Turning Nobel‑Prize MOF into a Liquid‑State Chip: The Key to Future Computing

Background

The 2025 Nobel Chemistry Prize honored pioneers of metal‑organic frameworks (MOFs), a class of porous crystalline materials once dismissed as impractical. MOFs act like molecular sieves or nanoscale LEGO bricks, allowing precise control of pore size and chemistry by selecting metal nodes and organic linkers. Their high surface area has driven extensive research in gas storage and catalysis, yet stability and cost have limited real‑world adoption.

Fabrication of a Liquid‑State Chip

Researchers addressed the “application gap” by constructing a fluidic transistor on a polymer membrane. They first created a nano‑scale channel (tens to hundreds of nanometres in diameter) and then grew a MOF crystal in situ within the channel. The MOF used contains zirconium (Zr) clusters and a sulfonated terephthalate linker (H2BDC‑SO3H). A amine‑functionalised “bullet‑shaped” nano‑channel membrane was sandwiched between two solution reservoirs: one with organic ligands, the other with metal salts. Molecules meet inside the confined channel, nucleating MOF seeds that elongate from one end, lining the channel with a porous crystal.

Device Structure

The resulting structure resembles a hierarchical “ion maze”: large pores function as highways, while smaller pores act as side streets, forming a graded network of nanometre‑ to Å‑scale channels. Growth control produced multiple internal interfaces: a ~100 nm one‑dimensional heterojunction between the polymer wall and MOF lining, and numerous sub‑nanometre junctions within the MOF itself, generated by two distinct Zr‑oxygen coordination modes.

Electrical Characterization

The MOF‑based nano‑fluidic transistor (h‑MOFNT) was immersed in electrolyte solutions and its current‑voltage (I‑V) behavior recorded. In 0.1 M HCl, where protons (H⁺) dominate, the I‑V curve showed strong non‑linearity: rapid current rise from 0 V to ~0.2 V, a slower increase from 0.3 V to 0.8 V, and saturation above ~0.9 V. This “triode‑like” response mirrors the turn‑on characteristic of a field‑effect transistor. By contrast, in KCl the device behaved as a simple diode with no clear threshold.

Proton‑Specific Switching Mechanism

The authors attribute the proton‑only threshold to the MOF’s built‑in electrostatic potential arising from fixed charges and the hierarchical pore network. At low bias the potential blocks proton passage; once the external voltage exceeds the threshold, protons flood the channel, further enhancing the internal potential and causing current saturation. Quantitative analysis showed that in HCl about 86 % of the total charge transport is due to protons, whereas in KCl roughly 81 % originates from K⁺, confirming the distinct ion‑selective behavior.

Memory Effect

When the voltage was cycled, the current response depended on the previous voltage history, producing a hysteresis loop characteristic of memristive behavior. Faster voltage sweeps amplified the hysteresis, while slow scans reduced it, indicating that the device “remembers” recent stimuli for several seconds. This short‑term memory parallels synaptic plasticity in biological neurons, where high‑frequency stimulation strengthens transmission.

Parallel Device Demonstration

To showcase scalability, five h‑MOFNTs were connected in parallel, forming a small fluidic circuit. Increasing the number of transistors yielded systematic, non‑linear changes in the overall I‑V curve, analogous to adjusting gate voltage in an array of electronic transistors. The inherent memory of each element endows the circuit with primitive neural‑network‑like properties, suggesting a route to programmable analog computation.

Implications and Remaining Challenges

The work proves that MOFs can serve as the active medium for liquid‑state chips that process information via ion flow, offering a potential platform for neuromorphic hardware that operates in a fluid environment and could interface directly with biological systems. However, large‑scale deployment will require improvements in device stability, reliability, and manufacturing cost.