.; Kou L.; Lang X.; Xia J.; Wang N.; Qin R.; Lu J.; Xu J.; Liao Z.; Zhang X. Electronic and Mechanical Coupling in Bent ZnO Nanowires. Adv. Mater. 2009, 21, 4937–4941. 10.1002/adma.200900956. [DOI] [PubMed] [Google Scholar]
  33. Kim Y.; Im H. S.; Park K.; Kim J.; Ahn J. P.; Yoo S. J.; Kim J. G.; Park J. Bent Polytypic ZnSe and CdSe Nanowires Probed by Photoluminescence. Small 2017, 13, 1603695. 10.1002/smll.201603695. [DOI] [PubMed] [Google Scholar]
  34. Sun F.; Sun L.; Zhang B.; Chen G.; Wang H.; Shen X.; Lu W. The Optical Waveguide of Buckled CdS Nanowires Modulated by Strain-Engineering. ACS Photonics 2018, 5, 746–751. 10.1021/acsphotonics.7b01348. [DOI] [Google Scholar]
  35. Sun L.; Kim D. H.; Oh K. H.; Agarwal R. Strain-Induced Large Exciton Energy Shifts in Buckled CdS Nanowires. Nano Lett. 2013, 13, 3836–3842. 10.1021/nl401860f. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Givargizov E. Mechanisms of Oriented Crystallization in Artificial Epitaxy (Graphoepitaxy). Thin Solid Films 1990, 189, 389–396. 10.1016/0040-6090(90)90468-S. [DOI] [Google Scholar]
  37. Bronshtain T.Guided Growth of Nanowires: Mechanism and Surface Pertubation Effects; Weizmann Institute of Science: Rehovot, 2012. [Google Scholar]
  38. Pevzner A.; Engel Y.; Elnathan R.; Tsukernik A.; Barkay Z.; Patolsky F. Confinement-Guided Shaping of Semiconductor Nanowires and Nanoribbons:“Writing with Nanowires. Nano Lett. 2012, 12, 7–12. 10.1021/nl201527h. [DOI] [PubMed] [Google Scholar]
  39. Xue Z.; Sun M.; Dong T.; Tang Z.; Zhao Y.; Wang J.; Wei X.; Yu L.; Chen Q.; Xu J. Deterministic Line-Shape Programming of Silicon Nanowires for Extremely Stretchable Springs and Electronics. Nano Lett. 2017, 17, 7638–7646. 10.1021/acs.nanolett.7b03658. [DOI] [PubMed] [Google Scholar]
  40. Yu L.; Chen W.; O’Donnell B.; Patriarche G.; Bouchoule S.; Pareige P.; Rogel R.; Claire Salaun A.; Pichon L.; Roca i Cabarrocas P. Growth-in-Place Deployment of in-Plane Silicon Nanowires. Appl. Phys. Lett. 2011, 99, 203104. 10.1063/1.3659895. [DOI] [Google Scholar]
  41. Yu L.; Oudwan M.; Moustapha O.; Fortuna F.; Roca i Cabarrocas P. Guided Growth of in-Plane Silicon Nanowires. Appl. Phys. Lett. 2009, 95, 113106. 10.1063/1.3227667. [DOI] [Google Scholar]
  42. Xu M.; Xue Z.; Yu L.; Qian S.; Fan Z.; Wang J.; Xu J.; Shi Y.; Chen K.; Roca i Cabarrocas P. R. Operating Principles of in-Plane Silicon Nanowires at Simple Step-Edges. Nanoscale 2015, 7, 5197–5202. 10.1039/C4NR06531J. [DOI] [PubMed] [Google Scholar]
  43. Zhao Y.; Yao J.; Xu L.; Mankin M. N.; Zhu Y.; Wu H.; Mai L.; Zhang Q.; Lieber C. M. Shape-Controlled Deterministic Assembly of Nanowires. Nano Lett. 2016, 16, 2644–2650. 10.1021/acs.nanolett.6b00292. [DOI] [PubMed] [Google Scholar]
  44. Hu Y.; Li J.; Tian J.; Xuan Y.; Deng B.; McNear K. L.; Lim D. G.; Chen Y.; Yang C.; Cheng G. J. Parallel Nanoshaping of Brittle Semiconductor Nanowires for Strained Electronics. Nano Lett. 2016, 16, 7536–7544. 10.1021/acs.nanolett.6b03366. [DOI] [PubMed] [Google Scholar]
  45. Durham III J. W.; Zhu Y. Fabrication of Functional Nanowire Devices on Unconventional Substrates Using Strain-Release Assembly. ACS Appl. Mater. Interfaces 2013, 5, 256–261. 10.1021/am302384z. [DOI] [PubMed] [Google Scholar]
  46. Ryu S. Y.; Xiao J.; Park W. I.; Son K. S.; Huang Y. Y.; Paik U.; Rogers J. A. Lateral Buckling Mechanics in Silicon Nanowires on Elastomeric Substrates. Nano Lett. 2009, 9, 3214–3219. 10.1021/nl901450q. [DOI] [PubMed] [Google Scholar]
  47. Xu F.; Lu W.; Zhu Y. Controlled 3d Buckling of Silicon Nanowires for Stretchable Electronics. ACS Nano 2011, 5, 672–678. 10.1021/nn103189z. [DOI] [PubMed] [Google Scholar]
  48. Xu J.; Oksenberg E.; Popovitz-Biro R.; Rechav K.; Joselevich E. Bottom-up Tri-Gate Transistors and Submicrosecond Photodetectors from Guided CdS Nanowalls. J. Am. Chem. Soc. 2017, 139, 15958–15967. 10.1021/jacs.7b09423. [DOI] [PubMed] [Google Scholar]
  49. Chou S. Y.; Krauss P. R.; Renstrom P. J. Imprint of Sub-25 nm Vias and Trenches in Polymers. Appl. Phys. Lett. 1995, 67, 3114–3116. 10.1063/1.114851. [DOI] [Google Scholar]
  50. Namatsu H.; Takahashi Y.; Yamazaki K.; Yamaguchi T.; Nagase M.; Kurihara K. Three-Dimensional Siloxane Resist for the Formation of Nanopatterns with Minimum Linewidth Fluctuations. J. Vac. Sci. Technol., B: Microelectron. Process. Phenom. 1998, 16, 69–76. 10.1116/1.589837. [DOI] [Google Scholar]
  51. Lan H. B.; Ding Y. C. Nanoimprint Lithography. Lithography 2010, 457–494. 10.5772/8189. [DOI] [Google Scholar]
  52. Neeman L.; Ben-Zvi R.; Rechav K.; Popovitz-Biro R.; Oron D.; Joselevich E. Crystallographic Mapping of Guided Nanowires by Second Harmonic Generation Polarimetry. Nano Lett. 2017, 17, 842–850. 10.1021/acs.nanolett.6b04087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Oksenberg E.; Martí-Sánchez S.; Popovitz-Biro R.; Arbiol J.; Joselevich E. Surface-Guided Core-Shell ZnSe@ ZnTe Nanowires as Radial p-n Heterojunctions with Photovoltaic Behavior. ACS Nano 2017, 11, 6155–6166. 10.1021/acsnano.7b02199. [DOI] [PubMed] [Google Scholar]
  54. Chen C.-C.; Yeh C.-C.; Chen C.-H.; Yu M.-Y.; Liu H.-L.; Wu J.-J.; Chen K.-H.; Chen L.-C.; Peng J.-Y.; Chen Y.-F. Catalytic Growth and Characterization of Gallium Nitride Nanowires. J. Am. Chem. Soc. 2001, 123, 2791–2798. 10.1021/ja0040518. [DOI] [PubMed] [Google Scholar]
  55. Ji H.; Kuball M.; Burke R. A.; Redwing J. M. Vibrational and Optical Properties of GaN Nanowires Synthesized by Ni-Assisted Catalytic Growth. Nanotechnology 2007, 18, 445704. 10.1088/0957-4484/18/44/445704. [DOI] [Google Scholar]
  56. Kuykendall T.; Pauzauskie P.; Lee S.; Zhang Y.; Goldberger J.; Yang P. Metalorganic Chemical Vapor Deposition Route to GaN Nanowires with Triangular Cross Sections. Nano Lett. 2003, 3, 1063–1066. 10.1021/nl034422t. [DOI] [Google Scholar]
  57. Kuykendall T.; Pauzauskie P. J.; Zhang Y.; Goldberger J.; Sirbuly D.; Denlinger J.; Yang P. Crystallographic Alignment of High-Density Gallium Nitride Nanowire Arrays. Nat. Mater. 2004, 3, 524. 10.1038/nmat1177. [DOI] [PubMed] [Google Scholar]
  58. Kuykendall T. R.; Altoe M. V. P.; Ogletree D. F.; Aloni S. Catalyst-Directed Crystallographic Orientation Control of GaN Nanowire Growth. Nano Lett. 2014, 14, 6767–6773. 10.1021/nl502079v. [DOI] [PubMed] [Google Scholar]
  59. Li Q.; Wang G. T. Improvement in Aligned GaN Nanowire Growth Using Submonolayer Ni Catalyst Films. Appl. Phys. Lett. 2008, 93, 043119. 10.1063/1.2965798. [DOI] [Google Scholar]
  60. Lim S. K.; Crawford S.; Gradečak S. Growth Mechanism of GaN Nanowires: Preferred Nucleation Site and Effect of Hydrogen. Nanotechnology 2010, 21, 345604. 10.1088/0957-4484/21/34/345604. [DOI] [PubMed] [Google Scholar]
  61. Wang G. T.; Talin A. A.; Werder D. J.; Creighton J. R.; Lai E.; Anderson R. J.; Arslan I. Highly Aligned, Template-Free Growth and Characterization of Vertical GaN Nanowires on Sapphire by Metal-Organic Chemical Vapour Deposition. Nanotechnology 2006, 17, 5773. 10.1088/0957-4484/17/23/011. [DOI] [Google Scholar]
  62. Zhou X.; Chesin J.; Crawford S.; Gradečak S. Using Seed Particle Composition to Control Structural and Optical Properties of GaN Nanowires. Nanotechnology 2012, 23, 285603. 10.1088/0957-4484/23/28/285603. [DOI] [PubMed] [Google Scholar]
  63. Schvartzman M.; Nguyen K.; Palma M.; Abramson J.; Sable J.; Hone J.; Sheetz M.; Wind S. Fabrication of Nanoscale Bioarrays for the Study of Cytoskeletal Protein Binding Interactions Using Nanoimprint Lithography. J. Vac. Sci. Technol., B: Microelectron. Nanometer Struct.--Process. Meas., Phenom. 2009, 27, 61–65. 10.1116/1.3043472. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Schvartzman M.; Wind S. J. Robust Pattern Transfer of Nanoimprinted Features for Sub-5-nm Fabrication. Nano Lett. 2009, 9, 3629–3634. 10.1021/nl9018512. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

nn9b00538_si_001.pdf (2.3MB, pdf)

Articles from ACS Nano are provided here courtesy of American Chemical Society

RESOURCES