By systematically probing the hyperspace of the Biginelli reaction, the authors uncovered a pseudo-seven-component transformation that yields structurally complex products with unusual supramolecular properties, demonstrating that unexplored regions harbor undiscovered reactivity.
Historically, discovering new reactions has propelled chemistry forward, unlocking molecular structures vital to medicines, materials, and catalysts. Such advances have largely relied on human intuition, incremental refinement, and chance observations.
No time to finish this now? Download a free PDF for later!
Today, chemical automation and mechanism-level artificial intelligence (AI) enable systematic probing of multidimensional condition hyperspaces, transcending conventional intuition to expose hidden reactivity even in well-studied transformations.
In this view, reactions are not linear processes but mechanistic networks that can be redirected into unexpected pathways under different conditions. Earlier work has demonstrated this, but it yielded only structurally simple, human-analyzable products.
The Biginelli multicomponent reaction, known since 1891 for producing biologically relevant dihydropyrimidinones, serves as a testing ground in this study. Robotic experimentation, with later evaluation by mechanism-aware AI algorithms, reveals an uncharted hyperspace region hosting an unprecedented pseudo-seven-component transformation, delivering polycyclic scaffolds of natural-product-like complexity with rare supramolecular behaviors.
How the Compounds Were Made
The synthetic procedures reveal two distinct protocols for accessing the newly discovered scaffolds. The first, yielding compounds 8a-i, involves reacting a dibenzylideneacetone derivative with urea in acetonitrile, using methanesulfonic acid as a catalyst, under heating at 100 °C for 24 hours in a sealed pressure tube.
After aqueous workup and dichloromethane extraction, the crude product is purified via recycling high-performance liquid chromatography (HPLC).
The second protocol, producing compounds 19a, 19b, 20a, and 20b, employs a dibenzylideneacetone derivative, ammonium acetate, and ethyl acetoacetate in ethanol, similarly heated in a sealed pressure flask.
Following workup, the material undergoes oxidation with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), is subsequently washed with sodium bicarbonate and brine, and passes through a short basic alumina column before final recycling HPLC purification.
To investigate the unusual supramolecular behaviors, proton nuclear magnetic resonance (1H NMR) titration and self-association studies were conducted according to established procedures in the literature, with data fitting performed using Bindfit software across multiple datasets simultaneously.
Ligand-metal-ion binding was evaluated against 1:1, 1:2, and 2:1 stoichiometric models, while self-association was analyzed using equal K-dimerization and coequal K models. Quantum chemical modeling supported these studies. Conformer ensembles were generated via CREST at the GFN2-xTB level, refined with B97-3c density functional theory (DFT), and electronic energies were calculated with ωB97X-3c.
Solvation-free energies employed the openCOSMO-RS method. Theoretical circular dichroism spectra were computed using time-domain DFT (TD-DFT) at the ωB97X-D/cc-pVDZ level, with transition energies empirically corrected, enabling confident assignment of the observed supramolecular phenomena.
Probing the Hidden Hyperspace
The hyperspace approach treats a reaction not as a single transformation but as a mechanistic network interrogated by varying substrate concentrations and analyzing the major detectable resulting species.
The custom-built Robowski platform automates reaction setup and ultraviolet-visible light (UV-Vis) spectral acquisition across hundreds of conditions. A closed-loop protocol then combines all crude mixtures, subjects them to iterative HPLC purification and NMR analysis, and fits calibrated UV-Vis spectra of isolated components against the entire hyperspace dataset.
Applied to the Biginelli reaction of p-methoxybenzaldehyde, ethyl acetoacetate, and urea under camphorsulfonic acid catalysis, the platform probed 960 conditions. Eight HPLC-NMR iterations identified 12 key components that satisfactorily fitted all spectra.
Reconstruction of the mechanistic network was straightforward for known products but challenging for the unprecedented compound 8a, whose structure was ultimately solved by X-ray diffraction. This bicyclic product incorporates seven components and forms via a pathway proposed by the MECH algorithm.
Yield mapping across 144 conditions showed that product 1 was favored at low acid and low aldehyde concentrations (up to 80%), product 2 at low acid and high aldehyde concentrations (up to 60%), and 8a reached only 9% at high acid.
To improve practicality, the pathway was split into two steps: first, pre-forming the dienone 7a, then reacting it with urea to yield 8a in 28% yield. Nine analogs with varied aromatic rings were synthesized, and the annulation pattern was extended to Hantzsch-type reactions, producing doubly and triply annulated pyridines.
Beyond synthesis, products 8a, 20a, and 20b displayed rare supramolecular behaviors. Concentration- and temperature-dependent NMR studies revealed hydrogen-bond-mediated dimerization of 8a, with aggregation increasing at lower temperatures. Metal-binding studies showed 20a and 20b selectively bind Ba2+ and Zn2+ as 2:1 complexes with exceptionally high positive cooperativity, while 8a and 8b exhibit broader but weaker binding profiles.
Most strikingly, 8a undergoes homochiral self-sorting upon crystallization, yet in solution Ba2+ induces exclusive heterochiral sorting, and Zn2+ promotes homochiral sorting.
Implications and Outlook
This study offers a fundamental reframing of how chemical reactivity can be viewed. By deploying robotic platforms guided by algorithmic decision-making, the researchers uncovered a pseudo-seven-component transformation hidden for over a century within the well-studied Biginelli reaction.
The resulting polycyclic products exhibited rare supramolecular behaviors suggesting potential utility in sensing, encapsulation, and materials design. More broadly, this work demonstrates that hyperspace exploration could help to unlock mechanistically distinct and functionally valuable chemistry, expanding the boundaries of what chemical reactivity can achieve.
Journal Reference
Matuszczyk, D. et al. (2026). Hyperspace exploration using robotics for the discovery of mechanistically distinct transformations and complex functional products. Nature Synthesis. https://www.nature.com/articles/s44160-026-01096-3.
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.