WO2008044050A2

WO2008044050A2 - Nanostructures

Info

Publication number: WO2008044050A2
Application number: PCT/GB2007/003899
Authority: WO
Inventors: Max Whitby; Milo Sebastian Peter Shaffer
Original assignee: Imperial Innovations Limited; Rgb Research Limited
Priority date: 2006-10-12
Filing date: 2007-10-12
Publication date: 2008-04-17
Also published as: WO2008044050A3; GB0620335D0

Abstract

Elongate arsenic nanostructures, and methods of making them comprising the steps of forming an arsenic vapour and contacting said vapour with a metal catalyst under an inert atmosphere or under vacuum, at a suitable temperature are disclosed.

Description

NANOSTRUCTURES

The present invention relates to tubular and/or rod-like nanostructures formed from elemental arsenic and to methods of forming such structures.

Recent advances in nanotechnology have resulted in the formation of a huge range of nanoscale structures from a wide variety of different chemical elements and compounds. The elongate versions of these nanostructures in particular have stimulated a very active field of research, especially relating to the synthesis and characterisation of nanowires, nanorods, nanotubes, nanowhiskers, nanofibres, and nanoscale needles, to name but a few. In particular, semiconductor nanowires and nanotubes have been heavily investigated (Semiconductor Nanowires and Nanotubes, Law et al., Annu. Rev. Mater. Res. (2004), 34, 83-122). These elongate structures have the potential to be useful in an array of possible applications including as wires, switches, sensors, and transducers in nanoscale electronic circuits, as electron emitters in flat panel displays, and as probe tips in some forms of probe microscopy.

Another active field of nanostructure research involves the study of carbon nanotubes. Carbon nanotubes are well known; they have nanoscaled diameters and a structure that can be visualised as one or more layers of graphite rolled to form seamless cylinder(s). They may be synthesised by a number of different methods, including electric arc evaporation or laser ablation of graphite, and catalytic decomposition of organic vapours. The layered structure of graphite, with planes of carbon atoms held together by weak interplanar bonds, enables the formation of tubular structures, since there is a single low energy surface once the layer is curved into a cylinder.

These carbon nanotubes show either metallic or semiconducting properties depending upon exactly how the graphene sheet is rolled up. Existing methods of manufacture of carbon nanotubes have difficulty selectively forming either metallic or semiconducting nanotubes preferentially.

Due to their unusual electronic properties, extraordinary thermal conductivity and high tensile strength and flexibility, many applications of carbon nanotubes have been proposed, including use as components in nanoscale electronic circuits, capacitor/fuel cell electrodes, and transparent antistatic coatings. However, for a number of these applications, the controlled production of either metallic or semiconducting nanotubes is required or desired.

Synthesis of non-carbon nanotubes is also known. For example nanotubes formed from P, Bi, Sb, B_xC_yN_z, MoS₂, WS₂, TiO₂, NiCI₂, MoSe₂, NbS₂, GaN, InS, ZnS and V₂O₅ have all been described, although elemental nanotubes are rare. However, elemental phosphorus nanotubes and a method for their synthesis are described in co-pending PCT application number PCT/GB2006/001277 (published as WO2006/106349).

Elemental arsenic is known to form a wide range of triply-coordinated compounds in its +3 or -3 oxidation states. By analogy with three-coordinate carbon in graphite and in carbon nanotubes, it has been suggested that it may be possible to form nanotubes from elemental arsenic, as well as crystalline nanorods and other nanostructures. In graphite, the carbon atoms are linked in a hexagonal array to form flat sheets; arsenic also forms sheet-like structures which may be important for nanotube formation, although in some allotropes the hexagonal rings are puckered out of the plane of the sheet (D. Schiferl and CS. Barret, J. Appl. Crystallogr., 1969, 2, 30). Seifert and Frauenheim have calculated that nanotubes formed from elemental phosphorus should be stable and they have suggested that similar results may be expected for elemental arsenic (J. Kor. Phys. Soc, 37(2), 89 (2000)).

Arsenic nanotubes have been studied theoretically using an iterative procedure based on the Monte Carlo-Metropolis method to minimise the energy of possible tubular forms of arsenic (S. Zamfira et al., Chalcogenide Letters Vol. 2, No. 6, June 2005, p. 55 - 61 ). This study indicated that tubular structures of arsenic are reasonably stable and might be expected to exist, although the smaller diameter arsenic nanotubes are calculated to be more stable than the corresponding nanotubes of larger diameter.

Nanorod structures are also known in a wide variety of systems, particularly including oxides (SiO₂, TiO₂, SnO, ZnO, In₂Os), and semiconductor compounds (Si, SiGe_x, InAs, GaAs, InP). Synthesis methods include, sol-gel chemistry, vapour transport, solvothermal treatment, and the vapour-liquid-solid (VLS) mechanism. The VLS method has been known in the production of (relatively large) silicon nanowires since the 1960s. More recently the method has been developed for other semiconducting systems, mostly based on gold catalyst particles (summarised in X. Duan, Adv. Mater. 2000, 12, No. 4).

The thermodynamically most stable form of arsenic is the grey allotrope. The structure of grey arsenic can be likened to that of black phosphorus in that it forms infinite sheets in which each atom has three nearest neighbours to give an extended hexagonal lattice. However, there are small differences between the way in which the six membered rings within the sheets are fused in black phosphorus and grey arsenic (The Allotropy of the Elements, W. E. Addison, Oldbourne Press, London, p.102-103). This layered structure of arsenic suggests that it should be possible, at least from a structural point of view, for a single layer of arsenic atoms to roll up to form a tubular structure.

There is a suggestion that some allotropes of arsenic can exhibit semiconducting behaviour (Amorphous Arsenic, Greaves et a/., Advances in Physics, 1979, 28(1),

49-141 ) whereas others, e.g. the grey allotrope, are metallic or semi-metallic (of the

Chemical Elements, 1968, Publisher: Reinhold Book Corporation, Ed. C. A. Hampel).

These different behaviours indicate that arsenic nanostructures may have very interesting and potentially useful conductivity properties.

The present inventors have developed a controllable synthesis of arsenic nanostructures.

Unless otherwise specified, the term "nanostructure" is used herein to mean any structure having at least one dimension on the nanoscale.

Unless otherwise specified, the term "nanoscale" is used herein to mean any measurement larger than about 0.1 nm and smaller than about 1 μm.

In a first aspect, the present invention provides arsenic nanostructures. These nanostructures may be hollow nanotubes, solid nanorods, needles, platelet structures, helices, and/or filaments.

In a one aspect, the arsenic nanostructures are elongate nanostructures (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments). Preferably the elongate nanostructures are nanotubes and/or nanorods, more preferably nanotubes. In some embodiment nanorods are preferred. Where the arsenic nanostructures are rod-like structures ("nanorods"), they are solid in cross section. In some cases, e.g. where the nanorods are particularly long, they may be termed "nanowires".

Where the arsenic nanostructures are nanotubes, they have a channel inside them running substantially parallel to and preferably substantially along, the principle axis of the nanotube.

Where the arsenic nanostructures are needles, they have a high aspect ratio and a substantially linearly decreasing diameter towards the tip. Scanning electron microscope (SEM) investigation of the needle structures reveals no evidence of VLS growth (e.g. no catalyst particles can be seen at the tip of the structure). The VLS (vapour-liquid-solid) mechanism (as originally proposed by Wagner and Ellis in relation to Si whisker growth - R.S. Wagner and W.C. Ellis, App. Phys. Lett, 4(5), (1964), 89), is an established route to high aspect ratio structures, and results in a characteristic, solidified droplet ('catalyst') at one end of the structure. However, it is possible that in some cases these arsenic needles are formed by a crystal growth method which does not involve a catalyst.

Where the arsenic nanostructures are rods, they have a high aspect ratio with a substantially constant diameter along their length. Growth along the rod is generally straight but bends, kinks and branches are observed in some cases. SEM investigation reveals evidence of VLS-growth (e.g., a catalyst particle present at least one end of the rod).

The term 'substantially' as used herein means where the diameter is as defined within a range of no more than ±10%, preferably ±5% and more preferably ±1%.

Where the arsenic nanostructures are platelets, they are crystalline in appearance and have polygonal flat faces. These platelets may grow in layers forming a 3-dimensional polyhedral structure. SEM investigation of the platelets in the example presented below reveals some evidence of the platelets growing by a VLS mechanism.

Where the arsenic nanostructures are helices, they have a high aspect ratio. These helix nanostructures are coiled around a central axis with various pitch and coil radii about the central axis. Helical arsenic nanostructures may extend from the tip of an aforementioned rod structure or may exist independently. The helical nature of these structures may be due to screw dislocations in the growing structure.

Where the arsenic nanostructures are filaments, they have a spidery appearance when observed by a light microscope. Small droplets of metal associated with the filament structures are visible, implying a possible VLS growth mechanism. The filaments may be more amorphous than the rods or needles and may grow in many different directions (i.e. not form one continuous structure). In preferred aspects, the filament structures form preferentially at higher temperatures (e.g., 500-550⁰C).

Elongate arsenic nanostructures (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments) may have a uniform circular or polygonal cross-section extended prismatically along the axis of the structure. More complex structures, in which nanorods or nanotubes change diameter, geometry, twist, join or branch may be derived from the basic structures.

Elongate arsenic nanostructures may be terminated, usually at one end, by a catalytic particle. Also, nanotubes are often intrinsically capped at one or both ends wherein the walls of the arsenic nanotube bend inwards to form an end-cap closing the nanotube.

Where the arsenic nanostructures of the present invention are elongate nanostructures (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments), they preferably have a relatively high aspect ratio. The aspect ratio of the nanostructures is defined as:

Aspect ratio = length/diameter

In preferred aspects, the aspect ratio of the nanostructures is greater than 50, preferably greater than 100 and more preferably greater than 200 and may be up to or greater than 1000.

The arsenic nanostructures may take any elongate form (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments). They may be substantially straight or may be curved or twisted in any direction. Furthermore, they may be branched structures. Preferably, the arsenic nanostructures are substantially straight.

In the graphite structure, the hexagons of carbon atoms lie flat within each plane, and hence carbon nanotubes have a 'smooth' outer surface. Similarly, arsenic atoms in arsenic nanotubes are three coordinate and can form hexagonal rings which may be either flat or puckered. This leads to arsenic nanotubes which may have either a 'smooth' outer surface (similar to that of carbon nanotubes) or may have a 'rough' outer surface (similar to that of phosphorus nanotubes which computer modelling suggests have a 'rough' outer surface due to puckering in the hexagons of phosphorus atoms).

The walls of arsenic nanotubes can be thought of as being formed from an extended hexagonal lattice of arsenic atoms, as described above, rolled substantially into a cylinder.

Defects may occur in the substantially cylindrical nanotube walls due to the presence of rings of arsenic atoms having more or less than six members, for example, 4, 5, 7 or 8 members, in the hexagonal lattice. These defects can result in, for example, changes in the direction of propagation of the nanotube, changes in the diameter of the nanotube along its length or can provide point defects at which the physical properties of the nanotube, such as conductivity or chemical reactivity, may be different from the rest of the nanotube. These defects may also provide for closure of the nanotube through the formation of conical or hemispherical caps. Alternatively the ends of the nanotubes may remain open.

The arsenic nanotubes may be formed from a single wall of arsenic atoms (single-walled nanotubes - SWNTs) or may have multiple walls of arsenic cylinders arranged concentrically inside each other in a "Russian-doll" formation (multiple-walled nanotubes - MWNTs). The MWNTs may have any number of walls, for example less than 10 walls, or between about 10 and about 20 walls, or more than about 20 walls (e.g. about 50 walls). Alternatively, the nanotubes may be formed from a single layer of arsenic atoms rolled to have a spiral arrangement in cross-section. Preferably the nanotubes have either a single wall or multiple walls arranged inside each other. More preferably, the nanotubes are single-walled tubes. The diameter of the nanostructures of the present invention may vary both between samples and within a given sample. However, the diameter of the nanostructures preferably lies in a range. The diameter range of the nanostructures may vary depending on whether the morphology of the structures. For example, SWNTs are likely to have a smaller average diameter than MWNTs.

For single-walled tubes, the nanotubes preferably have a diameter of between about 0.5 and 10nm, preferably 0.5-5nm, preferably 1-5nm.

For multiple-walled tubes, the nanotubes preferably have an external diameter of between about 2-200nm, preferably 5-100nm, preferably 5-50nm. The internal diameter (i.e., the diameter of the cavity inside the MWNT) is preferably between about 0.5 and 50nm, preferably 0.5-1 Onm, preferably 1-5nm.

The solid elongate nanostructures of the present invention (e.g. nanorods, nanowires, needles, helices, and filaments) may exist with a wide range of diameters. The lower limit of this range is preferably about 1nm, preferably about 5nm, preferably about 10nm or preferably about 50nm, or maybe about 1OOnm. The upper limit of the range is preferably about 5μm, preferably about 1 μm, preferably about 200nm, more preferably about 100nm or may be about 50nm.

In the above discussion of the diameters of the elongate nanostructures according to the present invention, any of the upper and lower values for the diameter range may be independently combined, i.e. the diameter range may have any one of the above mentioned lower limits and, independently, any one of the above mentioned upper limits. For example, the preferred range of diameters for a nanorod may be between about 50nm and 200nm.

Where the arsenic nanostructure is an elongate arsenic nanostructure, the diameter of the structure may determine whether it has a cavity or channel through it (i.e., it is a nanotube) or not (i.e., it is a solid nanostructure e.g. a nanofibre). Where the , nanostructure has a small diameter (e.g. less than about 10nm, or less than about 7nm, or less than about 5nm, or less than about 3nm), the relationship between the unfavourable strain energy of the structure (due to the high curvature caused by the low diameter) and the favourable surface energy (due to the formation of a continuous crystal structure and the elimination of so-called "dangling bonds" at the edges of a sheet structure) cause the nanostructure to take the form of a tube rather than a solid fibre or sheet structure. Therefore, in some cases, the observation of small diameter (e.g. less than about 10nm, or less than about 7nm, or less than about 5nm, or less than about 3nm) elongate arsenic nanostructures may indicate the presence of arsenic nanotubes (as opposed to solid nanostructures).

Individual arsenic elongate nanostructures (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments) may have different sections along their length which take the form of different elongate nanostructures. For example, an individual elongate nanostructure may have along its length a section that takes the form of a nanorod and a separate section which take the form of a nanotube.

The properties of the arsenic nanotube may change depending on how the 'sheet' of arsenic atoms is rolled up, i.e. which crystallographic vector in the plane of atoms lies parallel to the axis of the nanotube. In carbon nanotubes, this characteristic is known as chirality or helicity and it is known to strongly effect (opto)electronic properties of the nanotube. Therefore, with arsenic nanotubes of the present invention, this chirality or helicity may similarly define some electronic properties of the nanotube. The arsenic nanotubes may show metallic, semi-metallic, or semiconducting behaviour, and preferably show semiconducting behaviour.

Any of the elongate nanostructures (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments) of the present proposals may show metallic, semi-metallic, or semiconducting behaviour.

In one aspect, the arsenic nanostructures are air-stable in that they do not decompose in air, although an oxide layer may form on the outside of the nanostructures.

The arsenic nanostructures may exist alone or may be present with extraneous material which is not in the form of nanostructures. In a second aspect, the present proposals relate to a material containing greater than 5%, preferably greater than 10% or greater than 20% or greater than 30%, more preferably greater than 50% or greater than 70% or greater than 80% and maybe up to 95% elongate arsenic nanostructures. Preferably said extraneous material is bulk arsenic. Said extraneous material may comprise residual catalyst or unreacted starting materials or a side-product of the synthesis method. Furthermore, the second aspect may relate to arsenic, containing above 10%, or above 25%, maybe above 50% and advantageously above 75% or above 90% arsenic nanostructures.

In the second aspect of these proposals, the arsenic material present which is not present as elongate nanostructures may be present as any allotrope of arsenic, such as the stable silver-grey crystalline form, yellow arsenic, or black arsenic.

In a third aspect, the present invention provides a method for forming elongate arsenic nanostructures comprising the steps of forming an arsenic vapour and contacting said vapour with a metal catalyst under an inert atmosphere or under vacuum, at a suitable temperature.

By "inert" atmosphere is meant an atmosphere having a reduced reactivity compared to air to the reactants and intermediates in the method for forming elongate arsenic nanostructures and to the nanostructures themselves. Preferably this is a reduced-oxygen atmosphere, such as Ar gas. Preferably an "inert" atmosphere also has a reduced water content compared to air. More preferably an "inert" atmosphere as used herein is a reduced-oxygen atmosphere with a reduced water content compared to air, such as dry Ar gas.

In preferred methods, the concentration of oxygen in the inert atmosphere is less than 1%, preferably less than 0.1% and more preferably less than 0.01% by volume. The inert atmosphere may be any unreactive gas and may be selected from argon, carbon dioxide, nitrogen, helium, sulphur hexafluoride or a mixture of any two or more of these. Furthermore, the reaction may be performed under reduced pressure, for example less than 10^"2, less than 10^"4 or less than 10^'6 mbar.

Preferably, the inert atmosphere contains less than 1%, preferably less than 0.1%, more preferably less than 0.01 % water by weight.

The preferences given for the inert atmosphere preferably relate to the atmosphere prior to reaction i.e. as put into the reaction vessel.

In some embodiments of the third aspect of the present proposals, the arsenic vapour formed in the method used to synthesise elongate arsenic nanostructures may be any vapour containing arsenic atoms. The vapour containing arsenic atoms may be formed by vaporisation of any allotrope of arsenic, including silver-grey metallic arsenic, yellow arsenic, and black arsenic. Preferably, the vapour containing arsenic atoms is formed by vaporisation of metallic arsenic. In other embodiments of the third aspect of the present proposals the vapour may also be a CVD (chemical vapour deposition) source gas containing arsenic, such as arsine (AsH₃), and alkyl arsines (e.g. monoalkyl arsines, dialkyl arsines and trialkyl arsines).

Furthermore, in the methods of the third aspect of these proposals, the metal catalyst may be either solid or liquid, and is preferably liquid, at the synthesis temperature and pressure. More preferably, the metal catalyst is liquid when saturated with arsenic at the synthesis temperature and pressure.

The metal catalyst may be any metal but is advantageously a metal or alloy in which arsenic is at least slightly soluble; under growth conditions of temperature and arsenic concentration the catalyst metal or alloy is advantageously in its liquid form. More preferably the arsenic saturated catalyst metal or alloy is in thermodynamic equilibrium with solid elemental arsenic,_, at the synthesis temperature. Ideally this equilibrium should exist over a wide range of temperatures and metal/arsenic ratios. Preferably arsenic does not readily react with the catalyst to form intermetallic or other compounds. Most preferably, the catalyst metal is selected from one or more of the following non-limiting group of metals, mercury, bismuth, and antimony. Furthermore, in one aspect, the catalyst is a mixture or alloy comprising one or more of the following non-limiting group of metals, mercury, bismuth, lead, and antimony.

The metal catalyst may be present as one or more fragments, a melt, a vapour, or may be finely divided solid or molten particles or droplets any of which may be dispersed on a high surface area or functional support.

Suitable high surface area supports may include silica, alumina, zirconia, zeolites, glass wool, quartz wool, aerosil™, aerogel, dispersed silica, carbon black, and other fumed or sol-gel derived oxides.

Functional supports may include wafers for electronics applications, such as silicon (e.g., single crystal silicon), sapphire, GaAs, InP or GaP.

Preferred supports are selected from glass wool, quartz wool, and silicon (e.g., single crystal silicon). The method of the third aspect of the invention may also include arrangement of catalyst particles on the support material in order to direct the growth of the arsenic nanostructures to certain preferred regions. This arrangement may include patterning of the catalyst on the surface of the support (e.g., forming a pattern or array of bismuth catalyst particles on a silicon surface) which may be achieved either by selective deposition of catalyst particles onto certain areas or positions of the support surface, or may be achieved by forming a layer of catalyst on the support surface and then etching away selected regions of the catalyst layer to leave catalyst regions in the desired pattern or array. For example, this method could be utilised to grow an array of nanostructures upstanding from a support surface in a desired pattern, or could be used to direct the growth of the nanostructures in order to connect regions of the support surface (e.g., to act as wires in a nanoscale electronic circuit).

Preferably the method of the third aspect of the invention is performed at elevated temperature. Advantageously, the method is performed under temperature and pressure conditions at which the rate of growth of the arsenic nanostructures is greater than their rate of evaporation. In preferred aspects, the method is performed at above 100°C, preferably at above 250°C and more preferably at above 350°C. Preferably the method of the third aspect of the proposals is performed at up to 1000⁰C and maybe higher, preferably up to 800⁰C, preferably up to 600⁰C, preferably up to 550⁰C. Preferably the method of the third aspect is performed at about 400°C +/- 50°C, preferably 400⁰C +/- 20°C.

In preferred methods, the ratio of metal catalyst to arsenic (from which the arsenic vapour is formed) in the reaction vessel is as low as possible whilst still remaining effective to produce the desired arsenic nanostructures. This ensures growth of arsenic nanostructures with a minimum catalyst contamination of the final product. Preferably the ratio of metal catalyst to arsenic is between 1 to 1 and 1 to 1000, preferably to the lower end of this range such as between 1 to 100 and 1 to 1000, or between 1 to 500 and 1 to 1000 or may be between 1 to 800 and 1 to 1000 by weight. The concentration of metal catalyst present in the reaction vessel may be as low as 0.1-1 wt.%, and is preferably about 1-5 wt. %. However, nanostructures may also be formed with ratios of metal catalyst to arsenic between 1 to 1 and 1 to 100 and maybe between 1 to 1 and 1 to 50 or between 1 to 5 and 1 to 10 by weight, . Advantageously, the reaction is performed in a sealed vessel, e.g. a sealed glass ampoule or a sealed bomb apparatus. However, in some embodiments, the method is carried out in a flow apparatus.

In a fourth aspect, the present invention provides arsenic nanostructures obtainable by the methods of the third aspect.

Figures

Fig. 1 is a reacted ampoule following the method of the third aspect of the invention containing arsenic, glass wool, bismuth catalyst and a silicon wafer.

Fig. 2 is an optical micrograph of the reacted silicon wafer on which a rod-like structure of the invention is visible.

Fig. 3 is an optical micrograph of a catalyst cluster with high aspect ratio structures of the invention extending outwards from it. Fig. 4 is an optical micrograph of a multi-faceted platelet of the invention grown in the reaction ampoule.

Fig. 5 is an optical micrograph of a rod structure terminated in a helical structure, both of the invention.

Fig. 6 is an optical micrograph of a tangle of filament-like material of the invention found in a reaction ampoule.

Fig. 7 is a SEM image of a sample of the reaction product of the invention.

Fig. 8 is a SEM image of an arsenic elongate nanostructure according to the present invention.

Fig. 9 is a graph representing the Energy Dispersive X-ray microanalysis (EDX) data acquired from the body of the nanostructure in fig. 8.

Fig. 10 is a graph representing the EDX data acquired from the terminal droplet in fig.

8.

Example 0.41 g of metallic arsenic was washed with hydrogen peroxide (3x2ml) followed by distilled water (3x2ml) and finally ethanol (3x2ml) before being dried under vacuum in the presence of diphosphorus pentoxide for a period of an hour. The washed arsenic was added to a glass ampoule followed by a plug of borosilicate glass wool (fig. 1)

A crystalline sample of 99.99% bismuth metal (Zhuzhou Kete Metals Test Works, PRC.) was hammered on a steel anvil to obtain small particles of bismuth. 0.1g of this powder was finely applied to a silicon wafer (111 face) and heated at 380 degrees Celsius for ten minutes on a hotplate with a copper conductor in order to wet the surface of the silicon with molten bismuth. The droplets of bismuth can still be seen after reaction in figure 2 as described below.

The bismuth wetted silicon wafer was added to the ampoule under argon atmosphere and the ampoule was flame sealed under argon.

The sealed ampoule was placed in a steel bomb and the temperature was ramped up to 400 degrees Celsius over the course of an hour. The steel bomb was held at 400 degrees Celsius for 15 hours and the temperature was ramped down to room temperature over 4 hours.

Examination of the ampoule showed fine grey deposits on the silicon wafer surface around metallic appearing globules. Further examination under an optical microscope gave evidence of high aspect ratio features (figs. 2 - 6)

Samples were prepared for study by scanning electron microscope (SEM) by the following method:

The reaction ampoule was cracked open under an inert atmosphere and the silicon wafer within is transferred to a standard aluminium SEM stub. The stub was then placed in the viewing chamber of the SEM and imaged.

The samples were studied using a JEOL 201 OFX, fitted with an Oxford Instrμments EDX detector.

To prepare samples for Scanning Electron Microscopy; the reaction ampoule was sealed in a glove box containing an argon rich atmosphere (O₂ < 0.1%) and using pliers was delicately cracked. The contents were then sealed in separate vials under argon. Immediately prior to the microscopy study, the silicon wafer was transferred to an ordinary aluminium SEM stub and fixed using sticky carbon tape. The stub was fitted into a six-sample round-about style sample holder which was then placed in the viewing chamber.

To prepare samples for optical microscopy; the reaction ampoule is handled using nitrile gloves and attached to a glass slide. The optical microscope was an XSZ-20 series biological microscope with an eyepiece magnification of 16X and 4 objectives (4X, 10X, 4OX and 100X) giving a range of optical power from 64X to 1600X. The slide is placed on the instrument stage and the structures viewed directly through the glass wall of the ampoule. The sample is lit using a Fiber-Lite™ light pump and the stage backlight. Optical micrographs are recorded using a Nikon Coolpix™ 4500 camera which is mounted to the microscope using an adaptor. A further 4X optical power is available using the zoom function of the camera.

Fig. 1 shows a reaction ampoule containing metallic arsenic 1 , a plug of borosilicate glass wool 2, and a silicon wafer 3, after heating at 400⁰C for 15 hours. A deposit of arsenic metal 4 can be clearly seen coating the inner surface of the ampoule at the end of the ampoule containing the silicon wafer.

Fig. 2 shows the silicon wafer 3, after reaction, as seen under a light microscope. The surface of the silicon wafer is decorated with droplets of bismuth metal catalyst 5. An arsenic elongate nanostructure 6 can be seen in the centre of the figure projecting from a catalyst metal particle 7.

Fig. 3 shows a product of a reaction according to the present proposals as seen under a light microscope. A catalyst particle 8 can be seen with arsenic elongate nanostructures 9 projecting from it.

Fig. 4 shows a product of a reaction according to the present proposals as seen under a light microscope. A multi-faceted nanoscale platelet 10 can be seen in the product.

Fig. 5 shows a product of a reaction according to the present proposals as seen under a light microscope. An elongate structure 11 can be seen which comprises a straight section 12 and a helical section 13. The straight section 12 of this structure is thought to be an arsenic nanotube or nanorod with a helical nanostructure 13 joined on to one end of it. This demonstrates that the nanostructures according to the present proposals may be composite structures being made up from different possible nanostructural forms (e.g., hollow nanotubes, solid nanorods, needles, helices, and/or filaments).

Fig. 6 shows a product of a reaction according to the present proposals as seen under a light microscope. The reaction to produce this product was performed at a higher temperature (55O⁰C) than those producing the products shown in figs. 2-5. This figure shows a tangle of nanoscale filaments 14. Note: the elongate structures 15 are fragments of the borosilicate glass wool support on which the product was grown.

Fig. 7 shows a product of a reaction according to the present proposals as seen under a scanning electron microscope (SEM). Straight elongate structures 16 can be identified which may be any of solid nanorods, hollow nanotubes or needles. As the SEM technique only reveals the surface structure of the product, it is not possible to tell whether the structure contains a hollow cavity (i.e., a nanotube) or is solid throughout (i.e., a nanorod).

The reaction product in fig. 7 also shows examples of platelet structures 17 and elongate structures which abruptly change direction 18. These abrupt direction changes may be the result of the introduction of arsenic rings in the wall of the structure which have more or less than six members. The introduction of such rings leads to a change in the angle of propagation of the nanostructure.

Fig. 8 shows a product of a reaction according to the present proposals as seen under a scanning electron microscope (SEM). This product shows the end of an elongate nanostructure 19 having a body section 20 and a head section 21. As mentioned above, it is not possible to tell whether or not the body section has an internal cavity using the SEM technique. In this case, the nanostructure has a diameter of about 800nm so it seems likely that it is a solid nanorod or nanowire.

Fig. 9 shows the results of an EDX scan of the body section of the elongate nanostructure shown in fig. 8. This scan clearly shows a very large arsenic signal along with a much smaller signal that is characteristic of the bismuth catalyst used in the reaction. This strongly indicates that the body of the nanostructure is formed from elemental arsenic.

Note: the EDX technique relies on focussing a probe on an area of sample to be analysed. Whilst the centre of the probe is focussed over the area of interest (in this case the body section of the structure), some signal is also obtained from the periphery of the probe region. This explains the presence of the signal from the bismuth catalyst, because it was difficult to focus the probe solely on the body section of the structure. Fig. 10 shows the results of an EDX scan of the head section of the elongate nanostructure shown in fig. 8. This scan clearly shows a very large bismuth signal along with a much smaller signal that is characteristic of arsenic. This strongly indicates that the head section of the nanostructure is formed from bismuth catalyst.

The combination of the EDX results shown in figs 9 and 10 indicates that the nanostructure shown in fig. 8 has a body section composed of arsenic and a head section composed of bismuth catalyst. This strongly suggests a vapour-liquid-solid growth mechanism whereby the arsenic body section has grown from a molten bismuth catalyst droplet. This mechanism is well known for carbon nanotubes.

From observations made using a high-powered optical microscope, the arsenic nanostructures appear to air-stable. Indeed, the SEM pictures were achieved after transferring the material in air to the viewing chamber. Optically visible structures appear to be unchanged after ca. 4 months of oxygen exposure, although it is thought that an oxide layer may be present on the outside of the nanostructures.

Claims

I . Elongate arsenic nanostructures.

2. Nanostructures according to claim 1 , which are selected from solid nanorods, helical nanostructures, and needle nanostructures.

3. Nanostructures according to claim 2, which are solid nanorods.

4. Nanostructures according to any one of claims 1 to 3, wherein the structure is terminated at one end by a catalytic particle.

5. A material containing greater than 5%, of elongate arsenic nanostructures according to any one of claims 1 to 4.

6. A material according to claim 6, containing greater than 50%, of elongate arsenic nanostructures according to any one of claims 1 to 4.

7. A material according to either claim 5 or claim 6, wherein the remainder of the material comprises bulk arsenic.

8. A method for forming elongate arsenic nanostructures of any one of claims 1 to 4, comprising the steps of forming an arsenic vapour and contacting said vapour with a metal catalyst under an inert atmosphere or under vacuum, at a suitable temperature.

9. A method according to claim 8, wherein the inert atmosphere is a reduced-oxygen atmosphere.

10. A method according to either claim 8 or claim 9, wherein the inert atmosphere is argon gas.

I I . A method according to any one of claims 8 to 10, wherein the arsenic vapour is formed by vaporisation of metallic arsenic.

12. A method according to any one of claims 8 to 10, wherein the arsenic vapour is a CVD source gas containing arsenic.

13. A method according to any one of claims 8 to 12, wherein the metal catalyst is liquid at the synthesis temperature.

14. A method according to any one of claims 8 to 13, wherein metal catalyst is selected from: mercury, bismuth, lead, and antimony.

15. A method according to any one of claims 8 to 14, wherein the method is performed at between 350 and 450°C.

16. Elongate arsenic nanostructures obtainable by the method of any one of claims 8 to 15.