Zero-dimensional to three-dimensional nanojoining: current status and potential applications

Abstract

The continuing miniaturization of microelectronics is pushing advanced manufacturing into nanomanufacturing. Nanojoining is a bottom-up assembly technique that enables functional nanodevice fabrication with dissimilar nanoscopic building blocks and/or molecular components. Various conventional joining techniques have been modified and re-invented for joining nanomaterials. This review surveys recent progress in nanojoining methods, as compared to conventional joining processes. Examples of nanojoining are given and classified by the dimensionality of the joining materials. At each classification, nanojoining is reviewed and discussed according to materials specialties, low dimensional processing features, energy input mechanisms and potential applications. The preparation of new intermetallic materials by reactive nanoscale multilayer foils based on self-propagating high-temperature synthesis is highlighted. This review will provide insight into nanojoining fundamentals and innovative applications in power electronics packaging, plasmonic devices, nanosoldering for printable electronics, 3D printing and space manufacturing.

---

1 Introduction

Nanostructured materials with unique optical, electrical, and thermal properties have been widely explored with applications in various areas from biosensors to electronic nanodevices.^1–6 These unusual properties arise from the emergence of quantum confinement and surface state effects. Therefore, the shape, surface crystal orientation, internal structures and defects can significantly alter the properties of nanoparticles beyond the properties of the bulk material. For instance, silver nanoparticles exhibit a strong absorption band in the visible spectrum due to the surface plasmon oscillation modes of the conduction electrons, enabling significantly enhanced photocatalysis.^7,8 Furthermore, alloy, core–shell or bimetallic nanomaterials exhibit tunable properties beyond those of monometallic nanomaterials.^9–11 Nowadays, the syntheses of various nanomaterials, including nanoparticles, nanowires, nanotubes, multilayer films with nanoscale thicknesses, and core–shell nanostructures have been well developed.^12–15 Joining those nanoscopic building blocks becomes a challenging but crucial step for nanomanufacturing functional devices.¹⁶

Nanojoining is a bottom-up nanofabrication technique to join low-dimensional nanoscale building blocks with metallic or covalent bonds.^17,18 As devices become smaller and smaller, the use of nanoparticles and/or nanowires as building blocks is often preferred in order to lower energy and material costs.¹⁹ A variety of nanojoining methods have now been developed analogously to microjoining or conventional joining methods at macro-scales. However, due to size effects and high surface energy, thermodynamic behaviors of nanomaterials are different from their mesoscale counterparts. For instance, a nanoparticle can melt at a temperature much lower than the bulk melting point. Therefore, nanojoining methods should be designed and established to exploit low dimensional and nanoscopic functionalities. The appropriate nanomanipulation should be developed and the input energy for joining should be significantly decreased and locally delivered. In this work, we review the latest progress in nanojoining methods. Fundamentals for joining are analyzed, state-of the art applications are summarized, and the outlook of research and development and potential applications are discussed. This article is intended to be a useful primer for researchers who are exploring innovative nanojoining, for interconnection of nanomaterials at low temperatures or even room temperature, and for developing other new devices and applications.

2 Nanojoining based on nanoparticles

Unique thermodynamic, spectroscopic, electronic and chemical properties due to their small sizes and large surface/volume ratio.^20,21 Since surface atoms have fewer bonds than internal atoms, less energy is needed for them to leave the surface. Hence a large surface/volume leads to a size depression effect of the particle melting temperature. Accordingly to classical Gibbs–Thomson equation, the melting point of particles (T_M(d)) can be expressed as follows²²
where T_MB is the bulk melting point, σ_sl is the solid–liquid interface energy, H_f is the bulk heat of fusion, ρ_s density of solid and d is the particle size. In a practical situation, particle surfaces have a higher concentration of defects.²³ As a result, the surface atomic mobility is higher. Nanoparticles thus demonstrate more activated behaviors and have a much early onset of sintering temperatures. Collectively, the sintering temperature as a function of the melting temperature can written as

T_S(d) = αT_M(d)

where T_S(d) is the onset temperature of sintering. For microparticles, α ranges from 0.5 to 0.8,²³ but for nanoparticles, this value decreases to 0.1 to 0.3.^17,24

The joining process of nanoparticles can roughly classified into three major types according to the physical states of the parts to be joined and the degree of melting required: solid-state bonding, soldering/brazing, and fusion welding.²⁵ Fusion welding requires the melting of the bond surfaces.¹⁹ The heat necessary to melt the metal can be provided by an electrical arc or a laser beam, for example laser-induced nanowelding. Part of the inter-particle gaps were replaced by metallic necks. A solid-state bonding, on the other hand, never reaches the temperature necessary to melt the two objects being bonded. If nanoparticles are fully melt, their unique properties associated with their shapes and surface crystalline orientations certainly change. This poses a question to use conventional joining methods for nanojoining.²⁶ Obviously, due to the size depression effect, less energy is required for nanosintering. For a nanojoining, the input energy can be reduced from a couple nanojoules to a few eV level.²⁴ Therefore, the thermal side effects during nanojoining should be minimized. The highly localization of thermal effect to the particle surface is a challenge.

2.1 Cold welding

Cold welding at the nanoscale is one kind of solid-state bonding process which can occur at low temperature or even room temperature by minimization of the surface free energy through mass transfer in the form of atomic diffusion and surface relaxation.¹⁸ As aforementioned, at the nanoscale, the diffusion barrier of a single metal atom on a clean surface is low enough for a small thermal activation to trigger atomic diffusion. As a result, the melting point of nanoparticles can be far below that of pure bulk materials; indeed some nanoscale metals can flow like liquid at room temperature.^27,28 It is notable that once two particles fuse together, the mobility of surface atoms will decrease, i.e.: the joined particles will be more stable than two individual particles.²⁹ The mechanism of cold welding is based on this phenomenon of enhanced thermodynamic stability of fused particles.³⁰

Conventionally, particle fusion is a spontaneous process during which smaller particles coalesce to form larger particles, a so-called Ostwald ripening. The nanoparticle aggregation and subsequent grain growth due to particle fusion are demonstrated by Fig. 1. First, the capping ligands melt or desorb, which enables the nanoparticles to aggregate and join together. The aggregation of nanoparticles leads to the decrease of surface/volume ratio, thus significantly lowering the surface chemical potential.³⁰ At longer times, grain growth can be observed, and stacking fault densities decrease. The time scale of the grain growth process is significantly longer than that of the particle aggregation.³¹ The fusion-induced particle growth is thus – in principle – a cold welding.

Fig. 1 The processes of aggregation and subsequent grain growth.³¹

Liu et al. proposed a cold welding strategy to rapidly join together Au nanoparticles into two-dimensional continuous structures for enhancing the electrooxidation of carbon monoxide.³² As illustrated in Fig. 2, a mixture of ethanol and toluene was injected into the bottom of an 18 nm Au nanoparticle solution at a rate of 2.0 mL min⁻¹. The key factor to achieve the cold welding of Au nanoparticles is the volume concentration of ethanol in the rejected mixture of ethanol and toluene. At a relatively low ethanol concentration, isolated and slightly welded Au nanoparticles co-existed, as shown in Fig. 3A. As the ethanol concentration increases in the mixture, welding between neighboring Au nanoparticles occurred and gradually changed from a necklace-like structure (Fig. 3B) to a dendritic structure (Fig. 3C).

Fig. 2 Illustration of the cold welding and assembly process of Au nanoparticles.³²

Fig. 3 Transmission Electron Microscope (TEM) images of 18 nm Au nanoparticle monolayer films obtained by injecting mixtures with the ethanol : toluene volume ratios of 1 : 3 (A), 1 : 2 (B) and 1 : 1 (C), respectively. The injected volume of ethanol was fixed at 0.75 mL.³²

The coalescence of tetraoctylammonium bromide (TOAB)-stabilized gold nanoparticles has been investigated using high-resolution transmission electron microscopy (HRTEM).³³ Two types of coalescence have been observed. The first consists of fusion between two or more particles of different sizes to form a multiply-twinned structure in which smaller nanoparticles are consumed by a larger one, and the second consists of the ordered combination of two nanoparticles with similar size and shape through a common lattice, as shown in Fig. 4. Indeed the latter is similar to an attach-alignment growth mechanism of nanowires based on nanoparticles.³⁴

Fig. 4 Coalescence of two or more gold particles (without facets) through {111} twinning. (a) A particle with a single-twin structure. (b) A particle with a triple-twin structure. (c) A particle with a fivefold twinning configuration. (d) and (e) Coalescence of two particles with facets through a common lattice plane.³³

2.2 Laser-induced nanowelding

Ultrashort laser pulse (picosecond or sub-picosecond) irradiation provides a powerful tool for nanowelding since the interaction between laser pulses and materials at those timescales involves ultrafast electron excitation, nonthermal melting and nanoscopic surface morphology modification.^35,36 As the laser pulse-width is much shorter than the electron-lattice thermal coupling time (usually a few picoseconds), electrons do not have enough time to transfer energy to the lattice. The excited electrons are thus ejected and thereby weaken the chemical bonds of lattice atoms, leading to the melting of surface atoms, which can be used for the welding of nanoparticles.^37,38

Picosecond laser pulses have been employed successfully to adjoin, hold closely, and weld gold nanoparticles on carbon-coated copper grids.³⁹ The surfactant-free gold nanoparticles were irradiated by the second harmonic of a mode-locked Nd:YAG laser with a wavelength of 532 nm, a pulse duration of 30 ps, and a repetition rate of 10 Hz. The spot diameter of the irradiating beam was 4 mm. As shown in Fig. 5, gold nanoparticles are well contacted in a single phase to show the typical crystalline fcc structure of gold.

Fig. 5 HRTEM images of gold nanoparticles held together and nanowelded by irradiating laser pulses of 532 nm and 0.2 mJ for 10 min on a carbon-coated copper TEM grid.³⁹

Femtosecond laser irradiation has also been used for nanowelding. A. Hu et al. observed two effects by femtosecond laser (100 nm wavelength, 0.3 mJ per pulse, 1 kHz) irradiation of an Au nanoparticle solution at a laser intensity of 4 × 10¹⁴ W cm⁻² for 10 min: the generation of a large number of nanoparticles with a size of 1–3 nm and welding of 2–3 Au nanoparticles with a size around 15 nm at different laser powers, as shown in Fig. 6b and c.¹⁹ It can be deduced that the smaller nanoparticles are created by the fragmentation of Au nanoparticles through laser irradiation. It is evident in Fig. 6c where only welded Au nanoparticles are observed. Fig. 6d clearly shows that two Au nanoparticles are welded with metallic bonds. Similar results have been observed by H. Huang et al.⁴⁰ 35 fs, 800 nm, laser pulses with a 1 kHz rep rate were utilized to irradiate polyvinylpyrrolidone (PVP) coated Ag nanoparticles. After 20 min irradiation, the formation of Ag nanoparticle networks can be seen, but the shape of individual particles are not significantly affected by the irradiation process. When the irradiation fluence is relatively low, Ag nanoparticles are separated by a narrow gap and they appear to be bonded by a carbonaceous shell. As the fluence increases, part of the inter-particle gaps were replaced by metallic necks, acting as bridges between individual Ag nanoparticles.

Fig. 6 Nanowelded Au nanoparticles by femtosecond laser irradiation. (a) Pristine liquid, (b) irradiation at an intensity of 4 × 10¹⁴ W cm⁻², (c) irradiation at 3 × 10¹⁰ W cm⁻², (d) local structures of typical welded necks between two particles shown in (c). Note that the scales are 20 nm in (a), 50 nm in (b), 10 nm in (c) and 5 nm in (d).¹⁹

2.3 Applications of nanojoined particles: plasmonic devices, printable electronics and power electronics

2.3.1 Plasmonic devices. Localized surface plasmons (LSPs) are quasiparticles arising from the collective oscillation of conduction electrons at the surface of metallic nanostructures induced by incident electromagnetic (EM) waves. LSPs can enhance the local intensity of the EM field distribution well below the diffraction limit and thereby significantly enhance the scattering or absorption cross-sectional areas of EM waves, leading to non-linear optical properties of nanoparticles. This combination of nanoscale field confinement and high nonlinear efficiencies enables a new class of plasmonic nanoscale devices with applications including super-resolution imaging,^41–43 THz communications,⁴⁴ high efficient energy conversion⁴⁵ and ultra-sensitive sensing.^46–48 To construct the metallic nanostructures into a plasmonic device, one or more nanoparticles should be integrated together to meet different structural or functional requirements. In recent years, the optical properties of various plasmonic nanoparticles (NPs) with different elements, shapes, and sizes have been studied.^49–51 Control over a combination of NP morphology, spacing, and bonding is critical in the nanomanufacturing of scalable plasmonic structures with deterministic optical properties. Nanojoining at low temperatures minimizes unwanted changes in the NP structure, and thus could provide the desired control.

By changing the interconnection and configuration of two or more NPs, the LSP resonance can be tuned across the UV-VIS-NIR spectrum. For example, the dipolar LSP response of Au NPs can be tuned from a wavelength of 532 nm for a single NP to 745 nm by molecule gluing with a sub-5 nm gap, if 6 NPs were joined together with metallic bonds into a chain-like configuration a 1100 nm LSP resonance was observed as shown in Fig. 7a.⁵² Fig. 7b shows two joined Au NPs with different ‘necking’ sizes ranging from 15 to 30 nm. The experimental and simulation results both showed a red-shift of the LSP resonance with decreasing ‘necking’ size as shown in Fig. 7c. Ag NPs can also be held with a thin carbon layer (Fig. 8a and b) and then welded with metallic bonds (Fig. 8c and d) which are both achieved by femtosecond laser irradiation. Fig. 8e shows the demonstration of welded NPs for bio-sensing application.⁵³ Joining with metallic bonds therefore provides widely tunable linear and nonlinear optical properties of plasmonic nanostructures in addition to firm and reliable interconnection for devices.¹⁹

Fig. 7 (a) Charge-stabilized gold NPs exhibit single LSPs at 532 nm resonance. Particles glued into chains by CB molecules redshift the resonant absorption to 745 nm. The resonance further redshifts to 1100 nm after joining the NPs into a chain with a femtosecond laser. (b) SEM images of joined Au NPs with different ‘necking’ size. (c) The LSP response of joined Au NPs (dimer) with different ‘necking’ size (nanobridge thickens).⁵²

Fig. 8 (a and b) Ag NPs bonded by thin carbon layer and (c and d) Ag NPs welded by metallic bonds with femtosecond laser. (e) The Raman spectroscopic biosensor using such welded Ag NPs for sensing adenine.⁵³

2.3.2 Printable electronics. Typically, printed electronics refers to electronics fabricated with various functional inks and a printer.^54,55 NPs stably suspended in solutions with surfactants and organic bonder form inks with tunable optoelectronic properties. The sintering of printed metallic nanoparticles enables the fabrication of conducting electronic circuits,^24,56 especially for wearable electronics on flexible and stretchable substrates. At present, these ink-based printing methods mainly include screen,⁵⁷ ink-jet^58,59 and aerosol jet⁶⁰ printings, which are easy to control with multi-scale fabrication processes. Usually, NP inks form conducting networks after welding if the welding temperature is not too high, as shown in Fig. 9a. Fig. 9b shows the welded Cu NP 3D network through Cu nanoparticle sintering.⁶¹ These conductive patterns are the key component of flexible electronics. For conductive electrodes, the printable ink can contain one or more type(s) of NPs, including metal, semiconductor or carbon. Recently, a new direct-writing technique has been reported for fabricating flexible micro-electrodes for supercapacitors on paper by femtosecond laser induced in situ photoreduction of graphene oxide and Au NP networks with welded Au NPs.⁶² Fig. 10 shows the concept and the final device. It is notable that this technique enables a micromanufacturing of 3D devices,⁶³ and thus will enable cost-efficient and eco-friendly fabrication of on-chip energy devices for wearable electronics.

Fig. 9 (a) Typical microstructures of welded NPs, here in this case is Au, the welding temperature was 500 and 520 °C respectively.⁶³ (b) Printed conductive Cu patterns for electronics with Cu NP ink, inset is as-synthesized NPs.⁶¹

Fig. 10 Photoreduced graphene oxide and welded Au NPs electrodes for printing micro-supercapacitors on paper.⁶²

2.3.3 Power electronics packaging. Next-generation high-temperature power electronics are highly sought after for various applications including hybrid/full electric vehicles, renewable energy, high power semiconductor lasers, and aerospace applications. Packaging for power electronics is critical because of the need for a high operating temperature at a high power density. Wide bandgap semiconductors such as silicon carbide and gallium nitride are highly sought after active materials for next generation power electronics due to their ability to perform at high temperatures and higher breakdown voltages.^64–71 By extension, all packaging components must also be capable of high temperature performance, especially the die attach materials, electrical interconnections, and substrate materials, all of which can be seen in Fig. 11. The die attach material is the first means of thermal management in the power module; thus it must have a high thermal conductivity in order to dissipate heat in the die^72–74 as high temperatures in the device itself run the risk of damaging or melting smaller components in the die. The coefficient of thermal expansion must also match the die and substrate in order to prevent die or substrate fracture or fatigue.⁷⁵ Mechanically, the material must provide strong, reliable adhesion between the die and the substrate and be resilient enough to provide stress relaxation in the die. The substrate provides structural support and electrical insulation for the semiconductor device without sacrificing thermal conductivity. Direct-bonded metal substrates incorporating back metals on semiconductor materials are used more and more frequently for high temperature applications.^76,77 Materials used to attach the substrate to a baseplate or directly to the heat spreader have similar requirements as die attach materials. Conventional packaging materials include various tin-based or lead-based soldering alloys. However, due to the environmental concerns of lead and the lack of high temperature performance of tin-based soldering alloys, various alternatives are being explored.⁷⁸ Joined nanomaterials are of great interest for power electronics packaging, especially metal nanoparticles due to the previously mentioned low melting temperature of nanosized objects. The most widely investigated nanomaterials for power electronics packaging are Ag-based and Cu-based nanomaterials due to their excellent thermal and electrical conductivities and mechanical strength.⁷⁹ At the time of this review, Ag is more frequently chosen than Cu as a Pb-free soldering alternative.^80–82

Fig. 11 Basic construction of a power electronic device.

Research in using nanopastes for power electronics packaging has been focused primarily on lowering the required sintering temperature (<250 °C ambient temperature), decreasing the sintering time, and increasing the density of the sintered structure in order to increase mechanical bonding strength.⁸³ For metallic nanoparticles, joining at low temperatures can be challenging because high temperatures are typically needed for removing organic materials from the paste.^84,85 Recently, some sintering techniques have been developed that do not require an external heat source such as current-assisted local sintering and spark plasma sintering. Some popular “non-thermal” techniques such as flash photonic sintering are not practical for die attach technologies^56,86 because of a lack of optical access. Current assisted sintering and spark plasma sintering pass an electrical current directly through the joint area to generate heat with minimal heating in the semiconductor or ceramic. The high temperature in these techniques is localized to the joint, with the localized temperature reaching several hundred Kelvin.⁸⁷ Organic materials in the paste can easily be decomposed at these temperatures and the sintering process can be completed quickly.

For denser sintered structures, the two primary processing considerations are solid loading of the paste and controlling the densification step during sintering. Condensed nanopastes for power electronics packaging typically have solid loadings ≥60 wt%.^88,89 While having high solid loading is integral in achieving a low porosity sintered structure, die attachment procedures often apply pressure to the die to reduce the need for high temperatures and to increase the final grain density.^90,91 The general process for pressure-assisted joining techniques is as follows: (1) the nanopaste is applied to the bonding surface, (2) pre-dried at a low temperature to evaporate most of the solvent, and (3) the surface is pressed together and heated.⁷⁵ This process forces the particles closer together and facilitates multiple solid state bonding mechanisms. However, pressure-assisted techniques have risk factors related to automated processes and die integrity. Therefore, pressureless sintering has been receiving increasing attention.^92,93 Optimal strength using a pressureless technique can be achieved by using high-density pastes (>80 wt%) or sufficiently high sintering temperature (>200 °C) like Gorji et al. did with a silver–nickel composite paste.⁹⁴ A unique approach from M. Li et al. uses bimodal Ag NPs to make the nanopaste. The smaller nanoparticles (10 nm) fill the voids left by the larger particles (50 nm), providing the final densified structure with stable porosity (25.5%) and high shear strength (41.8 MPa).⁹⁵ Much progress has been made in pressureless sintering techniques from a materials approach and it appears that it will receive more attention in the future. The recent patent landscape study revealed that metal nanopaste for power electronic packaging has reached the technical transitional phase and large scale application has been under development by various stakeholders.⁹⁶

3 Nanojoining based on nanowires and carbon nanotubes

With the extensive research being conducted into nanoelectronic devices and nano-electromechanical systems (NEMS), interconnections at a nanoscale have become a necessity for extremely dense logic circuits. Bottom-up assembly for metallic one-dimensional nanostructures is one of the most efficient methods to construct nanocircuits.²⁶ In recent years, joining of individual low-dimensional nanostructures such as metal/semiconductor/ceramic nanowires, carbon nanotubes, or metal/semiconductor-filled carbon nanotubes has been successfully realized by solid-state bonding, nanosoldering or fusion nanowelding.^19,97–99

3.1 Solid state bonding

Solid state bonding includes ultrasonic welding, diffusion welding and cold welding.^100–102 Ultrasonic welding was developed to join single wall carbon nanotubes (SWCNTs) onto Ti electrodes for the development of carbon nanotube-based photovoltaic cells with high energy conversion efficiency.⁹⁹ By pressing SWCNTs against electrodes under an ultrasonic mechanical modulation, a stable low ohmic contact has been achieved between the Ti electrodes and SWCNTs resulting from the formation of covalent C–Ti bonds. As shown in Fig. 12, an SWCNT was hanging over the Ti electrodes before welding, while after welding, the ends of the SWCNT were embedded into the electrodes and the nanotube morphology was almost invisible on the electrodes. This is because during the ultrasonic welding, high-frequency ultrasonic energy softens the metal and causes plastic deformation of the metal under the clamping stress owing to the ‘acoustic softening effect’.¹⁰³ Multiwall carbon nanotube (MWCNT)–reinforced aluminum matrix composites can also be fabricated by ultrasonic welding.¹⁰⁴ The CNTs were confirmed to be embedded into the metal matrix while maintaining their multiwall structure.

Fig. 12 (a) and (b) SEM images of an individual SWCNT bridging the Ti electrodes before and after the ultrasonic nanowelding with an ultrasonic power of 0.07 W, respectively.⁹⁹

Cold welding has been used to join ultrathin gold nanowires at close to room temperature using only mechanical contact without affecting the mechanical and electrical properties of the nanowires.¹⁸ The welding process was monitored by a high-resolution transmission electron microscope (HRTEM) equipped with a scanning tunneling microscope (STM). Using a TEM-STM holder, three types of joining were performed: ‘head-to-head’, ‘side-to-side’ and ‘head-to-side’.

3.2 Nanosoldering

Recently, nanosoldering has drawn significant attention as a possible joining process in the assembly and integration of nanoelectronics devices. Interconnects by solder materials are extremely robust once formed. Moreover, due to the metallic nature of solder materials, a low contact resistance can be easily obtained, which is crucial for optimum performance of the target nanoelectronic systems. Therefore, nanosoldering has been considered as an effective alternative for assembly of nanocomponents and for interconnect formation at the nanoscale.¹⁰⁵

Y. Peng et al. reported a high spatial resolution nanowelding method to provide a controllable and reliable approach for assembling and welding metallic nanowire building blocks into complex three-dimensional structures using nanovolumes of Sn₉₉Au₁ solder.¹⁰⁶ Individual metallic nanowires were picked up by SEM nanomanipulators and assembled together. Then conductive nanocircuits were welded using gold nanowires or Sn₉₉Au₁ alloy nanowires. Once the welds formed, the joining of two nanowires exhibited high strength and excellent conductivity. The nanoscale weld resistances achieved just 20 Ω by using Sn₉₉Au₁ alloy solder.

Nanoscale solder can be electrodeposited on nanowires grown in nanoporous membranes and solder on those nanowires can be reflowed upon heating.¹⁰⁷ Fig. 13 demonstrated that two nanowires with diameter around 200 nm, each having one solder segment, can form a low-resistance ohmic solder joint. However, the solder joint breaks and results in an open circuit when a large amplitude of current is applied to the nanowire interconnect, indicating that solder joints are not effective under high current conditions.

Fig. 13 (A) SEM images of contact pads patterned on top of fused nanowires, connected by a solder joint, before (left) electrical testing and after (right) applying a current of approximately 10 mA across the pads. The high current caused a break between the wires and resulted in an open circuit. (B) A typical I–V curve that shows a low resistance ohmic contact obtained across fused nanowires. (C) A hypothetical curve showing the variation in resistance (R) per unit length (L) plotted against the “effective” contact radius for a metallic Sn-based joint. An electrical resistivity for Sn of 11 mΩ cm was used for the calculation.¹⁰⁷

Due to the high electrical and thermal resistances of CNT–CNT inter-nanotube junctions (NJs), the performance of CNT network devices is usually limited.^108,109 Through chemical vapor deposition (CVD), J. W. Do and his co-workers successfully reduced this resistance using Pd as solder metal.¹¹⁰ By passing current through devices like CNT networks and CNT crossbars, the junction resistance reduced as a consequence of self-aligned and self-limiting nanosoldering. A solution-based alternative to the CVD method was presented by the same group, which is also able to deposit nanoscale Pd selectively at the CNT junctions.¹¹¹ Fig. 14 shows SEM images of CNT networks before and after Pd nanosoldering.

Fig. 14 (a) SEM image of a CNT network after Pd deposition. Scale bar is 5 μm. (b) SEM image of a control device onto which precursor was applied without any current flow, showing no noticeable Pd particles. Scale bar is 5 μm. (c–i) Magnified SEM images of CNT network showing CNT junctions nanosoldered with Pd particles. The scale bar is 1 μm.¹¹¹

3.3 Fusion nanowelding

Fusion nanowelding processes utilize a variety of heating methods, including laser heating, joule heating, or even a resistively heated hot stage.^112–114 The joint is formed by the fusion of the molten metal at the surface by joule heating. Once the welding joint is formed, the voltage across the junction reduces slightly and then stabilizes, indicating the ohmic contact has been formed.

A. Vafaei et al. reported a procedure for joining polyol-synthesized silver nanowires in air using current-induced Joule heat welding.¹¹⁵ Photolithography is used to initiate contact to silver nanowire junctions using gold electrodes. Using a common probe station, current is driven through the electrodes and the welding process is completed using a common semiconductor analyzer. The SEM images in Fig. 15a and c shows that significant morphology change only occur at the junction, where the heat was concentrated. And the morphology of the larger nanowire underneath remains unchanged by the current flow, whereas the thinner one at the junction has changed significantly. Once the thinner nanowire has melted locally at the junction, the I–V curve became linear, indicating the ohmic contact has been formed.

Fig. 15 High-magnification images of the junction before (a), and after (c) the welding procedure was applied. (b) and (d) show the corresponding I–V curves, with the resistance decreasing from 2.4 × 10¹² Ω to 666 Ω.¹¹⁵

Light-induced plasmonic nanowelding also has been used for welding of nanowire junctions. Garnett and his co-workers use the effects of local heat generation in closely spaced metallic nanostructures to assemble metallic nanowires into large interconnected networks.¹¹⁶ The nanoscale gap between two crossed nanowires enable effective light concentration and heating at the point where the nanowires need to be joined together. Fig. 16a shows the schematic of the illumination geometry for silver nanowire junction plasmonic welding on a suspended silicon nitride membrane (blue) with a silicon wafer for structural support (not shown). The nanowire junctions naturally feature nanometer-scale gaps due to the presence of surface ligands on the nanowires. These small gaps enable extreme local heating due to the strong field concentration (red). Fig. 16b–e show TEM images before and after illumination. Evidence of twinning planes running along the length of each nanowire is visible through the junction in both directions (Fig. 16b). After plasmonic nanowelding, the twinning planes at melted junctions stopped abruptly in the bottom nanowire and only continued through the junction for the top nanowire, which indicated that the bottom nanowire recrystallized onto the top nanowire, as shown in Fig. 16c and d.

Fig. 16 (a) Schematic of the illumination geometry for silver nanowire junction; (b) TEM image of an as-made silver nanowire junction. Scale bar is 50 nm; (c) and (d) TEM images of silver nanowire junctions after optical welding. Scale bar is 50 nm; (e) lattice-resolved TEM image of the interface between the bottom nanowire and the junction, showing an abrupt change in crystal orientation. Scale bar is 5 nm.¹¹⁶

SWCNT network films on plastic substrates, such as polyethylene terephthalate (PET), polycarbonate (PC), etc., have proven to be good alternative electrodes for touch screen panels and flexible devices due to their excellent electrical and mechanical properties and their solution processability under ambient conditions. To top-coat or hybridize SWCNT films to the binder material and stabilize the CNT/binder interface, heating the films at an optimal temperature for long periods of time is required.¹¹⁷ Microwaves provided a possible method for thermal welding due to fast selective heating of CNTs on plastic substrates.¹¹⁸ J. T. Han et al. enhanced the environmental stability of transparent SWCNT network films using microwave heating to produce SWCNT film-substrate nanowelding without any chemicals.¹¹⁹ The selective heating leads to embedding the SWCNTs in the substrate, within 10 s without distortion of the plastic substrate.

3.4 Joined nanowires for applications: nanocircuits and molecular devices

1D nanowires have extensive applications in composite materials, electronic devices, optoelectronic devices, etc. However, the majority of nanowires in these applications are not metallically bonded, leading to reduced reliability. Herein, the two aspects, nanocircuits with primary bond (welding of nanowires) and molecular devices with secondary bond were both summarized in this section.

3.4.1 Nanocircuits. Transparent conductive films or electrodes are increasingly attracting attention as a result of the rapid increase in demand for touch panels and wearable electronics. Strictly, such films or electrodes are not nanocircuits. Nonetheless, they provide a framework for understanding the construction of nanocircuits. Silver and carbon materials are the two main groups used in transparent conductive films or electrodes including Ag NWs^120,121/troughs,¹²² CNTs¹²³/graphene,^124,125 etc. Ag NWs are easier to weld than CNTs using current techniques and have better conductivity and transparency making them more popular for industrial electronic applications. The welding of two individual nanowires to form a metallic bond is the key concern to gain good electrical conductivity.^26,115,126 Currently, controllable welding at a single location of two nanowires is still a challenge. Semiconductor nanowires and other semiconducting nanomaterials could be integrated into a p–n diode for field-effect transistors in applications of bio/chem-sensing. It has been demonstrated that p and n type Si nanowires could be interconnected to form a diode as shown in Fig. 17.¹²⁷ This junction could work as a probe to detect bio-molecules or record intra-cellular action potentials.¹²⁷ Plasmonic circuits can work as waveguides for optical communications.^128–130 Fig. 18a shows typical Ag NW structures with two junctions welded with femtosecond laser processing in which the morphology of the original NWs remained intact.¹³⁰ This is critical for plasmonic circuits because the plasmonic dispersion demonstrates a strong structural dependence. These welded circuits can provide strong signals as demonstrated in as-synthesized branched Ag NWs shown in Fig. 18b. The combination of semiconducting nanomaterials and Ag NW also has been studied in plasmonic circuits.^131,132 Since the semiconducting nanomaterials are difficult to weld, the introduction of metallic NWs might provide a new strategy to interconnect different waveguides together.

Fig. 17 Kinked Si nanowire p–n junction. (a) SEM images and (b) I–V curve of device.¹²⁷

Fig. 18 (a) Welded Ag NWs with femtosecond laser.¹³⁰ (b) Branched Ag NWs as a plasmon router.¹²⁹ (c–e) Optical routing circuit with SnO₂ and Ag NWs, (d) and (e) are without and with Ag NW between SnO₂ NWs.¹³¹

3.4.2 Molecular devices. Currently, Si-based transistors are smaller than 20 nm, and a combination of quantum tunneling effects, power consumption, and heat sink issues will likely prevent significant further reductions in size. Many new devices have been explored, with one potential solution to post-silicon computation lying in molecular electronics.^133–135 Molecular devices offer many unique properties in a nanoscale footpring. Usually, these devices contain metal^133,136,137 or semiconductor nanowires^138,139 and redox active molecules¹⁴⁰ that can charge or discharge as part of a binary logic platform. Fig. 19a and b show two different linear junctions with different metal nanowires and molecules. A cross junction is shown in Fig. 19c. Si, InP, GaN, etc. materials used in conventional transistors can also be used in such devices. The on/off state (Fig. 19d) and time response (Fig. 19e) of such devices suggest that they offer transistor functionality¹³⁸ and may also enable programmable logic arrays for future applications.¹⁴¹

Fig. 19 (a) A π-conjugated oligo(phenylene ethynylene) OPE junction between Au and Pd nanowires.¹³⁷ (b) A 16-mercaptohexadecanoic SAM junction between two Au nanowires.¹³⁶ (c and d) The demonstration of cobalt phthalocyanine-modified cross NW-FETs nonvolatile device where one NW is used as the gate (G) and the other NW is the active element. (e) Real-time switching of the device in (c).¹³⁸

4 Nanojoining based on 2D and 3D nanostructures

4.1 Nanojoining based on nanocones

In recent years, 3D integration for integrated circuit (IC) technologies has received growing interest since it provides great advantages such as high packing density, parallel processing, and short wire lengths. There are a variety of ways to compactly connect multiple chips using peripheral wiring technology, but the packing density of IC is in principle limited by the wire bonding technology.¹⁴²

Nanocone-arrayed structures prepared by electrodeposition methods using ethylenediamine dihydrochloride (EDA·2HCl) as modifying agent provide a new approach for reliable IC packaging.¹⁴³ Since conventional packages cannot provide enough adhesive strength between Pd preplated leadframes (Pd PPFs) and the epoxy molding compound (EMC), Ni nanocone arrays greatly increase the area of the interface between them with interlocking layers, as shown in Fig. 20. Shearing strength tests indicate that the adhesion between the EMC and nanocone-arrayed PPF is three to four times higher than that of the conventional PPF.

Fig. 20 Schematic images of a conventional Pd PPF and a nano-serrated Pd PPF.¹⁴³

With proper control of metal main salts, additives and deposition conditions (solution temperature, current density and deposition time), cone sizes can be varied from 50 nm to 1500 nm in height. W. Geng et al. developed a novel low temperature insertion bonding technology based on micro–nano cone arrays for 3D packaging.¹⁴⁴ Fig. 21 shows schematic diagram of low temperature insertion bonding technology based on micro–nano cone arrays. Hot-pressure insertion bonding was performed below the melting point of the soft solder (160 °C to 200 °C), the micro–nano cone arrays were embedded into the solder and a thin diffusion layer occurred along the interface. The shear strength test results exceed the industrial pass/failure criterion with low bonding temperature and short bonding time, providing a promising method for 3D packaging.

Fig. 21 (a) SEM image of Ni micro–nano cone arrays before bonding. (b) Hot-pressure insertion bonding.¹⁴⁴

Modifying the cone materials can also improve the bondability. For instance, Cu microcones have been used as a substitute for Ni microcones, and additional plating layers on the surface of Cu microcones also improved bondability.¹⁴⁵ F. Hu et al. coated Ni-21.4% W and Au on the surface of Cu microcones for low temperature solid state bonding.¹⁴⁶ The Ni–W(Au) alloy coating proved to be a helpful barrier layer for enhanced high-temperature corrosion resistance and retarded interfacial reactions. But the average interface shear strength of the solder and Au/Ni–W/Cu microcones is lower than that of pure Cu microcones.

4.2 Nano- and microjoining based on reactive multilayers

Reactive multilayer thin films are a relatively new form of energetic material that consist of a well-defined, heterogeneous structure with stored chemical energy. They have received intense scientific interest due to its potential application in diverse fields such as joining components, igniters, and power sources.^147–149 Self-propagating high-temperature synthesis (SHS) is a relatively novel and simple method to make certain advanced ceramic, composite and intermetallic compounds, and it has obtained considerable attention as an alternative to conventional furnace technology.¹⁵⁰ Compared to conventional joining processes, the reaction velocity of SHS can be as high as 30 m s⁻¹, and the maximum local temperature can reach 1400 °C prior to cooling to room temperature within 50 ms.¹⁵¹

4.2.1 SHS mechanism and assisted SHS reactions. Numerous studies of self-propagating high-temperature reactions were conducted before the discovery of reactive multilayer thin films and therefore form the basis of current research. In particular, the extensive work by Merzhanov et al. beginning in the 1960s demonstrated combustion synthesis of different reactive powder compacts.¹⁵² In thin film technology, the first SHS reaction was reported in 1978, where an Impulse Stimulated Crystallization (ISC) transformation occurred, inducing polycrystalline structure in an (In,Ga)Sb amorphous film.¹⁵³ Self-propagating reactions do not inherently depend on combustion with gaseous environments; they can be completed in vacuum or in an inert-gas environment. Generally, self-propagating reactions can proceed through the entire multilayer. Some of the reactions may even bypass obstructions such as particles.¹⁵⁴

The SHS reaction is driven by a reduction in atomic bond energy.¹⁵⁵ In order to be self-sustained, SHS materials must have an exothermal character and a high adiabatic reaction temperature. Once the reactions are ignited by an external source of energy, such as a small spark, a flame, or a focused laser beam, a small volume of reactants is mixed, generating heat locally. Then the generated heat spreads into neighboring domains, which causes additional mixing resulting in a self-sustained reaction wave.¹⁵⁶ Atomic diffusion occurs perpendicular to the layering with A–A and B–B bonds being exchanged for A–B bonds, as shown in Fig. 22. Thermal diffusion occurs parallel to the layering and heat conducted down the foil promotes the atomic mixing and compound formation, therefore establishing a self-propagating reaction along the reacted foil.¹⁵⁷

Fig. 22 Schematic drawing of a self-propagating reaction in a multilayer foil, showing a cross-sectional view of the atomic and thermal diffusion that enable reaction formation.¹⁵⁵

In earlier SHS studies, many intermetallic compounds were produced,¹⁵⁸ as shown in Table 1, and they demonstrated how heat released during a reaction could be used for joining dissimilar materials.^159,160 There are a number of reaction parameters that could affect SHS reactions, such as reactant particle size, thermal conductivity, ignition temperature, combustion temperature, heating and cooling rate and physical conditions of reactants.¹⁵⁸ G. M. Fritz et al. used pulses of electrical, mechanical, and thermal energy to determine the ignition thresholds of self-propagating reactions in Al/(Ni-7 wt% V) and Al/inconel multilayers.¹⁶¹ They proved that the ignition threshold was found to depend strongly on the technique used to ignite the multilayer. Both their experimental and predicted trends show that the ignition threshold rises with increasing bilayer thickness, increasing pre-reaction intermixed thickness and increasing electrode contact area.

Table 1 Some materials produced by SHS

Borides	CrB, HfB, NbB, NbB2, TiB, MoB, MoB2, WB, ZrB, VB, VB2
Carbides	TiC, ZrC, HfC, NbC, SiC, WC, TaC, Ta2C, Al4C, VC, Mo2C
Nitrides	Mg3N2, BN, AlN, SiN, Si3N4, TiN, ZrN, HfN, VN, NbN, Ta2N, TaN
Aluminides	NiAl, CoAl3, NbAl3
Silicides	TiSi3, Ti5Si3, ZrSi, Zr5Si3, MoSi2, TaSi2
Hydrides	TiH2, ZrH2, NbH2, PrH2
Intermetallics	CuAl, FeAl, NbGe, TiNi, CoTi
Cemented carbides	TiC–Ni, TiC–(Ni, Mo), WC–Co, Cr3C–(Ni, Mo)
Chalcogenides	MgS, MoS2, WS2
Binary compounds	TiB2–MoB2, TiC–WC, TiN–ZrN, WS2–NbS2
Composites	MoSi2–Al2O3, MoB–Al2O3, ZrO2–Al2O3–2Nb

---

Although some materials can be commercially produced by SHS, there still remain a number of technical problems to overcome, such as the control of product structure and the intrinsic creation of porosity. Microwave initiation of SHS reactions is one of the techniques being studied to overcome some of the inherent limitations of SHS.¹⁶² It is well known that during microwave heating, the center of a body becomes hotter than its surface.¹⁶³ Thus initiating SHS with microwaves typically caused the ignition in the center of the body and the subsequent rapid propagation of the combustion wavefront outward.¹⁶⁴ This phenomenon appears to offer the possibility of control over the progress of the combustion wavefront for the first time.¹⁶² Microwaves at 2.45 GHz have been used to ignite the SHS in a Ni and Al powder mixture to produce NiAl coatings on Ti.¹⁶⁵

In addition to microwave initiated control, centrifugal forces can also be applied to assist SHS on the composite coating of a titanium carbide aluminide–alumina–iron composite.¹⁶⁶ The centrifugal force significantly enhanced both metallurgical alloying and mechanical interlocking between different sample layers during product formation.

4.2.2 Reactive multilayers for micro- and nano-joining. Reactive nanoscale multilayer foils for joining based on SHS have been developed and partly commercialized in the last decades. Due to the high reaction rate and fast heat release, these functional foils can act as local heat sources for joining temperature-sensitive materials such as metal, ceramic and semiconductors or nano- and micro-scale components.^167–169 For joining purposes, the reaction multilayer foil is placed between two similar or dissimilar materials/components which are to be joined, along with two layers of preset or precoated solder. J. Wang et al. uses freestanding nanostructured Al/Ni multilayer foils as local heat sources to melt AuSn solder layers and thereby bond the components.¹⁷⁰ As shown in Fig. 23, the stainless-steel and Al specimens were coated by Ni and Au metallization. The joints were made using Incusil-coated Al/Ni reactive foils and AuSn solder layers under an applied pressure of 100 MPa. Their study proved that for stainless-steel and Al, the thickness of the multilayer foils much be at least 40 μm and 80 μm respectively, and the shear strengths are constant at approximately 48 and 32 MPa. Moreover, in order to form a strong joint, the AuSn solder should completely melt and fully wet the metallic samples. Similarly, W. Zhu and his coworkers developed a rapid and flexible Cu/Cu interconnection method using Al/Ni multilayer nanofilms as the heat source and Sn as inserted solder layers. The shear strength in these multilayers can reach 32 MPa, which enables the application of SHS bonding in high-power devices.¹⁵¹ The inserted solder layers can help to reduce the porosity through filling the voids and cracks in the joints and thus ensuring the strength and reliability of the resulting joints.

Fig. 23 Schematic showing the reactive joining of stainless-steel and Al shear-lap specimens.¹⁷⁰

Due to the growing demand for low temperature bonding processes in 3D integration and packaging of microelectronic or micromechanical components, typical bonding processes such as silicon direct bonding, anodic bonding, glass frit bonding or eutectic bonding are increasingly insufficient.^171–173 Therefore, reactive bonding which uses reactive multilayer foils as a local heat source has become a new bonding technique for the packaging of micro–electro–mechanical systems (MEMS).¹⁷⁴ X. Qiu uses reactive multilayer Ni/Al foils to melt solder layers and thus bond silicon wafers at room temperature.¹⁷⁵ The total thickness of the Ni/Al foils is 60 μm and the bilayer thickness is 40 nm. Two sheets of AuSn solder with thickness of 25 μm and one free-standing reactive Ni/Al foil were stacked between two silicon wafers. The bond strength of the silicon wafer proved to be larger than the strength of bulk silicon, and the leakage test in isopropanol alcohol (IPA) showed that reactive foil bonds possessed good hermeticity. In contrast with the Ni/Al systems, J. Braeuer et al. developed an integrated reactive system with the help of conventionally used processes steps, which not only reduce the stress within the multilayers but also the process costs with a total multilayer thickness of less than 5 μm.¹⁷⁴ Fig. 24 shows the process flow for processing the integrated reactive systems and applying the following reactive bonding process.

Fig. 24 Process flow of the reactive bonding procedure.¹⁷⁶

3D ordered macroporous (3DOM) materials with uniform pore size and well-defined periodic structure have shown potential as absorbents, catalysts, photonic crystals and lithium ion anodes.¹⁷⁷ W. Zhang developed a new method to form nanothermites by deposition of Al onto 3DOM Fe₂O₃ membranes.¹⁷⁸ As shown in Fig. 25, polystyrene spheres were deposited onto a substrate to obtain a 3D ordered latex template that was immersed in a Fe(NO₃)₃·9H₂O precursor solution for 10 min. After calcination at 500 °C for 5 h, a 3DOM Fe₂O₃ thin film was formed. Al was deposited onto the Fe₂O₃ membrane by thermal evaporation or magnetron sputtering. Through integration of the two exothermic peaks of the DSC measurement from 480 to 638 °C and 735 to 813 °C, heat outputs of 902 J g⁻¹ and 1929 J g⁻¹, respectively were obtained. The total heat release is 2.83 kJ g⁻¹, which provides an efficient method to produce micrometer-thick nanostructured thermite films.

Fig. 25 Schematic of the synthesis procedure for 3DOM nanothermite film.¹⁷⁸

4.3 Nanojoining for aerospace devices and space manufacturing

Material joining research in space and micro/zero gravity is relatively limited but crucial for space manufacturing. One such emergent application is for sample containment in microgravity environments. Future space missions to Mars, for example, seek to extract and store various samples from Mars and return them to Earth. The sample container design and sealing of the vessel must take the following into consideration: hermeticity, shock and vibration requirements, power requirements, and sterilization. A robust brazing material for the sample container is sought to protect the Martian samples from contamination from the environment and to contain any potential Martian microorganisms or volatile materials.^179,180 The brazed material must have the mechanical strength to sustain shock and vibrations due to rover movement on the Martian surface, and interplanetary transport. Sterilization of the container requires a temperature exceeding 500 °C to kill microbiomes and dissociate organic compounds. However, the sample inside must be shielded from these high temperatures in order to preserve possible biological samples within the container.^181,182 The required heat for brazing is delivered via induction heating built into the container.¹⁸³ Reactive multilayer materials are potential candidates for this application due to the exothermal nature of the reaction, as shown in Fig. 26. The reaction can potentially produce localized temperatures large enough for sterilization. There is a low power requirement for the brazing technique because the power delivered is limited to what is available on board the Mars rover.^184,185 Reactive multilayer nanomaterials can potentially reduce the power requirement. We are not aware of any data on reactive multilayer materials in a microgravity environment specifically. However, brazing experiments in microgravity environments have been conducted that suggest that the porosity of the brazed material will be much lower than that in normal gravity.^186,187

Fig. 26 The schematic breadboard representing the sample return container that is brazed via induction heating.¹⁸³

An additional potential application of nanobrazing is for brazing of aerospace structural components. Carbon fiber reinforced silicon carbide and other ceramic matrix composite materials are desirable alternatives for high temperature aerospace applications. Joining these composite materials to themselves and to metals remains a challenge because the materials must retain their unique thermal properties. A variety of new brazing materials and technologies are now being developed that are compatible with this material and for this application.^188–192 Nanobrazing can also potentially provide an alternative to traditional welding and brazing processes used for space station repair.^193–197 As previously mentioned, the use of reactive multilayer materials needs to be investigated in microgravity environments for these applications. Brazing materials for space craft must also take radiation and the stress of entering/exiting a planets orbit into consideration.¹⁹⁸

5 Summary and outlook

Nanojoining is currently a subject of intense scientific study. This review has summarized various nanojoining methods such as cold welding, solid state bonding, nanosoldering, and fusion nanowelding, classified from nanoparticles to nanoscale multilayer foils and nanocones. Low temperature nanosoldering by reactive nanoscale multilayer foils based on self-propagating reactions was highlighted. Since no external heat sources are required in this bonding approach, reactive bonding has become a novel and promising technology for joining temperature-sensitive materials and packaging functional nanodevices and nano-systems. Nanojoining technology based on reactive multilayers is still challenging, especially in fundamental research on dynamic processes that underlie ignition and self-sustained high-temperature reactions. Little is known about the size effect of nanoscaled multilayers on the exothermal character of the SHS reaction. Further research work is necessary to reveal details of reaction dynamics, phase transformations, heat release, the effects of cooling on phase formation and reaction wavefront instabilities as well as the bonded interface microstructure and properties. In addition, nano-manipulation is another challenging issue because nanoscopic building blocks have to be configured for most nanojoining techniques. In situ imaging and local energy delivery techniques are also crucial for practical nanojoining.

While nanojoining technology is still in its infancy, the application of nanojoining has become a strong driving force for relevant technological developments. Precise 3D printing is a typical example, in which line to line and layer to layer manufacturing involve nanojoining.¹⁹⁹ Self-assembly chemistry, especially DNA-templated assembly, has attracted much attention due to its appropriate molecular-recognition and mechanical properties. For instance, conductive silver wire connecting two gold electrodes can be prepared by self-assembly of a DNA template, the as-prepared wire was 12 μm length and 100 nm width, which was beyond the limit of conventional preparation method in microelectronic technology.²⁰⁰ Highly ordered nanostructures such as nanoparticles,²⁰¹ nanorods,²⁰² nanoclusters^203,204 have also been successfully assembled through the molecular-recognition of DNA template. The self-assembly was driven by the specific recognition of DNA molecular, thus by controlling the length, shape, and sequence of DNA, different functional nanostructures can be prepared. Self-assembly chemistry provided a potential method to fabricate nanodevices with specific properties or requirements,²⁰⁵ and also provided a promising prospect for establishing 3D ordered nanojoining. With the growing development and extensive application of nanowire-based sensors and molecular devices, nanojoining is becoming the cornerstone of nanomanufacturing critical to next generation technologies.

Acknowledgements

This work was supported in part by the Nature Science Foundation of China (NSFC) under grant number 51575016 and 51475007, and a strategic research grant (KZ20141000500, B-type) of Beijing Natural Science Foundation and a Joint Development Research and Development (JDRD) program between University of Tennessee Knoxville and Oak Ridge National Lab (ORNL). This work was also performed in part at Oak Ridge National Laboratory, operated by UT-Battelle for the U.S. Department of Energy under contract no. DE-AC05-00OR22725. B.L. acknowledges support from the Laboratory Directed Research and Development program.

References

1. M. Pumera, A. Ambrosi, A. Bonanni, E. L. K. Chng and H. L. Poh, TrAC, Trends Anal. Chem., 2010, 29, 954–965.
2. X. Lu, C. Wang and Y. Wei, Small, 2009, 5, 2349–2370.
3. Z. Zhu, L. Garcia-Gancedo, A. J. Flewitt, H. Xie, F. Moussy and W. I. Milne, Sensors, 2012, 12, 5996–6022.
4. M. Martín, P. Salazar, R. Villalonga, S. Campuzano, J. M. Pingarrón and J. L. González-Mora, J. Mater. Chem. B, 2014, 2, 739–746.
5. M. R. Kim, D. K. Lee and D.-J. Jang, Appl. Catal., B, 2011, 103, 253–260.
6. Y. Zhang, C. Ma, Y. Zhu, R. Si, Y. Cai, J. X. Wang and R. R. Adzic, Catal. Today, 2013, 202, 50–54.
7. P. Christopher, H. Xin and S. Linic, Nat. Chem., 2011, 3, 467–472.
8. M. Rycenga, C. M. Cobley, J. Zeng, W. Li, C. H. Moran, Q. Zhang, D. Qin and Y. Xia, Chem. Rev., 2011, 111, 3669–3712.
9. A. Holewinski, J.-C. Idrobo and S. Linic, Nat. Chem., 2014, 6, 828–834.
10. M. Zhu, C. Wang, D. Meng and G. Diao, J. Mater. Chem. A, 2013, 1, 2118–2125.
11. T.-J. Whang, M.-T. Hsieh and H.-H. Chen, Appl. Surf. Sci., 2012, 258, 2796–2801.
12. J. Hu, T. W. Odom and C. M. Lieber, Acc. Chem. Res., 1999, 32, 435–445.
13. S. Guo and E. Wang, Anal. Chim. Acta, 2007, 598, 181–192.
14. B. Yu, K. M. Leung, Q. Guo, W. M. Lau and J. Yang, Nanotechnology, 2011, 22, 115603.
15. Z. L. Wang, Mater. Sci. Eng., R, 2009, 64, 33–71.
16. Y. Zhou, A. Hu, M. Khan, W. Wu, B. Tam and M. Yavuz, J. Phys.: Conf. Ser., 2009, 1, 012012.
17. A. Hu, J. Guo, H. Alarifi, G. Patane, Y. Zhou, G. Compagnini and C. Xu, Appl. Phys. Lett., 2010, 97, 153117.
18. Y. Lu, J. Y. Huang, C. Wang, S. Sun and J. Lou, Nat. Nanotechnol., 2010, 5, 218–224.
19. A. Hu, P. Peng, H. Alarifi, X. Zhang, J. Guo, Y. Zhou and W. Duley, J. Laser Appl., 2012, 24, 042001.
20. C. Burda, X. Chen, R. Narayanan and M. A. El-Sayed, Chem. Rev., 2005, 105, 1025–1102.
21. L.-M. Lacroix, F. Delpech, C. Nayral, S. Lachaize and B. Chaudret, Interface Focus, 2013, 3, 20120103.
22. C. Yang, C. P. Wong and M. M. Yuen, J. Mater. Chem. C, 2013, 1, 4052–4069.
23. K. Lu, Int. Mater. Rev., 2008, 53, 21–38.
24. P. Peng, A. Hu, A. P. Gerlich, G. Zou, L. Liu and Y. N. Zhou, ACS Appl. Mater. Interfaces, 2015, 7, 12597–12618.
25. Y. Zhou and A. Hu, Open Surf. Sci. J., 2011, 3, 32–41.
26. P. Peng, L. Liu, A. P. Gerlich, A. Hu and Y. N. Zhou, Part. Part. Syst. Charact., 2013, 30, 420–426.
27. D. E. Sanders and A. E. DePristo, Surf. Sci., 1992, 260, 116–128.
28. N. A. Alcantar, C. Park, J.-M. Pan and J. N. Israelachvili, Acta Mater., 2003, 51, 31–47.
29. A. Hu, M. Rybachuk, Q.-B. Lu and W. Duley, Appl. Phys. Lett., 2007, 91, 131906.
30. D. V. Wagle and G. A. Baker, Mater. Horiz., 2015, 2, 157–167.
31. B. Ingham, T. H. Lim, C. J. Dotzler, A. Henning, M. F. Toney and R. D. Tilley, Chem. Mater., 2011, 23, 3312–3317.
32. C. Liu, Y.-J. Li, S.-G. Sun and E. S. Yeung, Chem. Commun., 2011, 47, 4481–4483.
33. Y. Wang, W. Liang and C. Geng, Nanoscale Res. Lett., 2009, 4, 684–688.
34. X. Huang, W. Dong, G. Wang, M. Yang, T. Li, Y. Feng and X. Zhang, J. Colloid Interface Sci., 2011, 359, 40–46.
35. A. Hu, G. Deng, S. Courvoisier, O. Reshef, C. Evans, E. Mazur, Y. Zhou, SPIE NanoScience + Engineering, International Society for Optics and Photonics, 2013, pp. 880907–880907.
36. A. Hu, Y. Zhou and W. Duley, Significance, 2011, 9, 10.
37. A. Hu, S. Panda, M. Khan and Y. Zhou, Chin. J. Lasers, 2009, 36, 3149–3159.
38. D. Von der Linde, K. Sokolowski-Tinten and J. Bialkowski, Appl. Surf. Sci., 1997, 109, 1–10.
39. S. J. Kim and D.-J. Jang, Appl. Phys. Lett., 2005, 86, 033112.
40. H. Huang, L. Liu, P. Peng, A. Hu, W. Duley and Y. Zhou, J. Appl. Phys., 2012, 112, 123519.
41. K. A. Willets and R. P. Van Duyne, Annu. Rev. Phys. Chem., 2007, 58, 267–297.
42. P. Mulvaney, Langmuir, 1996, 12, 788–800.
43. B. Rothenhäusler and W. Knoll, Nature, 1988, 332, 615–617.
44. M. L. Brongersma and P. G. Kik, Surface Plasmon Nanophotonics, Springer, 2007.
45. S. Pillai, K. Catchpole, T. Trupke and M. Green, J. Appl. Phys., 2007, 101, 093105.
46. J. Homola, S. S. Yee and G. Gauglitz, Sens. Actuators, B, 1999, 54, 3–15.
47. J. Homola, Anal. Bioanal. Chem., 2003, 377, 528–539.
48. B. Liedberg, C. Nylander and I. Lunström, Sens. Actuators, 1983, 4, 299–304.
49. S. Link and M. A. El-Sayed, J. Phys. Chem. B, 1999, 103, 4212–4217.
50. J. Mock, M. Barbic, D. Smith, D. Schultz and S. Schultz, J. Chem. Phys., 2002, 116, 6755–6759.
51. X.-Y. Zhang, A. Hu, T. Zhang, W. Lei, X.-J. Xue, Y. Zhou and W. W. Duley, ACS Nano, 2011, 5, 9082–9092.
52. L. O. Herrmann, V. K. Valev, C. Tserkezis, J. S. Barnard, S. Kasera, O. A. Scherman, J. Aizpurua and J. J. Baumberg, Nat. Commun., 2014, 5, 4568–4573.
53. H. Huang, L. Liu, P. Peng, A. Hu, W. Duley and Y. Zhou, J. Appl. Phys., 2012, 112, 123519.
54. Y. S. Rim, S. H. Bae, H. Chen, N. De Marco and Y. Yang, Adv. Mater., 2016, 28, 4415–4440.
55. M. Magliulo, M. Mulla, M. Singh, E. Macchia, A. Tiwari, L. Torsi and K. Manoli, J. Mater. Chem. C, 2015, 3, 12347–12363.
56. R.-Z. Li, A. Hu, T. Zhang and K. D. Oakes, ACS Appl. Mater. Interfaces, 2014, 6, 21721–21729.
57. A. H. Alshehri, M. Jakubowska, A. Młożniak, M. Horaczek, D. Rudka, C. Free and J. D. Carey, ACS Appl. Mater. Interfaces, 2012, 4, 7007–7010.
58. K. Fukuda, T. Sekine, D. Kumaki and S. Tokito, ACS Appl. Mater. Interfaces, 2013, 5, 3916–3920.
59. S.-P. Chen, Z.-K. Kao, J.-L. Lin and Y.-C. Liao, ACS Appl. Mater. Interfaces, 2012, 4, 7064–7068.
60. A. Mahajan, C. D. Frisbie and L. F. Francis, ACS Appl. Mater. Interfaces, 2013, 5, 4856–4864.
61. S. Jeong, K. Woo, D. Kim, S. Lim, J. S. Kim, H. Shin, Y. Xia and J. Moon, Adv. Funct. Mater., 2008, 18, 679–686.
62. R.-Z. Li, R. Peng, K. Kihm, S. Bai, D. Bridges, U. Tumuluri, Z. Wu, T. Zhang, G. Compagnini and Z. Feng, Energy Environ. Sci., 2016, 9, 1458–1467.
63. L. Schade, S. Franzka, M. Mathieu, M. M. Biener, J. R. Biener and N. Hartmann, Langmuir, 2014, 30, 7190–7197.
64. H. Zhang, L. M. Tolbert and B. Ozpineci, IEEE Trans. Ind. Appl., 2011, 47, 912–921.
65. K. Rajashekara, IEEE Journal of Emerging and Selected Topics in Power Electronics, 2013, 1, 3–10.
66. B. Whitaker, A. Barkley, Z. Cole, B. Passmore, D. Martin, T. R. McNutt, A. B. Lostetter, J. S. Lee and K. Shiozaki, IEEE Trans. Power Electron., 2014, 29, 2606–2617.
67. E. Higurashi, K. Okumura, K. Nakasuji and T. Suga, Jpn. J. Appl. Phys., 2015, 54, 030207.
68. B. L. Dokić and B. Blanuša, Power Electronics, Springer, 2015, pp. 559–592.
69. I. Colak, E. Kabalci, G. Fulli and S. Lazarou, Renewable Sustainable Energy Rev., 2015, 47, 562–579.
70. R. Khazaka, D. Henry and R. Hanna, IEEE Trans. Power Electron., 2015, 30, 2456–2464.
71. P. Ning, D. Zhang, R. Lai, D. Jiang, F. Wang, D. Boroyevich, R. Burgos, K. Karimi, V. D. Immanuel and E. V. Solodovnik, IEEE Industrial Electronics Magazine, 2013, 7, 6–17.
72. H.-M. Qu, X.-H. Yang, Q. Zheng, X.-T. Wang and Q. Chen, IEICE Electron. Express, 2014, 11, 20140965.
73. A. L. Moore and L. Shi, Mater. Today, 2014, 17, 163–174.
74. M. R. Panão, A. M. Correia and A. L. Moreira, Appl. Therm. Eng., 2012, 37, 293–301.
75. S. Wang, M. Li, H. Ji and C. Wang, Scr. Mater., 2013, 69, 789–792.
76. V. Wei, M. Huang, R. Lai and R. Persons, Microsystems, Packaging, Assembly and Circuits Technology Conference (IMPACT), 2014, pp. 141–145.
77. M. Fujino, K. Abe and T. Suga, Electronics Packaging (ICEP), 2014, pp. 648–651.
78. I. Ohnuma, M. Miyashita, K. Anzai, X. Liu, H. Ohtani, R. Kainuma and K. Ishida, J. Electron. Mater., 2000, 29, 1137–1144.
79. Y. Jianfeng, Z. Guisheng, H. Anming and Y. N. Zhou, J. Mater. Chem., 2011, 21, 15981–15986.
80. K. S. Tan and K. Y. Cheong, IEEE Trans. Compon., Packag., Manuf. Technol., 2014, 4, 8–15.
81. K. S. Siow, J. Alloys Compd., 2012, 514, 6–19.
82. Y. Yan, Y. Guan, X. Chen and G.-Q. Lu, J. Electron. Packag., 2013, 135, 041003.
83. H. Ikeda, S. Sekine, R. Kimura, K. Shimokawa, K. Okada, H. Shindo, T. Ooi and M. Nagata, Electronics Packaging and iMAPS All Asia Conference (ICEP-IACC), 2015, pp. 482–485.
84. R.-Z. Li, A. Hu, D. Bridges, T. Zhang, K. D. Oakes, R. Peng, U. Tumuluri, Z. Wu and Z. Feng, Nanoscale, 2015, 7, 7368–7377.
85. A. Hirose, N. Takeda, Y. Konaka, H. Tatsumi, Y. Akada, T. Ogura, E. Ide and T. Morita, Mater. Sci. Forum, 2012, 2962–2967.
86. M. Hösel and F. C. Krebs, J. Mater. Chem., 2012, 22, 15683–15688.
87. Y. Mei, Y. Cao, G. Chen, X. Li, G.-Q. Lu and X. Chen, IEEE Trans. Device Mater. Reliab., 2013, 13, 258–265.
88. Z. Liu, H. Zhou, L. Piao and S.-H. Kim, Adv. Sci. Lett., 2012, 10, 217–220.
89. K.-S. Kim, Y. Kim and S.-B. Jung, Nanoscale Res. Lett., 2012, 7, 1–6.
90. K. Suganuma, S. Sakamoto, N. Kagami, D. Wakuda, K.-S. Kim and M. Nogi, Microelectron. Reliab., 2012, 52, 375–380.
91. W. Sabbah, S. Azzopardi, C. Buttay, R. Meuret and E. Woirgard, Microelectron. Reliab., 2013, 53, 1617–1621.
92. S. Fu, Y. Mei, G.-Q. Lu, X. Li, G. Chen and X. Chen, Mater. Lett., 2014, 128, 42–45.
93. J. Yan, G. Zou, A.-p. Wu, J. Ren, J. Yan, A. Hu and Y. Zhou, Scr. Mater., 2012, 66, 582–585.
94. A. Gorji, A. Simchi and A. Kokabi, Ceram. Int., 2015, 41, 1815–1822.
95. M. Li, Y. Xiao, Z. Zhang and J. Yu, ACS Appl. Mater. Interfaces, 2015, 7, 9157–9168.
96. K. S. Siow and Y. Lin, J. Electron. Packag., 2016, 138, 020804–020817.
97. E. C. Garnett, W. Cai, J. J. Cha, F. Mahmood, S. T. Connor, M. G. Christoforo, Y. Cui, M. D. McGehee and M. L. Brongersma, Nat. Mater., 2012, 11, 241–249.
98. J. A. Rodríguez-Manzo, F. Banhart, M. Terrones, H. Terrones, N. Grobert, P. M. Ajayan, B. G. Sumpter, V. Meunier, M. Wang and Y. Bando, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 4591–4595.
99. C. Chen, L. Yan, E. S.-W. Kong and Y. Zhang, Nanotechnology, 2006, 17, 2192.
100. C. M. Schauerman, J. Alvarenga, J. Staub, M. W. Forney, R. Foringer and B. J. Landi, Adv. Eng. Mater., 2015, 17, 76–83.
101. Z. Hu, J. Zhang, Y. Yan, J. Yan and T. Sun, J. Mech. Sci. Tech., 2013, 27, 43–46.
102. T. Ishida, K. Kakushima and H. Fujita, J. Appl. Phys., 2011, 110, 104310.
103. S. Kang, P. Williams, T. McLaren and Y. Lee, Mater. Chem. Phys., 1995, 42, 31–37.
104. Z. Zhu, H. Mi, X. Wang, C. Xie and Z. Wu, Metall. Mater. Trans. B, 2012, 43, 1615–1621.
105. Q. Cui, F. Gao, S. Mukherjee and Z. Gu, Small, 2009, 5, 1246–1257.
106. Y. Peng, A. Cullis and B. Inkson, J. Phys.: Conf. Ser., 2010, 012083.
107. Z. Gu, H. Ye, D. Smirnova, D. Small and D. H. Gracias, Small, 2006, 2, 225–229.
108. A. V. Kyrylyuk, M. C. Hermant, T. Schilling, B. Klumperman, C. E. Koning and P. Van der Schoot, Nat. Nanotechnol., 2011, 6, 364–369.
109. J. Yang, S. Waltermire, Y. Chen, A. A. Zinn, T. T. Xu and D. Li, Appl. Phys. Lett., 2010, 96, 023109.
110. J.-W. Do, D. Estrada, X. Xie, N. N. Chang, J. Mallek, G. S. Girolami, J. A. Rogers, E. Pop and J. W. Lyding, Nano Lett., 2013, 13, 5844–5850.
111. J.-W. Do, N. N. Chang, D. Estrada, F. Lian, H. Cha, X. J. Duan, R. T. Haasch, E. Pop, G. S. Girolami and J. W. Lyding, ACS Nano, 2015, 9, 4806–4813.
112. M. Son, S. Jeong and D.-J. Jang, J. Phys. Chem. C, 2014, 118, 5961–5967.
113. H. Tohmyoh and S. Fukui, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 80, 155403.
114. Y. Wu and P. Yang, Adv. Mater., 2001, 13, 520–523.
115. A. Vafaei, A. Hu and I. A. Goldthorpe, Nano-Micro Lett., 2014, 6, 293–300.
116. E. C. Garnett, W. Cai, J. J. Cha, F. Mahmood, S. T. Connor, M. G. Christoforo, Y. Cui, M. D. McGehee and M. L. Brongersma, Nat. Mater., 2012, 11, 241–249.
117. J. T. Han, J. S. Kim, H. D. Jeong, H. J. Jeong, S. Y. Jeong and G.-W. Lee, ACS Nano, 2010, 4, 4551–4558.
118. T. Wu, Y. Pan, E. Liu and L. Li, J. Appl. Polym. Sci., 2012, 126, 283–289.
119. J. T. Han, D. Kim, J. S. Kim, S. K. Seol, S. Y. Jeong, H. J. Jeong, W. S. Chang, G.-W. Lee and S. Jung, Appl. Phys. Lett., 2012, 100, 163120.
120. C.-H. Liu and X. Yu, Nanoscale Res. Lett., 2011, 6, 75.
121. L. Hu, H. S. Kim, J.-Y. Lee, P. Peumans and Y. Cui, ACS Nano, 2010, 4, 2955–2963.
122. H. Wu, D. Kong, Z. Ruan, P.-C. Hsu, S. Wang, Z. Yu, T. J. Carney, L. Hu, S. Fan and Y. Cui, Nat. Nanotechnol., 2013, 8, 421–425.
123. D. Jiang, N. Wang, M. Edwards, W. Mu, A. Nylander, Y. Fu, K. Jeppson and J. Liu, Small, 2016, 12, 1521–1526.
124. X. Wang, L. Zhi and K. Müllen, Nano Lett., 2008, 8, 323–327.
125. W. H. Lee, J. Park, S. H. Sim, S. B. Jo, K. S. Kim, B. H. Hong and K. Cho, Adv. Mater., 2011, 23, 1752–1756.
126. F. Gao and Z. Gu, Nanotechnology, 2010, 21, 115604.
127. Z. Jiang, Q. Qing, P. Xie, R. Gao and C. M. Lieber, Nano Lett., 2012, 12, 1711–1716.
128. A. L. Pyayt, B. Wiley, Y. Xia, A. Chen and L. Dalton, Nat. Nanotechnol., 2008, 3, 660–665.
129. Y. Fang, Z. Li, Y. Huang, S. Zhang, P. Nordlander, N. J. Halas and H. Xu, Nano Lett., 2010, 10, 1950–1954.
130. L. Lin, L. Liu, P. Peng, G. Zou, W. W. Duley and Y. N. Zhou, Nanotechnology, 2016, 27, 125201.
131. R. Yan, P. Pausauskie, J. Huang and P. Yang, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 21045–21050.
132. O. Benson, Nature, 2011, 480, 193–199.
133. M. Austin and S. Y. Chou, J. Vac. Sci. Technol., B: Microelectron. Nanometer Struct., 2002, 20, 665–667.
134. C. Joachim, J. Gimzewski and A. Aviram, Nature, 2000, 408, 541–548.
135. N. I. Kovtyukhova and T. E. Mallouk, Chem.–Eur. J., 2002, 8, 4354–4363.
136. J. K. Mbindyo, T. E. Mallouk, J. B. Mattzela, I. Kratochvilova, B. Razavi, T. N. Jackson and T. S. Mayer, J. Am. Chem. Soc., 2002, 124, 4020–4026.
137. L. Cai, H. Skulason, J. Kushmerick, S. Pollack, J. Naciri, R. Shashidhar, D. Allara, T. Mallouk and T. Mayer, J. Phys. Chem. B, 2004, 108, 2827–2832.
138. X. Duan, Y. Huang and C. M. Lieber, Nano Lett., 2002, 2, 487–490.
139. E. Baek, S. Pregl, M. Shaygan, L. Römhildt, W. M. Weber, T. Mikolajick, D. A. Ryndyk, L. Baraban and G. Cuniberti, Nano Res., 2015, 8, 1229–1240.
140. H. Zhu, S. J. Pookpanratana, J. E. Bonevich, S. N. Natoli, C. A. Hacker, T. Ren, J. S. Suehle, C. A. Richter and Q. Li, ACS Appl. Mater. Interfaces, 2015, 7, 27306–27313.
141. Y. Chen, G.-Y. Jung, D. A. Ohlberg, X. Li, D. R. Stewart, J. O. Jeppesen, K. A. Nielsen, J. F. Stoddart and R. S. Williams, Nanotechnology, 2003, 14, 462.
142. L. Junhui, W. Ruishan, H. Lei, W. Fuliang and L. Zhili, Solid State Sci., 2011, 13, 72–76.
143. T. Hang, H. Ling, Z. Xiu, M. Li and D. Mao, J. Electron. Mater., 2007, 36, 1594–1598.
144. W. Geng, Z. Chen, A. Hu and M. Li, Mater. Lett., 2012, 78, 72–74.
145. F. Hu, S. Yang, H. Wang, A. Hu and M. Li, J. Electron. Mater., 2015, 44, 4516–4524.
146. F. Hu, H. Wang, S. Yang, A. Hu and M. Li, Appl. Surf. Sci., 2015, 353, 774–780.
147. C. Suryanarayana, J. J. Moore and R. P. Radtke, Adv. Mater. Process., 2001, 159, 29–31.
148. C. J. Morris, P. Wilkins, C. May, E. Zakar and T. P. Weihs, Thin Solid Films, 2011, 520, 1645–1650.
149. M. Ding, F. Krieger, J. Swank, J. Poret, C. McMullan, G. Chen, Proc. 43rd Power Sources Conference, 2008, p. 615.
150. P. Mossino, Ceram. Int., 2004, 30, 311–332.
151. W. Zhu, F. Wu, B. Wang, E. Hou, P. Wang, C. Liu and W. Xia, Microelectron. Eng., 2014, 128, 24–30.
152. E. I. Maksimov and A. Merzhanov, Fiz. Goreniya Vzryva, 1966, 1, 47–58.
153. C. Wickersham, G. Bajor and J. Greene, Solid State Commun., 1978, 27, 17–20.
154. J. Floro, J. Vac. Sci. Technol., A, 1986, 4, 631–636.
155. J. Wang, E. Besnoin, A. Duckham, S. Spey, M. Reiss, O. Knio and T. Weihs, J. Appl. Phys., 2004, 95, 248–256.
156. D. Adams, Thin Solid Films, 2015, 576, 98–128.
157. L. Zhang and Z. Wang, Exp. Therm. Fluid Sci., 2008, 32, 1255–1263.
158. J. J. Moore and H. Feng, Prog. Mater. Sci., 1995, 39, 243–273.
159. Z. A. Munir and J. Holt, Combustion and plasma synthesis of high-temperature materials, VCH, New York, etc., 1990.
160. V. Shcherbakov and A. Shteinberg, Int. J. Self-Propag. High-Temp. Synth., 1993, 2, 357–369.
161. G. M. Fritz, S. J. Spey Jr, M. D. Grapes and T. P. Weihs, J. Appl. Phys., 2013, 113, 014901.
162. J.-H. Peng, J. Binner and S. Bradshaw, Mater. Sci. Technol., 2002, 18, 1419–1427.
163. J. Binner and T. Cross, J. Hard Mater., 1993, 4, 177–185.
164. R. Dalton, I. Ahmad and D. Clark, Combustion and plasma synthesis of high-temperature materials, VCH, New York, 1990.
165. G. Cammarota, A. Casagrande, G. Poli and P. Veronesi, Surf. Coat. Technol., 2009, 203, 1429–1437.
166. R. Mahmoodian, M. Hassan, M. Hamdi, R. Yahya and R. Rahbari, Composites, Part B, 2014, 59, 279–284.
167. S. Simões, F. Viana, A. S. Ramos, M. T. Vieira and M. F. Vieira, Microsc. Microanal., 2015, 21, 132–139.
168. A. Duckham, S. Spey, J. Wang, M. Reiss, T. Weihs, E. Besnoin and O. Knio, J. Appl. Phys., 2004, 96, 2336–2342.
169. T. Weihs, O. Knio, SPIE proceedings series, 2003, pp. 235–239.
170. J. Wang, E. Besnoin, O. Knio and T. Weihs, J. Appl. Phys., 2005, 97, 114307.
171. F. Niklaus, G. Stemme, J.-Q. Lu and R. Gutmann, J. Appl. Phys., 2006, 99, 031101.
172. Q. Wang, S.-H. Choa, W. Kim, J. Hwang, S. Ham and C. Moon, J. Electron. Mater., 2006, 35, 425–432.
173. M. Alexe and U. Gösele, Wafer Bonding: Applications and Technology, Springer Science & Business Media, 2013.
174. J. Braeuer, J. Besser, M. Wiemer and T. Gessner, Sens. Actuators, A, 2012, 188, 212–219.
175. X. Qiu and J. Wang, Sens. Actuators, A, 2008, 141, 476–481.
176. O. Politano, F. Baras, A. Mukasyan, S. Vadchenko and A. Rogachev, Surf. Coat. Technol., 2013, 215, 485–492.
177. Y. Chu and Q. Pan, ACS Appl. Mater. Interfaces, 2012, 4, 2420–2425.
178. W. Zhang, B. Yin, R. Shen, J. Ye, J. A. Thomas and Y. Chao, ACS Appl. Mater. Interfaces, 2013, 5, 239–242.
179. P. Younse, T. De Alwis, P. Backes and A. Trebi-Ollennu, Aerospace Conference, 2012, pp. 1–11.
180. MEPAG Next Decade Science Analysis Group, 2008.
181. X. Bao, M. Badescu and Y. Bar-Cohen, SPIE Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring, 2015, p. 943527.
182. D. J. Des Marais, J. A. Nuth III, L. J. Allamandola, A. P. Boss, J. D. Farmer, T. M. Hoehler, B. M. Jakosky, V. S. Meadows, A. Pohorille and B. Runnegar, Astrobiology, 2008, 8, 715–730.
183. Y. Bar-Cohen, J. Wu, A. K. Olorunsola, T. P. Rivellini, S. Bley, J. Wincentsen and R. Gershman, Smart Struct. Mater., 2005, 261–266.
184. H.-P. Xiong, B. Chen, Y. Pan, H.-S. Zhao and L. Ye, J. Eur. Ceram. Soc., 2014, 34, 1481–1486.
185. P. Backes, P. Younse and A. Ganino, Aerospace Conference, 2013, pp. 1–10.
186. K. Frieler, N. Phlippovich, R. Stickler and W. Bathke, Adv. Space Res., 1981, 1, 117–120.
187. Y. Flom, Weld. J., 2005, 84, 25–29.
188. X. Hernandez, C. Jiménez, K. Mergia, P. Yialouris, S. Messoloras, V. Liedtke, C. Wilhelmi and J. Barcena, J. Mater. Eng. Perform., 2014, 23, 3069–3076.
189. E. J. T. Pialago, O. K. Kwon and C. W. Park, Ceram. Int., 2015, 41, 6764–6775.
190. W. Guo, L. Wang, Y. Zhu and P. K. Chu, Ceram. Int., 2015, 41, 7021–7027.
191. M. Halbig, R. Asthana and M. Singh, Ceram. Int., 2015, 41, 2140–2149.
192. R. Asthana and M. Singh, J. Eur. Ceram. Soc., 2008, 28, 617–631.
193. D. W. Dickinson, H. Babel, H. Conaway and W. Hooper, Technology for Space Station Evolution, 1990, vol. 5, pp. 207–267.
194. R. Anderson and P. Fuson, Seventh Annual Workshop on Space Operations Applications and Research, 1993, vol. 2, pp. 605–610.
195. M. E. Wilson, H. E. Cole, T. Rector, J. Steele and J. Varsik, 41st International Conference on Environmental Systems, 2011, p. 5006.
196. N. Werkheiser, Annual International Mars Society Convention, Washington DC, 13–16 Aug 2015.
197. T. Ohji, Weld. Int., 2006, 20, 355–360.
198. P. Craven, M. P. SanSoucie, B. Tomboulian, J. Rogers and R. Hyers, National Space and Missile Materials Symposium, Huntsville AL, 23–26 June 2014.
199. A. Hu, R. Li, D. Bridges, W. Zhou, S. Bai, D. Ma and P. Peng, J. Laser Appl., 2016, 28, 022602.
200. E. Braun, Y. Eichen and U. Sivan, et al., Nature, 1998, 391, 775–778.
201. M. N. O'Brien, M. R. Jones and B. Lee, Nat. Mater., 2015, 14, 833–839.
202. V. T. Milam, A. L. Hiddessen and J. C. Crocker, Langmuir, 2003, 19, 10317–10323.
203. M. M. Maye, D. Nykypanchuk and M. Cuisinier, Nat. Mater., 2009, 8, 388–391.
204. Q. Yao, X. Yuan and Y. Yu, J. Am. Chem. Soc., 2015, 137, 2128–2136.
205. M. M. Maye, M. T. Kumara and D. Nykypanchuk, Nat. Nanotechnol., 2010, 5, 116–120.

---