COMPOSITIONS AND METHOD OF PRINTING CERAMIC MATERIALS

FIELD OF THE INVENTION

The present invention relates to 3D printing compositions comprising ceramic particles and to methods of printing 3D ceramic objects. In one embodiment, the invention has been developed for printing biocompatible ceramic materials such as bone implants for medical uses. However, it will be appreciated that the invention is not limited to this particular field of use. In another embodiment, the invention has been developed for producing printable composite materials with advanced self-folding functionality.

BACKGROUND OF THE INVENTION

The following discussion of the prior art is provided to place the invention in an appropriate technical context and enable the advantages of it to be more fully understood. It should be appreciated, however, that any discussion of the prior art throughout the specification should not be considered as an express or implied admission that such prior art is widely known or forms part of common general knowledge in the field.

Stereolithography (SLA) is a form of 3D printing technology used for creating models, prototypes, patterns, and production parts in a layer by layer fashion using photopolymerization, a process by which light causes chains of molecules to link, thereby forming polymers. Those polymers make up the body of a three-dimensional solid. Modem stereolithography devices use a UV laser to photopolymerise and cross-link each layer of the article being printed. With the help of computer aided manufacturing or computer-aided design (CAM/CAD) software, the UV laser is used to draw a pre- programmed design or shape on to the surface of a photopolymer vat to produce a single layer of the desired 3D object. Then, the build platform lowers one layer and a blade recoats the top of the tank with resin. This process is repeated for each layer of the design until the 3D object is complete. It is also possible to print objects "bottom up".

3D printing has traditionally utilised polymeric materials and has to date enabled production of components with complex geometries through use of computer design programs. However, pure polymeric printed materials are limited in their physical and mechanical properties and hence their applications. Accordingly, there is a need to expand the range of printable compositions to enable production of printed parts having a wider variety of properties.

One such avenue is development of ceramic printing compositions that enable printing of ceramic objects. Although some ceramic-based 3D printing compositions are known, it remains a challenge to produce ceramic printing compositions that are 3D printable and offer versatility in terms of final printed material properties whilst being cost effective. Accordingly, there remains a need to develop alternative ceramic printing compositions.

Furthermore, unlike for pure polymer printed objects, ceramic printed objects are generally first printed in a slurry form and then the shaped object is sintered (e.g., heated to temperatures of > 1000 °C) to remove volatile carrier molecules and to densify the structure. Accordingly, unlike polymeric objects that are shaped in a single 3D printing step, in order to preserve the shape of a 3D printed ceramic object, compositions and methodologies to date have been designed to ensure minimal or no shape distortion during drying or sintering.

Current 3D printing apparatus, including those used to print ceramic materials, use a layer-by-layer building technique to create structures with a wide variety of shapes. However, fabrication of ceramic layers with complex shapes is still a major challenge in part due to the inherent resolution of 3D printers. Additionally, accommodating shapes that have curves, twists and bends using traditional 3D printing methods requires printing layers that abruptly change geometry (e.g., narrow or widen) in a step-wise fashion such that the maximum surface area contact between any two layers is determined by the surface area of the narrowest layer in the pair. This can lead to structural weaknesses and unacceptable surface roughness in the final printed object. Accordingly, there is a need in the art for alternative printing methodologies for producing complex shapes that reduce the likelihood of structural weaknesses and/or surface roughness being present in the final product.

One printing methodology that shows promise for producing more complex 3D printed shapes is so-called “4D printing”, which combines 3D printing with a fourth dimension “shape morphing” step that allows the original printed shape to change. The fourth-dimension step can be an external stimulus such as light, heat, stress, magnetic field, chemical exposure etc. under which the original printed shape changes either reversibly or permanently. Examples of 4D printable materials demonstrated to date include shape-memory alloys such as NiTi alloys, and shape memory polymers such as composite polymers comprising fibres in a matrix. However, successful 4D printing methodologies using ceramic materials are scarce. Indeed, only three successful studies on shape morphing (4D printing) of ceramics have been reported to date. The first of these studies operates on the principle of anisotropic direction of reinforcement, whereby adjacent layers of printed ceramic materials contain functionalised ceramic platelets magnetically aligned along different horizontal axes. Twisting and bending is induced in the printed materials during sintering due to orthogonal anisotropic shrinkage in the layers creating internal stresses that result in deformation. However, such a process requires each magnetically aligned platelet layer to be coagulated to fix the platelet orientation prior to application of a subsequent layer, which introduces manufacturing complexity. The second of these studies reports a flexible, stretchable elastomeric (silicone rubber matrix) ceramic composite that is printed then physically deformed (e.g., stretched or bent) into a variety of complex shapes prior to heat treatment, which locks in the final ceramic shape. The third of these studies uses a highly concentrated titanium hydride (TiH ₂ )/acrylate copolymer ink system that is extruded into a lattice-like fabric sheet (so-called “direct write assembly”), then folded into a complex shape and annealed under heat. However, in each of these methods, the ceramic composition (generally printed as a flat lattice or sheet) must be shaped prior to sintering, which also introduces manufacturing complexity. There is thus a need in the art for methods of creating complex ceramic shapes that offer simplicity in manufacturing.

In view of the foregoing, it is an object of the present invention to overcome or ameliorate at least one of the disadvantages of the prior art, or to provide a useful alternative.

It is an object of an especially preferred form of the present invention to provide a 3D printing composition that is able to accommodate a variety of ceramic materials.

It is an object of an especially preferred form of the present invention to provide a method for producing 3D printed ceramic objects with complex geometries.

SUMMARY OF THE INVENTION

In its broadest form, described herein is a sinterable composition for 3D printing an object, the sinterable composition comprising: a photosensitive resin; ceramic particles; a dispersant; and optionally a low viscosity additive.

According to a first aspect, the present invention provides a sinterable composition for 3D printing a ceramic object, the composition comprising: about 5-50% by weight of a photosensitive resin; about 30-75% by weight ceramic particles; and about 2-40% by weight of a dispersant. The sinterable composition optionally comprises from about 0- 30% by weight of a low viscosity additive. The ceramic particles preferably have a median particle size of between about 100 nm and about 20 mm.

In one embodiment of the first aspect, the present invention provides a sinterable composition for 3D printing a ceramic object, the composition comprising: about 10-25% by weight of a photosensitive resin; about 40-65% by weight ceramic particles having a median particle size of between about 100 nm and about 20 mm; about 3-6% by weight of a dispersant; and about 15-30% by weight of a low viscosity additive.

In one embodiment of the first aspect, the present invention provides a sinterable composition for 3D printing a ceramic object, the composition comprising: 14-24% by weight of a photosensitive resin; 40-65% by weight ceramic particles having a median particle size of between 100 nm and 16 mm : 3.5-6% by weight of a dispersant; and 17.5- 30% by weight of a low viscosity additive.

The following features may be used in conjunction with the first aspect above either individually or in any suitable combination.

The photosensitive resin may comprise one or more photosensitive monomers or oligomers capable of being polymerized by light having a wavelength of between 350 and 450 nm. The photosensitive resin may comprise one or more photosensitive acrylate or methacrylate monomers or oligomers.

The photopolymer can be a commercial photopolymer or a photopolymer formulated by mixing monomers, oligomers, photoinitiators, and other additives such as photo-absorbers, dyes, and inhibitors. The monomer/oligomer can be (but is not restricted to) an acrylate -based monomer, an acrylamide-based monomer, a polyether, a acryloyl morpholine, a pholyethylene glycol, an epoxy-based monomer, or a combination of these and other monomers.

Examples of acrylate-based monomers are acrylates (e.g., Behenyl Acrylate, or 2- Hydroxyethyl Acrylate), diacrylates (e.g., Poly ethylene glycol) diacrylate), triacryaltes (e.g., Trimethylolpropane triacrylate), tetraacrylates (e.g., Di(trimethylolpropane) tetraacrylate), and methacrylates (e.g., (Hydroxyethyl)methacrylate).

An example of a polyether is polypropylene glycol.

Examples of acrylamide-based monomer are acrylamide, and N,N'- Methylenebisacrylamide .

An example of an epoxy based monomer is epoxy cyclohexane carboxylate.

The photoinitiator can be (but is not restricted to) peroxides (e.g. Benzoyl peroxide), nitrogen dioxide, camphorquinone, molecular oxygen, azobisisobutyronitrile, lithium phenyl-2, 4, 6-trimethylbenzoylphosphinate, benzoin methyl ether, 2,2-dimethoxy- 2-phenylacetophenone, 2 -hydroxy-2 -methylphenylpropane- 1-one, a-hydroxy- acetophenone, bis(2,6-dimethoxybenzoyl)-2,4,4-trimethylpentylphosphine oxide, 2- hydroxy-2 -methyl- 1 -phenyl -propan- 1 -one, 2,4,6-trimethylbenzoyldiphenylphosphine oxide, or a combination of these and other photoinitiators.

The ceramic particles can be from any group of ceramics including but not restricted to oxides (e.g., alumina, zinc oxide, zirconia, ceria), silicates (e.g., calcium silicates), phosphates (e.g. calcium phosphates), nitrides (e.g. silicon nitride), carbides (e.g. silicon carbide), borides (e.g. silicon boride), any combinations of these groups (e.g. oxycarbides), any multiphasic ceramic (e.g. Sr-HT-gahnite), or any doped version of these ceramics (i.e., these ceramics doped with one or more other elements).

In one embodiment, the ceramic particles can be selected from the group consisting of oxides (e.g., alumina, zinc oxide, zirconia, ceria), silicates (e.g., calcium silicates), phosphates (e.g. calcium phosphates), nitrides (e.g. silicon nitride), carbides (e.g. silicon carbide) and borides (e.g. silicon boride) or any combinations of these groups (e.g., oxycarbides), any multiphasic ceramics (e.g., Sr-HT-gahnite) or any doped version of these ceramics (i.e., these ceramics doped with one or more other elements).

The ceramic particles may be biocompatible ceramic materials including but not restricted to calcium phosphates, calcium silicates, hardystonite, gahnite, porcelain, any doped version of these ceramics (e.g., Baghdadite and Baghdadite doped with one or more other elements such as Mg, and Fe), or any multiphasic ceramic (e.g., Sr-HT-gahnite)

In one embodiment, the ceramic particles may be biocompatible ceramic materials selected from the group consisting of calcium phosphates, calcium silicates, hardystonite, gahnite and porcelain, or any doped version of these ceramics (e.g., Baghdadite and Baghdadite doped with one or more other elements such as Mg, and Fe). In one embodiment, the ceramic particles may be selected from the group consisting of: baghdadite, alumina, hardystonite, gahnite and porcelain. In one embodiment, the ceramic particles may be selected from the group consisting of: zinc oxide, zirconia, ceria, a carbide, a boride, and a nitride.

The ceramic particles may have a median particle size of between 1.1 mm and 16 mm. The ceramic particles may have a median particle size of between 100 nm and 500 nm. The dispersant may be a non-ionic dispersant, an ionic dispersant or a combination of dispersants. For example, the dispersant can be a surfactant, such as a non-ionic surfactant, an ionic surfactant or a combination of surfactants.

For example, the dispersant may be a polyoxoethylene sorbitol ester, or polyethylene glycol sorbitan monolaurate. The dispersant may be (but is not restricted to) a polyalkylene glycol (e.g., polypropylene glycol), alkyl gallate molecules (e.g., propyl gallate, butyl gallate, octyl gallate, lauryl gallate, octadecyl gallate), sodium dodecyl sulfate, cetyltrimethyl ammonium bromide, polyoxy ethylene octylphenyl ether, polyacrybc acid-co-itaconic acid, sodium pyrophosphate, sodium citrate, sodium carbonate, diammonium hydrogen citrate, or a combination of these and other dispersants.

The low viscosity additive may comprise but is not restricted to an acrylate or a diacrylate compound. The low viscosity additive may comprise poly ethylene glycol) diacrylate (PEGDA).

According to a second aspect, the present invention provides use of a composition according to the first aspect above as a 3D printing feedstock.

According to a third aspect the present invention provides a 3D printed object comprising or printed from the composition according to the first aspect above. The 3D printed object may be an implantable medical device.

According to a fourth aspect the present invention provides a method of 3D printing an object, the method comprising: (a) providing a composition according to the first aspect above; and (b) 3D printing the object using the composition as a feedstock. Step (b) is optionally carried out at elevated temperature. The composition is optionally mixed during 3D printing.

The following features may be used in conjunction with the fourth aspect above either individually or in any suitable combination.

The method may further comprise: (c) sintering the object produced by step (b).

The sintering may comprise heating the object to a temperature of > 500 °C, such as between 500 and 2000 °C, such as between 500 and 1800 °C, e.g. between 500 and 1500 °C for a period of time sufficient to volatilise organic components present in the object and densify the ceramic particles.

According to a fifth aspect the present invention provides a 3D printed object produced by the method of the fourth aspect above. The 3D printed object may be an implantable medical device. The 3D printing compositions claimed herein advantageously accommodate a range of different ceramic particles by using a “base composition” (composition base) comprising a photopolymerisable resin, a dispersant, and optionally a low viscosity additive, thereby enabling 3D printing of objects having a variety of physical and chemical properties resulting primarily from the properties of the ceramic particles incorporated into the base composition. As the base composition remains relatively constant for a wide range of ceramic particles, the 3D printing compositions herein represent a cost-effective solution to printing a variety of ceramic objects.

According to a sixth aspect the present invention provides method of 4D printing a ceramic object, the method comprising:

(a) providing a first 3D printing feedstock composition comprising a first concentration of ceramic particles;

(b) providing a second 3D printing feedstock composition comprising a second concentration of ceramic particles different to the first concentration;

(c) 3D printing an object having an initial 3D shape, wherein a first portion of the object is printed with the first 3D printing feedstock composition and a second portion of the object is printed with the second 3D printing feedstock composition; and

(d) sintering the object of step (c) such that differential shrinkage of the first and second portions caused by the different concentration of ceramic particles produces a ceramic object having a final 3D shape which is different to the initial 3D shape.

The following features may be used in conjunction with the sixth aspect above either individually or in any suitable combination.

The first 3D printing feedstock composition may be a composition according to the first aspect above. The first 3D printing feedstock composition may comprise between 40 and 60% by weight ceramic particles having a median particle size of between 100 nm and 500 nm.

The second 3D printing feedstock composition may be a composition according to the first aspect above. The second 3D printing feedstock composition may comprise between 40 and 60% by weight ceramic particles having a median particle size of between 100 nm and 500 nm.

The first 3D printing feedstock composition may have a ceramic particle concentration that is up to 35% by weight lower than a ceramic particle concentration of the second 3D printing feedstock composition. The first 3D printing feedstock composition may have a ceramic particle concentration that is between 5 and 35% by weight lower than a ceramic particle concentration of the second 3D printing feedstock composition. The first 3D printing feedstock composition and the second 3D printing feedstock composition may comprise ceramic particles having an identical chemical composition. The first 3D printing feedstock composition and the second 3D printing feedstock composition may comprise an identical photosensitive resin, an identical low viscosity additive, an identical dispersant, and comprise ceramic particles of an identical composition and an identical median particle size.

The first portion of the object may comprise a band, a disc, a layer, or an edge, and the second portion of the object completes the initial 3D shape. The first portion of the object and the second portion of the object may be printed such that there is a discrete ceramic particle concentration boundary between the first and section portions.

The method of the sixth aspect may further comprise step (b2) between step (b) and step (c): (b2) providing a third 3D printing feedstock composition comprising a third concentration of ceramic particles different to the first and second concentrations; wherein step (c) comprises 3D printing the object having the initial 3D shape, wherein the first portion of the object is printed with the first 3D printing feedstock composition, the second portion of the object is printed with the second 3D printing feedstock composition and a third portion of the object is printed with the third 3D printing feedstock composition; and step (d) comprises sintering the object of step (c) such that differential shrinkage of the first, second and third portions caused by the different concentration of ceramic particles produces the ceramic object having the final 3D shape which is different to the initial 3D shape.

The initial 3D shape may be selected to produce a predetermined final 3D shape after sintering. The sintering may comprise heating the object to a temperature of > 500 °C, such as between 500 and 2000 °C, such as between 500 and 1800 °C, e.g. between 500 and 1500 °C for a period of time sufficient to volatilise organic components present in the object and densify the ceramic particles.

According to a seventh aspect the present invention provides a ceramic object 4D printed according to the method of the sixth aspect above.

The following features may be used in conjunction with the seventh aspect above either individually or in any suitable combination.

The ceramic object may have a porosity of between 10 and 80%. The ceramic object may have a compressive strength of between 5 MPa and 100 MPa. The ceramic object may have an elastic modulus of between 1 GPa and 300 GPa. The ceramic object may be a medical device such as a cellular structure or bone. The ceramic object may be a machinery component such as an engine part.

According to an eighth aspect the present invention provides use of a sinterable composition according to the first aspect above in the method of the sixth aspect above.

According to a ninth aspect the present invention provides use of the method of the sixth aspect above for producing a complex shape.

The present invention advantageously permits controlled shape-changing or morphing of a 3D printed ceramic material during sintering such that advanced geometries including curves, bends and twists can be imparted to the original 3D printed structure (referred to herein as “4D printing”). 3D structures can be produced during sintering with features having pre-determined relative conformations with surprising accuracy and consistency. The present inventors have discovered that including domains of relatively higher or lower concentrations of ceramic particles in a 3D printed object causes the object to shrink in a non-uniform but predictable manner during heating (sintering) and thereby impart curves, bends and twists into the object. This is an advance over the prior art that seeks to directly print a curved, bent, or twisted 3D object with vertically offset layers. Advantageously, objects shaped using the methods of the invention are more homogenous compared to prior art methods of layer-by-layer printing. To explain, printing a curved structure layer-by-layer requires each printed layer to be slightly offset with respect to the previous layer on which it is printed. Whilst there is necessarily some overlap of the layers, there is also a portion of the subsequent layer which is not in contact with the previous layer, (i.e., commonly known as “overhang”, which is when the printed layer of material is only partially supported by the layer below). Inadequate support provided by the surface below the build layer can result in poor layer adhesion, bulging, or curling. However, the present invention substantially ameliorates the “overhang” issue, since the object can be layer-by-layer printed with no overhang and then curved into the required shape by heating/sintering, thereby reducing the likelihood of structural weaknesses such as layer cleavage during and after sintering. In this regard, the present inventors note that ceramics printed by the 4D printing methods described herein are less vulnerable to fragmentation, cracks, microcracks or other signs of severe internal stresses than other sintered 3D printed materials reported in the prior art as the sintering and shape change steps occur concurrently in the methods herein, allowing any stresses to relax before the ceramic is completely solidified. Furthermore, the printing methods claimed herein avoid the need for an additional shaping step, as the shaping is inherently controlled by the printing and sintering steps to produce a pre-determined 3D structure that has a different shape to the initially printed 3D structure.

The 4D printing method of the present invention enables the production of sintered articles that cannot be produced by traditional 3D printing (or are significantly more difficult to produce by traditional 3D printing), for example the production of materials comprising a ceramic layer of thickness < 100 mm where the ceramic shape is not flat, or the production of complex shapes where very smooth surfaces are required. Due to the layer-by-layer nature of 3D printing, it is difficult to obtain curved, thin layers that retain sufficient connection strength between layers. Furthermore, the same layer-by-layer technique used in conventional 3D printing methods necessarily produces surfaces with a final surface roughness that reflects the layering process and is therefore not as smooth as the final curved surfaces achievable using the 4D printing methods described herein.

The method can be used for diverse applications including but not restricted to electronic packaging, protective equipment (e.g., armour applications), bone scaffolds, scaffolds for other musculoskeletal conditions, implants in general, high temperature applications (e.g., nozzles for combustion systems), space applications (e.g. ceramic protective tiling on spacecrafts), harsh environment applications (e.g. highly acidic or basic mediums).

BRIEF DESCRIPTION OF THE DRAWINGS Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

Figure 1 shows diagrammatically a bilayer sheet (a) prior to sintering; and (b) after sintering; where the upper layer has a relatively lower density of suspended ceramic particles and the lower layer has a relatively higher density of suspended ceramic particles. The mismatch in ceramic particle concentration causes the bilayer to bend. Figure 1 (c) shows the effect of increasing the extent of difference of ceramic particle concentration between layers in an alumina bilayer sheet after sintering (from left to right): from equal concentration of particles in each layer; the upper layer having 50 wt% ceramic particles and the lower layer having 60 wt% ceramic particles (10 wt% concentration difference); the upper layer having 40 wt% particles and the lower layer having 60 wt% ceramic particles (20 wt% concentration difference). Figure 2 shows diagrammatically a manner of bending /folding a bilayer sheet structure (a) prior to sintering; and (b) after sintering; whereby a region of relatively low density of ceramic particles is printed in the centre of the upper sheet of a bilayer, thereby causing bending /folding about the low density region.

Figure 3 shows diagrammatically in top view the initial 3D shape of a printed rectangular plate object; (a) shows the object printed with a first portion at the edges having a different (lower) concentration of ceramic particles to the second portion in the centre; (b) shows the object printed with a first portion at the edges having a different (lower) concentration of ceramic particles to the second portion in the centre and the region in between printed with sub-regions of gradually increasing concentration of ceramic particles going from edge to centre.

Figure 4 shows diagrammatically in side view the predicted 3D shape of the rectangular plate objects in Figure 3 after sintering; (a) shows the object of Figure 3(a) after sintering with sharp edges; (b) shows the object of Figure 3(b) after sintering with rounded edges.

Figure 5 a shows diagrammatically in side view a cylindrical stent prior to sintering, and Figure 5b shows the stent after sintering and having a waisted portion formed as a result of a (lower) concentration of ceramic particles in the middle portion of the stent compared to the ends.

Figure 6a shows a cross-sectional view of a cylindrical stent prior to sintering, and Figure 6b shows the stent after sintering and having 4 hemi-spherical lobes as a result of a lower concentration of ceramic particles at 4 locations at the interior of the stent.

Figure 7a shows a cross-section of a honeycomb-type structure which, when sintered, reduces in overall dimensions thereby forming intricate internal structures (Figure 7b) that have a similar pore connectivity and pore size to that of natural bone.

Figure 8 is similar to the embodiment shown in Figure 6. In the embodiment shown in Figure 8a, a relatively lower concentration of ceramic particles is printed at 4 interior locations within the cylindrical stent, and whereby each printed layer is radially offset from the previous layer (Figure 8a). During sintering, the cylindrical stent radially contracts at the locations having a relatively lower concentration of ceramic particles producing 4 hemi-spherical lobes (when viewed in cross section) and simultaneously axially contracts, thereby producing a helical or “twisted” structure (Figure 8b).

Figure 9 shows a number of 3D printed shapes produced using 3D printing compositions of the invention described herein: (a) a latticework disc; (b) a composite shape having a latticework prism base and cylindrical upper layer (combination of dense and porous materials); (c) a latticework cylinder; (d) a solid cylindrical spiral; and (e) a complex shape in the form of a bone.

Figures 10a and 10b show human-scale personalized 3D printed bioceramic scaffolds with complex anatomical shapes. Figure 10a shows a 2cm -long scaffold to repair segmental defects respectively in a human mandible. Figure 10b shows a 4cm-long scaffold to repair segmental defects respectively in a human femur.

Figure 11 shows an example of a 3D printing set up.

Figure 12 shows the geometry of bioceramic scaffolds. The optical and SEM images, CT reconstruction, and pore size distribution of five architectures: (a) replica of what can be achieved with extrusion-based techniques (b) The negative replica of architecture “a” (c) Cubic (d) Rotated cubic (e) Body centered cubic (bcc). The CT reconstructed samples agree well with the optical images (scale bar 2mm). The SEM side and top views of the samples show the uniformity of the structure through the thickness, i.e., the sedimentation and aggregation of particles during printing were negligible (scale bar 1mm).

Figure 13 shows mechanical properties of scaffolds. The images I-V show the transformation of the behaviour of the material after initiation of the cracks (scale bar 4mm) during compression testing.

Figure 14: Design and fabrication of Baghdadite scaffolds (a) A cubic unit cell with dies length / and pore size p. (b) A rotated cubic unit cell with the same dimensions and a 3x2x2 array of the unit cell showing how the structure looks like (c) Optical and SEM images (top and side views) of the Baghdadite scaffolds manufactured through stereolithography. Scale bar indicates (c) 2mm.

Figure 15: A simple mechanical model cross section of a single strut (a) pre- implantation, and (b) with the new bone post-implantation; both with the stress and strain profiles along the cross-section in bending.

Figure 16 illustrates control over the amount of bending by adjusting the mismatch of concentration between the layers.

Figure 17 illustrates folding in 4D printing where the level of folding is controlled by the mismatch of concentration.

Figure 18 shows a porous scaffold.

Figure 19 shows compressive stiffness and strength for ceramic scaffolds with 50% porosity (reaching) compressive stiffness of 9 GPa and strength of 100 MPa). Figure 20 shows baghdadite implants with a porous core and a dense shell.

Figure 21a, 21b and 21c illustrate the properties of the structures of Figure 20. The stiffness and strength of these baghdadite implants reached 13GPa and 250 MPa respectively.

DEFINITIONS

In describing and claiming the present invention, the following terminology will be used in accordance with the definitions set out below. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments of the invention only and is not intended to be limiting. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one having ordinary skill in the art to which the invention pertains.

The recitation of a numerical range using endpoints includes all numbers subsumed within that range (e.g., 1 to 5 includes 1, 1.5, 2, 2.75, 3, 3.80, 4, 5, etc.). Use of the term “between” in the context of a numerical range is not intended to exclude the endpoints of the range. For example, a range of “between 1 and 100” is to be construed as “from 1 to 100”, i.e. including 1 and 100 within the range.

The terms “preferred” and “preferably” refer to embodiments of the invention that may afford certain benefits, under certain circumstances. However, other embodiments may also be preferred, under the same or other circumstances. Furthermore, the recitation of one or more preferred embodiments does not imply that other embodiments are not useful, and is not intended to exclude other embodiments from the scope of the invention.

Unless the context clearly requires otherwise, throughout the description and the claims, the words ‘comprise’, ‘comprising’, and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in the sense of “including, but not limited to”.

Other than in the operating examples, or where otherwise indicated, all numbers expressing quantities of ingredients or reaction conditions used herein are to be understood as modified in all instances by the term “about”. The examples are not intended to limit the scope of the invention. In what follows, or where otherwise indicated, “%” will mean “weight %”, “ratio” will mean “weight ratio” and “parts” will mean “weight parts”.

Notwithstanding that the numerical ranges and parameters setting forth the broad scope of the invention are approximations, the numerical values set forth in the specific examples are reported as precisely as possible. Any numerical value, however, inherently contains certain errors necessarily resulting from the standard deviations found in their respective testing measurements.

The skilled person will appreciate the term “biocompatible” defining a two-way response, i.e. the body’s response to the ceramic material and the material’s response to the body’s environment. The biocompatibility of a medical device refers to the ability of the device to perform its intended function, with the desired degree of incorporation in the host, without eliciting any undesirable local or systemic effects in that host.

As used herein, an “implant” refers to an article or device that is placed entirely or partially into an animal, for example by a surgical procedure.

The terms “base resin”, “composition base” and “base composition” are used interchangeably.

The terms “photopolymer”, “photopolymer resin” and “photosensitive resin” are used interchangeably.

A dispersant is a substance that is added to a suspension to improve the separation of the particles and to prevent settling or clumping of the particles.

A dispersant can also be called a dispersing agent. A dispersant can be a surfactant.

Porosity is a measure of the void (i.e. "empty") spaces in a material, and is a % of the volume of voids over the total volume, between 0% and 100%. The skilled person is well aware of methods for determining porosity. For example, the porosity can be measured by computed tomography or average density characterization (e.g., by pycnometer).

Room temperature is understood to mean 25°C, unless otherwise specified.

Elevated temperature is understood to mean above room temperature (e.g. 30°C).

DETAILED DESCRIPTION OF THE INVENTION

Composition for 3D printing

In its broadest form, described herein is a sinterable composition for 3D printing an object, the sinterable composition comprising: a photosensitive resin; ceramic particles; and a dispersant.

The sinterable composition may optionally further comprise a low viscosity additive. Thus, also described herein is a sinterable composition for 3D printing an object, the sinterable composition comprising: a photosensitive resin; ceramic particles; a dispersant; and a low viscosity additive.

The present inventors have discovered that, generally speaking, ceramic particles having a lower median particle size (D50) are able to be included in 3D printing compositions at lower concentrations than ceramic particles having a greater median particle size (D50) and still produce 3D printed objects able to be sintered successfully. Furthermore, some ceramics are able to be 3D printed and subsequently sintered when used in lower concentrations in the 3D printing composition described herein than other ceramic particles. Accordingly, a range of 3D printable compositions comprising a photosensitive resin, ceramic particles, a dispersant, and optionally a low viscosity additive, in different proportions, can be made for a range of different ceramic materials and/or ceramic materials having a range of different median particle sizes.

Furthermore, the present inventors have discovered that different ceramic materials, having optionally different median particle sizes, can be incorporated in the compositions herein at different minimum concentrations and still be sinterable after printing. Finally, the present inventors have discovered that inclusion of a low viscosity additive as defined herein advantageously assists achieving a printable viscosity of the composition, and additionally may improve the resolution of 3D printing in light-activated 3D printing processes such as stereolithography (SLA). Furthermore, in contrast to common organic solvents, the low viscosity additive as defined herein advantageously does not impede the performance of photoinitiators in the composition.

The inclusion of a low viscosity additive is optional. The inclusion of a low viscosity additive is preferred when 3D printing is to be carried out at room temperature (25°C). When 3D printing is to be carried out at elevated temperature (e.g. 30°C), a low viscosity additive may not be required. In some embodiments, the low viscosity additive as described herein has a cross- linking functionality, which is to say that it is capable of copolymerising with the monomers/oligomers in the photosensitive resin during printing. In such embodiments, the low viscosity additive may provide additional benefits to the printing composition, such as increased printing resolution, and/or may allow a lower sinterable concentration of ceramic particles to be included in the composition relative to a composition comprising a low viscosity additive incapable of cross-linking.

In one embodiment there is provided a sinterable composition for 3D printing an object, the composition comprising: about 5-50% by weight of a photosensitive resin; about 30-75% by weight ceramic particles; and about 2-40% by weight of a dispersant.

The composition optionally comprises about 0-30% by weight of a low viscosity additive. The ceramic particles preferably have a median particle size of between about 100 nm and about 20 mm.

In one embodiment there is provided a sinterable composition for 3D printing an object, the composition comprising: about 10-25% by weight of a photosensitive resin; about 40-65% by weight ceramic particles having a median particle size of between 100 nm and 20 mm; about 3-6% by weight of a dispersant; and about 15-30% by weight of a low viscosity additive.

In another embodiment there is provided a sinterable composition for 3D printing an object, the composition comprising:

14-24% by weight of a photosensitive resin;

40-65% by weight ceramic particles having a median particle size of between 100 nm and 16 mm;

3.5-6% by weight of a dispersant; and

17.5-30% by weight of a low viscosity additive.

Also contemplated herein is use of compositions described herein for printing 3D ceramic objects. Further contemplated herein is use of a composition as described herein as a 3D printing feedstock. 3D printed objects comprising the composition described herein are also within the scope of the present invention.

Photosensitive Resin

The compositions herein may comprise between about 5-50% by weight of a photosensitive resin. For example, the photosensitive resin may be present in the composition for printing at a concentration of between 5 and 50% by weight, e.g. between 5 and 45%, between 5 and 40%, between 5 and 35%, between 5 and 30%, or between 5 and 25%, or between 10 and 25% by weight, e.g., at a concentration of about 5, 6, 7, 8, 9,

10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33,

34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49 or 50% by weight.

For example, the compositions herein may comprise between about 10-25% by weight of a photosensitive resin. For example, the photosensitive resin may be present in the composition for printing at a concentration of between 10 and 25% by weight, e.g., between 10 and 18%, between 14 and 18%, or between 15 and 20%, or between 14 and 24%, or between 18 and 24%, or between 16 and 22 %, or between 16 and 25 % by weight, e.g., at a concentration of about 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23 or 24% by weight.

As intended herein, the photosensitive resin is a polymer-based binder/matrix that is a composition capable of being polymerized or cured or crosslinked using light provided by a 3D printer having any suitable wavelength. For example, the photosensitive resin may be a polymer-based binder/matrix that is a composition capable of being polymerized or cured or crosslinked by light having a wavelength of between 350 and 450 nm, e.g., between 350 and 400 nm, or between 400 and 450 nm, or between 390 and 410 nm, e.g., light of wavelength 405 nm, which is a wavelength of light commonly used in commercial 3D printers. The photosensitive resin thus generally comprises one or more monomers with an absorption band that covers the laser wavelength of the 3D printer being used, and preferably, the photosensitive resin comprises monomer(s) having an absorption peak at the laser wavelength. Resins may include monomers, oligomers, prepolymers, or mixtures thereof. The composition may also include one or more photoinitiators, optionally one or more solvents, and optionally one or more other components such as dyes or pigments. The monomers, oligomers, or prepolymers may be any suitable monomers/oligomers/prepolymers, but are preferably acrylate- or methacrylate-based. Preferably the monomers/oligomers/prepolymers include one or more functional groups selected from the group consisting of vinyl, ethynyl, vinyl ether, vinyl ester, vinyl amide, vinyl triazine, vinyl isocyanurate, acrylate, methacrylate, diene, triene, and functional analogs thereof. A "functional analog" herein means that the functional group has similar chemical and reactive properties, with respect to the polymerization of the monomers/oligomers/prepolymers.

The photopolymer can be a commercial photopolymer or a photopolymer formulated by mixing monomers, oligomers, photoinitiators, and other additives such as photo-absorbers, dyes, and inhibitors. The monomer/oligomer can be (but is not restricted to) an acrylate -based monomer, an acrylamide-based monomer, a polyether, a acryloyl morpholine, a pholyethylene glycol, an epoxy-based monomer, or a combination of these and other monomers.

Examples of acrylate-based monomers are acrylates (e.g., Behenyl Acrylate, or 2- Hydroxyethyl Acrylate), diacrylates (e.g., Poly ethylene glycol) diacrylate), triacryaltes (e.g., Trimethylolpropane triacrylate), tetraacrylates (e.g., Di(trimethylolpropane) tetraacrylate), and methacrylates (e.g., (Hydroxyethyl)methacrylate).

An example of a polyether is polypropylene glycol.

Examples of acrylamide-based monomer are acrylamide, and N,N'- Methylenebisacrylamide .

An example of an epoxy based monomer is epoxy cyclohexane carboxylate.

Suitable monomers are selected from the group consisting of: methyl methacrylate (MMA), Tetrahydrofurfuryl Methacrylate (THFMA), 2-hydroxyethyl methacrylate (MEMA), Vinyl Acetate (VAC). Other monomers may be utilised.

Suitable polymeric binders are selected from the group consisting of: Polymethylmethacrylate (PMMA), Polyvinyl Acetate (PVAc). Other polymeric binders may be utilised.

The photoinitiator can be (but is not restricted to) peroxides (e.g. Benzoyl peroxide), nitrogen dioxide, camphorquinone, molecular oxygen, azobisisobutyronitrile, lithium phenyl-2, 4, 6-trimethylbenzoylphosphinate, benzoin methyl ether, 2,2-dimethoxy- 2-phenylacetophenone, 2 -hydroxy-2 -methylphenylpropane- 1-one, a-hydroxy- acetophenone, bis(2,6-dimethoxybenzoyl)-2,4,4-trimethylpentylphosphine oxide, 2- hydroxy-2 -methyl- 1 -phenyl -propan- 1 -one, 2,4,6-trimethylbenzoyldiphenylphosphine oxide, or a combination of these and other photoinitiators.

Suitable photoinitiators may be selected from the group consisting of 2,2- dimethoxy-2-phenylacetophenone, 2-hydroxy-2-methylpropiophenone, camphorquinone, bis(2,4,6-trimethylbenzoyl)-phenylphosphineoxide, benzophenone, benzoyl peroxide, dicumyl peroxide, 2,2'-azobisisobutyronitrile, camphorquinone, oxygen, nitrogen dioxide, and combinations or derivatives thereof. Other photoinitiators or thermal free-radical initiators may be utilised.

The photoinitiator may be present from about 0.001 wt % to about 15 wt % of the photopolymer, for example. In various embodiments, the photoinitiator may be present at about 0.002, 0.005, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 wt% of the photopolymer. A "3D-printing resolution agent" is a compound that improves print quality and resolution by containing the curing to a desired region of the laser or light exposure. The 3D-printing resolution agent may be present from about 0.001 wt % to about 10 wt % of the photopolymer, for example. In various embodiments, the 3D-printing resolution agent may be present at about 0.002, 0.005, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 wt% of the photopolymer, for example. The 3D-printing resolution agent may be selected from UV absorbers, fluorescent molecules, optical brighteners, or combinations thereof. In some embodiments, the 3D-printing resolution agent is selected from the group consisting of 2-(2-hydroxyphenyl)-benzotriazole, 2-hydroxyphenyl- benzophenone s, 2-hydroxyphenyl-s-triazine s, 2,2 '-(2, 5 -thiophenediyl)bis (5 -tert- butylbenzoxazole), ethenediyl bis(4,l-phenylene)bisbenzoxazole, and combinations or derivatives thereof.

Suitable solvents may include an alcohol (methanol, ethanol, propanol, isopropanol, etc.) acetone, hexane, ether, etc.

The photosensitive resin may be colourless. The photosensitive resin may alternatively be pigmented, e.g., be black, white, or coloured.

Examples of suitable photosensitive resins for use in the present invention include, but are not limited to, GPCL02 and GPCL04 by Formlabs, Inc., MA, USA. GPCL02 and GPCL04 are clear resins comprising a mixture of proprietary methacrylated oligomers, a proprietary methacrylated monomer, and proprietary photoinitiator(s).

Ceramic

The compositions herein may comprise between about 30 and 75% by weight ceramic particles. For example, the ceramic particles may be present in the composition for printing at a concentration of between 30 and 75% by weight, between 30 and 70% by weight, between 40 and 75% by weight, between 40 and 70% by weight or between 30 and 65% by weight.

The ceramic particles preferably have a median particle size of between 100 nm and 20 mm, e.g., of between 100 nm and 16 mm.

For example, the compositions herein may comprise between about 40 and 65% by weight ceramic particles having a median particle size of between 100 nm and 20 mm, e.g., of between 100 nm and 16 mm. For example, the ceramic particles may be present in the composition for printing at a concentration of between 40 and 65% by weight, e.g., between 40 and 55%, or between 45 and 65%, or between 45 and 55%, or between 50 and 60%, or between 55 and 65%, or between 45 and 60 % by weight, e.g., at a concentration of about 40, 42.5, 45, 47.5, 50, 52.5, 55, 57.5, 60, 62.5, or 65% by weight. The ceramic particles have a median (D50) particle size of between 100 nm and 20 mm. e.g., of between 100 and 500 nm, or between 250 and 750 nm, or between 100 and 1000 nm, or between 100 and 1600 nm, or between 500 and 1500 nm, or between 750 and 1500 nm, or between 400 and 800 nm, or between 1000 and 1500 nm, or between 1 and 10 mm, or between 5 and 15 mm, or between 10 and 16 mm, or between 1 and 16 mm, or between 1 and 20 mm, e.g., of 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 1100, 1200, 1300, 1400 or 1500 nm, or 2 mm, 4 mm, 6 mm, 8 mm, 10 mm, 12 mm, 14 mm, 16 mm, 18 mm or 20 mm.

The inventors have determined that ceramic particles having a larger median particle size (e.g. 5-20 mm, e.g. 10 mm) can be included in the composition when 3D printing is to be carried out with a printer equipped with a resin mixer.

The inventors have determined that ceramic particles having a smaller median particle size (e.g. 100nm-5 mm) can be included in the composition when 3D printing is to be carried out with a printer not equipped with a resin mixer

Methods of refining ceramics to conform to these particle size distributions will be known to those of skill in the art, but by way of example may include colloidal ball milling the raw ceramic material with different diameter zirconia balls and a volatile solvent such as ethanol for several hours and then drying the slurry to evaporate the solvent.

Any suitable ceramic material may be used in the compositions herein, and the selection of a particular ceramic material may be made having regard to the physical and chemical properties of the ceramic material and/or the intended application of the final 3D printed object. For example, any multiphasic or single-phase ceramic may be used. The skilled person will appreciate that the invention described herein is not limited to any one particular ceramic material.

By way of non-limiting example, it is envisaged that biocompatible ceramic materials such as baghdadite (see WO 2009/052583), hardystonite (see WO 2010/003191), gahnite (see WO 2012/162753), or alumina may be used. Such materials are advantageously used to fabricate objects for biological applications, e.g., producing implants or other objects intended for contact with biological tissues or systems. Other ceramic materials such as oxides including zinc oxide, zirconia, ceria, etc. or non-oxides including carbides, borides, nitrides, or traditional ceramics such as clays including kaolinite (e.g., porcelain), earthenware, bone china, etc. may also be used. Additionally, it is envisaged that complex ceramic systems including but not limited to oxynitrides, oxycarbides, or carbonitrides may also be used.

Dispersant

The compositions herein may comprise between about 2-40% by weight of a dispersant. For example, the dispersant may be present in the composition for printing at a concentration of between 2 and 40% by weight, e.g., between 3 and 40%, e.g., between 2 and 35%, or between 2 and 30%, or between 2 and 25%, or between 2 and 20%, or between 2 and 15%, or between 2 and 10%, or between 2 and 6%, or between 5 and 6% by weight, or between 10 and 20% by weight, or between 20 and 30% by weight, or between 30 and 40% by weight, e.g., at a concentration of about 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34,

35, 36, 37, 38, 39 or 40% by weight.

In one embodiment, the compositions herein may comprise between about 3-6% by weight of a dispersant (also called a dispersing agent). For example, the dispersant may be present in the composition for printing at a concentration of between 3 and 6% by weight, e.g., between 3.5 and 6%, or between 3.5 and 4.5%, or between 4 and 5%, or between 4 and 6%, or between 3.5 and 4%, or between 4 and 5.5%, or between 4.5 and 6%, or between 5 and 6% by weight, e.g., at a concentration of about 3.0, 3.25, 3.5, 3.75, 4.0, 4.25, 4.5, 4.75, 5.0, 5.25, 5.5, 5.75 or 6.0% by weight.

The compositions herein may comprise a dispersant, such as a surfactant.

The dispersant may be any suitable dispersant but is preferably a non-ionic dispersant.

For example, the dispersant may be a polysorbate compound, such as a polyoxoethylene sorbitol ester, or polyethylene glycol sorbitan monolaurate.

The dispersant may be (but is not restricted to) a polyalkylene glycol (e.g., polypropylene glycol), alkyl gallate molecules (e.g., propyl gallate, butyl gallate, octyl gallate, lauryl gallate, octadecyl gallate), sodium dodecyl sulfate, cetyltrimethyl ammonium bromide, polyoxy ethylene octylphenyl ether, polyacrybc acid-co-itaconic acid, sodium pyrophosphate, sodium citrate, sodium carbonate, diammonium hydrogen citrate, or a combination of these and other dispersants.

The dispersant preferably comprises an oligomeric poly ethylene glycol) moiety, e.g., having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 5, or 10, or 15, or 20, or 25. In one embodiment, the dispersant used herein is poly ethylene glycol) sorbitan monolaurate having 20 ethylene oxide units, 1 sorbitol, and 1 lauric acid as the primary fatty acid (e.g., sold as TWEEN® 20 by Sigma- Aldrich®). Another example of a suitable dispersant for use in the present invention may include sodium polyacrylate (e.g., from Vanderbilt Minerals, LLC USA).

Low viscosity additive

In one embodiment, the compositions herein optionally comprise a low viscosity additive.

The need for a low viscosity additive is influenced by the temperature at which 3D printing is to be carried out. For example, it may be desirable to include a low viscosity additive when 3D printing is to be carried out at room temperature (25°C). In contrast, when 3D printing is to be carried out at elevated temperature, a low viscosity additive may not be required.

When compositions herein comprise a low viscosity additive, the compositions herein may comprise between greater than 0 and 30% by weight of a low viscosity additive, e.g. between about 15-30% by weight of a low viscosity additive. For example, the low viscosity additive may be present in the composition for printing at a concentration of between greater than 0 and 30% by weight, e.g., between 0.1 and 30%, between 1 and 25%, or between 2 and 25%, or between 5 and 25%, or between 2 and 20%, or between 5 and 20%, or between 1 and 55%, or between 1 and 10%, or between 1 and 5% by weight, e.g., at a concentration of about 0.1, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30% by weight. For example, the low viscosity additive may be present in the composition for printing at a concentration of between 15 and 30% by weight, e.g., between 17.5 and 30%, between 17.5 and 20%, or between 20 and 25%, or between 23 and 27.5%, or between 17.5 and 25%, or between 20 and 30%, or between 25 and 30%, or between 27.5 and 30%, or between 20 and 27.5% by weight, e.g., at a concentration of about 15, 16, 17, 17.5, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 or 30% by weight.

The low viscosity additive preferably has a viscosity of between 85 and 125 mPa.s as measured using a rotational viscometer at room temperature (25 °C) [film thickness: 1 mm]. For example, the low viscosity additive may have a viscosity of between 85 and 100 mPa.s, or between 100 and 125 mPa.s, or between 90 and 110 mPa.s, or between 95 and 105 mPa.s, e.g., may have a viscosity of 85, 90, 95, 97.5, 100, 102.5, 105, 110, 115, 120 or 125 mPa.s, as measured using a rotational viscometer at room temperature (25 °C). Accordingly, in one embodiment, the low viscosity additive used in the compositions herein is an additive having a viscosity of between 85 and 125 mPa.s as measured using a rotational viscometer at room temperature (25 °C).

In one embodiment, the low viscosity additive described herein is incapable of cross-linking with monomer(s)/oligomer(s) of the photosensitive resin. In other words, in some embodiments, the low viscosity additive described herein is devoid of photopolymerisable groups such as vinyl (C=C) groups. In such embodiments, the low viscosity additive may comprise an oligomeric poly(alkylene glycol) moiety. For example, the low viscosity additive may comprise an oligomeric poly ethylene glycol) moiety, e.g., having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 575, 700, or 1,000. The low viscosity additive may comprise an oligomeric poly ethylene glycol) moiety having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 700. In another embodiment, the low viscosity additive described herein is an oligomeric poly(alkylene glycol). For example, the low viscosity additive may be an oligomeric poly(ethylene glycol), e.g., having an average number of ethylene glycol units M _n ([O- CH ₂ -CH ₂ ] _n ) of 575, 700, or 1,000. The low viscosity additive may be an oligomeric poly(ethylene glycol) moiety having an average number of ethylene glycol units M _n ([O- CH ₂ -CH ₂ ] _n ) of 700.

In another embodiment, the low viscosity additive described herein is capable of cross-linking with monomer(s)/oligomer(s) of the photosensitive resin. In other words, in some embodiments, the low viscosity additive described herein is or comprises a compound comprising two or more photopolymerisable groups, such as two or more vinyl (C=C) groups. In such embodiments, the low viscosity additive may comprise any suitable monomer/oligomer having a suitably low viscosity and at least two photopolymerisable groups. However, in preferred embodiments, the monomer/oligomer is acrylate -based, i.e., comprises two or more acrylate groups. For example, the low viscosity additive may be or comprise a diacrylate, tri -acrylate or tetra-acrylate compound. The low viscosity additive preferably also comprises an oligomeric poly(alkylene glycol) moiety. In one embodiment, the low viscosity additive comprises two or more acrylate groups and an oligomeric poly ethylene glycol) moiety, e.g., having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 575, 700, or 1,000. Preferably, the low viscosity additive comprises two or more acrylate groups and an oligomeric poly ethylene glycol) moiety having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 700. In one embodiment, the low viscosity additive used herein is poly ethylene glycol) diacrylate (PEGDA), and in a particularly preferred embodiment, is PEGDA having an average number of ethylene glycol units M _n ([O-CH ₂ -CH ₂ ] _n ) of 700. Other examples of suitable low viscosity additives having a cross-linking functionality for use in the present invention may include urethane dimethacrylate (UDMA) and compounds comprising same, neopentyl glycol propoxylate (DPNPG) and compounds comprising same, or co- polymers of either of these compounds.

Composition properties

The sinterable compositions described herein may have any suitable viscosity to enable 3D printing, and in particular, to enable 3D printing via light-based printing methods such as stereolithography (SLA) and digital light processing (DLP). For example, the sinterable compositions herein may have a viscosity of between 800 and 6000 mPa.s as measured, e.g., using a rotational viscometer at room temperature (25 °C). The present inventors have discovered that compositions having a viscosity of between 800 and 6000 mPa.s, e.g., as measured using a rotational viscometer at room temperature (25 °C) are particularly suitable for 3D printing ceramic materials using commercially available 3D SLA printers as the compositions are advantageously adequately viscous to enable printing as well as photo-crosslinking and sintering.

The compositions described herein are sinterable, by which it is meant that they comprise a suitably high ceramic particle concentration for any given ceramic particle chemistry and average ceramic particle size such that after 3D printing, the resultant object is able to be sintered (i.e., heated) such that the organic components pyrolyse and the ceramic particles coalesce to form a stable porous or dense ceramic structure.

Manufacture of composition for 3D printing

Disclosed herein is a method of manufacturing a composition as described herein for 3D printing, the method comprising: combining and mixing a photosensitive resin, ceramic particles, a dispersant and optionally a low viscosity additive.

The photosensitive resin, ceramic particles, a dispersant and optional low viscosity additive can be combined simultaneously or sequentially in any order.

The photosensitive resin, dispersant, low viscosity additive, and ceramic particles are as described hereinabove.

The amounts of photosensitive resin, ceramic particles, a dispersant and optional low viscosity additive required to manufacture the sinterable composition are as described hereinbefore.

Also disclosed herein is a method of manufacturing a composition as described herein for 3D printing, the method comprising:

(i) combining a photosensitive resin, a dispersant and optionally a low viscosity additive, thereby forming a base resin; and

(ii) mixing ceramic particles into the base resin.

The ceramic particles preferably have a median particle size as herein described.

In one embodiment, the method of manufacturing a composition for 3D printing comprises: combining about 40% by weight of a photosensitive resin, about 10% by weight of a dispersant, and about 50% by weight of a low viscosity additive, thereby forming a base resin; and mixing the base resin in an amount of between about 35 and about 65% by weight with ceramic particles having a median particle size of between 100 nm and 1.5 mm in an amount of between about 35 and about 65% by weight (wherein the % by weight are % by total weight of the composition).

Method of 3D printing

Disclosed herein are methods for printing a 3D ceramic object using the compositions as described herein. For example, there is disclosed a method of 3D printing an object, the method comprising:

(a) providing a composition as described in the section above entitled “Composition for 3D printing”; and

(b) 3D printing the object using the composition as a feedstock.

Step (b) is optionally carried out at elevated temperature (e.g. 30°C).

Also disclosed herein is a 3D printed object produced by the method as described in this section. In one embodiment, the 3D printed object is an implantable medical device. However, other applications and devices are discussed further below in the section entitled “Applications”.

It is intended that the composition provided in step (a) is used as-is as a 3D printing feedstock; however, it is within the scope of the present disclosure that the composition of step (a) may be modified such as by heating, cooling or diluting with solvent as necessary to facilitate 3D printing, depending on the 3D printing apparatus used.

The 3D printing may be carried out at room temperature (25 °C). Alternatively, the 3D printing may be carried out at from about 25°C to 70°C, e.g. 25 °C, 30°C, 35°C, 40°C, 45°C, 50°C, 55°C, 60°C or 65°C. The 3D printing may be carried out at elevated temperature (above room temperature) to reduce the viscosity of the resin, e.g. at least about 35°C or at least about 40°C.

Any suitable 3D printing apparatus may be used in conjunction with the 3D printing compositions described herein. For example, a stereolithographic 3D printer such as supplied by Formlabs Inc. “Form 1+” may be used print the ceramic/photopolymer composition.

The 3D printing apparatus may optionally comprise a resin mixer (e.g. comprise one or more motorised mixing blades). The 3D printing apparatus may optionally heat the printing composition to elevated temperature.

Furthermore, any suitable printing platform may be utilised, e.g., a silanised glass build platform may be used. Any suitable computer-aided design (CAD) software may be used to design the 3D printed object shape and/or architecture, e.g., as supplied by Solidworks (Dassault Systemes) (MA, US).

After a 3D printed object is produced in step (b), the object may be subjected to post-printing treatments to anneal or otherwise fix the structure, depending on the application. Therefore, the method of 3D printing an object may further comprise an additional step after step (b) whereby the 3D printed object is sintered, hydrated, coated, melted, and/or crosslinked, e.g., to increase its structural integrity and/or permanency.

Sintering

In one embodiment, disclosed herein is a method of 3D printing an object, the method comprising:

(a) providing a composition as described in the section above entitled “Composition for 3D printing”;

(b) 3D printing the object using the composition as a feedstock; and

(c) sintering the object produced by step (b).

As used herein, the term ‘sintering’ refers to a step of heating the 3D printed object at any suitable temperature to effect drying, organic molecule pyrolysis and densifi cation. Such processes will generally both occur at temperatures > 500 °C, such as between 500 and 2000 °C, e.g. between 500 and 1800 °C, e.g. between 500 and 1500 °C.

For example, the 3D printed object may initially be heated to any suitable temperature to dry the 3D printed object and pyrolyse organic molecules present in the object such as the dispersant, low viscosity additive and/or photosensitive resin components. Non-limiting examples of suitable temperatures for an initial drying and pyrolysis step include those between 200 and 700 °C, e.g., between 200 and 400 °C, or between 300 and 500 °C, or between 400 and 650 °C, or between 500 and 700 °C, e.g., 200, 300, 400, 500, 600, 650, or 700 °C. Any suitable residence time at this temperature may be used, e.g., the object may be maintained at between 200 and 650 °C for several hours, e.g., between 1 and 6 hours, or between 2 and 4 hours, or for about 3 hours, etc. Any suitable temperature ramp rate may be used to heat the sample from room temperature to up to the pyrolysis temperature, e.g., a ramp rate of between 0.5 and 5 °C/min, e.g., between 0.5 and 1 °C/min, or between 1 and 3 °C/min, or between 0.5 and 1.5 °C/min, or between 2 and 4 °C/min, or between 4 and 5 °C/min, e.g., a ramp rate of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0 or 4.5 °C/min may be used.

The object may then be further heated as part of the sintering step to density the residual ceramic structure. Densification of the structure may be effected at any suitable temperature, but by way of non-limiting example, may occur at temperatures of > 1000 °C, such as between 1000 and 2000 °C, between 1000 and 1800 °C, such as between 1200 and 1500 °C, or between 1000 and 1500 °C, or between 1300 and 1500 °C, or between 1400 and 1600 °C, or between 1500 and 1800 °C, or between 1350 and 1400 °C, or 1000, 1100, 1200, 1300, 1400, 1480 1500, 1600, 1700 or 1800 °C for any suitable residence time, e.g., for several hours, e.g., between 1 and 6 hours, or between 2 and 4 hours, or for about 3 hours, etc. Any suitable temperature ramp rate may be used to heat the sample from pyrolysing temperatures of up to 700 °C to up to densification temperatures of up to 1800 °C, e.g., a ramp rate of between 0.5 and 5 °C/min, e.g., between 0.5 and 1 °C/min, or between 1 and 3 °C/min, or between 0.5 and 1.5 °C/min, or between 2 and 4 °C/min, or between 4 and 5 °C/min, e.g., a ramp rate of 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0 or 4.5 °C/min may be used.

The sintering step may be conducted in any suitable atmosphere, but preferably the sintering herein is conducted in an air atmosphere.

Method of 4D printing

Disclosed herein is method of 4D printing a ceramic object, the method comprising:

(a) providing a first 3D printing feedstock composition comprising a first concentration of ceramic particles;

(b) providing a second 3D printing feedstock composition comprising a second concentration of ceramic particles different to the first concentration;

(c) 3D printing an object having an initial 3D shape, wherein a first portion of the object is printed with the first 3D printing feedstock composition and a second portion of the object is printed with the second 3D printing feedstock composition; and (d) sintering the object of step (c) such that differential shrinkage of the first and second portions caused by the different concentration of ceramic particles in the first and second portions produces a ceramic object having a final 3D shape which is different to the initial 3D shape.

In one embodiment, the method of 4D printing described herein is devoid of a manual shaping step. That is, the 3D printed object having an initial 3D shape is sintered without an intervening shaping step, or the 3D printed object having an initial 3D shape is sintered without use of external support means such as a wire scaffold, or the 3D printed object having an initial 3D shape is sintered without an intervening shaping step and without use of external support means such as a wire scaffold. In another embodiment, the method of 4D printing described herein is devoid of a step between step (c) and step (d). In such embodiments, there is no external treatment (e.g., by heat, chemical, or mechanical means) between step (c) and step (d). Put another way, in one embodiment, in the method of 4D printing described herein, step (d) comprises sintering the object of step (c) without further treatment of the object by heat, chemical or mechanical means such that differential shrinkage of the first and second portions caused by the different concentration of ceramic particles produces a ceramic object having a final 3D shape which is different to the initial 3D shape.

Compositions for 4D printing The printing methods herein may utilise the 3D printing compositions also described herein and doing so may advantageously allow a range of different ceramic loadings to be used in the same base resin without compromising print quality. However, the methods herein are not limited to using such compositions, and the inventors contemplate that any composition known in the art for 3D printing ceramic objects may be used in the methods herein provided the amount of ceramic incorporated in the composition can be adjusted.

Accordingly, in one embodiment, the first 3D printing feedstock composition is a composition for 3D printing as described above in the section entitled “Composition for 3D printing”. For example, in one embodiment, the first 3D printing feedstock composition comprises a first concentration of ceramic particles in the range of 40-65% by weight of the total composition wherein the ceramic particles have a median particle size of between 100 nm and 16 mm. In another embodiment, the first 3D printing feedstock composition is a composition for 3D printing as described above in the section entitled “Composition for 3D printing” comprising a first concentration of ceramic particles in the range of 40-60% by weight of the total composition wherein the ceramic particles have a median particle size of between 100 nm and 500 nm.

In one embodiment, the second 3D printing feedstock composition is a composition for 3D printing as described above in the section entitled “Composition for 3D printing” comprising a second concentration of ceramic particles different to the first concentration in the range of 40-65% by weight of the total composition wherein the ceramic particles have a median particle size of between 100 nm and 16 mm. In another embodiment, the second 3D printing feedstock composition is a composition for 3D printing as described above in the section entitled “Composition for 3D printing” comprising a second concentration of ceramic particles different to the first concentration in the range of 40-60% by weight of the total composition wherein the ceramic particles have a median particle size of between 100 nm and 500 nm.

In a preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise ceramic particles of the same chemical composition. In another preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise ceramic particles of the same median particle size. In yet another preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise ceramic particles of the same composition and the same median particle size. In yet a further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise ceramic particles of the same composition, the same median particle size and the same D10 and D90 particle sizes. In a still further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise the same photosensitive resin. In a still further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise the same low viscosity additive. In a still further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise the same dispersant. In a still further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise the same photosensitive resin, the same dispersant, and the same low viscosity additive. In a still further preferred embodiment, the first 3D printing feedstock composition and the second 3D printing feedstock composition comprise the same photosensitive resin, the same dispersant, and the same low viscosity additive, and comprise ceramic particles of the same composition, the same median particle size and the same D10 and D90 particle sizes. Accordingly, in one embodiment, the only difference between the first 3D printing feedstock composition and the second 3D printing feedstock composition is the concentration of ceramic particles present.

Ceramic loading

In the 4D printing method described herein, the second 3D printing feedstock composition comprises a second concentration of ceramic particles different to the first concentration of ceramic particles in the first 3D printing feedstock composition. In one embodiment, the first 3D printing feedstock composition has a ceramic particle concentration that is up to 35% by weight lower than a ceramic particle concentration of the second 3D printing feedstock composition. For example, the first 3D printing feedstock composition may have a ceramic particle concentration that is up to 35%, or up to 30%, or up to 25%, or up to 20%, or up to 15%, or up to 10%, or up to 5%, or up to 1% by weight lower than a ceramic particle concentration of the second 3D printing feedstock composition, e.g., that is between 0.1 and 35% by weight lower, or between 0.1 and 10% by weight lower, or between 1 and 5% by weight lower, or between 5 and 15% by weight lower, or between 10 and 20% by weight lower, or between 15 and 25% by weight lower, or between 5 and 20% by weight lower, or between 5 and 35% by weight lower, or between 25 and 35% by weight lower, or between 20 and 30% by weight lower, or between 30 and 35% by weight lower, e.g., that is 35, 33, 30, 27, 25, 23, 20, 17, 15, 13, 10, 7, 5, 3, 1, 0.5 or 0.1% by weight lower than a ceramic particle concentration of the second 3D printing feedstock composition. By way of example, the first 3D printing feedstock composition may have a ceramic particle concentration of 60 wt%, and the second 3D printing feedstock composition may have a ceramic particle concentration that is 33% lower than this, i.e., [ ( 100-33 )/ 100] 60 = 40 wt%, or the first 3D printing feedstock composition may have a ceramic particle concentration of 50 wt%, and the second 3D printing feedstock composition may have a ceramic particle concentration that is 20% lower than this, i.e., [( 100-20)/ 100]x = 40 wt%. Other examples of concentration combinations will be readily apparent to a person skilled in the art.

The present inventors have shown that 3D printed compositions comprising relatively lower concentrations of ceramic particles cause portions of 3D objects printed from those compositions to shrink more during sintering than portions of a 3D object printed from compositions comprising relatively higher concentrations of ceramic particles. It is this effect that, surprisingly, allows the 4D printing methods of the present invention to achieve complex shapes without the need for stepped layer printing as known in the art.

Ceramic domains

In the 4D printing methods described herein, the 3D printed object having an initial 3D shape is printed such that a first portion of the obj ect is printed with the first 3D printing feedstock composition and a second portion of the object is printed with the second 3D printing feedstock composition. It will be appreciated that the first portion and second portion of the object are not limited in their size, geometry or proportion. Accordingly, the first portion of the object may be a band, a disc, a layer, an edge, etc. and the second portion of the object may form the remaining part of the initial 3D shape.

In some embodiments, the 3D printed object will be printed with two discreet domains, i.e., the first portion of the object and the second portion of the object will be printed in a discontinuous manner, such that there is a specific ceramic particle concentration boundary between the first and section portions. This embodiment is depicted in one form in Figure 3(a), which shows diagrammatically in top view the initial 3D shape of a printed rectangular plate object 1 printed with first portions 5 at the edges having a different (lower) concentration of ceramic particles to the second portion 10 in the centre. Figure 4(a) shows diagrammatically in side view the predicted final 3D shape 3 of the rectangular plate object 1 of Figure 3(a) with sharp edges.

In other embodiments, the first and second portions of the object will be separated by a region of intermediate ceramic loading such that there is a continuous concentration gradient from a higher ceramic concentration portion printed from the first 3D printing feedstock composition to a lower ceramic concentration portion printed from the second 3D printing feedstock composition. This embodiment is depicted in one form in Figure 3(b), which shows diagrammatically in top view the initial 3D shape of a printed rectangular plate object 2 printed with first portions 5 at the edges having a different (lower) concentration of ceramic particles to the second portion 10 in the centre and the region in between printed with sub-regions 6, 7, 8, and 9 of gradually increasing concentration of ceramic particles going from edge most sub-region 6 to centre most sub- region 9. Figure 4(b) shows diagrammatically in side view the predicted final 3D shape 4 of the rectangular plate object 2 of Figure 3(b) with rounded edges. In Figure 3(a), the first portions 5 have been printed with a step change in ceramic particle concentration compared to the second portion 10, resulting in a sharp bend being produced in the final sintered material. In contrast, in Figure 3(b) there is a gradual transition in ceramic particle concentration from the first portions 5 to the second portion 10, resulting in a shallow bend being produced in the final sintered material. Controlling the change in ceramic particle concentration enables control over the degree to which a bend can be introduced, and therefore control over the shape of the final sintered material.

Although the 4D printing method is described herein as comprising two 3D printing feedstock compositions and two 3D printed portions, the method is not so limited. It is envisaged that any number of 3D printing feedstock compositions may be used to print any number of 3D printed portions. For example, in one embodiment, there is provided a method of 4D printing a ceramic object, the method comprising:

(a1) providing a first 3D printing feedstock composition comprising a first concentration of ceramic particles;

(b1) providing a second 3D printing feedstock composition comprising a second concentration of ceramic particles different to the first concentration;

(b2) providing a third 3D printing feedstock composition comprising a third concentration of ceramic particles different to the first and second concentrations;

(c1) 3D printing an object having an initial 3D shape, wherein a first portion of the object is printed with the first 3D printing feedstock composition, a second portion of the object is printed with the second 3D printing feedstock composition and a third portion of the object is printed with the third 3D printing feedstock composition; and

(d1) sintering the object of step (cl) such that differential shrinkage of the first, second and third portions caused by the different concentration of ceramic particles produces a ceramic object having a final 3D shape which is different to the initial 3D shape. In this embodiment, it is envisaged that further steps (b3), (b4), (b5), (b6), (b7) ... (bn) may be included (where n is any positive integer) and represents, respectively, the fourth, fifth, sixth, seventh, eighth ... (n+1) th 3D printing feedstock composition. Similarly, it is envisaged that step (cl) may comprise printing an unlimited number of portions of the 3D object with the 3D printing feedstocks from each step (b n). Of course, it is envisaged that there may be more than one portion of the 3D object that is printed using any given 3D printing feedstock composition, provided that overall there are at least two portions of the 3D object printed with at least two feedstock compositions having different ceramic particle concentrations.

It will be understood that the portions of the 3D objects printed using the 4D printing method described herein are not limited, and therefore that any arrangement of the low/high ceramic concentration printing feedstock compositions may be used when printing the initial 3D object shape. For example, the printing feedstock compositions may be layered on each other, e.g., in 2, 3, 4, 5, 10, 20, 50, 100, 200, 500, 1000, etc. layers, or may be printed as regions within a single layer, or printed as regions within a single layer and either duplicated or varied on a layer-by-layer basis.

Sintering parameters

The 4D printing method herein comprises step (d) sintering the object of step (c) such that differential shrinkage of the first and second portions caused by the different concentration of ceramic particles produces a ceramic object having a final 3D shape which is different to the initial 3D shape. The sintering used in the 4D printing may be as described above in the section entitled “Sintering”.

Shape control

As noted above, portions of 3D objects printed from 3D printed compositions comprising relatively lower concentrations of ceramic particles shrink relatively more during sintering than portions of a 3D object printed from compositions comprising relatively higher concentrations of ceramic particles.

On this basis, the present inventors envisage that complex shapes such as curves, bends and twists can be achieved by printing different portions with different ceramic concentration 3D printing compositions. Regions of a 3D object where relatively greater shrinkage is desired on sintering are printed from lower ceramic concentration feedstock, and regions of a 3D object where relatively less shrinkage is desired on sintering are printed from higher ceramic concentration feedstock. By way of example, bowl-type shapes may be achieved by printing a low ceramic concentration layer over a high ceramic concentration layer. Bent-type shapes may be achieved by printing a band of low ceramic concentration in the middle of a high ceramic concentration layer, e.g., as depicted in Figure 2.

The final shape of the 3D objects printed using the 4D printing methods of the invention described herein may be any suitable shape, and will be determined in part by the nature of the 3D printing feedstock compositions, in part by the initial 3D shape, in part by the distribution of the different concentration ceramic regions in the 3D object, and in part by the sintering conditions used.

In another embodiment, one or more complex shapes can be 3D printed with the first composition and interconnected with one or “bridges” or filaments of the second composition. During sintering, the bridge will act to “pull” the complex shapes together and intertwine, interconnect, interleave, or interdigitate them to produce a shape that cannot be 3D printed directly, or is difficult to do. In this case, the bridge has a relatively lower concentration of ceramic particles than the shapes to which it is connected. The bridge may be a permanent feature of the resulting sintered material, or may be sacrificial and can be mechanically removed by breaking the bridge away from the final printed shape. In an alternative embodiment, the bridge may “push” the complex shapes away from each other and expand the final printed shape. In this case, the bridge has a relatively higher concentration of ceramic particles than the shapes to which it is connected. Combinations of these embodiments are contemplated herein.

Other embodiments contemplated herein include printing a spiral structure which, when sintered, “collapses” into a more tightly wound spiral which mimics the concentric lamellae of human bone. In this case, the external surface of the spiral has a relatively higher concentration of ceramic particles than the internal surface of the spiral.

In another embodiment, as seen in Figure 5, a cylindrical stent can be printed whereby the internal surface of the middle portion of the stent has a relatively lower concentration of ceramic particles compared to the ends of the printed stent such that, once sintered, a waisted portion is formed forming a pre-determined constriction in the flow path through the stent.

In another embodiment, as seen in Figure 6, a cylindrical stent can be printed whereby the internal surface is printed with a relatively lower concentration of ceramic particles at 4 locations such that, once sintered, those 4 locations contract to a relatively greater degree and 4 hemi-spherical lobes are produced (when viewed in cross section). It will be appreciated that a greater number (or fewer) lobes can be produced in the sintered material by printing a greater number (or fewer) locations having a relatively lower concentration of ceramic particles. It will also be appreciated that a helical or “twisted” structure can be produced by printing the locations (having a relatively lower concentration of ceramic particles) at different positions in the subsequent layers of the printed structure (see Figures 8a and 8b).

In another embodiment, as seen in Figure 7, a honeycomb structure can be printed which, when sintered, reduces in overall dimensions thereby forming intricate internal structures that have a similar pore connectivity and pore size to that of natural bone. This embodiment is particularly useful for applications in which at least some of the internal dimensions are too small to print, or too small to print reliably. This embodiment is useful to produce ceramic materials that more closely mimic the natural porosity of bone.

Figures 10a and 10b show human-scale personalized 3D printed bioceramic scaffolds with complex anatomical shapes, i.e. the same anatomical shape as the natural tissues and a cubic internal architecture for reconstructing. These large defects could result from diseases such as cancer, accidents, sport incidents, postnatal trauma, and congenital defects. Figure 10a shows a 2cm-long scaffold to repair segmental defects respectively in a human mandible. Figure 10b shows a 4cm -long scaffold to repair segmental defects respectively in a human femur. The resin used to print these human- scale scaffolds is 65 wt% ceramic particles mixed with 17.5 wt% photosensitive resin (e.g., Formlabs clear resin V4) and 17.5 wt% dispersant (e.g., Polyethylene glycol sorbitan monolaurate or TWEEN® 20, Sigma Aldrich).

Concentration-shrinkage maps

Concentration-shrinkage maps may be constructed for each ceramic particle chemistry in order to guide initial 3D printing of objects to achieve a desired final 3D shape after sintering. A general method for constructing such concentration-shrinkage maps is described below:

A circle-shaped layer of material with thickness of 50 mm and diameter 10 mm is 3D printed using a series of 3D printing compositions having a range of ceramic concentrations, such as 40, 45, 50, 55, 60, and 65 wt%, and a range of median particle sizes D50, such as 0.13, 0.2, 0.5, 1, 5, 10, 15, 20 mm. Using this methodology, some combinations of particle size and concentration will be suboptimal due to high viscosity, inadequate photocuring, or inadequate sintering. The boundaries of these suboptimal regions determine the optimal range for compositions of each ceramic material. For each optimal composition, the % reduction in the diameter and thickness between the pre- and post-sintered sample is recorded. For each chemistry, the shrinkage is then plotted a function of particle concentration and median particle size and thence used to guide the theoretical predictions of shape changes in complex multilayered systems.

Porous scaffold with shell

The present invention further provides a 3D printed object comprising a porous scaffold at least partially enclosed within a dense shell (e.g. at least about 80%, at least about 90%, at least about 95% enclosed within a dense shell).

Figure 20 shows examples of a porous scaffold, a porous scaffold with a dense shell and a porous scaffold with a dense, multilayer shell.

Figures 21a, 21b and 21c illustrate the properties of the structures of Figure 20.

The dense shell optionally comprises a plurality of layers (e.g. 2 or more, such as 2, 3, 4, 5, 6, 7, 8, 9 or 10 layers).

The porosity of the porous scaffold is higher than the porosity of the shell, e.g. at least about 5% higher than the porosity of the shell, e.g. at least about 10% higher, e.g. at least about 20% higher, e.g. at least about 30% higher, e.g. at least about 40% higher, e.g. at least about 50% higher, e.g. up to about 80% higher. The % difference in porosity is calculated as the porosity of the porous scaffold (%) subtracted from the porosity of the shell (%). So, for example, if the porous scaffold has a porosity of 80% and the shell has a porosity of 30%, the difference is 50%, i.e. the porosity of the scaffold is 50% higher than the porosity of the shell).

The porous scaffold is preferably 3D or 4D printed in accordance with the present invention.

The dense shell is preferably 3D or 4D printed in accordance with the present invention.

The porous scaffold and shell can be 3D printed together. Alternatively, the porous scaffold and shell can be 3D printed separately, optionally sintered separately and then assembled.

When the shell comprises a plurality of layers, the layers can be made either from the same printing feedstock composition or from different printing feedstock composition. The layers can be printed all together or printed separately and then assembled. In one embodiment, the porous scaffold and dense shell are printed using the same 3D printing feedstock composition.

In one embodiment, the porous scaffold and dense shell are printed using the different 3D printing feedstock compositions (e.g. different photosensitive resin, different ceramic, different median particle size of ceramic particles, different dispersant and/or different optional low viscosity additive).

The porous scaffold with a shell optionally comprises a crack-deflection interface as described below.

Crack-deflection interface

In one embodiment, a porous scaffold with a shell according to the present invention comprises a crack-deflection interface between the porous scaffold and the shell.

Figure 20 shows an example comprising a crack-deflection interface. The crack- deflection interface is visible in Figure 20 around the perimeter of the porous scaffold, i.e. at the boundary between the porous scaffold and the shell.

A crack impinging an interface joining two dissimilar materials may arrest or may advance by either penetrating the interface or deflecting into the interface. The purpose of a crack-deflection interface is to arrest a crack so that a crack which penetrates the porous scaffold does not penetrate the shell (or vice versa).

The crack-deflection interface comprises a crack-deflection material (e.g. a rubbery material such as a polymer, e.g. chitosan).

The porous scaffold and shell can be 3D or 4D printed together. Alternatively, the porous scaffold and shell can be 3D or 4D printed separately and then assembled.

In one embodiment the porous scaffold and the shell are 3D or 4D printed and sintered and then a crack-deflection material (e.g. a polymer) is injected to afford the crack-deflection interface.

In another embodiment the porous scaffold and the shell and the crack-deflection interface are 3D printed.

Applications and properties of 3D printed and 4D printed objects

Objects printed using the 3D printing compositions described herein and/or objects produced from the 4D printing methods described herein may have any suitable properties.

In one embodiment objects printed using the 3D printing compositions described herein and/or objects produced from the 4D printing methods described herein can have any suitable porosity, such as a porosity of between about 20 to about 30%. However, it will be appreciated that the objects could be configured to have lower or greater porosity according to the intended or desired use, and any porosity range would be within the purview of the present invention. For example, porosities of between 10 and 80%, e.g., of between 10 and 40%, or between 1 and 30%, or between 20 and 60%, or between 40 and 80%, or between 60 and 80%, e.g., of 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or 80% are possible. In some embodiments, the porosity achievable for 3D printed objects may be up to about 80%, e.g., between 10 and 40%, or between 20 and 60%, or between 40 and 80%, or between 60 and 80%. In other embodiments, the porosity achievable for 4D printed objects may be up to about 30%, e.g., between 1 and 30%, e.g., between 1 and 10%, or between 5 and 25%, or between 10 and 30%.

The pore size of objects printed using the 3D printing compositions described herein and/or objects produced from the 4D printing methods described herein may be any suitable pore size, e.g., the pore size may be between about 75 to about 200 mm. However, it will be appreciated that objects could be configured to have lower or greater pore size according to the intended or desired use, and any pore size would be within the purview of the present invention. For example, pore sizes of between 20 and 500 micron, or between 50 and 150 micron, or between 100 and 250 micron, or between 150 and 400 micron, or between 250 and 500 micron, or between 300 and 400 micron, or between 350 and 500 micron, or of 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 310, 320, 330, 340, 350, 360, 370, 380, 390, 400, 410, 420, 430, 440, 450, 460, 470, 480, 490, or 500 micron are possible. As the skilled person will appreciate, the porosity of ceramics can be adjusted by controlling the content and size of porogens.

The compressive strength of the ceramics of the invention, including porous ceramics, may be any suitable strength. For example, the compressive strength is preferably between 1 and 5000 MPa, such as between 1 and 6 MPa for highly porous structures (such as >60% porous structures, e.g., 65-8-% porous structures); or between 5 and 20 MPa for moderately porous structures (e.g., ~50% porous structures), e.g., between 5 and 15 MPa, or between 10 and 20 MPa, or may be between 50 and 250 MPa for dense structures, e.g., between 50 and 100 MPa, or between 80 and 120 MPa, or between 100 and 200 MPa, or between 150 and 250 MPa. It will be readily appreciated that the compressive strength of a structure will depend on its cross-sectional surface area. Accordingly, the compressive strengths in this paragraph may refer to the compressive strength of a square prismatic or cylindrical specimen having a radius of between 0.5 and

1 mm. In other embodiments, and with other sized specimens, other compressive strengths may be achieved. For example, compressive strengths of between 150 and 200 MPa, or between 125 and 225 MPa, or between 200 and 250 MPa, or between 175 and 250 MPa, or between 500 and 1000 MPa, e.g., between 500 and 750 MPa, or between 750 MPa and 100 MPa, or between 1000 and 5000 MPa, e.g., between 1000 and 3000 MPa, or between 3000 and 5000 MPa, or between 3000 and 4000 MPa, or between 3500 and 4500 MPa, or between 4000 and 5000 MPa may be achieved. Accordingly, compressive strengths of, e.g., 1, 5, 10, 50, 80, 100, 150, 200, 500, 750, 1000, 2000, 3000, 4000 or 5000 MPa may be achieved.

The elastic modulus of the ceramics of the invention, including porous ceramics, may be any suitable modulus. For example, the elastic modulus is preferably between 2 and 15 GPa for moderately porous structures (e.g., ~50% porous structures), e.g., between

2 and 7 GPa, or between 5 and 10 GPa, or between 7 and 15 GPa, or may be between 5 and 30 GPa, or may be between 50 and 150 GPa, or may be between 30 and 450 GPa for dense structures, e.g., between 30 and 100 GPa, or between 80 and 120 GPa, or between 100 and 200 GPa, or between 150 and 250 GPa, or between 200 and 300 GPa, or between 250 and 400 GPa, or between 350 and 450 GPa, or between 400 and 450 GPa. In other embodiments, the elastic modulus may be between 5 and 500 GPa, e.g., between 5 and 15 GPa, or between 12 and 18 GPa, or between 15 and 30 GPa, or between 20 and 40 GPa, or between 50 and 200 GPa, or between 100 and 300 GPa, or between 200 and 450 GPa, or between 250 and 500 GPa, such as 10, 20, 50, 100, 150, 200, 250, 300, 400 or 500 GPa.

These abovementioned compressive strengths and/or elastic moduli may be for ceramics with any suitable porosities, e.g., porosities of between about 20 to about 30%, or of between 10 and 80%, e.g., of between 10 and 40%, or between 20 and 60%, or between 40 and 80%, or between 60 and 80%, e.g., of 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75 or 80%. In one embodiment, the compressive strength of a structure of the invention composed of baghdadite may be between 1.8 and 5.1 MPa with porosities of between 65 to 78%. This is ideal for scaffolds to be placed in load-bearing applications as the strength of the natural bone is within this range. In another embodiment, the modulus and strength of a dense ceramic strut printed from a composition of the invention comprising baghdadite ceramic particles with a radius of between 0.5 and 1 mm is 250GPa and 80 MPa, respectively, which is very close to those of cortical bone. In yet a further embodiment, the modulus and strength of an object printed from a composition of the invention comprising baghdadite ceramic particles with 50% porosity and a cubic cellular structure has a modulus and strength of 5GPa and 10 MPa, which are close to those of tubercular bone . Other ceramics (such as Sr-Ht gahnite or alumina) will result in structures with higher mechanical properties than analogous structures composed of baghdadite.

Objects printed using the 3D printing compositions described herein and/or objects produced from the 4D printing methods described herein may be used to manufacture products suitable for use as medical devices such as implants, including high mechanical strength, resistance to fatigue, corrosion resistance, and/or biocompatibility.

For example, where the 3D printing compositions described herein comprise biocompatible ceramic particles, objects printed from such compositions may be used as medical devices. Similarly, where the 4D printing methods described herein utilise 3D printing feedstocks comprising biocompatible ceramic particles, objects printed using such methods may be used as medical devices. The medical devices may be a 3D implantable scaffold, an orthopaedic implant, e.g., for reconstructive surgery, a dental implant/prosthesis, a spine implant, an implant for craniofacial reconstruction and/or alveolar ridge augmentation, an implant for cartilage regeneration, an osteochondral defect implant, a strut, a stent or a stent-graft. However, it will be appreciated that there are many other devices which could be manufactured from the compositions and methods of the present invention.

The skilled person will readily appreciate how to manufacture a medical device from the compositions and methods of the invention. Suitable medical devices may include a bone implant, or a tooth filling implant, or a biocement, etc. Bone implants include, for example mandible sections and femur sections. Other bone implants for any part of the musculoskeletal system (e.g. craniofacial, tibia, etc) are also contemplated. Implants manufactured according to the methods of the invention may be implanted in animals, non-limiting examples of which include reptiles, birds, and mammals, with humans being particularly preferred. They may be implanted into a body in different ways, including, but not limited to subcutaneous implantation, implantation at the surface of the skin, implantation in the oral cavity, use as sutures, and other surgical implantation methods. In one embodiment, the compositions and methods of the invention are used to produce biocompatible ceramic objects that are fully synthetic bone graft substitutes.

In some embodiments the compositions and methods of the invention may be used to produce a biocompatible medical grade or implant grade object. In one embodiment, the object after sintering is essentially “pure” ceramic material from the ceramic particles, e.g., comprising greater than about 95%, and more preferably greater than about 99% ceramic material. For example, the purity may be between 50 and 100%, e.g., between 50 and 75%, or between 60 and 80%, or between 75 and 90%, or between 80 and 100%, or between 90 and 100%, or between 95 and 100%, or between 95 and 99%, e.g., the purity is greater than about 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99%, e.g., is about 50, 55, 60, 65, 70, 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100%.

In other embodiments, objects printed using the 3D printing compositions described herein and/or objects produced from the 4D printing methods described herein may be used to produce products suitable for industrial uses, e.g., components for machinery such as engines, including jet engines, building or construction parts such as pipes, linings, tiles, etc., homewares, etc.

Enhancing bone ingrowth and transforming the mechanics of bioceramic scaffolds through designed architecture and stereolithography 3D printing

The internal architecture of the bioceramic scaffolds plays a significant role in successful regeneration of bone tissue in critical size defects - an unmet clinical need. To achieve a precise control over the architecture, the inventors developed a design strategy and a stereolithographic free-form fabrication technique using a basic desktop printer not equipped with a resin mixer or elevated printing temperature. The effects of surface convexity, scaffold isotropicity, and relative orientation of scaffolds with respect to the host bone on the mechanical and fluidic properties of the scaffolds, and in turn quality and quantity of the new bone formation after 12-weeks implantation (in rabbit calvaria 10mm defects) were identified. The new bone formation significantly reduced the sensitivity of the scaffolds to cracks and transformed them into tough and strong bone-like materials.

Architectural characteristics of bioceramic scaffolds such as porosity, pore-size and permeability have shown to play significant roles on the volume and quality of new bone ingrowth. These characteristics have been controlled through fabrication techniques such as polymer foam replication, foaming, sacrificial templating, and more recently through direct or indirect 3D printing. 3D printing has the obvious advantage of more control over the architecture and in turn over pore connectivity and permeability. The most common 3D printing techniques for bioceramics are extrusion-based techniques (e.g. robocasting), laser sintering, photo-polymerization, and powder bed binding. Photo- polymerization includes techniques such as stereo-lithography (SLA), digital light processing (DLP), or 2-photon polymerization (2PP) and usually consists of slicing a 3D geometry into layers followed by sequentially depositing the layers through photo- crosslinking of a photosensitive ceramic resin. Particle -laden ceramic resins (also called slurry-based resins) usually constitute ceramic particles, a photosensitive polymer, and a dispersant. After 3D printing, high temperature is usually used to eliminate the organic constituents of these resins and to sinter the particles.

Photo-polymerization can result in freeform fabrication with high resolution circumventing (a) the characteristic geometrical limitations of extrusion-based technologies, and (b) the high energy laser requirement of the laser sintering systems. However, despite the use of photo-polymerization for ceramics such as alumina, silicon oxdie, carbide, or nitride, and zirconia, their strength in freeform fabrication is not yet exploited to study the effects of architecture (i.e., pore shape) on the bone regeneration capability and mechanics of bioceramic scaffolds. These limitations persisted mainly due to two main challenges. The first is developing a stable photocurable ceramic resin that can be printed using a desktop SLA or DLP printer and that can be used for different ceramics/bioceramics. SLA or DLP of ceramics with limited compositions (Alumina, or Zirconia) has been reported but usually require access to printers equipped with resin mixers to prevent sedimentation of particles and segregation of constituents, and with resin bath heaters to reduce the viscosity of the resin. The second challenge is isolating architectural effects in scaffolds: i.e., designing scaffolds with the same porosity, pore size and surface to volume ratio but different internal architectures. In this work, the inventors addressed these challenges and used combinations of mechanical tests to reveal the effects of architecture on the mechanics and bone regeneration capacity of bioceramic scaffolds.

Design and 3D printing of bioceramic scaffolds

The inventors developed a photo-polymerization based 3D printing technique that can be used for any ceramic material and can be implemented with a table-top SLA or DLP 3D printer (as disclosed herein). The success of the printing depends on two main factors. The first is optimum concentration and size of ceramic particles. Low concentration results in unsuccessful sintering and high concentration results in a viscous paste that cannot easily flow during the printing process. Also, the particles must be fine enough to prevent segregation or sedimentation during printing and in turn to prevent inhomogeneities in the printed material. The second factor is presence of enough photopolymer for effective photo-crosslinking of the resin. Personalized Baghdadite scaffolds: stereolithography, mechanics and in vivo testing

Photopolymerization-based 3D printing can enable development of large-scale personalized bioceramic scaffolds but are currently limited to few ceramic compositions. The inventors have developed stereolithography printing of Baghdadite (Ca-Mg-silicate) scaffolds with excellent fidelity to the shape of natural bone and precise internal architectures using a desktop printer, and explored controlled release of pharmaceuticals to promote bony ingrowth. Mechanical tests pre- and post-implantation, and mechanical modelling were used to assess bone formation and ingrowth and its effects on the mechanics of the scaffolds. The results revealed the clinical promise of the printing technology.

Photopolymerization-based 3D printing has a significant potential for making bioceramic scaffolds with anatomical shapes and precisely controlled internal architectures but its use has been limited to few ceramic compositions (e.g. alumina, SiC and zirconia) offered commercially at exorbitant cost. Also, currently bone morphogenetic protein BMP-2 is used in resorbable collagen scaffolds but the matrix has poor mechanical properties. The inventors have developed an inexpensive high resolution stereolithography printing method for ceramic (Baghdadite) scaffolds using a desktop printer.

While bone autograft remains the current gold standard in bone tissue engineering, it has limited availability in some patients and can lead to complications associated with donor site morbidity. Porous and interconnected ceramic scaffolds made of hydroxyapatite (HA), b-tricalcium phosphate (b-TCP), bioactive glasses, and more recently calcium silicates (not yet in clinical use) have gained significant interest as an alternative to autografts. Synthetic bone scaffolds can be made using a variety of techniques such foaming, sponge templating, and 3D printing. 3D printing is particularly appealing as it can produce controlled and predictable architecture and stmt topography that can consequently result in controlled permeability and pore interconnectivity crucial for bone regeneration. Other techniques can result in poorly defined and irregular internal architecture that may not provide proper mechanical support and may inhibit nutrient supply and vascularization, which consequently restricts the overall regenerative capacity of the scaffold.

Among printing techniques, photopolymerization through stereolithography (SLA) or digital light processing are particularly attractive. Firstly, these methods do not have the geometrical limitations of the extrusion-based robocasting techniques in which the geometry is limited by the shape of the nozzle. Secondly, they do not require high power lasers, such as those used in laser bed sintering systems. Thirdly, they are well- suited for both mass production of standard geometries as well as customized implants. Nevertheless, the use of photopolymerization for ceramics/bioceramics has been limited to a few compositions such as alumina, silicon carbides and zirconia due to stringent requirements for the photocurable ceramic resins. These requirements include low viscosity to allow flow of the resin, high stability to avoid segregation of components or sedimentation of ceramic particles, and high content of ceramics particles with tailored size for successful sintering.

The inventors have developed an SLA technique to print Baghdadite scaffolds. Baghdadite is a calcium-zirconium-silicate with chemical structure Ca ₃ ZrSi ₂ O ₉ and is used as the bioceramic material here because it has shown high levels of bioactivity: Baghdadite scaffolds previously produced by the sponge template method were able to promote bone regeneration in critical-sized radial defects in the rabbit (15 mm defect). The inventors looked at the effects of bone formation on the mechanics of scaffolds through compression testing and mechanical modeling.

EXAMPLES

The present invention will now be described with reference to the following examples which should be considered in all respects as illustrative and non-restrictive.

Example 1 : 3D printing composition Methodology

A commercial SLA 3D printer (Form 1+, Formlabs, US) was used. The photo- polymer was provided by the Formlabs (clear resin, code: FLGPCL02). This photopolymer resin was mixed with ceramic particles, a non-ionic dispersant (TWEEN® 20) and a low viscosity polymer (poly ethylene glycol) diacrylate). The resulting mixture was then 3D printed to variety of shapes (cellular structures, or shape of bone) and sintered to make ceramics with a variety of shapes.

The ceramic particles (baghdadite) which were received from Fraunhofer institute (Germany) had a median particle size D50 of ~10 mm measured using laser diffraction analysis (LA-960 HORIBA, Japan). These particles were refined to median particle size D50 ~1.3 mm using colloidal ball milling (Retsch PM 400, Germany) using the following process: 20wt% particles were mixed with 60 wt% zirconia balls and 20 wt% ethanol. The zirconia ball had three different diameters of 20, 3 and 1 mm in the ratio of 80 wt%, 15 wt%, and 5 wt%, respectively. The mixture was ball-milled for 5 hours at 250 rpm. The refined ceramic slurry was then dried at room temperature for one day.

The base resin was prepared by mixing 40 wt% photosensitive resin (FLGPCL02, Formlabs Inc., MA, USA) with 10% non-ionic dispersant (polyethylene glycol sorbitan monolaurate or TWEEN® 20, Sigma Aldrich) and 50 wt% poly ethylene glycol) diacrylate (PEGDA, Sigma Aldrich) to form a base resin (composition base). 45-65wt% of the dried baghdadite particles was mixed with 35-55 wt% of the base resin (composition base) to prepare the 3D printing composition.

A stereolithographic 3D printer (Formlabs, Form 1+) was used to print the ceramic/photopolymer mixture. A silanized glass build platform was used to improve the adhesion of the 3D printed part to the platform. To silanize glass, 3 vol% solution of 3- (trimethoxysilyl)propyl acrylate was prepared in a 5 vol% acetic acid mixed with 92vol% methanol. The 1mm thick glass slides were first washed with methanol three times and were then incubated in the silanization bath for 45 minutes. The slides were then washed with methanol and ethanol and dried by air pressure.

Solidworks (MA, US) was used to design the 3D shapes and the architectures. After the 3D printing was done, the samples were sintered in air. The green samples were heated at 1 °C/min to 620 °C at which temperature the samples were kept for three hours so that all the organic molecules were eliminated. The temperature was then increased at and 1 °C/min to 1400 °C at which temperature the samples were densified for 3 hours. The slow heating rate was important in obtaining ceramic parts with less internal voids.

Results

Ceramics with variety of external shapes (e.g. shape of femur) and internal architectures (e.g. cubic unit cells) were successfully 3D printed and tested in compression. A selection of such printed shapes are shown in Figure 9. The elastic modulus and compressive strength of dense ceramic struts made from 3D printed baghdadite with radius of 0.5-1 mm was found to be 250 GPa and 80 MPa, respectively, which is very close to those of cortical bone. The elastic modulus and compressive strength of a 3D printed baghdadite scaffold with 50% porosity and with a cubic cellular structure reached 5 GPa and 10 MPa, which are close to those of tubercular bone. Example 2: 4D printing (single & dual layer!

Methodology

The procedure was based on tailoring concentration of particles within a single or multiple layers of material using a DLP 3D printer. Different concentrations (between 40 and 60 wt%) of 100-500 nm sized alumina particles were mixed with the base resin (composition base) described in Example 1 (40 wt% photosensitive resin (FLGPCL02, Formlabs Inc., MA, USA) with 10% dispersant (polyethylene glycol sorbitan monolaurate or TWEEN® 20, Sigma Aldrich) and 50 wt% poly ethylene glycol) diacrylate (PEGDA, Sigma Aldrich) to form a base resin (composition base).

For this particular ceramic (alumina) and particle size range, at ceramic concentrations of >60 wt%, the high viscosity of the mixture prevented successful printing, and at concentrations of <40 wt%, the sintering did not succeed.

Results

In a bilayer case, two layers were printed from two different compositions, each composition comprising a different concentration of ceramic, the material bent during sintering because the layer containing lower concentration of ceramic particles contracted more during sintering than the layer containing higher concentration of ceramic particles. The results also demonstrated that the level of bending is a function of the concentration differential between the two layers: the bilayer with 40 wt% ceramic particles on the top and 60 wt% ceramic particles on the bottom (a 20% difference) bent more than the bilayer with 50 wt% ceramic particles on the top and 60 wt% particles on the bottom (10 wt% difference).

Example 3 : Human-scale scaffolds with complex anatomical shapes Methodology

The composition of the resin can contain for example 65 wt% ceramic particles mixed with 17.5 wt% photosensitive resin (e.g., Formlabs clear resin V4) and 17.5 wt% dispersant (e.g., Polyethylene glycol sorbitan monolaurate or TWEEN® 20, Sigma Aldrich). The intersection between the porous scaffolds, the CAD model of the human scale bone with complex geometry was used to design a porous structure with complex and large shapes. The software Solidworks was used for this purpose but any other design software can be used as well. The printing methodology was the same as discussed: the complex geometry was divided into layer (each 50 mm) and the layers are deposited on each other.

Results

The human-scale scaffolds obtained are shown in Figures 10a and 10b.

Example 4: Design and 3D printing of bioceramic scaffolds

The inventors developed a photo-polymerization based 3D printing technique that can be used for any ceramic material and can be implemented with a table-top SLA or DLP 3D printer (as disclosed herein). As noted above, the success of the printing depends on two main factors. The first is optimum concentration and size of ceramic particles. The second factor is presence of enough photopolymer for effective photo-crosslinking of the resin.

To satisfy these requirements, the inventors formulated a ceramic resin by mixing 47 wt% refined bioceramic (Baghdadite) particles (median size D50 1 .3 mm), with 20 wt% commercial photopolymer, 25 wt% a low viscosity monomer, and 8 wt% dispenser. The process resulted in a low viscosity ceramic resin that could be printed using a commercial SLA printer (Fig. 11; Formlabs Form 1+, MA, US). The viscosity in 24 °C (room temperature) was 1.38 Pa.s (Physica MCR 302 rheometer, Anton Paar, Austria) that is very close to the viscosity of commercial photopolymers. Also, the viscosity was found to be constant in shear rates ranging from 15 to 200 s ^-1 . While the developed printing process is versatile and can be used for any ceramic, the inventors used the bioceramic material “Baghdadite” (a Zr-doped calcium silicate: Ca ₃ ZrSi ₂ O ₉ ) because it has shown high levels of osteo-conductivity, -inductivity and -integration in different animal models.

To show the versatility of the shapes and architectures that can be printed, a few examples are included in Figures 9a to 9e.

The inventors used the freeform capability of their printing technique to study architectural effects in bioceramic scaffolds. Five different internal architectures were explored: (A) An architecture made of cylindrical struts (that can also be made using the more common robocasting procedure, Fig. 12a); (B) Negative replica of the architecture “A” (Fig. 12b); (C) Cubic (Fig. 12c); (D) Rotated cubic (Fig. 12d); and (E) Body centered cubic (bcc, Fig. 12e). Architectures B-E cannot be made with the common robocasting procedure, demonstrating the strength of the printing technique to determine architectural effects that were otherwise not possible to explore. The inventors implanted these scaffolds into rabbit calvarial defects and characterized their morphologies and mechanics pre- and post-implantation to answer five questions:

(i) Architecture “A” is made of cylindrical rods (i.e., convex internal surfaces), while its negative replica (architecture “B”) is essentially a solid material with cylindrical holes (i.e., concave internal surfaces). Comparison between these two architectures reveal whether the mechanics of initial scaffolds and quality/quantity of the new bone are affected by these two simple but opposite design pathways.

(ii) While previous studies show that permeability of scaffolds affect the bone formation, other variable such as porosity and pore size usually also change with the permeability (i.e., the effects of permeability was not isolated from other variables). By changing the orientation of the scaffolds with respect to the orientation of the bone (architecture “C” versus “D”), the inventors changed the permeability of the scaffolds along the direction of bone ingrowth while keeping all other parameters constant since the internal architecture of the scaffolds remained unchanged. The rotation is done in a way so that the strut perpendicular to the lattice plane [ 1 0 0] in architecture “C” is positioned perpendicular to the lattice plane [1 1 1] in architecture “D”. Fluid dynamic simulations showed that this rotation reduces that permeability of the scaffolds along both the direction of bone ingrowth (along x) and perpendicular to it (along z) by ~ 50%. Here, the hypothesis is tested whether bone ingrowth is the same in these scaffolds is spite of change in the permeability in direction of bone ingrowth.

(iii) Does increased isotropicity in cell geometry (i.e., same properties and permeability along x, y, and z directions) that is found in a cubic architecture “C” result is higher performance compared with the anisotropic architectures than can be achieved using robocasting (architecture “A”), and in which the permeability is 3x lower in direction of bone ingrowth (x or y) compared to z (see supplemental info of permeability calculations).

(iv) How is the performance the bcc architecture “E” compared to a simpler cubic architecture “C”?

(v) How is the performance of our printed bioceramics compared with a negative control (no scaffolds) and a positive control (a commercially available bio-glass).

To factor-out variables other than the architectural effects, it was necessary to keep the mean pore size, porosity, and surface area constant across different designs. It is however theoretically impossible to keep all these characteristics simultaneously constant. Therefore two sets of experiments were developed: in the first and main set, the inventors fixed the porosity and the average pore size across different architectures. The first step was to calculate volume fraction of the solid f (and in turn porosity that P = 1 - F ) for different architectures, which the inventors did by adding up the volume of all the struts followed by subtracting the volume of their intersections; the intersections were calculated following Angell et al. (I.O. Angell, M. Moore, Symmetrical intersections of cylinders, Acta Crystallographica Section A: Foundations of Crystallography 43(2) (1987) 244-250; for the notations, see Fig. 2):

The values of porosity obtained from equation 1 were the same as those obtained in in the computer aided design (CAD) models. To compute pore size, the inventors used a 3D definition following Hildebrand and Ruegsegger (T. Hildebrand, P. Ruegsegger, A new method for the model - independent assessment of thickness in three - dimensional images, Journal of microscopy 185(1) (1997) 67-75); a 2D definition can be problematic because the pore size can be very different when scaffolds are seen in different planes. For example, architecture “A” has pore size of 250 ± 18.7 mm when seen in xy plane and a pore size of 560 ± 42.7 mm when seen in yz plane. The inventors therefore defined the pore size as the maximum diameter of a virtual sphere which can fit within the architecture at any given point within the pore space. This definition represents both the largest diameter of the cell than can migrate through the scaffolds and the available volume for tissue ingrowth at any given point within the pore space. Before proceeding with fabrication and by using the CAD models, the inventors calculated the pore size distribution and consequently average pore size using the software CTan (MI, US).

The calculations for pore size and porosity were used to design CAD models of the scaffolds with 50% porosity and average pore size of 500 mm. This average pore size was selected as it was within the optimum range of pore size reported in the previous in- vivo studies. After fabrication, the structure of implants was analyzed using mCT (Fig. 12). To ensure consistency, the same scanning parameters, reconstruction parameters, and grayscale thresholding values were used for all the scaffolds. The samples were reconstructed in CTan (MI, US) to get the actual values of pore size distribution and porosity. The inventors realized that both porosity and pore size of the samples were 4- 8% less than the designed values, which the inventors attributed to the light scattering in the ceramic resin. The initial designs were therefore corrected accordingly to make scaffolds with the desired porosity and average pore size (Table 1). While average pore size was kept constant, each architecture still had a specific pore size distribution, that varied across different architectures (Fig. 12). Specific surface areas also varied slightly across different architectures. For completeness, the inventors considered a second set of experiments where porosity and specific surface area were kept constant between the architecture “a” and “b” (Fig. 12), while the mean pore size varied between the two architectures.

Table 1. the average pore size, porosity, and the surface to volume ratio of the scaffolds

Mechanical properties and permeability

The inventors used a combination of mechanical tests to determine the mechanical properties (pre— implantation). First, the modulus and strength of the base material (3D printed baghdadite) was determined: beams with thickness t = 0.3 mm, and width w = 3 mm were 3D printed, and tested in 3-point bending (N = 5). The thickness of the beam was selected to be in the same range as that of 3D printed struts. The modulus and strength of the baghdadite beams was found to be E ₀ = 28.1 ± 3.2 GPa and S ₀ = 30.2 ± 5.8 MPa respectively. The scaffolds were tested in compression pre- and post-implantation. The samples were fixed in 10% buffered formaldehyde. Previous studies showed that this fixation has negligible effects on the quasi-static properties of bone. If the properties of new bone are slightly degraded, the actual improvements after implantation could be more significant than the inventors found here; however, the current experiments are still very revealing to (a) determine the changes in the overall behaviour of the material after implantation, and (b) to determine the differences between the groups. To prepare the samples after implantation, six scaffolds for each group were harvested from rabbits using a 10mm diameter trephine bur. The bottom and top surfaces of all the scaffolds (pre- and post-implantation) were polished until a uniform surface was achieved. They were then placed on test fixtures with polished flat surfaces. The compression tests were performed at a constant displacement rate of 5 mm/sec using an Instron 5690 dual column instrument (MA, US). Optical images were taken every 5 seconds and were used to compute the compressive strains by manual image correlation: the crosshead displacements recorded in the optical images were used to compute strains. Fig. 13 shows snapshot images at five points during one of the tests (I to V for before implantation).

Pre-implantation, the stress increased in the I-II region due to elastic resistance of the material until the inventors reached the peak stress (i.e., strength) at which point a few struts broke (point II, cracks are shown on Fig. 13). The stress then decreased slowly as the struts continued to break progressively (region II- V) till almost all the struts were broken (point V). The stress re-increased after point V because the broken struts were compacted and crushed on each other (the densification effect). The tests stopped when this densification was observed. Pre -implantation, the trends for the stiffness can be explained by domination of compressive versus bending stresses in the scaffolds: the stiffness of the compression-dominated scaffolds (“A”, “B” and “C”) were almost the same and higher than the bending -dominated scaffold (“D”). For bcc architecture (“E”) that has both compression and bending dominated regions, the stiffness was between these two groups.

Pore size distribution calculations : The input to the software (CTan) is stack of images, each representing a layer of material (in this case 25 mm thick), that is sliced parallel to the horizontal plane. Prior to fabrication, the CAD (computer aided design) model of the scaffolds that was prepared using Solidworks (MA, US) was converted to a stack of images using an open source sheer software (Formlabs, MI, US). These images were then used in the software CTan to analyze the pore-size distribution. To build trust in the pore size distribution obtained from the software, the inventors tested simple designs such as a solid with cylindrical or spherical holes with a fixed diameter. The pore size obtained from the software matched the size of the pores in the design. mCT analysis before implantation: For each architecture, three disk-shape scaffolds (N = 3) with diameter and thickness of 10mm and 3mm, respectively, were scanned. Scaffolds were scanned using Zeiss Xradia Micro CT-400 scanner (Oberkochen, Germany) with 140 kV and 60 mA at a 24 mm voxel size resolution. 450 images were taken for each CT scan (every 0.8°), and Xradia software was used for reconstruction of 3D geometry from the images.

Polishing scaffolds for mechanical tests : A polishing holder that contained holes the same diameter as that of the scaffolds (10 mm) but with different depths (3.2mm and 2.8mm). Initially the samples were put in the deeper hole and were polished. Then the inventors rotate the sample and put it inside the hole that is 2.4 mm deep. To remove the samples, there is smaller hole at the other side of the holder: the inventors use a small pin to remove the sample after polishing is done.

Summary

(1) the inventors provided a fast and efficient way by using topology analysis to design scaffolds with the same porosity and average pore size (or same porosity and surface area).

(2) the inventors developed a SLA techniques to print bioceramic scaffolds with any internal architecture. The printing method can be implemented by any table-top SLA or DLP printer (as opposed to expensive ceramic 3D printers).

(3) the inventors showed that freeform fabrication achieved through SLA can be used to decrease the internal stresses and improve strength of the scaffolds: the design based on holes within a solid material showed higher strength compared with a design based on cylindrical struts with the same porosity and average pore size.

(4) the stiffness of the scaffolds were only a function of whether they were bending or stretch dominated.

(5) the strength, permeability, and bone regeneration capability of the scaffold with cylindirical holes (B) or cubic architecture (B) were higher than what could be achieved with robocasting (A) demonstrating the shows the strength of photo-printing in development of scaffolds with improved performance for tissue engineering.

Example 5 : SLA technique to print Baghdadite scaffolds

The inventors have developed an SLA technique to print Baghdadite scaffolds. Baghdadite is a calcium-zirconium-silicate with chemical structure Ca ₃ ZrSi ₂ O ₉ and is used as the bioceramic material here because it has shown high levels of bioactivity: Baghdadite scaffolds previously produced by the sponge template method were able to promote bone regeneration in critical-sized radial defects in the rabbit (15 mm defect). Synthesis of Baghdadite powder

Solid state synthesis was used to prepare the powder. It was preferred over the sol gel method because it is more suitable for development of large quantity of the powder needed for 3D printing. The precursors were calcium oxide CaO, zirconium oxide ZrO ₂ , and silicon oxide S _i O ₂ , all sourced from Sigma Aldrich. Stochiometric amounts of precursors were mixed with ethanol and zirconia balls with ratios of 20 wt%, 20 wt%, and 60 wt% respectively. The zirconia ball had three different diameters of 20, 3 and 1 mm with ratios of 80 wt%, 15 wt%, and 5 wt% respectively. The mixture was ball-milled for 5 hours at 200 rpm (Retsch, PM 400, Düisseldorf, Germany). The refined ceramic slurry was dried at 80 °C for one day. The dried material was manually ground with a mortar and pestle, which was then filtered through a 200 mm sieve to remove the large clusters. Pallets were prepared using uniaxial dry pressing under ~ 0.15 GPa for 2 min (Hydraulic press). They were heated to (and kept 1 hour at) 600 °C at 2 °C/min to remove organic residues (if they exist). The temperature was then increased to 1400 °C at 2 °C/min, at which temperature the samples were kept for 3 hours for the reaction between the precursors to be completed and the sample to be fully sintered. X-ray diffraction analysis (XRD) was used to confirm that the developed composition is Baghdadite: the diffraction pattern matched that from the database. The sintered Baghdadite pallets were broken and ground using ball milling at 150 rpm for 8 hours in ethanol medium. At this stage, the same ratios of 20 wt%, 20 wt%, and 60 wt% were used for the baghdadite, ethanol and Zirconia balls respectively but only 20mm-diameter balls were used. The particles were then dried at 80 °C. These ball milling parameters were used to obtain a medium particle size D ₅₀ = 9.52 mm required for successful printing/sintering. The particles size measurement was done using laser scattering particle size analysis (Horiba, Japan).

Stereolithography 3D printing

The inventors developed a photosensitive ceramic resin (as described herein) for stereolithography printing of Baghdadite scaffolds using a desktop printer. SLA of ceramics involves dividing a complex 3D geometry into an array of layers, and depositing the adjacent layers through photo-polymerization of a photosensitive ceramic resin followed by sintering. The technique provides a powerful pathway for freeform fabrication of bioceramics but it has been mostly used indirectly for bioceramic scaffolds; i.e. a polymeric mold is made by SLA, is filled with the bioceramic resin and is then degraded at high temperatures to yield the bioceramics scaffold. Direct SLA printing of bioceramics has been hindered by problems such as segregation of resin constituents, particle sedimentation, and high viscosity of resin. The inventors avoided these problems by:

(a) formulating a ceramic resin with optimized particle size and concentration of constituents: our resin was composed of 65wt% bioceramic (Baghdadite) particles refined to median particles size D50 = 9.52 mm, 17.5% a commercial photopolymer (Clear resin V4, Formlabs, MI US), and 17.5% dispersant (Tween 20, Sigma Aldrich). Smaller particle size while keeping the same concentration (needed for successful sintering) resulted in a viscous paste that was hard to flow/print. Larger particle size on the other hand resulted in low mechanical properties after sintering or unsuccessful sintering. The concentration of the photopolymer and the dispersant were crucial as well. Lower concertation of the photopolymer resulted in unsuccessful photo-crosslinking, while lower concentration of dispersant resulted in increased resin viscosity and in turn significant decrease in printing success.

(b) using a desktop 3D printer (Form 2, Formlabs, US) equipped with a resin mixer and an elevated working temperature: mixing the resin after printing each layer prevented sedimentation of the particles during printing, resulted in re-mixing the potentially segregated components, and displace the partially cured resin, ultimately improving the quality of the prints. Increasing the temperature of the resin on the other hand reduced the viscosity and in turn facilitated the printing. The viscosity of resin at 31 °C (the operating temperature of the utilized desktop 3D printer: Formlab, Form2) was found to 4 Pa.s at 100 s ^-1 (Physica MCR 302 rheometer; Anton Paar, Austria), that is lower than its viscosity at room temperature (25 ^° C) and is in the range of other photosensitive ceramic resins.

(c) using an effective mixing procedure: the components were mixed using ball milling (Retsch PM 400, Germany) at 150rpm for 30 min with resin/ball weight ratio of 2/1 (20 mm zirconia balls were used) - a process that resulted in a uniform paste.

The scaffolds were deign using software Solidworks (MA, US), and were printed using layer thickness of 50 mm. The 3D printed components were first heated at 2 °C/min to 600 °C at which temperature the samples were kept for three hours to eliminate the organic components of the resin. The samples were then heated to 1400 °C at the same rate and were kept at this temperature for three hours for sintering the ceramics particles. Mechanics: experiments and model

To examine the effects of new bone formation on the mechanics of the scaffolds, the inventors tested the scaffolds in quasi-static compression under in-situ imaging pre- implantation. The top and bottom surfaces of the scaffolds were polished and made parallel to avoid stress concentrations. Polishing holders made of stainless steel were made for this purpose. The initial height of the scaffolds were all made to 5.5 cm and a displacement rate of 5 mm/s was used in the tests. The in-situ optical images were used to calculate the actual values of average strain using the crosshead displacements.

Results

Manufacturing of 3D printed Baghdadite scaffolds

The scaffolds were designed to have cylindrical shape (dimeter 4m; height 6 mm) a rotated cubic architecture (Fig. 14 a, b). The architecture is constructed by rotating the cubic architecture (Fig. 14a) 45 ^° around the z axis, and then 30 ^° around the x axis, effectively locating the strut that was along x axis (vector [1 00]) in the cubic architecture along [1 1 1] direction in the rotated cubic. While cubic is a compression/tension dominated architecture when loaded vertically (along the z axis), this rotation procedure makes it bending-dominated along the z axis. This would make the scaffolds less stiff along z, that is desired here because the differences in the new one formation would have a more pronounced effect on the mechanical properties of the scaffolds. The scaffolds were designed to have porosity of 50% and pore size p = 400 mm that is within the optimum range of pore size for new bone formation. The CT reconstruction of the scaffolds and the SEM images showed the porosity P = 52 ± 4.3 % and pore size p = 400 ± 27.2 mm - values very close to the designed ones (number of samples N= 5).

To demonstrate the strength of the inventors’ manufacturing approach for development of scaffold with complex anatomical shapes, human-sized 2cm- and 4cm- long scaffolds to repair segmental defects respectively in human mandible and femur are printed (Figures 10a and 10b). These large defects could result from diseases such as cancer, accidents, sport incidents, postnatal trauma, and congenital defects.

Example 6: Modelling Scaffold Biomechanics

To capture the effect of bone formation on the mechanical properties of the scaffolds, the inventors developed a mechanical model. Both stmts and the new bone around them were modeled as linear elastic materials with elastic modulus E _s and E _b respectively. Since the scaffolds in this study (rotated cubic) were bending dominated, the properties of a single stmt in bending was compared before and after bone formation. The stmt with the new bone was modeled as a composite beam made of bioceramic and the new bone. To simplify the problem, the bone formed around the stmts were assumed to follow the circular shape of the struts (Fig. 15a). Under bending, the strain of the composite beam linearly increases from the neutral axes ( e ₀ =0) to the edge that has maximum strain e _max (Fig. 15b) so that strain e(r,q) is stated as:

Since the materials are modelled as linear elastic, the stress s(r,q) is stated as: s( r,q) = e(r,q)E (2)

This implies that the stress is discontinuous and is higher for the section with higher elastic modulus (that is usually the ceramic stmts compared with the new bone). This stress distribution results in a bending moment:

Combining equations 1-3 affords:

Where is the second moment of inertia of the cross-section areas for each material. For the same amount of load on the scaffolds pre- and post-implantation, the stmts would be under the same moment but different strains and stresses. By using equation 4 and writing the second moment of inertia of the stmt as the maximum strain pre-implantation would be:

Post-implantation (assuming that the mechanical properties of the stmts is not affected by ion release during implantation), the maximum strain for the same amount of load would reduce to:

It is shown in the previous work that the failure of brittle scaffolds such as the Baghdadite scaffolds in this study is governed by the maximum principal stress within the material. To predict the level of improvement as a function of amount of bone formation, the inventors therefore obtained the maximum principal stress in the scaffold pre- implantation and post-implantation can be written as (Fig. 15b):

The ratio of these two stresses shows the level of improvement that can be achieved in the strength of the scaffolds because of new bone formation by combine g equations 5- 7, this ratio is stated as:

The model shows the dependence of the strength amplification (i.e. reduction of maximum stress) to two nondimensional parameters E _b /E _s that shows the linear dependence to the material properties and (r _b /r _s ) ⁴ that shows the nonlinear dependence to the amount of new bone formed within the material. It is reasonable to assume that the amount of new bone formed within the scaffolds scales with r (i.e. with the area of the newly formed bone). The ratio is 1 when there is no bone formed at the interface (r _b r _s ) and increases nonlinearly with the increase in the radius of the new bone. To compare the model with our experiments, the inventors measured the elastic modulus of the struts (E _s ) by 3 pint bending test on 3D printed beams that had the same thickness as the struts and were sintered using the same sintering parameters.

Although the invention has been described with reference to specific examples, it will be appreciated by those skilled in the art that the invention may be embodied in many other forms. In particular, features of any one of the various described examples may be provided in any combination in any of the other described examples. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment(s) disclosed, but that the invention will include all embodiments falling within the scope of the appended claims.