A Theoretical Investigation of the Thermal and Photochemical Mechanisms of Ethylbenzene Dehydrogenation on Rutile TiO$_{2}$(110)

This master's thesis utilizes a dual-methodological quantum chemical approach to reveal that ethylbenzene dehydrogenation on rutile TiO$_{2}$(110) proceeds via proton-coupled electron transfer on stoichiometric surfaces but shifts to a more efficient direct hydrogen atom transfer mechanism on oxidized surfaces, with photon energy determining whether the reaction bypasses ground-state kinetic barriers through excited-state persistence.

The Gist

The Big Picture: Making Styrene Without the Heat

Imagine you are trying to build a specific type of plastic (styrene) by taking a hydrogen atom off a molecule called ethylbenzene. Currently, the chemical industry does this by heating the mixture to extremely high temperatures (like a very hot oven, 550–650°C) using iron catalysts. It works, but it's energy-hungry and messy, like trying to cook a delicate soufflé in a blast furnace.

This paper asks: Can we use light instead of heat? Specifically, can we use a semiconductor material called Titanium Dioxide (TiO2) to act as a catalyst that uses sunlight (or UV light) to pull that hydrogen off gently and efficiently?

The author, Nico Yannik Merkt, used powerful computer simulations to figure out exactly how the atoms move and interact during this process.

The Stage: The Catalyst Surface

Think of the TiO2 surface as a dance floor.

• The Dancers: The ethylbenzene molecule (the guest) and the atoms on the TiO2 floor (the hosts).
• The Floor: The specific "dance floor" used in this study is a very flat, orderly section of the crystal called the (110) surface. It has rows of oxygen atoms and titanium atoms.

The Two Ways to Dance: Thermal vs. Photochemical

1. The Thermal Way (The "Slow Walk")

If you just heat the floor (no light), the reaction is slow and difficult.

• The Problem: The hydrogen atom is tightly holding onto the carbon. To break this bond, the floor has to act like a polite but firm host. It tries to pull the hydrogen away as a proton (a positive charge) while the electron stays behind. This is called Proton-Coupled Electron Transfer (PCET).
• The Analogy: Imagine trying to pull a heavy suitcase out of a tight suitcase rack. You have to wiggle it, pull the handle, and push the wheels all at once. It takes a lot of effort (high energy/heat).
• The Result: The second hydrogen is even harder to remove. The process gets stuck, requiring high temperatures to finish the job.

2. The Photochemical Way (The "Lightning Strike")

Now, shine a light on the floor.

• The Magic: When a photon (a particle of light) hits the TiO2, it kicks an electron out of its seat on the floor and sends it flying to a different spot. This leaves behind a "hole" (a missing electron), which acts like a super-charged vacuum cleaner.
• The Mechanism: This "hole" is so aggressive that it doesn't need to be polite. It grabs the hydrogen atom whole (proton + electron together) in a single, swift motion. This is called Hydrogen Atom Transfer (HAT).
• The Analogy: Instead of wiggling the suitcase, you use a magnet to yank the whole thing out instantly. It's much faster and requires less heat.

The Wavelength Mystery: Why Brighter Light Works Better

The paper investigates a real-world puzzle: Why does shining a specific high-energy light (257 nm, which is deep UV) produce seven times more styrene than a lower-energy light (343 nm)?

• The Low-Energy Light (343 nm): This is like giving the dancer a gentle nudge. It gets them moving, but they quickly get tired and fall back into a "resting" state (the ground state) before they can finish the dance. They hit a wall (an energy barrier) and can't finish the second step of the reaction.
• The High-Energy Light (257 nm): This is like giving the dancer a massive boost of adrenaline. The energy is so high that the dancer stays in a "super-activated" state the whole time. They can jump over the walls that stopped the low-energy dancers. They don't fall back to the resting state until the dance is completely finished.
• The "Hot Hole" Theory: The paper supports the idea that these high-energy "holes" are "hot" (full of extra energy) and can do work before they cool down.

The Twist: The Oxidized Floor

The paper also looked at what happens if the dance floor is "oxidized" (has extra oxygen atoms stuck to it).

• The Change: On a normal floor, the host has to be very careful and polite (PCET). On an oxidized floor, the extra oxygen acts like a pre-charged battery or a "hydrogen scavenger."
• The Result: The reaction becomes much easier. The extra oxygen grabs the hydrogen immediately (HAT), and the whole process speeds up. This explains why experiments show that pre-treating the catalyst with oxygen makes it four times more efficient.

The Computer Tools: The "Microscope"

To see all this, the author used two types of computer tools:

1. DFT (Density Functional Theory): Like a high-resolution camera. It's great at seeing the shape of the molecules and where they sit on the floor. However, it sometimes misses the complex "ghostly" interactions between electrons when bonds are breaking.
2. CASSCF (Multi-reference method): Like an X-ray vision that sees the quantum nature of electrons. It is much harder to use and takes a long time, but it is necessary to see what happens when the electrons get "confused" or "entangled" during the bond-breaking.

The Finding: The author found that the "camera" (DFT) often underestimated how stable the final product was and missed the complex electron dance. The "X-ray" (CASSCF) showed that the reaction involves a complex "biradical" state (two unpaired electrons dancing together) that the camera couldn't see clearly.

Summary of Conclusions

• Light is better than heat: Using light allows the reaction to happen at much lower temperatures.
• More energy is better: High-energy light (257 nm) keeps the reaction "alive" and moving, whereas lower energy light causes the reaction to stall.
• Oxygen helps: Adding extra oxygen to the catalyst surface acts as a shortcut, making the hydrogen removal much faster and more efficient.
• It's complicated: The reaction isn't a simple straight line; it involves electrons jumping between the molecule and the surface, creating temporary radical states that require advanced math to understand.

The paper concludes that to make this process a reality for industry, we need to understand these quantum steps to design better catalysts that can harness light efficiently without needing extreme heat.

Technical Summary

1. Problem Statement

The industrial production of styrene from ethylbenzene (EB) is a critical process for the polymer industry, currently relying on thermal dehydrogenation at high temperatures (550–650°C) over iron oxide catalysts. This method suffers from immense energy consumption, catalyst deactivation due to coking, and thermodynamic equilibrium limitations.

Photocatalysis using titanium dioxide (TiO2) offers a sustainable alternative, enabling reaction under milder conditions. However, the fundamental mechanisms governing this process remain poorly understood. Key challenges include:

• High Activation Barriers: The second C-H bond cleavage (at the $\beta$-carbon) is kinetically limited by a high activation barrier (~1.2 eV).
• Wavelength Dependence: Experimental data shows a significant increase in styrene yield when using higher energy photons (257 nm) compared to lower energy ones (343 nm), contradicting simple "hot carrier" relaxation models.
• Surface Sensitivity: The oxidation state of the TiO2 surface drastically alters activity, yet the electronic reasons for this enhancement are unclear.
• Methodological Limitations: Standard Density Functional Theory (DFT) often fails to accurately describe the multi-reference character (static correlation) inherent in radical formation, bond breaking, and excited states involved in photocatalysis.

2. Methodology

The study employs a hybrid quantum chemical approach to model the reaction on the rutile TiO2(110) surface.

• Surface Model: A "charged cluster" model (Ti27O54[O34]$^{-68}$) embedded in a point charge field (PCF) and effective core potentials (ECPs) was used. This approach captures the long-range electrostatic Madelung potential of the crystal while allowing for computationally intensive multi-reference calculations, which are infeasible with periodic boundary conditions for such large systems.
• Computational Methods:
  • DFT (UKS-PBE-D3): Used for extensive geometry optimizations, transition state searches (NEB method), and initial screening of reaction pathways. The PBE functional was chosen for its balance of accuracy and cost, with D3 corrections for dispersion forces.
  • Multi-Reference Methods (CASSCF/SA-CASSCF): Essential for describing excited states, biradical intermediates, and conical intersections. Active spaces of CAS(12,12) (stoichiometric surface) and CAS(14,14) (oxidized surface) were employed to capture the transfer of electrons between the organic $\pi$-system, C-H $\sigma$-bonds, and Ti(3d) orbitals.
  • Validation: The study attempted NEVPT2 (dynamic correlation) but found it numerically unstable for this specific charged cluster model due to artificial quasi-continuum states. Consequently, SA-CASSCF was used as the reference for electronic topology, while DFT provided geometric trends.
• Scope: The investigation covers thermal pathways, low-energy photoexcitation (343 nm), high-energy photoexcitation (257 nm), and the influence of surface oxidation (pre-adsorbed oxygen species).

3. Key Contributions

• Mechanistic Differentiation: The thesis distinguishes between Proton-Coupled Electron Transfer (PCET) on the stoichiometric surface and Hydrogen Atom Transfer (HAT) on the oxidized surface.
• Wavelength Dependence Explanation: It provides a theoretical basis for the "hot hole" theory, demonstrating that high-energy photons allow the system to navigate excited state manifolds (S1, S2, T2, T3) to bypass ground-state kinetic barriers, rather than simply generating charge carriers that relax to the band edge.
• Role of Surface Oxidation: It elucidates how pre-adsorbed oxygen ($O_{Ti}$) acts as a "hydrogen scavenger" and a photosensitizer, drastically lowering the excitation threshold and stabilizing radical intermediates.
• Methodological Insight: The work highlights the limitations of single-reference DFT in describing the deep thermodynamic sinks of styrene-surface complexes and the necessity of multi-reference methods to capture the biradical nature of the reaction.

4. Key Results

A. Adsorption and Electronic Structure

• Stoichiometric Surface: EB physisorbs via $\pi$-interaction with 5f-Ti sites. Upon photoexcitation, an Adsorbate-to-Surface Charge Transfer (ASCT) occurs, forming a spatially separated radical pair: an organic radical cation ($EB^{\bullet+}$) and a reduced $Ti^{3+}$ center.
• Oxidized Surface: The presence of $O_{Ti}$ introduces gap states, narrowing the band gap and lowering the vertical excitation energy from >3.5 eV to ~1.29 eV. The $O_{Ti}$ species acts as a radical center, facilitating direct HAT.

B. Reaction Mechanisms

• Thermal Pathway (Ground State):
  • Step I: $\alpha$-C-H cleavage is favored over $\beta$-cleavage.
  • Step II: The second cleavage ($\beta$-C-H) is the rate-determining step with a high barrier (~0.85 eV in DFT, ~1.26 eV in CASSCF).
  • Mechanism: Proceeds via PCET, where the electron transfers to Ti and the proton to bridging oxygen ($O_{br}$).
• Low-Energy Photoexcitation (343 nm):
  • The system is excited to S1 but rapidly relaxes into the ground state biradical manifold (S0/T1) before the second step.
  • The reaction is forced to overcome the high thermal barrier of the second step, resulting in low quantum yield.
• High-Energy Photoexcitation (257 nm):
  • The system accesses higher excited states (S2, T3) and remains on these excited potential energy surfaces throughout the reaction coordinate.
  • This "high-road" trajectory allows the system to bypass the ground-state rate-determining barrier, explaining the sevenfold increase in yield observed experimentally.
• Oxidized Surface Mechanism:
  • Step I: Proceeds via direct HAT to the pre-adsorbed $O_{Ti}$ radical, forming a stable intermediate.
  • Step II: The reaction branches into two pathways ("diff" to $O_{br}$ or "same" to $O_{Ti}$). Both pathways exhibit significantly lower barriers on the oxidized surface compared to the reduced surface.
  • Selectivity: While yield increases fourfold on the oxidized surface, selectivity drops due to side reactions (dimerization), but the primary mechanism is driven by the radical nature of $O_{Ti}$ preventing recombination.

C. DFT vs. CASSCF Comparison

• DFT Limitations: DFT (PBE) underestimates the stabilization energy of the product (styrene-surface complex) and fails to correctly track the lowest triplet roots in the presence of strong static correlation. It incorrectly predicts the "diff" pathway as kinetically favored on the oxidized surface, whereas CASSCF shows both pathways are nearly degenerate.
• CASSCF Accuracy: Correctly identifies the deep thermodynamic sink (~-3.9 eV) and the multi-reference character of the biradical intermediates, providing a more accurate description of the electronic landscape.

5. Significance and Outlook

This thesis provides a comprehensive quantum mechanical explanation for the efficiency of photocatalytic ethylbenzene dehydrogenation.

• Theoretical Impact: It validates the "hot hole" mechanism, showing that photon energy is not just a trigger for charge generation but a tool to navigate excited state topologies, avoiding kinetic traps.
• Catalyst Design: The findings suggest that surface pre-oxidation is a critical strategy to lower excitation thresholds and facilitate HAT mechanisms.
• Future Directions: The author suggests investigating other TiO2 phases (anatase), the role of transition metal dopants to promote $H_2$ desorption (avoiding water formation), and the use of heterojunctions to improve charge separation and selectivity.

In conclusion, the work demonstrates that the synergy between high-energy photons and specific surface oxidation states creates a unique reaction environment where excited-state dynamics override thermal limitations, offering a pathway for more energy-efficient styrene production.