Water friction in nanofluidic devices made from two-dimensional crystals

This study employed microfluidic technologies to investigate the relationship between water–surface friction and the electronic, structural, and chemical properties of confining atomically flat materials in nanofluidic devices.

This short review is based on the publication entitled “Water friction in nanofluidic channels made from two-dimensional crystals” by Ashok Keerthi, Solleti Goutham, Yi You, Pawin Iamprasertkun, Robert A. W. Dryfe, Andre K. Geim & Boya Radha published in May 2021 in Nature Communications.

Abstract

Water friction at nanoscale is a topical issue in membrane-based applications from osmotic power generation, desalination and molecular separation. Decreasing water friction in nanosized channels would be very beneficial for those applications. Nevertheless, the processes at stake in fast water flows are yet to be elucidated. This work exploits nanometer scale capillaries made from atomically flat crystals to evaluate the effect of confining walls’ material on water friction.

The study stresses a strong difference observed between channels made from isostructural graphite and hexagonal boron nitride. This difference in fluid friction is attributed to different electrostatic and chemical interactions at the solid-liquid interface. To investigate further these differences, precision gravimetry and ion streaming measurements were performed to measure the slip length. The slip length, measure of water friction, was studied as a function of the electrical conductivity, wettability, surface charge and polarity of the confining walls. This study shows that water friction can be tuned with hybrid capillaries with different slip lengths at opposing walls.

Introduction to water friction study in nanofluidic devices

The challenges encountered in the displacement and flow of water from an inlet to an outlet through porous media and membranes can be found in many fields of application such as molecular transport in biology, osmotic power generation, membrane-based separation and water purification processes [1-5]

It was reported in the literature that the use of micro- or nano- 2D channels made of different materials can lead to water flow enhancement. This result was primarily attributed to the material property of the micro or nano device made of the atomically flat graphite, thus, limiting the friction at the channel wall. [6-9]. Nonetheless, due to a lack of controlled and reproducible experimental data, it remains unclear which factors and interactions contribute to the ultrafast water flow under a confined environment. 

Therefore, this work aims to investigate the relationship between water–surface friction and the electronic, structural, and chemical properties of confining atomically flat materials such as graphene, hexagonal boron nitride (hBN), and MoS2, based on the method previously developed with nanoscale capillaries. [10]

Aim & objectives

The primary objectives of this study are the following:

• Investigate the water-surface friction in various experimental conditions. 
• Study the evolution of water friction for various confining flat materials.

Materials & methods

The nanoscale capillary device is represented in Figure 1. It comprised three layers made from 2D crystals such as graphite, hBN or MoS2 stacked vua mechanical-transfer techniques. For more information, please refer to the original paper available here.  

Figure 1: a) Schematics of our capillary devices. b) Optical image of a hBN capillary device on a silicon nitride membrane (seen in light green). The top hBN layer is contoured with the dotted curve for visibility. Five parallel rectangular holes (light purple color) open to the other side of the wafer. The direction of the channels underneath the top hBN layer is indicated by the red dotted lines with a defined length L. Courtesy of the authors. CC. by 4.0.

The water flow measurement through the capillaries were performed by microgravimetry. As water evaporates through capillaries, a water weight loss was measured as a function of time.

In addition, contact-angle measurements were realised on the exfoliated 2D crystals. Lastly, streaming current measurements were made with custom-made electrochemical cells. Two 2mL reservoirs sealed with acid-resistant O-rings (James Walker UK, Ltd) were employed.

Two electrodes (Ag/AgCl) could be inserted through pressure inlet and outlet pipes (Fig. 3a). The pressure is applied via a microfluidic pump with step size of about 1 mbar (AF1 Dual, Elveflow) and run by using ESI elveflow software interface.

Streaming current measurements were performed as a function of the applied pressure varying from 0 to 250 mbar. The pressure is applied via the μm-size hole on SiNx substrate.

Key findings

In this work, the effect of several factors on the slip length were examined for two structurally similar atomically flat surfaces of graphite and hBN. 

Based on the accurate control and measurement performed with the Elveflow microfluidic pump, streaming experiments were performed. Those measurements showed that the graphite surface presents much lower friction (high slip) compared to hBN. 

The experiments suggested that hBN–water friction comes from electrostatic interaction of polar water molecules with OH− ions adsorbed on the heteropolar hBN surface, which may include the generation of immobile water clusters. [11]

Previous work suggested that all flat surfaces that are hydrophobic could provide little friction for water flow. This work highlights that friction is mainly governed by electrostatic and similar interactions of flowing molecules with confining surfaces

Figure 2: Schematics of the experimental  microfluidicsetup to measure ionic streaming currents. Dotted curves are a guide to the eye. c) Error bars indicate the SD of the measured streaming currents.  d) Error bars indicate the uncertainty in the fits of streaming currents vs. pressure. Courtesy of the authors. CC. by 4.0

As described in Figure 2, a known pressure ΔP is applied on KCl solution (equimolar on both sides of membrane) from one end and the pressure-induced current was measured by using Ag-AgCl electrodes. 

Figure 2 b) illustrates the ion streaming currents as a function of applied pressure, for a hybrid device (GB-G) shown in the inset. At a given pressure (shown normalized per 1 µm channel length; blue colored y-axis), Istr was recorded for one minute. As soon as the pressure is applied the current overshoots sharply and then stabilizes. 

The KCl concentration here was 0.3 M. In Figure c), for each concentration, Istr was normalized in the number of channels n and plotted against the pressure difference across the channel; the lines represent the best linear fits. 

Figure 2 d) depicts the streaming (electrokinetic) mobility µ for the GB-G device compared to the values obtained for graphite and hBN devices. GB-G and B-B devices show similar behavior. 

Conclusions

This work brought more insight into the relationship between water–surface friction and the properties of confining atomically flat materials in nanofluidic devices. The results obtained with the help of Elveflow’s autonomous microfluidic pump here are important for the development of nanofluidic channels providing ultrafast flows.

Indeed, this study is key for the evaporation-driven technologies such as distillation-based separation because it shows that a much higher driving pressures and ultrafast flows can be achieved by covering capillary exit surfaces with low-friction graphene layer.

The present work brings forward the many benefits of nanofluidics to design smart membranes and to mimic manifold machinery of biological channels.

References