Excited-State Intramolecular Proton Transfer and Competing Pathways in 3-Hydroxychromone: A Non-adiabatic Dynamics Study

Using mixed quantum-classical non-adiabatic dynamics simulations, this study elucidates the microscopic origin of the dual time scales in 3-hydroxychromone's excited-state intramolecular proton transfer by revealing that a competitive out-of-plane hydrogen torsional motion generates a slower pathway alongside the ultrafast canonical proton transfer.

The Gist

Imagine a molecule called 3-Hydroxychromone (3-HC) as a tiny, energetic acrobat. When you shine a specific light on it, the acrobat gets a sudden burst of energy and starts to dance wildly. This is the "excited state."

The main trick this acrobat is famous for is ESIPT (Excited-State Intramolecular Proton Transfer). Think of this as the acrobat grabbing a small ball (a proton/hydrogen atom) from one hand and quickly tossing it to the other hand. This move changes the acrobat's outfit entirely (a chemical change called "tautomerization"), and when they settle down, they glow with a different color of light.

For a long time, scientists knew this "ball toss" happened incredibly fast—like a blink of an eye (femtoseconds). But experiments showed something weird: sometimes the acrobat took a second, slower path to do the same trick, taking about a thousand times longer (picoseconds). Why? Nobody could explain the "slow motion" version until now.

This paper is like a high-speed camera recording of that acrobat, revealing exactly what's happening. Here is the story in simple terms:

1. The Setup: Two Doors, One Stage

When the light hits the molecule, it jumps onto a stage with two very similar "floors" (energy states).

• Floor 1 (The Bright Floor): This is where the acrobat usually starts. It's easy to jump here, and the "ball toss" (proton transfer) is a straight, smooth path.
• Floor 2 (The Dark Floor): This is a tricky, dark floor right next to the first one.

The researchers found that the acrobat doesn't just stay on one floor. It bounces back and forth between them almost instantly.

2. The Mystery: Why the Delay?

Scientists noticed that while most acrobats did the ball toss instantly, a huge group (about 58%) seemed to hesitate. They didn't just toss the ball; they seemed to get distracted.

The paper explains that this "distraction" is a twist.
Imagine the acrobat is supposed to toss the ball straight across. But instead, they sometimes decide to do a cartwheel or a spin (a twisting motion of the hydrogen atom) before they toss the ball.

• The Fast Path: The acrobat sees the ball, grabs it, and tosses it immediately. (This takes ~25 femtoseconds).
• The Slow Path: The acrobat grabs the ball, but then gets caught in a twisting spin. They have to finish the spin, stabilize, and then toss the ball. This extra dance move adds a delay, making the process take about 500 femtoseconds (or even up to a few picoseconds).

3. The "Trap" in the Dance

The researchers discovered a specific "trap" or "pit" on the dance floor. After the acrobat jumps, they often fall into this pit where they are forced to spin (the "torsional minimum").

• If they fall into the pit, they have to climb out of it before they can toss the ball. This climbing and spinning is the slow component.
• If they miss the pit, they toss the ball immediately. This is the fast component.

4. The Big Picture: A Map of the Dance

The authors created a complete "map" (a reaction network) of all the possible moves the acrobat can make.

• They realized that the "slow" delay isn't because the ball toss is hard; it's because the acrobat is busy doing a twisting dance move that competes with the toss.
• It's like a runner who is supposed to sprint to the finish line. Some runners sprint straight there. Others get distracted by a side track where they have to do a few cartwheels before rejoining the race. Both reach the finish, but one gets there much later.

Why Does This Matter?

Understanding this "twist vs. toss" competition is like having the instruction manual for the acrobat.

• For Science: It solves a decades-old mystery about why some molecules glow in two different ways (dual fluorescence).
• For Technology: These molecules are used in things like OLED screens (your phone or TV) and medical sensors. If we know exactly how the "twist" slows things down, we can design better molecules that glow brighter, faster, or more reliably.

In a nutshell: The molecule isn't just a simple ball-tosser. It's a complex dancer. Sometimes it dances straight to the finish line, and sometimes it gets caught in a spinning groove, which explains why the "glow" happens at two different speeds. This paper finally caught the dancer in the act and explained the spin.

Technical Summary

1. Problem Statement

The study addresses the mechanistic origin of the dual time constants observed in the Excited-State Intramolecular Proton Transfer (ESIPT) of 3-hydroxychromone (3-HC).

• Experimental Context: Upon photoexcitation to the lowest "bright" singlet excited state ($S_1$, $\pi\pi^*$), 3-HC exhibits two distinct proton-transfer timescales: an ultrafast component (femtoseconds) and a slower component (picoseconds).
• The Mystery: While the ultrafast component is well-understood, the microscopic origin of the slower (picosecond) component has remained hypothetical. Previous hypotheses suggested it might arise from solvent effects, intramolecular vibrational relaxation (IVR), or transient access to a trans-enol conformation.
• The Gap: Previous theoretical studies (specifically Perveaux et al. and Anand et al.) identified a conical intersection (CI) between the $S_1$ ($\pi\pi^*$) and $S_2$ ($n\pi^*$) states near the Franck-Condon (FC) point, suggesting it could divert population away from the reaction coordinate. However, no study had successfully reproduced both time constants explicitly via non-adiabatic dynamics simulations or provided a unified reaction network linking these pathways.

2. Methodology

The authors employed mixed quantum-classical non-adiabatic dynamics simulations to model the coupled electron-nuclear motion of 3-HC.

• Electronic Structure Method:
  • Theory: Time-Dependent Density Functional Theory (TD-DFT) using the PBE0 functional.
  • Basis Set: cc-pVDZ (validated with cc-pVTZ and aug-cc-pVDZ to ensure basis set convergence).
  • States: The first three electronic states ($S_0, S_1, S_2$) were considered.
  • Validation: The method was benchmarked against higher-level methods (CC2, ADC(2), EOM-CCSD) and experimental absorption spectra. The $S_1$/$S_2$ energy gap at the FC point was calculated as ~0.12 eV, consistent with literature.
• Dynamics Simulation:
  • Software: SHARC (Surface Hopping Including Arbitrary Couplings) version 3.0.1.
  • Algorithm: Trajectory Surface Hopping (TSH) with the Granucci-Persico decoherence scheme.
  • Initial Conditions: 10,000 geometries sampled from a 0 K Wigner distribution of the ground-state (cis-)enol conformer. Trajectories were selected within a 0.2 eV window centered at 3.9 eV (~320 nm) to match experimental excitation.
  • Ensemble: 311 trajectories initiated on $S_1$ and 86 on $S_2$ (due to intensity borrowing).
  • Parameters: 0.5 fs time step; energy threshold of 0.10 eV for hops to the ground state (to mitigate TD-DFT deficiencies at ground-state CIs).
• Analysis:
  • Reaction Coordinate: Monitored via O–H bond distances ($d(O_d-H)$ and $d(H-O_a)$).
  • Torsional Coordinate: Monitored via dihedral angles $\phi_1$ and $\phi_2$ (out-of-plane hydrogen motion).
  • Statistical Tools: Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) were used to identify dominant collective motions and separate trajectory subsets.

3. Key Contributions

1. Explicit Observation of Dual Time Constants: The simulations successfully reproduced both the ultrafast (sub-100 fs) and the slower (picosecond) ESIPT components, a feat not achieved in previous non-adiabatic dynamics studies of this system.
2. Identification of the Competing Pathway: The study definitively attributes the slower time constant to a competitive out-of-plane hydrogen torsional motion rather than solvent effects or IVR alone.
3. Construction of a Unified Reaction Network: The authors mapped the full excited-state landscape, delineating the interplay between the canonical ESIPT pathway and the torsion-mediated pathway, including specific Minimum Energy Conical Intersections (MECIs).

4. Key Results

A. Electronic State Dynamics

• $S_1$/$S_2$ Coupling: Upon excitation, the system rapidly accesses a Minimum Energy Conical Intersection (MECI) between $S_1$ ($\pi\pi^*$) and $S_2$ ($n\pi^*$) located ~0.10 eV below the FC point.
• Population Transfer: Within the first 25 fs, significant population transfer occurs between $S_1$ and $S_2$. The system relaxes efficiently to the $S_1$ state (now with $n\pi^*$ character) before proton transfer occurs.
• No Ground State Decay: No trajectories returned to the ground state ($S_0$) within the simulation window, consistent with experimental fluorescence lifetimes.

B. The Two ESIPT Time Constants

Kinetic modeling of the enol-to-keto population decay revealed a bi-exponential behavior:

• Fast Component ($\tau_1$): ~24.8 fs. This corresponds to the direct ESIPT pathway on the $S_1$ surface, driven by the low barrier (~0.04 eV) along the proton transfer coordinate.
• Slow Component ($\tau_2$): ~510 fs (simulated) vs. ~5.5 ps (experimental). The discrepancy is attributed to the limited simulation time and subtle solvent effects, but the qualitative existence of a picosecond component is confirmed.

C. Origin of the Slow Component: Torsional Competition

• Mechanism: After crossing the $S_1/S_2$ seam near the FC point, a significant fraction of trajectories (54.4%) explores a torsional minimum on the $S_1$ surface where the transferring hydrogen rotates out-of-plane ($|\phi_1| \approx 21.6^\circ$).
• Barrier Crossing: To reach the trans-enol conformer ($|\phi_1| > 130^\circ$), the system must overcome a modest barrier (0.09 eV).
• Correlation: Trajectories that visit this torsional region exhibit delayed proton transfer.
  • Trajectories without significant torsion ($|\phi_1| < 21.6^\circ$) convert rapidly to the keto form.
  • Trajectories with torsion ($|\phi_1| \ge 21.6^\circ$) show a delayed ESIPT, matching the slower time constant.
• Conclusion: The slower ESIPT component arises because the system is temporarily "trapped" in the torsional well, delaying the proton transfer until the hydrogen re-orients.

D. Reaction Network

The study constructed a complete reaction network involving eight distinct states (combinations of cis-enol, tor-enol, trans-enol, and keto forms across $S_1$ and $S_2$).

• Key finding: Population transfer between $S_1$ and $S_2$ also occurs after proton transfer (in the keto region) via non-planar MECIs, further complicating the relaxation landscape.

5. Significance

• Mechanistic Resolution: This work resolves the long-standing debate regarding the dual time constants in 3-HC, proving that the slower component is an intrinsic molecular property driven by out-of-plane torsion, not an artifact of solvent interactions.
• Design Principles: By identifying the torsional motion as a competitor to ESIPT, the study provides a rational basis for designing 3-HC derivatives. Modifying the molecular structure to restrict out-of-plane motion could potentially enhance the quantum yield and speed of ESIPT, improving their utility as fluorescent probes, OLED materials, and photostabilizers.
• Methodological Benchmark: The successful reproduction of multi-scale dynamics (fs to ps) using TD-DFT based surface hopping demonstrates the capability of these methods to capture complex non-adiabatic phenomena in medium-sized organic molecules.

In summary, Nardi and Vacher provide a unified mechanistic framework where canonical ESIPT and torsion-mediated pathways coexist, explaining the complex photodynamics of 3-hydroxychromone through explicit non-adiabatic simulations.