This year's Nobel Prize-winning achievement was once questioned for being useless, but a Chinese team used it to develop the "key" to future chips.

[Introduction] The achievement of this year's Nobel Prize in Chemistry, MOF, was once questioned for its lack of practicality. However, a research team almost entirely composed of overseas Chinese has discovered that this achievement can be used to manufacture "fluid chips", breaking through the bottlenecks of current semiconductor chips! This achievement was recently published in a sub - journal of Science.

The 2025 Nobel Prize in Chemistry was awarded to the pioneers of metal - organic framework (MOF) materials.

This material is known as a "molecular sieve" because it is like a sieve or sponge in the microscopic world, filled with nanoscale pores inside, capable of screening and storing specific molecules.

Meanwhile, MOF is like the "nano - Lego bricks" in the chemical world. We can use different metal nodes and organic molecules as building blocks to freely piece together various pore structures, precisely controlling the pore size and chemical functions.

With these characteristics, a large number of papers on MOF have been published in fields such as gas storage and catalysis. However, due to reasons such as poor stability and high cost, there were few practical applications for a long time, which can be described as "much ado about nothing".

The Nobel Prize Committee also carefully stated in the award speech that MOF has "great potential" and provides unprecedented opportunities for new functional materials.

This also implies that although MOF is very promising, a truly applicable scenario still needs to be found.

Now, an application scenario for MOF seems to have emerged: Scientists have actually used MOF materials to create a nanoscale liquid chip. It can not only perform logical operations like a conventional electronic chip but also "remember" the previous signal states, producing a memory effect similar to that of brain neurons!

Transforming MOF into a chip: 'Growing' circuits in nanopores

Most of the chips we usually refer to are silicon - based electronic chips, which rely on the movement of electrons in wires and transistors to transmit and process information.

The "fluid chip" here is completely different: it uses the flow of ions in liquid to simulate electronic circuits.

Sounds like science fiction, doesn't it?

The research team designed a special nanofluidic transistor, and the core material is the newly - famous Nobel Prize "internet celebrity" - MOF.

This chip is about the size of a coin, but its structure is quite delicate.

Researchers first created a nanoscale pore channel (with a diameter of only tens to hundreds of nanometers) on a polymer film, and then cleverly grew MOF crystals in - situ inside the channel.

This is like cultivating a porous crystalline "sponge" inside an extremely thin straw.

This MOF sponge has regularly arranged micropores and nanopores, so a complex structure with intertwined large and small channels is formed in the entire channel: the large pores are like highways, and the small pores are like alleys. The two form a hierarchical "ion maze" for charged ions to pass through.

A schematic diagram of the structure of the MOF nanofluidic transistor device shows that there are two scales of heterojunctions in this device: one is a one - dimensional heterogeneous interface formed at the junction of the polymer nanopore (the gray pipeline on the left) and the MOF crystal (the green part); the other is a three - dimensional junction network formed by the connection of different structural units inside the MOF crystal (the yellow interface inside the green crystal in the figure). These hierarchical interfaces and channels together create special ion - transport properties.

The researchers selected a MOF material containing zirconium (Zr) metal clusters and terephthalic acid derivatives (H2BDC - SO3H, with sulfonic acid groups).

They sandwiched a bullet - shaped nanopore membrane modified with amino groups between two solution pools: one pool contained an organic ligand solution, and the other contained a metal salt solution.

The molecules of the two solutions met and combined in the narrow channel. Initially, tiny "seeds" of MOF were formed, and then MOF crystals gradually grew on the side near the tip of the channel.

In this way, the MOF crystal grew into the channel from one end, adding a "lining" to the channel.

Due to the careful control of the growth process and conditions, this section of the MOF crystal does not have a single, unchanging structure. Instead, multiple interfaces are spontaneously formed inside.

On the one hand, a one - dimensional heterojunction interface with a diameter of about 100 nanometers appears between the originally smooth polymer channel and the MOF crystal lining;

On the other hand, the MOF crystal itself is composed of two types of zirconium - oxygen cluster units with different connection methods, which is equivalent to forming numerous tiny interfaces and secondary channels inside the crystal.

These hierarchical channels and interfaces from the nanoscale to the angstrom scale (0.1 nanometer scale) are the secrets behind the uniqueness of this MOF chip.

Ion transistor: Controlling proton flow like a triode

Now that the structure is there, how does it actually exhibit the function of a "chip"?

The researchers immersed the prepared MOF nanofluidic transistor (abbreviated as h - MOFNT) in solutions containing different ions and measured its current - voltage characteristics.

This is actually to see how many ion currents can pass through the MOF channel when different voltages are applied.

The experimental results are exciting: when a hydrochloric acid (HCl) solution is used (where the main cation is the proton H⁺), the current of the chip shows a strong nonlinear change with the voltage.

At low voltages (0 - 0.2V), the current rises rapidly; at medium voltages (0.3 - 0.8V), the rise slows down; and when the voltage exceeds about 0.9V, the current growth gradually saturates and no longer increases significantly.

This "fast - then - slow" current - voltage curve is similar to the characteristic curve of a triode or field - effect transistor in electronics, which means that the proton flow is controlled by a threshold, a bit like water only gushing out in large quantities after a water valve is turned to a certain degree.

Scientists vividly call it a "triode - like" proton transport.

In contrast, for other larger metal ions (such as potassium ions K⁺), the device only shows an ordinary "diode - like" rectification behavior, without an obvious threshold switching effect.

Only the smallest and lightest ions, protons, trigger a special nonlinear switch.

Why are protons so special? The secret lies in the nanoscale to angstrom - scale channels and interfaces inside the MOF crystal.

The fixed charges and structure in MOF create a "built - in potential" barrier for protons.

Normally, this "gate" restricts the passage of protons; only when the applied voltage exceeds a certain threshold can the built - in potential be overcome, allowing protons to pass in groups.

Once the threshold opens the gate, more protons flowing through will further enhance the built - in potential, causing the current to tend to saturate.

This process is similar to how we control water flow with a sluice in our daily lives: the gate remains closed when the water pressure is insufficient. Once the water pressure is large enough, it pushes open the gate to release water, and the impact of the water flow may also keep the gate in a certain open state, and the water flow no longer increases.

In this way, the MOF nanopore channel realizes the switching control of proton flow, which is equivalent to an ionic version of an electronic transistor.

The researchers verified by changing the concentration of the solution and other means that this nonlinear conduction is indeed a phenomenon unique to protons and has nothing to do with the solution environment.

They also conducted an "attribution analysis" of the ions flowing through the chip and found that about 86% of the charge in the HCl solution is contributed by protons, while K⁺ contributes about 81% in the KCl solution.

This once again proves that the nonlinear "switch" conduction of protons is completely different from the linear behavior of ordinary ions.

Through the unique hierarchical channels of the MOF material, the chip realizes the selective transport of different ions: protons take the "fast lane" and enjoy threshold - controlled passage like traffic lights; while other ions are restricted to the "slow lane" and can only move at a constant speed.

This is the origin of the "selective ion transport" in the title of the paper.

Ion channels with "memory": Imitating the learning effect of neurons

Even more surprisingly, this MOF ion transistor can not only control the on - off of ion flow but also exhibits a "memory" ability!

When the researchers applied a cyclically changing voltage to the chip, they found that the current response does not simply depend on the instantaneous voltage but also on the previously experienced voltages.

The current magnitudes obtained when the voltage is increased from 0 to a certain value and then decreased back to 0 are different from those when the voltage is directly increased from 0 to the same value.

This means that the conductive state of the chip has a memory effect on historical signals.

The plotted current - voltage loop shows an obvious hysteresis loop, which is a typical characteristic of a memristor.

Moreover, the faster the voltage is cycled, the more obvious the hysteresis phenomenon; conversely, the hysteresis decreases when the voltage is scanned slowly.

That is, the chip has a "better memory" for rapid and continuous stimuli and "easily forgets" stimuli with long intervals.

The experiment shows that this memory effect of retaining the influence of the previous voltage can last for several seconds.

Doesn't it sound very similar to the characteristics of short - term memory in the brain?

In fact, the synaptic connections of brain neurons are also strengthened after short - term high - frequency stimulation (so - called synaptic plasticity), forming memory traces from seconds to minutes.

This MOF fluid chip exhibits a learning behavior similar to that of neural synapses at the ionic level. Its response to electrical signals changes with the past stimulation history and has simple brain - like learning functions.

The researchers further proved the usability of this "ionic memory": they connected five such MOF ion transistors in parallel to form a small fluid circuit to simulate more complex signal processing.

It was found that as the number of parallel components increases, the output curve of the entire circuit shows a series of regular nonlinear changes, similar to controlling the output of an electronic transistor array by adjusting the gate voltage.

That is to say, we can combine multiple ion transistors to achieve programmable analog computing, just like designing an electronic circuit.

Even better, since each component has a short - term memory, such an ionic circuit naturally has some characteristics of a neural network and is expected to be used to construct a brain - like computing system.

As the author said, this may be a small attempt towards a new generation of computers: if we can design functional materials (such as MOF) that are only a few nanometers thick, we will have the opportunity to manufacture advanced fluid chips to make up for or even surpass the limitations of today's semiconductor chips.

Great potential at the molecular scale

Using molecules and ions to build information - processing circuits sounds like a plot from a science - fiction novel, but this cutting - edge research shows us the prototype.

By using the MOF "molecular sieve" material as the basic unit to "assemble" a nanoscale ion - channel network, the research team successfully realized the logical and storage functions of a circuit at the molecular scale.

This achievement was published in the sub - journal of Science, Science Advances, in September 2025.

It not only proves the great usefulness of MOF materials (providing a strong response to the previous doubts about its "uselessness") but also opens up new possibilities for "liquid electronics" (ionics).

Imagine that in the future, computer chips may no longer be limited to silicon semiconductors. There may also be components like "liquid" brains: they operate in a liquid medium, use ion flow to process information, can more efficiently simulate