Corresponding application

The present application claims priority to earlier international patent application PCT/IB2018/057401 filed on September 25, 2018 in the name of ECOLE POLYTECHNIQUE FEDERALE DE LAUSANNE (EPFL) the content of this earlier application being incorporated by reference in its entirety in the present application.

TECHNICAL FIELD AND BACKGROUND ART

Micro- and nanoscale metallic glasses (or MGs) offer exciting opportunities compared to their bulky counterparts, both for fundamental research and applications in health care, micro-engineering or optics and electronics. Thus far however, the typical approaches to structure metallic glasses at the nanoscale have entailed complicated micro- and nanofabrication techniques resulting in metallic glasses with limited geometries and aspect-ratios, hard to interface and exploit, which has severely restricted their potential applications. The present patent application presents a simple and scalable approach for the production of very high aspect ratio, well-ordered and uniform metallic glasses of arbitrary geometry and spanning an a wide range of sizes from micrometer scale down to 40 nm. We observe size-dependent crystallization kinetics, revealing the mechanical breakup of metallic glasses ribbons with thicknesses below around 30 nm, via in situ heating in a transmission electron microscope. This approach also enables the fabrication of composite and structured metallic glasses with arbitrary transverse geometries previously unachievable with existing methods. Integrating these metallic glasses with other functional materials in fibers forms all-in-fiber monolithic devices, facilitating their straightforward practical use. The present application demonstrates the versatile capability of these metallic glass- based devices in optics, electronics, optoelectronics and neuroscience.

Micro- and nanoscale metallic glasses have generated tremendous interest due to the intriguing fundamental effects they exhibit such as size-dependent mechanical properties, see references [1] to [4] and crystallization, see references [5] and [6] They also offer great opportunities in technological applications that include biodegradable implants, see reference [7], electrocatalysts, see references [8] and [9], and microelectromechanical devices, see reference [10] Thus far however, functional applications of micro- and nanoscale metallic glasses (MGs) remain underexploited. The most common approach for the fabrication of micro- and nanoscale metallic glasses relies on nanofabrication techniques such as focused ion beams to carve out the desired structures, see references [1] and [2] While these subtractive methods allow for fine architectures, many challenges remain regarding irregular geometries (tapering and top curvature of pillars), limited resolution, surface damage, elementary contamination and still a low throughput. On the non-subtractive side, thermoplastic forming processes such as micro- and nanoimprinting exploit the superplasticity of metallic glasses in the supercooled liquid region. They enable the structuring of multiple metallic glass (or MG) compositions on the micro- and nanoscale with high-precision surface finishing, small feature sizes and relatively high yields. These techniques however still require high-resolution patterning and etching steps. The imperfect wetting behavior between the metallic glasses and the molds has also thus far restricted metallic glass structures to low dimensional accuracy short lengths. Other processing methods include the fabrication of uniform metallic glass nanopillars in PMMA templates via an electroplating process , see reference [11], which suffers from stringent process conditions; metallic glass rod arrays via physical vapor deposition, see reference [12], which results in highly nonuniformity and short length; and entangled metallic glass nanowires via gas atomization, see reference [13], which features a large dispersion in the size distribution. Direct attempts to fabricate long metallic glass micro- and nanowires by pulling their bulky counterparts in the supercooled liquid region could only generate simple structures with uncontrollable size and reproducibility, see references [14] and [15] Therefore, producing long micro- and nanoscale metallic glasses with complex geometric features and unlimited aspect-ratios is technologically challenging using the above-mentioned approaches. Furthermore, the use of the resulting structures fabricated by all of these existing techniques requires complicated post-fabrication treatments, see references [16] and [17] , transfer processes, see reference [8], and integration with macro systems see references [10] and [18], thereby restricting their potential applications.

The present application presents novel and inventive embodiments of thermal size-reduction methods to produce micro- and nanoscale metallic glasses in a flexible polymer fiber in an unprecedented, simple, and scalable way inspired by optical fiber manufacturing. By controlling the crystallization and the in-fiber fluid instability-driven breakup, indefinitely long, well-ordered, uniform metallic glasses spanning an exceptionally wide size range, and with ribbon thicknesses down to around 30 nm depending on the glass composition and properties, have been created in some embodiments of the invention. The study on crystallization via in-situ heating Transmission Electron Microscopy (TEM) of these metallic glass ribbons reveals unusual size-dependent crystallization kinetics that directly result in mechanical breakup of ribbons with thickness below around 30 nm. In embodiments of the invention, designing the structures in a macroscopic scaled- up model of the fiber allows to fabricate composite and structured metallic glasses with arbitrary transverse geometries previously unachievable with existing methods. The resulting metallic glasses are embedded inside flexible polymers, or can straightforwardly be monolithically integrated with other functional materials forming novel all-in-fiber devices according to embodiments of the invention. They can be removed from the embedding polymer mechanically or chemically. The versatile capability of these in-fiber metallic glasses are illustrated by means of two applications: metallic glass ribbons serving as unique conductors interfaced with semiconducting nanowires for high-performance fiber optoelectronics, and multiple metallic glass wires embedded in a biocompatible polymer for simultaneous neural stimulation and recording in the deep brain of freely moving animals. These and other embodiments of the invention are described in more detail hereunder.

In embodiments the invention concerns a method for producing a polymer fiber comprising a metallic glass (so called "MG") via thermal drawing and scale-down from a macroscale to a nanoscale, the method comprising at least the following successive steps

-) preparing a macroscopic scale preform assembled from compatible materials for co-drawing such as a metallic glass structure encased in a supporting first cladding made of a polymer or glass;

-) thermally drawing the first preform into a first fiber.

In embodiments the method further comprises the steps of:

-) encapsulating at least a part of the first fiber in a second cladding made of a polymer or glass,

-) thermally drawing the encapsulated first fiber to obtain a second fiber with a reduced thickness.

In embodiments the method further comprises the steps of:

-) encapsulating the second fiber in a third cladding made of a polymer or glass; -) thermally drawing the second fiber to obtain a third fiber with a further reduced thickness.

In embodiments the method comprise further or additional successive steps of encapsulating the fiber in a cladding and thermally drawing the fiber.

The claddings may be made of a thermoplastic polymer for example.

In embodiments the initial metallic glass is a 60-pm-thick ribbon. Other sizes are of course possible.

In embodiments the step of thermally drawing the first fiber results in a ribbon thickness of a few hundreds of nanometers.

In embodiments the step of thermally drawing the second fiber results in a ribbon thickness of a few tens of nanometers.

Subsequent thermal drawings as described herein result in a reduction of the ribbon thickness.

In embodiments the initial metallic glass structure is a cylinder with an arbitrary cross-sectional shape.

In embodiments, an array of metallic glass cylinders of arbitrary size and shape, and organized in different structures, are initially fabricated in a preform, and thermally drawn into a fiber. In embodiments the extended first fiber is cut before being thermally drawn. This cutting operation may be repeated on the drawn fibers.

In embodiments the claddings may be removed after the last thermal drawing step or they may be removed after each drawing step.

In embodiments the invention concerns a method of manufacturing a product using a fiber produced with a method as described in the present application. The product may be used in different applications and or have different uses as will described in more detail hereunder.

In embodiments the metallic glass materials of the fiber may be used as or form electrodes of an electrochemical sensor.

In embodiments the metallic glass materials of the fiber may be used as or form electrodes wherein at least one of the electrodes act(s) as a reference in an electrochemical sensor.

In embodiments the metallic glass materials of the fibers may be used as electrodes in a neurological probe for stimulation and recording in the deep brain or other tissues.

In embodiments the metallic glass materials of the fiber may be interfaced with semiconducting materials and used as electrodes for example in an optoelectronic fiber device.

In embodiments the metallic glass materials of the fiber may form a ring to make a waveguide for light transmission.

In embodiments the metallic glass materials of the fibers are structured to make a metamaterial with electronic, magnetic or optical properties.

In embodiments the method comprises the step of slicing at least one fiber formed by the method as defined herein to form thin metamaterials with optical, magnetic or electrical properties.

In embodiments, the invention concerns a product such as a fiber, obtained by a method as defined in the present application or the use of such fiber in a device.

Accordingly, the present invention concerns methods, products and uses as described hereunder in detail by the way of exemplary and non-limiting embodiments.

DETAILED DESCRIPTION OF THE INVENTION Figure 1 | Production of indefinitely long, highly uniform, well-ordered micro- and nanoscale metallic glasses (MGs) via thermal drawing according to embodiments of the invention.

a, Schematic of the fabrication scheme showing a macroscopic metallic glass ribbon encapsulated in a PEI cladding reduced to a nanoscale ribbon via multi- step drawing.

b, SEM micrographs (cross-sectional view) of metallic glass ribbons with the thickness of ~ 8 pm, 1.2 pm, 390 nm, 160 nm, 75 nm and 40 nm.

c, left image: optical image (global view) of 28-meter-long metallic glass ribbons with thicknesses of ~ 2.5 pm by drawing a 5-cm-long metallic glass ribbon with a thickness of 60 pm. metallic glass ribbons are extracted by peeling off the PEI cladding; middle image: SEM micrograph (global view) of ribbons of -500 nm thickness produced by drawing a 5-cm-long once-drawn fiber with a thickness of 10 pm; right image: SEM micrograph (global view) of ribbons of - 50 nm thickness by drawing a 5-cm-long twice-drawn fiber with a thickness of 1 pm. Metallic glass ribbons in the middle and right are extracted by chemically dissolving the PEI cladding, resulting in globally aligned, freestanding micro- and nanoribbons.

Figure 2 1 Breakup of in-fiber metallic glasses and their electron microscope characterization a according to embodiments of the invention.

a, Optical micrograph (top view) of a fiber, showing the breakup of the metallic glass ribbon.

b, Selected-area electron diffraction (SAED) pattern of a broken fragment.

c, Dark-field TEM image of a broken fragment, showing nanocrystals with orientation along the fiber axis.

d, High-resolution TEM image in the crack region of a fragment, showing nanocrystal formation along the crack.

Figure 3 | Size-dependent crystallization kinetics investigated via in situ TEM technique according to embodiments of the invention.

a, In situ TEM characterization of the Tx of nanoribbons with varying thicknesses. Snapshot SAED patterns of nanoribbons (with thicknesses of 45, 90, 105 nm) when heated with a constant ramping rate of 40 °C/s. The red rectangles indicate the onset of crystallization.

b, Quantification of T _x of these nanoribbons based on the intensity change of diffraction pattern contrasts. The abrupt increase of the relative intensity (denoted by the arrows) indicates the onset of crystallization, which is defined as T _x (see Methods).

c, In situ TEM characterization of the crystallization time of nanoribbons with varying thicknesses. Snapshot SAED patterns of nanoribbons (with thicknesses of 45, 90, 105 nm) when isothermally heated at 260 °C. The rectangles indicate the onset of crystallization.

d, Quantification of crystallization time of these nanoribbons based on the intensity change of diffraction pattern contrasts. The abrupt increase of the relative intensity (denoted by the red arrows) indicates the onset of crystallization, which is defined as crystallization time (see Methods). The fluctuation of the normalized relative intensity for these samples is due to the instability of the recording system in the TEM. Figure 4 | Composite and structured metallic glasses in fibers according to embodiments of the invention.

a, SEM micrograph (cross-sectional view) of a fiber that consists of ten alternating layers of the MG and the dielectric PEI encapsulated by the PEI cladding. Inset: High-magnification SEM image of the interface region.

b, SEM micrograph (side-view) of a fiber that consists of metallic glass arrays exposed on the fiber surface.

c, SEM micrograph (tilted view) of a fiber containing a metallic glass slotted- cylinder. Inset: EDX mapping of Pt.

d, left: SEM micrograph (tilted view) of a fiber consisting of a hollow air core surrounded by a metallic glass full cylinder that is encapsulated by a thin PEI layer. Inset: EDX mapping of Pt; right: High-magnification SEM image confirming the fiber architecture; the air core is filled with epoxy to facilitate the cross-section polishing and SEM imaging e, left: SEM micrograph (cross-sectional view) of a fiber containing three metallic glass ribbons and two metallic glass rods. The metallic glass ribbons are indicated with red rectangles and their thicknesses are highlighted in the high-magnification SEM images on the right.

Figure 5 | High-performance optoelectronic fibers using metallic glass ribbons as electrodes according to embodiments of the invention.

a, Optical micrograph (cross-sectional view) of a fiber that consists of an amorphous Se core sandwiched by two metallic glass ribbons in contact with carbon-loaded polycarbonate nanocomposite domains.

b, SEM micrograph of the fiber tip after sonochemical treatment, showing the Se nanowire network interfaces with metallic glass electrodes.

c, Photosensitivity defined by the ratio of photocurrent over dark current measured at different voltages (V = 2, 5, 10 V) and under monochromatic illumination (l = 532 nm) as a function of power.

d, Photodetecting fibers embedded in a fabric.

Fig. 6 | Neural stimulation and recording using MG-based fiber probes. a, Photograph (total view) of a fiber implant. Inset: Optical micrograph (cross- sectional view) of the fiber tip showing the fiber diameter and electrode diameters. b, Photograph of a fiber implanted in the MLR region of the rat brain.

c, Fibers implanted in the MLR region of the rat brain are fully functional and capable of in vivo stimulation and recording. EMG traces (muscle activity) recorded from left and right hindlimb muscles (TA - tibialis anterior muscle) during 40 Hz stimulation of the MLR show that locomotion is initiated shortly after stimulation is turned on (blue arrow) and stops as soon as stimulation is turned off (red arrow).

d, A representative example of the unitary activity recorded from the MLR during spontaneous locomotion at 4 weeks post-implantation. Trace shows numerous action potentials recorded from the same neuron by one electrode in an MLR implanted fiber.

e, Representative immunohistochemistry images taken from 12 week post implantation tissue highlight the inflammatory processes occurring at the electrode implantation site, with microglia and macrophages (Iba1 , red) being recruited to the site.

f, A glial scar of reactive astrocytes (GFAP, red) forming.

g, A detailed close up of the glial scar highlights that the tip of the implant was surrounded by reactive astrocytes (GFAP, red) and there were no remaining neuronal cell bodies (NeuN, green) in the immediate vicinity.

Figure 7. Temperature dependence of viscosities of typical thermoplastics PEI, PSU, PMMA and PC used as claddings for thermal drawing as well as some typical bulk metallic glasses (Zr4i.2Tii3.8Cui2.5NiioBe22.5, Pd43Cu27l\lhoP2o, Pt57.5Cui4.7Ni5.3P22.5, Au49Ag5.5Pd2.3Cu26.9Sh6.3, Zr35Ti30Cu8.25Be26.75,) the data on these metallic glasses are from reference [33]

Figure 8. DSC heating curves of PEI and bulk Pt57.sCui4.7Ni5.3P22.5 metallic glass at a rate of 20 K/min. Onset of glass transition and crystallization temperatures are indicated by arrows.

Figure 9. Time-T emperature-T ransformation (TTT) diagram of bulk metallic glass alloy Pt57.5Cui4.7l\li5.3P22.5. For DSC measurements bulk pieces between 5-20 mg mass were removed from glassy rods. While standard deviations of measured time for 5 and 95% crystallization fractions are shown by error bars, ±2°C accuracy of DSC measurements is not indicated in the figure.

Figure 10. a-c, Selected-area electron diffraction (SAED) patterns of the metallic glass ribbon drawn once, twice and third time, respectively, according to embodiments of the invention.

Figure 11. a-c, Photograph of the metallic glass-based fiber drawn once (2.5 pm ribbon thickness), twice (500 nm ribbon thickness) and third time (50 nm ribbon thickness), respectively, according to embodiments of the invention.

Figure 12. Photograph of a hand drawn PMMA and Au49Ag5.sPd2.3Cu26.9Sii6.3 preform. Inset: SEM micrograph of the fiber cross-section.

Figure 13. a, The profile of a metallic glass section, according to embodiments of the invention.

b, Temperature distribution along the MG section in the neck-down region. c, Velocity profile in the neck-down region.

d, Diameter of the MG section in the neck-down region.

Figure 14. a, . ) ^f (*· ¾~/¾-<) -I versus x for the Sn/PSU system.

b, Instability time versus Sn diameter. versus x for the MG/PEI system.

d, Instability time versus MG diameter e, Instability time in the neck down region. Figure 15. a, Schematic of the preform making for the fabrication of the metamaterial fiber (Fig. 4a) consisting of ten alternating layers of metallic glass and PEI . The same approach was exploited to make fibers where metallic glasses are exposed on the fiber surface (Fig. 4b), according to embodiments of the invention.

b, Schematic of the preform making for the fabrication of fibers containing a metallic glass slotted-cylinder (Fig. 4c) or a hollow air core surrounded by a metallic glass full cylinder (Fig. 4d), according to embodiments of the invention. c, Schematic of the preform making for the fabrication of the hybrid fiber that integrates a PEI/PES optical fiber and three metallic glass ribbons as well as two metallic glass rods (Fig. 4e), according to embodiments of the invention.

d, Schematic of the preform making for the fabrication of the fiber for neural stimulation and recording (Fig. 6), according to embodiments of the invention.

Figure 16. 1-V curves of the optoelectronic system under different light powers at l = 532 nm.

Figure 17. I-V curve of a metallic glass ribbon that was drawn from a 60-pm- thickness ribbon.

Figure 18. a, Representative Electrochemical Impedance Spectra (EIS) measured in vitro and in vivo on 30-pm-diameter metallic glass fiber electrodes, according to embodiments of the invention.

b, Electrochemical Impedance Spectroscopy (EIS) - phase, measured in vitro in phosphate buffered saline solution on two 30-pm-diameter metallic glass electrodes, according to embodiments of the invention. The plot shows the transition from a more resistive impedance at high frequency to a capacitive- dominated impedance towards the mid-range.

c, Voltage Transient (VT) curve, measured in vitro in phosphate buffered saline solution on a fiber electrodes of 30 pm diameter, according to embodiments of the invention. The current waveform has 50 pA amplitude and 300 ps pulse width, and is symmetric charge-balanced with the cathodic phase first. The plot shows that both fiber electrodes enable injecting 50 pA (0.015 pC per phase) with a voltage drop that is lower than 1 V, suggesting that stimulation in vivo may be possible without incurring large requirements in terms of voltage supply.

d, Representative example of a Cyclic Voltammetry (CV) curve measured in vivo 1 week post implantation. The plot shows the second CV cycle in the session. The greyed region is the area used for the calculation of the cathodal Charge Storage Capacity (CSCc).

Micro- and nanoscale metallic glass fabrication via thermal drawing, according to embodiments of the invention

Figure 1 outlines an illustrative embodiment to micro- and nanoscale metallic glass fabrication. It begins with the preparation of a macroscopic scale model, called a "preform" assembled from materials encased in a supporting cladding. The resulting preform is then thermally drawn into an extended fiber, similar to the optical fiber manufacturing process. The cladding is commonly made out of thermoplastic polymers or glasses, which support most of the draw stress and thus is processed at a high viscosity, and undergoes a continuous plastic deformation. Likewise, the metallic glass reaches a viscous state in the supercooled liquid region above the glass transition temperature (T _g ). This is in contrast with conventional metals, which must be heated above their melting temperatures to obtain similar processing capability. The draw stress applied to the cladding then plastically deforms the metallic glass core. Therefore, the general conditions required for co-drawing a metallic glass and a thermoplastic cladding include: the two materials must exhibit compatible viscosities at the drawing temperature; the processing window (between the glass transition and the crystallization temperatures) is comparable; and the metallic glass exhibits a sluggish crystallization behavior featuring a long crystallization time at the drawing temperature. Summarizing the viscosities of typical thermoplastics for thermal drawing and some bulk metallic glasses in Figure 7, we thus choose polyetherimide (PEI) as the cladding and Pt57.5Cui4.7l\li5.3P22.5 metallic glass as the core for the demonstration of producing micro- and nanoscale metallic glass using our method. This alloy also exhibits a large supercooled liquid region (83 °C), a compatible T _g with PEI (see the DSC curve in Figure 8), a higher viscosity than that of PEI (Figure 7), and a long crystallization time at the drawing temperature that allows us to process it multiple times (see the temperature dependent crystallization time in Figure 9).

According to embodiments of the invention, one then designed a multistep thermal size- reduction method to scale down a metallic glass ribbon (fabricated by rapid solidification with melt spinning technique, see the Methods section) encapsulated in the PEI cladding from the macro- to nanoscale.

A first feature of the approach of the present invention is that it can produce uniformly sized metallic glass ribbons or other shapes over an extremely wide range of thicknesses. Figure 1a schematically illustrates embodiments of the fabrication process. Beginning with a 60-pm-thick metallic glass ribbon, an initial draw reduces the metallic glass thickness to a few micrometers (see figure 1a on the left side). A piece of the millimeter-long fiber obtained from the first step is cut and encapsulated in a PEI jacket, consolidated, and redrawn (see figure 1 a in the middle). This second step reduces the ribbon thickness to a few hundreds of nanometers. The drawing step is repeated a third time (see figure 1a on the right side) using the fiber obtained in the second draw to achieve ribbons with thicknesses of a few tens of nanometers. The metallic glass in each step remains highly amorphous (see Figure 10). The feature size in each step is adjustable depending on the scale-down-ratio (between ~5 and -100 for example). Experimentally, fibers with metallic glass thicknesses ranging from -8 pm down to -30 nm have been drawn, confirmed by microscope imaging on the fiber cross- sections (see Figure 1 b). It is remarkable that the in-fiber metallic glass ribbons on different scale levels maintain the radial and axial uniformity and integrity. The second appealing aspect of the present invention is its scalability, namely the ability to manufacture metallic glass ribbons or other shapes whose feature sizes span over an extremely wide range of scales with a large-quantity. Starting with a 5-cm-long bulky metallic glass ribbon with the thickness of 60 pm, it has been possible to produce 28-meter-long metallic glass ribbons with thicknesses of ~ 2.5 pm (see Figure 1 1a), and 20-meter-long metallic glass ribbons with thicknesses of -500 nm by drawing a 5-cm-long once-drawn fiber with a thickness of 10 pm (see Figure 11 b), and 20-meter-long metallic glass ribbons with thicknesses of -50 nm by drawing a 5-cm-long twice-drawn fiber with a thickness of 1 pm (see Figure 1 1c). These metallic glass ribbons can be easily extracted when needed via mechanical peel-off of the PEI cladding due to the weak adhesion between the PEI and the metallic glass or by selectively dissolving the PEI with an organic solvent (N-Methyl-2-Pyrrolidone, NMP).

The resulting globally aligned, free-standing micro- and nanoribbons are shown in Figure 1c. In principle, one may obtain 72000-meter-long 50-nm-thickness metallic glass fibers by drawing a 5-cm-long bulky metallic glass ribbon. This change corresponds to 3 orders of magnitude in linear thickness. Assuming that the surface area of this 5-cm-long bulky metallic glass ribbon is 5 cm ² , one can get 6000 cm ² of surface area from this 72000-meter-long fiber. The approach according to the present invention can also be applied to other MGs, e.g., Au49Ag5.5Pd2.3Cu26.9Si 16.3 (Tg = 128 °C) can be co-drawn with PMMA (see Figure 12) .

Feature size and mechanical breakup according to embodiments of the invention

Continuity of the metallic glass ribbons along the fiber length has been achieved for ribbon thicknesses as low as 30 nm, which is a significant improvement from conventional crystalline metal wires that longitudinally break up into fragments when their diameters reach a few micrometers during thermal drawing. This unique feature of in-fiber metallic glasses may be appreciated by modelling the metallic glass fiber core as a viscous fluid thread surrounded by another viscous immiscible fluid, and applying the classical Tomotika linear instability theory. At elevated temperature, a sinusoidal perturbation appears on the surface of the metallic glass due to the difference in surface tension between the two materials. Such a perturbation grows exponentially over time with an instability time scale t that is linearly proportional to the core diameter and inversely dependent on the maximum of a function associated with the core/cladding viscosity ratio Hcore/qciadding (see further description hereunder). The perturbation at the maximum of this function dominates the instability, ultimately leading to breakup. For a typical Sn/PSU (tin/polysufone) system, the maximum of this function reaches the upper limit unity due to an extremely low Hcore/qciadding, resulting in an instability time of a few seconds for a one-pm-diameter Sn wire. For the metallic glass/PEI system, this maximum value is two to three orders of magnitude lower at the drawing temperature, which gives rise to a long instability time for the metallic glass wires with submicron feature sizes below submicron. If the instability time exceeds the processing time (the time when the materials dwell in the neck-down region, see more details hereunder), the metallic glass keeps the continuity. Otherwise, the instability causes breakup. Simulation results show that the instabilities should not grow for a metallic glass section with a characteristic feature size spanning from 900 nm to 30 nm (see further description hereunder). In this scenario, the visco-elastic force that maintains the integrity beats the surface tension that induces the perturbation, leading to the longitudinal continuity of the metallic glass. However, it has been discovered that the metallic glass ribbon breaks when the thickness is around, sometimes higher than 30 nm (Figure 2a). Since this fragmentation is not created by the fluid instability at the interface, especially given the failure behaviour that is more reminiscent of a mechanical break, it is assumed that the crystallization, which is detrimental to the superplasticity of metallic glasses, might be enhanced at this thickness scale. Indeed, selected-area electron diffraction (SAED) from transmission electron microscope (TEM) characterization shows broad features on the second diffusive ring of SAED pattern taken from a fragment (Figure 2b), and a dark field TEM image formed by the broad ring reveals a highly dense cluster of nanocrystals with lengths of below 10 nm that preferentially orientate in the drawing direction (Figure 2c). Once the volume of these nanocrystals relative to the amorphous matrix is large enough in some regions, the breakup starts to occur due to an enhanced viscosity and hence stress build-up during drawing. This mechanical breakup originating from crystallization is further confirmed by the high-resolution TEM (HRTEM) imaging of a cracked region that shows many nano crystals (Figure 2d).

Size-dependent crystallization kinetics revealed via in situ TEM

According to embodiments of the invention, the ability to produce metallic glass ribbons with tunable thicknesses on the nanoscale allows to gain deeper insight into the size-dependent breakup. Hence, in-situ heating TEM experiments we performed to study the crystallization kinetics of metallic glass ribbons with thicknesses of 45, 95, and 105 nm. One first investigates the crystallization temperature T _x variation via heating the samples from room temperature to 500 °C (see Methods). One observes that the crystallization temperature T _x decreases with decreasing ribbon thickness, as shown by the snapshot SAED patterns in Figure 3a. Figure 3b presents the quantitatively measured T _x of ribbons of the three thicknesses (see Methods). These results are consistent with the detected T _x from SAED, further confirming the size-dependent T _x . Remarkably, crystallization of a 45 nm ribbon occurs at ~ 268 °C, which is approaching the drawing temperature (~ 260 °C). In order to estimate the crystallization rate of these samples at the drawing temperature, one then investigates their crystallization time (t _c ) via isothermal holding at 260 °C (see Methods). As indicated by the snapshot SAED patterns in Figure 3c, t _c decreases rapidly with decreasing ribbon thickness. The corresponding quantitative measurement of tc of these ribbons is presented in Figure 3d. The slope variation of the curves after crystallization further demonstrates that the crystallization rate increases when reducing the ribbon thickness. In particular, the t _c of a 45 nm ribbon is as low as ~32 s. The size-dependent crystallization kinetics observed here might be rationalized by recently reported nanoscale confinement effects, see reference [5] These findings reveal that a ribbon with a thickness above 45 nm remains highly amorphous owing to sluggish crystallization kinetics at 260 °C and thus maintains the continuity, whereas the crystallization kinetics become more rapid when the ribbon thickness approaches 40 nm, leading to an increase in nanocrystal formation and mechanical breakup.

Composite and structured metallic glasses according to embodiments of the invention

The third peculiar attribute in some embodiments is that the approach according to the present invention allows to produce structured metallic glasses. As the preform is structured at the macroscale, complex cross-sections with multiple materials can be easily designed and a subsequent thermal drawing that overcomes crystallization-induced and capillary-induced breakup enables the production of fibers with the desired transverse geometry while maintaining the integrity. One first demonstrates the fabrication of a metamaterial fiber consisting of ten alternating layers of the metallic glass and the dielectric PEI encapsulated by a PEI cladding (cross-section shown in Figure 4a). The interface between the two materials is intimate (the inset of Figure 4a), and the number and feature size of these layers are tunable simply by designing a desired preform, which might enable the scalable fabrication of fiber-based negative refractive index metamaterials, or other optical effects commonly associated with optical metamaterial. Besides embedding these metallic glass ribbons inside the polymer matrix, they can also be exposed on the fiber surface. As shown in Figure 4b, a metallic glass array with a periodicity of 4.5 pm spans along the fiber length. This might open a new way for the fabrication of optical gratings, see references [19] and [20] One then demonstrates that circular metallic glasses with large curvatures can also be structured in fibers. Figures 4c and 4d show a fiber containing a metallic glass slotted-cylinder and a fiber consisting of a hollow air core surrounded by a full metallic glass cylinder that is encapsulated by a thin PEI layer (shown by the magnified SEM image in Figure 4d), respectively. The integrity of both structures is confirmed by the EDX mapping shown in the insets. The former fiber might find application as a magnetic resonator, see reference [21] and the latter might be used for transmitting high-power light at long wavelengths, see reference [22] These examples indicate a strategy for the construction of metallic glass-based fibers with increasingly sophisticated architectures. In Figure 4e, one shows a hybrid fiber that is comprised of a PEI/PES waveguide surrounded by three metallic glass ribbons (thicknesses of 2.5 pm, 400 nm and 53 nm) and two metallic glass cylindrical rods (diameters of 26 pm) encapsulated in a PEI cladding. One predicts the use of such multi functional fibers for simultaneous optical and electrical in vivo dissection of neural circuitry, see reference [23] In particular, these metallic glass domains acting as highly conductive electrodes can be shaped into arbitrary geometry and at feature sizes previously unachievable with conventional metals or conductive nanocomposites, see reference [24], enabling interrogation of brain at sites of interest with adjustable temporal resolution. The fiber architectures proposed above are prototypical structures from which more complex geometries that have more stringent requirements on the shape, feature size, and number of metallic glass domains may be designed. Fabricating these architectures with conventional metals remains a significant challenge, owing to uncontrollable fluidic instabilities during thermal drawing. The metallic glass-based structure proposed here has a wide-range of potential uses in different areas such as electronics, optoelectronics, and bioengineering, as described in the following sections.

Finally, it is also possible to design complex cross-sectional architectures and then cut the formed fiber into thin slices for example, 100 micrometer thin slices of a one kilometer fiber would produce 10 million of small devices that integrate complex metal-polymer metamaterial.

Optoelectronic applications in some embodiments of the invention

The ability to structure fibers having complex transverse geometries with metallic glasses, together with their superior properties, allow this one-dimensional, flexible, and thin platform to have many potential applications. One first demonstrates the use of an optoelectronic fiber that monolithically integrates metallic glass ribbons and a selenium (Se) core acting as highly conductive electrodes and a photosensitive semiconductor, respectively, to efficiently sense photons, as summarized in Figure 5. The metallic glass electrodes owning a high viscous state are capable of directly interfacing with molten Se during thermal drawing, which addresses a long-lasting challenge in the functional fiber field: semiconductors and crystalline metals that are thermally co-drawn will mix due to their molten states, leading to the disruption of the device structure. However, in using metallic glasses as electrodes, our configuration forms intimate contacts between the metals and the semiconductor in the as-drawn fiber (Figure 5a). One then creates a monocrystalline Se nanowire/metallic glass optoelectronic system at the fiber tip by leveraging our recent developed sonochemical method, see reference [25] (Figure 5b, and see Methods). The two attractive attributes of this optoelectronic system - high-quality homogeneous junctions at the interfaces (see the linear l-V curves in Figure 16) and highly conductive metallic glass electrodes (see the electrical conductivity in Figure 17) - enable high- performance optoelectronic properties. Indeed, the device exhibits extraordinary photosensitivity that is orders of magnitude higher than that of some planar wafer- based devices. Optoelectronic fibers with a functionalized tip can find several applications in remote detection and sensing, optoelectronic probes, and minimally invasive in situ and in vivo bio-compatible probing and imaging of biological tissues, as well as smart textiles. The small cross-section and large aspect ratio of this high-performance fiber allows access to remote and confined environments where rigid and planar point photodetectors are unable to reach.

Neural stimulation and recording applications according to embodiments of the invention

In addition to a high conductivity, the metallic glass alloy in some embodiments also exhibits other interesting features, such as oxidation and electrochemical reaction resistance. Fibers with several MG electrodes (at least 2) could be used for electrochemical sensing in a variety of configurations. Note that this alloy is rich in platinum, a standard material used for clinical implants in the treatment of Parkinson’s disease, see reference [26] and that PEI is a biocompatible material commonly used in medical devices. Therefore, the second demonstration is in neuroscience. In particular, we demonstrate the capability of a fiber probe that integrates four metallic glass cylindrical electrodes embedded in the PEI (see Figure 6a) for long-term neural stimulation and recording in the deep brain of freely moving rats. Neural stimulation and recording are important means for understanding the information transfer and processing within the mammalian nervous system.

The fiber probes are mechanically flexible, and the implanted portions (~8 mm in length, see Figure 6a,) have a miniaturized dimension (-300 pm in diameter, inset of Figure 6a), facilitating their implantation into the brain at multiple sites (see Figure 6b). The un-implanted portion has a relative larger size (-900 pm in diameter), allowing wire connection with external electronics (see Figure 6a). Prior to implantation, electrochemical characterization in vitro, in terms of impedance modulus, phase and voltage transient, shows their excellent functionality and viability for recording and/or stimulation (see Figure 18a-c).

In order to assess the functionality and longevity of implanted fibers, electrical stimulation and recording sessions were performed on a weekly basis for up to 12 weeks post-implantation. The fiber was implanted in the Mesencephalic Locomotor Region (MLR) of the rat brain (see Figure 6b), for a number of reasons. Firstly, the neurons of this region are known to be highly active during locomotion, with minimal spontaneous neuronal activity when the animal is at rest. Thus, determining whether electrodes are detecting actual neuronal activity is easier, as recordings of neuronal activity should be closely linked with locomotor activity. Secondly, if the neuronal cell bodies located in this region are artificially activated using electrical stimulation, the animal produces locomotor activity despite having been in a resting state prior to stimulation (i.e. stimulation forces the animal to walk). Therefore, this gives a clear and obvious behavioral response to stimulation if electrodes are functional. In all animals (n = 6), 40 Hz stimulation of the MLR using the implanted devices induced a stereotypical forced locomotion response (see Figure 6c) at all time-points assessed (i.e. up to 12 weeks post-implantation). The electrochemical properties of these implanted electrodes remained excellent, as shown by the characterization on impedance modulus and cyclic voltammetry in vivo (see Figure 18a and d).

For the recording of neural activity, one were immediately able to record endogenous multi-unit activity (the action potentials of numerous individual neurons in close proximity to the electrode) in the MLR during periods of spontaneous locomotor activity in all animals at the earliest time-point of 1 week post-implantation, and again during weekly recording assessments up to 2 months post-implantation (see Figure 6d). Generally, the ability to record this activity was gradually lost between 6 and 8 weeks post-implantation. Whilst it is disappointing that one could no longer record multi-unit neuronal activity beyond 2 months post-implantation, this was not unexpected and is in line with the typical signal degradation that is often observed with these types of recordings (particularly using electrodes in a single shank) over a period of days to months (see references [27] to [29]). As appears to be the case with the electrodes according to the present invention, signal degradation is generally due to the immune response of the host brain tissue, particularly reactive gliosis (see Figure 6e-g) (see reference [30]).

Overall the results highlight that, using the novel technique in accordance with embodiments of the invention, one is able to produce implantable fibers containing multiple electrodes that are fully functional and are capable of in vivo stimulation and recording over longer periods of time than has previously been reported using nanocomposite electrodes, see reference [23] (up to 2 months for recording and >3 months for stimulation).

This in vivo data acts as proof-of-principal that these fibers according to embodiments of the invention are fully functional for electrical recording and stimulation, although further functionality could also be added to the fibers. For example, in other embodiments, inclusion of optic fibers would allow optical stimulation whilst recording from the same region using the normal electrodes (simultaneous recording is not possible with electrical stimulation, as the stimulation artefact would mask electrical recordings). This could be very advantageous in certain situations, as optogenetic experiments allow very specific activation of neuronal subtypes within chosen brain regions (see reference [31]).

Methods and examples according to embodiments of the invention

Metallic glass alloy preparation. The Pt57.5Cui4.7l\li5.3P22.5 MG ribbons were prepared by melt spinning of rods of Pt57.sCui4.7Ni5.3P22.5 alloy (with a diameter of 10 mm, purchased from PXGroup, Switzerland) on a Cu wheel with a rim speed of 20 ms-1. The alloy was melted in a quartz tube and then ejected onto the wheel by high-pressure argon. The approximate thickness and width of the ribbons were 60 pm and 5-6 mm, respectively. Glassy rods of Pt-based alloy with a diameter of 1 and 3 mm were prepared by suction casting in an arc melter.

Fabrication of micro- and nanoscale metallic glasses. The fabrication of microscale metallic glasses (see Figure 1) began with preparation of a rectangular preform where a bulky Pt57.sCui4.7Ni5.3P22.5 MG ribbon (with a thickness of ~ 60 pm) was encapsulated between two PEI plates (24 x 7 x 15 mm). Subsequently, the assembly was consolidated in a hot press at ~ 240 °C for 30 minutes. The composite structure was then thermally drawn in a fiber drawing tower into microscale metallic glass-based fibers (once-drawn fiber). Starting with the preparation of another preform that encapsulated a piece of once-drawn fiber, a second draw of this preform reduced the feature size of metallic glass into hundreds of nanometers (twice-draw fiber). A third draw of a preform that encapsulated a piece of twice-draw fiber further reduced the feature size of MG into tens of nanometers.

SEM and TEM samples preparation and imaging. All the SEM samples for cross-section imaging were prepared by mechanically cutting using ultramicrotomy (diamond blade). The SEM samples were then coated with a 10 nm carbon film. The SEM images were taken with a Zeiss Merlin field emission SEM (Zeiss, Gottingen, Germany) equipped with a GEMINI II column operating at 3.0 kV with a probe current of 150 pA. TEM samples in Figure 2 were prepared by dissolving the PEI cladding using N-Methyl-2-Pyrrolidone, NMP. The free- standing metallic glass ribbons were then cleaned with ethanol five times before they were transferred on a carbon/ Cu grid support (300 mesh). The TEM images and SAED patterns were taken using a Talos F200X operating at 200 kV.

In-situ TEM samples preparation and characterization. All the TEM specimens for in-situ TEM in elevated temperatures were prepared by dissolving the PEI cladding using N-Methyl-2-Pyrrolidone (NMP) followed by cleaning the metallic glass ribbons with ethanol five times before they were transferred on MEMS chips (DENSsolutions, through hole). After monitoring their resistance, the chips were inserted in the TEM. All data were acquired in a ThermoScientific Titan Themis operated at 300 kV. The samples were first monitored in high-angle annular dark field (HAADF) STEM/EELS mode for t/l (thickness divided by the mean free path) mapping of the area of interest. The absolute thicknesses were calculated after relating the mean free path with a thickness measured from a 120 nm metallic glass ribbon measured by SEM. Bright-field diffraction imaging was then optimized for all samples at a dose of 2 nA and the selected area probed in each case was 500 nm in diameter. Diffraction patterns were acquired in series with a frame size of 1 kx1 k for 0.98 sec dwell time each. The imaging was synchronized with a chosen temperature ramp profile (Heated to 170 °C within 0.5 s from room temperature, and then a ramp rate of 40 °C/s was applied from 170 to 500 °C) for the crystallization temperature measurements and a constant temperature of profile 260 °C for the crystallization time measurements.

Quantification of T _x and t _c of nanoscale metallic glass ribbons according to embodiments of the invention.

The quantification of Tx and tc of the metallic glass ribbons with diameters of 45, 95, 105 nm was carried out based on the contrast change in SAED patterns. In order to better measure the contrast change of the diffusive rings in the SAED patterns, polar transformation was applied to the patterns. Thus, the curved, broad spots in the diffusive rings indicating crystallization appeared as straight, broad lines after transformation. In the polar transformed patterns two rectangular regions of interest were selected to quantify the T _c : one on the straight, broad line and the other on the beam-stop as the background (the size of both regions was the same). Pixel counts from the two regions were performed for all the SAED patterns as a function of imaging time. The relative intensity for each pattern was calculated based on the equation:

Where are the total pixel count of the spot and background, respectively. At the onset of crystallization, the relative intensity increases abruptly from a constant background (Figure 3b). The T _x was defined at the intercept of the constant background and the linear fit to the increasing relative intensity. To quantify the t _c , every ten SAED frames was treated as one frame by averaging the contrast in order to reduce the noise of each frame. The same polar transformation and relative intensity calculation was applied to each frame. The t _c was defined at the intercept of the constant background and the linear fit to the increasing relative intensity.

Selenium nanowire growth and optoelectronic characterization.

The as-drawn fiber was first mechanically polished to get a smooth surface before it was cleaned within an ultrasonic bath. Then the fiber tip was immersed in 1- propanol for four days. The amorphous bulk Se in the fiber transformed into a mesh of nanowires in the solvent, directly in contact with metallic glass electrodes, forming an optoelectronic device at the fiber tip. To characterization the optoelectronic properties of the device, the laser beam generated by a SuperK Extreme from NKT was incoming perpendicularly to the fiber cross section. The electrical- and photoresponse were measured using a Keithley 6517B.

Fiber implantation in Mesencephalic Locomotor Region (MLR).

Adult female Lewis rats (Janvier, France) were group-housed, maintained on a 12h light/dark cycle and had access to food and water ad libitum. For fiber implantation, rats were surgically anaesthetized using 1.5-2.5% gaseous isofluorane in medical oxygen. The skull was exposed and a small hole was drilled over the appropriate area of the brain. The fiber was then implanted into the MLR using coordinates taken from the atlas of Paxinos and Watson ( see reference [32]): -8 mm from Bregma, anterior-posterior (AP); 2 mm from the midline, medio-lateral (ML); and a depth of -6.5 mm, dorso-ventral (DV). Ground and reference wires were attached to skull-fixed screws and the implant was then permanently stabilized using dental cement. Animals were provided with suitable analgesia post-surgery (buprenorphine, 0.01 mg/kg s.c.), and antibiotics (amoxicillin) were given in drinking water for one week to prevent infection due to implantation.

Stimulation and MLR recordings using implanted fiber.

In order to assess the functionality and longevity of the implanted fiber, stimulation and recording sessions were performed on a weekly basis for up to 12 weeks post-implantation. During stimulation sessions the behavioral response to MLR stimulation was tested (40 Hz train of 200 ps duration biphasic pulses, with an amplitude of 50-250 pA; A-M Systems isolated pulse stimulator). Animals were placed on a one-meter long runway and stimulation was initiated, resulting in forced locomotor activity (typically after 1-2 seconds of stimulation) which was halted as soon as stimulation was turned off. Recording sessions during spontaneous locomotor behavior involved the recording of extracellular voltage signals, which were pre-amplified, digitalized, sampled at 24 kHz, and stored using a BioAmp processor (Tucker-Davis Technologies, USA). The channels average was subtracted offline from each trace to reject common mode noise. MLR multi-unit activity consisted of all field potential stochastic events that crossed the threshold value of three standard deviations of the potential signal.

Histological processing.

At study end-point animals were terminally anaesthetized with sodium pentobarbital (80 mg/kg i.p.) and transcardially perfused with phosphate buffered saline (PBS) followed by 4% paraformaldehyde in PBS. Immediately after perfusion the brain was dissected free and post-fixed overnight at 4°C, cryoprotected in 30% sucrose in PBS for 72 hours, then embedded and frozen in OCT. The relevant region of the brain was then cut in 40 pm serial sections using a cryostat. Sections were then immunostained using the following primary antibodies: rabbit anti-GFAP (1 : 1500; Dako) to label astrocytes, rabbit anti-lba1 (1 : 1500; Wako) to label microglia and macrophages, and guinea pig anti-NeuN (1 :3000; Millipore) to label neuronal cell bodies. Complementary secondary antibodies were then used. Images were acquired using an Olympus VS120 and a Leica LSM-880.

1. Viscosities of typical thermoplastics and bulk metallic glasses (see figure 7)

Viscosity measurement of typical thermoplastics: The rheological properties of typical thermoplastics (polyetherimide (PEI), polysulfone (PSU), poly(methyl methacrylate) (PMMA), polycarbonate (PC)) were measured with a rheometer (TA AR2000) in flow mode, with shear rates varying from Is ^-1 and 2.5 s ^_1 so that the polymer flow was in Newtonian regime.

2. Thermoanalysis and Time-Temperature-Transformation of bulk Pt57.5Cui4.7Ni5.3P22.5 metallic glass (figures 8 and 9)

Differential Scanning Calorimetry (DSC) characterization: Differential scanning calorimetry (DSC) measurements of PEI and Pt57.sCui4.7Ni5.3P22.5 were performed using Mettler-Toledo DSC 1/700 under 30 ml/min Ar flow. Their heating curves at a rate of 20 K/min presented in Fig.8 demonstrate the onset of glass transition (T _g ) and onset of crystallization (T _x ) temperatures

Time-Temperature-Transformation (TTT) diagram of bulk Pt57.5Cui4.7Ni5.3P22.5 metallic glass: DSC measurements were also performed to construct the low temperature part of the TTT diagram for Pt57.sCui4.7Ni5.3P22.5. The temperature of metallic glass specimens were increased from room temperature to isothermal treatment temperatures in the supercooled region (250, 260, 270, 280 and 290 °C) at a rate of 60 K/min. They were held at these temperatures until they are fully crystallized. The time required for 5% and 95% crystal phase fractions were determined for each isothermal temperature using total enthalpy of crystallization. At least five measurements were done for each temperature. 3. TEM characterization on the metallic glass ribbon drawn at each step (figure 10) according to embodiments of the invention.

TEM sample preparation and characterization: The TEM samples of the metallic glass ribbon drawn once and twice were prepared by embedding the metallic glass fiber in epoxy resin followed by sectioning thin slices (~70 nm) which were transferred on a carbon/Cu grid support (300 mesh). The SAED patterns were taken using a Talos F200X operating at 200 kV. The TEM sample of the metallic glass ribbon drawn third time was prepared by dissolving the PEI cladding using NMP followed by cleaning with ethanol 5 times before the ribbon was transferred on a MEMS chip (commercially available wildfire, through hole, DENS solutions). The SAED patterns was taken using a ThermoScientific Titan Themis operated at 300 kV.

4. Scalability of thermal drawing of the metallic glass ribbon (see figure 11 ) according to embodiments of the invention

5. Preliminary results of thermal drawing of PMMA and AU49Ag5.5Pd2.3CU26.9Sil6.3

According to the viscosity measurements (see Fig. 7), PMMA and Au49Ag5.5Pd2.3Cu26.9Si 16.3 have a similar viscosity between 150 and 175°C and could be another pair of suitable materials to be co-drawn. We successfully deformed a PMMA preform containing a 80 pm thick Au49Ag5.sPd2.3Cu26.9Sii6.3 ribbon by heating it on a hot plate set at 230°C and pulling it by hand. Fig. 13 shows the deformed preform with in inset a SEM micrograph (cross-sectional view) of the most deformed part. Due to the limited time before crystallization in this temperature range and the broad heating zone of the furnace of our drawing tower, we were still not able to draw a continuous fiber by bypassing crystallization. However, it is believed that with a more precise temperature control in the bait-off zone, these two materials could be co-drawn.

6. Modelling of the axial velocity and diameter of a MG ribbon in the neck- down region (see figure 13)

The tensile stress applied on the preform that is heated above its T _g leads to the thermoplastic deformation of the preform. The deformation creates a neck-down region, as shown in Figure 13(a). The axial velocity of one domain in this region along the drawing direction z can be expressed by (see reference [34]):

where v^ v^are the feeding speed of the preform and the drawing speed of the fiber, respectively; L is the length of the neck-down region, where the temperature is above the T _g of the metallic glass so that it can deform; h( _z ) is the viscosity of the Pt57.5Cui4.7Ni5.3P22.5 metallic glass and we assume that it is only temperature dependent and can be obtained from Figure 7.

Assuming the conservation of mass for the system, the relation between the radius along the drawing direction z and the initial radius can be expressed by:

(Supp. Eq.2)

where is the initial radius, R(z) is the radius along the z axis, v _z0 is the initial velocity.

In order to get the viscosities of the metallic glass core, we precisely measured the temperature distribution along the preform when it is soaked in the furnace of fiber draw tower using a thermocouple. The measurements were fitted by the quadratic function T(z ) = _z ² + T _max , with T _max = 258 °C, T _g =

228°C and L = 6 cm. The result is plotted in Fig. 13b, where the position having the highest temperature is set as the origin.

One considers that a 900-nm-diameter metallic glass rod cladded by a PEI cladding is drawn into a 30-nm-diameter metallic glass fiber with a scale down ratio of 30 with typical v _f = 1 mm/min, v _d = 0.9 m/min. Based on Supp. Eq.1 , one obtains the velocity profile in the drawing direction (see Fig. 13c) and based on Supp. Eq.2, we obtain the diameter of metallic glass rod in the neck-down region (see Fig. 13d) and the profile of the metallic glass section (see Figure 13a).

7. Modelling of the instability time of a metallic glass ribbon in the neck- down region (see figure 14)

Classical Plateau-Rayleigh instability theory points out that, when a cylindrical column of liquid surround by another substance, a sinusoidal perturbation wave would appear at the heterogeneous interfaces between the two materials and may finally leads to the breakup of the liquid. Here, we consider the metallic glass core heated at temperature T as a viscous cylindrical thread (diameter 2 R and viscosity ¾ _ore ) encased by an infinite viscous cladding PEI (viscosity ¾ _ad ) and the core is incompressible, Newtonian and isotropic. The time scale determining the growth of the perturbation can be obtained using Tomotika’s linear theory, see reference [35]:

(Supp. Eq.3) Where y \s the interfacial energy between the cladding and the core metal, 2 p

x =—— (2 is the wavelength of a perturbation), Ois an explicitly known function

A

see reference [35]:

where b (x) and h(x) are the modified Bessel functions of the 0th and 1th order, respectively. Di, D2, D ₃ , D ₄ are the functions of x expressed in determinantal forms as follows:

The perturbation with the max (l-x ² )o(x,^ _core /¾ _arf )J will dominate the instability and will lead to breakup (this is a very conservative estimate). We first model the instability time of a typical metal, e.g., Sn codrawn with a PSU cladding { h _¥k =

10E-3 Pa-s, h _άaMh}, = 10E5 Pa-s, g = 0.1 N/m). The

reaches unity when x is around 0.015 (Fig. 14a). The instability time versus radius given by Eq. 3 is plotted in Fig. 14b. For the metallic glass/PEI system, however, max (ΐ - h _¥h Ih _άaά ) is almost three orders of magnitude lower compared to the Sn/PSU system (Fig. 14c). This creates a much longer instability time shown in Fig. 12d, e.g., the instability time of the metallic glass/PEI system is two orders of magnitude longer than that of the Sn/PSU system when the diameter is 1 pm. Using the same drawing parameters and temperature profile in section 3 and assuming that the interfacial energy is dominated by the surface tension of Pt57.5Cui4.7l\li5.3P22.5 (y I N/m, a typical value for BMGs, see reference [36], which is another conservative estimate considering that the interface between metallic glass and PEI is intimate), one obtains the instability time at each position in the neck-down region (see Fig. 14e).

In order to determine the stability of the metallic glass rod, it is required to compare the processing time (the time when the material dwells in the neck-down region before it exits furnace) with the capillary instability time. To do so, we use the growth factor G defined as follow, see reference [37]:

(Supp. Eq.8)

dz

where r(z) is the instability time and—— is the dwelling time at each position in v(z)

the neck-down region. G » 1 corresponds to break-up while G « 1 indicates a stable draw.

With our parameters we obtain:

G = 0.79

Though this value does not completely fulfill the criteria of G « 1, many assumptions made are conservative. Therefore, one can expect a stable draw under these conditions. 8. Preparation of complex and structured metallic glasses (see figure 15) according to embodiments of the invention

Various methods were used to make macroscopic preforms from which complex and structured metallic glasses fibers were fabricated. To make the preform shown in Fig. 4a, ten alternating layers of metallic glass and PEI were placed in the groove milled in one PEI plate before it was consolidated via hot pressing with another PEI plate. The thickness of these metallic glasses was ~10 pm obtained by the thermal drawing of a ~60 pm-thickness metallic glass ribbon. The same approach was used to make the preform of the fiber shown in Fig. 4b. Here, the stack of alternating layers of metallic glass and PEI was placed on the edge of the two PEI plates. To make the preform shown in Fig. 15b, the metallic glass ribbon with the desired thickness was inserted between PEI thin films that were tightly rolled around a PEI rod (to make the fiber shown in Fig.4c) or a ceramic rod that can be removed after consolidation (to make the fiber shown in Fig. 4d). To make the preform shown in Fig. 15c, an optical waveguide structure consisting of PEI core and polyethersulfones (PES) cladding was first made via thin-film rolling technique. More PEI films were wrapped around the structure before the three metallic glass ribbons with desired thicknesses were introduced. After consolidating the structure, two pieces of metallic glass rod-based fibers drawn from a larger metallic glass rod were inserted into the channels of the solid. To make the preform shown in Fig. 15d, four pieces of MG rod-based fibers were inserted into the channels of the PEI rod. All the consolidations were performed under vacuum at a temperature above the glass transition temperature of the PEI .

9. Optoelectronic characterization on the MG/Se nanowire optoelectronic system at the fiber tip according to embodiments of the invention (see figure 16)

10. Electrical conductivity of a MG ribbon drawn once (see figure 17)

To measure the conductivity of the metallic glass ribbon, four electrical contacts were made by removing the PEI cladding followed by wrapping around the exposed MG with metallic wires. The four-probe method was applied and l-V curve was recorded using a Keithley 6517B. 11. In vivo and in vitro electrochemical properties of in-fiber MG electrodes

(see figure 18) according to embodiments of the invention

In vitro electrochemical characterization of the fiber electrodes: In vitro measurements were taken before implantation to verify the functionality of the fiber electrodes. The fiber under test was immersed in phosphate buffered saline solution (Gibco PBS, pH 7.4, 1X), together with counter (platinum rod) and reference (Metrohm, El. Ag/AgCI DJ RN SC: KCI) electrodes. In this three- electrode configuration, Electrochemical Impedance Spectroscopy (EIS) were taken using a Gamry 600 potentiostat, in order to characterize the electrochemical properties of fiber electrodes. EIS spectra were acquired by injecting sinewave signals of 0.1 V amplitude at the 1 Hz - 1 MHz frequency range, with 10 data points per decade. In the same, three-electrode setup, Voltage Transients (VT) measurements following constant current were also taken to assess the stimulation properties of the fibers. Constant current, symmetric, biphasic, charge balanced, cathodic-first pulses were applied between the electrode under test and the counter with an A-M Systems 2100 Isolated Pulse Stimulator (300 ps/phase pulse width, 1 s inter-pulse period, 50 pA amplitude), while measuring the voltage across the working and reference electrodes with an oscilloscope.

Fig. 18a reveals a typical capacitive spectrum from low frequency to about 10 kHz, with a less steep slope at frequencies above visible in vitro, indicating, as expected, that the impedance becomes more dependent on the solution resistance, see reference [38] This is also suggested by the phase plots in vitro shown in Fig. 18b. The magnitude of the recorded impedances at 1 kHz is in the range of 25 - 60 kOhm, comparable to some of the devices of similar geometry reported in the literature, see reference [39] Voltage Transient measurements were also taken in vitro to evaluate the stimulation performance of the fibre electrodes. The measurements shown in Fig. 18c for two different fibers demonstrate the ability to inject a representative cathodic current pulse of 50 pA and 300 ps with less than 1 V overall voltage drop.

In vivo electrochemical characterization of the fiber electrodes according to embodiments of the invention: Once implanted, the fiber electrodes were characterized by taking EIS and CV measurements in vivo. The methodology is similar to that followed in vitro, but using a two-electrode configuration, where a wire (Cooner wire) attached to a skull-fixed screw acts as both counter and reference electrode. The measurements were taken using a portable PalmSens potentiostat. The parameters used are as follows: EIS - signal amplitude 0.1 V, frequency range 1 Hz - 50 kHz, 21 data points per decade; CV - voltage window -0.6 V - 0.8 V, scan rate 100 mV/s, 3 cycles per measurement. The CV scans were used to assess the cathodal Charge Storage Capacity (CSCc) of some electrodes in vivo. This is calculated as the time-integral of the negative-current part of the CV curve, where the time-voltage relationship is given by the 0.1 V scan rate.

Fig. 18d shows an example of the CV measurement, with a calculated CSCc of about 6 mC/cm ² for a fiber electrode of 30 pm diameter. This is comparable to the figures reported for typical electrode materials, see reference 40, and could be certainly improved by optimizing (roughening, structuring, etc.) the exposed surface of the fiber.

In summary, the invention and its embodiments propose a simple, unique, and scalable method for producing indefinitely long, highly uniform, and well-ordered micro- and nanoscale metallic glasses via thermal drawing. The approach of the invention is considered effective for the processing of a variety of bulk metallic glass systems. Fundamental study of the crystallization kinetics of nanoscale metallic glass ribbons via in situ heating TEM reveals size-dependent breakup. Overcoming crystallization-induced and capillary-induced breakup enables the production of continuous metallic glass-based fibers with arbitrary cross-sections previously unachievable with existing methods. The resulting micro- and nanoscale metallic glasses are embedded in a flexible polymer fiber matrix, facilitating their handling, direct integration with other functional materials, as well as interfacing with external macro-systems. We demonstrate the versatile capability of these in-fiber metallic glasses by means of applications in optoelectronics and neuroscience. Our work highlights the unique platform that our approach offers for opportunities in fundamental research (e.g. size- dependent crystallization or plastic deformation) as well as in functional applications (e.g. electronics, metamaterial optics, optoelectronics, and bioengineering).

As has been understood from the above description, the present invention not only is about a method but also products, such as a fiber, as defined in the present application and as obtained by the methods and processes described herein.

Non-limiting examples and embodiments of products as indicated hereunder:

-) the product may comprise at least a micro and/or nanoscale metallic glass part; -) the metallic glasses may be in a flexible cladding, for example a polymer fiber; -) the product may comprise metallic glass ribbons serving as unique conductors interfaced with semiconducting nanowires for high-performance fiber optoelectronics;

-) the product may comprise multiple metallic glass wires embedded in a biocompatible polymer used for example for simultaneous neural stimulation and recording in the deep brain of freely moving animals;

-) the product as defined above may act as an electrochemical sensor;

-) the product may comprise a metallic ring of various thicknesses closed or partially opened embedded at different position in a fiber that can act as an optical waveguide, or a magnetic device.

-) the product may comprise micro and nanostructured MG in a polymer fiber that form a metamaterial structure with various optical and electronical properties -) the product obtained from the slicing of the fiber as defined above, to form thin metamaterials with optical (lens... ) or electrical properties.

Other products and applications may of course be envisaged in the frame of the present invention, such as heat sensors or heat generators, scaffolds with metallic glasses domains, etc.

The description is neither intended nor should it be construed as being representative of the full extent and scope of the invention. The present invention is set forth in various levels of detail and the description as well as in the attached drawings and in the detailed description of the invention and no limitation as to the scope of the invention is intended by either the inclusion or non-inclusion of elements, components, etc. Additional aspects of the invention have become more readily apparent from the detailed description, particularly when taken together with the drawings.

The exemplary embodiments have been described to provide an overall understanding of the principles of the structure, function, manufacture, and use of the systems and methods disclosed herein. One or more examples of these embodiments have been illustrated in the accompanying drawings. Those skilled in the art will understand that the systems and methods specifically described herein and illustrated in the accompanying drawings are non-limiting exemplary embodiments and that the scope of the present invention is defined not solely by the claims. The features illustrated or described in connection with one exemplary embodiment may be combined with the features of other embodiments. Such modifications and variations are intended to be included within the scope of the present invention. A number of problems with conventional methods and systems are noted herein and the methods and systems disclosed herein may address one or more of these problems. By describing these problems, no admission as to their knowledge in the art is intended. A person having ordinary skill in the art will appreciate that, although certain methods and systems are described herein with respect to embodiments of methods and products, the scope of the present invention is not so limited. Moreover, while this invention has been described in conjunction with a number of embodiments, it is evident that many alternatives, modifications and variations would be or are apparent to those of ordinary skill in the applicable arts. Accordingly, it is intended to embrace all such alternatives, modifications, equivalents and variations.

REFERENCES

(1) Guo, H.; Yan, P. F.; Wang, Y. B.; Tan, J.; Zhang, Z. F.; Sui, M. L.; Ma, E. Tensile Ductility and Necking of Metallic Glass. Nat. Mater. 2007, 6, 735-739.

(2) Jang, D.; Greer, J. R. Transition from a Strong-yet-Brittle to a Stronger-and- Ductile State by Size Reduction of Metallic Glasses. Nat. Mater. 2010, 9, 215— 219.

(3) Jang, D.; MaaB, R.; Wang, G.; Liaw, P. K.; Greer, J. R. Fatigue Deformation of Microsized Metallic Glasses. Scr. Mater. 2013, 68, 773-776.

(4) Kumar, G.; Desai, A.; Schroers, J. Bulk Metallic Glass: The Smaller the Better.

Adv. Mater. 2011, 23, 461-476.

(5) Sohn, S.; Jung, Y.; Xie, Y.; Osuji, C.; Schroers, J.; Cha, J. J. Nanoscale Size Effects in Crystallization of Metallic Glass Nanorods. Nat. Commun. 2015, 6, 1-6.

(6) Sohn, S.; Xie, Y.; Jung, Y.; Schroers, J.; Cha, J. J. Tailoring Crystallization Phases in Metallic Glass Nanorods via Nucleus Starvation. Nat. Commun. 2017, 8, 1-8.

(7) Zberg, B.; Uggowitzer, P. J.; Loffler, J. F. MgZnCa Glasses without Clinically Observable Hydrogen Evolution for Biodegradable Implants. Nat. Mater. 2009, 8, 887-891.

(8) Doubek, G.; Sekol, R. C.; Li, J.; Ryu, W. H.; Gittleson, F. S.; Nejati, S.; Moy, E.;

Reid, C.; Carmo, M.; Linardi, M.; et al. Guided Evolution of Bulk Metallic Glass Nanostructures: A Platform for Designing 3D Electrocatalytic Surfaces. Adv. Mater. 2016, 28, 1940-1949.

(9) Hu, Y. C.; Wang, Y. Z.; Su, R; Cao, C. R; Li, F.; Sun, C. W.; Yang, Y.; Guan, P.

F.; Ding, D. W.; Wang, Z. L.; et al. A Highly Efficient and Self-Stabilizing Metallic-Glass Catalyst for Electrochemical Hydrogen Generation. Adv. Mater. 2016, 28, 10293-10297.

(10) Lin, Y. C.; Tsai, Y. C.; Ono, T.; Liu, P.; Esashi, M.; Gessner, T.; Chen, M. Metallic Glass as a Mechanical Material for Microscanners. Adv. Funct. Mater. 2015, 25, 5677-5682.

(11) Chen, D. Z.; Jang, D.; Guan, K. M.; An, Q.; Goddard, W. a.; Greer, J. R.

Nanometallic Glasses: Size Reduction Brings Ductility, Surface State Drives Its Extent. Nano Lett. 2013, 13, 4462-4468.

(12) Liu, Y.; Liu, J.; Sohn, S.; Li, Y.; Cha, J. J.; Schroers, J. Metallic Glass Nanostructures of Tunable Shape and Composition. Nat. Commun. 2015, 6, 1-6.

(13) Nakayama, K. S.; Yokoyama, Y.; Wada, T.; Chen, N.; Inoue, A. Formation of Metallic Glass Nanowires by Gas Atomization. Nano Lett. 2012, 12, 2404-2407. (14) Nakayama, K. S.; Yokoyama, Y.; Ono, T.; Chen, M. W.; Akiyama, K.; Sakurai, T.; Inoue, A. Controlled Formation and Mechanical Characterization of Metallic Glassy Nanowires. Adv. Mater. 2010, 22, 872-875.

(15) Yi, J.; Xia, X. X.; Zhao, D. Q.; Pan, M. X.; Bai, H. Y.; Wang, W. H. Micro-and Nanoscale Metallic Glassy Fibers. Adv. Eng. Mater. 2010, 12, 1117-1122.

(16) Kumar, G.; Tang, H. X.; Schroers, J. Nanomoulding with Amorphous Metals.

Nature 2009, 457, 868-872.

(17) Schroers, J.; Nguyen, T.; O’Keeffe, S.; Desai, A. Thermoplastic Forming of Bulk Metallic glass— Applications for MEMS and Microstructure Fabrication.pdf. Mater. Sci. Eng. A 2007, 449-451, 898-902.

(18) Jeong, H. W.; Hata, S.; Shimokohbe, A. Microforming of Three-Dimensional Microstructures from Thin-Film Metallic Glass. J. Microelectromechanical Syst. 2003, 12, 42-52.

(19) Saotome, Y.; Fukuda, Y.; Yamaguchi, E; Inoue, a. Superplastic Nanoforming of Optical Components of Pt-Based Metallic Glass. J. Alloys Compd. 2007, 434-435,

97-101.

(20) Chu, J. P.; Wijaya, H.; Wu, C. W.; Tsai, T. R; Wei, C. S.; Nieh, T. G.; Wadsworth, J. Nanoimprint of Gratings on a Bulk Metallic Glass. Appl. Phys. Lett. 2007, 90, 3-6.

(21) Tuniz, A.; Lwin, R.; Argyros, A.; Fleming, S. C.; Pogson, E. M.; Constable, E.;

Lewis, R. a.; Kuhlmey, B. T. Stacked-and-Drawn Metamaterials with Magnetic Resonances in the Terahertz Range. Opt. Express 2011, 19, 16480-16490.

(22) Ibanescu, M.; Fink, Y.; Fan, S.; Thomas, E. L.; Joannopoulos, J. D. An All- Dielectric Coaxial Waveguide. Science (80- ). 2000, 289, 415-419.

(23) Canales, A.; Jia, X.; Froriep, U. P.; Koppes, R. a.; Tringides, C. M.; Selvidge, J.;

Lu, C.; Hou, C.; Wei, L.; Fink, Y.; et al. Multifunctional Fibers for Simultaneous Optical, Electrical and Chemical Interrogation of Neural Circuits in Vivo. Nat. Biotechnol. 2015, 33, 277-284.

(24) Canales, A.; Jia, X.; Froriep, U. P.; Koppes, R. a; Tringides, C. M.; Selvidge, J.;

Lu, C.; Hou, C.; Wei, L.; Fink, Y.; et al. Multifunctional Fibers for Simultaneous

Optical, Electrical and Chemical Interrogation of Neural Circuits in Vivo. Nat. Biotechnol. 2015, 2, 1-10.

(25) Yan, W.; Qu, Y.; Gupta, T. Das; Darga, A.; Nguyen, D. T.; Page, A. G.; Rossi, M.;

Ceriotti, M.; Sorin, F. Semiconducting Nanowire-Based Optoelectronic Fibers. Adv. Mater. 2011, 1700681, 1700681. (26) Lacour, S. P.; Courtine, G.; Guck, J. Materials and Technologies for Soft Implantable Neuroprostheses. Nat. Rev. Mater. 2016, 1, 16063.

(27) Leach, J. B., Achyuta, A. K. H. & Murthy, S. K. Bridging the divide between neuroprosthetic design, tissue engineering and neurobiology. Frontiers in Neuroengineering 2, (2010).

(28) Kozai, T. D. Y., Jaquins-Gerstl, A. S., Vazquez, A. L., Michael, A. C. & Cui, X.

T. Brain tissue responses to neural implants impact signal sensitivity and intervention strategies. ACS Chem. Neurosci. 6, 48-67 (2015).

(29) Etemadi, L. et al. Embedded ultrathin cluster electrodes for long-term recordings in deep brain centers. PLoS One 11, (2016).

(30) Polikov, V. S., Tresco, P. A. & Reichert, W. M. Response of brain tissue to chronically implanted neural electrodes. Journal of Neuroscience Methods 148, 1- 18 (2005).

(31) Deisseroth, K. Optogenetics: 10 years of microbial opsins in neuroscience. Nature Neuroscience 18, 1213-1225 (2015).

(32) Paxinos, G. & Watson, C. The rat brain in stereotaxic coordinates .

(33) E. Bryn Pitt, G. Kumar, J. Schroers, Temperature dependence of the thermoplastic formability in bulk metallic glasses, J. Appl. Phys. 110 (2011) 043518. doi: 10.1063/1.3624666.

(34) E. Pone, C. Dubois, N. Gu, Y. Gao, A. Dupuis, F. Boismenu, et al., Drawing of the hollow all-polymer Bragg fibers, Opt. Express. 14 (2006) 5838-5852. doi: 10.1109/ECOC.2006.4801421.

(35) S. Tomotika, On the Instability of a Cylindrical Thread of a Viscous Liquid Surrounded by Another Viscous Fluid, Proc. R. Soc. A Math. Phys. Eng. Sci. 150 (1935) 322-337. doi: l0.l098/rspa. l935.0l04.

(36) G. Kumar, J. Schroers, Write and erase mechanisms for bulk metallic glass, Appl.

Phys. Lett. 92 (2008) 031901. doi: l0T063/l.28347l2.

(37) D.S. Deng, J.C. Nave, X. Liang, S.G. Johnson, Y. Fink, Exploration of in-fiber nanostructures from capillary instability, Opt. Express. 19 (2011) 16273-16290. doi: 10.1364/OE.19.016273.

(38) S.F. Cogan, Neural stimulation and recording electrodes, Annu. Rev. Biomed. Eng.

10 (2008) 275-309. doi: l0.l l46/annurev.bioeng. l0.06l807T605 l8.

(39) A. Canales, X. Jia, U.P. Froriep, R. a Koppes, C.M. Tringides, J. Selvidge, et al., Multifunctional fibers for simultaneous optical, electrical and chemical interrogation of neural circuits in vivo, Nat. Biotechnol. 2 (2015) 1-10. doi: l0T038/nbt.3093. (40) S. Venkatraman, J. Hendricks, Z. a King, A.J. Sereno, S. Richardson-bums, D. Martin, et ak, In Vitro and In Vivo Evaluation of PEDOT Microelectrodes for Neural Stimulation and Recording, 19 (2011) 307-316.