Jump to content
[dismiss]
Molecupedia is a parody encyclopedia written by Molecular Reality Corporation. It is not Wikipedia and is not affiliated with the Wikimedia Foundation. About

Microfluidics

Page protected
From Molecupedia, the encyclopedia of very small things
This is the current revision of this page. The present address (URL) is a permanent link to this version.
Science and technology of fluids in channels tens to hundreds of micrometers wide

Microfluidics is the study of liquids in channels with at least one dimension below about a millimeter, usually tens to hundreds of micrometers, and the technology built on it. At that size the physics of a liquid changes character. Flow is smooth and layered, two streams side by side do not mix except by diffusion, a bubble acts as a spring, and the walls matter more than the bulk. A nanopore instrument has to bring salt solution and sample to a membrane a few tens of nanometers thick without breaking it, and the pipework that does this is microfluidic.

Physics

Laminar flow

The ratio of inertial to viscous forces in a flow is the Reynolds number, Re = ρvDh/μ, where ρ is the density of the liquid, v its mean speed, Dh the hydraulic diameter of the channel and μ the viscosity. Pipe flow turns turbulent at a Reynolds number of the order of 2,000. Channels a few tens of micrometers wide carrying water at ordinary speeds sit at 1 to 10 or well below, where viscosity decides everything and the flow is laminar. Edward Purcell's account of swimming at such numbers, in which a swimmer that reverses its motion only retraces its path, has become the standard introduction.[1]

The pressure needed to drive a flow Q through a channel is ΔP = Q Rhyd, an analogue of Ohm's law with flow in place of current. For a round tube the hydraulic resistance is 128μL/(πD4), so halving the diameter multiplies it by 16. For a rectangular channel of width w and depth h, with h smaller than w, it is approximately 12μL/[wh3(1 − 0.63h/w)], which rises with the cube of the depth.[2] A liquid in a channel also has a capacitance: a bubble of volume V trapped at atmospheric pressure P0 gives the system a compliance of about V/P0, and with the resistance this produces a time constant. A single microliter of air behind a resistance of 4.5 × 1014 Pa s/m3 takes about 4 seconds to settle.[a]

Diffusion

A rectangular channel cross-section 50 micrometers wide and 25 micrometers deep, and a top view of two streams flowing side by side with a blurred boundary for potassium ions and a sharp one for a DNA coil
Laminar flow and diffusion in the channel of the MR1 design note: 50 µm wide, 25 µm deep, 10 µL/min. Two streams side by side stay separate for the 0.15 s a molecule takes to travel 20 mm; in that time a potassium ion diffuses sideways about 24 µm and a DNA coil about 0.4 µm. The drawing is computed from the stated values and is not to scale along the channel.

Without turbulence, mixing is diffusion alone. A particle with diffusion coefficient D spreads over a distance x in a time of about x2/2D. A potassium ion, with D near 2 × 10−9 m2/s, crosses a channel 50 µm wide in about 0.6 s. A coil of λ-DNA, with D near 0.5 µm2/s, takes about 2,500 s, which is 42 minutes.[b] The ratio of transport by flow to transport by diffusion is the Péclet number, vw/D. At 0.13 m/s in a 50 µm channel it is about 3,400 for the ion and 13 million for the DNA, A stream of ions mixes across the channel within about 9 cm of travel at that speed, while the DNA would need about 330 m, so two streams of different molecules run side by side through any chip.

History

An early microfabricated fluidic device, often counted as the first, was a gas chromatograph etched into a silicon wafer, described by Terry, Jerman and Angell in 1979.[3] In 1990 Manz, Graber and Widmer proposed the "miniaturized total chemical analysis system", a whole laboratory on a chip.[4] In 1998 a group at Harvard showed how to design and make a microfluidic device in the silicone elastomer PDMS in less than 24 hours, against the days or weeks that silicon and glass need.[5] In 2000 stacked layers of elastomer gave valves and pumps made entirely of soft material, which shrank device areas by more than two orders of magnitude compared with silicon.[6] George Whitesides's review of 2006 defines the field as the manipulation of fluids in channels with dimensions of tens of micrometers, and a review by Squires and Quake of 2005 gives the physics.[7][8]

Pumps

Fluids are moved by pressure from a syringe or a peristaltic pump, by an electric field in electroosmotic flow, by capillary action, or by rotating the chip. A peristaltic pump squeezes a flexible tube with rollers, which makes it simple and sterile and also delivers the liquid in pulses. Where the signal is a picoampere current, the pulses can be seen in it, and recordings are often made with the flow stopped.

Fluidics for nanopore measurements

A flow cell for a solid-state nanopore has two chambers, one on each side of the membrane, each with an electrode and each with inlet and outlet ports. The tasks are to fill both without trapping air, exchange the buffer, add the sample on one side only and keep the pressure on the two sides equal, since the membrane is a few tens of nanometers thick. Wetting is the first obstacle. A pore a few nanometers wide does not always fill by itself, because air held in it by surface tension can stay in place (see Nanopore wetting and Nanobubble).

An exchange also takes longer than the volume suggests. In a well-mixed chamber of volume V flushed at flow Q, the old solution falls as e−Qt/V. A chamber of 12.5 µL flushed at 10 µL/min is replaced by one volume of fresh liquid every 75 seconds, and 63% of the old solution is gone after that time. Five volumes, 6 minutes and 15 seconds, leave 0.7%.

A leak is a fluidic failure that shows up in the electronics. The open design of the MR1 Nanopore Reader surrounds the wet zone with a perimeter moat of copper, 0.25 mm from the ground plane, whose resistance is above 10 GΩ when dry and falls when salt solution reaches it; an October 2026 engineering change specifies that a leak cuts the power of the converter, and that the firmware then stops the pump driver within 10 ms (see Weep moat).[9]

At Molecular Reality

The MR1 fluidics module

The MR1 design pairs the reader with a pump board. Two peristaltic pumps, each driven by a DRV8837C motor driver, deliver 45 µL per motor revolution over a range of 0.5 to 150 µL/min. Two pressure sensors read 0 to 15 psi on the feed side and 0 to 5 psi on the other, and the firmware stops the pumps at 10 psi and 4 psi.[10] At the nominal 10 µL/min the pump turns once every 4.5 minutes.

The repository's design note works through a channel 50 µm wide, 25 µm deep and 20 mm long, with 1 M potassium chloride taken as 1 mPa s. The flow is 0.133 m/s and the hydraulic diameter 33.3 µm. The note puts the Reynolds number at 4.4 × 10−3 and the pressure drop at 6.2 kPa. Recomputing from the note's own inputs gives a Reynolds number of 4.4, a thousand times larger, and a hydraulic resistance of 4.5 × 1014 Pa s/m3, about twelve times the note's figure, so the pressure drop is 75 kPa, 10.9 psi.[c] The flow is still laminar, and 10.9 psi is more than the 10 psi at which the pump firmware trips.[11] The same note requires the channels on the two sides of the wafer to be hydrodynamically balanced, so that no pressure difference reaches the membrane. The repository's status line reads "Designed. Simulated. Built and measured: not yet."[12]

The first laboratory

The fluidics of the company's first pores were done by hand. In a post of March 29, 2026, the founder recalls the best afternoon of data from the days at Edison Way in Menlo Park, California, a recording from a wet, stable 200 nm pore in a silicon chip that had cost about $100 and followed a week or two of what he calls "dumb problems". "The biggest, ugliest, dumbest one is: how do you get water through the pore?"[13] The company's term for such problems has its own article, Dumb problems.

See also

Notes

  1. ^ Calculated: τ = RhydV/P0 = (4.5 × 1014 Pa s/m3)(10−9 m3)/(1.01 × 105 Pa) = 4.4 s, for an isothermal gas. The resistance is the one worked out for the channel described below.
  2. ^ Calculated: (50 × 10−6 m)2/(2 × 1.96 × 10−9 m2/s) = 0.64 s for K+, and (50 × 10−6 m)2/(2 × 5 × 10−13 m2/s) = 2,500 s for λ-DNA. The diffusion coefficients are textbook values for water at 25 °C.
  3. ^ The Reynolds number is (1,000 kg/m3)(0.133 m/s)(33.3 × 10−6 m)/(10−3 Pa s) = 4.4. The resistance is 12(10−3 Pa s)(0.020 m)/[(50 × 10−6 m)(25 × 10−6 m)3(1 − 0.63 × 0.5)] = 4.5 × 1014 Pa s/m3, and 10 µL/min is 1.67 × 10−10 m3/s, which gives 75 kPa. Other documents in the repository give other channel sizes, so these figures are an example and not a property of the device.

References

  1. ^ Purcell, E. M. (1977). "Life at low Reynolds number". American Journal of Physics 45 (1): 3-11. doi:10.1119/1.10903.
  2. ^ Bruus, H. (2008). Theoretical Microfluidics. Oxford University Press. ISBN 978-0-19-923509-4.
  3. ^ Terry, S. C.; Jerman, J. H.; Angell, J. B. (1979). "A gas chromatographic air analyzer fabricated on a silicon wafer". IEEE Transactions on Electron Devices 26 (12): 1880-1886. doi:10.1109/T-ED.1979.19791.
  4. ^ Manz, A.; Graber, N.; Widmer, H. M. (1990). "Miniaturized total chemical analysis systems: a novel concept for chemical sensing". Sensors and Actuators B: Chemical 1 (1-6): 244-248. doi:10.1016/0925-4005(90)80209-I.
  5. ^ Duffy, D. C.; et al. (1998). "Rapid prototyping of microfluidic systems in poly(dimethylsiloxane)". Analytical Chemistry 70 (23): 4974-4984. doi:10.1021/ac980656z.
  6. ^ Unger, M. A.; et al. (2000). "Monolithic microfabricated valves and pumps by multilayer soft lithography". Science 288 (5463): 113-116. doi:10.1126/science.288.5463.113.
  7. ^ Whitesides, G. M. (2006). "The origins and the future of microfluidics". Nature 442 (7101): 368-373. doi:10.1038/nature05058.
  8. ^ Squires, T. M.; Quake, S. R. (2005). "Microfluidics: fluid physics at the nanoliter scale". Reviews of Modern Physics 77 (3): 977-1026. doi:10.1103/RevModPhys.77.977.
  9. ^ "Engineering Change Order: Weep Sense N-FET Buffer & Negative Excursion Clamp (ECO-2026-004)". GitHub. Molecular Reality Corporation. October 2026.
  10. ^ "config.h (Pump_firmware)". GitHub. Molecular Reality Corporation. October 2026.
  11. ^ "OpenAuto-CFD: Co-Design Architecture, 3D Visualizer & KiCad Plugin Guide". GitHub. Molecular Reality Corporation. October 2026.
  12. ^ "MR1 Nanopore Reader". GitHub. October 16, 2026.
  13. ^ Kemmishtree (March 29, 2026). "Focused Ion Beam pore". EQBio-OS.
Retrieved from "https://molecularreality.com/wiki/Microfluidics"
Last edited 1 hour ago by Debye Length Enjoyer