WO2025262569A1 - Ceramic nanocomposite material - Google Patents

Abstract

A ceramic nanocomposite material comprising a mixture of boron suboxide with boron carbide, and corresponding method of manufacture for the ceramic nanocomposite material.

Description

CERAMIC NANOCOMPOSITE MATERIAL

FIELD OF THE INVENTION

The invention relates to a novel ceramic nanocomposite material, in particular to a material suitable for use, amongst other applications, as a ballistic armour material, as well as a neutron-irradiation resistant, hard wearing refractory ceramic, and a corresponding method of manufacture thereof.

BACKGROUND TO THE INVENTION

It is known that boron suboxide (BeO) is one of the most stable composites of B and 0. The crystal structure of BeO has an a-rhombohedral boron as its basic framework, i.e., eight boron icosahedra (B12) located at the apexes of the R3m unit cell, then paired with two oxygen atom chains (0-0) inserted in the interstices along the rhombohedral direction. The short and strong covalent bonds provide BeO with excellent physicochemical properties, including high hardness, low density, high thermal conductivity, high chemical stability, excellent wear resistance, high tensile strength, as well as large values of bulk modulus and bandgap. Although the synthesis of BeO powder is relatively simple and has been mass-produced, densified BeO ceramics possessing improved mechanical properties have not yet been available due to the poor sinterability of the BeO powder.

The mechanical properties of ceramics are closely related to sintering densification, where higher densification denotes better mechanical properties. For BeO, it is extremely challenging to obtain dense sintering of the powder due to its low diffusion coefficient and relatively high vapour pressure at the densification temperature. In short, BeO is a material typically hard to sinter at atmospheric pressure. This is especially so since atmospheric pressure is extremely unfavourable for the sintering of BeO nanocrystalline and submicron crystalline ceramics, where full densification is difficult to accomplish, and excessively high sintering temperatures will lead to rapid grain growth. Applying pressure during the sintering process can trigger rearrangement, sliding and plastic deformation of ceramic powder particles as well as promoting pore contraction, and is beneficial for the densification of ceramics. BeO ceramics prepared by hot-pressure and spark plasma sintering have higher densification (98.2-99.1 %) and enhanced mechanical properties (HV = 35-40 GPa, Kic = 1.3-3.1 MPa m^{0 5}) compared to those prepared by pressure-less sintering. In recent years, there has been some progress in enhancing the sinterability and fracture toughness of BeO by introducing second phases, but inevitably, the introduction of these second phases can potentially lead to a loss of hardness and increase in density of the synthetic ceramics.

US11274066B1 discloses a method of making a ceramic matrix composite (CMC) part in which a mixture, including a preceramic polymer, particles such as ceramic microparticles and/or nanoparticles, and organic compounds such as a surfactant and a solvent, are mixed to form a paste and printed or moulded. The CMC can contain silicon carbide, boron carbide, boron suboxide, alumina, or any other ceramic.

It has now been discovered that it is possible to prepare a ceramic nanocomposite material, a Carbon Reinforced Boron-subOxide Nanocomposite (CaRBON) material, which overcomes the problems of the prior art. Such a material is envisaged to be suitable for a range of uses including, but not limited to, ballistic armour applications, industrial abrasives, neutron moderation and absorption applications, or other hypothetical uses, such as in supercapacitance and hydrogen storage technology owing to the interesting physical and chemical properties of boron-based compounds.

SUMMARY OF THE INVENTION

A first aspect of the invention relates to a ceramic nanocomposite material comprising a mixture of boron suboxide with boron carbide.

Ceramic nanomaterials according to the invention exhibit optimal mechanical properties over pure BeO when made under similar conditions, including the potential for ultrahigh hardness at low density. Furthermore, it is possible to prove that the material’s chemical properties under local deformation are different to pure boron carbide (B4C), which undergoes a phenomenon called ‘shock amorphisation’ (whereby it’s crystal structure, under dynamic loading, collapses into a more dense, amorphous form, thereby effecting resultant mechanical properties). Finally, CaRBON is able to demonstrate resilience against damage and irradiation caused by fast neutron bombardment over a two-week period, exhibiting no signs of damage and a radiation level of ~0.2 mSv (approximately equivalent to that of two bananas) after the test period.

Preferably, the boron carbide is derived from graphene oxide. Graphene oxide features multiple hydrogen-bonding functional groups on the basal hexagonal carbon layers of the material. This means it is possible to readily prepare specific masses of the substance and blend them intimately with other materials in aqueous suspension, which is straightforward to process with minimal chemical, supply, and process-based risks. Preferably, the graphene oxide is reduced graphene oxide. More preferably, the graphene oxide thermally decomposes under the elevated temperature conditions inherent to ceramics sintering in order to form reduced graphene oxide. Reduced graphene oxide is the easiest form of nano-carbon to be homogeneously and readily dispersed in other suspended materials with water.

Preferably, the amount of graphene oxide is present in an amount from 0-5 wt.%. This allows a sufficient amount of carbon-containing material to be present within the material blend so as to allow it to react with the surrounding BeO matrix and aid the formation of the composite, as well as achieve densification, whilst also permitting the addition of potential surplus levels of rGO if it’s decided that doing so may provide benefit to the resultant composite when used in a specific application. Preferably, the amount of graphene oxide is present in an amount of at least 0.125 wt.%, preferably at least 0.25 wt.%, preferably at least 0.375 wt.%, preferably at least 0.5 wt.%, preferably at least 0.625 wt.%, preferably at least 0.75 wt.%, preferably at least 0.875 wt.%, preferably at least 1 .0 wt.%, preferably at least 1 .125 wt.%, preferably at least 1 .25 wt.%, preferably at least 1 .375 wt.%, preferably at least 1 .5 wt.%, preferably at least 1 .625 wt.%, preferably at least 1 .75 wt.%, preferably at least 1 .875 wt.%, preferably at least 2.0 wt.%, preferably at least 2.125 wt.%, preferably at least 2.25 wt.%, preferably at least 2.375 wt.%, or preferably at least 2.5 wt.%.

Preferably, the amount of graphene oxide is present in an amount of at most 2.5 wt.%, preferably at most 2.625 wt.%, preferably at most 2.75 wt.%, preferably at most 2.875 wt.%, preferably at most 3.0 wt.%, preferably at most 3.125 wt.%, preferably at most 3.25 wt.%, preferably at most 3.375 wt.%, preferably at most 3.5 wt.%, preferably at most 3.625 wt.%, preferably at most 3.75 wt.%, preferably at most 3.875 wt.%, preferably at most 4.0 wt.%, preferably at most 4.125 wt.%, preferably at most 4.25 wt.%, preferably at most 4.375 wt.%, preferably at most 4.5 wt.%, preferably at most 4.625 wt.%, preferably at most 4.75 wt.%, preferably at most 4.875 wt.%, or preferably at most 5.0 wt.%.

Most preferably, the amount of graphene oxide is present in an amount of less than 1 wt.%. These conditions were empirically determined to be the most conducive to the formation of more densified, more mechanically robust and adequately reacted composite samples.

Preferably, the ceramic nanocomposite material has a density greater than 90% of the composite’s theoretical density, compared to pure boron suboxide. This is due to the exponential levels of contribution that densification provides to both the mechanical properties of the composite after sintering, as well as to the assurance of adequate levels of chemical reaction between its constituent components. Preferably the ceramic nanocomposite material has a density greater than 90.5%, preferably greater than 91 %, preferably greater than 91 .5%, preferably greater than 92%, preferably greater than 92.5%, preferably greater than 93%, preferably greater than 93.5%, preferably greater than 94%, preferably greater than 94.5%, preferably greater than 95%, preferably greater than 95.5%, preferably greater than 96%, preferably greater than 96.5%, preferably greater than 97%, preferably greater than 97.5%, preferably greater than 98%, or preferably greater than 98.5% of the composite’s theoretical density, compared to pure boron suboxide. Most preferably, the ceramic nanocomposite material has a density greater than 98.5% of the composite’s theoretical density, compared to pure boron suboxide. This is because, in many structural applications, materials exhibit optimised properties when produced to theoretically near-full levels of density, accounting for possible errors in rounding (to the nearest integer) upon sample production.

Preferably, the ceramic nanocomposite material has a radioactivity of less than 1 mSv upon exposure to 2 x 4 consecutive day periods over a two-week period of a neutron fluence of 39.6 ± 3.41 n/cm² with energies of 13.7 ± 1 .5 MeV. This ensures that, under such conditions, the material exhibits a level of radioactivity that would still otherwise be deemed, according to known universal standards, to be legally and practically safe for exposure. For reference, 1 mSv (equivalent to 10 bananas) is an entirely legally permissible level of irradiation and takes into account of any wild fluctuations from the tested "norms" derived from experiments that may ever arise under similar circumstances due to any unknown additional factors (e. g. sample size, etc.). For further reference, the average radiation exposure of a person in the UK is ~2.7 mSv/year.

Preferably, the ceramic nanocomposite material has a Vickers hardness of greater than 4 GPa as measured by ASTM 384-11e1 Standard Test for Knoop and Vickers Hardness of Materials, preferably greater than 5 GPa, preferably greater than 6 GPa, preferably greater than 7 GPa, preferably greater than 8 GPa, preferably greater than 9 GPa, preferably greater than 10 GPa, preferably greater than 11 GPa, preferably greater than 12 GPa, preferably greater than 13 GPa, preferably greater than 14 GPa, preferably greater than 15 GPa, preferably greater than 16 GPa, preferably greater than 17 GPa, preferably greater than 18 GPa, preferably greater than 19 GPa, or preferably greater than 20 GPa. This ensures that the composite has a sufficient level of hardness for a wide range of various applications that it may see in practical use, whilst ensuring enough mechanical robustness to allow it to survive various means through which it might be handled (e.g. logistically or in machining and forming).

More preferably, the ceramic nanocomposite material has a Vickers hardness of greater than 30 GPa. This is to ensure that the composite provides the optimal level of mechanical and physical behaviour that would likely be required from it in its intended applications.

Preferably, the ceramic nanocomposite material is suitable for use in ballistic armour applications. This is because this material is likely to exhibit both optimised physical properties over existing armour ceramics, particularly with respect to hardness and low-density, whilst in practice suffering little to none of their respective disadvantages when in use. A second aspect of the invention relates to a method of making a ceramic nanocomposite material, the method comprising the steps of: mixing boron suboxide with reduced graphene oxide into an aqueous suspension with deionised water so as to produce a blended substrate; filtering, drying and de-agglomerating the blended substrate; and consolidating the blended substrate using spark plasma sintering (SPS) so as to produce the ceramic nanocomposite material.

Preferably, the boron suboxide and reduced graphene oxide are mixed into an aqueous suspension for up to 0.25 hr, preferably up to 0.5 hr, preferably up to 0.75 hr, preferably up to 1 hour, preferably up to 1 .25 hr, preferably up to 1 .5 hr, preferably up to 1 .75 hr, or preferably up to 2 hr. This ensures that the two materials have had sufficient time to not only be mixed together homogenously but undergo sufficient levels of the electrostatic coagulation that is most likely thought to take place between them when in suspension.

Preferably, the blended substrate is dried for at least 2 hours at 125 °C. This ensures that the substrate is sufficiently dry to undergo deagglomeration. In so doing, any further levels of intended agglomeration, or lack thereof, can be intentionally re-incorporated into the material depending on the subsequent product or application it may be used for.

Preferably, the step of SPS comprises holding the substrate in a vacuum at less than 10’² bar. This ensures both an adequate level of pore closure within the ceramic during sintering, as well as sufficient extraction of any byproducts formed from material outgassing processes taking place during these consolidation stages.

Preferably, the step of SPS further comprises applying a minimum contact force of 1 kN to the substrates, preferably a minimum contact force of 2 kN, preferably a minimum contact force of 3 kN, preferably a minimum contact force of 4 kN, or preferably a minimum contact force of 5 kN to the substrates. This ensures that a minimum level of electrical contact is made between the SPS press and the sample during sintering. Preferably, the step of SPS further comprises heating the substrates from room temperature to a peak temperature and being held at the peak temperature for at least 1 minute. This ensures that the material composite can be sufficiently densified for any potential application that it may be used for.

Preferably, the peak temperature is at least 1000 °C, preferably at least 1050 °C, preferably at least 1100 °C, preferably at least 1150 °C, preferably at least 1200 °C, preferably at least 1250 °C, preferably at least 1300 °C, preferably at least 1350 °C, preferably at least 1400 °C, preferably at least 1450 °C, preferably at least 1500 °C, preferably at least 1550 °C, preferably at least 1600 °C, preferably at least 1650 °C, , preferably at least 1700 °C, preferably at least 1750 °C, or preferably at least 1800 °C. More preferably, the peak temperature is at least 1300 °C. Most preferably, the peak temperature is at least 1570 °C. This provides comprehensive cover to all temperature ranges required to ensure an adequate degree of material densification, and/or potential reaction, for any application that it may be used for.

Preferably, prior to reaching, or upon reaching the peak temperature, the minimum contact force is increased to at least 10 kN, preferably at least 11 kN, preferably at least 12 kN, preferably at least 13 kN, preferably at least 14 kN, preferably at least 15 kN, preferably at least 16 kN, preferably at least 17 kN, preferably at least 18 kN, preferably at least 19 kN, preferably at least 20 kN, preferably at least 21 kN, or preferably at least 22 kN. This provides comprehensive cover to all pressure ranges required to ensure an adequate degree of material densification, and/or potential reaction, for any application that it may be used for.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be described in more detail, by way of example, with reference to the accompanying drawings in which:

Figure 1 shows a hypothetical caged B4C nanostructure within BeO;

Figure 2 shows different proposed sintering processing routes for CaRBON;

Figure 3 shows Vickers Hardness values for samples as a function of GO content; Figure 4 shows Vickers Hardness values for samples including BeO sintered at 1850 °C;

Figure 5 shows SEM micrographs of BeO sintered at 1570 °C (top) and 1850 °C (bottom);

Figure 6 shows SEM micrographs of BeO + 1wt.% GO sintered at 1570 °C for 15 mins (top) and 30 mins (bottom);

Figure 7 shows the XRD patterns for sintered B4C and CaRBON samples (29: IQ- 90⁰, 0.16167s);

Figure 8 shows FTIR spectra for sintered CaRBON samples (wavenumber: 500- 3500 cm-¹);

Figure 9 shows FTIR spectra for sintered B4C analogues (wavenumber: 500-3500 cm^-1);

Figure 10 shows Raman spectra for CaRBON samples, B4C analogues and indents (w=100-2000 cm-¹);

Figure 11 shows Raman spectra for sintered CaRBON samples and B4C analogues (w=100-2000 cm’¹);

Figure 12 shows the amorphisation map for B4C analogue (Left: plane-sided; Right: indented sample) - 532nm laser, 50x Objective, 2 pm Spot Size;

Figure 13 shows the amorphisation map for BeO sintered at 1570 °C for 30 mins (Left: plane-sided; Right: indented sample);

Figure 14 shows the amorphisation map for BeO sintered at 1850 °C for 30 mins (Left: plane-sided; Right: indented sample);

Figure 15 shows the amorphisation map of BeO + 1 wt.% GO sintered at 1570 °C for 15 mins (Left: plane-sided; Right: indented sample);

Figure 16 shows the amorphisation map of BeO + 1 wt.% GO sintered at 1570 °C for 30 mins (Left: plane-sided; Right: indented sample), and;

Figure 17 shows neutron fluence over time exposure for CaRBON samples and B4C analogues.

DETAILED DESCRIPTION OF THE INVENTION

One of the key areas of interest for CaRBON, as a material, is in the field of armour systems against ballistics threats. In particular, the threat posed by armour piercing, high velocity bullet cores made from materials such as cemented tungsten carbide (WC-10Co; 10wt.% cobalt is used as a sintering aid and additive to increase the toughness of the material and consolidate it at lower temperature). These projectiles, commonly used in firearms such as standard assault rifles and anything more powerful, can pose a distinct threat and challenge to most conventional armour materials and ceramics e.g., alumina (AI2O3), silicon carbide (SiC) and boron carbide.

A key trade-off to make for most armour systems and in most applications is that between protection and weight. When not in active use, most armour systems are effectively considered to be ‘parasitic’ - they add weight onto the system without necessarily providing benefit until the point of necessity. Furthermore, by making use of lighter materials that offer equal or greater protection than other alternatives, one can increase the manoeuvrability of the platform they’re placed on (e.g., foot soldiers, light ground, and rotary-wing vehicles, etc.).

Existing armour ceramics, such as AI2O3, SiC and B4C, have higher hardness and lower mass than metallic alternatives (~16 GPa, ~4.54 g/cm³; ~24 GPa, ~3.23 g/cm³; and ~30 GPa, ~2.52 g/cm³ respectively). The logic behind using such materials is that they are harder than the tip of the penetrator, thereby causing it to mechanically deform, dissipate its energy and make it less effective at defeating the material (WC- 10Co has a hardness of ~14 GPa and a density of ~14 g/cm³ - roughly twice that of steel). However, each material is not without its advantages and drawbacks. AI2O3, for one, is a very inexpensive, widely manufactured engineering ceramic. This makes it cheap, effective, and easy to manufacture at scale. However, compared to other potential materials, it is dense and must be used in more considerable quantities to render it effective against hard, high-energy threats. SiC is much more capable of dealing with such threats in lower quantities and is less dense. B4C is harder and lighter still but undergoes ‘shock-amorphisation’ when impacted by hard, high-energy threats, deteriorating its mechanical properties to a similar performance to materials like AI2O3.

Shock induced amorphisation in B4C occurs owing to the severance of the C-B-C atomic chain linking the B12 icosahedral structures in the material’s rhombohedral unit cell (which can also be viewed as hexagonal from alternative origin points). This causes a Lewis acid-base interaction to occur between the severed C-B radical from the chain and the parent icosahedron it is bonded to which, in effect, collapses the crystal to form a more dense, disordered structure.

BeO, as a material, is theorised to be of a similar density to B4C (2.55-2.62 g/cm³) with a hardness in the region of >35 GPa (some estimates suggest as high as 45-55 GPa, which would make the material the joint second hardest material known to mankind alongside cubic boron nitride, c-BN). However, preliminary evidence suggests that it doesn’t experience the same kind of amorphisation to B4C. This is thought to be because the severance of the 0-0 inter-icosahedral chain doesn’t cause a similar Lewis acid-base interaction with its parent B12 icosahedron, thereby maintaining the overall mechanical and structural integrity of the crystal lattice around the impact region.

Much of the recent literature over the past two decades has investigated the toughening of ceramics through multiple different strategies. The first involves reducing the grain size of such materials down to the nanoscale (<500 nm). This increases the overall number and density of grain boundaries that a crack must propagate through, and the mismatch of the interfacial energy between a grain boundary compared to its bulk thereby causes the crack to have to take a more torturous path as it travels through the material. This occurs up to grain sizes of approximately <10 nm for most materials, whereby the density of grain boundaries is thought to be equal to or exceed that of the material’s grains - which is known as the ‘Inverse Hall-Petch Effect’.

Another involves adding dopants to different material systems to substitute atoms in a material’s crystal structure, and thereby change its resultant mechanical properties (as well as potentially act as a sintering aid by increasing the interfacial energy of the system, and thereby increasing its Gibbs Free Energy, making it less stable under conditions of elevated pressure and temperature and more likely to transition in phase to a sintered form). This can also cause reactions to occur, forming secondary phases with different mechanical properties to that of either parent phase. The final way involves engineering the materials microstructure to form composite systems, that cause multiple, more convoluted failure mechanisms to take place for a crack to propagate through it. Such systems include the use of fibrous or particulate additives with similar or different mechanical properties to the matrix material. This causes any crack to have to firstly de-bond the matrix from the additive before enough energy can be transferred across it so it can continue to propagate, and then cause a failure of the additive particle, which then bridges the crack in the matrix. Such additives can be macro-particles (e.g., fibre-matrix systems) or nanoparticles.

In addition to this, certain additives can change the failure mechanics within a composite system. For example, porosity, as well as hard particles (which don’t have as much of a detrimental effect on the overall system’s strength and hardness properties) are capable of ‘blunting’ cracks. This is when the crack hits the interfacial region between the two zones (matrix and pore or second particle) and experiences a significant reduction in ‘acuity’ (sharpness) ahead of its tip. Thereby, more energy is needed to drive the propagation of the crack. If they experience different mechanical failure properties to the matrix phase, they can also cause crack deflection, which also dissipates its energy.

As a concept, CaRBON intends to make use of a number of these materials and composite formation principles and merge them together into a single, useable product. This is intended to be done by forming a nanocomposite of the novel, experimental ceramic BeO, with an additive of B4C that is formed in-situ during consolidation due to the reaction of BeO with elemental carbon in the form of rGO. BeO, as with most oxides, has a high reactive affinity for carbon-based materials. Under high temperature conditions, the carbon atom will typically reduce the oxide in a ‘carbothermal reduction’ reaction and substitute itself in the materials crystal lattice.

As a precursor, rGO was chosen for several reasons. Namely, it is the easiest form of nano-carbon to be homogeneously and readily dispersed in other suspended materials with water. This is because it originates from GO, which features multiple hydrogen-bonding functional groups on the basal hexagonal carbon layers of the material. This means it is possible to readily prepare specific masses of the substance and blend them intimately with other materials in aqueous suspension, which is straightforward to process with minimal chemical, supply, and processbased risks. It then reduces exothermally under temperatures exceeding ~200 °C (max. ~240 °C) to form rGO (effectively a defective, holey version of few-layer graphene), which can then react with the BeO under sufficient reaction and consolidation conditions (in particular, under both heat and pressure). This, in principle, should form a finely dispersed phase of B4C between the BeO grains which, owing to their propensity to undergo amorphisation under critical shock stresses, can potentially readily deflect any cracks that may propagate through them over the BeO matrix phase, as illustrated in Figure 1 . Oftentimes, ultrahard materials are also known to be extremely brittle, which is why it’s difficult to make particularly large parts from very hard minerals and ceramics like diamond or c-BN. However, by making use of such a composite, it may be possible to ‘cement’ the BeO together whilst still maintaining its ultrahard properties in a similar fashion to WC-10Co.

Although CaRBON could potentially be formed by BeO that is made in-situ within the apparatus that is used to sinter it from its known chemical precursors, a-B and B2O3, attempting such a reaction whilst ensuring the production of high quality, consistent products has proven to be prohibitively challenging in practice. Therefore, the ‘ex- situ’ production route of purchasing the BeO from a known chemical supplier, blending it together with the GO and forming the composite during sintering seems to be the favoured route to be proposed for the project, as shown in Figure 2.

And as well as being a sintering aid to the material, the multiple decomposition stages of the GO, and its subsequent reaction to the BeO, are very exothermic. This can therefore raise the likelihood of being able to form an ultralight, ultrahard material with multiple other bespoke properties at much lower temperatures and pressures than other engineering ceramics of its kind. This lessens both the resource use and intensity of the conditions associated with its manufacture.

Neutron Absorbent, Moderation and

Compounds of boron and carbon are currently widely used for many applications in nuclear physics, both for civil and defence purposes, to act as absorbers and shields for radiation (in the prevailing case of boron) or as moderators (in the case of carbon) or both (e.g., boron carbide, etc.). This is particularly so with neutrons, which are widely known - particularly after long term exposure - to cause the greatest level of structural and activity damage to other materials.

This behaviour is primarily for two reasons. The first is due to neutrons having a net lack of polar charge. This means that, except for materials that diffract them via inelastic scattering, nothing can deflect their motion. The second reason is due to their mass and kinetic energy in most nuclear applications. This causes them to impact the crystallographic structure of most materials and potentially displace their atoms from the lattice, giving rise to Frenkel and Schottky defects (where an atom in the lattice gets substituted, however asymmetrically, by another nearest neighbour, or moved into one of the interstitial voids in the lattice structure). Over time, this not only causes great damage to materials facing prolonged exposure to high neutron fluence but can cause an accumulation of Wigner Energy (stored up potential energy caused by the continuous displacement of atoms in a microstructure). If not controlled, this Wigner Energy can spontaneously release when energetically favourable to do so. This was the case in the Windscale nuclear disaster in Sellafield in 1957, where the graphite control rods to the reactor caught fire, causing mass radiation pollution on a wide scale. This can also cause most materials with higher atomic numbers to potentially become radioactive. This is of particular concern in all kinds of civilian nuclear reactors, where neutrons of a high enough temperature can potentially transmute otherwise benign surrounding materials into radioisotopes. Therefore, typically, both fission and experimental fusion reactors require a modest degree of shielding to all the various forms of radiation that can be produced from them, but most notably neutrons.

Compounds of boron and carbon are favoured for use in moderation, absorption and/or shielding applications due to their high respective neutron cross sections. This means that, owing to their low atomic mass and bonding, they approach the mass of neutrons themselves and can ‘capture’ them inelastically with both minimal damage to the material or risk of it becoming radioactive, even after prolonged exposure. This means that if nuclear control rods were to be made from such materials in fission reactors, for example, that the ability to dispose of such materials at their end of life becomes much shorter and easier than other potential candidates. It is also known that oxygen shares similar absorption and moderation properties to carbon. Such compounds include graphite or B4C, the latter is particularly favoured in pressurised water reactors (PWRs). One potential advantage that BeO could have over both materials is that, as well as exhibiting many of the positive properties of B4C (high hardness, mechanical robustness, lack of flammability after prolonged use), it has 1.5 times more B atoms per mole of material than the former. This means that it could, in principle, exhibit a greater commensurate level of neutron absorption behaviour for the same overall mass of material used, but still exhibit some moderation behaviour owing to the presence of oxygen in the material’s inter- icosahedral chains (and due to the potential presence of intergranular B4C that’s still found in the composite). There is also reason to believe that these chains could also be self-healing, by allowing further atomic substitution to take place if other atoms from the surrounding lattice get bombarded out of their respective positions.

This means that it can be used to make more efficient nuclear control rods for fission reactors. It may also serve as a useful radiation shield cladding to protect the surrounding parts of both fission and experimental fusion reactors. Nuclear fusion reactors typically require a toroidal plasma of deuterium/tritium to be kept in electromagnetic or inertial confinement under very substantial pressures and temperatures to fuse, whilst also facing high prolonged neutron fluence. Most materials cannot survive such conditions for very long without decay or failure. This has meant that the standard materials under investigation have, to date, been very expensive and difficult to process alloys of tungsten, and occasionally boron carbide. CaRBON could provide utility in these applications by already existing as a refractory ceramic with a high neutron cross section. It may also be useable in smaller quantities to existing alternatives to protect the surrounding parts of such a reactor and to allow a sustainable, prolonged nuclear fusion process to take place and produce the energy it needs to both sustain itself, as well as any surplus power that would be required for civilian electrical power production.

Abrasives and Hard-’

BeO is thought to be the joint second hardest material known to mankind, alongside c-BN (~55 GPa) and diamond (>100 GPa). So, if it’s able to be produced in large scale quantities, and potentially be used to form large parts, it may stand as a promising candidate for use in industrial abrasives and other wear resistant applications. The entire premise of such applications rests on the use of a harder material (less susceptible to localised plastic deformation under quasistatic loading) to either wear, cut through or withstand wear from a softer material. Although many materials exist that can either be used on their own or in conjunction with another material (e.g., diamond plated cutting disks, c-BN cutting tools), they are typically limited in use both by their costs, as well as the inability to manufacture large parts from them (e.g., pipes, industrial components, wear resistant turbines, etc.). Where CaRBON could be useful, in this regard, is the ability to have large scale parts made from it both under a potentially limited production cost compared to other super-hard ceramics and minerals, whilst also exhibiting equal or superior properties. The presence of rGO to act as both a reactant and sintering aid with the BeO means that the temperatures required to process the material can potentially be lessened by several hundreds of degrees Celsius. This may allow certain parts to be formed from it by either using less pressure, or entirely without it. In so doing, the variety of part geometries that can be made using it greatly increases.

In addition to these applications, recent research has also been conducted in the use of boron oxycarbide nanostructures for use in supercapacitors. Supercapacitors have a wide range of growing current, as well as potential uses in both energy efficiency, as well as sustainable energy applications (e.g., regenerative breaking in hybrid/electric vehicles, large-scale, high-power energy storage, etc.). Published literature suggests that such materials, made with similar starting materials (albeit not at all the exact same materials and process) to CaRBON were able to operate over a voltage window of 2.0 V in tetraethylammonium tetrafluoroborate (TEABF4) electrolyte. The example in that study displayed a capacitance of 7.26 mF/cm², an energy density of 14.53 mJ/cm², as well as a lower equivalent series resistance of 1 .44 Q, a higher power density of 373 mW/cm² and a longer charge/discharge cycle life than other existing supercapacitor variants using silicon or carbon.

Solid State

The oxidation of hydrogen to produce water and energy, in the form of both heat and electrical current using a fuel cell, is being seen as one of the perennial challenges in developing a sustainable energy economy for the future. Current solutions of hydrogen gas pressurisation and containment exhibit difficulty in allowing the sufficient levels of hydrogen to be stored to render such vehicles competitive with existing fossil fuel powered, or even electrical variants. The liquification of the gas would pose additional risks in both safety (against potential leakage or fires), as well as material and economic resources to keep the hydrogen as a liquid. The favoured solution to this issue in the research community, therefore, is to store hydrogen in solid state form within the chemical composition of materials containing the element (e.g., ammonia diborane, NH(BHs)2) or in the interstitial crystal structure vacancies of other materials (e.g., magnesium alloys or lithium borohydride, LiBH4). To date, there haven’t been a wide variety of studies looking into the use of refractory ceramics like BeO as hydrogen storage materials. However, due to their large crystal structures and complex bonding, there may be reason to believe that BeO could at least be potentially tested as candidate hydrogen storage materials using methods such as Intelligent Gravimetric Analysis (IGA) and TGA/DSC.

Powder Blending and Precursor Synthesis

Commercial grade boron suboxide powder (dso <1 pm; Fraunhofer IKTS, Hermsdorf, Germany) and aqueous graphene oxide suspension (1 g/100ml, GoGraphene, William-Blythe Ltd., Accrington, UK) was acquired to form the colloidal aqueous CaRBON suspension. To make these up, Type I deionised water, available from a laboratory supply (Millipore Milli-Q RIOS, Advantage and Q-Pod, MilliporeSigma Ltd., MA, USA; ELGA Purelab DV35, Elga LabWater Ltd., High Wycombe, UK) was used. Suspensions of the BeO and GO were diluted and made according to these concentrations:

• BeO - 10 mg/ml.

• GO - 0.4 mg/ml.

Upon formation, these would then require stirring for 1 h to homogenise both blends. Table 1 shows the necessary amounts of each material required to make up the suspensions. Table 1 : Preparation of GO suspension

Table 2 shows the masses of BeO added to the deionised water to make up the required levels of suspension for each 4g composite sample, along with their respective volumes of additional GO suspension. Three samples at each composition were made, except for the 0 and 1 wt.% GO containing sample, which had an extra 4g of material made up to conduct iterative tests on the sintering conditions and parameters required for SPS.

Table 2: Compositions, Masses and Volumes of BeO and GO

The mixed suspensions then require an additional hour of stirring before being passed through a series of Nalgene PES Buchner funnels (0.2 pm, 90 mm diameter, 1 L capacity, Merck Life Sciences Ltd., Gillingham, UK) linked to a vacuum pump (KNF Lab Laboport, KNF Neuberger Ltd., Witney, UK) to help draw through the suspension.

A different filter funnel was used, according to the concentration of the suspension that was passed through it. A wetted nylon filter membrane (0.45 pm, Merck Life Sciences Ltd.) was added for each filtration to both minimise the level of cross contamination from one stage of filtering to the next, as well as to provide an extra level of mitigation against any of the GO in suspension from potentially being pulled through the filter. A magnetic stirring plate (Coming Ltd, US) with corresponding stirring beads were used for all stages of the colloidal processing. Once filtered, the powdered materials were then dried in a drying oven (AX120, Carbolite-Gero Ltd., Hope Valley, UK) or box furnace (CWF1300, Carbolite-Gero Ltd.) at temperatures of ~125 °C for ~2 h using a Temperature Control Unit (3216, Eurotherm Ltd., Worthing, UK) until they were dry. The materials were then lightly de-agglomerated in a pestle and mortar until they were rendered flowable and bottled in sample vials ready for delivery to Nanoforce Ltd., an assigned subcontractor operating out of Queen Mary University of London (QMUL). To obtain characterisation data of the powder blend before and after GO reduction, samples of the BeO containing 1 wt.% GO and 4 wt.% GO were used, both as produced and after heat treated in a tube furnace (STF 15/100, Carbolite-Gero Ltd.) using a temperature controller (2132 & 3216, Eurotherm Ltd.) at 250 °C at 5 °C/min from room temperature under 30 cm³/min of flowing Argon (BOC Ltd, Reading, UK). The tube was purged of air over a 15 min cycle to stop sample oxidation.

Spark Plasma Sintering (SPS)

Sintering is a predominately solid-state mass diffusion process that is used to combine a body consisting of many discrete, individual particles into a larger, single, chemically bonded polycrystalline mass. This is typically achieved using heat (sintering temperatures normally exceed 60% of the materials melting temperature, Ts > 0.6Tm) and pressure (the application of pressure increases the Gibbs free energy of the system, thereby thermodynamically driving it towards the formation of reaction products). Sintering is the standard method through which various ceramic materials are produced. There are a variety of methods that can be used to do this, such as pressure-less sintering (placing a ‘green body’ of the formed and/or compacted ceramic, typically mixed with a ‘binder’, in a kiln at high temperature for extended time periods until densified) or hot isostatic pressing (placing the preformed material within an envelope under elevated temperature and pressure, applied on all sides with even force through the use of pneumatics or hydraulics). CaRBON samples were prepared using SPS. SPS is a method that uses double ended, uniaxial pressing in conjunction with an electrically conductive die (graphite parts for lower pressure, high temperature experiments or WC-10Co for low temperature, high pressure use) and a high-power electrical current that’s passed through or around the sample (depending on whether the sample is an electrical conductor or not). This gives rise to sparking across the two ends of the press, which effectively act as electrodes, and joule (electrically resistant) heating within the sample die and, when conductive, the samples themselves. This allows the formation of highly dense samples with very controlled grain sizes (due to the limited ability for asymmetric grain growth and coarsening mechanisms to occur) to be made in a matter of minutes, as opposed to more conventional mechanisms which can typically take hours or days. This method was seen as the ideal one to use for early-stage studies due to the use of heat, pressure, and electrical current to assist with the consolidation of the sample. Owing to the in-situ formation of rGO within the sample, CaRBON effectively becomes electrically conductive during SPS. This, combined with the additives effect as a sintering aid and the use of pressure, helps to minimise the theoretical sintering temperature and time needed to consolidate the samples.

To consolidate the samples, an SPS Press (HP D 25, FCT Systeme GmbH, Rauenstein, Germany) was used. The powders were placed inside a 20mm diameter graphite die (Goodfellow Ltd., Huntingdon, UK) made of a cylindrical mould with 2 punch pieces, which the sample sat between. Both ends of the press face and inner circumference were lined with graphite foil. This was to aid the electrical current conduction through and/or around the sample. All samples were sintered under a <10^-2 bar vacuum to aid with pore closure and the evacuation of gaseous byproducts from the sample. For all samples, a minimum contact force of 5 kN (16 MPa) was applied whilst they were heated from room temperature to 1300 °C and held for 5 mins. This step was to help sublimate any residual B2O3 off the BeO grains and aid the materials sinterability. After this stage, the sample temperatures were increased to the peak levels and hold times seen in Table 3. These parameters were chosen from the literature for BeO, as well as by trial and error from previous experiments.

Table 3: Compositions, Masses and Volumes of BeO and GO After reaching peak temperature, the pressure on all the samples was increased to 16 kN (50 MPa) over 5 mins before commencing the hold times listed in Table 3, to aid with sample densification. In practice, the graphite dies can survive pressures of up to 25 kN (80 MPa). However, for most applications in engineering, the applied pressure rarely ever exceeds 50 MPa so the dies can be reused without the risk of damage. A heating and cooling rate of 50 °C/min was used for all samples. Live information on changes in system control temperature, applied force, chamber pressure, and die displacement were also measured during sintering to gain information on sample densification.

Post Sintering Sample Preparation

To prepare each of the sintered samples for further characterisation, they needed to be ground to optical flatness and polished. To do this, they first had to be encased in cold mount resin (Epofix, Struers Ltd., Denmark) before being loaded on an automatic polishing machine (Tegrapol-35 and Tegradoser-5, Struers Ltd., Rotherham, UK) and being ground and polished to a 1 pm finish using a series of successively fine diamond plated polishing discs. For microstructural characterisation, the samples were etched for 1 min in hot 6M hydrochloric acid (Honeywell Fluka, USA) kept to a temperature of ~75 °C using a hotplate, before being quenched in deionised water.

Microstructural Characterisation

To characterise the microstructures of both the powdered and sintered samples, a JSF 7000F SEM was used. For analysis of any radiation induced damage after fast neutron bombardment, a Neoscope 7000 was used (both SEMs were manufactured by Jeol Ltd., Welwyn Garden City, UK). All samples were analysed with an acceleration voltage of 15 kV, a beam current of ~30 pA and a working distance of ~10-12mm for imaging and EDX. Images were taken of all analysed samples at 30x, 200x, 1 ,000x, 2,000x, 5,000x and 10,000x magnification using secondary electron (SE) and backscattered electron (BSE) imaging. Many of these samples were also analysed using EDX (X-Max 80, Oxford Instruments Ltd., Oxford, UK - the EDX unit on the Neoscope was inbuilt) to check for both point and areal composition. The point measurements were taken over an average of 5 readings (each done for 60s) with compositional error margins calculated via standard deviation. All samples were adhered onto a stub with a carbon sticker prior to analysis.

To gain information on both the crystallography of the sample, as well as serve purposes of phase identification, all samples were subjected to XRD analysis. A D2 Diffractometer (Broker GmbH, Karlsruhe, Germany) was used with a Cuka source between a Bragg angle range of 10-90° gathered over 43 mins using a step rate of 0.16167min (Dw slits = 1 mm; 3mm air screen; 1.5Ni KB filter, 2.5mm Soller slits, 5.6mm PsD opening). The patterns obtained from the software were analysed both as. xy files and as .raw files loaded up into TOPAS software (Broker GmbH).

Laser Scattering Particle Size Analysis (LSPSA)

LSPSA was carried out with the as-received BeO powder using a Mastersizer 3000 (Malvern Panalytical, Malvern, UK) under both wet and dry analysis. The BeO was diluted in de-ionised water and dispersed with a surfactant to aid the analysis. It was decided not to conduct the same analysis to the composite powder, either before or after reduction, due to the combination of the probable equiaxed structure of the BeO particles with the platelike particles of the (r)GO. Owing to programmable algorithms in the machine’s analysis software, this would introduce substantial errors in the measurements to discern either material, and pose significant difficulties in accurately determining the plate morphology of the (r)GO samples (much greater in diameter than thickness).

Thermoqravimetric Analysis/Differential Scanning Calorimetry (TGA/DSC)

TGA/DSC was used to gain inferences on both the heat flow, as well as the mass changes occurring to the powdered samples as a function of temperature, using an STA449 (Netzsch Thermal Instruments Ltd., Wolverhampton, UK) making use of a Platinum (Pt) furnace. These tests were conducted under vacuum between 10-1500 °C at a ramp rate of 10 °C/min.

Vickers Hardness Testing

A Vickers hardness indenter (CV450AAT, CV Instruments Ltd., Hoddesdon, UK) was used to assess the mean average hardness of the sintered samples across 10 indentation points. The errors associated with each samples recorded hardness values were gathered via standard deviation. A load force of 10N was applied to each indent over a dwell time of ~10s.

To gain insight into the surface level chemistry of both the powdered and sintered samples, optical spectroscopy was performed on them using both an FTIR (Vertex 70V, Broker GmbH) and Raman spectrometer (Invia, Renishaw Ltd., Wootton- Under-Edge, UK). Due to exclusion rules, the former method was used to gain information on vibrational modes corresponding to symmetrical bonds, whereas the latter was used to gain information on those exhibiting asymmetric bonds. The FTIR was used with an Attenuated Total Reflection (ATR) stage with a 4 cm^-1 resolution.

A series of single point Raman spectral scans were performed using a 532 nm laser through an 1800 lines/mm diffraction grating. The microscope was mounted with a 5x objective (and a 10x camera/eyepiece magnification) for large areas of the powdered and solid samples (~20 pm laser spot size). Single point scans were done between 0-2000 cm’¹, each lasting 60s and done over a total of 10 accumulations. A range of laser powers between 0.05-10% were chosen, depending on the signal: noise response from the test sample undergoing analysis.

Mapping analyses were also subsequently carried out on plane-sided and indented regions of the sintered composite samples using the 50x objective (~2 pm spot size) between 1 -10% power. This was carried out to investigate the presence and extent of amorphisation in the CaRBON samples relative to a series of sintered B4C analogues made during the authors doctoral thesis. These analogues were fabricated via similar processing methods to the ones employed in this study (e.g., via SPS). To distinguish the matrix phase from the amorphised regions, and the intensity to which this process took place within the samples, a series of colourgradient maps were superimposed on images of the plane-sided and indented areas taken from the machine’s microscope. The modes corresponding to B4C or BeO were designated to be <1200 cm’¹ and were coloured blue. Those corresponding to amorphisation were designated to be >1200 cm’¹ and coloured with a green-to- yellow-to-red ‘heat map’ according to the level and intensity of amorphisation in those regions.

Accelerator Source Pulsed

One of the key properties of interest to be determined during the CaRBON study is the materials interaction behaviour with neutrons - in particular, fast neutrons with energies exceeding 14 MeV. All the elements making up the ceramic composite, such as B, C and 0 are all light elements, and act as ‘neutron poisons’ (they inelastically scatter neutrons owing to their high capture cross-sections. This either causes them to be slowed down, diffracted, or stopped altogether). What is more, every mole of BeO contains 1 .5 times more B atoms - which is the lightest of the 3 constituent elements, and therefore possesses the highest capture cross section - than B4C. The two materials also share similar density values and crystal structures to each other. This therefore raises the prospect of BeO-based materials acting as superior neutron poisons to monolithic B4C. To test this, the two best quality samples were subjected to a continuous two week-long bombardment of fast neutrons produced by deuterium/tritium (D/T) fusions. In essence, this experiment entailed affixing the two samples alongside two comparable B4C analogues around the water coolant line, facing opposite the detectors. Thereby, the tested samples could be continuously subjected to the radiation and assessed for damage, if any was accumulated, after the two-week period. This could then also be compared to the B4C analogues. Additionally, any changes in neutron fluence (flow, intensity, and energy) with each of the tested samples could be measured across the rig and either compared with each other, or to the levels obtained under a control experiment (run with no loaded samples).

Results And Discussion

SEM/EDX Characterisation

SEM images, gathered at 1 ,000x and 5,000x magnification, show that the powder exhibits a lot of agglomeration. The diameter of these agglomerates can be >30 pm, but those of the equiaxed primary particles are <1 pm. Judging from first principles, this agglomeration is likely to take place due to the physisorption of atmospheric moisture on the powder’s surface.

SEM micrographs for the reduced powder containing 4 wt.% GO, gathered under identical conditions are shown in Figure 6. By comparison, these micrographs show a distinct lack of primary particle agglomeration. Once again, using the same data corroboration methods, this is thought to be due to the reduction of the GO to rGO, which causes the volatile production of H2O, CO and CO2, which blows the primary particles apart from each other.

Sintered Sample Characterisation

After sintering, several samples remained physically and geometrically intact, whereas others were either too brittle to be retrieved, or simply reacted with the graphite tooling in the die, thereby misshaping them and rendering the sample too unreliable to process in practice. These are illustrated in Table 4, where all the pure BeO samples failed upon retrieval. By contrast, many more of the composite samples were able to be retrieved intact and survived. This may give reason to suggest that the presence of rGO, whether it reacts with the BeO or not, may act as a ‘cementing agent’ in a similar way that cobalt does to tungsten carbide in WC-1 OCo.

Table 4: Conditions and Prognosis for the Densification of Samples in Phase I Study

Failed = sample broke upon retrieval and/or caused reaction with the die. Succeeded = sample able to be retrieved from die whole and intact with minimal wider damage. Densification Analysis of CaRBON Samples

Theoretical densities were calculated in according with the standard equations outlined in the Annexes appended to this description. The densification results for the samples in this study can be seen in Table 5.

Table 5: Densification of Samples in Phase I Study

Of all these samples, it was the pure samples that achieved the highest level of densification. However, as mentioned in Table 5, all these samples failed upon retrieval by both reacting with the die and/or subsequently breaking. Therefore, the optimal practical results that have been attained from this study have been gathered from the 1 wt.% GO containing samples.

The case for suggesting that rGO could act as both a sintering aid and cementing agent is further strengthened by the fact that the 1 wt.% GO containing samples, sintered at 15 and 30 mins, both exhibited greater degrees of densification (90.82% and 92% respectively) than the pure BeO sample sintered at the same temperature for 30 mins (89%), and both remained intact, unlike the latter. In practice, there is a trade-off between adding enough GO to ensure the adequate consolidation of the BeO as a sintering aid, but not adding enough to cause a failure by way of sample or die breakage.

Vickers Hardness Testing

A Vickers hardness indenter (CV450AAT, CV Instruments Ltd., Hoddesdon, UK) was used to assess the mean average hardness of the sintered samples across 10 indentation points. The errors associated with each samples recorded hardness values were gathered via standard deviation. A load force of 10N was applied to each indent over a dwell time of ~10s. Table 6 shows the mean average Vickers hardness readings for samples described above, with standard deviation error margins.

Table 6: Vickers Hardness of CaRBON Samples

These results can also be seen in plot format on Figure 3. What they show is that, under similar processing conditions, as well as being able to achieve higher density levels and retain better geometry, the composite samples, most notably those with 1 wt.% GO, achieve statistically more significant hardness values under similar processing conditions to the pure BeO, indicating optimised mechanical properties from said samples. The average hardness of the BeO, sintered at 1850 °C, has been added to Figure 4.

As seen in Table 5, the level of densification from this sample is considerably greater than most of the others (~95% of TD). As a result, its average hardness increases accordingly (193.6 Hv ± 40.63 Hv across 5 indents). Estimates suggest these values to be slightly greater than B4C (~182.7 Hv ± ~27.4 Hv, >98.5% of TD), made under similar conditions in the authors doctoral thesis from commercial grade material (HS Grade, dso ~1 pm, H.C. Starck GmbH, Germany - measured to be 32.5 GPa ± 1 GPa via micro-Vickers indentation. The corresponding hardness for diamond is ~100 GPa+). Considering that two of the non-fully dense 1 wt.% GO containing materials achieve hardness values that are almost half this, their approximate corresponding micro-Vickers hardness values would likely be ~16 GPa (greater than fully dense AI2O3 or WC-Co). It can therefore be inferred that, if brought close to theoretical density (TD, ideally >98.5%), CaRBON could be capable of displaying ultrahigh hardness, whilst also being able to be consolidated to large geometric forms, potentially with and without the use of pressure in sintering. SEM/EDX Analysis

Figure 5 shows the SEM micrographs for the pure BeO samples sintered at 1570 °C and 1850 °C respectively. The former exhibits a far greater level of porosity, both in terms of size and population, than the latter material. This is to be expected, owing to the lower temperature used to sinter the sample and the lack of use of sintering aids in either material. The 200x micrographs were also taken with Vickers indents, present from the hardness tests that were carried out before analysis. At the edges of an indent in the material consolidated at 1570 °C, there are no cracks. But very long radial cracks can be seen in the comers of the indent in the material sintered at higher temperature conditions. This is also to be expected, as porosity acts as a ‘crack blunter’, meaning that it dissipates energy ahead of its tip by making it less sharp. This can increase the overall ‘toughness’ of the material (the amount of energy, under dynamic loading, that a material can absorb before plastic failure), but typically does so at the expense of most of the materials other mechanical properties (especially ‘hardness’, the materials resistance to localised plastic deformation under quasi-static force loading conditions, and ‘strength’, the amount of quasi-static force that can be loaded over a cross sectional area of the material before deformation). Further evidence of this crack blunting effect can also be seen in the micrographs for the composite samples featuring 1 wt.% GO, seen in Figure 6.

Table 7 shows the EDX results for the sintered CaRBON samples imaged under the SEM. For the BeO sintered at 1570 °C, the results describe an outcome that would be expected for the material. With XRD results stating that the sintered material can be positively identified as BeO, the division of the molar atomic ratios clearly shows that the sample, with high accuracy, conforms to this formula (<6 parts boron, ~1 part oxygen).

Table 7: EDX Results for the Sintered Samples with Standard Deviation Error Values These results begin to change for the BeO sample sintered at 1850 °C, which exhibits a near 100 at. % B reading with <1 at. % for 0 with high accuracy (<0.5 at. % error). The addition of rGO to the material, however, begins to change the chemical characteristics of the BeO. After 15 mins of sintering, the 1 wt.% GO containing sample features an atomic percentage of boron that is ~76%, ~21 % of carbon and a small remainder of oxygen. As the material’s sintering time is doubled, so is its respective carbon content, with a commensurate decrease in the materials boron content (the error margins associated with the two readings also nearly double). These findings could potentially be explained if the interfacial regions between the BeO and the carbon in the 1 wt.% GO containing sample react to form B4C under prolonged periods of sintering. By contrast, the samples held for a shorter time at peak temperature would contain a mixture of BeO and B4C phases. If this reflected in the EDX data, this would be seen in a carbon reading of ~20 at. % and ~80 at. % for boron. Whilst these results potentially give an indication of gradual carbon enrichment of the surface layer (perhaps akin to a ‘bark’ layer), it is worth noting that the signal peaks seen on the energy spectrum produced by EDX become more characteristic as the difference between the atomic numbers of two elements increase. Because carbon only has one more proton than boron, there may be some ambiguity in being able to ascertain these results with accuracy. However, owing to the fidelity of the 80mm EDX detector used in this study, these results can be believed to be of at least a moderate degree of precision.

XRD Results

Results obtained via XRD, as shown in Figure 7, give conclusive evidence that by sintering BeO with carbon, the material reacts in situ during sintering to form B4C. This can be seen in the patterns for the 1 wt.% GO containing sample after 15 and 30 minutes of sintering (both of a similar level of densification and hardness), whereby the sample sintered after 15 minutes consists of a mixture of the peaks associated with B4C and BeO, whereas the latter transforms almost entirely into B4C after 30 minutes.

In addition, when pure BeO was sintered at 1850 °C for 30 minutes, it did not exhibit the characteristic pattern for either BeO (or any variant of the material found in current literature) or B4C. The sample’s pattern most closely correlated to that of a B4C-like configuration, with a particularly accentuated peak intensity (often corresponding with strong dimensional anisotropy of the materials grains, indicating preferential orientation and alignment) in either the (102) or (002) crystallographic plane (depending on whether it is modelled under an R3m or a P63/mmc space group. This could potentially indicate the synthesis of a presently unknown and novel material phase that may only either form during the CaRBON synthesis process, or once the existing material composites achieve higher levels of densification.

FTIR Spectroscopy Results

Table 8 and Figure 8 show the FTIR spectra for all the BeO based samples analysed in this study.

Table 8: FTIR Vibrational Modes of Sintered BeO-Containing Samples

With the exceptions of the pure BeO sample sintered at 1570 °C and the 1 wt.% GO containing sample sintered for 30 mins, most of the spectra appear to be nearly identical to one another. Due to the novelty of BeO, as a material, it is difficult to fully characterise all the peaks arising from its vibrational modes. A few of the peaks between -1200-1450 cm’¹ could be characterised with relative certainty, however these modes get dampened in most of the other samples.

Table 9 and Figure 9 show the FTIR spectra for a series of B4C sample analogues. Owing to the similar rhombohedral/hexagonal crystal structure of B4C and BeO (the main difference between the two materials being the presence of a bonding oxygen atom between the B12 icosahedra in the BeO unit cell, as opposed to an intericosahedral C-B-C chain linking the B12 icosahedra in B4C), the vibrational spectra between the two materials do appear to be rather similar to one another (with perhaps some slight shifts in Full-Width-Half-Maxima wavenumber).

Table 9: FTIR Vibrational Modes of Sintered B4C-Containing Samples

One of the key distinguishing features between the B4C and BeO samples is the presence of the peak at ~1600 cm’¹, most likely corresponding to the presence of the G Peak (graphitic peak - caused by transverse optical phonons transmitting in the vertical c-axis of the material’s Ce hexagonal rings). In conjunction with all of the characterisation data gathered to date, the lack of this peak, or an associated D peak (disorder peak, ~1330 cm’¹, caused by longitudinal phonons scattering in the lateral axes of the Ce rings) indicate a lack of presence of C in any of the BeO containing samples, suggesting that it probably does get reacted and consumed during extended sintering.

Raman Spectroscopy Results

Due to mutual exclusion phonon scattering rules seen between compounds with symmetric bonds, relative to those with asymmetric bonds, FTIR (mostly used to detect the former kind), as a technique, is best used in conjunction with Raman spectroscopy (able to detect the latter).

Table 10: Raman Vibrational Modes of Sintered B4C-Containing Samples

The modal peaks and the spectra for the sintered materials, analysed in this study, can be seen in Table 10 (for B4C), Table 11 (for the BeO-containing samples) and Figure 10 respectively.

Table 11 : Raman Vibrational Modes of Sintered BeO-Containing Samples

Due to the similar crystallographic structures of the B4C and the BeO, many of the spectral features between the two materials appear to be effectively similar, with some slight shifts in wavenumber values owing to the atomic structural differences between the samples. Two features that can be seen to be different between the two materials, except for the 1 wt.% GO containing sample after 30 mins of sintering, is the presence of the D and G peaks. These indicate the presence of graphitic carbon. Data from XRD indicates that a substantial amount of the BeO composite’s surface layer transforms into B4C, hence justifying the presence of the peaks.

Amorphisation Mapping Analysis

Figures 12-16 show the amorphisation maps gathered over the Raman spectrometer for the B4C and CaRBON samples for both plane sided samples, as well as those which had Vickers microindents placed in them for hardness testing. A multivariate analysis was performed, whereby all the peaks below 1200 cm’¹, assigned in one colour, were assigned to either B4C or BeO, and any peaks above 1200 cm’¹ were associated to sample amorphisation (the D and G peaks of carbon form at 1330 cm’¹ and 1605 cm^-1 respectively). A variable gradient shaded ‘heat map’ was used to highlight, on an increasing level, the corresponding intensities of the amorphisation in each of the samples.

Figure 12 shows the amorphisation maps for the B4C analogue. It can be seen that a very strong and evident degree of amorphisation takes place within the indent. This is to be expected, as the high local pressure field in this region is thought to be the key driving mechanism for amorphisation in B4C and has been observed in the literature. The degree of circumferential ‘penny-cracking’ around the indent, as well as the linear cracks emanating from its corners, is also a notable indicator of the material’s brittle behaviour and low toughness.

Figure 13 is the amorphisation map for the pure BeO sintered at 1570 °C for 30 mins. It shows a slight degree of sample amorphisation, or the various other mechanisms that can be associated with possible crystallographic changes (e.g., residual stress accumulation) within the indented material. However, the point of note is that this occurrence mostly takes place in the regions surrounding the indent, particularly around it’s perimeter and corner regions, as opposed to the indent cavity, or its apex - the regions associated with the application of the highest levels of local stress. As well as indicating that BeO is more impervious to amorphisation than B4C, it also suggests that these structural changes may not be mainly driven by applying greater levels of local pressure.

These phenomena are made all the clearer in Figure 14, which shows the amorphisation map for the pure BeO, sintered at 1850 °C. This sample also shows a significant lack of amorphisation in the indent region, with the comparatively greater levels of structural change occurring in the surrounding ‘half-moon’ crack regions.

Figure 15 shows the amorphisation map for the BeO + 1 wt.% GO sample sintered for 15 mins. This image shows an almost total absence of any amorphisation within the indent region, and a limited extent of it occurring in the regions surrounding. As well as due to the matrix phase predominately consisting of BeO, this could also be in part due to the still considerable levels of porosity in the sample (only ~90%+ of TD). Pores, although making the material softer and tougher, also act as crack blunters, and would therefore serve to confine the local stress field around the region of the indent. This level of amorphisation begins to increase steadily in the sample of the same composition after it has been sintered for 30 mins, as shown in Figure 16, but remains confined to the area around the indent as opposed to the region within it. XRD data shows that a full conversion to B4C occurs in the interface regions between the samples and the graphite SPS tooling, although the intensity of these readings is still rather low compared to the pure B4C analogue. This could suggest that the extent of this B4C formation is largely confined to the sample’s surface regions and doesn’t permeate the bulk.

The key finding from all these readings are the confirmation of the hypothesis that BeO, both in pure and composite form, experiences a limited or potentially no significant degree of amorphisation compared to B4C. As the degree to which it interacts with carbon increases with respect to content, temperature, and time - either in the interfacial regions between the sample and the graphite SPS tooling, or with rGO additive - the extent of this amorphisation also increases, which could also allude to the greater level of B4C formation taking place in these sample areas.

ASP Fast Neutron Bombardment Experiments

The experimental parameters used in the ASP/DD test can be seen in Table 12. The key parameters to note are the values of the neutron fluence (flow per unit area) and their respective energy levels. Over the test period (2x4 consecutive working days over the course of 2 weeks), each of the four tested samples were exposed to fast neutrons under the following conditions.

Table 12: ASP/DD Experimental Parameters Figure 17 shows the variation in neutron fluence over the course of each day. The plot, as well as the error value for the measurement shown in Table 12, shows that there is quite a wide error margin in neutron fluence over the course of the test period. For the first two days of the test period, the fluence levels were <1 O¹⁰ n/cm² With a slight deviation to a level of <10^{7 5} n/cm² on day 3 of testing, the fluence levels eventually converged to a level of <10⁹ n/cm² between days 4-8 of the test. This variation of fluence is to be expected owing to the stochastic and random nature of both the neutron production process, as well as their general direction of travel after formation. Neutrons are uncharged atomic particles. Therefore, at any moment in time, they will not only be produced because of a variable probability of successful D/T fusion events, but their likelihood of travel towards the DD experimental setup will also vary. But, due to the limited number of variables in the physical situation and the scale of the experiment, these fluence and energy levels will adopt a thin tailed (typically Gaussian) distribution. This allows the general variation of the neutron fluence and energy to generally become rather predictable over the whole history of the experiment.

After the two-week neutron bombardment period, the samples needed to have their radioactivity levels assessed to determine whether they were physically safe to release. High energy neutrons can increase the atomic mass of different materials by causing them to form different isotopes. Depending on the neutron scattering properties of the material, this can in turn cause them to develop unstable nuclei and subsequently become radioactive as the isotope decays to a more stable form. Materials with more inelastic neutron scattering properties, and high neutron capture cross sections, tend to be less susceptible to this effect than those with heavier elements.

Table 12 shows that, by the end of the 2-week test period, all four of the samples had a radioactivity of 0.2 mSv. This is the equivalent of two average bananas (a Banana Equivalent Dose of 2, or 2 BED), and were therefore deemed to be safe for authorised release. Upon release, the samples were then inspected for any signs of irradiation damage by eye and via SEM. Neutron irradiation, from a structural perspective, is known to be the most damaging form of radiation for most materials. The reason for this is because neutrons, as a basic atomic particle, are just the right size to impact with the nuclei of various atoms within a crystal lattice and impart their kinetic energy to them under various levels of energy, depending on how they are produced. The higher the energy of the neutron, the deeper it is capable of penetrating into the material’s lattice, thereby causing these damage effects to be observable deeper within the materials bulk thickness within a set frame of time. However, neutrons that travel with lower energies experience a lower penetration depth and are therefore much more likely to be successful in imparting their energy to the nuclei they do collide with. The damage that they cause is therefore more widespread and intense at the exposed materials surfaces but doesn’t permeate as deeply within it. After the test period, none of the samples experienced any greater level of damage, as analysed by eye or under SEM. As a material, B4C is widely used as a neutron poison due to the high capture cross section of the atoms within the compound, as well as its other generally appropriate mechanical properties (e.g., hardness and strength). It was therefore chosen as a good analogue to benchmark the performance of the CaRBON samples. As can be seen, over the 8-day test period, the sample accumulated no detectable levels of neutron irradiation damage. Furthermore, because the sample was near-fully dense and polished to a 1 pm finish before the test, the profilometry displays no significant change in the materials surface roughness. Any variation in height across the sample can only be attributed to the sample being slightly non-level when loaded onto the SEM test stub.

In summary, the study was able to demonstrate that under two consecutive 4-day periods of exposure to high energy fast neutron irradiation over the timespan of 2 weeks, neither the CaRBON samples under analysis, nor the B4C analogues, exhibited any signs of damage or practically significant increased levels of radioactivity after testing. This indicates that CaRBON may very well be a viable material candidate for applications involving any exposure to high energy neutron radiation. This prospectively opens the materials applications up for use in a variety of civilian and defence nuclear applications (e.g., deflection and shielding purposes), as well as a material that - upon further optimisation - will most likely also display ultralight and ultrahard mechanical properties as well. Annexes

A1 : Archimedes Method and Theoretical Density Determination

PTD = Theoretical density of composite (g/cm³)

A = Mass of the sample in air (g)

B = Mass of the sample in reagent grade (>99% pure) ethanol (g) po = Theoretical density of the ethanol at the time of the test (g/cm³)

%TD = Percentage densification of sample

A2: Rule of Mixtures for Composite Theoretical Density Calculation

PB6O = Theoretical density of boron suboxide (2.62 g/cm³)

VfB6o = Volume fraction of boron suboxide

PGO = Theoretical density of graphene oxide (2.2 g/cm³)

VfGo = Volume fraction of graphene oxide = /V

M = Mass of sample

V = Volume of sample

Vf of components was calculated by: • Dividing the known mass of either the BeO or GO used to make a single 4g sample, at its particular GO wt.% content, by its respective theoretical density value; • Adding the resultant theoretical volumes of BeO and GO together to calculate the composite’s theoretical total volume; and

• Dividing the previously calculated theoretical volumes for BeO or GO by the total composite volume value.

Claims

CLAIMS:

1 . Ceramic nanocomposite material comprising a mixture of boron suboxide with boron carbide.

2. Ceramic nanocomposite material according to claim 1 , wherein the boron carbide is derived from graphene oxide.

3. Ceramic nanocomposite material according to claim 2, wherein the graphene oxide is reduced graphene oxide.

4. Ceramic nanocomposite material according to claim 2 or claim 3, wherein the amount of graphene oxide is present in an amount from 0-5 wt.%.

5. Ceramic nanocomposite material according to claim 4, wherein the amount of graphene oxide is present in an amount of less than 1 wt.%.

6. Ceramic nanocomposite material according to any preceding claim, having a density greater than 90% of the composite’s theoretical density, compared to pure boron suboxide.

7. Ceramic nanocomposite material according to claim 6, having a density greater than 98.5% of the composite’s theoretical density, compared to pure boron suboxide.

8. Ceramic nanocomposite material according to any preceding claim, having a radioactivity of less than 1 mSv upon exposure to 2 x 4 consecutive day periods over a two week period of a neutron fluence of 39.6 ± 3.41 n/cm² with energies of 13.7 ± 1.5 MeV.

9. Ceramic nanocomposite material according to any preceding claim, having a Vickers hardness of greater than 4 GPa as measured by ASTM 384-11e1 Standard Test for Knoop and Vickers Hardness of Materials.

10. Ceramic nanocomposite material according to claim 9, having a Vickers hardness of greater than 30 GPa.

11 . Ceramic nanocomposite material according to any preceding claim, suitable for use in ballistic armour applications.

12. Method of making a ceramic nanocomposite material, the method comprising the steps of: mixing boron suboxide with reduced graphene oxide into an aqueous suspension with deionised water so as to produce a blended substrate; filtering, drying and de-agglomerating the blended substrate; and consolidating the blended substrate using spark plasma sintering (SPS) so as to produce the ceramic nanocomposite material.

13. Method of making a ceramic nanocomposite material according to claim 12, wherein the boron suboxide and reduced graphene oxide are mixed into an aqueous suspension for up to 0.25 hour.

14. Method of making a ceramic nanocomposite material according to claim 12 or claim 13, wherein the blended substrate is dried for at least 2 hours at 125 °C.

15. Method of making a ceramic nanocomposite material according to any of claim 12 to claim 14, wherein the step of SPS comprises holding the substrate in a vacuum at less than 10’² bar.

16. Method of making a ceramic nanocomposite material according to any of claim 12 to claim 15, wherein the step of SPS further comprises applying a minimum contact force of 5 kN to the substrates.

17. Method of making a ceramic nanocomposite material according to any of claim 12 to claim 16, wherein the step of SPS further comprises heating the substrates from room temperature to a peak temperature and being held at the peak temperature for at least 1 minute.

18. Method of making a ceramic nanocomposite material according to claim 17, wherein the peak temperature is at least 1300 °C.

19. Method of making a ceramic nanocomposite material according to claim 18, wherein the peak temperature is at least 1570 °C.

20. Method of making a ceramic nanocomposite material according to any of claim 17 to claim 19, wherein upon reaching the peak temperature, the minimum contact force is increased to at least 16 kN.