Mechanically Strong, Optically Transparent, Giant Metal Superlattice Nanomembranes From Ultrathin Gold Nanowires
Article first published online: 18 SEP 2012
Copyright © 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Volume 25, Issue 1, pages 80–85, January 4, 2013
How to Cite
Chen, Y., Ouyang, Z., Gu, M. and Cheng, W. (2013), Mechanically Strong, Optically Transparent, Giant Metal Superlattice Nanomembranes From Ultrathin Gold Nanowires. Adv. Mater., 25: 80–85. doi: 10.1002/adma.201202241
- Issue published online: 2 JAN 2013
- Article first published online: 18 SEP 2012
- Manuscript Revised: 24 JUL 2012
- Manuscript Received: 4 JUN 2012
- gold nanowires;
- transparent membranes
Lipids and polymers are classic materials to be used in natural and artificial membrane systems, respectively. In contrast, recently there is burgeoning interest in constructing new class of membranes from nanoscale optoelectronic building blocks.1–11 In particular, it is possible to fabricate free-standing periodic nanoparticle arrays (superlattices) which behave as mechanically strong nanomembranes with Young's moduli of several GPa.3, 6, 11 Such superlattice nanomembranes differ from solid-supported superlattice films12–15 in that they exist as free-standing, isolated forms in their final stage of growth or processing. Remarkably, both optoelectronic and mechanical properties of such superlattice membranes are tunable by varying materials types, adjusting ligand length, and regulating lattice structures, etc. Hence, engineering superlattice nanomembranes constitutes an exciting route to integrate unique properties of optoelectronic nanomaterials with striking mechanical robustness and flexibility into one tailorable multifunctional system, leading to lightweight metamaterials and devices with new functions for novel applications in optoelectronics,9, 16, 17 electrocatalysis,18 and ultrafiltration.19
Despite the exciting progress in the fabrication of free-standing metallic superlattice nanomembranes, their practical applications are often hampered by the difficulty in fabricating defect-free superlattice membranes at large area. In particular, it is highly desirable to develop the ability to fabricate giant metal superlattice nanomembranes (two-dimensional, ordered metallic nanoparticle arrays forming ultrathin membrane of a few nanometers thick but with macroscopic lateral dimensions). This is because nanoscale thickness and macroscopic lateral dimensions are apparently conflicting characteristics difficult to achieve simultaneously.2 Previous attempts with sphere-like hard building blocks only led to superlattice membranes restricted to small lateral dimensions.3–6, 9, 10
Here, we report on the fabrication of giant superlattice nanomembranes from soft 1D building blocks—ultrathin single-crystalline gold nanowires (AuNWs)—via Langmuir–Blodgett (LB) techniques. Although LB techniques have been widely used for the assembly of nanowires,20–25 the ability to form mechanically-robust free-standing nanomembranes has not yet been reported, to the best of our knowledge. This could be due to the fact that our gold nanowires are thinner and more flexible, contributing to maintaining the integrity of the nanomembranes at large area. Specifically, we show a single-layer AuNWs superlattice nanomembrane is transparent, conductive and mechanically robust, with an optical transmittance of 90–97% over a wide spectral window of 300–1100 nm, an electrical resistance of ∼1142 kΩ sq−1, and a breaking strength of ∼14 N m−1 with a typical atomic force microscope probe. Such single-layer nanowire membranes are transferable to any arbitrary substrates, facile to be integrated into lightweight, foldable optoelectronic devices with low consumption of materials and energy. Our methodology may serve as a model system, extendable to superlattice nanomembranes from other materials for a myriad of applications in construction of new classes of two-dimensional metamaterials and devices.
We began with the large-scale synthesis of high-quality single crystalline ultrathin AuNWs by adopting the recently reported approaches.26–28 In brief, gold precursor (HAuCl4) was gradually reduced by triisopropylsilane (TIPS) in the presence of oleylamine (OA). After aging the solution at room temperature for 2 days, the solution turned from yellow to dark red (Figure 1a,b), indicating complete reduction of Au+-amine complex into OA-capped AuNWs dispersible in hexane27 (Figure 1c). The nanowires were precipitated out by adding ethanol and centrifugation, and then washed twice with ethanol and re-dispersed in chloroform as concentrated stock solution for further use.
In the purified nanowires, the OA molecules contribute to 18% of the total weight as demonstrated by thermogravimetric analysis (Supporting Information, Figure S1). Thus, the OA capping effectively stabilizes nanowires and prevents them from deterioration in various characterization and processing in the fabrication of nanomembranes. Electron microscopy studies show each nanowire is single crystalline growing along the  direction (Figure 1d) with an interfringe distances of ∼0.23 nm and ultrathin with a width of ∼2.5 nm. The nanowires have enormous aspect ratios, extending to tens of micrometers in length. These structural characteristics give nanowires high mechanical robustness and flexibility, leading to hairy morphology (Figure 1e). Due to strong hydrophobic interactions between OA ligands, the nanowires tended to form parallel bundle-like strands (Figure 1f,g).
Furthermore, we applied the as-synthesized AuNWs to fabricate giant superlattice nanomembranes by LB techniques. Typically, one droplet (∼4–6 μL) of nanowires dispersed chloroform solution was carefully spread onto the water surface in a LB trough. The droplet quickly spread and evaporated in less than 1s, distributing the nanowires uniformly on the water surface. By moving LB baffle, the free-floating nanowires were forced to approach each other with gradual phase transitions (Supporting Information, Figure S2). In particular, a Mott-insulator-to-metal transition was noted at the surface pressure of ∼13 mN m−1. This transition was indicated by the appearance of red-colored metallic sheen, and further demonstrated by turning the slope in the typical π-A isotherm.22 A further compression led to rapid pressure buildup from 13 mN m−1 to 38 mN m−1 until collapse, and in this process a condensed nanowire membrane formed on the water surface. Figure 2a shows a giant nanowire superlattice membrane covering an entire trough area of ∼35 cm2. The membranes were mechanically strong yet flexible, transferable to a variety of substrates (including silicon, glass slides, paper, plastics, TEM copper grids, and PDMS, etc) without cracking or rapturing (Figure 2b and Supporting Information, Section 2). The coating could substantially alter the wetting properties. For example, the contact angle of a glass slide increased from ∼26° to ∼103° after nanomembrane covering (Figure 2c).
Remarkably, the nanomembranes consisted of dominantly monolayered nanowires aligned in parallel. Figure 2d shows a typical atomic force microscopy (AFM) image of AuNWs nanomembranes transferred onto a silicon wafer. The membrane is uniformly continuous extending to centimeter-scale with little defects. Further AFM line scanning gave an average thickness of 2.2 ± 0.1 nm (Figure 2e). We found that monolayered superlattice nanomembranes could withstand large lateral compression forces up to ∼38 mN m−1. We transferred the membranes to lacey/carbon copper grid at specific surface pressures of 20 mN m−1 and 30 mN m−1, and characterized their morphologies (Figure 2f,g). Both images indicated predominantly the monolayered AuNWs, in agreement with the above AFM characterizations. Nevertheless, under high compression forces, the membranes tended to be crumpled by piling (Figure 2g) or bending (Supporting Information, Figure S5). Nevertheless, the membranes remain intact without any cracking or rupturing during the LB compression process.
To quantitatively determine the mechanical strength of free-standing superlattice nanomembranes, we performed AFM nanoindentation experiments. Firstly, the superlattice membranes were transferred onto holey silicon nitride substrate patterned with periodic 2 μm-diameter holes. The superlattice membranes could span over the holes. Majority of membranes (∼80%) remained intact after transfer and survived from vacuum drying (Supporting Information, Figure S6). By high-resolution imaging, we located specific intact membranes (Figure 3d) and then indented at the center of microholes. Force-displacement curves were recorded with stepwise increase of force until rupturing the membrane (Figure 3a). Notably, the ruptured holes were V-shaped memorizing the shapes of pyramid AFM probes (Figure 3e,f), and the adjacent holes didn't coalescence (Supporting Information, Figure S9), indicating paper-like mechanical properties. We observed consistent force curves for both multiple indents within a specific holey region, as well as other holey areas. Fitting of the force curves allowed us to derive a Young's modulus of ∼5.2 ± 0.4 GPa using the similar model reported earlier3 (Supporting Information, Section 4). This value is comparable to those for DNA-nanoparticle membrane3 and alkyl-nanoparticle membrane.6, 11
We further estimated the breaking strength when using a typical AFM probe. Based on the models used for monolayer graphene,29 we assume our nanowire superlattice nanomembrane as a true 2D material given the ultrathin nature and large aspect ratio of our nanomembrane. Thus, the strain energy density can be normalized by the area rather than the volume of the nanomembrane. Hence, its mechanical behavior under AFM probe indentation is properly described by a 2D stress, σ2D and 2D modulus, E2D. Both have units of force/length rather than force/area for normal 3D material systems. The force-displacement behavior for a clamped circular membrane under central point loading can then be approximated as:
where F is the applied force, σ2D0 is the pretension in the membrane, R is the radius of circular membrane, δ is the deflection at the center, E2D is the 2D modulus, and q is a constant depending on poisson's ratio, ν, by the equation, q = 1/(1.05-0.15ν-0.16ν2). Numerical fitting of 10 sets of force curves from different locations under different loads gave σ2D and E2D of 1.4 ± 0.1 N/m and 52 ± 23 N/m, respectively (Figure 3b and c). With these values on the basis of the continuum model used for 2D material,29 we could then estimate the breaking strength of 14.3 ± 1.4 N/m with a typical AFM probe with a tip radius of 24 nm. Following the identical treatment, we also estimated our previous nanoparticle membranes from DNA of 30 and 50 bases long, giving breaking strengths of 4.5 N/m and 1.8 N/m, respectively. Considering the fact that 100 nm-radius probe was used in DNA-based nanomembranes, the above results show that the nanowire superlattice membrane is much stronger than DNA-nanoparticle membrane. This may originate from the 1D nanowire building blocks used in this study. It is harder to break gold nanowires than tangled DNA strands.
The monolayer AuNWs nanomembrane exhibited exceptionally high optical transmittance of 90–97% with an almost flat line over a fairly wide spectral range of 300∼1100 nm (Figure 2b, and Figure 4). The observed high optical transparency is due to the ultrathin nature of our superlattice nanomembranes, which is in agreement with the theoretical prediction. We estimated the theoretical transmittances of nanomembranes with different wire diameters by calculating the decay of an electromagnetic wave after propagating through a medium (Supporting Information, Section 5). The results showed an evident decreasing trend of the transmittance with increasing diameters (Supporting Information, Figure S10). Note that the calculated theoretical transmittance of 2.5 nm AuNWs monolayer nanomembrane provides a reasonable fit of our measured profile, which further proved the monolayer structure of our nanomembrane. The slight decrease in the region of 500–550 nm is due to the surface plasmon resonance of spherical nanoparticles existing as impurity in the nanomembrane as seen in Figure 1e.30
In addition, monolayered AuNWs superlattice membranes are conductive. The square resistances of a set of superlattice membranes from 15 batches of stock nanowire solutions were measured by a four-point probe station, giving a value of 1142±16 kΩ sq−1. This value is substantially higher than that given by a theoretical modeling, in which nanowires are assumed to be infinitely long and ideally aligned to the device scale. This model predicted a sheet resistance of 15–200 Ω sq−1 (Supporting Information, Section 6) along the direction of the AuNWs. The discrepancy may be due to the fact that AuNWs have finite length and the alignment is limited to micrometer-scale, and there may exist a large contact resistance at the joint of two nanowires.31 We further investigated relationship between optical transmittance and electrical conductivity by multilayer coating. As the coating layer of nanomembrane increased, both the transmittance and sheet resistance decreased (Supporting Information, Figure S11). Similar to transparent electrode from other materials,31, 32 the relationship between the transmittance and sheet resistance can be described by Equation S4 in supporting information. However, the square resistance for multilayered nanomembrane remains very high. We applied the normal pressing forces to the mulilayered membranes and observed improvement in conductivity (15–19%), which may be due to the decrease of wire-to-wire spacing during mechanical compression.33 Nevertheless, the conductivity remains far from theoretical prediction, indicating that the densely-packed capping oleylamine molecules surrounding AuNWs may substantially prevent electron-hopping from wire to wire.
Despite high sheet resistance, superlattice nanomembrane could withstand mechanical bending for hundreds of cycles. We monitored the sheet resistance of superlattice nanomembranes (Supporting Information, Figure S12). In particular, the sheet resistance for monolayer nanomembrane only increased by about 14% after 400 bending cycles.
In summary, this paper presents a simple yet efficient method to fabricate giant metallic superlattice nanomembrane of about 2.5 nm thick but with macroscopic lateral dimensions. To the best of our knowledge, such superlattice nanomembrane represents the thinnest version of metallic membranes known to date.3, 6, 11 Remarkably, the ultrathin superlattice nanomembranes are mechanically strong, optically transparent and electrically conductive. Notably, both synthesis of gold nanowires and fabrication process of superlattice nanomembranes are scalable, which render our methodology promising for practical applications in future lightweight foldable optoelectronics, such as touch-screen devices.33, 34 Our methodology can potentially be extended to fabricate nanomembranes from other types of nanowires or multiple components, therefore, it may become a general modular approach to integrate optoelectronic properties with mechanical robustness into a single lightweight, foldable 2D materials systems for promising applications in future foldable electronics.
Materials: Gold (III) chloride trihydrate (HAuCl4·3H2O, ≥99.9%), Triisopropylsilane (99%) and Oleylamine were purchased from Sigma Aldrich. Hexane and chloroform were obtained from Merck KGaA. All chemicals were used as received unless otherwise indicated. All aqueous solutions were made using demineralized water, which was further purified with a Milli-Q system (Millipore). All glassware used in the following procedures was cleaned in a bath of freshly prepared aqua regia and rinsed thoroughly in H2O prior to use.
Holey Lacey Formvar/Carbon films (300 mesh and irregular pores with sizes varying from less than a quarter of 1 μm to more than 10 μm) were purchased from Ted Pella. Quantifoil holey carbon films (2 μm-diameter hole, 2 μm in space) were purchased from Electron Microscopy Sciences. Holey silicon nitride support films (2-μm-diameter hole, 4 μm pitch) were purchased from SPI supplies.
Characterization: Morphology characterization was carried out using a Philips CM20 TEM at a 200 kV accelerating voltage, Hitachi H-7500 field emission TEM operating at 80 kV, and JEOL JSM-7001F FEG SEM. High-resolution TEM images were performed on a JEOL JEM 2011 TEM operated at an acceleration voltage of 200 kV. Stepwise focusing with low beam currents helped to avoid distortion of AuNWs by the electron beam. The transmittance of superlattice nanomembrane is measured with Agilent 8453 UV–vis spectrophotometer, and the transparent images were recorded by a CCD camera on the J&M MSP210 Microscope Spectrometry System, while the membrane was transferred onto glass slide and illuminated by a high-intensity fiber light source under a 40× objective. Thermogravimetric analysis (TGA) was carried out using a Mettler Toledo TGA-SDTA851 analyzer in the temperature range from 25 to 700 °C at a heating rate of 10 °C/min. The runs were performed under N2 gas, and the gas was switched to air for 10 min at the end of the run at 700 °C to ensure complete combustion of the organic material. N2 and air gases were used at a rate of 50 mL/min.
Force-displacement curves from nanoindentation and topography images were obtained with a Veeco Dimension Icon AFM in tapping mode using Bruker silicon probes (MPP-21100-10). The spring constant for cantilever was 3 N m−1. The typical tip speed for the indentation was 500 nm s−1. Each membrane was first imaged by AFM, then gradually increased the applied force and performed multiple indentations with the AuNWs superlattice membrane to collect the force-indentation curve, and finally imaged again to check for rupture of membrane. See Supporting Information Section 4 for a detailed analysis.
Synthesis of Ultrathin AuNWs: Large-scale synthesis of ultrathin AuNWs was performed according to the literature.26 HAuCl4·3H2O (53 mg) and hexane (50 mL) were added into a 100 mL Schott/Duran bottle, followed by addition of Oleylamine (OA, 1.8 mL) as both stabilizer and growth template. After completely dissolving of gold precursor, Triisopropylsilane (TIPS, 2.5 mL) was added into above auric solution and allowed to stand for 2 days without disturbing at room temperature. Finally, the solution turned into dark-red, indicating the formation of AuNWs (0.58 mg mL−1). The residue chemicals were removed by repeated centrifuging and thorough washing by ethanol/hexane (1/3, v/v) and finally concentrated into 5 mL that dispersed in chloroform (5.82 mg mL−1). Statistical evaluation from TEM images of as-prepared AuNWs shows an estimated yield of ∼90%. The surfaces of the as-made AuNWs were capped by oleylamine and indicated as hydrophobic property.
Fabrication and Transfer of AuNW Superlattice Nanomembrane: AuNWs superlattice membrane was prepared using a Langmuir-Blodgett trough (Nima Technology) at 25 °C. The Telflon-coated trough and barrier were wiped with CHCl3 and then purged with ethanol to remove any dust or organic contaminants. In a typical experiment, 10 mL purified AuNWs (0.58 mg mL−1) was centrifuged and redispersed in 1 mL CHCl3 (5.82 mg mL−1). Then the concentrated AuNWs solution was added dropwise to the air-water interface of the LB trough. The droplet quickly spread over the water surface in less than 1s and evaporated to leave an isotropically distributed AuNWs floating on the water surface. The trough was then covered to prevent fluid flows and solvent fluctuations, and allowed to equilibrate for 30 min. The AuNWs were then isothermally compressed by moving the two opposing barriers towards each other while the surface pressure was monitored with a Wilhelmy plate. At a target surface pressure (e.g., 20 mN m−1 for monolayered AuNWs superlattice membrane), the pressure was kept constant at least 1 h. After the equilibration, the AuNWs superlattice membrane was transferred onto hydrophilic substrates by slowly pulling the substrate out of the aqueous subphase with a vertical dipping speed of 5 mm min−1, and transferred to hydrophobic substrates by horizontal deposition through Langmuir-Schaefer technique. Also, transfer of multilayer AuNWs nanomembrane could be realized by using layer-by-layer transfer method. See Supporting Information Section 2 for a detailed description.
Supporting Information is available from the Wiley Online Library or from the author.
This work is financially supported by the New Staff Member Research Fund awarded by the Faculty of Engineering, Monash University and ARC discovery project DP120100170. The authors also acknowledge the use of facilities in the Monash Centre for Nanofabrication by the program of MCN technology fellows. Y.C. acknowledges the graduate student scholarship (MGS & IPRS) from Monash University. Z.O. and M.G. acknowledge the support from the Victoria-Suntech Advanced Solar Facility (VSASF) funded under the Victoria Science Agenda (VSA) scheme.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.