Device with an arrangement of nano-columns and method for forming an arrangement of nano-columns

Abstract

Device (1100) comprising: an inlet (1105) configured to receive a fluid; an outlet (940, 945) configured to discharge the fluid; and an arrangement (320) of nano-columns (314) connected to the inlet (1105) and the outlet (940, 945), wherein the arrangement (320) of nano-columns (314) is configured to allow the fluid to flow from the inlet (1105) to the outlet (940, 945); wherein the arrangement (320) of nano-columns (314) includes nano-columns (314) arranged to separate entities according to size;wherein the nano-pillars (314) have an inwardly curved shape on the order of nanometers in the center of the nano-pillars (314), forming a non-uniformity of diameter, wherein the arrangement (320) of nano-pillars (314) has an oxide layer (316) applied to the nano-pillars (314), the oxide layer (316) filling the inwardly curved shape in the center and straightening the nano-pillars (314) with the oxide layer (316) applied to them into a cylindrical shape, wherein the nano-pillars (314) are arranged such that they have a space (G5) that separates one nano-pillar (314) with the oxide layer (316) applied to it from another nano-pillar (314) with the oxide layer (316) applied to it; and wherein the space (G5) is designed to be on the order of nanometers.;

Description

BACKGROUNDThe present invention relates to a nanocolumn array device for separating biological entities and a method for forming an array of nanocolumns.Separation and sorting of biological entities such as cells, proteins, deoxyribonucleic acid (DNA), ribonucleic acid (RNA), etc. is important for a wide variety of biomedical applications including diagnostics, therapeutics, cell biology, and proteomics.Protein and DNA / RNA separation for analytical purposes is conventionally performed by gel electrophoresis in which a protein mix is subjected to a strong electric field (typically 30 volts per centimeter (V / cm)). Proteins or DNA / RNA move through the gel at a rate dependent on their dimension and surface charge. The gels are prepared from agarose or acrylamide polymers known to be toxic. The result of the electrophoresis experiment is optically obtained from staining of the proteins with a dye or from staining of the DNA / RNA with ethidium bromide, which is extremely carcinogen. Gels require sufficient amounts of material to allow the result of electrophoresis to be detectable, but poor cross-linking in the gel matrix often results in unaccounted for results and complete loss of samples. If the dimension of the gel matrix is not matched to the dimension of the sample molecules, or if electrophoresis is run too long, the sample is also lost.Gel electrophoresis is widely used for separation of macromolecules, such as DNA, RNA, proteins and fragments thereof. Gel electrophoresis currently has a market with world-wide turnovers of more than $1 billion dollars per year. Gel electrophoresis used in medical diagnostics represents a market of many billion dollars.Compared to conventional techniques, silicon (Si) nano-manufacturing technology provides a much more precise and accurate control in nano-structural dimensions and positioning thereof, and thus may result in reliable sorting of particles based on their dimensions. To date, Si-based lab-on-a-chip approaches using Si pillar arrays have been promising. Using these techniques, only micron-scale sorting (10 6 or micron (μm) range) has been documented, but this does not provide access to the nanometer dimensions required for sorting DNA, proteins, etc.US 2003 / 0 049 563 A1 relates to an apparatus for fractionating a sample into microstructures of different sizes, comprising a fractionating unit having a region in which the sample can migrate and at least one colony of microbodies which serve as obstacles to migration of the microstructures which are formed in a region of the region and define a labyrinth for capturing microstructures of small size, such that a remaining region of the region serves as a path for microstructures of large size.The publication "Clog-free translocation of long DNA in nanofluental pillar arrays and 30 nm wide channels: a fabrication and hydrodynamic study" relates to the production of nanofluentic chips with critical dimensions of only 30 nm without electron beam writing and the investigation of DNA translocation and hydrodynamic interactions with nanocolumns and nanochannels. The geometric dimensions and the arrangement of the nanostructures have a considerable influence on the translocation process and the stretching of DNA molecules (WANG, Chao [et al.]: Clog-free translocation of long DNA in nanofluental pillar arrays and 30 nm wide channels: a fabrication and hydrodynamic study. In: 18th International Conference on Miniaturized Systems for Chemistry and Life Science, October 26-30 2014, pp. 1347-1349. - isbn 978-0-9798064-7-6 - url: https: / / www.rsc.org / images / loc / 2014 / PDFs / Papers / 445_0190.pdf [retrieved on 27.8.2020]).US 2004 / 0 125 266 A1 relates to a functional element provided with a group of columnar microcolumns of an organic polymer which are arranged in a predetermined pattern on a matrix of an organic polymer which is provided on a carrier element. The microcolumns have an equivalent diameter of 10 nm to 500 μm, a height of 50 nm to 5000 μm, and an aspect ratio of 4 or more. The equivalent diameter of the tip ends of the microcolumns is smaller than the equivalent diameter of the bottom surface.US 2014 / 0 256 028 A1 relates to a semiconductor microanalysis chip for detecting fine particles in a sample liquid. The chip includes a semiconductor substrate, a first flow channel provided in the semiconductor substrate, the sample liquid being introduced into the first flow channel, and a plurality of columnar structures laid entirely in the first flow channel at predetermined arrangement intervals.US 2002 / 0 079 490 A1 relates to a structure for use in DNA separation. The structure includes: a semiconductor substrate; and a plurality of pillars formed on the semiconductor substrate and at least one surface layer of which is made of an oxide of the semiconductor, wherein passage of DNAs through the pillars enables separation of the DNAs.The publication "Three-dimensional nanostructured substrates town efficient capture of circulating tumor cells" relates to a cell capture platform based on nanostructured 3D substrates, which has efficient cell capture. This is probably due to enhanced local topographical interactions between nanostructured 3D substrates and extracellular extensions as well as to the biological recognition of anti-epithelial cell adhesion molecules / epithelial cell adhesion molecules (anti-EpCAM / EpCAM). With these nanostructured substrates, it is possible to reliably trap cancer cells from artificial blood samples with circulating tumor cells. The viability of the cells is 84-91% with this platform, which allows subsequent release of the cells, their cultivation and the performance of molecular biological diagnoses (WANG S, WANG H, JIAO J, et al.: Three-dimensional nanostructured substrates toward efficient capture of circulating tumor cells. Angew Chem Int Ed Engl. 2009; 48(47):8970-8973).US 2013 / 0 209 988 A1 relates to selective capture of components, such as living cells, from biological samples, such as blood.SUMMARYThe invention relates to an apparatus and a method for forming an array of nanocolumns, the features of which are set forth in the respective independent claims. Embodiments of the invention are set forth in the dependent claims.An apparatus includes an inlet configured to receive a fluid, an outlet configured to discharge the fluid, and an array of nano-columns connected to the inlet and to the outlet. The array of nanocolumns is configured to allow the fluid to flow from the inlet to the outlet. The array of nanocolumns includes nanocolumns arranged such that entities are separated by size. The nanocolumns are arranged to have a space separating one nanocolumn from another nanocolumn. The clearance is designed to be on the order of nanometers.A method for providing a fluid device includes providing an inlet configured to receive a fluid, providing an outlet configured to discharge the fluid, and coupling an array of nano-columns to the inlet and to the outlet. The array of nanocolumns is configured to allow the fluid to flow from the inlet to the outlet. The array of nanocolumns includes nanocolumns arranged such that entities are separated by size. The nanocolumns are arranged to have a space separating one nanocolumn from another nanocolumn. The clearance is designed to be on the order of nanometers.A method for forming an array of nano-pillars includes arranging a hard mask layer on a substrate and patterning a resist layer into a pattern of the array of nano-pillars, the resist layer having been arranged on the hard mask layer. The resist layer is used to pattern the hard mask layer into the nano-pillar array structure such that both the resist layer and the hard mask layer have the nano-pillar array structure. The substrate is patterned into the nanocolumn array structure so as to form the nanocolumn array. The resist layer and the hard mask layer are removed, and nano-pillars in the array of nano-pillars have a first gap dimension relative to each other in a side-by-side arrangement relationship. The first pitch dimension is reduced to a second pitch dimension by disposing an oxide layer on the array of nano-pillars.An apparatus includes an inlet configured to receive a fluid, an outlet configured to discharge the fluid, and an array of nano-columns connected to the inlet and to the outlet. The array of nanocolumns is configured to allow the fluid to flow from the inlet to the outlet. The array of nanocolumns includes nanocolumns arranged such that entities are separated by size. The nanocolumns are arranged to have a space separating one nanocolumn from another nanocolumn. The clearance is designed to be on the order of nanometers. A monomolecular layer is attached to the nanocolumns, and the monomolecular layer reduces a dimension of the gap.The techniques of the present invention realize further features and advantages. Other embodiments and aspects of the invention are described in detail herein and are considered part of the claimed invention. For a better understanding of the invention with the advantages and features, see the description and drawings.BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGSThe subject matter which is considered to be the invention is particularly pointed out and clearly claimed in the claims at the end of the specification. The foregoing and other features and advantages of the invention will be apparent from the following detailed description taken in conjunction with the accompanying drawings, in which: FIG. 1 is a schematic illustration of a deterministic lateral displacement (DLD) array showing definitions of the parameters of the array; FIG. 2A shows a schematic representation of particle trajectories at the interface between a neutral region and an element made of microfluidic metamaterial; FIG. 2B illustrates the simplest element of metamaterial consisting of an asymmetric arrangement of posts tilted at an angle +α relative to the channel walls and a main fluid flow; FIG. 2C is a cross-sectional SEM image showing the arrangement of microfabricated stands; FIG. 2D illustrates an equivalent microfluidic birefringence based on a particle dimension showing the time trace of a 2.7 μm red fluorescent substance crossing the interface and deflected from the normal; FIGS. 3A to 3G illustrate schematics of a process flow for the fabrication of an array of nano-pillars according to an embodiment in which: FIG. 3A illustrates a hard mask layer disposed on a substrate; FIG. 3B illustrates disposing a resist layer on the hard mask layer; FIG. 3C illustrates patterning of the resist layer; FIG. 3D illustrates patterning of the hard mask layer; FIG. 3E illustrates etching the substrate to the array of pillars; FIG. 3F illustrates the arrangement of pillars with the hard mask pattern removed; and FIG. 3G illustrates disposing an oxide layer on the array of pillars; FIGS. 4A and 4B are scanning electron micrographs of the same wafer, according to an embodiment, to show the result of a reactive ion etching process before the hard masks are removed; FIGS. 4C and 4D are scanning electron micrographs of a parallel processed wafer according to an embodiment to show the result of a reactive ion etching process after the hard masks are removed; FIGS. 5A and 5B are scanning electron micrographs of another wafer to show a fabricated array of nanocolumns without a thermal oxide, according to an embodiment; FIGS. 5C, 5D, and 5E are scanning electron micrographs of a parallel processed wafer according to an embodiment to show the influence of growth of a thermal oxide on nanocolumn arrays; FIGS. 6A and 6B are scanning electron micrographs of another wafer to show starting with a smaller gap dimension, according to an embodiment; FIGS. 6C and 6D are scanning electron micrographs of a parallel processed wafer according to an embodiment to show the oxidation process when the initial gap dimension is small; FIG. 7A illustrates a general chemical scheme of chemical modification to an array of columns according to an embodiment for forming sorting array surfaces; FIG. 7B illustrates a chemical scheme for chemical modification by attaching a metal to an array of pillars according to an embodiment for forming sorting array surfaces; FIGS. 8A to 8D are cross-sectional views illustrating chemical modification of sorting assemblies according to an embodiment as a means for modifying the gap dimension between columns in which: FIG. 8A illustrates the intercolumn gap dimension prior to chemical modification; FIG. 8B illustrates the reduced intercolumn gap dimension after chemical modification; FIG. 8C is an enlarged view of a reactive site in FIG. 8A ; and FIG. 8D is an enlarged view of the monomolecular layer in FIG. 8B; FIG. 9A is a plan view illustrating a particle flow in a chemically modified sorting arrangement according to an embodiment with particles having no affinity to the monomolecular surface layer compared to particles having an affinity to the monomolecular surface layer; FIG. 9B is an enlarged view of a cross section of the nano-column, the monomolecular layer, and the particle having an affinity according to an embodiment; FIG. 10A is a cross-sectional view illustrating pillars according to an embodiment with variation of the clearance; FIG. 10B is a cross-sectional view illustrating the oxidation process that eliminates the variation of the clearance according to an embodiment; FIG. 11 is a plan view illustrating a chip (a fluid unit) having the array of pillars according to an embodiment; FIG. 12 is a method for providing a fluid device (e.g., a chip) according to an embodiment; FIG. 13 is a method for forming an array of nano-pillars, according to an embodiment.DETAILED DESCRIPTIONSorting in the micrometer (10 6 μm) range has been documented using Si-based lab-ona-chip approaches. Further information in this regard is further discussed in a paper entitled "Hydrodynamic Metamaterials: Micro-fabricated Arrays To Steer, Refract, And Focus Streams Of Biomaterials" by Keith J. Morton et al. in PNAS 2008, 105(21), 7434-7438 (published before a print on May 21, 2008), which is incorporated herein by reference.The publication "Hydrodynamic Metamaterials: Microfabraded Arrays To Steer, Refract, And Focus Streams Of Biomaterials" discusses that their understanding of optics resulted from viewing light as particles that moved in straight lines and refracted into media in which the speed of the light was material dependent. The publication showed that objects moving in a laminar flow with a high Peclet number through a structured anisotropic hydrodynamic medium move along trajectories resembling light rays in the optics. One example is the periodic microfabricated array of posts, known as the deterministic lateral displacement (DLD) array, a high resolution microfluidic particle sorter. This arrangement of posts is asymmetric. Each subsequent downstream row is shifted relative to the previous row such that the axis of the array forms an angle α relative to the channel walls and the direction of fluid flow, as shown in Figure 1. During operation, particles having dimensions greater than a certain critical dimension are laterally displaced by a stand at each row and follow a deterministic path through the array in the so-called "jumping (impact)" mode. The trajectory of the impacting particles follows the angle α of the axis of the array. Particles whose dimensions are smaller than the critical dimension follow the flow lines of the stream, "weaving" in a periodic "zigzag" mode by the arrangement of uprights.FIG. 1 is a schematic illustration of a deterministic lateral displacement (DLD) array showing definitions of the parameters of the array. The uprights are periodically arranged at a distance λ, and each downstream row is laterally offset from the previous row by the amount δ, breaking the symmetry of the arrangement. This axis of the arrangement forms an angle α = tan -1( δ / λ) = tan -1( ε) with respect to the duct walls and therefore with respect to the direction of the fluid flow. Due to the asymmetry of the arrangement, the fluid flow in the spaces between the stators G is divided into 1 / ε gaps. Each of these columns repeats every 1 / ε row so that the current through the array is generally straight. Particles traversing the gap near a stator may be displaced into an adjacent streamline (from a gap 1 to a gap 2) when the particle radius is greater than the width of the gap in the gap. Therefore, larger particles are deterministically shifted at each stator and travel at an angle α to the current. Smaller particles simply follow the streamline paths and flow in the direction of fluid flow through the array.Figure 2A shows a dimensional birefringence of particles passing through hydrodynamic medium of cross-channel microfabricated stators. Two particles of different dimensions impinge perpendicularly on an interface between a symmetrical array of uprights (left half of the channel) and an asymmetrical array of uprights (right half). Pressurized fluid flow through the assemblies is from left to right with its overall direction determined by the larger microfluidic channel. FIG. 2B shows a schematic illustration of particle trajectories at the interface between a neutral region and an element made of a microfluidic metamaterial. Particles whose dimensions are greater than a critical dimension follow the asymmetry of the arrangement, while smaller particles follow the fluid stream. Figure 2B illustrates that the simplest element of metamaterial is an asymmetric arrangement of posts tilted at an angle +α relative to the channel walls and the main fluid stream. A scanning electron microscope (SEM) photograph in plan view from the interface between a neutral arrangement (α=0°) and an arrangement having an arrangement angle α=11.3° is shown (the pitch G=4 μm and the pitch of the stands λ=11 μm are the same for both sides). FIG. 2C illustrates a cross-sectional SEM image showing the arrangement of the microfabricated posts. FIG. 2D illustrates an equivalent microfluidic birefringence based on the particle dimension, showing the time trace of a 2.7 μm red fluorescent substance crossing the interface and deflected from normal. Smaller green 1.1 μm beads are not deflected at the interface.Elements of an array can be tailored to direct specific particle dimensions at an angle to the current by making arrays with design parameters shown in FIG. 1 that include an obstruction dimension D, a distance G between the posts, and a grid dimension λ of the posts. The asymmetry is determined by the extent of the shift δ from row to row and is characterized by the slope ε = δ / λ. The definitive angle of the arrangement is then α=tan -1( ε). For a given angle of placement, the critical particle dimension for the jumping mode is determined by the ratio between the particle diameter and the spacing or spacing of the posts. This critical particle dimension has been described above for a range of placement angles between 1.0° and 16°. For a given clearance dimension, the critical dimension is larger for more steeply angled rebound. Using these design criteria, streams of beads, cells and DNA were all deterministically moved for dimensional-based separation applications. For the example indicated in FIG. 1, which has an angle of arrangement of 11.3°, an intermediate space G=4 μm and a grid dimension λ=11 μm of the stands, the limit value of the particle dimension is ≈2.4 μm. Thus, 2.7 μm red beads in the jumping mode travel along the angle of the axis of the array, and the 1.0 μm green beads travel in the zigzag mode along streamlines as shown. The elements of the array and any auxiliary microfluidic channels and reservoirs are fabricated in silicon wafers using conventional photolithography and etching microlithography techniques. Assemblies can also be molded in PDMS using a silicon master manufactured in a similar manner. For the silicon etch, an optimized deep reactive ion etch (DRIE) is used to obtain smooth vertical sidewalls that ensure a uniform top-to-bottom spacing between posts, as shown in FIG. 2C.Unlike the prior art, embodiments are designed to produce in-line arrays of silicon pillars with uniform inter-pillar clearances (also referred to as pillars) having sub-100 nanometer (nm) dimensions. These arrays of columns can be used, for example, in a configuration for a bumper array as described above for the sorting and separation of biological entities at these dimensions, such as DNA, RNA, exosomes, single proteins and protein complexes. In particular, the arrays of pillars are designed with an oxide coating, such as a SiO 2- coating, which can be used to "heal" a variation in the gap dimension along the entire axis of the pillars. Uniform gap dimensions are used to achieve efficient sorting, e.g., to sort a 20 nm particle from a 10 nm particle. This presents a particular challenge for sub-100 nm gaps where there is an inherent variation in gap dimension that is greater than the dimensions of the particles to be sorted, which is limited at this order by the reactive ion etching (RIE) process. Documented sorting gaps of columns found in the prior art have dimensions in the micrometer range, and therefore the prior art cannot sort close to this fineness of dimension disclosed in embodiments. Also, for an array of columns having a very small angular pitch (also referred to as an angle of an array and a critical angle) of, for example, 0.57 degrees at which a sorting efficiency is the highest, only a particle whose dimension is larger than 12% of the dimension of the clearance is sorted. Therefore, uniform nanometric spaces are required to sort a protein complex, for example. Sorting of individual proteins (e.g. in a dimensional range of 1 to 10 nm) is conventionally performed using ion exchange chromatography or gel electrophoresis, which are load-and-sort techniques rather than a solution based on Si with a continuous stream. However, the prior art method does not have an existing solution for sorting entities on the order of 10 to 100 nm, but the embodiments provide a solution in both of these ranges (e.g., the range of 1 to 10 nm and the range of 10 to 100 nm). Embodiments also include chemically modifying the sows via attachment and / or attachment of molecules to further reduce a given space to a customized dimension.For ease of understanding, sub-headers may sometimes be used. It should be noted that the sub-headings are for purposes of explanation only and not limitation.Production of an Arrangement of ColumnsFIGS. 3A through 3G illustrate schematics of a process flow for fabricating an array of nano-pillars, according to an embodiment. In FIG. 3A, a process flow 301 illustrates a substrate 302. A hard mask 304 is disposed on top of the substrate 302. The substrate 302 may be a wafer, such as a silicon (Si) wafer. The oxide hard mask 304 may be made of silicon dioxide (SiO 2) used for an etching process. Although an oxide is an example, a nitride or other hard material may be used. The oxide hardmask 304 may be deposited and / or grown on a bulk silicon (substrate 302). The thickness of the oxide hardmask 304 may be in a range of tens to hundreds of nanometers depending on the etch depth necessary to generate the height of the pillars and the selectivity of the RIE chemistry for the substrate 302 to the material of the hardmask 304. Other materials may be used for the substrate 302 and the hard mask layer 304.In FIG. 3B, process flow 303 illustrates placing a resist 306 on top of oxide hardmask 304. The resist 306 may be a positive resist or a negative resist. The thickness of the resist 306 may be in a range of 100 nm to 1 μm depending on the etch selectivity of the resist 306 / hard mask ( 304), the thickness of the hard mask 304, and the required pitch resolution of the nano-pillars. For narrow sub-100 nm gaps and small column depths, a resist thickness range of 100 to 500 nm is used to achieve higher resolution elements with less gap dimension variability. The resist 306 may also be a multi-layer resist stack consisting of two or more layers each having different etching selectivity to improve resolution.In FIG. 3C, process flow 305 illustrates patterning resist 306 into a resist pattern 308. The resist pattern 308 may be defined using, but is not limited to, electron beam lithography, nanoimprint lithography, interference lithography, extreme ultraviolet lithography, and / or deep ultraviolet lithography, or a combination of these techniques. The resist pattern 308 is formed into resist pillars in the pattern of the future array of nano-pillars. In one case, the resist pattern 308 may include multiple patterns for different nanocolumn arrangements.The process flow 307 illustrates pattern transfer from the resist pattern 308 to the oxide hardmask 304 to result in the etched hardmask pattern 312 in FIG. 3D. Pattern transfer to hard mask 304 may be performed using reactive ion etching (RIE). The process flow 307 shows the resist pattern 308 on top of the corresponding etched hard mask pattern 312.In FIG. 3E, the process flow 309 illustrates patterning the nano-pillars 314 to be defined in the substrate 302 below the etched hard mask pattern 312. The nano-pillars 314 may be etched using reactive ion etching. The resist pattern 308 may be removed from the top surface of the etched hard mask pattern 312 before the nano-pillars 314 are patterned in the substrate 302 or after the nano-pillars 314 are patterned. Removal of the resist pattern 308 may be performed after the etching of the nano-pillars 314 because it may serve to avoid erosion of the hard mask pattern 312 that may occur during the RIE process of the nano-pillars 314. Erosion of the hard mask may in turn result in pillars with a deformed (undesirable) sidewall angle.The process flow 311 in FIG. 3F illustrates removal of the hard mask pattern 312. The hard mask pattern 312 may be removed in dilute hydrofluoric acid (DHF acid) when the material of the hard mask layer is SiO 2. The process flow 311 shows a nano-pillar array 320 of nano-pillars 314.To further reduce the dimension of spaces between the nano-pillars 314 respectively and to reduce variation in spaces, the process flow 313 of FIG. 3G illustrates arranging an oxide 316 to cover the surface of the nano-pillar array 320 formed in the substrate 302. In one case, thermal oxidation may be used to grow silicon dioxide 316 to cover the surface of the nanocolumn array 320 to narrow the gaps. In another case, the oxide 316 may be deposited using atomic layer deposition on the array 320 of nano-pillars (made of silicon), for example.Generally, pillar arrays include a dense array of silicon pillars defined by RIE followed by an oxidation process (e.g., process flow 313) that serves to narrow the gaps between the pillar pillars and minimize variation in the gaps. The fabrication of an array of nanocolumns may also include an optional chemical modification process in which further gap scaling (i.e., a reduction in dimension) may be required. These arrangements of columns and / or spaces may be implemented in angled column designs to concentrate a sample or separate a heterogeneous mixture of biological entities at the level of individual molecules, similar to the work documented by the publication "Hydrodynamic Metamaterials: Microfabilitated Arrays To Steer, Refract, And Focus Streams Of Biomaterials" for sorting cells or large particles. The process flow for fabricating an array of nano-pillars in FIGS. 3A and 3B may be used to generate arrays of nano-pillars 314 shifted at any desired spacing of spaces G between the nano-pillars 314, a desired pitch λ of the pillars, a desired row-to-row shift δ, and a desired array angle α (also referred to as the critical angle α) (as shown in FIG. 1 ).Multiple nano-pillar arrays 320 (e.g., 1-N, where N is the last number of nano-pillar arrays 320) may be fabricated on the same substrate 302, as discussed with reference to FIGS. 3A and 3B. The first nanocolumn arrays 320 may have a first set of parameters (a desired spacing of the spaces G between the nanocolumns 314, a desired pitch λ of the columns, a desired row-to-row shift δ, and a desired array angle α). The second nanocolumn arrays 320 may have a second set of parameters (a desired spacing of the spaces G between the nanocolumns 314, a desired pitch λ of the columns, a desired row-to-row shift δ, and a desired array angle α), wherein one or more parameters of the first set of parameters may be different from that of the second set of parameters. The third nano-pillar arrays 320 may have a third set of parameters (a desired spacing of the gaps G between the nano-pillars 314, a desired pitch λ of the pillars, a desired row-to-row shift δ, and a desired array angle α), wherein one or more parameters of the first set of parameters may be different from some parameters of the second set of parameters and / or may be the same or the same as some parameters of the second set of parameters and one or more parameters of the third set of parameters may be different from some parameters of the first and second sets of parameters and / or may be the same or the same as some parameters of the first and second sets of parameters. The same analogy may apply up to the last (N) arrangements 320 of nano-pillars, which may have a last (N) set of parameters (a desired spacing of the spaces G between the nano-pillars 314, a desired pitch λ of the pillars, a desired row-to-row shift δ, and a desired arrangement angle α), wherein one or more parameters of the last set of parameters may be different from any parameter of the first, second, third, and N-1th sets of parameters and / or may be the same or the same parameters as any parameter of the first, second, third, and N-1th sets of parameters.To define the pillars and spaces, a nano-scale negative lithography technique may be better to ensure a patterned space dimension of less than (<)100 nm, e.g., to begin with the pillars and spaces defined in the resist pattern 308, as shown in process flow 305. Electron beam lithography is an option where patterns of pillar arrays are smaller. However, the more production-oriented approach of nanoimprint lithography as well as lithography in the extreme ultra violet (EUV) range and in the deep ultra violet (DUV) range under conditions with a well-controlled dose can also be used. In order to obtain pillars with a high aspect ratio, the written pattern (i.e., the resist pattern 308) needs to be transferred onto the hard mask 304 (the hard mask pattern 312) before etching the (Si) substrate 302. Columns with a high aspect ratio allow for higher fluid throughput and may reduce problems of clogging associated with micro / nano fluid elements. Columns with a high aspect ratio are therefore a useful feature as long as the gap dimension between adjacent columns can be maintained. By first defining the pillars in the resist pattern 308 and transferring them to the etched hard mask 312, the advantage of etch selectivity increases aspect ratio while maintaining a more uniform gap dimension when performing the etching of the array (320) of pillars.Some experimental data are discussed below as exemplary implementations. The experimental data is illustrative and not restrictive. In this case, electron beam lithography has been used to define the dimensions of the pillars (e.g., the resist pattern 308) in hydrogen silsesquioxane (HSQ) as part of a two-layer resist stack (e.g., resist 306), which is then transferred to a 150 nm low-temperature oxide (LTO) hard mask (e.g., the etched hard mask pattern 312). Hard masks of a densified LTO, a thermal oxide, and / or SiO 2 / SiN / SiO 2 may also be contemplated. The experiment then used a Si etch process based on RIE to define the pillars (e.g., pillars 314) in the substrate. Further details of the RIE process will now be described.Details of RIE process: Dry etching was performed in an Applied Materials DPSII ICP etching chamber for pattern transfer to fabricate Si columns of 400 nm height from the electron beam resist pattern. First, the developed negative electron beam resist (HSQ) is used to etch through an organic planarization layer (OPL) mask using an N 2 / O 2 / Ar / C 2 H 4- chemism at a source power of 400 watts (W), a bias power of 100 W, and a pressure of 4 millitorr (mTorr) at 65° C. Then, the pattern is further transferred into a hard mask of SiO 2 using a CF 4 / CHF 3- Chemism at a source power of 500 W, a bias power of 100 W and a pressure of 30 mTorr at 65° C. The hard mask of carbon is then stripped using an O 2 / N 2- Chemism in an Applied Materials Axiom downstream incinerator at 250°C. The Si pillars are etched to a depth of 400 nm using the hard mask of SiO 2 using the DPS II first through a breakdown step with CF 4 / C 2 H 4 and then a main etch with Cl 2 / HBr / CF 4 / He / O 2 / C 2 H 4 at 650 W source power, 85 W bias power, and 4 mTorr pressure at 65° C. Note that three masks were used to finally etch the pillars, and the three masks were the developed HSQ electron beam resist (mask), the OPL mask, and the SiO 2. hard mask.Gap AnalysisFigures 4A, 4B, 4C and 4D are scanning electron micrographs of the result of this RIE process for two separate cases. FIGS. 4A and 4B illustrate the pillars (e.g., pillars 314) before the hard mask (e.g., hard mask pattern 312) is removed (such as in process flow 309), and the top ends of the pillars (pillars 314 with hard mask pattern 312 on the top end) have a rounded shape. The hard mask of a 150 nm LTO (not densified) was used in FIGS. 4A and 4B along with a RIE etch to create the pillars 314. FIGS. 4C and 4D illustrate pillars (e.g., pillars 314) after removal of the hard mask (e.g., hard mask pattern 312) using dilute hydrofluoric acid performed on different wafers, and the top ends of pillars 314 in FIGS. 4C and 4D are flat. In both cases, the Si pillars flex inward in the center due to the high density of the pillars in the array. That is, the gaps between the pillars broaden at the center of the pillars 314 because the diameter of the pillars is reduced at the center. The pillars have an inwardly bent shape or an hourglass shape. It should be noted that columns at the boundaries of the array are very vertical (not shown). This emphasizes the problem of gap non-uniformity on the order of nanometers, with gap dimensions of about (~) 100 nm having a variation in gap from the top of one column to the bottom of the same column (i.e., depth or height of one column) of about 50 nm, as can be seen from Figures 4C and 4D. The close proximity of the pillars in the array as defined by the gap caused the pillars to bow inward midway, creating a variation in the gaps that prevents further scaling. This effect was observed for the etching process described above (i.e., prior to disposing the oxide layer 316) at gap dimensions having dimensions of 250 nm and below.FIGS. 5A and 5B are scanning electron micrographs of the fabricated array of nanocolumns of a wafer 5 without a thermal oxide having a thickness of 50 nm. FIGS. 5C, 5D, and 5E are scanning electron micrographs of a wafer 7 showing the influence of growing a thermal oxide having a thickness of 50 nm (e.g., oxide layer 316) on nano-pillar arrays embedded in Si. On the side of the columns there is a right wall 505 (shown in Figures 5A and 5C), a bottom 510 of the substrate and a left wall 515 (shown in Figure 5B).The processing of the columns in FIGS. 5A and 5B for wafer 5 is identical to the processing of the columns on wafer 7 in FIGS. 5C, 5D, and 5E, except for the last oxidation step (which is performed only on wafer 7 in FIGS. 5C to 5E ). In the case of Fig. 5B (wafer 5), there is a variation of 26 nm for the gap dimension of about 186 nm, while Fig. 5D (wafer 7) after oxidation with the narrowing of the gap dimension to about 138 nm shows only a variation of 13 nm in the gap dimension in this case. This curative effect of oxidation occurs as a result of the non-uniformity of the oxide on these non-planar structures (i.e., the pillars), as shown in Figure 5E. FIG. 5E shows that the gap dimension relative to two pillars (from a side-by-side perspective in the direction of the x-axis) between these two pillars may vary from top to bottom (i.e., along the vertical axis of the y-axis) only by 13 nm, as the oxide has been filled into the inwardly curved shape. When the etching process applied to FIGS. 5A and 5B is used (wafer 5), it is determined that uneven oxidation on the pillar members "heals" the variation in the gaps as shown in FIGS. 5C, 5D, 5E (wafer 7) because the oxidation proceeds more rapidly in the middle of the pillars (rather than at the top and bottom), and this is further shown in FIGS. 10A and 10B.FIGS. 6A and 6B (wafer 5) illustrate starting with a smaller dimension of the gap, such as 80-89 nm (variation around 9 nm), in which no oxide is placed to fill the hourglass shape. FIGS. 6C and 6D (wafer 7) illustrate the 50 nm oxidation step employed when the original gap dimension is 80 to 89 nm (variation by 9 nm). The effect of the oxidation is very evident in Figures 6C and 6D, where the same 50 nm oxidation step (discussed above for Figures 5C, 5D, 5E) reduces the gap dimension from 80 to 89 nm to just 21 to 25 nm (variation of the gaps of 4 nm) with a ratio of 12:1 (depth:gap). As can be seen from Figures 6C and 6D, oxidation at lesser starting gap dimensions (such as 80-89 nm (or less) prior to the oxidation step for gap narrowing and inward deflection elimination) yields gaps of about 25 nm with a variation of only a few nanometers (4 nm) over an etch depth of about 300 nm, with the depth to gap ratio of 300:25 resulting in the ratio of 12:1. This small value for the variation of the spaces (e.g., 4 nm) and this process offer the possibility of producing custom tunable space dimensions, especially when these nanocolumns are combined with chemical modification processes. The term high aspect ratio may refer to structures with a depth to space ratio of greater than 4:1, and may be difficult to achieve in an on-line process of this order of magnitude.When the oxide is disposed on the array of pillars as discussed herein, embodiments are configured to provide an array of pillars having a pitch dimension that is uniform along the vertical axis (i.e., depth) of two pillars that are adjacent to one another (e.g., the pitch dimension between the two adjacent pillars varies by less than 5 nm (such as 4 nm, 3 nm, 2 nm)). For example, FIGS. 10A and 10B are cross-sectional views illustrating the healing process that eliminates (reduces) the variation of the gaps and creates a uniform gap dimension in the array 320 of pillars according to an embodiment. For purposes of illustration only, two columns 314 are shown side-by-side, but the illustration applies to all of the columns 314 in the array 320 of columns. The height of the pillars 314 is shown on the y-axis and the width / diameter is shown on the x-axis. The z-axis represents the length of the array 320, and further columns 314 (not shown) in the array are positioned in front of and behind the two columns 314. FIG. 10A shows two example pillars 314 made of their substrate material (substrate 302). The pillars 314 are bent inward to have the shape of an hourglass. In FIG. 10A, two gap dimensions G 1 and G 2 are shown, however, additional gap dimensions may be present between the gap dimensions G 1 and G 2. The gap dimension G 1 is at (near) the top and bottom ends of the pillars 314. The gap dimension G2 is located at (near) the center of the pillars 314. The close proximity of the pillars 314 as defined by the gap dimension G 1 in the array 320 may cause the hourglass shape due to dimensional constraints on the gap dimension G 1 imposed on the impinging flow of reactive ions during the RIE process.FIG. 10B shows the two example pillars 314 after disposing the oxide layer 316. Due to the non-planar architecture, the nano-size of the pillars, and the close nano-spacing between the pillars 314 in the array 320 of pillars, the oxide layer 316 does not uniformly spread on the pillars 314. Instead, more oxide 316 is formed in the center (voids) of the pillars 314 more quickly than at the upper end and the lower end of the pillars 314 in the y-axis direction. In other words, the inwardly bent centers are filled at a faster rate than the top ends and bottom ends of the pillars 314. This uneven distribution of the oxide 316 formed on the pillars 314 serves to straighten the individual pillars 314 while changing from the shape of an hourglass to a cylinder-like shape, which in turn makes the gap dimension G 5 between the two pillars 314 (and any other two pillars 314 that are adjacent to each other in the direction of the x-axis) uniform. Accordingly, all of the spaces G (where G represents the overall space dimension of the array) are uniform throughout the array 320 of columns.Chemical ModificationInteraction between the particles to be sorted and the surfaces of the array can be tailored using chemical modification. Generally, it may be associated with the attachment and / or attachment of molecules to the surfaces of the array of columns by physical adsorption and / or formation of chemical bonds. Additionally, chemical modification of the array of pillars may include attachment of a layer(s) of a material such as a metal, polymer, and / or ceramic coating, as well as changes in the oxidation state of the array surface. Surfaces (for chemical modification) may include the surfaces of the sorting columns, the walls, the ceiling, and / or the bottoms of the fluid column assembly. Moreover, the chemical modification may occur on any surfaces present in the inlets, outlets, drive mechanisms, and / or fluid channels attached to the nano fluid unit (e.g., the one or more arrays of columns).Although chemical modification may be employed as discussed above, the better application is chemical modification of the sorting columns themselves, since this allows for the design of the interactions between the particles and the surfaces of the sorting assembly.In one example, a small organic molecule or polymer referred to as a ligand may be chemically attached to the surface of the columns, such as by condensation of chlorosilane and / or alkoxysilanes on the native silica of the columns, as shown in FIG. 7A. In addition, the ligand can be chemically attached to the surface of the columns, such as by thiols, amines, and / or phosphines on columns coated with a thin layer (e.g., 10 nm) of gold or silver, as shown in FIG. 7B. The resulting layer of ligand molecules has a thickness of a single molecule, i.e., a monomolecular layer. The end groups of the monomolecular layer that are in direct contact with the fluid and particles determine the physochemical interactions that the particles experience as they pass through the array. Changing the end group of the ligand therefore allows for customising surface interactions within the array.FIGS. 7A and 7B illustrate the general chemical scheme of methods for chemical modification of sorting array surfaces, according to an embodiment. Referring to FIG. 7A, for a generic substrate (a column of array), reactive sites (X) on the surface may be used to form chemical bonds and / or physical absorption of ligands with small molecules. The attachment of ligands to the surface forms a new layer having a thickness of a single molecule (i.e., the monomolecular layer). A general ligand consists of (i) a bonding group (Z) interacting with the reactive site (X) of the substrate, (ii) a backbone consisting of a number of spacer molecules (n) that determine largely the thickness of the final monomolecular layer, and (iii) an end group (A) interacting with the interface between the monomolecular layer and the fluid / particles in the array. The end group (A) interacts with the particles to be sorted. Although FIG. 7A shows the bonding group Z and the reactive site X, this is only an example, and the chemical modification is not intended to be limited to the one type of reaction mechanism in this example. There are two other general mechanistic possibilities: (1) the formation of a direct bond, i.e., the Z group binds to the reactive site X in a Z-X bond, and / or (2) the formation of a bond with a deletion, i.e., the reactive group Z-W binds to the reactive site X-V in a Z-X bond, thereby deleting the byproducts W, V. For example, the reaction of chlorosilanes R-Si-Cl with a silanol on the silica surface, H-O-Si, forms the R-Si-O-Si bond, thereby removing the HCl.Referring to FIG. 7B, monomolecular layers may be formed on metal layers (M) previously deposited on the array of columns. For example, one or more metal layers (M) may be deposited on the pillars (e.g., after the oxidation process that produces the uniform gap dimension) such that the pillars now have a metal surface (M) over the substrate (and / or over the oxide layer that has been filled inward into the bend). In Figure 7B, the linking group is labeled 'Q' as opposed to 'Z' in Figure 7A. Ligands (e.g., having the bonding group (Q)) can form coordination complexes directly with the metal surface (M) of the array of columns, forming a densely packed monomolecular layer.A chemical modification can be used to tune the array of columns to sort smaller particles by reducing the gap dimension, as shown in FIGS. 8A and 8D. The surfaces of the sorting columns 314 may be modified with molecules of varying lengths that include aliphatic or aromatic oligomers / polymers that effectively increase the thickness of the columns and thereby decrease the inter-compartment distance therebetween. By selecting longer ligands, the gap dimension can be decreased, thereby decreasing the effective boundary particle dimension (i.e., smaller particles can be sorted). The backbone of the ligand can be selected to provide a range of mechanical properties between a rigid, densely packed molecular layer and / or a flexible, disordered layer. Ligands may include small organic molecules, proteins, peptides, nucleic acids, oligosaccharides, and / or synthetic polymers. In one example, the surfaces of the columns are modified with oligomers of polyethylene glycol (PEG) by siloxane compounds. With about 0.36 nm per ethylene oxide residue for a 12 residue PEG oligomer, this produces a gap dimension reduction of about 9 nm, for a 20 residue PEG oligomer, this means a gap dimension reduction of about 14 nm.FIGS. 8A to 8D illustrate schemes of chemical modification of sorting assemblies as a means for modifying the inter-column gap dimension according to an embodiment. Referring to FIG. 8A, for pillars 314 with their native oxide, a grown oxide layer, and / or a deposited layer of an alternative material, e.g., a metal, a ceramic, a polymer, reactive sites (X) are present on the surface of the pillars. The pillars 314 have an initial gap width, denoted g. There is a bottom 805 of the array (which is the bottom of the substrate 302 on which the columns stand). FIG. 8C shows an enlarged view 820 showing the empty reactive site (X) in FIG. 8A. In view 820, the reactive site (X) is not attached to a ligand, but the ligand is to be applied to the array 320 of columns as shown in Figure 8B.Referring to Figure 8B, chemical attachment of the ligand 810 to the surfaces of the columns forms a monomolecular layer 815 having a thickness determined by the characteristics of the ligand package. The added thickness of the monomolecular layer 815 reduces the width of the gap (from an initial width g of the gap) to a new effective width (g e) of the gap. Adjusting the structure of the ligand, particularly the backbone, as well as the packing and defect density of monomolecular layer 815 may tailor the thickness of monomolecular layer 815 and thus tailor the effective gap (g e). The effective gap (g e) is the new physical gap dimension experienced by the particles as they pass through the array 320 and is formed from the combination of the physical barrier of the pillars plus the added steric barrier of the monomolecular layer. The effective clearance is generally an approximate value depending on the structural, mechanical and dynamic properties of the monomolecular layer under the operating conditions of sorting the particles. FIG. 8D shows an enlarged view 820 in which the reactive sites (X) have been attached to the ligands 810, thereby increasing the diameter of the columns 314.Further improvement and development of the sorting assembly may be introduced by the end group(s) (A) of the ligands, which may be chosen to have specific interactions with the fluid and / or particles to be sorted, as shown in the schematic representation of Figure 9A. As the particles pass through the array 320 of columns, interactions with the end groups of monomolecular layer 915 result in increased adhesion and temporary retention on the column walls of columns 314. These interactions slow the flow of particles as well as causing the particles to be more positioned on average on the walls of the columns, thereby reducing the size of the flow field that scans it. Because the pitch of the array is asymmetric with respect to the mean fluid flow, particles (such as particles 910) that traverse through and are retained between the columns 314 are effectively moved along the critical angle of the array and are sorted out. In one example, thiol end groups at the end of PEG-type ligands can be used for disulfide bonds formed between traversing particles, such as proteins or other molecules to which thiols are attached. In combination with a suitable catalyst agent in the fluid, as proteins (such as particles 910) pass through assembly 320, these may form disulfide bonds with columns 314, with their flow being temporarily halted. In another example, small segments of a chemically stable nucleic acid, such as a peptide nucleic acid (PNA), can be attached to the walls of the columns to selectively slow down and sort out DNA or RNA analytes by reversible base pairing. In another example, patches of hydrophobic ligands embedded in hydrophilic monomolecular layers can be placed on the columns, such a pair consisting of aliphatic hydrocarbon ligands and PEG. The hydrophobic patches can be used to interact with hydrophobic domains on proteins to selectively sort them from solution.FIG. 9A is a diagram showing a flow of particles in a chemically modified sorting arrangement including particles 905 having no affinity for the monomolecular surface layer 915 and particles 910 interacting with the monomolecular layer 915. The particles 905 without affinity follow the flow lines through the array 320 (i.e., show a zig zag mode) and are not subject to any strong interactions with the columns 314. The trajectory of these particles 905 is on average unaffected and they flow without sorting in the array 320. Particles 905 flow into an outlet 940, for example. However, the particles 910 having a physochemical affinity caused by molecules on the surface thereof undergo interactions with the molecules of the monomolecular layer 915 on the surface of the pillars 314. The interactions may temporarily bind these particles 910 to the surface of the pillars 314 and cause the particles 910 to remain on average closer to the pillar walls of the pillars 314. By multiple sequential binding and dissociation events, the particles 910 are transferred along the direction of the columns 314 (i.e., they show a bump mode in the direction of the critical angle α) and are sorted by the array 320 due to chemical affinity. The particles 910 are sorted into an outlet 945. FIG. 9B is an enlarged view of a cross section of the nano column 314, the monomolecular layer 915, and the particle 910 having an affinity according to an embodiment.To chemically modify the array 320 of pillars, the ligand may be introduced by chemical vapor deposition (CVD) and / or by wet chemistry. To attach the metal, CVD, sputtering, and / or wet chemistry may be used. Provided below are two detailed examples of chemically modifying the pillars 314 by adding a monomolecular layer discussed for purposes of illustration and not limitation, as well as the two examples using wet chemistry.For purposes of illustration, an exemplary method for modifying a microfluidic device using poly(ethylene oxide)(PEG) ligand modifiers is provided below. All glass to be exposed to chlorosilanes is first cleaned in an isopropanol bath saturated with potassium hydroxide for at least 24 hours, then thoroughly rinsed with deionized water, and dried in an oven at 140° C. for 12 hours.A 100 ml round bottom flask is removed from the oven at 140°C and rapidly sealed with a septum. Nitrogen gas purging is established using needles through the septum and the flask is allowed to purge for 10 minutes. 30 ml of anhydrous toluene are transferred to the flask via a cannula. Via syringe, 600 μl of n-octyldecyltrichlorosilane is injected to form a 49 millimolar solution. The flask is gently agitated to mix the reagents homogeneously. This forms the passivation solution. A 500 ml reaction vessel and a 3 necked head are removed from the 140°C furnace and then rapidly sealed together with each inlet sealed with a septum. Nitrogen gas purging is established using needles through the septum and the flask is allowed to purge for 10 minutes. Via a cannula, 20 ml of the passivation solution in the 100 ml flask is transferred to the reaction vessel.The reaction vessel is gently shaken to thoroughly sweep the passivation solution around the walls of the reaction vessel. The same is done using the remaining 100 ml passivation solution for the flask. This gentle shaking is repeated for 1 hr every 10 to 15 minutes. Shaking allows the glass to stand at ambient temperature. This procedure serves to passivated the walls of the glass against further silizanization. Thereafter, the passivation solution is discharged from the reaction vessel, and the reaction vessel is sequentially cleaned 3X each with toluene, acetone, isopropanol, and deionized water. The same is done for the flask at 100 ml. Then, the glass is returned to the oven at 140°C and allowed to dry for 12 to 14 hours.The 100 ml round bottom flask is removed from the 140°C oven and rapidly sealed with a septum. Nitrogen gas purging is established using needles through the septum and the flask is allowed to purge for 30 min. 100 ml of anhydrous toluene are transferred to the flask via a cannula. Via syringe, 100 μl of 2-(methoxypoly(ethyleneoxy) 6-9 propyl)dimethylchlorosilane is injected to form an approximately 2 millimolar solution. The flask is gently agitated to mix the reagents homogeneously. This modification solution is used within the day of its preparation.Silica / silicon based microfluidic devices (chips) are cleaned for 30 min in an oxygen plasma to remove organic surface contamination. The chips are then transferred to a 0.1 molar aqueous nitric acid solution for 10 min to hydrolyze any surface siloxane bonds with silanols. Next, the chips are cleaned sequentially using a spray bottle jet in deionized water, acetone, ethanol, and then isopropanol. The chip is then placed face up on a fresh Texwipe and immediately dried under a stream of nitrogen gas forcing the solvent out the center of the chip. Thereafter, the chips are placed on a custom glass holder (which places the chips horizontally / face up inside the reaction vessel, as described below).A 500 ml reaction vessel and a 3 necked head are removed from the 140°C furnace. A stirring bead including the glass holder and the chips are placed in the reaction vessel and then rapidly sealed together, each inlet being sealed with a septum. Nitrogen gas purging is established using needles through the septum and the reaction vessel is allowed to purge for 30 min.The modifying solution (with the ligand) is transferred via cannula into the reaction flask until the solution level is above the chips. A positive nitrogen pressure is then maintained using a stirrer. The reaction is allowed to run for 2 hours with stirring at ambient temperature. The reaction vessel is then opened and the chips are cleaned (sequentially) by sequential rinsing using a spray bottle jet, toluene, acetone, isopropanol, then deionized water. The chip is then placed face up on a fresh Texwipe and immediately dried using a stream of nitrogen gas forcing the solvent out the center of the chip. The chips are then placed in a glass holding vessel with a septum. Nitrogen gas purge is established using needles through the septum and the chips are allowed to dry overnight (about 12 to 14 hours).The use of the sub-headers is now terminated. FIG. 11 illustrates a chip 1100 (a fluid unit) including the array 320 of pillars according to an embodiment. The chip 1100 has an inlet 1105 to receive the fluid containing the particles of different dimensions (i.e., biological entities) to be sorted. The inlet 1105 may be an opening or hole in the walls around the nanocolumn array 320, or it may span the width of the nanocolumn array 320 through which the fluid (e.g., water, electrolyte solutions, organic solvents, etc.) and the particles (e.g., the biological entities) may flow. The particles having a dimension greater than the critical dimension flow in the critical angle direction (i.e., bump mode) through the array 320 of pillars, and those particles having a dimension greater than the critical dimension are collected at an outlet 940. The critical dimension is the dimension (e.g., diameter) of a round shaped particle and / or persistence length of a chain structure, such as DNA, that is too long to zig-zag traverse the nano-array 320. Particles having a dimension less than the critical dimension pass through the array 320 of columns in the direction of fluid flow in a zig zag (i.e., zig zag mode) and these smaller particles are collected at the outlet 945. The particles having the dimension less than the critical dimension follow the direction of fluid flow and are sorted through the outlet 945. In one case, the pillars 314 may include the chemical modification as discussed herein, which may further reduce the gap dimension and / or sort particles having an affinity for the chemical modification. The outlets 940 and 945 may be openings through which the sorted particles can flow and, after sorting, can be collected into containers.FIG. 12 is a method 1200 of providing a fluid device 1100 (e.g., a chip 1100) provided according to one embodiment. Reference may be made to Figures 1-11 discussed above. At block 1205, the inlet 1105 is configured to receive a fluid. At block 1210, the outlet (e.g., outlets 940, 945) is configured to discharge the fluid. The nanocolumn array 320 is coupled to the inlet and the outlet, and the nanocolumn array 320 is configured to allow the fluid to flow from the inlet to the outlet at block 1215.At block 1220, the array 320 of nanocolumns includes nanocolumns 314 that are arranged to separate biological entities (particles) by size. At block 1225, the nano-pillars 314 are arranged to have a gap G separating one nano-pillar 314 from another nano-pillar 314, and the gap is configured to be in a nano-range (e.g., sub-100 nm).The one nano-column is located on the side of the other nano-column such that the space G is interposed therebetween. The gap between the one nano-column and the other nano-column is uniform along a vertical axis of the one nano-column and the other nano-column (such as the gap G 5 as shown in FIG. 10B ).The nanocolumn array includes an oxide layer 316 disposed on the nanocolumns, and the oxide layer 316 causes the gap to be uniform along a vertical axis of the one nanocolumn and the other nanocolumn (e.g., the gap G 5 in FIG. 10B is uniform up and down in the space between the two nanocolumns 314).The oxide layer 316 causes a dimension of the gap (e.g., the gap G 5) to be only about 20 nanometers, with the gap remaining uniform along the vertical axis (e.g., the y-axis in FIG. 10B ). The oxide layer 316 causes unevenness of a diameter (e.g., the diameter of the pillar 314 in FIG. 10A is not uniform) of the nano-pillars in FIG. 10B to become uniform, resulting in the gap being uniform along the vertical axis of the one nano-pillar and the other nano-pillar. Increasing a thickness of the oxide layer 316 causes a reduction in a dimension of the gap.In one case, the dimension of the clearance is in a range of 20 to 300 nm. In another case, the dimension of the gap may be formed to be less than 100 nm, may be less than 80 nm, may be less than 60 nm, may be less than 40 nm, may be less than 30 nm, may be less than 25 nm, etc. For example, particles of 100 nm may be sorted / separated according to a desired dimension of the particles to be separated.A monomolecular layer (e.g., monomolecular layer in FIGS. 7A, 7B, monomolecular layer 815 in FIG. 8B, and / or monomolecular layer 915 in FIG. 9A ) is attached to nano columns 314 to reduce a dimension of the gap. The gap with a reduced dimension is configured so that smaller entities are separated compared to when the monomolecular layer is not attached to the nanocolumns.FIG. 13 is a method 1300 of forming an array 320 of nano-pillars, according to an embodiment. Reference may be made to Figures 1 to 12.At block 1305, the hard mask layer 304 is disposed on the substrate 302. At block 1310, the resist layer 306 is patterned into a pattern (to the resist pattern 308) of the nanocolumn array 320 where the resist layer 306 has been disposed on the hard mask layer 304.At block 1315, the resist layer (resist pattern 308) is used to pattern the hard mask layer 304 to the pattern (hard mask pattern 312) of the nano-pillar array 320 such that both the resist layer and the hard mask layer have the pattern of the nano-pillar array 320.At block 1320, the substrate 302 is patterned into the structure of the nano-pillar array 320 such that the nano-pillar array 320 is now formed, with the resist layer (the resist structure 308) and the hard mask layer (the hard mask structure 312) removed, and with the nano-pillars 314 in the nano-pillar array having a first gap dimension relative to each other in a side-by-side arrangement relationship (e.g., the gap dimension G 1 and / or G 2 in FIG. 10A ). At block 1325, the first gap dimension is reduced to a second gap dimension (e.g., the gap dimension G 5) by disposing the oxide layer 316 on the nano-pillar array 320.The resist layer is patterned by at least one of electron beam lithography and / or nanoimprint lithography or other form of lithography to the pattern(s) of the array 320 of nano-pillars.Using the resist layer to pattern the hard mask layer to the nano-pillar array structure includes performing a reactive ion etching operation to etch the hard mask to the pattern (hard mask pattern 312) of the nano-pillar array 320.Patterning the substrate 302 into the nanocolumn array structure to form the nanocolumn array includes performing a reactive ion etching process to etch the substrate into the nanocolumn array 320.Reducing the first gap dimension (e.g., gap dimension G 1 and G 2) to the second gap dimension (gap dimension G 5) by disposing oxide layer 316 on nano-pillar array 320 includes reducing the second gap dimension (e.g., to less than 300 nanometers, to less than 100 nanometers, etc.).Reducing the first gap dimension to the second gap dimension by disposing the oxide layer on the nano-pillar array causes each of the nano-pillars to have a uniform shape, and causes the second gap dimension to be uniform throughout the nano-pillar array for the side-by-side arrangement relationship of the nano-pillars (as shown in FIGS. 10A and 10B ). Before reducing the first pitch dimension to the second pitch dimension by disposing the oxide layer, the nano-pillars in a center of the nano-pillars have a shape bent inward in an order of nanometers. By reducing the first space dimension to the second space dimension by disposing the oxide layer, both the inwardly bent shape is filled in the middle and the nano-pillars are straightened into a cylinder-like shape.Deposition is any process that grows a material on, coats, or otherwise transfers a material to the wafer. Available technologies include, but are not limited to, thermal oxidation, physical vapor deposition (PVD), chemical vapor deposition (CVD), electrochemical deposition (ECD), molecular beam epitaxy (MBE), and more recently atomic layer deposition (ALD), among others.Removal is any process that removes a material from the wafer: examples include etching processes (either wet or dry), chemical mechanical planarization (CMP), etc.Patterning is the forming or altering of deposited materials, and is generally referred to as lithography. In conventional lithography, for example, the wafer is coated with a chemical referred to as a photoresist; then, an apparatus referred to as a stepper focuses, aligns, and moves a mask, exposes selected areas of the wafer underneath with short wavelength light; the exposed areas are removed using a developer solution. After etching or other processing, the remaining photoresist is removed. Patterning also includes electron beam lithography, nanoimprint lithography, and reactive ion etching.The flowchart and block diagrams in the figures depict the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or range of instructions that includes one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur in a different order than noted in the figures. For example, depending on the functionality involved, two blocks shown in succession may actually be executed substantially simultaneously, or the blocks may sometimes be executed in the reverse order. It should also be noted that each block of the block diagrams and / or flowchart illustration, as well as combinations of blocks in the block diagrams and / or flowchart illustration, may be implemented by special purpose hardware-based systems that perform the specified functions or acts or execute combinations of special purpose hardware and computer instructions.

Claims

An apparatus (1100) comprising: an inlet (1105) configured to receive a fluid; an outlet (940, 945) configured to discharge the fluid; and an array (320) of nano-pillars (314) connected to the inlet (1105) and the outlet (940, 945), the array (320) of nano-pillars (314) configured to allow the fluid to flow from the inlet (1105) to the outlet (940, 945); wherein the array (320) of nano-pillars (314) includes nano-pillars (314) arranged to separate entities by size; wherein the nano-pillars (314) have an inwardly bent shape on the order of nanometers at a center of the nano-pillars (314) forming a unevenness of a diameter, wherein the array (320) of nano-pillars (314) has an oxide layer (316) disposed on the nano-pillars (314), wherein the oxide layer (316) fills the inwardly bent shape at the center and straightens the nano-pillars (314) with the oxide layer (316) disposed thereon into a cylinder-like shape, wherein the nano-pillars (314) are disposed to have a gap (G5) separating a nano-pillar (314) with the oxide layer (316) disposed thereon from another nano-pillar (314) with the oxide layer (316) disposed thereon; and wherein the clearance (G5) is designed to be on the order of nanometers.The apparatus of claim 1, wherein the one nano-column (314) is located on a side of the other nano-column (314) such that the space (G5) is interposed therebetween.The device of claim 1, wherein the oxide layer (316) causes a dimension of the gap (G5) to be uniform along a vertical axis of the one nano-pillar (314) and the other nano-pillar (314) and to vary between two adjacent nano-pillars (314) by less than 5 nm.The device of claim 3, wherein the oxide layer (316) causes a dimension of the gap (G5) to be only about 20 nanometers.The device of claim 3, wherein the oxide layer (316) causes a diameter of the nano-pillars (314) to vary by less than 5 nm.The apparatus of claim 3, wherein an increase in thickness of the oxide layer (316) causes a reduction in a dimension of the gap (G5).The apparatus of claim 1, wherein a dimension of the gap (G5) is less than 100 nanometers.The apparatus of claim 1, wherein a dimension of the gap (G5) is less than 300 nanometers.The device of claim 1, wherein a monomolecular layer (815) is attached to the nano-pillars (314) to reduce a dimension of the gap (G5); and wherein the gap (G5) having a reduced dimension (g e) is configured to separate smaller entities compared to when the monomolecular layer (815) is not attached to the nano-pillars (314).A method (1300) of forming an array (320) of nano-pillars (314), the method comprising: arranging (1305) a hard mask layer (304) on a substrate (302); patterning (1310) a resist layer (306) into a pattern (308) of the array (320) of nano-pillars (314), wherein the resist layer (306) has been arranged on the hard mask layer (304); using (1315) the resist layer (306) to pattern the hard mask layer (304) to the pattern (308) of the array (320) of nano-pillars (314) such that both the resist layer (306) and the hard mask layer (304) comprise the pattern (308) of the array (320) of nano-pillars (314); patterning (1320) the substrate (302) to the pattern (308) of the array (320) of nano-pillars (314) such that the array (320) of nano-pillars (314) is formed, wherein the resist layer (306) and the hard mask layer (304) are removed, and wherein nano-pillars (314) in the array (320) of nano-pillars (314) have a first gap dimension (G 1, G 2) relative to each other in a side-by-side array relationship; reducing (1325) the first gap dimension (G1, G2) to a second gap dimension (G5) by disposing an oxide layer (316) on the array (320) of nano-pillars (314), wherein the nano-pillars (314), before the first gap dimension (G1, G2) is reduced to the second gap dimension (G5) by disposing the oxide layer (316), have an inwardly curved shape on the order of nanometers at a center of the nano-pillars (314); wherein reducing the first gap dimension (G1, G2) to the second gap dimension (G5) by disposing the oxide layer (316) both fills the inwardly bent shape in the center and straightens the nano-pillars (314) to a cylindrical shape.The method of claim 10, wherein the resist layer (306) is patterned to the pattern (308) of the array (320) of nano-pillars (314) using at least one of electron beam lithography and nanoimprint lithography.The method of claim 10, wherein using the resist layer (306) to pattern the hard mask layer (304) to the structure (308) of the array (320) of nano-pillars (314) comprises performing a reactive ion etching operation to etch the hard mask layer (304) to the structure (308) of the array (320) of nano-pillars (314); and wherein patterning the substrate (302) to the structure (308) of the array (320) of nano-pillars (314) such that the array (320) of nano-pillars (314) is formed comprises performing a reactive ion etching operation to etch the substrate (302) to the array (320) of nano-pillars (314).The method of claim 10, wherein reducing the first gap dimension (G1, G2) to the second gap dimension (G5) by disposing the oxide layer (316) on the array (320) of nano-pillars (314) comprises reducing the second gap dimension (G5) to less than 100 nanometers.The method of claim 10, wherein reducing the first gap dimension (G1, G2) to the second gap dimension (G5) by disposing the oxide layer (316) on the array (320) of nano-pillars (314) causes each of the nano-pillars (314) to have a uniform shape, and causes the second gap dimension (G5) to be uniform throughout the array (320) of nano-pillars (314) for the side-by-side arrangement relationship of the nano-pillars (314) and to vary along the vertical axis of two nano-pillars (314) that are adjacent to each other by less than 5 nm.

Images