Text from the interactive explanation
Bimodal multiphoton catalysis via structural regeneration
Cristian Rosso, Giorgia Barison, Federico Droghetti, Nicola Michelazzo, Andrea Sartorel, Giorgio Pelosi, Marco Bortolus, Sara Bonacchi, Stefano Visentini, Alessia Marrese, Giovanni Bistoni, Sofia Lerda, Vittoria Burigana, Elisabetta Collini, Paolo Costa, Mirco Natali, Luca Dell’Amico
Original paperNature Catalysis · 2026 · Rosso, Barison et al.
Violet light splits phoenix1 into an electron that reduces and a fragment that oxidises. Two opposite jobs in one flask, a 5.7-volt reach from visible light. Then the halves rebuild the catalyst.
400 nm photon
solvated electron
radical 11
acridinium 5 +
1 → radical 12
2 → radical 13
product 3 , new bond
phoenix1 intact
Every atom is placed from the authors' DFT-computed structures; the motion between them is interpolated. Drag the chamber to turn it. Switch to One photon at a time to be the light source.
01 · The redox window
One photon is not enough
A molecule's redox potential (in volts against a reference electrode, SCE) says how hard it pushes electrons out (very negative) or pulls them in (very positive). The test reaction needs both: an electron pushed into 1 beyond −1.8 V to cut its C–Cl bond, and an electron pulled out of 2 beyond +1.7 V to cut its C–Si bond. That 3.5 V span is more than one 400 nm photon carries (3.10 eV). Pick a catalyst and see what the authors measured.
Beyond the one-photon line
Redox power each coupling demands (Fig. 6), against the energy of a single 400 nm photon. Tap a product.
Two photons pay for one turnover
02 · Selectivity
Sorting radicals without a sorter
A radical is a fragment with an unpaired electron, eager to bond with anything. This flask makes four kinds at once, so random pairing should give a mess. Two features keep it clean: the acridine adducts re-open under light, so only product 3 is a permanent exit, and radical 11 can briefly hold a benzyl radical as 10 , a shuttle that later hands it to its partner.
What the NMR saw
A fluorine-tagged silane 15 lets 19F-NMR follow each product over four hours (Fig. 5a). Slide through time.
03 · Evidence
Caught in the act, across 20 orders of magnitude
The cycle above is the authors' proposal, assembled from measurements that each see one moment: femtosecond lasers see the electron leave, nanosecond flashes see radicals and electrons, spin traps catch them, electrodes interconvert phoenix1 and 5 , and NMR follows the flask for hours.
Adding the electron scavenger KNO3 collapses the yield: free solvated electrons do the reducing.
EPR spin trapping shows two radicals after light: the trapped electron (nine lines) and radical 11 (six lines). Nothing in the dark.
The X-ray structure of phoenix1 confirms the C–C-linked dimer, and phoenix1 is recovered intact at the end of the reaction.
04 · Why it matters
Organic dyes usually fail by falling apart. phoenix1 falls apart on purpose, and puts itself back together: one visible-light catalyst that reduces and oxidises in the same flask, cutting C–Cl, C–Br and C–Si bonds to build 52 products, including boronic esters and all-carbon quaternary centres, in yields up to 78%.
Source
Bimodal multiphoton catalysis via structural regeneration. Cristian Rosso, Giorgia Barison, Federico Droghetti, Nicola Michelazzo, Andrea Sartorel, Giorgio Pelosi, Marco Bortolus, Sara Bonacchi, Stefano Visentini, Alessia Marrese, Giovanni Bistoni, Sofia Lerda, Vittoria Burigana, Elisabetta Collini, Paolo Costa, Mirco Natali & Luca Dell’Amico. Nature Catalysis 9, 1079–1088 (2026). doi:10.1038/s41929-026-01593-w
Open access under CC BY 4.0. 3D molecules are rendered from the authors' DFT coordinates (Supplementary Data Table 1); atom order was re-mapped and structures rigidly aligned, and motion between computed structures is interpolated for illustration. Charts are redrawn from Fig. 2, Fig. 5a, Fig. 6, Supplementary Fig. 54 and Supplementary Tables 1 and 10. Figure images are crops of Fig. 3. The flask is an illustrative particle model built on the proposed mechanism. 3D rendering uses three.js (MIT licence).
Topics in this iPaper
Chemistry
we report a purely organic photocatalyst that operates via a two-photon excitation mechanism, enabling both oxidative and reductive transformations within a unified platform
Abstract, page 1079
photoredox catalysis
electron transfer, commonly referred to as photoredox catalysis, has become a major strategy
Introduction, page 1079
multiphoton excitation
In this Article, we introduce bimodal multiphoton catalysis, which transcends these limitations.
Introduction, page 1080
photoionization
femtosecond transient absorption spectroscopy unambiguously established the photoionization of phoenix1 upon 400 nm excitation
Mechanistic insights from spectroscopic measurements, page 1081
solvated electrons
These solvated electrons are responsible for the reduction of 1, having an E = −3.3 V/−3.5 V versus the SCE
Dynamic photocatalysis, page 1082
transient absorption spectroscopy
Femtosecond pump–probe and nanosecond LFP studies of transient photoexcited species.
Fig. 3 caption, page 1083
radical cross coupling
These reduced and oxidized species undergo radical combination to deliver the corresponding coupling products (52 examples, up to 78% yield).
Conclusions, page 1087
density functional theory
The catalytic cycle in Fig. 3e was supported by calculations carried out at the M062X/def2-TZVP-CPCM (iPrOH) level of theory
Dynamic photocatalysis, page 1082