To produce Au nanocondensates, Au (99.99% pure, 0.3 mm in thickness) foil immersed in de-ionized (DI) water within a glass beaker was subjected to energetic Nd-YAG-laser (Lotis, 1,064 nm in wavelength, beam mode: TEM00) pulse irradiation for up to 3,000 pulses inside an ablation chamber under the laser parameters compiled in Table 1. A relatively high power density of 1.4 × 10¹² W/cm²(i.e., pulsed energy 650 mJ/pulse; pulse duration 16 ns; beam size 0.03 mm²; fluence 2.2 kJ/cm²; frequency 10 Hz under Q-switch mode) caused a larger yield of atom clusters as a colloidal solution. The reddish solution was then settled in capped vial for up to 40 days in desiccators or open air in order to study the influence of moisture on the assembly and migration of the resultant MWTs.

FRandQSdenote free run and Q-switch mode, respectively.MTPdenotes multiply twinned particle,Ccluster,fccface-centered cubic, andTtubes

^* Based on SAED lattice parameter of the fcc/MTP nanocondensates and the Birch-Murnaghan equation of state of fcc gold with relevant bulk modulus and its pressure derivative (cf. text)

The optical absorbance of the as-deposited nanocondensates and further developed MWTs in solution with specified dwelling times was acquired by a UV–vis spectrophotometer (U-3900H, Hitachi) operating at an instrumental resolution of 0.1 nm in the range of 300–800 nm. The powders recovered from such samples were dried for microstructure observations using optical polarized microscopy and scanning electron microscopy (SEM, JEOL JSM-6700F, 10 kV, 10 μA). The crystal structure of the MWTs was determined by X-ray diffraction (XRD, Siemens D5000, Cu Kα, 40 kV, 30 mA, at 0.05° and 3 s per step from 2θ angle for 30° up to 90°). The d-spacings measured from XRD trace were used for least-squares refinement of the lattice parameters with an error ±0.0001 nm using bulk gold reflections as a standard. Field emission transmission electron microscopy (TEM, FEI Tecnai G2 F20 at 200 kV) with selected area electron diffraction (SAED), and point-count energy dispersive X-ray (EDX) analysis at a beam size of 1 nm was used to study the composition and phases of the tubular walls. Lattice imaging coupled with two-dimensional (2-D) Fourier transform and inverse transform were used to study the rolling planes of the MWT and their partial epitaxy relationship with the associated MTP/fcc nanoparticles. Powdery sample mixed with KBr was studied by FTIR (Bruker 66v/S) for the extent of OH⁻signature on MWT. As for the SERS effect, the powdery MWTs settled on a vitreous SiO₂substrate were studied by micro-Raman in a backscattering geometry by a Jobin Yvon T64000 system working in the triple-subtractive mode.

The colloidal Au solution as formed via PLAL is reddish under the naked eye (Fig. 1). The corresponding surface plasmon absorbance near 525 nm is similar to that of the Au nanoparticles and clusters in aqueous or solvent-protected solution 15 16. Room-temperature aging of the solution for up to 40 days caused progressive lowering of the absorbance peak because more opaque materials were deposited at the bottom and even along the humidified wall of the capped bottle especially when desiccated (Fig. 1). The opaque materials are in fact mesomorphic lamellae and MWTs as indicated by the following electron microscopic observations.

The powdery materials retrieved from the colloidal Au solution having been processed by PLAL were agglomerated as nanoparticle chain aggregate (NCA), as shown by the TEM bright field image (BFI) in Fig. 2a. The individual atom clusters and MTP/fcc particles within the NCA were imaged (Fig. 2b, c) using the low-angle scattering halo (Fig. 2d) and the nearly superimposed strong {111}/{200} diffractions of the MTP/fcc (Fig. 2d), respectively. The diffraction intensity profile (Fig. 2e) further showed that the low-angle scattering of the atom clusters falls in the range of 1.46–1.78 Q⁻¹ full width at half-height, analogous to that of the polished Beilby layers 17 or Au foil on the verge of melting 18 . EDX analysis on the nanocondensates (Fig. 2f) showed Au peaks, with the remaining peaks from the supporting carbon-coated collodion film and Cu grid. This result indicates that oxidation or contamination of the Au nanocondensates is negligible.

After aging for 10 days in water, the Au atom clusters were self-assembled as corrugated lamellae as shown by the TEM BFI in Fig. 3a. The SAED pattern (Fig. 3b) and corresponding diffraction intensity profile (Fig. 3c) showed a broad diffraction near 1.63 nm⁻¹for atom clusters, a sharp diffraction at 2.84 nm⁻¹for Au nanotubes and broad diffractions near 4.21, 4.65, 6.96, 7.96 nm⁻¹for (111), (200), (220), and (311) of MTP/fcc nanoparticles, respectively. Lattice images further indicated the lamellae are bifurcated and entangled (Fig. 3d), yet having lamellar layer nearly parallel to (200) and/or (111) of MTP (Fig. 3e, f), i.e., following partial epitaxy relationship.

The final products after aging for 40 days in water are tubular Au wires which are about 2–5 μm in width and several hundreds of micrometers in length as indicated by SEM observation (Fig. 4a). The microtubes showed extensive branches and entanglement having walls decorated with Au nanoparticles when imaged at a higher magnification (Fig. 4b). TEM further showed that the microtubes are either uncapped or capped (Fig. 4c, d) having randomly oriented fcc-type domains in the wall, and a hollow core of such tubes was manifested by a relative weak diffraction in the center of the tubes. XRD trace (Fig. 4e) and the structure refinement of the d-spacings further indicated that the fcc-type wall newly assembled from atom clusters has a rather relaxed lattice parameter (a = 0.4075 nm) close to the ambient value (a = 0.4079 nm, JCPDS file 04-0784). By contrast, there is a much smaller SAED lattice parameter (0.402 ± 0.002 nm) for the as-formed fcc nanocondensates than the relaxed fcc-type wall of the microtubes (a = 0.408 ± 0.002 nm) implying a significant internal stress as discussed later.

Further TEM scrutiny revealed a typical nanotube with a varied diameter ranging from ~100 nm on the one end to less than 10 nm on the other end (Fig. 5a). The BFIs magnified from the specified regions of the nanotube showed encapsulated fcc and MTP nanoparticles which have much stronger diffraction contrast than the lamellar wall (Fig. 5b, c). The identity of lamellae within the corrugated wall of the MTP/fcc encapsulated nanotube was confirmed by high-resolution TEM image (Fig. 5d) and corresponding SAED pattern (Fig. 5e) coupled with intensity profile (Fig. 5f). Point-count EDX analysis on the lamellar wall with weak Bragg diffraction contrast, i.e., immune from fcc/MTP nanoparticles, showed Au counts without appreciable oxygen and impurities.

The hollow core and optically isotropic wall of the Au microtubes were manifested by optical polarized microscopic observation under a single polarizer and crossed polarizers, respectively (not shown). The wall of such microtubes showed significant light reflection despite the dark appearance of the sediments under the naked eye. The absence of metallic luster for the microtubes can be attributed to the surface roughness and defects of the wall.

The Raman shift of the as-formed Au condensates and subsequently assembled microtubes on a substrate of vitreous silica showed a significant SERS effect (Fig. 6). The vitreous silica alone is characterized by the vibrational D₁ and D₂ Raman bands at 488 and 599 cm⁻¹ (Fig. 6a), which have been assigned to small planar ring 19 and/or paracrystalline cluster interface 20 . When such a substrate was overlaid with the as-formed Au condensates (blue trace in Fig. 6b) or subsequently assembled Au microtubes (red trace in Fig. 6b), both D₁, D₂, and minor unidentified bands became one to two-orders-of magnitude higher. In both cases, the D₁ mode shift significantly because of the selection rules 21 . It should be noted that the SERS effect is more pronounced to have a higher signal to noise ratio for the assembled Au microtubes than the as-formed Au condensates. The OH⁻ signature of the microtubes was indicated by a typical FTIR spectrum showing a significant peak near 3,432 cm⁻¹ (Fig. 6c) which can be assigned as OH stretching vibration 22 .

In comparison with that fabricated by PLA 9 , The present Au condensates via PLAL have much more atom clusters than MTP and fcc. This can be attributed to an order-of-magnitude higher heating-cooling rate (ca. 1 × 10¹⁰ K/s) for PLAL 23 than PLA process and hence a pressure effect to form smaller sized nanocondensates. Base on Birch–Murnaghan equation of state and relevant bulk modulus B₀ = 166.7 GPa as well as its pressure derivative B₀′ = 5.48 for bulk fcc Au at zero pressure and 298 K 24 25, the internal compressive stress of the as-formed fcc nanocondensates was calculated to increase up to ~8 GPa with the increase of PLAL power density (Table 1).

A high quenching rate of laser-induced plasma in confining liquid was known to cause metastable phase, such as diamond, that is thermodynamically stable in high-temperature high-pressure regime 26 . The estimated compressive stress level up to ca. 8 GPa is not high enough to stabilize the high-pressure hcp phase, which typically occurs above ~240 GPa at ambient temperature 27 . The reasons to retain a high internal stress for the fcc nanocondensates may include a very high quenching rate, the capillarity force under nanosize effect, and the constraint exerted by partially epitaxial MTPs. It should be noted that the internal stress remained nearly unchanged for the fcc nanocondensates after room-temperature aging for 40 days in water (not shown). The relaxed fcc-type wall of the microtubes thus has nothing to do with such nanocondensates. Instead, they were derived from atom clusters through an intermediate lamellar phase. The driving force for the formation of such a mesomorphic phase was likely total surface reduction under the influence of water for hydrogen-bonding related coordination change of Au atoms.

The present Au nano- and micro-tubes were able to develop from atom clusters and then lamellae for further rolling in aqueous solution without the assistance of artificial surfactant for the following reasons. The atom clusters were likely self-assembled into lamellae under the influence of hydration effect, as indicated by OH stretching mode in the FTIR spectrum (Fig. 6c), in order to reduce the exposed surface area and to modify the structure of nanoparticles as the case of other materials 28 . The mesomorphic rolling of the Au lamella was then triggered by the mismatch strain of its structure units with considerable distortion and residual stress due to a dynamic PLAL process. The lamellae could possibly be made of atom clusters in the form of trigonal planar and pentagonal pyramid with electron positive ligands that donate electron density to the relativistically contracted and stabilized Au 6s orbital, and thus enhance the aurophilic interactions 29 , or alternatively distorted hexagon as a precursor of fcc structure motif. In any case, the anisotropic strain among the randomly distributed atom clusters with directional ligands throughout the network would cause the lamella to roll up in order to minimize the dangling bonds analogous to the case of carbon nanotubes 30 . Such energetic favorable rolling of the structural units is also adopted by viruses, Buckminster fullerene 30 , and chrysotile, a well-known fiber tubular mineral having is curvature determined by the degree of mismatch between the cations occupying the tetrahedral and octahedral layers 31 .

It should be noted that the lamellar wall turned into fcc-type structure when the nanotubes were coarsened up to microns in diameter. This indicates that size dependent chemical free energy is of concern to form a more stable close packed structure for the tube wall. A partial epitaxial nucleation route by (111) or (200)_MTP//lamella layer may also lower the activation energy for the lamellar wall to change into fcc structure.

The mesomorphic Au microtubes were able to migrate along the humidified wall of the capped bottle especially when desiccated. This indicates that the outer surface of the Au microtube is hydrophilic. This hydration nature is in drastic contrast with the case of carbon nano-onions floating on the water surface or falling to the bottom of the container 10 , and carbon nanotubes by PECVD or pyrolysis in the absence of surfactant 32 33. Still, it is not clear if the Au nanotube has hydrophobic channel for tight hydrogen-bonding network facilitated conduction of water analogous to the case of carbon nanotube with the channel occupancy and conductivity tunable by the change in the local polarity and solvent condition 34 .