CORDIS - EU research results

Active nanofluidics towards ionic machines

Periodic Reporting for period 3 - SHADOKS (Active nanofluidics towards ionic machines)

Reporting period: 2021-07-01 to 2023-03-31

Filtering and water purification rely traditionally on the concept of passive sieving across properly decorated nanopores. Such basic separation principle contrasts with the highly advanced membrane processes existing in Nature, which harness the full subtleties of active transport across channels. This involves advanced functions like ionic pumps, ultra-high selective channels, or voltage-gated nanopores, which all play a key role in many vital needs and neuronal functions.

The Shadoks project is at the frontier of the field of nanofluidics, which explores the transport of fluids and ionic species at the nanometric scales. Over the last 10 years, the domain has undergone a quantum leap with the development of individual artificial channels with nanometric and even sub-nanometric size with manifold geometries. Many ‘exotic’ properties have been indeed unveiled over the recent years [Bocquet, Nature Materials 2020].

The Shadoks projects builds on this flourishing know-how to push the frontiers of nanofluidics in order to invent and develop the concept of artificial ionic machines based on active nanofluidic transport.

The Shadoks project is an experimental project involving a strong theoretical counterpart, essential to experimental advances and prototyping. We investigate a wealth of strongly nonequilibrium transport phenomena occurring at the nanoscales, taking advantage of our unique know-how in building nanofluidic heterostructures, in particular made of carbon and boron-nitride. We target ionic Coulomb blockade, stimulated transport and gated nanopore, ionic pumps, dynamical osmosis. These processes allow to tune ionic fluxes against the gradients and induce out-of-equilibrium charge separation, hereby conceiving active sieving as a novel route for separation and desalination. Those new building blocks will subsequently be assembled to create advanced bioinspired membrane functionalities. Furthermore, we use the active nanofluidics building blocks to mimic a basic machinery of neuronal processes toward primitive artificial ionic machines.
Since the beginning of the project, manifold tasks towards the development of active nanofluidics have been accomplished. However, several key milestones have been already achieved:

(1) First, we have developed a detailed understanding of the ionic Coulomb blockade [Kavokine et al., Nature Nano 2019], showing that Coulomb blockade occurs in ionic systems due to a fractional Wien effect, associated with the dissociation of Bjerrum pairs with the gate charge. This mechanism leads unexpectedly to quantized ionic transport in a non-quantum system. Furthermore this active device paves the way to fabricate artificial ionic pumps.

(2) On the theoretical side, we made a breakthrough in the understanding of the ‘bizarre’ water-carbon interface. It emerges from our (and others) experimental reports in distinct systems that water behaves in a most peculiar way close to graphitic interfaces, with numerous examples showing how graphite ‘outperforms’, in some way or the other, alternative confining materials. We have demonstrated that these specific properties take their roots in quantum effects at the water carbon interface. This is a groundbreaking step and in the context of active nanofluidics, reveiling that a ‘quantum engineering’ of fluid and ion transport is possible. This paper was published in Nature in 2022.

(3) Second, a strong experimental effort has been achieved to develop active nanofluidic devices, in particular the so-called “Single Digit Nanopores” (SDN), with a size below 10nm, which have highlighted many peculiar transport properties. This relies on our expertize to fabricate 1D and 2D nano-assemblies. A key experimental result in both 1D and 2D SDNs is the demonstration of stimulated transport [Mouterde et al., Nature 2019; Marcotte et al., Nature Materials, 2020; Emmerich, Nature Materials 2022]. The ionic conduction is shown to be non-linearly modulated by the mechanical stimuli, here the applied pressure, akin a mechanical transistor. These responses echo directly the behaviour of some biological channels, which exhibit activated responses under various stimuli, such as the mechanosensitive channels such as Piezos, which are involved in touch sensing and in hearing.

(4) Then, we have thoroughly explored the non-linear ionic response in two dimensions - in direct line with the experimental systems – and we have unveiled a very specific Wien effect in the 2D ionic systems. A groundbreaking outcome in 2D is that the non-linear ionic transport results in memory dependent conduction, due to the slow time dynamics of the pairing. In other words, these 2D systems behave as memristors. Coupling two of these ionic 2D-memristors, in the spirit of Hodgkin-Huxley model of neuronal dynamics, we have shown that this nanofluidic device highlight autonomous spiking, in full analogy to real neurons. This serendipitous outcome is a quantum leap to develop artificial neural machines, in particular in view of the accessibility to fabricate such 2D systems in the lab. This paper was published in the review Science in 2021.

(5) Last, and a kind of climax of the project, a strong experimental effort has been achieved to explore these phenomena in experimental nanofluidic devices. We have extended our know-how to fabricate 2D systems and heterostructures, in the footsteps of recent advances achieved on 2D slits. A key experimental result is that we fully confirmed the predictions for the occurrence of ionic memristors in confined 2D systems. Going beyond, we showed that these systems allowed to perform Hebbian learning, just like biological synapses. This is the first ionic analogue of such an advanced functionality. This paper was published in the review Science in 2023.

Altogether this fully achieves the original goals of Shadoks: developing active nanofluidics to created biomimetic ionic machines mimicking their biological (in particular neuronal) counterparts.

Nanofluidics is an emerging field with high potential in terms of fundamental science, but also in terms of applications at the water-energy nexus.

In terms of dissemination, a series of lectures explaining the emerging field of nanofluidics and the molecular mechanics of fluids has been realized at College de France in 2023 (
The general objective of the Shadoks projects was to unveil new nanofluidic behaviors, in relation to non-equilibrium and active transport, and harness these behaviors as building-blocks towards a ionic machinery.
The project has fully reached its goals: new fluidic behaviors were shown to emerge at the nanoscales , ionic machines were designed and then demonstrated experimentally. This allowed us to build the first 'ionic memristor' made of water and salts, mimicking the basic Hebbian learning of neuronal counterpart.

This layes the foundations to develop a primitive neuronal structure coupling several nanofluidic devices. This paves the way to mimicking elementary neuronal functionalities in an artificial device, which is now one of our grand challenge.
shadoks pumping