FIELD OF THE INVENTION

[ 0001 ] The present invention relates to methods and apparatus for separation of molecules, ions, colloids and/or particles and in particular to methods and apparatus for separation of nano-scale molecules, ions, colloids and/or particles.

[ 0002 ] The invention has been developed primarily for use as method and apparatus for separation of nano-scale molecules, ions and/or particles for in-line particle separation applications and will be described hereinafter with reference to this application. However, it will be appreciated that the invention is not limited to this particular field of use (e.g., such a device could also be a method & apparatus for the concentration of nano-scale molecules, ions, colloids and/or particles).

BACKGROUND

[ 0003 ] Any discussion of the background art throughout the specification should in no way be considered as an admission that such background art is prior art nor that such background art is widely known or forms part of the common general knowledge in the field.

[ 0004 ] Separation science is of great importance to numerous industries (e.g., pharmaceuticals/medical, mining, electroplating, agriculture, desalination, and others). Known separation methods are typically performed using complex and expensive methods that can be difficult to implement outside of the laboratory or in routine 'inline' processes or in situ. Further to this, known separation methods are typically batch type separation methods where only a small sample of the mixture is separated at a time usually because the separation occurs in the same direction as the flow making it impossible to continuously add sample. Conversely, a continuous flow separation is one in which the separation occurs at an angle to the flow direction allowing continuous sample addition, collection, analysis and monitoring and are well suited to 'in-line' processes. Figures 1A and 1B depict the typical prior art approaches to particle separation i.e., membrane separation/filtration (Figure 1A) or column separation (Figure 1B). Membrane filtration/separation (Figure 1A) uses a membrane 101 to selectively allow some of the molecules in the mixture 105 to pass through 104. Membranes can become clogged unless pre-filtration is used and the separation of small molecules requires high pressures/energies (e.g., as in desalination processes). This can be used for batch or continuous flow separations.

[ 0005 ] Column separation methods as depicted in Figure 1B, where a small amount of a sample 105 is added to the top of a column 107. The sample 105 in the present example comprises two constituents 102 and 104. The mixture 105 flows along column 107 and the separation of constituents 102 and 104 occurs in the direction 106 of the flow such that constituents 102 and 104 exit the column 107 at different times. This technique is time consuming, can require expensive equipment/columns (e.g., high- performance liquid chromatography, gas chromatography) and is ultimately a batch separation technique.

[ 0006 ] Ratchet-type separation methods are known and the general operation of a ratchet type continuous flow separation device is shown in Figure 1C. In Figure 1C (i) there is flow through a tube but no separation of components of the mixture, but in Figure 1C (ii), an array of obstacles 111 combined with diffusion or flow lanes may be used to separate molecules in a mixture 105 at an angle to the flow as they are forced through the obstacle array 111. With the obstacle array 111 there is separation of the constituents of the mixture at an angle to the flow. This technique is continuous and typically requires less energy for the separation. However, current ratchet-type devices are limited by lithography techniques which are prohibitively expensive and limited as to the obstacle size that can be fabricated. This means that with the obstacle dimensions available in current devices, ratchet-type separation methods are only suitable for large molecules (e.g. DNA) and suspensions.

SUMMARY

[ 0007 ] Disclosed herein is an improved ratchet type separation mechanism that is cheap, fast, easily implemented 'in-line' to existing production processes requiring particle separation and may be easily scalable to suit separations of components where the sizes may range from metal ions up to large particles. In a particular arrangement disclosed in greater detail herein, nano-materials are used to achieve the new ratchet mechanism, removing the limitations of lithography (and similar methods) which are expensive and prevent current ratchet type devices separating small molecules/ions. The separation medium of the ratchet separation apparatus disclosed herein is expected to be significantly cheaper and far easier to make compared to lithography methods. The ratchet apparatus disclosed herein enables continuous flow separation of small molecules, ions or particles and has potential for applications such as speciation analysis or quality control monitoring where oxidation species are important (e.g., mining waste waters, electroplating industries, pharmaceutical industries, medical industries, and environmental analyses among many other applications).

[ 0008 ] 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.

[ 0009 ] It is an object of the invention in its preferred form to provide improved methods and apparatus for separation of molecules, ions and/or particles in particular nano-scale molecules, ions and/or particles.

[ 0010 ] According to a first aspect of the invention there is provided a ratchet-type separation apparatus. The apparatus may comprise at least one inlet port. The at least one input port may be adapted for receiving a mixture comprising at least two constituents of different sizes or having different diffusive properties. The apparatus may further comprise a porous or phase-separated medium. The porous or phase separated medium may be adapted for spatial separation of the at least two constituents. The apparatus may further comprise at least two outlet ports. Each of the outlet ports may be adapted to output one of said spatially separated constituents. In particular arrangements, the outlet ports may output at least a portion of one or more of the constituents in the mixture. Each of the constituents in the mixture may be output from the apparatus through one or more of the outlet ports. The spatially separated constituents may be output from the apparatus through a plurality of the outlet ports, wherein each constituent is output according to a distribution across the plurality of outlet ports. Each of the outlet ports may be adapted to output a portion of one or more of the mixture constituents. The portion of each of the mixture constituents which is output from a selected outlet port may be determined by a distribution. The distribution of the mixture constituents at the outlet ports may be determined by either or both of the size or the diffusive properties of the mixture constituents. [ 0011 ] The porous or phase-separated medium may comprises a cross-linked polymerised liquid crystal material or liquid crystal templated material.

[ 0012 ] According to a particular arrangement of the first aspect, there is provided a ratchet-type separation apparatus comprising: at least one inlet port adapted for receiving a mixture comprising at least two constituents of different sizes; a porous or phase-separated medium adapted for spatial separation of the at least two constituents; and at least two outlet ports, each port adapted to output one of the at least two spatially separated constituents; wherein, the porous or phase-separated medium may comprises a cross-linked polymerised liquid crystal material or liquid crystal templated material.

[ 0013 ] The cross-linked polymerised liquid crystal material or liquid crystal templated material may be a cross-linked polymerised hexagonal phase material.

[ 0014 ] In particular arrangements, the separation apparatus may be a continuous flow separation apparatus. In alternate arrangements, the separation apparatus may be a continuous flow batch-process separation apparatus. In alternate arrangements, the separation apparatus may be a continuous flow particle concentrator (i.e. if the sample is added to the entire top of the array then the particles will be concentrated at one side of the apparatus).

[ 0015 ] The porous or phase-separated separation medium may be a bulk lyotropic liquid crystal medium.

[ 0016 ] The apparatus may further comprise means of providing a driving force to the mixture to generate movement of the mixture through the separation medium. The apparatus may further comprise at least two electrodes adapted to provide a driving force in the form of an electrophoretic driving force to said mixture thereby to generate drift of the mixture through the separation medium. The driving force may alternatively be hydrostatic pressure or gravity depending on the particular appraatus configuration.

[ 0017 ] The porous or phase-separated medium may be formed from a polymerisable surfactant/lyotropic liquid crystal (LLC) monomer. [ 0018 ] According to a second aspect, there is provided a method of separating at least two constituents from a mixture. The at least two constituents may be either of different physical sizes or have different diffusive properties. The method may comprise the step of providing an inlet port adapted to accept said mixture comprising said at least two constituents. The method may comprise the further step of providing a porous or phase-separated separation medium located downstream from said inlet port. The method may comprise the further step of providing a driving force to generate drift of the mixture, the mixture being received at the inlet port, through the porous or phase-separated separation medium. The method may comprise the further step of providing at least two outlet ports located downstream of the porous separation medium. Each of the at least two outlet ports may be adapted to output at least a portion of the one or more of spatially separated constituents of the mixture. The method may comprise the further step of inputting the mixture to the inlet port. The method may comprise the further step of providing a downstream driving force to the mixture such that the mixture is propelled with the driving force through the porous or phase-separated separation medium. Each of the constituents of the mixture, being of different size and/or diffusive properties, may experience unique drift substantially perpendicular (or at some other angle between 0° and 90°) to the direction of flow of the mixture constituents through the separation medium such that on exiting the porous separation medium, the at least two constituents of different sizes are at least partially spatially separated. The method may further comprise the step of outputting each of the at least partially spatially separated mixture constituents substantially through one or more of said at least two outlet ports.

[ 0019 ] According to a particular arrangement of the second aspect, there is provided a method of separating at least two constituents from a mixture, said at least two constituents being either of different physical sizes or having different diffusive properties, the method comprising: providing an inlet port adapted to accept said mixture comprising said at least two constituents; providing a porous or phase- separated separation medium located downstream from said inlet port; providing a downstream driving force to the mixture such that the mixture is propelled with the driving force through the porous or phase-separated separation medium; providing at least two outlet ports located downstream of the porous separation medium, each of said at least two outlet ports being adapted to output at least a portion of the one or more of spatially separated constituent of the mixture; inputting the mixture to said inlet port; providing a downstream driving force to the mixture such that the mixture is propelled with the driving force through the porous or phase-separated separation medium; whereby said each constituent of different size and/or diffusive properties comprised in the mixture experiences unique drift perpendicular (or at some other angle between 0° and 90°) to the direction of flow through the porous separation medium such that on exiting the porous separation medium, the at least two constituents of different sizes are at least partially spatially separated; and outputting each of the at least partially spatially separated mixture constituents substantially through one or more of said at least two outlet ports.

[ 0020 ] The porous or phase-separated separation medium may comprise a polymerised hexagonal phase material.

[ 0021 ] The at least partially separated constituents may exit through a plurality of the outlet ports. Each constituent may output according to a distribution across the plurality of outlet ports.

[ 0022 ] The porous separation medium may comprise hexagonally ordered, cylindrical nanopores/nanorods. The porous separation medium may comprise a cross-linked polymerised hexagonal phase material. The porous separation medium may be a bulk lyotropic liquid crystal medium.

[ 0023 ] The driving force may comprise an electrophoretic force provided by at least two electrodes respectively located upstream and downstream with respect to the porous separation medium.

[ 0024 ] In a further aspect, the ratchet-type separation apparatus may be adapted for concentrating particles of one size, and accordingly would comprise one inlet and one outlet port designed to concentrate particles that enter the device less frequently (i.e. a low concentration suspension or simply a less frequent addition of particles to the entry - both are essentially just a lower entry frequency). BRIEF DESCRIPTION OF THE DRAWINGS

[ 0025 ] Preferred embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which:

[ 0026 ] Figures 1A, 1 B & 1 C are schematic depictions of prior art separation methods;

[ 0027 ] Figure 2 is a schematic of the operating methodology of a proposed ratchet- type separation apparatus based on a hexagonal phase as disclosed herein;

[ 0028 ] Figure 3A is a schematic depiction of a ratchet-type separation apparatus based on a hexagonal phase as dislcosed herein;

[ 0029 ] Figure 3B is a depiction of a porous or phase-separated separation medium of the ratchet-type separation apparatus based on a hexagonal phase of Figure 3A;

[ 0030 ] Figure 4 is an exploded schematic depiction of the ratchet-type separation apparatus of Figure 3A;

[ 0031 ] Figures 5A and 5B depict the operating principle of the ratchet-type separation apparatus of Figure 3A operating in separation mode;

[ 0032 ] Figure 6 is a depiction of the analytical solution from a simple approach of the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array, on the flow angle, SFIOW, for a distance travelled in the flow direction, arrangement, flow rate and for two particle sizes, for the ratchet-type separation apparatus as disclosed herein;

[ 0033 ] Figure 7 is a depiction of the approximate analytical solution from a simple approach of the dependence of the optimum flow angle for maximum shift, 6>FI ₀ W, on the ratio of the particle radius to the cylinder radius for the ratchet-type separation apparatus as disclosed herein;

[ 0034 ] Figure 8 is a depiction of the analytical solution from a simple approach of the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array on the flow rate, v, for a distance travelled in the flow direction, arrangement and particle size, for the ratchet-type separation apparatus as disclosed herein; [ 0035 ] Figure 9 is a depiction of the analytical solution from a simple approach of the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array on the hexagonal lattice parameter, aHe*, for a distance travelled in the flow direction, arrangement, flow rate and particle size, for the ratchet-type separation apparatus as disclosed herein;

[ 0036 ] Figure 10 is a depiction of the analytical solution from a simple approach of the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array on the cylinder radius, Rc _y i, for a distance travelled in the flow direction, arrangement, flow rate and particle size, for the ratchet-type separation apparatus as disclosed herein;

[ 0037 ] Figure 11 is a depiction of numerical simulation results from random walk simulations (data points) compared to the analytical solution from a simple approach (solid lines) for a flow distance of oV = 200 nm, for an arrangement, flow rate and particle size, for the ratchet-type separation apparatus as disclosed herein;

[ 0038 ] Figure 12 is a depiction of numerical simulation results from random walk simulations (data points) compared to the analytical solution from a simple approach (solid lines) for a flow distance of dv = 1 μΐτι, for a flow rate, particle size and arrangements, for the ratchet-type separation apparatus as disclosed herein;

[ 0039 ] Figure 13 is a prediction from the analytical solutions from a simple approach of the separation capability of Pb ²⁺ , Fe ²⁺ , Fe ³⁺ and Al ³⁺ ions for the ratchet-type apparatus disclosed herein;

[ 0040 ] Figure 14 show photographs of a 3D-printed scaled-up version of a ratchet- type separation apparatus depicting the operating principle as disclosed herein;

[ 0041 ] Figure 15 shows photographs of a 3D-printed scaled-up version of a ratchet- type separation apparatus as in Figure 14 but disassembled to show the components;

[ 0042 ] Figure 16 shows photographs of the technical aspects relating to a 3D-printed scaled-up version of a ratchet-type separation apparatus as in Figure 14 and Figure 15 and corresponding schematics highlighting parameters describing the array of cylinders (i.e., the cylinder radius, Rc _y i, the hexagonal lattice parameter, aHex, and the array orientation, and the collection chamber (i.e., bin separation and number); [ 0043 ] Figure 17 shows photographs of the size and shape of the two beads used to test the separation using the 3D-printed scaled-up version of a ratchet-type separation apparatus as in Figures 14, 15 and 16;

[ 0044 ] Figure 18 shows photographs of the entry chute diameters of the model of Figure 14 with and without an entry chute size modifier in place in the 3D-printed scaled-up version of a ratchet-type separation apparatus as in Figure 14;

[ 0045 ] Figure 19 shows a histogram of the results and fitted Gaussian curves for the separation of beads in Figure 17 when passed through a 3D-printed scaled-up version of a ratchet-type separation apparatus as in Figure 14 without the entry chute modifier in Figure 18;

[ 0046 ] Figure 20 shows a prediction of the results in Figure 19 with modified simple models (i.e., for bin number and position with respect to the entry point) showing the need for further modification of the simple models for the scaled-up version of the ratchet-type separation apparatus as disclosed herein;

[ 0047 ] Figure 21 shows a histogram of the results and fitted Gaussian curves for the separation of beads in Figure 17 when passed through a 3D-printed scaled-up version of a ratchet-type separation apparatus as in Figure 14 with and without the entry chute modifier in Figure 18 in place for the smaller of the two beads;

[ 0048 ] Figure 22 shows a prediction of the results in Figure 21 with modified simple models (i.e., for bin number and position with respect to the entry point) and correction to the injection width for particle size showing the need for further modification of the simple models or calibration using empirical parameters for the scaled-up version of the ratchet-type separation apparatus as disclosed herein; and

[ 0049 ] Figure 23 shows a histogram of the results and fitted Gaussian and reflected Gaussian curves for the 'reverse' entry chute showing no shift and no separation of beads in Figure 17 when passed through a 3D-printed scaled-up version of a ratchet- type separation apparatus as in Figure 14 with and without the entry chute modifier in Figure 18 in place for the smaller of the two beads. DEFINITIONS

[ 0050 ] The following definitions are provided as general definitions and should in no way limit the scope of the present invention to those terms alone, but are put forth for a better understanding of the following description.

[ 0051 ] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms used herein should be interpreted as having a meaning that is consistent with their meaning in the context of this specification and any relevant art will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. For the purposes of the present invention, additional terms are defined below.

[ 0052 ] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the singular articles "a" , "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise and thus are used herein to refer to one or to more than one (i.e. to at least one) of the grammatical object of the article. By way of example, "an element" refers to one element or more than one element.

[ 0053 ] The term "about" is used herein to refer to quantities that vary by as much as 30%, preferably by as much as 20%, and more preferably by as much as 10% to a reference quantity. The use of the word 'about' to qualify a number is merely an express indication that the number is not to be construed as a precise value.

[ 0054 ] Throughout this specification, unless the context requires otherwise, the words "comprise" , "comprises" and "comprising" will be understood to imply the inclusion of a stated step or element or group of steps or elements but not the exclusion of any other step or element or group of steps or elements.

[ 0055 ] The term " real-time" for example "displaying real-time data," refers to the display of the data without intentional delay, given the processing limitations of the system and the time required to accurately measure the data. DETAILED DESCRfPTiO

[ 0056 ] Briefly, the limitations to 'top-down' approaches, e.g., photolithography and soft lithography/molding, of fabricating ratchet type separation devices are being approached. Such approaches are expensive and difficult, requiring specialised equipment to make devices with dimensions smaller than sub-micrometer to micrometer dimensions. This ultimately limits their applicability to the separation of large macromolecules and suspensions. To develop the ratchet-type devices further has required the development of a new mechanism and a new 'bottom-up' approach resulting in the methods and apparatus' disclosed herein (for the specific case of separations of small ions/molecules, the separation apparatus has been termed the 'NanoRatchet').

[ 0057 ] A new ratchet type separation mechanism has been developed as presently disclosed herein with the use of nano-materials (in a particular example arrangement, lyotropic liquid crystal polymers/templated materials using lyotropic liquid crystals are described or other suitable 'bottom-up' approach) to employ this new ratchet separation mechanism for the separation of small molecules/ions which has significant implications for the industrial separation procedures & applications. Current types of ratchets (e.g., geometric (2D) Brownian ratchets and Deterministic (microfluidic) ratchets) remain an active area of research. Similarly the area of lyotropic liquid crystals membranes is also a very active research area. The new ratchet mechanism disclosed herein seems to fall in between geometric (2D) Brownian ratchets and Deterministic (microfluidic) ratchets and allows the proposed materials to be used for this purpose.

[ 0058 ] With regard to the NanoRatchet, the anisotropy imparted on the molecules/ions' mobility in the macroscopically aligned systems is exploited to generate the ratcheting effect. In present arrangements disclosed herein, the aligned hexagonal phase which ideally consists of a hexagonal lattice of circular cylinders is employed. In practice, such systems typically comprise features that are roughly circular, however hexagonal phases are commonly described as comprising a hexagonal lattice of circular cylinders. During separation, the flow of the mixture containing disparate constituents (i.e. molecules/particles/etc) to be separated is set at an angle to the cylinder alignment of the apparatus which results in the molecule/ions being shifted by the cylinder tilt and hence separated from the mixture.

[ 0059 ] The separation technique disclosed herein can be used to separate: very small ions such as, for example, hydrated metal ions; and different oxidation states of the same ion (i.e. speciation). The main advantage of this method over others is the simplicity and continuous flow nature while other small ion separation techniques are batch processes. This has applications as a continuous flow separation/monitoring device suitable for use in industries such as mining, electroplating, environmental monitoring (e.g., waste water), pharmaceutical etc. However, the separation mechanism employed in the currently disclosed ratchet apparatus is readily scalable to be applicable for separation of larger (e.g. sub-micron and micron scales and beyond) extending the applicability of the technique to a wide range of industries including, for example, among others: laboratory equipment/chemical processing and analysis industry; pharmaceutical/medical industry; mining industry; electroplating industries; agriculture industries, and potentially even as an alternative to current methods for desalination. The feature size and spacing of the cylinders can typically be manipulated during fabrication of the separation medium by use of different lyotropic liquid crystal constituent molecules or mixtures of different constituent molecules. For example, the dimensions of the features within the separation medium are dependent on the surfactant type and hydrophilic/solvophilic and hydrophobic/solvophobic lengths of the constituent molecules. There might be a family of surfactants for example that vary from one to the next by increasing the length of the hydrophobic/solvophobic and/or hydrophilic/solvophilic region which would increase the size of the cylinders and/or change the cylinder spacing. Alternatively, different surfactants may require different amounts of solvent to make the phase structure which would change the dimensions in the phase.

[ 0060 ] The size of the bulk separation material may be scaled depending on the sample alignment technique used for fabrication of the material. For example, if the constituent molecules and phase are readily alignable with magnetic fields then it could be that the array size is easily made to be large (i.e., the cylinder radius and spacings are still the same but the size of the final bulk material is large and contains more cylinders).

[ 0061 ] Furthermore, the size of the individual features (e.g. cylinders) in the separation medium may also be scaled depending upon the available surfactants, polymers etc that have suitable characteristics for formation of the requisite liquid crystal phases/phase separated regions in the material and it can be appreciated that other phases with anisotropy but not necessarily consisting of cylinders may also be utilised (e.g. lamellar phases).

Limitations of current top-down fabrications used for current ratchet devices

[ 0062 ] The limitations of lithographic techniques (e.g., using a top-down approach) means that the separation of smaller molecules/ions is currently not possible with ratchet type continuous flow separation devices (e.g., geometric (2D) Brownian ratchets and Deterministic (microfluidic) ratchets) simply because the separation characteristics for a given particle/molecule size depends on the obstacle dimensions and precise spatial arrangement. The feature size and depth expected from current micro- fabrication techniques is typically ~>50 - 100 nm (smaller feature sizes may be achieved for example, with soft lithography techniques or more specialised techniques) but the results suffer from deformation problems and the fabrication techniques are typically either slow, or are surface-only methods or are capable of only producing shallow obstacle arrays. Further development to extend such techniques for fabrication of ratchet-type devices applicable for smaller constituents will be difficult and/or expensive to implement. This limits the currently available ratchet-type separation devices to the separation of large particles/molecules such as, for example, DNA, polymer spheres, suspensions etc, where the constituents to be separated are typically in the sub-micron to micron size range. An additional problem for these type of analysis/separation devices is the depth of the obstacles because, for shallow arrays, there are sample concentration limitations (i.e. as they are typically used for the separation of large molecules/particles such as DNA, polymer spheres etc and the characteristic dimensions, like separation of the obstacles, is of a comparable size, then unless the array is deep there is the problem of clogging/blocking of the gaps in the array). The use of a polymerised aligned lyotropic liquid crystal phases or templated polymers/gels/silicon arrays embedded in gels using aligned lyotropic liquid crystal phases as disclosed herein provides the advantages of a bottom-up approach and does not face the same limitations as current lithographic methods (for example, materials with small obstacles - on the order of a few nm's - but in a macroscopic or bulk material are possible with the methods and apparatus disclosed herein) and also allows for the possibility of separation and analysis of smaller molecules/ions continuously.

Improved Ratchet-type continuous flow separation mechanism

[ 0063 ] Geometric Brownian ratchets require a mechanism comprising ideal obstacle shapes, sizes or orientations. Also, Deterministic (microfluidic) ratchet separation devices in bulk material require specific lattice orientations.

[ 0064 ] While it is relatively easy experimentally to obtain uniformly aligned hexagonal phases for templating/polymerising to obtain nano-structured membranes/materials, it is more difficult to experimentally obtain aligned compressed hexagonal phases (i.e. ribbon phases where the cylindrical obstacles are non-circular but aligned with the flat sides of the 'ribbon' in one orientation) with either ideal obstacle shapes, sizes or orientations (which is required if they are to be used as new geometric Brownian ratchets) or specific single lattice orientations - i.e. if they are to be used as a new Deterministic (microfluidic) ratchet-type separation device). Note they are referred to herein as 'new' because they would contain smaller characteristic dimensions (i.e., obstacles and spacing are on the nm range) than presently available ratchet-type devices which allows for the continuous separation and/or concentration of smaller molecules or particles and also would be large in all three dimensions and not just two which is the case for the current geometric (2D) Brownian ratchets and Deterministic (microfluidic) ratchets for which the obstacle array height is typically only on the μπι range. In respect of membrane filtration systems, these are very thin layers on porous membrane supports. Previous LLC membranes have been based on the hexagonal phase (inverse - where the cylinders are pores and so the separation occurs because the pores limit which molecules in the mixture can go through) and also on a bicontinuous cubic phase which is isotropic and so the pores are channels in a specific path for the bicontinuous cubic structure. The main difference here is that the membranes use the size exclusion based on the pore size to effect separation. The NanoRatchet apparatus as disclosed herein differs in two aspects here: it is typically going to be bulk materials not thin membranes and it is the anisotropy of the hexagonal (or other) phase which is utilised to effect separation.

[ 0065 ] With respect to other known applications of LLC materials: Bulk materials with macroscopic alignment have been used for low dimensional conductors where the anisotropy means that conduction is preferable along one axis in the anisotropic material. The NanoRatchet apparatus as disclosed herein differs from this in that the interest in the conductors is not separation and the anisotropy is at an angle to the flow not parallel to it (though it could be but the figures show the expected loss of separation at the extremes).

[ 0066 ] Drug delivery with LLC phases typically the interest is in the phases for slow drug release.

[ 0067 ] There are a plethora of applications of these materials. The NanoRatchet apparatus as disclosed herein exploits the anisotropy to effect separation using bulk macroscopically aligned LLC phase materials and the separation mechanism using the anisotropy is something different.

[ 0068 ] It will be appreciated that any suitable poiymerisable molecule can be used in the invention. However, an especially preferred poiymerisable molecule may comprise a lyotropic liquid crystal (LLC) monomer. Lyotropic liquid crystal (LLC) monomers are amphiphilic molecules containing one or more hydrophobic/solvophobic regions/tails and one or more hydrophilic/solvophilic regions/headgroups. The amphiphilic character of these molecules encourages them to self-organize into aggregate structures, with the boundary of the hydrophobic/solvophobic regions and the hydrophilic/solvophilic regions defining the interface of phase-separated domains. These aggregates may be relatively simple individual structures such as micelles and vesicles or highly ordered yet fluid condensed assemblies with specific nanometer-scale geometries known collectively as LLC phases.

[ 0069 ] LLC phases are well-suited for the production of nanostructured organic materials. Their architectures may incorporate hydrophobic/solvophobic and hydrophilic/solvophilic (or charged) compounds in separate domains with well-defined nano-scale geometries, and may be especially attractive for the production of nanostructured materials, with only the caveat that LLC phases are inherently fluid and therefore lack the robustness required for most materials applications. Thus, the present invention may use polymerisable LLC surfactants to form nanoporous polymers.

[ 0070 ] Polymerisable surfactants may comprise molecules having a pair of hydrophobic/solvophobic and hydrophilic/solvophilic components together with one or more polymerisable groups in their structure. These polymerisable surfactants may be used to form surfactant phases to produce useful materials with highly regular nano- scale architectural features (i.e. pores, cylinders etc.).

[ 0071 ] It will further be appreciated that many different kinds and types of LLCs can be used, and the choice of surfactant molecule will enable control of the array characteristics (obstacles, spacings, pore size, etc). It will further be appreciated that the surfactant molecules can be single molecular weight molecules, distributions of molecular weights, or mixtures comprising members of the general families of surfactants (anionic, cationic, no-ionic, and zwitterionic). An example could be non- ionic surfactants like polyethylene glycol monoalkyl ethers but including a polymerisable group such as a diene moiety.

[ 0072 ] If these nano-structured materials are to be used as ratchet-type continuous flow separation devices a new type of ratchet mechanism has to be developed. Known geometric Brownian ratchets use an array of obstacles with asymmetry with respect to the flow direction, and the molecules are separated through different shifting probabilities. This is related to the diffusion coefficient of the particles, where those with a higher diffusion coefficient (small particles) are shifted more than those with lower diffusion coefficients (large particles). Similarly, known Deterministic ratchets use an array of obstacles of specific arrangement and the molecules are separated into different flow lanes by being 'bumped' from one flow lane to the next by the obstacle. This is related to the size of the particles, where the large particles are 'bumped' more than the smaller particles and so the larger particles shift greatest. [ 0073 ] Both of the above known Ratchet mechanisms are dependent on careful array orientation or obstacle shapes, however the NanoRatchet as disclosed herein makes use of the overall alignment direction and is not as dependent on the obstacle shape and lattice orientation as known ratchet separation mechanisms other than through the anisotropy of the array. The differences are that the two types of known ratchet mechanisms mentioned above are concerned with a 2D arrangement, whereas the NanoRatchet disclosed herein is concerned with a 3D arrangement; the NanoRatchet disclosed herein has obstacles arranged so there is some asymmetry with respect to at least one axis or the flow direction but could be symmetric from with respect to other axes; simulations of the separation medium of the Nanoratchet disclosed herein suggest that particles of different sizes are separated with the larger particles shifting the most because of the greater anisotropy experienced by those larger particles.

t 0074 ] A new continuous flow separation/concentration device is still possible with aligned hexagonal phases/aligned cylinders (i.e. along the z-axis) with random lattice orientations (i.e. in the x-y plane) throughout the sample (other phase structures might also be suitable for similar separations and should not be ruled out but the development of analytical theories for modelling the suitability of such alternate structures will depend on the particular phase structure). One further example phase structure which would be suitable for separation purposes would be the lamellar phase. The lamellar could be tilted with respect to the flow direction so that the highest diffusion direction is at an angle to the flow. For this, the transport in the direction of the flow would be enabled through defects present in the phase structure (i.e., porous lamellar sheets) or via the presence of micro-domain boundaries. Along similar lines, a mesh phase (perforated lamellar phase) liquid crystal material could also be useful (but this is an intermediate phase so it may be hard to find a surfactant that displays a large region of mesh phase). Another alternative is an inverted hexagonal phase whereby the mixture flows down the inside of the cylinders instead of around them and where the defects (such as bridges between the cylinders) and micro-domain/crystallite sizes could allow transport along the flow direction.

[ 0075 ] A schematic of the operating methodology of a proposed ratchet-type separation apparatus (an example arrangement of which is as seen in Figure 3A) referred to herein as the NanoRatchet, is shown in Figure 2. The particular arrangement of the NanoRatchet device disclosed herein takes advantage of the anisotropic properties of aligned hexagonal phases in lyotropic liquid crystals, however it will be appreciated that suitable polymerisable surfactant could be used as discussed above. Referring to Figure 2, the hexagonal lattice arrangement of the present porous (or phase-separated) material used for the NanoRatchet device comprises a lattice of cylinders 201 aligned along the z axis 206 with flow v at an angle to the cylinder axes (i.e. 203 is defined as the angle the flow makes with the y axis 204 in the -z plane). The hexagonal lattice is characterised by the hexagonal lattice parameter, aHox 205. The cylinders are circular with radius Rc _y i 207. In general, the spacing between the cylinders limits the particle radius (actually in some of the example simulations the particle radius, Rpsmae 209, was approximately 1/3 of the cylinder radius, ¾ _y i 207, but this depended on the cylinder spacing). The cylinders could be small and comparable to the particle radii and so the anisotropy might not be as great but still large enough over a large array size to still provide sufficient separation ability for the device to be useful. So, in general the particle radius, Rps Me 209, is limited to less than a maximum of (aHex -

[ 0076 ] The distance travelled by the particles in the flow direction (having a flow rate v 211) from the start point zstan) 213 is denoted . Point ) 214 is the reference position if no shift at an angle FIOW to the flow occurs during the flow time and 216 is the shifted position. The shift relative to the flow line, ds m 215, is positive if z > z and > yena. The x-y plane is perpendicular to the cylinder axes. The lines, L 217 and L> 218 are perpendicular to the flow direction, and cross the start and end points, respectively; and the line L OW 204 denotes the direction along which flow occurs. The array orientation (ΘΑ™ _υ 221) is the angle of rotation of the hexagonal lattice 202 in the x plane and is between 0° and 30° for the simulations because of symmetry. The hexagonal lattice has symmetry that means the only angles that are important for purposes of simulation of the flow characteristics of the NanoRatchet device are those between 0 and 30 degrees. In practise, the actual rotation angle of the hexagonal lattice 202 (i.e. 221) is not expected to be important since (given the lattice symmetry) the flow/diffusion characteristics are not dependent on the array orientation for nanoscale systems - given that real systems are likely to have all micro- domains present (i.e., all rotation angles, signifying a complete 360 degree revolution of the hexagonal lattice, but because of the lattice symmetry, all angles can be simulated by considering lattice rotations in the range of between 0° - 30° and mirroring). As discussed below, the analytical expressions based upon a simple approach support this finding as the hexagonal lattice rotation angle (i.e. 221 ) does not appear in the final equations. In alternative arrangements, however, if the device scaled up for separation of larger particles, the larger dimensions may mean that the angle of the hexagonal array is selectable/tunable (i.e. easier to make a separation medium having one specific orientation).

[ 0077 ] The aligned cyl inders 201 are tilted with respect to the flow along line LFIOW 204, where the cylinder axis is along z 206, and since there are gaps between the cylinders, this imparts anisotropy onto the molecules/particles as they are forced through the array (via electrophoretic mobility, hydraulic or hydrostatic pressure or other suitable driving force). Note too that there is no requirement for the dimensions of the system to be on the nano-scale as such a separation/concentration should also occur for larger particles and larger device dimensions (for larger dimensions, it may be preferable to use the lattice orientation of 0° to ensure there is displacement even when there is little or no diffusion). Also note that the lattice of this kind of separation device might not necessarily be required to be hexagonal and other lattice types might also be used, for example square, rectangular lattice types may be developed with suitable surfactants or for scaled up devices. However, the possibility of using aligned hexagonal phase materials shows promise for continuous flow separation/concentration of small molecules because of their nm characteristic dimensions (i.e. cyl inder radius and hexagonal lattice parameter) and essentially unlimited bulk dimensions (i.e. the alignment can readily be made to occur over macroscopic dimension materials and exhibits anisotropy despite any microscopic crystal sizes/defects that make up the bul k material).

[ 0078 ] An analytical analysis, from a simple approach, of the characteristics of the device is discussed below to describe some considerations and example performance characteristics for the NanoRatchet made using polymerised lyotropic liquid crystal hexagonal phases (or polymers/silicon arrays in gels/gels templated using them) as disclosed herein. For such a system, electrophoresis might be the most suitable mechanism to achieve flow of the particles/ions/molecules. This is because the linear flow rate through cross-linked lyotropic liquid crystal materials is likely to be low with hydrodynamic flow/hydraulic pressure which can be seen when considering the typical values of permeability/flux for membranes based on these materials (i.e., membranes for vapour barriers, nanofiltration and reverse osmosis) but also for typical reverse osmosis/nanofiltration/desalination membranes. However, the simulations are random walk simulations with a fixed increment along the flow direction. These are simple simulations and so the flow in the simulation could be from any suitable driving force.

[ 0079 ] Referring to Figure 2, there is depicted a stylised conceptual model of the separation method according to the present disclosure where the porous or phase- separated medium is made using polymerised lyotropic liquid crystal hexagonal phases (or polymers/silicon arrays in gels/gels templated using them). A hexagonal lattice arrangement of cylinders 201 aligned along the z axis 206 with-flow v at an angle 203 to the cylinder axes (i.e., angle 203 is defined as the angle the flow makes with the y- axis in the yz plane). The hexagonal lattice 202 is characterised by the hexagonal lattice parameter, 205. The cylinders 201 are circular with radius Rc≠ 207. The particle radius, RpartMe, 209, is simply added to the cylinder radius 207 for the simulation calculations discussed below. The distance travelled in the flow direction 211 from the start point 213 zstan) is dv. Reference co-ordinate zi¾f) is the reference position if no shift at an angle to the flow occurs during the flow time and is the shifted position. The shift relative to the flow line 204 is cfshm 215 and is positive if and The xy plane is perpendicular to the cylinder axes which is aligned along the z-axis 206. The lines, L\ 217 and Li 218, are perpendicular to the flow direction 211 and cross the start and end points, respectively; and the line 204 is the line along which flow occurs. The array orientation 221 is the angle of rotation of the lattice 223 in the x-y plane with respect to the 0° array 202. In the numerical simulations discussed below for this lattice structure, 221 is maintained between 0° and 30° for the numerical simulations because of symmetry. In practice 221 may range between 0° and 360° however, due to symmetry of the hexagonal lattice, all algles can be simulated simply by considering angle between 0° and 30°.

[ 0080 ] An example NanoRatchet device 300 util ising the separation method described in Figure 2 is depicted in Figures 3A and 3B. The device arrangement 300 shown in Figure 3A is particularly suited to be used as a continuous-flow-device for continuous quality control/monitoring, speciation analysis, pre-concentration, etc, or other molecular or particulate separation application. In practise, the NanoRatchet device may vary slightly as would be appreciated by the skilled addressee e.g. for the pre-concentration mode, the device would have a modified input channel so that the sample covers the entire top of the array.

[ 0081 ] Figure 3B depicts the main component of the separation apparatus 300, the separation region 350 formed by a porous or phase-separated medium. The limitations on the dimensions of the separation medium which may be used in the NanoRatchet device will typically be dependent on factors such as, for example, the specific particles the device is designed to separate and the flow rate of the particles through the separation medium are required to be known in order to select the width and height of the array (i.e., to be able to achieve sufficient separation in accordance with requirements). Also, the depth of the array is an important consideration: i.e. the larger the array depth, the more sample that can be added to the top of the device. This is dependent on the ability to make the bulk material to that size (typically dependent upon practical fabrication methods) but is also likely to be limited (partly) on the heat dissipation characteristics (i.e., Joule heating) of the separation medium 350 as larger cross-sectional area (i.e., depth x width) will increase resistive (Joule) heating through a reduction of flow resistance. So unless adequate cooling is used to protect the separation medium (if needed), this may limit the practical depth of the separation medium in the NanoRatchet device 300.

[ 0082 ] The porous/phase-separated separation medium is shown in the present example (as seen in Figure 3) as an aligned polymerised lyotropic bulk liquid crystal material 350 with hexagonal phases or, alternately, templated polymers/gels/silicon arrays of cyl inders 351 embedded in gels using them (where the slanted cylinders 351 - i.e. cylinders 201 of Figure 2 - are the ratchet obstacles). The alignment of the slanted cylinders 351 is oriented so that cylinders 351 are at an angle to the flow of mixture through the separation medium 350 (the optimal angle for separation is easily found using the theories developed below complemented with random walk simulation results). A key feature of separation medium 350 is the size of the obstacles i.e. cylinders 351 , their spacing and final array dimensions compared to what is currently available with micro-fabrication/lithography techniques. The obstacles and the spacings of separation medium 350 are in the low nm range and the width, depth and height of the final array is easily made to any size within practical fabrication constraints, for example, the array dimensions (length, width and depth) can be made quite large with a magnetic field alignment procedure for aligning the phase structure of the medium during fabrication - but note that while the cyl inder alignment, for example, over the entire array is along one axis, the bulk material still has numerous micro-domains (crystallites) that have individual hexagonal lattice orientations and defects are likely present across the final material too (but all of these are only likely to average or have l imited effect on the results or aid the results for other phase structures). Other alignment methods could be useful as well as would be appreciated by the ski lled addressee.

[ 0083 ] Figure 5A depicts one possible implementation 300 of the NanoRatchet using a polymerised lyotropic liquid crystal material separation medium 350 utilising electrophoresis as a driving force to push an input mixture 360 admitted through an inlet port (i.e. inlet ports 301a or 301 b) of device 300 and through the separation medium 350 to one of a plurality of outlet ports 305 in a 'lab-on-a-chip' format. However, since the NanoRatchet device 300 disclosed herein is readily scalable for separation of molecular or particulate mixtures comprising constituents of varied size then it is foreseeable that flow in an array on a larger scale (i.e., not using lyotropic liquid crystal materials) with greater permeability could be generated via hydraulic or hydrostatic pressure or gravity as an alternate driving mechanism for the device 300.

[ 0084 ] Figure 4 depicts an exploded view of the concept model of the NanoRatchet separation device 300 of Figure 3A and using a polymerised lyotropic liquid crystal material with hexagonal phases (or polymers/silicon arrays in gels/gels templated using them) as the separation medium 350. For this system, the most likely method for flow generation is electrophoretic based on the permeability of membranes made from lyotropic liquid crystal materials but for scaled up forms of this type of ratchet and further developed materials the flow could be generated via a hydraul ic or hydrostatic pressure or gravity as well as would be appreciated by the skilled addressee. Highlighted in Figure 4 are some components/features of this particular NanoRatchet arrangement 300 including multiple inlet channels 301a and 301 b; and a plurality of possible outlet ports 305; separation medium / obstacle array 350; electrodes 307a and 307b to provide the electrophoretic drivi ng force; optional porous supports 311a and 311 , optional inlet 309a and outlet 309b ports for an optional buffer chamber located underneath the buffer ports ; and a microfabricated channel system portion 315 with barriers 313a and 313b between the slots for the electrodes and the rest of the channel system. The purpose of the sample input channel having an inlet 301a and an outlet 301 b is so that the device can be used in 'continuous flow' monitoring applications where only a small amount of sample is required so the device 300 can be placed 'inline' with a process. The purpose of the buffer chamber is to allow the electric field (in this particular arrangement where electrophoresis is used as the driving force) to be across the entire separation channel where the sample inlet is only on the top corner of the obstacle array for separation mode. The buffer chamber has an inlet 309a and outlet 309b so that the buffer can be refreshed easily if required while retaining the conti nuous operation. The multiple outlet channels 305 are a requirement due to the continuous flow separation operation because they allow for the different sample fractions, separated laterally, to be collected or taken to the next process or detector individually and free from the other fractions. The purpose of the multiple outlet channels 305 having an inlet 305a and outlet 305b is so the sample can be removed from the outlet chambers with buffer flushes and so that the device can be used in 'pre-concentration' type applications (or if the sample needs to be concentrated before analysis) where the sample is not removed from the outlet channel until a sufficient amount is collected. There are also porous supports 311a and 311 b at the top and the bottom of the separation medium 350, and barriers 313a and 313b between the electrodes 307a and 307b and the buffer/sample solutions to prevent electrolysis. In particular arrangements of the NanoRatchet, the microfabricated channel system 315 could have small or large inputs with numerous or few outputs depending on requirements, and that the chamber for the separation medium/obstacle matrix 350 could have any dimensions depending on the requirements of particular applications. Also, the separation medium 350 could be made of several blocks of arrays of different characteristic sizes stacked i n the flow direction to modify the separation characteristics as the molecules/ions/particles are forced through the arrays (this is similar to particular applications of geometric (2D) Brownian ratchets and Determi nistic (microfluidic) ratchets), however, the smallest characteristic gap size cannot be smaller than the largest molecule/ion/particle.

[ 0085 ] Figures 5A and 5B depict the operating principle of the NanoRatchet device 300 (of Figures 3 and 4) operating in separation mode (although the separation mode operates similarly to other operating modes including pre-concentration or continuous monitoring etc). A mixture 360 comprising two molecular or particulate constituents 361 and 363 of different size, enters the device 300 at inlet port 301a where it is forced through the separation medium 350 such that the constituents 361 and 363 are separated based on anisotropic diffusion and array dimensions of the separation medium 350, and two separate (or at least partially separate) particle distributions exit the array through exit ports 365 and 366 for collection, analysis or transport for use in a downstream process.

[ 0086 ] Figure 5B depicts a schematic of the nature of the continuous flow separation in the NanoRatchet device 300. A narrow particle/molecule distribution 371 of the input mixture 360 is forced through the array of the separation medium 350 where the obstruction and anisotropy of the medium 350 result in the particles/molecules of different sizes shifting off the flow/drift line to different extents resulting in separation at the exit of the array between the differently sized molecular components 361 and 363 of the input mixture 360.

[ 0087 ] As mentioned above, the NanoRatchet device 300 can operate in different modes, much the same as other types of ratchets, for example, geometric (2D) Brownian ratchets and Deterministic (microfluidic) ratchets. Two common modes are 'separation' and 'concentration'. [ 0088 ] In 'separation mode' there is continuous flow separation of molecular or particulate constituents of a mixture whereby the mixture is injected into the array at one of the top corners, and the molecules/particles/ions are separated as they flow through the array of the separation medium and exit at different locations on the bottom edge or side of the array.

[ 0089 ] In 'concentration mode' the sample is added to the entire top edge of the array and the molecules/particles/ions will be concentrated at one side edge or at a bottom corner of the separation medium depending on the direction of the shift.

[ 0090 ] The 'separation mode' requires a small injection width, as with most separation technologies, to ensure the best resolution. Small structures/channels with dimensions of the μτη range (or even sub-μιη sized dimensions), can be made using currently known methods. As such, an input channel for a NanoRatchet separation device such as device 300 discussed above made using polymerised lyotropic liquid crystal hexagonal phases (or polymers/silicon arrays in gels/gels templated using them) can easily be made to have a small width (e.g., it is not unreasonable to assume an initial width of < 100 μιη, even < 50 μιη) minimising the effect of the initial distribution width.

Analytical Analysis of the NanoRatchet Device

[ 0091 ] Below are some equations derived from simplified theory for the ratchet mechanism and proposed NanoRatchet device as disclosed herein. Whilst the NanoRatchet device may be configured to operate slightly differently given the simple approach taken for this analytical analysis, the simulations discussed below describe the expected form of the device characteristics in comparison to random walk simulations.

[ 0092 ] The width, w, of the particle distribution at the bottom edge/exit of the array can be described by: w(D d _v , v, a ₀ , e _FLow ,

¾lnj - ^Particle)' ⁺ 8 [D„ ( _HeX (si Π (θ _Ρ , _ονν )) ² + (cOs(0 _F)OW )) ² )] ^

where D\\ is the diffusion (in units of m ² s ^-1 ) parallel to the cylinder axes (and already including the reduction due to the matrix); dv is the distance (in units of m) travelled i n the flow direction if there is no shift and depends on the flow rate, v (m s ') and the time taken, t (in units of s); σο is the standard deviation due to the injection port width corrected for the particle radius (in units of m), σο,ΐηί is the standard deviation from the injection port (may be taken as 1/2 the injection port width in the direction of the separation), Θπ is the angle of the flow in the y-z plane with respect to the y-axis (rad), and frtex which is the obstruction factor describing the obstruction to diffusion by the cylinders and depends on the hexagonal lattice parameter, an (in units of m); the particle radius, (in units of m), and the cylinder radius, Rc≠ (in units of m). Note that the width could be reduced to lots of SHH by dividing the result by 3Η«<. The above equation can also be written in terms of the diffusion coefficient of the solute at infinite dilution, Do (in units of m ² s ^-1 ), provided a model for the obstruction factor from the matrix surrounding the cylinders, f atrix, e. ., the polymer chains (for a separating medium made from polymerised lyotropic liquid crystal phases or for cylinders templated into polymer gels) which may also include a correction that represents additional averaging due to binding events, is known. This equation could be improved and scaled up versions of the apparatus and method may requi re modifications to this equation to account for other effects (such as for example bouncing, no diffusion etc) which are observed at larger scales.

[ 0093 ] The centre of the output distribution after a given flow time (i.e., the shifted amount.where cfehift, is the perpendicular distance to LBOW from (yEnd, ZEnd)) can be found from:

, d., ) = d tan n

2π{ ^ _≠ A', , _{; i :,} ){ / . , + ^, _ar „ _de cos(g _FlmT ))

: vt tan 2 π

and like the width, this could be reduced to being in the units of how many lots of am _* the centre of the distribution has shifted. This equation could be improved and scaled up versions of the apparatus and method may require modifications to this equation as noted above.

[ 0094 ] The predicted output for the simulations and analytical models can be visualised using a normalised Gaussian function (i.e., normalised to the number of particles) and could be normalised to an∞.

[ 0095 ] Further to the above equations, if the driving force is, for example, an electrophoretic force, then similar procedures can be used to predict the electrophoretic mobility parallel to the intended flow direction, (m ² V ^-1 ), assuming that similar relationships exist between electrophoretic mobility and obstruction as do with diffusion and obstruction.

[ 0096 ] The peak position of the exit location (i.e., for a distribution is not at the same flow angle for all particles. The optimum flow angle for a given particle, for the maximum x _c can be found by differentiating the shift equations and setting it equal to zero. Given the simple theory developed it is also conceivable that there is a (i.e., an angle of the hexagonal array between 0 and 30° (due to symmetry of the hexagonal lattice) for maximum shift or separation). That is, while the separation and shift using the simple theory seems independent of the array orientation, it is conceivable that there may be an optimum array orientation in some cases.

[ 0097 ] The following Figures 6 to 13 were made using the simplified analytical equations discussed above and represent the separation in the example NanoRatchet device 300 to predict the device operational outcomes. The following figures show some examples of the characteristic trends expected for the NanoRatchet device 300 (i.e., they are only given for a limited number of conditions and are intended to show the expected form of the performance characteristics).

[ 0098 ] Figure 6 depicts the dependence of the centre of the distribution, c, and its width, w, at the exit of the array on the flow angle, GROW, with the distance travelled i n the flow direction of ώ = 0.1 m, a flow rate of v = 200 pm s ^" \ the hexagonal lattice parameter of a** = 5.4 nm and a cylinder radius of Rc _y i = 1.4 nm for two particles. Line 601 shows is a particle with diffusion coefficient parallel to the cylinders of D = 6.8 * 10 ^{" 11} m ² s ¹ and radius of Rpartwe = 0.32 nm. Line 603 is a particle with D\\ = 4.03 * 10 ^"11 m ² s and Rparucie = 0.43 nm. Note the standard deviation at injection was zero for this figure.

[0100] Figure 7 depicts the dependence of the optimum flow angle for maximum shift, ffnow. Max, on the ratio of the particle radius to the cylinder radius (i.e., =

Rpartiele/Rcyl).

[0101] Figure 8 depicts the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array on the flow rate, v, with the distance travelled in the flow direction of = 0.1 m, a flow angle of = 45 °, the hexagonal lattice parameter of aHex = 5.4 nm and a cylinder radius of Rc _y i = 1 .4 nm for a particles with diffusion coefficient parallel to the cylinders of D\\ = 6.8 * 10 " m ² s ¹ and radius of Rpanido = 0.32 nm. Note the standard deviation at injection was zero for this figure.

[0102] Figure 9 depicts the dependence of the centre of the distribution, x _c , and its width, w, at the exit of the array on the hexagonal lattice parameter, aHex, with the distance travelled in the flow direction of - 0.1 m, a flow rate of v - 200 pm s ¹ , a flow angle of f v = 45 ° and a cylinder radius of Rc _y i = 1.4 nm for a particles with diffusion coefficient parallel to the cylinders of D\\ = 6.8 * 10 ^"1 ' m ² s ¹ and radius of particie = 0.32 nm. Note the standard deviation at injection was zero for this figure.

[0103] Figure 10 depicts the dependence of the centre of the distribution, x=, and its width, w, at the exit of the array on the cyli nder radius, Rc _y i, with the distance travelled in the flow direction of = 0.1 m, a flow rate of v - 200 pm s ¹ , a flow angle of 0 - 45 ⁰ and a hexagonal lattice parameter of aHex = 5.4 nm for a particles with diffusion coefficient parallel to the cylinders of D\\ = 6.8 χ 10 ¹¹ m ² s ^"1 and radius of RparMe = 0.32 nm. Note the standard deviation at injection was zero for this figure. [0104] Some figures below show the numerical (random walk simulations) and analytical predictions:

[0105] Figure 11 depicts the results of random walk simulations for a NanoRatchet device 300 compared to the analytically derived expressions for the centre of the distribution, _c in lots of aHex (i.e., normalised to the hexagonal lattice parameter, anex), and the width of the distribution, w in lots of anex. The distance travelled in the flow direction was dv - 200 nm, the particle had a diffusion coefficient parallel to the cylinder axes of D\\ = 6.8 * 10 ^'11 m ² s ^" \ the particle radi us was Rpar«ci _e = 0.32 nm, the flow rate was v = 5 mm s \ the hexagonal lattice parameter was aHex = 5.4 nm, the cylinder radius was Rcyi = 1.4 nm.■ and ® are the results of random walk simulations for c in lots of aHex and w in lots of aHex, respectively. The random walk simulations were performed with an array orientation of ΘΑ™ _υ = 0°, three flow angles of 0FI™ = 0°, 45° and 90°, 1000 particles were simulated, the step size in the flow direction was 0.003 nm, twice the maximum step size due to diffusion was 0.989 nm, the estimated number of steps was 66667. Note that the large step sizes were done to minimise simulation time, but note that the simulations can be done with these high step sizes without sacrificing the accuracy of the results greatly. Line 1102 (with square points 1101) and line 1104 (with circular points 1103) are the predictions from the analytically derived equations for Xc in lots of anex and w in lots of anex , respectively, where the solid l ines represent the predictions from numerical modelling and the points are simulation results.

[0106] Figure 12 depicts the results of random walk simulations for a NanoRatchet device 300 compared to the analytically derived expressions for the centre of the distribution, x _c in lots of anex (i.e., normalised to the hexagonal lattice parameter, aHex), and the width of the distribution, w in lots of ¾. The distance travelled in the flow direction was dv - 1 μηι, the particle had a diffusion coefficient parallel to the cylinder axes of D|| = 6.8 χ 10 ¹¹ m ² s ¹ , the particle radius was RpartMe = 0.32 nm, the flow rate was v - 20 mm s ^' (note that this flow rate the flow rate is probably unrealistic, but was assumed to make the simulations complete in a reasonable time for comparison of simulation results to the simplified theory developed), the hexagonal lattice parameter was aHex = 5.4 nm, the cyl inder radius was Rcyi = 1.4 nm. Points marked with symbols:

■ K/B/ _anc | #/β/Φ/€ are the results of random walk simulations for x _c in lots of aHex and w in lots of respectively and the different symbols represent different step sizes used. The random walk simulations were performed with an array orientation of 0Array = 0°, three flow angles of 0 = 0°, 10°, 20°, 45°, 75° and 90°, the step size in the flow direction was 0.05/0.0125/0.025/0.035 nm, twice the maximum step size due to diffusion was 2.018/1.009/1.427/1.688 nm, the estimated number of steps was 20000/80000/40000/28570, 1000 particles (for 0 _ow = 0°, 45° and 90° with step size in the flow direction of 0.05 nm) and 5000 particles (for 6> = 10°, 20° and 75° with step size in the flow direction of 0.05 nm) and 1000 particles (for all others) were simulated. Note that the large step sizes were done to minimise simulation time, but also note that the simulations can be done with these high step sizes without sacrificing the accuracy of the results greatly. Line 1202 (with square points 1201) and l ine 1204 (with circular points 1203) are the predictions from the analytically derived equations for c in lots of aHe and w in lots of ate, respectively, where the solid lines represent the predictions from numerical modelling and the points are simulation results.

[0107] Since the analytically derived expressions are shown to be good representations of the device characteristics by numerical simulations (as evidenced i n Figures 11 and 12), they were applied to test scenarios for ion separation and speciation. An example of the predicted separation (i.e., from the analytical expressions of the simple theory) for Pb ²⁺ , Fe ^2t , Fe ^3t and AP ions with the assumed and calculated parameters provides for the separation of ions where the smallest difference in radii is 0.024 nm and a largest difference in radii of 0.062 nm between adjacent particles (based on the best available information for the hydrated size of the ions). The results, as depicted in Figure 13, show the possibility of good separation after being forced electrophoretically through a 10 cm long NanoRatchet with aHex = 4 nm and Rc _y i = 1.5 nm. The time taken for the Pb ^2t after injection to reach the exit is 3.5 min, for Fe ²⁺ it is 5 min, for Fe ³⁺ it is 4.9 min and for Al ⁺ is 5.9 min, under an electric field of 671 V cnv' (i.e., 6.71 kV applied across the 10 cm). This would be the expected separation time in a batch separation scenario and is comparable to what can be expected from capillary electrophoresis for example (but the continuous operation means this time is only important for the description of a 'lag time', i.e., the time from injection to the time when the ions first exit the separation medium). [0108] Figure 13 shows the predicted separation of Pb ^2* 401, Fe ²⁺ 402, Fe ^3t 403 and AP 404 ions (present in equal concentration) for the selected characteristic ratio (i.e., the ratio of the hexagonal lattice parameter, anex = 4 nm, to the cylinder radius, Rc _y i = 1.5 nm; and assuming all ions 'see' the same characteristic ratio), xc is the shift from the flow line. The assumed radius for Pb ²⁺ was Rpamae - 0.261 nm, and the calculated diffusion parallel to the cylinders of D\\ = 1.049 * 10' ¹⁰ m ² S ^"1 (using the known free diffusion coefficient and obstruction factors from theory) and electrophoretic mobility in free solution at infinite dilution of με = 7.2 ^χ 10 ⁸ m ² s ^_ V ¹ (note that the electrophoretic mobility is reduced from με in the system and was assumed reduced by obstruction factors similar to as for diffusion); the assumed radius for Fe ^2* was = 0.291 nm with calculated D| = 7.216 * 10' ¹¹ m ² s ¹ and με = 5.55 χ 10 ^s m ² s ¹ V ¹ ; the assumed radius for Fe ³ⁱ was Rp _a rticie = 0.353 nm with calculated D\\ = 4.921 ^χ 10 ¹¹ m ² s ¹ and με = 7.05 * 10 ⁸ m ² S ^'1 V ¹ ; and the assumed radius for Al ³⁺ was Rparti e = 0.377 nm with calculated D \\ - 4.099 x 10" ¹¹ m ² s' ¹ and με = 6.32 10" ^a m ² s-' V ^"1 . The results were predicted for a distance travelled in the flow direction of dv = 0.1 m, a flow angle of 0 = 45°, the flow rate of Pb ²⁺ was set to v = 476 μιτι s ¹ which required an electric field of £ = 671 V cm ¹ which is a voltage across dv of V - 6.71 kV giving a flow rate for Fe ^2t of v = 332 pm s ¹ , a flow rate for Fe ³⁺ of v = 342 pm S ^'1 , a flow rate for Al ³⁺ of v = 282 pm s ¹ ; the standard deviation at injection was σο = 100 pm (so the injection width is 200 pm). 401 is the predicted result for the Pb ²⁺ alone, 402 is the predicted result for the Fe ²ⁱ alone, 403 is the predicted result for the Fe ³⁺ alone, 404 is the predicted result for the Al ³⁺ alone and is the combined result as would be expected for the separation from the Nanoratchet device 300. Note that these model predictions still include several assumptions and the appearance could improve with better models.

'MAC ORATCHET' EXAMPLE

[0109] To test the scalability of the NanoRatchet, a 3D printer (Model: uPrint SE Plus, Stratasys Inc.) was used to produce a working model made from acrylonitrile butadiene styrene (ABS) plastic and is shown in Figures 14 to 16. The stereolithography (.stl) file was generated from 3D models drawn in Google Sketchup 8 (free version; Trimble Sketchup v. 8.0.15158), and were converted to the toolpath (printer head path) file using CatalystEX v. 4.3 (Dimension, Stratasys Inc.). The stated layer resolution was 0.254 mm. The total print time was on the order about 50 h, and the time for the solvent wash bath (to remove the support material) was simi lar to this.

[01 10] There were three main components to the 3D pri nted scaled-up NanoRatchet (see Figure 15 but also Figure 14), the separation chamber casing with particle guides at the outlet (i.e., guides that match the bi ns used to collect the particles after exiting the array), the collection chamber for collecting the particles in different bins (i.e., integration of the particle distribution) and the separation chamber with entry chutes attached. The overall size of the assembled 3D printed scaled-up NanoRatchet was -16 ^χ 5 (10 for the widest part of the base) ^χ 22 cm (length * depth ^χ height). The separation chamber (i.e., the cylinder array) was 14.4 * 3.5 ^χ 15 cm (i.e., dv - 15 cm) and the arrangement of the cylinders was such that when upright it had a (9 - 50°, 6 ray - 20° (i.e., when measured from the cylinder i n the centre of a hexagon and in the xy plane, there is a cylinder 10° from the y axis and 20° from the x axis; see Figure 16), Rc was 2.5 mm and was 1 1 mm. This means that the maximum radius of particles that could be separated was 3 mm. There were two entry chutes with a radius at the point of entry to the array of 3 mm (one for 'normal' operation where the cylinders are tilted to give shift away from the side wall near the entry chute and one for 'reverse' operation where they are tilted to give shift towards the wall near the entry chute). But there was a larger radius funnel before this to help direct particles into the entry point.

[01 11] Two entry points were printed for this example because they were attached to the top of the cylinder array to fix their positions relative to the array but other arrangements could be printed.

[01 12] Figure 14 depicts a photograph of a 3D-printed scaled-up NanoRatchet 1400 (i.e., a MacroRatchet). Vertical arrow 1401 in Figure 14A indicates the direction of the flow (i.e., the particles fall through the device under gravitational force). Vertical arrow 1401 as shown, is pointing down the 'normal' entry chute that is at the top corner of the cyl inders in the separation chamber whereby the cylinders are tilted from top left to bottom right in the images. Horizontal red arrow 1403 in Figure 14(A) and 14(B) indicate the direction of separation (for this arrangement, and with the bins in the collection chamber (Figure14(B)(i)) horizontal). Also shown in Figure 14(B) is an example of the output for 40 particles (beads) with 1.97 mm average radius, after falling down the 'normal' entry chute marked in (A) with the vertical arrow 1401 into the collection chamber (Figure 14(B)(i)). Line 1405 in Figure14(B) is a depiction of a Gaussian curve that is used to describe the output particle distribution. The second entry point (i.e., the 'reverse' entry chute) on the top right corner in Figure 14 was 3D printed for a test to confirm that there was no shift if the particles are input via this entry. Note that this 3D printed scaled up Nano atchet is only one possible arrangement and size.

[0113] Figures 15(A) to (F) show a series of photographs of this example MacroRatchet device 1400 of Figure 14 depicting details of the components of the 3D printed scaled NanoRatchet 1400. Figure 15(A) shows the assembled NanoRatchet example device 1400 with the chute lids 1409 taken off. Figure 15(B)(i) shows a view into device 1400 from the base of the separation chamber. Collection bin dividers are visible in the collection chamber (Figure 15(B)(ii)), and in the base of the separation chamber casing at the bottom of the cylinders (Figure 15(B)(i)). The separation chamber containing the tilted cylinders can be removed from the separation chamber casing (Figure 15(C)). Figures 15(D) to 15(F) show the removed separation chamber (i.e., the array of tilted cylinders 1411 in a hexagonal lattice, for this example). Locator mark 1501 marks the same corner of the separation chamber.

[0114] Details of the collection bins are shown in Figure 16B. There was a distance of 7 mm between the centre of the bin dividers (or the centre of the bins) and the bin dividers were 2 mm thick, giving a collection bin width of 5 mm for the particles to enter a bin. That is, the bins integrate regions of 7 mm at a time. It should be noted that the last bin only had a 4 mm width for the particles to enter this bin because of the dimensions to the device (i.e., this is the left over space available for the length of the array printed - so from the centre of the last bin divider to the wall is 5 mm). This means the centre of the last bin is 0.5 mm before the centre if it was a 5 mm wide bin but this is of no real consequence since the bin is counting the particles that come after the previous bin divider and as this is the last bin there is no other bin to move to. For assigning the bin numbers, the bin that is only 4 mm wide is bin 21. With the separation chamber in place the divider between bins 1 and 2 is approximately directly below the entry chute. [0115] Figure 16 shows the definition of the parameters which describe the 3D- printed scaled NanoRatchet 1400 and its collection chamber 1407. Specifically, Figure 16(A) depicts the orientation of the array of cylinders 1411 in the xy-plane 1601 in this particular example (i.e., when viewed down the cylinder axes) as characterised by = 20° for this example.

[0116] Figure 16(A)(i) is a photograph of the 3D printed separation chamber (i.e., the array of tilted cylinders 1411 in a hexagonal lattice, for this example) viewed from the bottom and down the cylinder axes. Note that in the photo the cylinders 1411 appear to shift slightly off the hexagon and their diameter seems to change but this is due to perspective.

[0117] Figure 16 (A)(ii) A schematic showing the values for some parameters,

Rcyi and with with respect to the xy-plane 1601 and a hexagon 1603 ( ) representing the array with = 0° is shown for reference along with two cylinders 1411 (i.e., circles) shown on both the 0° and 20° hexagons. The y-axis shown matches the y-axis in Figure 2 and that if the x-axis was reversed in the schematic, the same goes for the x-axis in the Figure 2, the resulting would still be 20° (i.e., not 40°) since ranges from 0° to 30° because of symmetry.

[0118] Figure 16 (B) depicts the collection chamber 1407 with bins 1408 suited to hold particles up to 5 mm diameter - except the last bin (shaded) which is only suitable for 4 mm particles. The bin dividers (the vertical lines within the large rectangle shown in Figure 16(B)(ii) and as also seen in the separation chamber in Figure 15(B)(i)) are 2 mm thick, the vertical lines at the end of the rectangle shown in Figure 16{B)(ii) can also be considered as 2 mm thick, with the vertical lines in Figure 16(B)(ii) representing the centre of the bin dividers. Note that the entry point is approximately above the bin divider for the end of bin 1 and corresponds to ~0 cm shift. Bin 21 is shaded grey to highlight that this bin is 1 mm smaller than the others due to the number of bins and the width of the array 3D printed (which was limited by the 3D print envelope volume).

[0119] To test the separation of the particles through the 3D printed scaled-up NanoRatchet, beads of different sizes were passed through it. Two groups of differently sized particles/beads were tested (the larger beads were from Ribtex International Pty Ltd and smaller beads were from Darice Inc). To test the separation (i.e., through the 'normal' entry chute 1401), 800 particles of each size were passed through the device in lots of 40, and their output locations (i.e., collection bin number) were recorded (e.g., see the result for 40 large particles in Figure 14(8)(i)). Note that the particles of different sizes were added separately (i.e., not mixed) since this made the counting easier but there is still expected to be separation for a mixture. If a mixture was input it was difficult to see clearly the separation because large particles fill up the bin more than the smaller particles making it hard to compare the two distributions (i.e., compare the size of 40 particles of each size in the glass jars (A)(i) and (A)(ii) in Figure 17). The results were analysed and discussed in terms of the simple theory developed for the NanoRatchet. In some rare instances, the particles (particularly the larger ones) did not enter the bins and this is because of the arrangement and the depth of the array whereby the array orientation meant that there were cylinders near some bins that could pin the particle to the wall of the separation chamber. If this happened the particle was counted in the bin that it was above (if pinned just before the exit location) or it was redone. This usually only happened for 1 particle in a few of the runs of 40 particles. To test that the device would not separate or shift particles when the particles were added via the 'reverse' entry chute, the separation chamber was reversed in the casing so that the 'reverse' entry chute was now vertically above the bin divider for bins 1 and 2, and 40 particles of each size were input and the results were recorded. While the particles could have been added to the 'reverse' entry chute without reversing the separation chamber in the casing, this was not done due to the slightly smaller size of bin 21 , and reversing the separation chamber meant that the results were better comparable to the 'normal' entry chute results.

[0120] The particle radii used for the predictions were the average radii based on the particle/bead shape as they were not spherical (see Figure 17(B)). The reasoning for this was as follows: the variation of the particle radius due to the shape of the particles/beads (i.e., a spherical segment; capped spheres; spheres with a flat top and bottom) may have resulted in an averaged shift (e.g., c. Fit) other than predicted using the average radius and the simple theory, but because of the tumbling of the particle it may be represented adequately by the average radius from consideration of the shape. That is, the shift would vary more if particles constantly only hit on one side (i.e., one radius, larger or smaller than the average radius) all the way through the array with different particles consistently hitting on the same side (but a side with a different radius compared to others) so that some particles seem larger and some smaller - but this is very unlikely and the average radius from tumbling is therefore more acceptable. The average radius was calculated by using the average radius for the spherical section (i.e., average of the value found for a few beads) and the angular dependence of the average distance from the centre of the capped sphere to the plane at the top or bottom (i.e., angular average of the particle radius). The average radius for the large particles was 1.97 mm and for the small particles this was 0.92 mm.

[0121] It was noted when comparing the simple theory to the experimental results that it was necessary to consider/clarify the effect of particle size on the standard deviation of the injection (i.e., the standard deviation from injection was taken to be the input channel radius but this needed to be corrected for the effects of the particle radius). To test the effect of entry radius (and the correction for the particle radius) a smaller entry chute/chute modifier 1801 was 3D printed as well (as shown in Figure 18). The printer tolerance affected the inner diameter of the smaller entry chute/chute modifier where the printed inner diameter was smaller than the intended value because of the small value of the intended chute diameter - the smallest inner diameter value of the smaller entry chute/chute modifier was 2.1 mm (i.e., measured for the 3D printed component). The height of the modified entry chute was 9 mm (with a 1 mm lip at the top) and the outer diameter of the insert was about 5.5 mm (to ensure that it fitted in the larger entry port easily) and the outer diameter of the lip at the top of the cylinder was about 6.5 mm. This entry chute modifier was used to measure the results for 800 of the smaller particles input via the 'normal' entry chute and 40 smaller particles input via the 'reverse' entry chute for comparison to the results obtained for the larger entry chute.

[0122] Figure 18 shows a series of photographs of the entry chute width modifier 1801 (shown in Figure 18(C)) to test the effect on the width of the resulting distribution at the exit for the smaller particles in the 3D printed scaled NanoRatchet or 'MacroRatchet' (the chute modifier was also 3D printed and fits inside the entry chutes to reduce the entry point diameter. Figure 18(A) shows the entry chute that is suitable for both large and small particles tested (having a radius of about 3 mm) and Figure 18(B) shows the entry chute reduced to a radius of about 1.05 mm to reduce the standard deviation of the entry point for the small particles.

Results & Discussion

[0123] The important points arising from the scaled-up 'MacorRatchet' example described above are:

[0124] As can be seen in Figure 19, there is separation of the two bead sizes (only slight given the dimensions of the printed ratchet).

[0125] As can be seen in Figure 21, the width of the smaller bead distribution is improved (narrowed from line 2107 goes to line 2103) using a more controlled entry chute (meaning that in an upscaled device, the entry point needs to be controlled well for narrow distributions), but the scaled version with less control (i.e., Figure 19) would still give separation if the dimensions (i.e., the number of bins) were increased.

[0126] Simple theory needs modifications to include other effects in scaled up versions (equations not shown).

[0127] There is no separation if the arrangement angle is reversed (i.e., opposite tilt with respect to the entry point) as can be seen in Figure 23.

[0128] The results for separation after passing approximately 800 of each of the two differetly sized particles (beads) through the 3D printed scaled-up NanoRatchet are shown in Figure 19 (these are the results with the entry chute size suited for addition of both particles). It is seen that there is slight separation of the two particle sizes with the larger particles being shifted more than the smaller particles (i.e., the two distributions don't share the same centre). More separation is expected for larger separation chamber/array dimensions but this was limited for the presented case by the available 3D print envelope volume.

[0129] Comparison of these experimental results to the results predicted from the simple analytical treatment with the initial assumption that the diffusion coefficient is zero (a fair assumption on the macroscopic scale), shows that the 3D printed scaled-up NanoRatchet gives very different results (compare Figure 19 (experimental) with Figure 20 (predictions)) but after consideration, the differences can be accounted for (though in a qualitative way for some).

[0130] Figure 19 shows experimental results for the separation of two particles (beads) of different sizes and Gaussian curves fitted to the results. The result for the large particles are given by point 1905 and the corresponding curve-fit 1907, and the results for the small particles are given by points 1901 and curve-fit 1903. The data points (i.e., points 1901 and 1905) are experimental data points in this figure. A full Gaussian curve, (N _pit /(w _Fit [n/2 ) exp (— 2(x - x _c ,Fit† / ^w _t ) _> ^w 'th Npamaes = 800 (i.e., peak area; total number of particles) was fitted to each distribution where Xc. m was 16.15 ± 0.16 bins and iv™ was 4.72 + 0.26 bins (large particles; R ² = 0.929); and Xc was 15.47 + 0.19 bins and ¾ was 6.45 ± 0.31 bins (small particles; R ^z = 0.933) . The entry chute radius was 3 mm and was the same for both particles. The vertical grid lines in the figure represent the centres of the bin dividers.

[0131] Figure 20 shows the expected outcome for a total of 800 of each size particle (bead) predicted using slightly modified equations from the simple analytical models derived. The predicted results shown are the expected Gaussian distribution and corresponding discrete distribution (i.e., histogram) results. The result for the large particles are given by points 2005 (histogram format via integration for each bin of the predicted Gaussian curve) and curve-fit 2007 (predicted Gaussian curve); and the results for the small particles are given by points 2001 and curve-fit 2003. The entry chute radius was 3 mm (i.e., this was taken to be σο and was the same for both particles - but consideration for this showed this is not the correct ao and needed modification). The vertical grid lines in the figure represent the centres of the bin dividers.

[0132] It is noted from Figure 20 that the particles are not expected to have as large a shift and width as observed in the experimental results (Figure 19). In addition to this, the two distribution for the two differently-sized particles have different widths in the experimental results of Figure 19 leading to different peak heights, while in the predicted results (Figure 20) the peak widths and heights are the same. This is an important observation since it leads to a necessary clarification for the meaning of σο in the simple analytical models which can involve a simple and justifiable modification to the equation describing w. It is necessary to correct for RpartMe in the standard deviation of the injection channel/entry chute and this is especially relevant when Rpani e becomes larger or more comparable to the standard deviation of the entry chute, σο (i.e., approximated as ½ injection channel/entry chute width in the direction of the separation).

[0133] To test that this effect has a significant effect on the outcomes of the 3D printed scaled-up NanoRatchet and predictions, a smaller entry chute/chute modifier with a smaller inner diameter (i.e., smaller σο, Actual) suitable for the smal ler particles only was also 3D printed. This fitted in the larger (3 mm radius) entry chute reducing its radius to 1.05 mm. This means that the corrected injection standard deviations (i.e., σο) for both entry chutes were 1.03 mm (large particles, 3 mm radius entry), 2.08 mm (small particles, 3 mm radius entry) and 0.13 mm (small particles, 1.05 mm radius entry).

[0134] Figure 21 shows the experimental results highlighting the influence of the particle size on the width of the output distri butions for the separation of two particles (beads) of different sizes and Gaussian curve s fitted to the results. The result for the large particles with entry chute radius of 3 mm are given by points 2101 and the corresponding curve-fit 2103; and the results for the small particles with entry chute radius of 3 mm are given by 2105 and curve-fit 2107 (note that the data points 2105 and 2101 and respective curve-fits 2103 and 2107 are the same as in Figure 19). The results for the smaller particles with the entry chute modifier so that the entry chute radius is 1.05 mm are given by points 2109 and corresponding curve-fit 2111. A ful l Gaussian curve, exp [-2(x - x ^w ' ^tn /particies = 800 (i.e., peak area; total number of particles) was also fitted to each distribution and c, was 16.15 ± 0.16 bins and ιν was 4.72 ± 0.26 bins (large particles, 3 mm radius entry; R ² = 0.929); c was 15.47 ± 0.19 bins and wm was 6.45 ± 0.31 bins (small particles, 3 mm radius entry; R ² = 0.933) and with the entry chute modifier Xc, was 14.89 ± 0.14 bins and was 4.95 ± 0.23 bins (small particles, 1.05 mm radius entry; R ² = 0.941 ). The vertical grid l ines in the figure represent the centres of the bin dividers.

[0135] Figure 22 shows the expected outcome for a total of 800 of each size particle (bead) predicted using the simple analytical models with standard deviation of the entry chute modified to be the particle appropriate val ue (i.e., σο = σο, Actual - Rparticie) with the assumption of zero diffusion and using slightly modified equations for the expected w, and for the expected The predicted results shown are the expected Gaussian distribution and the expected discrete (i.e., histogram) results. The result for the large particles are given by points 2201 (histogram format via integration for each bin of the predicted Gaussian curve) and line 2203 (the predicted Gaussian curve) with

3 mm, and the results for the small particles are given by points 2205 and line 2207 with = 3 mm. The results for the small particles with - 1.05 mm are given by points 2209 and line 2211. The vertical grid lines in the figure represent the centres of the bin dividers. Empirical fit constants (equations not shown) can be included in the simple models to shift the predictions of Figure 22 to desribe the results of Figure 21 and could be calibrated as a characteristic of a scaled-up NanoRatchet using the separation results for particles/beads of known size or the theory could be extended to include other factors important for scaled-up NanoRatchets.

[0136] To confirm that the separation and shift of the particles is due to the cylinders and that no separation/shift occurs when the particles are input via the opposite entry chute, the separation chamber was turned around in the casing (i.e., so as to align the opposite entry chute with the bin labelled as bin 1 in the collection chamber for comparison to the results with the 'normal' entry chute (i.e., see Figure 19 and Figure 21), and 40 beads of each particle size were put through the device (note that the smaller particles were done twice - once for each size of entry chute). The tilt of the cylinders under this 'reverse' entry chute are the opposite way compared to under the 'normal' entry chute and because of this the particles just remain against the edge of the separation chamber below the entry point and do not shift (see Figure 23) (actually there is a small negative shift, i.e., shift into bin 1 which is less than the 0 cm shift, due to the 0 cm shift position being approximately aligned with the bin divider between bins 1 and 2). The results were fitted with a reflected Gaussian curve (Figure 23(A)) and a full Gaussian curve (Figure 23(B)). Interestingly the "diffusion" (i.e., a random component) noted to be present in the separation experiment was also noted here but it is much less (i.e., less broadening, probably due to the particles being kept close to the wall by the cylinders). If there wasn't a random component present, the particles would travel straight down with minimal shift (i.e., slight negative shift as mentioned) and minimal broadening. Whi le the fits are expected to be improved by increasing the number of particles, the main aim of this test was only to show that there was no shift for the opposite entry chute and this was demonstrated in the figures.

[0137] Figure 23 Experimental results and fitted curves for 40 of each particle size entering the array from the opposite side (i.e., the 'reverse' entry chute) showing that there is no separation or shift (actually there is a slight shift in the negative direction- given that bin 1 is before the 0 cm shift location) if the cylinders are tilted the opposite way. The result for the large particles with entry chute radius of 3 mm are given by points 2301 and corresponding curve-fit 2303, the results for the small particles with entry chute radius of 3 mm are given by points 2305 and curve-fit 2307, and the results for the smaller particles with the entry chute modifier so that the entry chute radius is 1.05 mm are given by points 2309 and curve-fit 2311 (note that line 2311 is underneath line 2307 in Figure 23(A)). A reflected Gaussian curve with centre 1 was fitted to the data as shown in Figure 23(A) and the corresponding equation is given by exp[-2{x - 1 ) ² /w _it ] where Npanides is 40. The value of IVRI was 2.15 ± 0.15 bins (large particles; R ² - 0.879), 3.55 ± 0.30 bins (small particles without modified entry chute; R ² = 0.845) and 3.52 ± 0.22 bins (small particles with the modified entry chute; R ² = 0.905). A full Gaussian curve was also fitted for comparison and is shown in Figure 23(B) and the corresponding equation is given by exp[-2(x - 1 ) ² /w| _it ] and wm was 1.22 ± 0.05 bins (large particles; R ² = 0.974), 1.61 + 0.12 bins (small particles without modified entry chute; R ² = 0.902) and 1.89 ± 0.12 bins (small particles with the modified entry chute; R ² = 0.941 ). The vertical grid lines in the figure represent the centres of the bin dividers.

[0138] It will be appreciated that the methods & apparatus described/illustrated above at least substantially provide methods, systems & apparatus for separation of nano-scale molecules, ions and/or particles.

[0139] The methods, systems & apparatus described herein, and/or shown in the drawings, are presented by way of example only and are not limiting as to the scope of the invention. Unless otherwise specifically stated, individual aspects and components of the methods, systems & apparatus may be modified, or may have been substituted therefore known equivalents, or as yet unknown substitutes such as may be developed in the future or such as may be found to be acceptable substitutes in the future. The methods, systems & apparatus may also be modified for a variety of applications while remaining within the scope and spirit of the claimed invention, since the range of potential applications is great, and since it is intended that the present methods, systems & apparatus be adaptable to many such variations.