Nanopore wetting

Nanopore wetting is the step of getting liquid into a nanopore so that ions can cross it. A pore that is not wet is not a sensor, and a pore that is half wet is a noisy one: a bubble sitting in the opening changes the resistance, moves, and makes the current wander. Kent Kemmish, the founder of Molecular Reality Corporation, ranks it first among his dumb problems: "The biggest, ugliest, dumbest one is: how do you get water through the pore?"^[1] The literature agrees on the difficulty, and it has a few remedies, none of them certain.

Whether water enters a narrow channel by itself depends on the contact angle θ it makes with the wall. The pressure across the meniscus in a cylindrical pore of radius r is

where γ is the surface tension of water, 0.072 N/m. For a perfectly wetting wall (θ = 0) and a pore 5 nm in radius, water is pulled in with 28.8 MPa, about 284 atmospheres. For a wall with a contact angle of 110°, about that of Teflon, the sign reverses: water is held out until it is pushed with roughly 10 MPa, nearly 100 atmospheres. The pore on the company's live stream is 200 µm across, a radius of 100 µm, and for a wetting wall its pressure is 1.4 kPa, a twenty-thousandth of the nanopore's.^[2] The arithmetic is why a pore that gives no trouble at the micrometer scale becomes difficult below it.

A clean silicon nitride or silica surface is hydrophilic, so a clean pore might be expected to fill by itself. In practice air stays in it. Gas trapped in a pore can leave only by dissolving, and water that is already saturated with air, as water from a bottle or a tap is, cannot take it. Marion and colleagues found in 2021 that improper degassing of solutions raised the likelihood of wetting problems. When they used a solution that had not been degassed, the capacitance often returned close to its original value, which the authors took to mean that gas pushed in by compression had come back out as bubbles once the pressure was released.^[3]

The bubble is the common failure. In 2006 Smeets and colleagues moved solid-state pores through the focus of an infrared laser while recording the conductance. The current noise power of the same pore varied by up to five orders of magnitude between measurements, and the conductance showed an unexpected double peak. Both fit a model of a cylindrical pore containing a single nanobubble, and the authors concluded that nanometer-scale gas bubbles are the largest source of noise in solid-state nanopores.^[4] See Nanobubble.

Marion and colleagues showed the same thing from the other side. A silicon nitride pore 75 nm across, freshly treated with oxygen plasma, was obstructed, apparently by a bubble whose pinning the authors found unclear, and pressure differences of up to 3 bar switched it between the obstructed state and an open one. In the open state its resistance was 1.61 MΩ, within 10% of the value expected for the pore, so plasma cleaning had not kept the bubble out. They concluded that the lowest noise appears only once the whole pore region is wet.^[3] The same group monitored wetting on the membrane itself: the capacitance of a suspended membrane falls with the area that is not in contact with liquid, so a rise in capacitance under pressure shows the film wetting.

Pores with hydrophobic walls can keep water out for good reasons of thermodynamics. In molecular dynamics simulations of short hydrophobic pores 0.35 to 1.0 nm in radius, Beckstein and Sansom found water flickering on a nanosecond scale between a liquid-filled and a vapor-filled pore. Above a radius of about 0.55 nm the liquid state becomes the stable one and the vapor state vanishes.^[5] Biology uses the effect. Aryal, Sansom and Tucker reviewed the evidence that ion channels exploit "hydrophobic gating", in which a dewetted stretch of a narrow hydrophobic pore forms an energetic barrier to ions.^[6]

In 2011 Powell and colleagues showed that single hydrophobic pores in a polymer film switch reversibly between conducting and non-conducting states, by condensation and evaporation of water inside, under a transmembrane voltage.^[7] In 2023 a group used a hydrophobically gated protein pore as a memristor, a resistor with memory, which learned and forgot in the manner of a synapse.^[8] Thin membranes of molybdenum disulfide are hydrophobic enough to cause trouble, with water contact angles above 70° reported, and Marion and colleagues used the material to show how bubbles and wetting can distort a conductance measurement.^[3]

Protocols combine several measures. The first is cleaning. Solvent washes and oxygen plasma strip the organic films that make an oxide surface repel water, and so does piranha solution, a hot mixture of sulfuric acid and hydrogen peroxide. Plasma helps but does not guarantee anything, as the 75 nm pore above shows.

The second is to wet the pore with a liquid that wets it. Ethanol and isopropanol have a surface tension of about 22 mN/m against 72 for water, and enter a pore that water refuses; the pore is then exchanged into the aqueous electrolyte. Marion and colleagues, who tried a 50% ethanol exchange, give a simplified account of why the exchange can also go wrong: a liquid with more capacity for gas is replaced by one with less, and on a hydrophobic surface the first solvent can itself act as a catalyst for the nucleation of bubbles.^[3]

The third is to take out the gas. Degassed electrolyte dissolves residual pockets, and in a sealed cell built to hold more than 10 bar the same group compressed degassed solution to 7 bar and watched hydrophilic membranes fill within tens of seconds. They cite earlier work on microfluidic chips that found compression comparable in effectiveness to alcohol wetting.^[3][9]

The fourth is an electric field. A voltage can wet a pore, as Powell and colleagues showed, and the firmware of the open MR1 Nanopore Reader has a default pulse "for pore wetting / unclogging" of 800 mV for 250 ms.^[10] The fifth avoids the question. Pores made by controlled dielectric breakdown are formed in the electrolyte itself, so the liquid is there before the pore is.^[11]

The cleaning sequence in the design of the reader, solvent, water and a three-minute plasma at 30 W "for bubble-free priming", is described in Dumb problems. The first recordings in the company's public archive of live streams are of air.^[2]

In the company's account of what a finished instrument should be, "in a box" means "standardized, pre-wetted and self-calibrating, like a thermostat".^[12] Pre-wetting moves the step from the user's bench to the maker's.

• Solid-state nanopore
• Clogging (nanopore)
• Nanopore noise
• Nanopore fabrication
• Graphene nanopore