Microfluidic chips for scale-independent, continuous, end-to-end manufacturing of nanomaterials

Abstract

Microfluidic mixer apparatuses and methods having a high aspect ratio and specific dimensions may achieve nearly complete (e.g., 85% or more, 90% or more, etc.) mixing efficiency within microseconds relatively low pressure (e.g., 75 psi or less) and at relatively high flow rates (e.g., flow rate of between 100 ml / hour and greater than 2 L / hour), without clogging. In general, the methods and apparatuses described herein may include two or more inputs, an output, and a mixing channel extending between the inputs and the output that have a high aspect ratio and a defined curvature and cross-sectional dimensions configured to provide efficient mixing while avoiding clogging, allowing these mixers to operate for longer periods of time at higher efficiencies than conventional mixers.

Description

MICROFLUIDIC CHIPS FOR SCALE-INDEPENDENT, CONTINUOUS, END-TO- END MANUFACTURING OF NANOMATERIALSCLAIM OF PRIORITY

[0001] This patent application claims priority to U.S. provisional patent application no. 63 / 555,877, titled “MICROFLUIDIC CHIPS FOR SCALE-INDEPENDENT, CONTINUOUS, END-TO-END MANUFACTURING OF NANOMATERIALS,” filed February 20, 2024, which is herein incorporated by reference in its entirety.BACKGROUND

[0002] Microfluidics offers significant advantages for both research and industrial-scale manufacturing of nanoparticles and other nanomaterials, including those comprised of biomolecules, polymers, and inorganic components. Microfluidic-based mixing techniques may yield monodisperse products with precisely controlled critical quality attributes (CQAs). Maintaining control over CQAs is crucial for ensuring the effective and safe functionality of nanomaterials (e.g., vaccines). Moreover, the inherent channel dimensions of microfluidics, typically ranging from tens to hundreds of micrometers, enable the minimization of valuable reagent consumption to small volumes, a feature particularly valuable for formulation discovery and other small-scale manufacturing applications. Additionally, fabrication methods such as injection molding with plastics facilitate mass production and relatively affordable access to microfluidic chips.

[0003] The ability to apply microfluidics to all scales of biomedical nanomaterial manufacturing, coupled with next-generation good manufacturing practice (GMP) workflows, could significantly reduce development times and regulatory approval hurdles. However, several challenges have impeded the widespread application of microfluidics to such scale-independent, GMP-compliant, and end-to-end continuous manufacturing processes. Primarily, scaling up typical microfluidic operations often involves either increasing the flow rate in a given device, leading to high shear conditions and potential degradation of CQAs, or operating multiple devices concurrently. The latter becomes cumbersome beyond a few devices operated in parallel due to the proportional increase in required space and fluidic connections. Additionally, many microfluidic chips are operated in batch mode or for a predetermined duration due to issues such as product aggregation within the channels and degradation of CQAs during chip operation. Such reproducibility issues hinder GMP process qualification, and the inability to operate microfluidic chips continuously and at scale has hindered their integration with other continuous processes.Finally, microfluidic chips made from plastics are susceptible to leaching of plasticizers, monomers, and other additives into the final product, along with the retention of residual solvents within the device, particularly when non-aqueous solvents are used in the formulation process. Thus, novel microfluidic strategies and technologies are necessary to address these challenges for scalable, GMP-compliant, and continuous nanomaterial manufacturing. Described herein are methods and apparatuses that may address these needs.SUMMARY OF THE DISCLOSURE

[0004] Described herein are methods and apparatuses (e.g., devices and systems, including chips and microfluidic mixers) that may address the problems described above. In particular, described herein are apparatuses and methods configured to achieve greater than 90% mixing efficiency within less than a second (e.g., less than 900 milliseconds, less than 800 milliseconds, less than 700 milliseconds, less than 600 milliseconds, less than 500 milliseconds, less than 400 milliseconds, less than 300 milliseconds, less than 200 milliseconds, less than 100 milliseconds, less than 50 milliseconds, etc.) at 75 psi or less, and at high flow rates (e.g., a flow rate of between 100 ml / hour and greater than 2 L / hour), without clogging.

[0005] These microfluidics-based techniques may provide valuable tools in industry and academia for producing monodisperse nanomaterial formulations with precisely controlled critical quality attributes. Achieving desirable and reproducible nanomaterial critical quality attributes at arbitrary formulation throughputs and process durations may reduce time and costs associated with scaling up formulation processes, developing next generation end-to- end continuous workflows, and achieving regulatory approval. Current microfluidic approaches do not readily and simultaneously enable scale-independent, continuous, and end- to-end manufacturing of nanomaterials that is readily aligned with good manufacturing practices (GMP). To address these challenges, the methods and apparatuses described herein may generally include a microfluidic high aspect ratio (HAR), in which the aspect ratio of the continuous length of mixing channel is configured to be dimensioned so that multiple vertically arranged vertically-arranged vortexes (e.g., Dean vortices) are formed along the length of the channel to provide rapid and thorough mixing. The high aspect ratio may refer to the ratio of height of the channel (perpendicular to the plane of device) and width of the channel (e.g., in the plane of the device). As will be described herein, the dimensions of the channel, including the aspect ratio (height / width), and length, may enable the microfluidic mixers described herein to achieve specific operating targets and process requirements. Further, these methods and apparatuses may allow arbitrary scale-up. In addition, alsodescribed herein are methods and apparatuses for fabrication that allow the production of highly rectangular microfluidic channels with tunable aspect ratios. In general, these apparatuses may be part of a single chip that may facilitate the simultaneous operation of many microfluidic mixing units in parallel. Operating in the mixing regime, which may be referred to herein as a high aspect ratio (HAR) regime, may provide clog-free, continuous, and stable microfluidic operation. The robustness of the HAR strategy demonstrated herein through continuous operation for hours (e.g., 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, 8 hours, etc.) of continuous operation, through end-to-end operation with a downstream purification process, and with reproducible operation over time, over orders of magnitude of formulation throughputs, and across different manufacturing runs. The methods and apparatuses described herein link microfluidic mixer design to smaller-scale formulation and to larger-scale, continuous and end-to-end manufacturing that is typically found in commercial manufacturing settings.

[0006] The strategies described here, including HAR strategies, may provide a microfluidic formulation and scale-up roadmap that is compatible with GMP, continuous, and end-to-end nanomaterial manufacturing. Specifically, this work demonstrates a microfluidic fabrication and design strategy for specifying and tuning the formulation throughput, while maintaining the same or comparable footprint and inlet pressures, that is based on designing and precisely fabricating mixing channels with pre-determined aspect ratios. Extending this strategy with on-chip parallelization enables throughputs to be scaled from 100 mL / hr. to at least 15 L / hr. on a single microfluidic chip, while maintaining the CQAs of the formulated product. Furthermore, the microfluidic platform can maintain CQAs over extended operation, produce nanomaterials with reproducible CQAs, and be integrated with downstream processes, such as tangential flow filtration. Finally, further amenability to safe and aseptic manufacturing is provided because the microfluidic chips are constructed entirely from silica- based glass, without any organic leachable chemicals, in a cleanroom environment. These strategies thus include the collection of design, fabrication, and operational considerations described here.

[0007] For example, the microfluidic configurations, designs, and manufacturing techniques described herein may provide for scaling out the formulation of nanoparticles from small scales (<100 mL / hr) to large scales (>100 L / hr or more). These methods and apparatuses may allow a single microfluidic chip to be used at every desired formulation flow rate. These methods and apparatuses may provide flow conditions (e.g., average channel velocity and shear) and high mixing that are comparable at every desired formulation flow rate. When used to manufacture nanoparticles, these methods and apparatuses may providenanoparticles with consistent, high-quality CQAs at every desired formulation flow rate. In addition, these methods and apparatuses may provide similar input fluidic pressures and device footprint are required at every desired formulation flow rate. Finally, these methods and apparatuses may provide linear scalability in throughput. For a given layout of planar mixing channels, virtually any desired increase in formulation flow rate may be selected a priori and achieved by creating rectangular microchannels with the appropriately scaled aspect ratio (i.e. channel height) and degree of on-chip parallelization, as will be described in greater detail herein.

[0008] The methods and apparatuses may therefore provide rapid mixing of two or more fluids from low to high flow rates at similar input pressures and device footprint. These methods and apparatuses may also provide continuous, clog-free manufacturing of nanomaterials for extended periods of time, and integration of the microfluidic formulation chips in end-to-end manufacturing workflows. In addition, these methods and apparatuses are compatible with GMP workflows.

[0009] Also described herein are microfluidic-based nanoparticle formulation chips that may be designed based on external system requirements (e.g., to meet target flow rates and satisfy pressure constraints), contrasting with the conventional approach of designing an external system to meet the requirements of a microfluidic chip. Furthermore, these apparatuses may rapidly achieve high mixing (~ milliseconds) at each throughput and enable continuous nanoparticle formulation for extended times (> 1 hr.). These features may also enable the integration of microfluidics-based nanoparticle formulation in novel continuous, end-to-end manufacturing workflows where the fluidic integration of multiple, optimized unit operations - without an intermediate reservoir - requires balancing the constraints (e.g., flow rates, pressures, etc.) of each unit operation. One example of such a downstream purification process may include a single-pass tangential flow filtration (SPTFF), where the required input flow rate of an optimized process scales proportionally to the manufacturer-specific step-increases in available filtration sizes.

[0010] The mixers and design strategy described here could achieve these target flow rates while also linearly scaling the throughput (from lOOmL / hr to greater than 2L / hr), at similar fluidic pressures and device footprint, to meet the input requirements of each scale of SPTFF.

[0011] In general, the methods and apparatuses described herein may be configured to achieve efficient mixing of two or more fluids, in a small footprint, based on the formation of secondary Dean flows. These secondary Dean flows may result in multiple, vertically- arranged vortexes. The channel forming the mixing region may be a planar, curved channelwith inner and outer radii of curvature that, while constant in magnitude, periodically invert along the direction of flow. The channel length of the mixing region may be divided into stages, where the length of each stage may be between about eight- and 10-fold of the channel width. In some examples, the stages may be maintained at a constant curvature that is inverted in the next stage. For example, the inner radii of curvature may be between about 20 mm to 200 mm, the outer radii of curvature may be between about 40 mm to 400 mm, and the channel widths may be between about 20 mm to 200 mm.

[0012] In some examples the inner radii of curvature may be approximately equal to the channel width, while the outer radii of curvature may be approximately twice the channel width. The mixer may have any appropriate number of stages. For example, the mixer may include at least one stage (e.g., 2 or more stages, 3 or more stages, 4 or more stages, 5 or more stages, 6 or more stages, 7 or more stages, 8 or more stages, 9 or more stages, 10 or more stages, 11 or more stages, 12 or more stages, 13 or more stages, 14 or more stages, 15 or more stages, 16 or more stages, 17 or more stages, 18 or more stages, 19 or more stages, up to 20 or more stages, etc.).

[0013] Any number of fluid inputs may be received into the mixer. For examples, the mixer may receive two or more fluid streams. The inputs may be oriented from 0 to 90 degrees relative to the channel entrance.

[0014] Within the other constraints of the channel design described herein (e.g., high aspect ratio, etc.), the channels may have any appropriate height, width and length. For example, the channel heights may be increased, e.g., while maintaining the same channel width, including having a channel height that is between less than 40 mm to greater than 700 mm (e.g., between 400 mm to 800 mm, between 400 mm to 700 mm, etc.).

[0015] In general, the aspect ratio of the channel may be 0.4 or greater, e.g., 0.5 or greater, 0.6 or greater, 0.7 or greater, 0.8 or greater, 0.9 or greater, 1 or greater, 1.1 or greater, 1.2 or greater, etc.). In some examples, the channel may have an aspect ratios (channel height / channel width) of 0.7 or greater.

[0016] The microfluidic mixers described herein may be used to rapidly mix two or more fluids for the formulation of nanoparticles. For example, for a given 2D (“XY”) design of a planar microfluidic mixer, a minimum channel aspect ratio and mixing flow rate may be maintained to rapidly (~ milliseconds) achieve high mixing (> 90% mixing efficiency) at low-to-moderate fluidic pressures (< 75 psi).

[0017] In general, these methods and apparatuses may be configured to prevent clogging, allowing them to be used with high efficiency for long periods of time (hours). In general, these apparatuses (mixers) may be configured to have a rectangular cross-section that has anapproximately constant cross-sectional area and / or shape along the length from the two or more inputs to the output. The constant cross-sectional area and / or shape may be, e.g., constant within + / - 5%, 4%, 3%, 2%, 1%, 0.9%, 0.8%, 0.7%. 0.6%, 0.5%, 0.4%, 0.3%, 0.2%, 0.1%, etc. along the length of the channel(s) forming the mixer, e.g., from input(s) to output. In some variations the cross-sectional area and / or shape may be relatively uniform along the length of the channel(s), e.g., having minimal to no variation. In general, the work described herein has determined that local variation in cross-sectional area and / or shape (e.g., increases or decreases in cross-sectional area) may result in local changes in pressure, such as low pressure regions, that can correspond with solid-phase aggregation and clogging.

[0018] In additional, these apparatuses may be configured to provide rapid and high (e.g., >80%, >85%, >90%, etc.) mixing; when high mixing is not achieved rapidly, nanoparticle precursors may aggregate within the channels. This may cause device clogging, preventing long-term device operation, and may result in rapid increases in device pressure. Although high mixing can be achieved in a low aspect ratio channel by increasing the mixing flow rate, this will lead to high shear forces and high pressures, resulting in the degradation of CQAs and conditions that are difficult to maintain with typical pharmaceutical process pumps or integrate continuously with upstream or downstream processes. In contrast, the methods and apparatuses described herein may achieve mixing and formulation scale-out through the use of individual chips having mixers with channels having and high channel aspect ratio (e.g., from 0.4 to 4) paired with a corresponding linear increase in the formulation throughput. The mixing chips (at each scaled channel height) may be operated in the laminar flow regime, defined by Reynolds numbers of less than 2000.

[0019] For example, described herein are devices with 200 mm channel widths and a critical aspect ratio of 0.4 may provide a formulation throughput to be scaled out from a minimum flow rate of 200 mL / hr. In some examples devices with 80 mm channel widths and critical aspect ratios of 1.0 may provide a formulation throughput to be scaled out from a minimum flow rate of 150 mL / hr. In some examples devices with 40 mm channel widths and critical aspect ratios of 2.0 may provide a formulation throughput to be scaled out from a minimum flow rate of 100 mL / hr. Dean numbers of at least 100 may be achieved at the critical aspect ratio.

[0020] In general, these methods and apparatuses may enable the design, manufacture, and operation of a single microfluidic chip for producing nanomaterials with known CQAs at a linearly scalable formulation throughput and with similar fluidic pressures and device footprints by specifying an aspect ratio of the mixing design, including on-chip parallelization as needed.

[0021] In general, the mixers described herein may be formed of a silica-based substratum. The channels may be precisely controlled to have rectangular cross-sectional sides formed as curved regions, as described herein. For example, also described herein are methods of forming any of these mixers. In general, these methods may include forming these channels by deep reactive ion etching (DRIE) manufacturing process that may specify and create rectangular microfluidic channels of a set width and arbitrary depth (i.e., aspect ratio, where aspect ratio is defined as the ratio of the channel depth to channel width (d / w)). The manufacturing process described herein may produce devices, regardless of channel depth, by tuning only one manufacturing process parameter (DRIE) to create channels with the desired aspect ratio. This process may allow channel depths to be specified, with submicrometer resolution, from less than 10 mm to millimeter-scale values limited by the thickness of the silica-based substratum. This allows devices with the same “X-Y,” or “topview,” footprint of the channels regardless of user-specified increases in channel depth. This further enables the formulation throughput to be scaled out in two steps: for a given X-Y microfluidic mixing design, first, increase the channel aspect ratio while maintaining the same average flow velocity, then, second, incorporate on-chip parallelization as needed to achieve the desired throughput. The methods described here also enable on-chip parallelization that is independent of the desired channel aspect ratio.

[0022] For example, described herein are microfluidic mixer apparatuses. These apparatuses may be configured to achieve 90% or more mixing efficiency within less than 1 second at 75 psi or less at a flow rate of between 100 ml / hour and greater than 2 L / hour without clogging. For example, an apparatus may include: two or more inputs; an output; a channel extending between the two or more inputs and the output that is planar and curving, the channel having a rectangular cross-section transverse to a length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, further wherein the rectangular cross-section has an approximately constant cross-sectional shape and / or area along the length from the two or more inputs to the output.

[0023] The channel may include between 2 and 20 stages. The stage length of each stage may be between 8 and 10 times a width of the channel in the plane of the microfluidic mixer. The inner radius of curvature may be between 20 pm and 200 pm. The outer radius of curvature may be between 40 pm and 400 pm. The width of the channel (e.g., in the plane of the microfluidic mixer) may be between 20 pm and 200 pm. The inner radius of curvaturemay be approximately equal to the width of the channel in the plane of the microfluidic mixer. The outer radius of curvature may be approximately equal to twice a width of the channel (e.g., in the plane of the microfluidic mixer). Each of the two or more inputs may be between 0 and 90 degrees relative to an entrance into the channel. The height of the channel (e.g., perpendicular to the plane of the microfluidic mixer) may be between about 40 pm and 700 pm. The aspect ratio of the rectangular cross-section may be between about 0.4 and about 4 or greater. For example, the aspect ratio may be 0.7 or greater. The channel may be formed in a silica-based substrate.

[0024] In any of these apparatuses, the output may have a larger width than the channel.

[0025] For example, a microfluidic mixer apparatus configured to achieve 90% or more mixing efficiency within less than 1 second at 75 psi or less at a flow rate of between 100 ml / hour and greater than 2 L / hour without clogging may include: a planar substrate; two or more inputs formed in the planar substrate; an output formed in the planar substrate; a channel through the planar substrate and extending along a length between the two or more inputs and the output that is curving, the channel having a rectangular cross-section traverse to the length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, further wherein the rectangular cross-section has an approximately constant cross-sectional shape and / or area along the length of the channel from the two or more inputs to the output.

[0026] Also described herein are methods of mixing using any of these apparatuses.

[0027] For example a method of microfluidic mixing may include: inputting a first fluid into a first input of a microfluidic mixer apparatus; inputting a second fluid into a second input of the microfluidic mixer apparatus; passing the first and second fluids from the first and second inputs in a channel extending along a curving length to an output at a flow rate of between 100 ml / hour and greater than 2 L / hour, wherein the channel has a rectangular crosssection traverse to the length and an aspect ratio of greater than 0.4, and wherein the channel comprises one or more stages, further wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, so that multiple vertical vortexes are formed within the channel as the fluids are passed along the length of the channel to achieve 90% or more mixing efficiency of the first and second fluids within less than 1 second at 75 psi or less.

[0028] The rectangular cross-section of the channel may have an approximately constant cross-sectional shape and / or area along the length of the channel from the first and second inputs to the output, to prevent clogging.

[0029] Any appropriate materials may be mixed. In some cases the method may include forming nanoparticles encapsulating a polynucleotide. For example, the first fluid may comprise a nanoparticle. The second fluid may comprise a polynucleotide. Passing the first and second fluids may comprise encapsuling a polynucleotide within a nanoparticle while mixing the first and second fluids.

[0030] Any of these methods may include passing the first and second fluids continuously for greater than one hour (e.g., greater 2 hours, greater than 3 hours, greater than 4 hours, greater than 5 hours, greater than 6 hours, greater than 7 hours, greater than 8 hours, greater than 9 hours, etc.). For example, the method may include passing the first and second fluids continuously for greater than six hours. Passing may comprise passing through between 2 and 20 stages of the channel. Passing may comprise passing down each of the one or more stages having a stage length for each stage of between 8 and 10 times a width of the channel. In any of these methods, passing may comprise passing down the channel having an inner radius of curvature is between 20 pm and 200 pm. Passing may comprise passing down the channel having an outer radius of curvature is between 40 pm and 400 pm. In some cases passing comprises passing down the channel having a width of the channel between 20 pm and 200 pm. Passing may comprise passing down the channel in which the inner radius of curvature is approximately equal to a width of the channel. In some examples passing may comprise passing down the channel in which the outer radius of curvature is approximately equal to twice the width of the channel. For example, passing may comprise passing down the channel in which a height of the channel is between 40 pm and 700 pm. Passing may comprise passing down the channel in which the aspect ratio of the rectangular cross-section is between 0.4 and 4 or greater. In some cases passing comprises passing down the channel in which the aspect ratio is 0.7 or greater.

[0031] Also described herein are methods of designing and / or fabricating microfluidic mixers, including microfluidic chips incorporating one or more of these mixers (e.g., including parallel mixers). In particular, described herein are methods for end-to-end integration of the microfluidic mixers described herein. Specifically, described herein are methods for designing chips to be integrated with the process steps described herein, upstream and / or downstream.

[0032] For example, any of these methods may include receiving a target set of parameters for a downstream microfluidics process comprising one or more process inputs; and outputting a microfluidics chip design, wherein the microfluidics chip design comprises a mixing channel extending between the two or more inputs and an output, wherein the mixing channel is planar and curving and has a rectangular cross-section transverse to a length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the mixing channel has an output that is configured to match the required process inputs, and wherein the mixing channel comprises a channel aspect ratio configured to achieve a 90% or more mixing efficiency using the one or more process inputs.

[0033] An of these methods may include providing the microfluidics chip based on the microfluidics chip design. Chips may be manufactured using any of the techniques described herein, including in particular, deep reactive ion etching (DRIE). In any of these examples, the one or more process inputs may comprise one or more of: flow rates and pressures; for example, the one or more process inputs may comprise a pressure 75 psi or less and / or a flow rate of between 100 ml / hour and greater than 2 L / hour. In some examples the channel may comprise one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages. The rectangular cross-section may have a constant cross-sectional area along the length from the two or more inputs to the output.

[0034] All of the methods and apparatuses described herein, in any combination, are herein contemplated and can be used to achieve the benefits as described herein.BRIEF DESCRIPTION OF THE DRAWINGS

[0035] A better understanding of the features and advantages of the methods and apparatuses described herein will be obtained by reference to the following detailed description that sets forth illustrative embodiments, and the accompanying drawings of which:

[0036] FIG. 1 shows an example of schematics of representative mixers compatible with the high aspect ratio (HAR) approach. (A)-(D) Dean-based high aspect ratio mixers: (A) D200L, (B) D200S, (C) D80, and (D) D40 mixers. (E)-(I) Concentric circle-based high aspect ratio mixers: (E) C1E0, (F) C1E1, (G) C1E3, (H) C1E6, (I) C0E7.

[0037] FIG. 2 illustrates Dean-based high aspect ratio (D-HAR) design and chips. (A)-(E) Top-down reflected light micrographs of the design of D200 designs: (A) D200L-0.18, (B) D200L-0.4, (C) D200L-0.7, (D) D200S-0.7, (E) D200S-3.5. (F)-(I) Cross-sectional reflectedlight micrographs of 150pm wide channels that were 35, 85, 140, and 700 pm deep, respectively. (J) Comparison of microfluidic chip using the mixer in (A) to a U.S. quarter. The mixing region is highlighted in yellow. Scale bars are 200pm.

[0038] FIG. 3 illustrates tunability in device throughput with the HAR approach. (A) Aqueous input pressure versus total flow rate for D200L-0.18, D200L-0.4, and D200L-0.7.(B) Aqueous input pressure versus total flow rate for D200S-0.7 and D200S-3.5. (C) Total flow rates achieved for D200L-0.18, D200L-0.4, and D200L-0.7 at aqueous input pressures between 45 and 75 psi. (D) Total flow rates achieved for D200S-0.7 and D200S-3.5 at aqueous input pressures between 45 and 75 psi.

[0039] FIG. 4 shows mixing in D200 chips with different aspect ratios. (A) Mixing achieved in D200L at aqueous input pressures from 45 to 75 psi and at 1mm along the transverse length of the channel. (B) Mixing achieved in D200S at aqueous input pressures from 45 to 75 psi and at 1mm along the transverse length of the channel. (C) Mixing vs. total flow rate for D200L-0.7. (D) Mixing vs. Dean number of the D200L devices. (E) Epifluorescence micrograph of mixing in D200L-0.7 at a total flow rate of 400mL / hr. Scale bar is 200pm.

[0040] FIG. 5 illustrates LNPs formulated with the D200 chips. (A) LNP Z-average diameter, (B) PDI, and (C) EE achieved within the D200 devices.

[0041] FIG. 6 shows D-HAR devices with reduced channel widths: D80 and D40. (A)-(C) Top-down reflected light micrographs of the design of D80 designs: (A) D80-0.4, (B) D80-1.0, and (C) D80-1.8. (D)-(F) Top-down reflected light micrographs of the design of D40 designs: (D) D40-0.88, (E) D40-2.0, and (F) D40-3.0. (G)-(I) Cross-sectional reflected light micrographs of 40pm wide channels that were 35, 85, and 120pm deep, respectively. Scale bars are 100pm.

[0042] FIG. 7 illustrates throughput and mixing in D80 and D40 chips at different aspect ratios. (A) Total flow rates achieved for D80 devices at aqueous input pressures between 45 and 75 psi. (B) Total flow rates achieved for D40 devices at aqueous input pressures between 45 and 75 psi. (C) Mixing achieved in D80 devices at aqueous input pressures from 45 to 75 psi and at 1mm along the transverse length of the channel. (D) Mixing achieved in D40 devices at aqueous input pressures from 45 to 75 psi and at 1mm along the transverse length of the channel. (E) Mixing vs. Dean number in the D80 devices. (F) Mixing vs. Dean number in the D40 devices.

[0043] FIG. 8 illustrates mixing in HAR devices at low flow rates. (A) Reflected light image of a co-flow input to a D40 device. (B) Reflected light image of a flow-focusing input to a D40 device. (C) Mixing vs. channel aspect ratio for D-HAR devices with co-flow inlets.(D) Mixing vs. channel aspect ratio for D-HAR devices with flow focusing inlets. All mixing values were measured at 1mm along the transverse length of the channel. Scale bars are 100pm.

[0044] FIG. 9 illustrates LNPs formulated with the D80 and D40 devices. (A) LNP Z- average diameter, (B) PDI, and (C) EE achieved within the D80 devices. (D) LNP Z-average diameter, (E) PDI, and (F) EE achieved within the D40 devices.

[0045] FIG. 10 shows examples of C-HAR devices containing a concentric circle and one expansion chamber. (A)-(C) Top-down reflected light micrographs of C1E1 designs: (A) C1E1-0.2, (B) C1E1-0.8, and (C) C1E1-3.2. (D)-(F) Cross-sectional reflected light micrographs of 150pm wide channels that were 35, 140, and 560pm deep, respectively. Scale bars are 200pm.

[0046] FIG. 11 illustrates throughput and mixing in C-HAR chips at different aspect ratios. (A) Aqueous input pressure versus total flow rate for C1E1-0.2, C1E1-0.8, and C1E1- 3.2. (B) Total flow rates achieved for C1E1 devices at aqueous input pressures between 45 and 75 psi. (C) Mixing achieved in C1E1 devices at aqueous input pressures from 45 to 75 psi and at the output from the final expansion chamber. (D) Epifluorescence micrograph of mixing in C1E1-0.8 at a total flow rate of 400mL / hr. (E) Epifluorescence micrograph of mixing in C1E3-0.8 at a total flow rate of 400mL / hr. Scale bars are 200pm.

[0047] FIG. 12 shows LNPs formulated with the C-HAR devices. (A) LNP Z-average diameter, (B) PDI, and (C) EE achieved within the C1E1 devices.

[0048] FIG. 13 shows examples of clogging in staggered herringbone mixers. (A) Reflected light micrograph of an SHM device operating for 5 minutes at 120mL / hr and with one-third of the concentration of lipids and PolyA. (B) Aqueous input pressure measured during SHM device operation. (C) Z-average diameter and (D) PDI of LNP samples during SHM operation. Scale bar is 100pm.

[0049] FIG. 14 illustrates one example of a HAR approach to reduce clogging and extend device operation. (A) Reflected light micrograph of D200L-0.18 after 5 minutes of operation. (B) Reflected light micrograph of D200L-0.7 after one hour of operation. (C) Aqueous input pressure measured during D200L device operation. (D) Z-average diameter and (E) PDI of LNP samples during D200L device operation. (F) Reflected light micrograph of C1E1-0.18 after 10 minutes of operation. (G) Reflected light micrograph of C1E1-0.7 after one hour of operation. (H) Aqueous input pressure measured during C1E1 device operation. (I) Z-average diameter and (J) PDI of LNP samples during C1E1 device operation. Scale bars are 200pm.

[0050] FIG. 15 shows an example of LNP formulation during 8 hours of continuous operation. (A) LNP Z-average diameter, (B) PDI, and (C) encapsulation efficiency achievedover an 8-hour period of continuous manufacturing. PolyA from Source 2 was used for this trial. (D) Image of LNP volume produced over the course of the 8-hour operation. (E) Reflected light micrograph of C1E3-0.7 after 8 hours of continuous manufacturing. Scale bars is 200pm.

[0051] FIG. 16 shows an example of a strategy for scale-independent manufacturing. (A) Schematic of approach for scaling up the formulation output. (B) Comparison (from left to right) of P1-C1E1-0.8, P1-C1E1-3.2, and P4-C1E1-3.2 microfluidic chips. (C) Schematic for parallelizing 20 mixing units in a single chip. (D) Comparison of a P20 parallelized chip to a U.S. quarter. (E) Expected throughput versus fold change in channel area (by increasing the channel height or parallelizing the mixers) using this scale-up strategy.

[0052] FIG. 17 illustrates the throughput capability of the scale-up strategy. (A) Aqueous input pressure versus total flow rate for Pl-D200s-0.7, Pl-D200s-3.5, and P4-D200s-3.5. (B) Aqueous input pressure versus total flow rate for P1-C1E1-0.8, P1-C1E1-3.2, and P4-C1E1- 3.2. (C) Total flow rates achieved for Pl-D200s-0.7, Pl-D200s-3.5, and P4-D200s-3.5 at aqueous input pressures between 45 and 75 psi. (D) Total flow rates achieved for P1-C1E1- 0.8, P1-C1E1-3.2, and P4-C1E1-3.2 at aqueous input pressures between 45 and 75 psi. (E)- (G) Epifluorescence micrographs of mixing in (E) Pl-D200s-0.7 at 700mL / hr, (F) Pl-D200s- 3.5 at 3.6L / hr, and (G) P4-D200s-3.5 at 700mL / hr. (H) Epifluorescence micrographs of mixing in P20-C1E3-0.8. Scale bars are 200pm.

[0053] FIG. 18 shows examples of LNP properties achieved using the scale-up strategy. (A)-(C) Properties of LNPs formulated with the scaled C1E1 devices: (A) Z-average diameter, (B) PDI, and (C) encapsulation efficiency. (D)-(F) Properties of LNPs formulated with the scaled D200S devices: (D) Z-average diameter, (E) PDI, and (F) encapsulation efficiency.

[0054] FIG. 19 illustrates LNP properties produced by scaled-up chips over time. (A) LNP Z-average diameter, (B) PDI, and (C) encapsulation efficiency achieved in P1-C1E1- 0.8, P1-C1E1-3.2, and P3-C1E1-3.2 over time.

[0055] FIG. 20 shows example workflows for end-to-end integration that are enabled by the HAR technology. (A) Integration with an intermediate reservoir, in which LNP formulation via a microfluidic HAR chip is stabilized by the addition of dilution buffer and stored prior to downstream processing. (B) Integration without an intermediate reservoir, in which the formulated LNPs are diluted with buffer and continuously fed into the SPTFF module, where the formulation output is specifically chosen to match the optimal feed flow rate for tangential flow filtration.

[0056] FIG. 21 illustrates example integration with single-pass tangential flow filtration (SPTFF). (A) Schematic and (B) LNP size distributions when an intermediate reservoir was used prior to SPTFF with a Cadence module. (C) Schematic and (D) LNP size distributions when an intermediate reservoir was not used prior to SPTFF with a Cadence module. (E) Schematic and (F) LNP size distributions when using three Pellicon capsules at a throughput of approximately 1g of PolyA-encapsulated LNP / hr. (G) Schematic and (H) LNP size distributions when using one Pellicon capsule to process 10 mg of PolyA-encapsulated LNP.

[0057] FIG. 22 shows reproducibility of manufacturing between runs. (A) LNP Z-average diameter, (B) PDI, and (C) EE across different manufacturing runs when formulated using D200L-0.7. (D) LNP Z-average diameter, (E) PDI, and (F) EE across different manufacturing runs when formulated using C1E1-0.8. (G) LNP Z-average diameter, (H) PDI, and (I) EE across different manufacturing runs when formulated using C1E1-3.2.

[0058] FIG. 23 illustrates LNPs formulated with different molecular compositions. (A) TEM image of LNPs formulated in the C1E1-0.8 device. (B) LNP Z-average diameter and (C) PDI versus concentration of encapsulated PolyA using the MC3-based formulation. (D) LNP Z-average diameter and (E) PDI versus concentration of encapsulated PolyA using the SM-102 and ALC-0315-based formulations. (F) LNP Z-average diameter and (G) PDI when formulated using different molar ratios of PEG in the lipid solution. (H) LNPs formulated when replacing 1,2-DSPC with 100% DOPE and (I) 50% / 50% DOPE / DSPC in the standard MC3 lipid formulation. Scale bars are 50nm.

[0059] FIG. 24 shows an example of LNPs formulated by changing process parameters. (A) LNP Z-average diameter, (B) PDI, and (C) EE versus total flow rate for three different LNP formulations using the D200S-0.7 device. (D) LNP Z-average diameter, (E) PDI, and (F) EE for the MC3 formulation created with the D200S-3.5 device.

[0060] FIG. 25 is a schematic of integration strategy for end-to-end processing. (A) Examples of downstream processes and process parameters that must be matched with the LNP output from the HAR devices. (B) Plot depicting achievable throughput versus device aspect ratio and degree of parallelization. (C) Plots depicting the flow rate output from the chips versus the required operational pressures.DETAILED DESCRIPTION

[0061] The microfluidic mixer apparatuses and methods described herein may achieve nearly complete (e.g., 85% or more, 90% or more, etc.) mixing efficiency within microseconds relatively low pressure (e.g., 75 psi or less) and at relatively high flow rates (e.g., flow rate of between 100 ml / hour and greater than 2 L / hour), without clogging. Ingeneral, these methods and apparatuses may include two or more inputs, an output, and a mixing channel extending between the inputs and the output that have a high aspect ratio and a defined curvature and cross-sectional dimensions. For example, the mixing channel extending between the two or more inputs and the output may be planar and curving, and may have a rectangular cross-section (transverse to a length of the channel), wherein the rectangular cross-section having a high aspect ratio (e.g., greater than 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 1.1, etc.), and wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages. The rectangular cross-section may have an approximately constant cross-sectional shape and / or area along the length from the two or more inputs to the output.

[0062] The method and apparatuses described herein may achieve robust nanomaterial formulation in a scale-independent manner that is compatible with GMP, continuous, and end-to-end integrated manufacturing and may include microfluidic mixers that achieve sufficiently high mixing while requiring a minimal footprint, may require low-to-moderate input fluid pressures for operation, and may exhibit comparable flow dynamics as the channel height is increased.

[0063] In general, these methods may include a sufficiently high mixing with a minimal footprint. Achieving rapid mixing may be crucial for controlling the formulation process, achieving desirable CQAs, and producing monodisperse nanomaterials. Mixing in a small footprint may minimize the likelihood of significant reagent aggregation within the channels and, by extension, minimize the likelihood that the CQAs of the nanomaterial will degrade during extended, continuous operation. Aggregation of formulation product and intermediates during the operation of a microfluidic chip may generally be a major concern. Microfluidic designs that minimize or prevent channel clogging are thus highly desirable. A small footprint is also desirable for maximizing the potential for further process scale-up via on-chip parallelization, because a greater number of smaller units may fit within one chip.

[0064] In general, these methods and apparatuses may operate at low-to-moderate input fluid pressures. Second, and intimately related to minimizing the footprint of the channels, is the goal to minimize the pressure drop within the microfluidic device. Achieving sufficiently high mixing at low to moderate pressures would allow for direct integration with other downstream fluidic processes and provide compatibility with a variety of process pumps typically used in GMP manufacturing.

[0065] These methods and apparatuses may provide comparable flow dynamics as the channel height is increased. As described herein, mixers that exhibit comparable flow dynamics as the channel height is increased may provide an opportunity to scale the formulation throughput, provided the device fabrication process is robust, tunable, and capable of generating high aspect ratio channels. In some microfluidic mixing cases, such as traditional flow focusing designs, sheath flow develops as the aspect ratio of the channel is increased, resulting in comparable, purely diffusive mixing. These methods and apparatuses may not provide flow focusing (e.g., sorting), but may instead provide microfluidic geometries that provides similar mixing when the channel height and total flow rate are comparably increased, whether that is due to internal, three-dimensional flows that arise based on the channel architecture, or due to a rapid and efficient “sheath-like” merging of multiple fluid streams. One additional advantage of increasing the throughput by increasing the channel aspect ratio is that similar shear is expected across scaled devices. Considering a rough approximation based on Poiseuille flow, shear stress at the wall is expected to vary as piv / w, where pi is the fluid viscosity, w is the channel width for a high aspect ratio device, and vis the average velocity through the channel. If pi, rand w are maintained as the channel height is increased, then shear of a similar order of magnitude would be expected as the aspect ratio is increased over different devices.

[0066] As a set of initial examples, these microfluidic mixing units may be expected to achieve rapid mixing through the formation of secondary Dean vortices (FIGS. 1 A-D). These vortices are internal, three-dimensional flows that arise from centrifugal forces in a curved channel. Enhanced mixing is observed in Dean-based microfluidic mixers as the non- dimensional Dean number (De) is increased.Here, Re is the Reynolds number, DHis the hydraulic diameter of the channel, and R is the radius of curvature of the inner (concave) wall. Thus, the Dean number may be increased, for a given set of fluids, by increasing the hydraulic diameter of the microfluidic channel while maintaining the average flow velocity through a channel with a pre-defined radius of curvature. By increasing the channel height (H) while maintaining the channel width (VF), identical to increasing the channel aspect ratio (H / W) for a given channel width, the hydraulic diameter, and thus the Dean number, will monotonically increase.

[0067] For this reason, it is expected that by determining a minimum channel aspect ratio and flow rate that produces sufficiently high mixing at moderate input pressures, a critical Dean number can be identified that correlates with the formation of secondary vortices and enhanced mixing for that channel width. From that starting point, comparably high mixing, at moderate input pressures, is expected to be achieved as the Dean number, formulation throughput, and channel aspect ratio are correspondingly increased.

[0068] To demonstrate the applicability of the strategy to other microfluidic mixing modes, a novel set of microfluidic mixing designs were created that consist of up to one fluidic channel in the shape of a concentric circle and up to multiple expansion chambers (FIGS. 1E-1I). It is expected that the channel in the shape of a concentric circle would serve to initially merge the two input fluidic streams, while the expansion chambers would provide additional mixing via the formation of vortices that result when the cross-sectional shape and / or area of the channel abruptly changes. In this case, treating the input fluid streams as sheaths, increasing the channel height is expected to effectively scale the mixing phenomena in the z-direction if the average fluid velocity is maintained within the channel. As before, identifying a minimal channel height that produces sufficiently high mixing at moderate input pressures is the starting point to this scale-up strategy.

[0069] Variations of the two types of designs described above are depicted in FIG. 1. The first type, in which mixing is driven by Dean flows, are referred to as Dean-based high aspect ratio (D-HAR) mixers. The representative devices considered here consist of channel widths ranging from 40pm to 200pm and, for simplicity, the radius of curvature of the concave wall is specified to be identical to the channel width. In the rest of this work, these mixers are named according to the channel width and aspect ratio, defined as the channel height divided by the channel width.DiChannel Width}-} Aspect Ratio!

[0070] As an example, FIG. 1 thus depicts schematics for D200 (FIGS. 1 A-B), D80 (FIG. 1C), and D40 (FIG. ID) mixers. Subscripts will be used to delineate between the D200 mixers with the longer, 4mm channel length (D200L, FIG. 1 A) and shorter, 2mm channel length (D200s, FIG. IB). The second type of mixing unit, containing the concentric circle and expansion chambers, are referred to as concentric circle based high aspect ratio (C-HAR) mixers. These mixing designs could consist of just one concentric circle (FIG. IE), one concentric circle with multiple expansion chambers (FIGS. 1F-H), or just expansion chambers (FIG. II). Here, these mixers are named according to the number of concentric circles, expansion chambers, and aspect ratio.C{Number of concentric circles, 0 or l }E{Number of expansion chambers}-) Aspect Ratio}

[0071] In this case, the representative width is taken to be the width of the channel in the concentric circle region. FIG. 1 thus depicts schematics for C1E0 (FIG. IE), C1E1 (FIG. IF), C1E3 (FIG. 1G), C1E6 (FIG. 1H), and C0E7 (FIG. II) mixers. The results show that microfluidic high aspect ratio (HAR) results in significant mixing.

[0072] To realize the design considerations described above, a fabrication strategy rooted in semiconductor manufacturing was used to create the D-HAR mixers with precisely defined channel aspect ratios in silicon and glass microfluidic chips. This strategy used deep reactive ion etching (DRIE) to produce channels in photolithographically defined regions on a silicon wafer; the silicon wafer was then bonded to borosilicate glass to create microfluidic chips with fully enclosed internal channel regions. As an initial example, D200L -0.18 (FIG. 2A), D200L-0.4 (FIG. 2B), D200L-0.7 (FIG. 2C), D200s-0.7 (FIG. 2D), and D200s-3.5 (FIG. 2E) microfluidic chips were created. Images of the cross-sections of these chips depicted highly rectangular channels in each of these devices (FIG. 2F-I). Specifically, channels with heights of 35pm (FIG. 2F), 80pm (FIG. 2G), 140pm (FIG. 2H), and 700pm (FIG. 21) were created, corresponding to the channel aspect ratios of 0.18, 0.4, 0.7, and 3.5, respectively. Thus, it was observed that the fabrication strategy could produce microfluidic chips with an identical X / Y footprint for the mixing region, but with highly tunable channel aspect ratios. A representative microfluidic chip was less than the size of a U.S. quarter (FIG. 2J).

[0073] The HAR strategy enabled microfluidic chips with throughputs that were tunable, for the same mixer footprint and input fluidic pressures, by specifying the device aspect ratio. To operate these chips, an aqueous input stream was mixed at a 3 : 1 flow rate ratio relative to an ethanolic input stream. These two input fluids were chosen because of wide interest in creating lipid nanoparticles (LNPs), which may be formed via microfluidic nanoprecipitation when ethanolic lipids are mixed with an aqueous buffer. Operating these devices across a range of pressures revealed a unique set of flow rates that were achievable for each of the D200L (FIG. 3 A) and D200s (FIG. 3B) devices. When each device was operated with similar input aqueous pressures (45psi to 75psi), the lowest aspect ratio (D200L-0.18, FIG. 3C) and highest aspect ratio (D200s-3.5, FIG. 3D) devices could achieve total flow rates of up to lOOmL / hr and 5L / hr, respectively. For each of the D200L (FIG. 3C) and D200s (FIG. 3D) designs, the throughput scaled approximately linearly with the aspect ratio of each set of devices. Note that this moderate pressure range was considered as a reference throughout this work because it is achievable with pumps that are often used in GMP processes. Also, the aqueous input pressure was always higher than the ethanolic input pressure in these operating conditions and so the aqueous input pressure was thus viewed as a limiting factor for operation.

[0074] Robust mixing was observed in the D200 chips, for the same mixer footprint and required fluidic pressures, provided the devices were created with channels above a threshold aspect ratio. When the D200L (FIG. 4A) and D200s (FIG. 4B) devices were operated at input pressures between 45psi and 75psi, robust mixing was observed in devices created with a channel aspect ratio of at least 0.4. That is, robust mixing was observed in the collection of D200 devices at throughputs ranging from 200mL / hr to 5L / hr. Note that a high degree of mixing could be achieved at lower flow rates as well, and these required input pressures less than 45psi (FIG. 4C). Furthermore, high mixing was observed in these channels at Dean numbers greater than 150 (FIG. 4D), which could not be achieved in the low aspect ratio device (D200L-0.18) at the moderate input pressures considered here. A representative epifluorescence micrograph of D200L-0.7 in operation depicts the high mixing that was achieved in these devices (FIG. 4E). For comparison between all D-HAR devices, mixing was considered 1mm along the transverse length of the channel. Taken together, these data show that by varying the aspect ratio of a D200 mixer above 0.4, the throughput can be tuned while maintaining high mixing, the same footprint, and comparable input pressures.

[0075] Formulating LNPs in the D200 mixers revealed that uniform CQAs were achievable at throughputs from 200mL / hr to 5L / hr by choosing devices with an appropriate channel aspect ratio. Here, the CQAs of interest for LNP formulation were Z-average diameter, poly dispersity index (PDI), and encapsulation efficiency (EE). Throughout this work, LNPs were formulated by mixing an aqueous solution containing polyadenylic acid (Poly A) and an ethanol solution containing a lipid mixture. Unless stated otherwise, the lipid solution consisted of DLin-MC3-DMA (MC3), 1,2-DSPC, Cholesterol, and DMG- PEG(2000) dissolved in ethanol at a molar ratio of 40: 10:48:2 (MC3: 1,2- DSPC:Cholesterol:DMG-PEG(2000)). This solution was rapidly mixed with lOOmM citrate buffer containing 320ug / mL Poly A, representing a typical nucleic acid cargo, at a final N / P ratio of 6 to produce LNPs with a final PolyA concentration of 240ug / mL. The D200 devices created with a channel aspect ratio of at least 0.4 produced LNPs with a Z-average diameter of approximately 60nm (FIG. 5 A), PDI less than 0.1 (FIG. 5B), and EE of approximately 90% or higher (FIG. 5C). LNPs with these CQAs were produced over throughputs ranging from 240mL / hr (D200L-0.4) to 5L / hr (D200s-3.5) by appropriately scaling the total flow rate and maintaining comparable input pressures based on the selected device. Larger, more poly disperse LNPs were formed in D200L-0.18, which did not meet the minimum aspect ratio required to achieve rapid mixing (FIG. 4).The HAR strategy with smaller channel widths

[0076] The HAR strategy, not limited to channels greater than 100pm in width, was further used to create 40pm and 80pm-wide channels with controllable aspect ratios. Specifically, D80-0.4 (FIG. 6A), D80-1.0 (FIG. 6B), D80-1.8 (FIG. 6C), D40-0.88 (FIG. 6D), D40-2.0 (FIG. 6E), and D40-3.0 (FIG. 6F) microfluidic chips were created using the same fabrication strategy. As previously, highly rectangular channels were observed in these devices (FIGS. 6G-I). In the case of the D40 devices, for example, 40pm-wide channels with heights of 35pm (FIG. 6G), 80pm (FIG. 6H), and 120pm (FIG. 61) were created. It was thus observed that the fabrication strategy could produce microfluidic mixers with controllable aspect ratios, without changing the mixing footprint, when the channel widths ranged from less than 50pm to at least 200pm.

[0077] When the HAR approach was applied to D80 and D40 devices, high mixing was again achieved for the same footprint and a comparable required input pressure, provided the device exceeded a threshold aspect ratio. When the D80 chips were operated at fluidic input pressures between 45 and 75 psi, total flow rates from 60mL / hr to 280 mL / hr were achieved by varying the aspect ratio from 0.4 to 1.8 (FIG. 7A). For the D40 chips operating with these pressure conditions, total flow rates from 40mL / hr to 180mL / hr were achieved by varying the aspect ratio from 0.88 to 3.0 (FIG. 7B). High (95%) mixing was demonstrated in the D80 devices created with a channel aspect ratio of at least 1.0 (FIGS. 7C), and in D40 devices created with a channel aspect ratio of at least 2.0 (FIGS. 7D). The minimum throughputs in this pressure range were 150mL / hr and lOOmL for the D80 and D40 devices, respectively. The HAR strategy thus achieved high mixing in these devices at reduced total flow rates compared to the D200 devices, but at similar input pressures. In the D80 and D40 mixers, 95% mixing was consistently observed at Dean numbers greater than approximately 150 and 100, respectively (FIGS. 7E-F), though the D40-3.0 devices exhibited high mixing at lower Dean numbers as well. As with D200, the lowest aspect ratios devices could not reach a sufficiently high Dean number at the moderate input pressures considered here. Furthermore, high mixing could be achieved in D-HAR devices at throughputs of less than 50mL / hr by increasing the channel aspect ratio and considering alternate inputs to the mixers (FIG. 8). These data show that increasing the channel aspect ratio to 5, by decreasing the channel width to 20pm, resulted in high mixing below 30mL / hr, with flow-focusing inlets (FIG. 8B, D) generally achieving higher mixing compared to co-flow inlets (FIG. 8A, C). Unless otherwise specified, the work described here considers devices with co-flow inlets as representative examples (FIG. 1).

[0078] Formulating LNPs in the D80 and D40 mixers revealed that uniform CQAs could be achieved at throughputs of less than lOOmL / hr, and up to 300mL / hr, by choosing deviceswith an appropriate channel aspect ratio. Using the D80 devices, LNPs with a Z-average diameter of approximately 60nm (FIG. 9A), PDI less than 0.1 (FIG. 9B), and EE greater than 90% (FIG. 9C) were formed in devices with a channel aspect ratio of at least 1.0. Similarly, with the D40 devices, LNPs with a Z-average diameter of approximately 60nm (FIG. 9D), PDI of less than 0.1 (FIG. 9E), and EE greater than 90% (FIG. 9F) were formed in devices with a channel aspect ratio of at least 2.0. In D80-1.0 and D40-2.0, LNPs with these CQAs were produced at throughputs of 160mL / hr and 80mL / hr, respectively. Larger, more polydisperse LNPs were formed in D80-0.4 and D40-0.9, which did not achieve the same degree of mixing as the other D80 and D40 devices, respectively (FIG. 7). Thus, the design of these D-HAR mixers also required surpassing a minimum channel aspect ratio to produce LNPs with uniform properties across throughputs, though the range of achievable throughputs was expanded to lower values compared to the D200 devices.The HAR strategy with C-HAR devices

[0079] The HAR fabrication strategy was further applied to an additional set of mixers, in which microfluidic channels with pre-determined aspect ratios were again created while maintaining the same footprint. Specifically, a novel set of microfluidic mixers were developed that consisted of no more than one concentric circle and up to many expansion chambers (FIG. 1). By applying the same fabrication strategy, C1E1-0.2 (FIG. 10A), C1E1- 0.8 (FIG. 10B), and C1E1-3.2 (FIG. 10C) microfluidic chips were created. These C-HAR devices exhibited rectangular channels with heights of 35pm (FIG. 10D), 140pm (FIG. 10E), and 560pm (FIG. 10F), respectively. While these representative mixers contained one concentric circle and one expansion chamber, devices with any number of expansion chambers, with or without the concentric circle, were also created.

[0080] As with the D-HAR devices, the C-HAR chips demonstrated high mixing at throughputs determined by the aspect ratio of the device, for the same footprint and a comparable required input pressure. In the case of the C1E1 chips, total flow rates from less than lOOmL / hr and up to 2.5L / hr were achieved by varying the aspect ratio from 0.2 to 3.2 (FIG. 11 A-B). As previously, the throughputs that were achieved between 45 and 75 psi varied linearly with the aspect ratio of the device. The higher aspect ratio devices considered (C1E1-0.8 and C1E1-3.2) achieved 95% mixing by the output from the expansion chamber; the lower aspect ratio device (C1E1-0.2) did not achieve this degree of mixing (FIG. 11C). Representative epifluorescence micrographs of C1E1-0.8 (FIG. 11D) and C1E3-0.8 (FIG.1 IE) operating at a total flow rate of 400mL / hr confirmed the homogenous dye signal at the output from the final expansion chambers. Robust mixing was thus achievable in a C-HARdevice at a moderate input pressure, provided it was created with channels of a sufficiently high aspect ratio.

[0081] The novel C-HAR mixers produced LNPs with uniform CQAs over a range of throughputs comparably to the D-HAR mixers, without changing the footprint or required pressures, when the devices exceeded the critical aspect ratio. For the higher aspect ratio devices (C1E1-0.8 and C1E1-3.2), uniform LNP properties were achieved over a four-fold range in throughputs (FIGS. 12A-C). In these cases, LNPs with a Z-av erage diameter of approximately 60nm (FIG. 12A), PDI less than 0.1 (FIG. 12B), and EE greater than 90% (FIG. 12C) were formed at throughputs from 500mL / hr to 2.2L / hr. The device with channels of a lower aspect ratio, C1E1-0.2, produced larger, polydisperse LNPs, though the size was smaller than the D-HAR cases. This device did not mix the two fluidic streams as effectively as the higher aspect ratio devices (FIG. 11). Thus, the novel mixers designed here also produced LNPs with uniform CQAs over a range of throughputs, provided the HAR approach was followed.Continuous, clog-free, and stable operation with microfluidic HAR devices

[0082] Continuous and stable operation of a microfluidic device for nanomaterial production, in which CQAs of the formulated product were maintained during the production run, required that a sufficiently small degree of aggregation occurred on the channel surfaces during operation. To illustrate when this condition was not satisfied, LNPs were formulated in a staggered herringbone mixer (SHM), a popular microfluidic mixing unit known to aggregate lipids and nucleic acids within the out-of-plane herringbone grooves. Despite attempting to reduce device clogging by decreasing the lipid and PolyA concentrations to one-third of that used with the HAR devices, lipid aggregation was still observed within 10 minutes in the herringbones (FIG. 13 A). During the operation of this device, the required input fluid pressure steadily rose to compensate for the build-up of material within the channels and maintain the total flow rate of 120 mL / hr (FIG. 13B). Collecting samples over time revealed that while the Z-average diameter remained approximately constant (FIG.13C), the PDI increased until the device operation had to be stopped (FIG. 13D). Due to the steadily increasing input pressure and the degradation of LNP CQAs, the SHM device could not operate stably for an extended period.

[0083] Contrasted with the SHMs, devices produced with the HAR approach operated stably and continuously for extended periods with reduced clogging. To evaluate the effect of channel aspect ratio on the tendency for a device to clog and become inoperable, low aspect ratio devices (D200L-0.2 and C1E1-0.2) were compared with higher aspect ratio counterparts (D200L-0.7 and C1E3-0.8). In the case of D200L-0.2, significant aggregation occurred in thechannels (FIG. 14A) and the required input pressures rapidly rose over 5 minutes of operation (FIG. 14C). Analyzing the LNPs at this time revealed that large, polydisperse particles were formed (FIGS. 14D-E). On the other hand, when the D200-0.7 device was operated at 400mL / hr for one hour, only minor aggregation was observed in the channels (FIG. 14B) and the aqueous input pressure did not change over time (FIG. 14C). Importantly, the LNP CQAs did not change over time (FIGS. 14D-E): LNPs with a Z-average diameter of 60nm (FIG. 14C) and PDI less than 0.1 (FIG. 14D) were consistently produced over these 60 minutes. Similar observations were made with the C-HAR chips: significant lipid aggregation was observed (FIG. 14F), and the input pressure did not stabilize (FIG. 14H), when C1E1-0.2 was operated for ten minutes. Conversely, C1E3-0.8 produced uniform LNPs for one hour with only minor aggregation in the channels (FIG. 14G) and a stable pressure profile (FIG. 14H). In the latter case, LNPs with a Z-average diameter of 60 nm (FIG. 141) and PDI less than 0.1 (FIG. 14 J) were produced during the continuous operation.

[0084] The utility of the HAR approach was further demonstrated by operating a chip continuously and clog-free for eight hours, without the degradation of LNP CQAs. The eight- hour duration was chosen because it corresponds to one manufacturing shift in a pharmaceutical manufacturing environment. During this manufacturing run, LNPs were formulated to contain 240ug / mL PolyA (Source 2, see Materials and Methods) at 400mL / hr with C1E3-0.8. Over the course of the eight hours, LNPs were consistently produced with a Z-average diameter of approximately 75nm (FIG. 15 A), PDI less than 0.1 (FIG. 15B), and EE greater than 90% (FIG. 15C). Upon completion of the run, 3.2L of LNP product was produced (FIG. 15D) while the device did not exhibit noticeable lipid aggregation (FIG.15E). Taken together, these data demonstrate the ability of HAR devices to operate continuously and clog-free for manufacturing-scale durations.Scale-independent manufacturing

[0085] The HAR strategy enabled scale-independent microfluidic manufacturing of nanomaterials by combining throughput tunability, via specifying the aspect ratio of a device, with on-chip parallelization. That is, for a given mixing design, low throughputs could be achieved by operating a version of the device with a small channel aspect ratio (that still met the minimum aspect ratio threshold); the throughput could be scaled by first increasing the aspect ratio of the device, and then parallelizing that mixing unit on a single chip (FIG. 16 A). To demonstrate the fabrication of such chips, the C1E1 mixing design was incorporated into three devices capable of increasingly greater throughputs: a C1E1-0.8 chip, a C1E1-3.2 chip, and a C1E1-3.2 chip with four mixing units arranged in parallel (FIG. 16B). Importantly, these three chips required just one ethanolic input and one aqueous input. For convenience,parallelized devices were named according to the number of mixers arranged in parallel and the name of the mixing unit.P{Number of mixing units in parallel} -{Device name}

[0086] Additional parallelization was achieved with the VLSMI strategy by producing, as a representative example, a microfluidic chip with 20 microfluidic mixing units arranged in parallel (FIG. 16C). In each of these chips, the total size of the chip was still approximately that of a U.S. quarter (FIGS. 16B, D). Since the basis of this strategy was to maintain a comparable average velocity through the mixing channels across these different devices, the throughputs of devices with further parallelization were estimated (FIG. 16E). For example, a P100-C1E1-3.2 device, requiring the space of approximately five quarters, would be expected to formulate nanomaterial at over 200L / hr.

[0087] The HAR strategy enabled the tuning of the formulation throughput over orders of magnitude while exhibiting comparable mixing and requiring similar input pressures. To demonstrate this, the scale-up strategy was applied to the D200 and C1E1 designs. In both cases, P1-D200-0.7 and P1-C1E1-0.7 operated at total flow rates of 100s of mL / hr, whereas P4-D200-3.5 and P4-C1E1-3.2 devices achieved total flow rates on the order of lOL / hr (FIGS. 17A-B). Comparing throughputs at input pressures between 45psi and 75psi, the devices with different scales (P1-D200-0.7, P1-D200-3.5, and P4-D200-3.5) operated at approximately 700mL / hr, 3.5L / hr, and 14L / hr, respectively (FIG. 17C). P1-C1E1-0.8, Pl- C1E1-3.2, and P4-C1E1-3.2 achieved a similar linear scalability in throughput, reaching 500mL / hr, 2.2L / hr, and 8.5L / hr, respectively (FIG. 17D). Furthermore, mixing visualization revealed comparable mixing characteristics across P1-D200-0.7 (FIG. 17E), P1-D200-3.5 (FIG. 17F), and P4-D200-3.5 (FIG. 17G), as well as across C-HAR mixers in P10-C1E3-0.8 (FIG. 17H).

[0088] LNPs with consistent properties were produced across manufacturing scales with the HAR scale-independent manufacturing strategy. To demonstrate this, the C1E1 (FIGS. 18A-C) and D200 (FIGS. 18D-F) devices described in FIG. 17 were used to produce LNPs at total flow rates from 500mL / hr to 15L / hr. In both cases, LNPs with a Z-average diameter of approximately 60nm (FIG. 18 A, D), PDI less than 0.1 (FIG. 18B, E), and EE greater than 90% (FIG. 18C, F) were produced across this range of throughputs. Taken together, these data demonstrate how the HAR strategy can be used to specify a desired LNP formulation throughput over orders of magnitude, for the same microfluidic mixing design, by specifying the channel aspect ratio and degree of parallelization.

[0089] Combining scale-independent manufacturing with stable, continuous operation, the scaled HAR microfluidic mixers produced LNPs with uniform properties over extendedperiods of chip operation. To evaluate this, P1-C1E1-0.8, P1-C1E1-3.2, and P2-C1E1-3.2 were operated at 500mL / hr, 2.2L / hr, and 4.4L / hr, respectively, and the properties of the produced LNPs were analyzed for up to 40 minutes of device operation. These experiments revealed that consistent LNP CQAs were achievable over time and across the throughput scales (FIGS. 19A-C). Specifically, LNPs were consistently produced with a Z-average diameter of approximately 60nm (FIG. 19A), PDI of less than 0.1 (FIG. 19B), and EE greater than 90% (FIG. 19C). Taken together, these data reinforce the ability of the HAR strategy to produce microfluidic chips for scale-independent and continuous manufacturing of nanomaterials.End-to-end integration

[0090] The microfluidic HAR strategy is well-suited for end-to-end integration with downstream process equipment due to its ability to A) specify the throughput for a given mixing design at a moderate input pressure, and B) operate continuously. One way to perform such an integration would be to produce LNPs via a microfluidic HAR chip, stabilize the LNPs by mixing the formulation output with an aqueous buffer, and then store the diluted formulation in an intermediate reservoir prior to downstream processing (FIG. 20A). As an example, the downstream processing step(s) may include single-pass tangential flow filtration (SPTFF), a common filtration process used to reduce the residual solvent in a formulation. In this case, a HAR chip may be selected to minimize the LNP formulation time, while the downstream SPTFF process may be initiated simultaneously to, or after, the formulation step. Alternatively, the HAR strategy also enables the integration of formulation and SPTFF without the use of an intermediate reservoir (FIG. 20B). Here, the formulation output may be specifically chosen to match the optimal feed flow rate and pressure for SPTFF, thereby negating the need for intermediate storage.

[0091] Demonstrating the integration of HAR with downstream processing, LNP CQAs were maintained from formulation through SPTFF, with and without an intermediate reservoir. First, LNPs were formulated using the C1E1-3.2 device, diluted 12X in PBS, and collected in an intermediate reservoir. SPTFF was performed using a 30kDa, 0.24m2SPTFF module to concentrate the diluted LNPs back to the original concentration. In this case, LNPs exhibited similar properties prior to and after SPTFF (FIG. 21 A, B), with a Z-average diameter of approximately 60 nm and PDI less than 0.1 measured in both cases. To demonstrate end-to-end integration without an intermediate reservoir, formulated LNPs were diluted with buffer and the diluted formulation was continuously fed at a feed flow rate of 400mL / min into the same type of SPTFF module. As before, comparable LNPs were measured with and without SPTFF (FIG. 21C, D), exhibiting a Z-average diameter ofapproximately 60nm and PDI less than 0.1. Additionally, to show the compatibility of HAR with a variety of SPTFF modules, LNPs were formulated, collected in an intermediate reservoir, and fed into three lOOkDa 0.1m2SPTFF capsules arranged in series. Consistent with the previous module, LNPs with a size of approximately 65nm and PDI of 0.1 were maintained throughout the SPTFF process (FIG. 2 IE, F). Finally, to show the compatibility of HAR with SPTFF at reduced formulation throughputs, LNPs were formulated at 120 ug / mL, stored in an intermediate reservoir, and fed into a single lOOkDa 0.1m2SPTFF capsule. In this case, LNPs with a size of approximately 70nm and PDI of less than 0.1 were maintained throughout the SPTFF process (FIG. 21G, H). These final two tests represented LNP processing at approximately 1g of PolyA-encapsulated LNP per hour (FIG. 2 IF) and with a batch of lOmg of LNP drug product (FIG. 21H).Reproducibility and GMP -forward

[0092] In addition to demonstrating reproducible operation over time and across formulation throughputs, the HAR microfluidic mixers demonstrated reproducible operation over different manufacturing runs. To evaluate this, D200L-0.7 (FIG. 22A-C), C1E1-0.8 (FIG. 22D-F), and C1E1-3.2 (FIG. 22G-I) were used to formulate LNPs on three separate days. For both devices, the average values across the respective three runs were consistent with the values achieved previously: a Z-av erage diameter of approximately 60nm (FIGS. 22A, D, G), PDI less than 0.1 (FIGS. 22B, E, H), and EE greater than 90% (FIG. 22C, F, I). The coefficient of variation across the measured Z-average diameter and EE was less than 5% for each device. While the coefficient of variation for the measured PDI values were higher, the absolute value of the PDI only varied from 0.03 to 0.05 across the six different runs, indicating that highly monodisperse LNP populations were produced in all runs and for each mixer.Other nanoparticle manufacturing

[0093] The HAR mixers described here were not restricted to MC3 -based formulations and were also used to create LNPs with different compositions. Beyond using MC3 as the ionizable lipid (FIG. 23 A-C), LNPs were also produced with lipid mixtures based on SM-102 or ALC-0315 (FIG. 23D-E). To compare the LNPs produced with these different mixtures, LNPs were formulated with different concentrations of lipids and PolyA at 600mL / hr using the D200s-0.7 device while maintaining a fixed N / P ratio of 6. In the MC3 formulation, LNPs with a Z-average diameter of approximately 60nm (FIG. 23B) and PDI less than 0.1 (FIG. 23C) were formed across the range of concentrations tested. The SM-102 and ALC-0315 formulations exhibited comparable, but slightly reduced, Z-average diameters (FIG. 23D) and similar PDI (FIG. 23E) values across the range of concentrations tested. Furthermore, whenthe HAR mixers were used to formulate LNPs with an increasing molar ratio of PEG, LNPs with decreasing Z-av erage diameter (FIG. 23F) and a small change in PDI (FIG. 23 G) were produced. Finally, LNPs with a different phospholipid, DOPE, were formulated by replacing 100% (FIG. 23H) or 50% (FIG. 231) of DSPC with DOPE. The HAR devices were thus used to produce LNPs with a variety of compositions.

[0094] LNP properties were also tunable, at multiple scales of formulation throughput, by changing critical process parameters when operating the HAR devices. To evaluate this, a lower throughput HAR device (D200s-0.7) and a higher throughput HAR device (D200s-3.5) were used to formulate LNPs at different flow rates. When using D200s-0.7 with the MC3, SM-102, and ALC-0315 formulations, increasing the total flow rate resulted in a decrease in the Z-average diameter of the LNPs (FIG. 24A), a decrease in the PDI (FIG. 24B), and EE at or above 90% (FIG. 24C) as the total flow rate was increased to 500mL / hr. Above this flow rate, the LNP properties remained relatively constant. When LNPs were created with the MC3 formulation and using the scaled up microfluidic mixer, D200s-3.5, increasing the total flow rate from 2000mL / hr to 5000mL / hr resulted in a decrease in LNP Z-average diameter (FIG. 24D), comparable PDIs (FIG. 24E), and encapsulation efficiency above 90% (FIG. 24F). Taken together, these data demonstrate that LNP properties may be tuned by adjusting the total flow rate of both scaled-down and scaled-up HAR mixers.

[0095] Described herein are microfluidic, scale-independent, continuous, end-to-end, and GMP-compatible manufacturing of nanomaterials. These methods include the precise design and fabrication of microfluidic chips, based on a two-dimensional mixing geometry of interest, for specific operating conditions and process requirements. As described herein, once a sufficiently high aspect ratio was determined for achieving rapid mixing in that unit, the throughput that was achievable could be arbitrarily scaled, potentially orders of magnitude higher, by specifying the channel aspect ratio and degree of on-chip parallelization. Determining the minimum aspect ratio for a given device was important for clog-free, continuous, and stable microfluidic operation, and an eight-hour manufacturing run was demonstrated for a chip designed with the HAR strategy. Furthermore, the ability to operate a given microfluidic mixer continuously and with a tunable throughput made the HAR strategy well-suited for integration with downstream processes, which was demonstrated by operating SPTFF immediately downstream of the chip. Finally, the HAR strategy enabled reproducible nanomaterial formulation over time, over orders of magnitude of formulation throughputs, and across different manufacturing runs. This reproducibility, along with the glass-based fabrication strategy that did not include the leachable components found in plastics, highlighted avenues in which the HAR strategy aligned with requisites forGMP. Taken together, the HAR strategy addresses current challenges and provides a roadmap for producing GMP-ready microfluidic chips for targeted formulation throughputs that are capable of continuous and end-to-end operation.

[0096] To demonstrate the HAR strategy, two distinct sets of microfluidic mixers were designed and characterized in which rapid mixing and uniform LNP production were observed above a critical aspect ratio at low-to-moderate input fluidic pressures. In D-HAR devices that did not satisfy this critical aspect ratio, it was not possible to achieve a sufficiently high Dean number to produce rapid mixing, which correlated with channel fouling during LNP formulation. While the Dean number could be increased in these devices by operating at higher flow rates, this would require A) much higher input pressures, making the operation of the microfluidic chip difficult to integrate with downstream processes and become potentially dangerous, as well as B) increasing the shear experienced by the formulation components. Previous reports estimate a critical Dean number for the formation of secondary Dean vortices to be on the order of 50-200, which aligned with the threshold Dean numbers of approximately 150 observed here. When the critical aspect ratio criterion was satisfied, and rapid mixing was achieved, the devices reproducibly generated LNPs, and scale-up could be achieved by proportionally increasing the channel aspect ratio and total flow rate. These observations also suggest that the formulation throughput may be decreased in devices near the critical aspect ratio by decreasing the channel width for a given height to avoid unstable device operation and the introduction of undesirable CQAs. Other designs considerations, such as optimizing the inlets used in the device, may also reduce the critical Dean number and provide an additional route for achieving high mixing at low formulation throughputs.

[0097] In addition to the D-HAR devices, a unique set of C-HAR mixers were designed to demonstrate the widespread nature of the HAR strategy. These microfluidic mixers consisted of up to one concentric circle and up to many expansion chambers. These mixers differ from the bifurcating mixers demonstrated previously, as the design and mechanism of mixing are fundamentally different. In the bifurcating mixers, mixing is facilitated by repeated cycles of fluid exchange between two distinct fluidic paths with different impedance, and at least one pair of toroids is present. In contrast, the mixers created here do not contain pairs of toroids - no more than one concentric circle was present - and mixing occurred near the entrance to the concentric circle and via multiple cycles of expansion vortices in the downstream chambers. Also, in the previous work, the toroids were oriented by an angle of 120 degrees, whereas the concentric circle was oriented relative to the expansion chambers by 60 degrees here. In other previous work, a high aspect ratiomicrofluidic hydrodynamic focusing device based exclusively on diffusive, flow focusing mixers demonstrated a throughput of up to lOOmg (lipid) per hour when formulating liposomes. Contrasting with that work, a fundamental aspect of the HAR strategy is the ability to achieve scale-independent manufacturing and specify a target throughput in the chip design stage. As a result, the HAR strategy was used to tune the LNP formulation throughput to much higher rates, demonstrating at least 90 g (lipid) per hour. Furthermore, the HAR strategy does not rely on diffusive mixers - it was applied to different microfluidic mixing designs here - and it demonstrated clog-free, continuous, and long-term operation. In fact, the HAR strategy is expected to apply to any microfluidic mixing unit in which comparable mixing can occur as the channel height and throughputs are correspondingly increased.

[0098] By enabling the user to specify the channel aspect ratio and degree of on-chip parallelization, the HAR strategy enabled scale-independent manufacturing of nanomaterial formulations. Here, LNPs with similar CQAs were achieved at throughputs that spanned orders of magnitude, owing to the use of the same two-dimensional mixing design in devices of each scale. In fact, because the mixing units achieved rapid mixing in a very small footprint, there is potential for additional, massive scale-up. For example, more than 1,000 of the ~1 mm2units used here could fit in a single 4-inch silicon wafer for parallel operation, whereas more than 2,500 and 10,000 units could fit in a 6-inch and 12-inch silicon wafer, respectively. Thus, this strategy could be used to reach greater than 100,000 L / hr throughputs, require less than a cubic foot of space, and be created with standard semiconductor manufacturing equipment. Finally, the scale-up demonstrated by the HAR strategy was achieved in a straightforward manner. The formulation throughput could be increased in other ways, such as by increasing the flow rate through a given device or by connecting and operating multiple chips simultaneously. These approaches, however, run into quality and logistical constraints: A) increasing the flow rate will ramp up the fluidic pressures and shear within the device, potentially affecting product quality; and B) operating multiple devices simultaneously will require space and fluidic connections that scale linearly with the number of devices used. Alternatively, a device could be made with wider channels, but this may prevent adequate mixing, desirable CQAs, or continuous, clog-free operation from being achieved, especially if a sufficiently high channel aspect ratio is not created. The HAR strategy circumvents these issues by scaling the total flow rate proportionally to the channel height and degree of parallelization to produce comparable average velocities through devices of all scales. As a result, this approach achieves scale-up with similar input pressures, a comparable mixing footprint, and only one set of fluidic connections for each of the formulation input phases.

[0099] Furthermore, the HAR strategy demonstrated clog-free, continuous, and stable microfluidic operation. When the HAR strategy was followed, clogging was not observed, and devices could operate for extended periods without the degradation of CQAs or dramatic changes in the required operating pressures. Such stable operation correlated with the uniformity of CQAs over time, which was observed in devices operating at each of the different throughput scales. Additionally, a HAR device was used to produce LNPs with uniform CQAs over eight hours of continuous LNP manufacturing. To the authors’ knowledge, this is the longest demonstrated duration of continuous LNP production using a microfluidic chip. In contrast, unstable operation, clogging, and undesirable CQAs were observed when the HAR design criteria was not met and when SHMs were used for LNP formulation.

[0100] The capacity for scale-independent and continuous operation lends the HAR strategy to end-to-end integration with downstream processes (FIG. 25). Such downstream processes include TFF or SPTFF, further fluidic mixing, sterile filtration, fill finish, etc. (FIG. 25A). In these cases, it is important for the chip to accommodate the throughput required for the workflow, whether that consists of producing a small batch (e.g., less than lOmg) of LNPs (FIG. 21H) or continuous LNP manufacturing. In these cases, the output from the chip must occur at the desired flow rate, while controlling for the fluidic pressure and achieving target CQAs. The ability of the HAR strategy to specify and scale the throughput of a given microfluidic mixer aligns well with the requirements for end-to-end integration. For example, a P1-C1E1-0.8 device may be designed to produce an output of 500 mL / hr, or a Pl-Cl-3.2 device may be selected if a throughput on the order of 2L / hr is desired (FIG. 25B). In both devices, LNPs with comparable CQAs were produced at similar input pressures and requiring the same footprint. Furthermore, it was observed that LNP properties tend to plateau above a certain flow rate (FIG. 24); thus, one potential way to accommodate lower pressure requirements is to reduce the total flow rate to the minimum required in that HAR device to produce the desired LNPs (FIG. 25C). Finally, as a proof-of-concept, end-to-end integration was demonstrated by feeding the output from the microfluidic chip directly into the SPTFF module. The HAR strategy thus enabled the removal of an unnecessary reservoir and the reduction of the fluidic residence time within the closed system, important considerations for GMP process qualification.

[0101] Lastly, the HAR microfluidic strategy is readily compatible with GMP manufacturing due to the inherent cleanroom-based fabrication process, the lack of plastic components, and the reproducible operation that demonstrated across runs, over time, and across formulation throughputs. The silicon-based fabrication process was chosen for theHAR strategy because it enabled the user to specify the desired channel aspect ratio a priori depending on the desired throughput, and then readily fabricate those highly rectangular channels within the glass-based devices. This technique overcomes limitations with other techniques, such as injection molding with plastics, where high aspect ratio, rectangular channels are difficult to produce. Furthermore, plastic devices contain additives, residues, and other byproducts of the injection molding process that may leach into the drug product, especially over long-term operation and in the presence of organic solvents. An alternative to injection molding is micro-machining stainless steel, though this process also faces limitations with producing micron-scale high aspect ratio and rectangular channels, as well as due to the susceptibility to surface defects. Silicon-based microfluidics thus provides a unique solution for the HAR strategy: it can be used to produce precisely defined microfluidic devices with high aspect ratio, micron-scale channels; it does not contain organic leachable materials; and the devices may be produced with standard semiconductor manufacturing equipment and in commercial foundries. Note that alternative methods to form channels in glass, such as isotropic wet etching, cannot produce the high aspect ratio channels described here. Finally, in addition to fabrication details, CQAs produced with the HAR strategy were reproducible across process runs, over formulation throughputs, and over time. The GMP- readiness of the HAR strategy is supported by this reproducibility, the material properties of the device, and the ability to use the same number of wetted fluidic connections regardless of the throughput.Device fabrication

[0102] A chips, consisting of microfluidic channels fully enclosed within permanently bonded silicon and glass, may be formed by etching two layers in silicon: mixing channels (Layer 1) and through-silicon vias (Layer 2); in the competed chips, the vias in Layer 2 provide access to the mixing channels (Layer 1) from the backside of the silicon wafer. For each layer, photomasks were designed using AutoCAD 2022 (Autodesk, Inc.) and created by writing the design files on chrome-coated soda lime photomasks (AZ 1500) using a DWL 66+ mask writer (Heidelberg Instruments) with a 10mm write head. Following exposure, the photomasks were developed in MF 319 developer for one minute and immersed in chrome etchant for one minute. Residual photoresist was removed via sonication in Microposit Remover 1165 (Dow) for five minutes at 60°C. During the fabrication process, the two layers were lithographically patterned, and the channels were etched using DRIE (SPTS Rapier Si DRIE). For all layers, SI 805 photoresist was mixed with acetone (1 :8) and the resist was spray-coated (Suss Microtech) to the required thickness; after developing the photoresist, the wafers were then cleaned in a spin rinse dryer in preparation for DRIE.

[0103] To prepare the first layer, the wafer was spray-coated with 16pm of photoresist and soft baked at 90°C for four minutes. Photolithography was performed with the photomask for the mixing channels (Layer 1), and the wafer was left at room temperature for one hour for rehydration.

[0104] The wafer was then developed in MF 319 for two minutes, cleaned in a spin rinse dryer, and kept at 100°C for five minutes. After cleaning the wafer again in the spin rinse dryer, the mixing channels were etched to the target etch depth in DRIE. The etched wafer was kept in Nanostrip for one hour, and cleaned in the spin rinse dryer, after which 10pm of SiO2 was deposited on the wafer using plasma enhanced chemical vapor deposition. To prepare the through-silicon vias (Layer 2), the wafer was spray-coated with 12pm of photoresist and soft baked at 130°C for five minutes. Photolithography was performed with the photomask for the through-silicon vias, and the wafer was left at room temperature for 10 minutes, developed in MF 319 for two minutes, cleaned in the spin rinse dryer, and kept at 100°C for a further five minutes. After an additional cleaning of the wafer in the spin rinse dryer, the through-silicon vias were etched with DRIE.

[0105] Finally, the processed wafer was permanently bonded to a 4” diameter Borofloat 33 glass wafer to encapsulate the microfluidic channels. To accomplish this, the wafers were cleaned in acetone, isopropanol, and deionized water for 5 minutes each, immersed in nanostrip and piranha solution for one hour each, rinsed in deionized water for five minutes, and cleaned in the spin rinse dryer. The cleaned wafers were anodic bonded by applying a force of 100N and a voltage of 800V for an hour in an EVG 510 anodic bonding tool. The bonded wafers were then diced into individual microfluidic chips using an ADT 7100 Dicing Saw.Device operation

[0106] Microfluidic chips were integrated with the pumps, sensors, and other fluidic components via a custom-designed aluminum or stainless-steel chip holder. This chip holder incorporated built-in fluidic connections for integrating the tubing and provided a fluid-tight seal via inert gaskets.

[0107] Fluid was delivered to the devices by either A) pressurizing reagent vessels and using an OBI pressure-driven flow controller (Elveflow, Paris, France) communicating with flow sensors (Bronkhorst), or B) using Quattroflow QF30 or QF150 diaphragm pumps (PSG Dover, Oakbrook Terrace, IL). The total flow rate of operation was determined by the experiment and ranged from 100 mL / hr to 15 L / hr. The flow rate ratio, defined as the ratio of the aqueous phase to the ethanol phase, was typically 3: 1. Fluidic operation of the device wasvisualized with a DM2500 upright microscope (Leica) integrated with a CMOS camera (ThorLabs CS126CU).Mixing quantification

[0108] Mixing was visualized by operating the microfluidic devices with deionized water and dye-loaded ethanol at 3 : 1 flow rate ratio and the desired total flow rates. The dye used was rhodamine B dissolved to 5-20 pM in ethanol. Images were acquired at the desired flow conditions and the red channel of the RGB image was isolated for subsequent analysis. The acquired image was then registered to an image of the device filled with ethanol-dye, and grey values of the acquired image were normalized by the measured grey values of an image with just the dye to account for any variations in measured intensity across the device or image. Channel cross-sections perpendicular to the direction of flow were obtained from this normalized image at the desired location and mixing was quantified with Equation 1, a formula used in the literature.Here, xtrefers to the pixel in the i ’th position of the channel cross-section, xj represents the mean of all pixels in the channel cross section, and N is the total number of grey values in the cross-section.LNP materials and formulation

[0109] The lipid solution typically used in the LNP formulation examples described herein consisted of MC3, 1,2-DSPC, Cholesterol, and DMG-PEG(2000) dissolved in ethanol at a molar ratio of 40: 10:48:2 (MC3: 1,2-DSPC :Cholesterol:DMG-PEG(2000)). Lipid solutions using SM-102 consisted of SM-102, 1,2-DSPC, Cholesterol, and DMG-PEG(2000) dissolved in ethanol at a molar ratio of 50: 10:38.5: 1.5 (SM-102: l,2-DSPC:Cholesterol:DMG- PEG(2000)). Lipid solutions using ALC-0315 consisted of ALC-0315, 1,2-DSPC, Cholesterol, ALC-0159 dissolved in ethanol at a molar ratio of 46.3:9.4:42.7: 1.6 (ALC- 0315: l,2-DSPC:Cholesterol:ALC-0159). The appropriate lipid solution was rapidly mixed with 100 mM citrate buffer, pH 6.0 containing PolyA. Unless otherwise stated, LNP formulation was typically carried out at a 3 : 1 flow rate ratio of aqueous to ethanol streams, Source 1 was used for PolyA, and LNPs with a final PolyA concentration of 240pg / mL were formed at a final N / P ratio of 6. When formulating LNPs with varying ratios of DMG- PEG(2000), the molar ratios of the standard MC3 formulation were changed to (41.5: 10.5:47: 1) for 1% PEG and (40: 10:45:5) for 5% PEG.Particle-size characterization

[0110] Formulated LNPs were typically diluted 10-15X in IX PBS immediately after formulation. The size of these diluted LNPs were analyzed using dynamic light scattering using a Zetasizer Ultra Red label DLS (Malvern Panalytical) according to the manufacturer’s specifications. Size and PDI measurements were performed in backscatter mode. The built-in refractive index of water was used for the refractive index of the dispersant. For comparison, LNPs were dialyzed against IX PBS for two hours and particle size was measured in the same way. Comparing the measured sizes of LNPs produced in this manner often resulted in differences in Z-average diameter of less than 5nm.Encapsulation efficiency quantification

[0111] Encapsulation efficiency experiments were performed using the Quant-it™ RiboGreen RNA Assay (Invitrogen, Waltham, MA) according to the manufacturer’s specifications. Briefly, a positive reference sample was prepared by diluting formulation buffer and pure ethanol in IX PBS to the same dilution factor as during the formulation experiment. A negative reference sample was prepared by diluting formulating buffer, without Poly A, in the same way as that of the positive reference sample. The positive reference sample contained PolyA at the same concentration of total PolyA as in the LNPs after formulation and dilution, but without the lipids; the negative reference sample did not contain PolyA. These two references samples were diluted in A) TE buffer, and B) TE buffer + 0.5% Triton-X, and separate standard curves for TE buffer and TE buffer + Triton-X were prepared in a 96-well microplate. To perform the assay, formulated LNPs were diluted separately in TE buffer and TE buffer + Triton-X, these samples were incubated for five minutes in darkness, and fluorescence measurements were performed with a Hl Synergy (Agilent) or CLARIOstar (BMG-Labtech) microplate reader. Encapsulation efficiency was quantified from the measured fluorescence of a well containing the LNP sample diluted in Triton, xTriton, and a different well corresponding to the same sample diluted only in TE buffer, as in Equation 2.EE=fxTriton-xNo Triton\x 10 Q%(2)'xTriton 'Three replicates were performed for each measurement.Single-pass tangential flow filtration

[0112] SPTFF was first performed with a 7-in-series, Cadence T02 SPTFF module (Cytiva, Marlborough, MA). In both modes of integrating the SPTFF module with a HAR chip, i.e., with an intermediate reservoir and without an intermediate reservoir, the formulated LNPs were diluted 12X in IX PBS prior to SPTFF. For operation with an intermediatereservoir, feed material was input into the module via a diaphragm pump (Quattroflow, QF150) at a flow rate of 22L / hr. For operation without an intermediate reservoir, the microfluidic chip was operated at 2100mL / hr, IX PBS was mixed with the output from the chip via a T-junction, and the combined flow of approximately 25L / hr was fed directly to the SPTFF module. SPTFF was also carried out with 0.1m2Ultracel lOOkDa C-Screen Pellicon capsules (Millipore Sigma, Burlington, MA). When three capsules were arranged in series, feed flow rates of 300-400mL / min were delivered to the module. When one capsule was used, a feed flow rate of 140mL / min was delivered to the module. In both cases, an intermediate reservoir was used and 12X concentration factors were achieved by modulating a backpressure valve on the retentate. Output flow rates in all cases were measured by flow sensor or by manually measuring the volume of the output from the chip.

[0113] All publications and patent applications mentioned in this specification are herein incorporated by reference in their entirety to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference. Furthermore, it should be appreciated that all combinations of the foregoing concepts and additional concepts discussed in greater detail below (provided such concepts are not mutually inconsistent) are contemplated as being part of the inventive subject matter disclosed herein and may be used to achieve the benefits described herein.

[0114] Any of the methods (including user interfaces) described herein may be implemented as software, hardware or firmware, and may be described as a non-transitory computer-readable storage medium storing a set of instructions capable of being executed by a processor (e.g., computer, tablet, smartphone, etc.), that when executed by the processor causes the processor to control perform any of the steps, including but not limited to: displaying, communicating with the user, analyzing, modifying parameters (including timing, frequency, intensity, etc.), determining, alerting, or the like. For example, any of the methods described herein may be performed, at least in part, by an apparatus including one or more processors having a memory storing a non-transitory computer-readable storage medium storing a set of instructions for the processes(s) of the method.

[0115] While various embodiments have been described and / or illustrated herein in the context of fully functional computing systems, one or more of these example embodiments may be distributed as a program product in a variety of forms, regardless of the particular type of computer-readable media used to actually carry out the distribution. The embodiments disclosed herein may also be implemented using software modules that perform certain tasks. These software modules may include script, batch, or other executable files that may be stored on a computer-readable storage medium or in a computing system. In someembodiments, these software modules may configure a computing system to perform one or more of the example embodiments disclosed herein.

[0116] As described herein, the computing devices and systems described and / or illustrated herein broadly represent any type or form of computing device or system capable of executing computer-readable instructions, such as those contained within the modules described herein. In their most basic configuration, these computing device(s) may each comprise at least one memory device and at least one physical processor.

[0117] The term “memory” or “memory device,” as used herein, generally represents any type or form of volatile or non-volatile storage device or medium capable of storing data and / or computer-readable instructions. In one example, a memory device may store, load, and / or maintain one or more of the modules described herein. Examples of memory devices comprise, without limitation, Random Access Memory (RAM), Read Only Memory (ROM), flash memory, Hard Disk Drives (HDDs), Solid-State Drives (SSDs), optical disk drives, caches, variations or combinations of one or more of the same, or any other suitable storage memory.

[0118] In addition, the term “processor” or “physical processor,” as used herein, generally refers to any type or form of hardware-implemented processing unit capable of interpreting and / or executing computer-readable instructions. In one example, a physical processor may access and / or modify one or more modules stored in the above-described memory device. Examples of physical processors comprise, without limitation, microprocessors, microcontrollers, Central Processing Units (CPUs), Field-Programmable Gate Arrays (FPGAs) that implement softcore processors, Application-Specific Integrated Circuits (ASICs), portions of one or more of the same, variations or combinations of one or more of the same, or any other suitable physical processor.

[0119] Although illustrated as separate elements, the method steps described and / or illustrated herein may represent portions of a single application. In addition, in some embodiments one or more of these steps may represent or correspond to one or more software applications or programs that, when executed by a computing device, may cause the computing device to perform one or more tasks, such as the method step.

[0120] In addition, one or more of the devices described herein may transform data, physical devices, and / or representations of physical devices from one form to another. Additionally or alternatively, one or more of the modules recited herein may transform a processor, volatile memory, non-volatile memory, and / or any other portion of a physical computing device from one form of computing device to another form of computing deviceby executing on the computing device, storing data on the computing device, and / or otherwise interacting with the computing device.

[0121] The term “computer-readable medium,” as used herein, generally refers to any form of device, carrier, or medium capable of storing or carrying computer-readable instructions. Examples of computer-readable media comprise, without limitation, transmission-type media, such as carrier waves, and non-transitory-type media, such as magnetic-storage media (e.g., hard disk drives, tape drives, and floppy disks), optical-storage media (e.g., Compact Disks (CDs), Digital Video Disks (DVDs), and BLU-RAY disks), electronic-storage media (e.g., solid-state drives and flash media), and other distribution systems.

[0122] A person of ordinary skill in the art will recognize that any process or method disclosed herein can be modified in many ways. The process parameters and sequence of the steps described and / or illustrated herein are given by way of example only and can be varied as desired. For example, while the steps illustrated and / or described herein may be shown or discussed in a particular order, these steps do not necessarily need to be performed in the order illustrated or discussed.

[0123] The various exemplary methods described and / or illustrated herein may also omit one or more of the steps described or illustrated herein or comprise additional steps in addition to those disclosed. Further, a step of any method as disclosed herein can be combined with any one or more steps of any other method as disclosed herein.

[0124] The processor as described herein can be configured to perform one or more steps of any method disclosed herein. Alternatively or in combination, the processor can be configured to combine one or more steps of one or more methods as disclosed herein.

[0125] When a feature or element is herein referred to as being "on" another feature or element, it can be directly on the other feature or element or intervening features and / or elements may also be present. In contrast, when a feature or element is referred to as being "directly on" another feature or element, there are no intervening features or elements present. It will also be understood that, when a feature or element is referred to as being "connected", "attached" or "coupled" to another feature or element, it can be directly connected, attached or coupled to the other feature or element or intervening features or elements may be present. In contrast, when a feature or element is referred to as being "directly connected", "directly attached" or "directly coupled" to another feature or element, there are no intervening features or elements present. Although described or shown with respect to one embodiment, the features and elements so described or shown can apply to other embodiments. It will also beappreciated by those of skill in the art that references to a structure or feature that is disposed "adjacent" another feature may have portions that overlap or underlie the adjacent feature.

[0126] Terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. For example, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items and may be abbreviated as " / ".

[0127] Spatially relative terms, such as "under", "below", "lower", "over", "upper" and the like, may be used herein for ease of description to describe one element or feature's relationship to another element(s) or feature(s) as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if a device in the figures is inverted, elements described as "under”, or "beneath" other elements or features would then be oriented "over" the other elements or features. Thus, the exemplary term "under" can encompass both an orientation of over and under. The device may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein interpreted accordingly. Similarly, the terms "upwardly", "downwardly", "vertical", "horizontal" and the like are used herein for the purpose of explanation only unless specifically indicated otherwise.

[0128] Although the terms “first” and “second” may be used herein to describe various features / elements (including steps), these features / elements should not be limited by these terms, unless the context indicates otherwise. These terms may be used to distinguish one feature / element from another feature / element. Thus, a first feature / element discussed below could be termed a second feature / element, and similarly, a second feature / element discussed below could be termed a first feature / element without departing from the teachings of the present invention.

[0129] In general, any of the apparatuses and methods described herein should be understood to be inclusive, but all or a sub-set of the components and / or steps may alternatively be exclusive and may be expressed as “consisting of’ or alternatively “consisting essentially of’ the various components, steps, sub-components or sub-steps.

[0130] As used herein in the specification and claims, including as used in the examples and unless otherwise expressly specified, all numbers may be read as if prefaced by the word "about" or “approximately,” even if the term does not expressly appear. The phrase “about” or “approximately” may be used when describing magnitude and / or position to indicate that the value and / or position described is within a reasonable expected range of values and / or positions. For example, a numeric value may have a value that is + / - 0.1% of the stated value (or range of values), + / - 1% of the stated value (or range of values), + / - 2% of the stated value (or range of values), + / - 5% of the stated value (or range of values), + / - 10% of the stated value (or range of values), etc. Any numerical values given herein should also be understood to include about or approximately that value, unless the context indicates otherwise. For example, if the value " 10" is disclosed, then "about 10" is also disclosed. Any numerical range recited herein is intended to include all sub-ranges subsumed therein. It is also understood that when a value is disclosed that "less than or equal to" the value, "greater than or equal to the value" and possible ranges between values are also disclosed, as appropriately understood by the skilled artisan. For example, if the value "X" is disclosed the "less than or equal to X" as well as "greater than or equal to X" (e.g., where X is a numerical value) is also disclosed. It is also understood that the throughout the application, data is provided in a number of different formats, and that this data, represents endpoints and starting points, and ranges for any combination of the data points. For example, if a particular data point “10” and a particular data point “15” are disclosed, it is understood that greater than, greater than or equal to, less than, less than or equal to, and equal to 10 and 15 are considered disclosed as well as between 10 and 15. It is also understood that each unit between two particular units are also disclosed. For example, if 10 and 15 are disclosed, then 11, 12, 13, and 14 are also disclosed.

[0131] Although various illustrative embodiments are described above, any of a number of changes may be made to various embodiments without departing from the scope of the invention as described by the claims. Optional features of various device and system embodiments may be included in some embodiments and not in others. Therefore, the foregoing description is provided primarily for exemplary purposes and should not be interpreted to limit the scope of the invention as it is set forth in the claims.

[0132] The examples and illustrations included herein show, by way of illustration and not of limitation, specific embodiments in which the subject matter may be practiced. As mentioned, other embodiments may be utilized and derived there from, such that structural and logical substitutions and changes may be made without departing from the scope of this disclosure. Such embodiments of the inventive subject matter may be referred to hereinindividually or collectively by the term “invention” merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept, if more than one is, in fact, disclosed. Thus, although specific embodiments have been illustrated and described herein, any arrangement calculated to achieve the same purpose may be substituted for the specific embodiments shown. This disclosure is intended to cover any and all adaptations or variations of various embodiments. Combinations of the above embodiments, and other embodiments not specifically described herein, will be apparent to those of skill in the art upon reviewing the above description.

Claims

CLAIMSWhat is claimed is:

1. A microfluidic mixer apparatus configured to achieve 90% or more mixing efficiency within less than 1 second at 75 psi or less at a flow rate of between 100 ml / hour and greater than 2 L / hour without clogging, the apparatus comprising: two or more inputs; an output; a channel extending between the two or more inputs and the output that is planar and curving, the channel having a rectangular cross-section transverse to a length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, further wherein the rectangular cross-section has a constant cross-sectional area along the length from the two or more inputs to the output.

2. The apparatus of claim 1, wherein the channel comprises between 2 and 20 stages.

3. The apparatus of claims 1 or 2, wherein a stage length of each stage is between 8 and 10 times a width of the channel in the plane of the microfluidic mixer.

4. The apparatus of any of claims 1-3, wherein the inner radius of curvature is between 20 pm and 200 pm.

5. The apparatus of any of claims 1-4, wherein the outer radius of curvature is between 40 pm and 400 pm.

6. The apparatus of any of claims 1-5, wherein a width of the channel in the plane of the microfluidic mixer is between 20 pm and 200 pm.

7. The apparatus of any of claims 1-6, wherein the inner radius of curvature is approximately equal to a width of the channel in the plane of the microfluidic mixer.

8. The apparatus of any of claims 1-7, wherein the outer radius of curvature is approximately equal to twice a width of the channel in the plane of the microfluidic mixer.

9. The apparatus of any of claims 1-8, wherein each of the two or more inputs is between 0 and 90 degrees relative to an entrance into the channel.

10. The apparatus of any of claims 1-9, wherein a height of the channel perpendicular to the plane of the microfluidic mixer is between 40 pm and 700 pm.

11. The apparatus of any of claims 1-10, wherein the aspect ratio of the rectangular crosssection is between 0.4 and 4 or greater.

12. The apparatus of any of claims 1-11, wherein the aspect ratio is 0.7 or greater.

13. The apparatus of any of claims 1-12, the channel is formed in a silica-based substrate.

14. The apparatus of any of claims 1-13, wherein the output has a larger width than the channel.

15. A microfluidic mixer apparatus configured to achieve 90% or more mixing efficiency within less than 1 second at 75 psi or less at a flow rate of between 100 ml / hour and greater than 2 L / hour without clogging, the apparatus comprising: a planar substrate; two or more inputs formed in the planar substrate; an output formed in the planar substrate; a channel through the planar substrate and extending along a length between the two or more inputs and the output that is curving, the channel having a rectangular cross-section traverse to the length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, further wherein the rectangular cross-section has a constant cross-sectional area along the length of the channel from the two or more inputs to the output.

16. A method of microfluidic mixing, the method comprising:inputting a first fluid into a first input of a microfluidic mixer apparatus; inputting a second fluid into a second input of the microfluidic mixer apparatus; passing the first and second fluids from the first and second inputs in a channel extending along a curving length to an output at a flow rate of between 100 ml / hour and greater than 2 L / hour, wherein the channel has a rectangular cross-section traverse to the length and an aspect ratio of greater than 0.4, and wherein the channel comprises one or more stages, further wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages, so that multiple vertical vortexes are formed within the channel as the fluids are passed along the length of the channel to achieve 90% or more mixing efficiency of the first and second fluids within less than 1 second at 75 psi or less.

17. The method of claim 16, wherein the rectangular cross-section of the channel has a constant cross-sectional area along the length of the channel from the first and second inputs to the output, to prevent clogging.

18. The method of any of claims 16-17, wherein the first fluid comprises a nanoparticle.

19. The method of any of claims 16-18, wherein the second fluid comprises a polynucleotide.

20. The method of any of claims 16-19, wherein passing the first and second fluids comprises encapsuling a polynucleotide within a nanoparticle while mixing the first and second fluids.

21. The method of any of claims 16-20, further comprising passing the first and second fluids continuously for greater than one hour.

22. The method of any of claims 16-21, further comprising passing the first and second fluids continuously for greater than six hours.

23. The method of any of claims 16-22, wherein passing comprises passing through between 2 and 20 stages of the channel.

24. The method of any of claims 16-23, wherein passing comprises passing down each of the one or more stages having a stage length for each stage of between 8 and 10 times a width of the channel.

25. The method of any of claims 16-24, wherein passing comprises passing down the channel having an inner radius of curvature is between 20 pm and 200 pm.

26. The method of any of claims 16-25, wherein passing comprises passing down the channel having an outer radius of curvature is between 40 pm and 400 pm.

27. The method of any of claims 16-26, wherein passing comprises passing down the channel having a width of the channel between 20 pm and 200 pm.

28. The method of any of claims 16-27, wherein passing comprises passing down the channel in which the inner radius of curvature is approximately equal to a width of the channel.

29. The method of any of claims 16-28, wherein passing comprises passing down the channel in which the outer radius of curvature is approximately equal to twice a width of the channel.

30. The method of any of claims 16-29, wherein passing comprises passing down the channel in which a height of the channel is between 40 pm and 700 pm.

31. The method of any of claims 16-30, wherein passing comprises passing down the channel in which the aspect ratio of the rectangular cross-section is between 0.4 and 4 or greater.

32. The method of any of claims 16-31, wherein passing comprises passing down the channel in which the aspect ratio is 0.7 or greater.

33. A method, the method comprising: receiving a target set of parameters for a downstream microfluidics process comprising one or more process inputs; and outputting a microfluidics chip design, wherein the microfluidics chip design comprises a mixing channel extending between the two or more inputs and an output, wherein the mixing channel is planar and curving and has a rectangular cross-section transverse to a length of the channel, the rectangular cross-section having an aspect ratio of greater than 0.4, wherein the mixing channel has an output that is configured to match the required process inputs, and wherein the mixing channel comprises a channel aspect ratio configured to achieve a 90% or more mixing efficiency using the one or more process inputs.

34. The method of claim 33, further comprising providing the microfluidics chip based on the microfluidics chip design.

35. The method of any of claims 33-34, wherein the one or more process inputs comprises one or more of: flow rates and pressures.

36. The method of any of claims 33-35, wherein the one or more process inputs comprises a pressure 75 psi or less.

37. The method of any of claims 33-36, wherein the one or more process inputs comprises a flow rate of between 100 ml / hour and greater than 2 L / hour.

38. The method of any of claims 33-37, wherein the channel comprises one or more stages, wherein, for each stage, the channel has an inner radius of curvature and an outer radius of curvature that each have a constant magnitude and wherein the inner and outer radius of curvature invert along the length between adjacent stages.

39. The method of any of claims 33-38, further wherein the rectangular cross-section has a constant cross-sectional area along the length from the two or more inputs to the output.

40. The method of any of claims 33-39, further comprising fabricating the microfluidics chip by deep reactive ion etching (DRIE).

Images