Every living cell depends on shape-shifters. Proteins contort themselves to grasp specific molecules. DNA coils and uncoils to regulate gene expression. RNA folds into compact structures for transport, then unfurls to transcribe genetic instructions. This structural choreography isn't merely elegant—it's essential.

Chemists have long sought to replicate this adaptive behavior in synthetic molecules. The goal is to create artificial hosts that can accommodate different guest molecules without redesigning the entire structure each time. Such versatility could revolutionize chemical purification, sensing, and catalysis. But there's a fundamental tension in the design: these molecular cages must be rigid enough to assemble reliably, yet flexible enough to adapt.

Now, researchers at the University of Cambridge have solved this puzzle with a cube-shaped molecular cage that reconfigures its faces to accommodate guests ranging from tiny adamantane molecules to bulky organic anions nearly four times larger. The structure achieves what most synthetic hosts cannot—true induced-fit binding, where the host reshapes itself around each guest rather than forcing guests into a fixed cavity.

The cage is built from zinc ions and organic ligands incorporating naphthalene groups—flat, two-ring aromatic structures that can rotate around specific bonds. Think of each face of the cube as a panel with rotating shutters. These naphthalene groups can flip between two positions: endo, pointing inward toward the cavity, or exo, pointing outward. Each of the six faces operates independently, creating what the researchers describe as a quantized system with multiple distinct volume states.

When empty, the cage adopts an all-endo conformation. This configuration minimizes cavity volume, releasing the maximum number of solvent molecules into solution—an entropically favorable arrangement. The cavity measures 389 cubic angstroms, roughly the volume of a small vitamin molecule.

But introduce a guest, and the cage responds. Nuclear magnetic resonance spectroscopy revealed that binding occurs slowly on the NMR timescale, suggesting the cage undergoes substantial structural reorganization. The team used selective irradiation experiments to determine which naphthalene protons sat closest to bound guests. For small guests like adamantane, protons on the inward-facing portions of endo naphthalenes showed proximity signals. For larger guests like tetrakis(4-chlorophenyl)borate, the pattern flipped—outward-facing protons from exo naphthalenes appeared near the guest instead.

The most striking evidence came from ion mobility mass spectrometry. This technique measures how ions drift through an inert gas under an electric field, revealing their collision cross-section—essentially their physical size. The researchers analyzed the cage with seventeen different guests and found the collision cross-section values clustered into five distinct tiers, separated by increments of 9.2 square angstroms.

This quantization suggests that faces flip from endo to exo in discrete steps rather than continuous deformation. A cage with zero, one, two, three, or four exo faces would occupy progressively larger volume states. The five tiers observed experimentally align beautifully with this model. The fact that no intermediate values appeared supports the interpretation that individual faces switch cleanly between conformations rather than gradually bending.

Computational simulations traced a possible pathway for the all-endo to all-exo transformation. The calculations revealed that naphthalene groups can rotate individually or in concert, expanding the cavity incrementally as guest size demands. Energy barriers for these rotations proved modest—low enough that guest binding itself likely stabilizes intermediate structures and accelerates the transformation.

The range of guests successfully bound is extraordinary. Adamantane, at 178 cubic angstroms, occupies just 46 percent of the empty cage's cavity volume. At the opposite extreme, tetrakis(4-chlorophenyl)borate measures 599 cubic angstroms—154 percent of the original cavity. This nearly fourfold span exceeds what other flexible synthetic cages have achieved.

Crucially, the cage maintains its structural integrity throughout these transformations. The metal-to-metal distances remain constant whether faces adopt endo or exo orientations, preventing strain accumulation in the framework. This design insight—decoupling face rotation from vertex geometry—may prove broadly applicable to other cage architectures.

Diffusion measurements in solution painted a slightly different picture than gas-phase mass spectrometry. While collision cross-sections showed clear quantization, diffusion coefficients varied more smoothly with guest size. This difference likely reflects the different conditions: in solution, acetonitrile molecules solvate the cage and may stabilize multiple conformations simultaneously. These microstates would interconvert rapidly, yielding an average diffusion rate. In the gas phase, without solvent, van der Waals interactions between naphthalene groups would favor the minimum number of exo faces needed to accommodate each guest, producing discrete size states.

The work represents a conceptual advance in supramolecular chemistry. Most artificial hosts follow the lock-and-key principle: a rigid binding site either fits a guest or doesn't. Proteins often work differently, adopting what Daniel Koshland termed induced fit in 1958—the binding site reshapes itself around the substrate. This flexibility allows a single protein to process multiple substrates or adapt binding strength through conformational selection.

Creating induced fit in wholly synthetic systems has proven difficult. Some approaches use mechanical bonds like rotaxanes, where rings thread onto molecular axles. Others employ light-triggered switches or stimuli that interconvert between cage forms. But these strategies typically toggle between two states rather than offering continuous adaptability across a size range.

The naphthalene-hinged cage achieves true conformational adaptation through inherent structural properties rather than external triggers. Each face possesses two stable states, and the energetic difference between them is small compared to the binding energy of most guests. This means guest encapsulation provides sufficient thermodynamic driving force to flip whatever number of faces is needed.

From a practical standpoint, such adaptive hosts could enable more general-purpose molecular containers. Current applications often require synthesizing a new cage for each target molecule. A single adaptive host capable of binding molecules across a broad size range would streamline chemical separations, purification of reaction products, or selective extraction of contaminants.

There are broader implications for understanding molecular recognition. The quantized expansion observed here suggests that even in flexible systems, discrete conformational states may dominate rather than continuous deformation. This principle might inform the design of sensors where distinct binding modes produce distinct signals, or catalysts where cavity size controls reaction selectivity.

The researchers estimate the cage can access ten possible combinations of face orientations, yielding seven distinct cavity volumes when symmetry is considered. Only five tiers appeared in the experimental data, meaning the largest guests tested didn't require all six faces to flip exo. Identifying guests that demand full expansion would test the limits of the cage's adaptability.

Questions remain about the detailed mechanism. Do faces flip one at a time in a specific sequence, or can multiple faces rotate simultaneously? Does the guest molecule actively template the transition, or does the cage sample different conformations spontaneously and guests simply stabilize certain states? Variable-temperature NMR experiments suggested activation barriers of 4 to 9 kilojoules per mole for conformational interconversion—low enough for rapid exchange at room temperature, consistent with the fast equilibration observed.

The computational simulations indicated that aromatic stacking interactions between naphthalene groups stabilize the endo conformation when the cavity is empty. Guest binding disrupts these stacking arrangements, shifting the equilibrium toward exo forms. This interplay between intramolecular stabilization and host-guest interactions fine-tunes the system's response to different guests.

Looking forward, the design principle—incorporating rotational freedom into cage faces while maintaining rigid vertices—should generalize beyond this specific structure. Different aromatic linkers, alternative metal ions, or varied ligand geometries could produce adaptive cages with different size ranges, selectivity profiles, or functional capabilities. The key insight is balancing pre-organization for assembly with conformational freedom for adaptation.

The work also highlights how quantitative techniques like ion mobility mass spectrometry can reveal subtle structural features invisible to conventional methods. The clear tier structure in collision cross-section data provided evidence for discrete conformational states that wouldn't have been obvious from binding studies or NMR alone. As instrumentation advances, such measurements may become routine tools for characterizing dynamic supramolecular systems.

Biology has exploited conformational adaptability for billions of years. Enzymes shift shape to stabilize transition states. Chaperone proteins unfold and refold client proteins. Molecular motors change conformation to generate mechanical work. This new molecular cage demonstrates that chemists can now engineer comparable sophistication into wholly synthetic systems—structures that sense their environment and respond by reshaping themselves.

The difference between a rigid container and an adaptive one parallels the difference between a box and a hand. One holds objects that happen to fit. The other adjusts its grip to whatever it grasps.

Credit & Disclaimer: This article is a popular science summary written to make peer-reviewed research accessible to a broad audience. All scientific facts, findings, and conclusions presented here are drawn directly and accurately from the original research paper. Readers are strongly encouraged to consult the full research article for complete data, methodologies, and scientific detail. The article can be accessed through https://doi.org/10.1038/s41557-024-01708-5