Epitaxial CeO2 Films as a Host for Quantum Applications

This study demonstrates that high-quality, isotopically pure CeO2 thin films grown by pulsed laser deposition serve as an effective host for quantum applications, exhibiting significantly longer photoluminescence lifetimes for Er-doped samples compared to Tm-doped ones due to the latter's non-radiative recombination pathways caused by strong O 2p and Tm 4f orbital overlap.

The Gist

Imagine you are trying to build a super-precise clock that can keep time for the entire universe without ever ticking wrong. In the world of quantum computing, these "clocks" are called quantum emitters. They are tiny particles that hold information (like a memory bit) in a state of perfect order, known as coherence.

However, there's a problem. Imagine trying to keep a spinning top perfectly still on a table that is shaking violently. In quantum materials, the "shaking" comes from the atomic nuclei of the material itself. Many atoms have tiny magnetic spins (like tiny magnets) that wiggle around and mess up the quantum information. This is called decoherence.

The Perfect Host: A Silent Room

To fix this, scientists usually have to build their clocks out of "isotopically purified" materials—essentially, they have to buy a very expensive, custom-made table where every single atom is a non-magnetic type.

This paper introduces a new, naturally perfect table: Cerium Oxide (CeO₂).

• The Analogy: Think of Cerium Oxide as a library where every single book is written by an author who has no voice. Because the atoms in this material have zero magnetic spin, they don't wiggle or make noise. It's a "magnetically silent" room. This means quantum information can sit there without being disturbed by the environment, potentially for a very long time.

The Experiment: Two Different Keys

The researchers wanted to see if they could put "keys" (quantum emitters) into this silent library. They chose two types of rare-earth ions as keys: Thulium (Tm) and Erbium (Er). They grew thin films of Cerium Oxide on silicon and other substrates using a high-tech laser process (like a very precise 3D printer for atoms).

Here is what they found:

1. The "Perfect Fit" (Structural Quality)

First, they checked if the keys actually fit into the holes in the library shelves.

• The Result: They used powerful microscopes and X-rays to look at the films. The results showed that the Tm and Er atoms were sitting perfectly in the spots where Cerium atoms should be. The film was smooth as glass and perfectly aligned. It was a high-quality, single-crystal structure.

2. The "Short-Lived" vs. "Long-Lived" Keys

Next, they tested how long the keys could hold their "quantum memory" (how long they glowed after being lit up).

• The Erbium (Er) Key: This one was a superstar. It glowed for a long time (about 2.9 to 5.3 milliseconds). In the quantum world, that's an eternity! It's like a lightbulb that stays on for a long time after you flip the switch.
• The Thulium (Tm) Key: This one was much shorter-lived. It only glowed for a tiny fraction of a second (microseconds). It was like a lightbulb that flickers and dies almost instantly.

Why the difference?
The researchers used a supercomputer to simulate the atomic world (DFT calculations) to find the answer.

• The Metaphor: Imagine the Erbium atom is wearing a thick, soundproof coat (its electron shells). It sits in the library, and the "noise" from the oxygen atoms in the Cerium Oxide can't reach it. It stays isolated and calm.
• The Thulium Problem: The Thulium atom's coat has a tear in it. Its electrons are "leaking" out and mixing with the oxygen atoms in the library. This mixing creates a "shortcut" for the energy to escape. Instead of glowing (radiative decay), the energy leaks away as heat or vibration (non-radiative decay). This is why the Thulium key dies so fast.

The Up-Conversion Surprise

They also discovered something cool with the Thulium. Even though it dies fast, they found a way to make it glow brighter by hitting it with two low-energy infrared lasers at once. It's like pushing a swing twice in just the right rhythm to make it go higher. This "up-conversion" could be useful for specific types of quantum communication.

The Big Picture

This paper is a roadmap for building better quantum computers.

1. The Host: Cerium Oxide is a fantastic, naturally quiet host for quantum bits because it doesn't have magnetic noise.
2. The Lesson: Just because a material is quiet doesn't mean any key will work. You have to pick the right key (dopant) that doesn't "leak" into the host.
3. The Future: By choosing the right combination (like Erbium in Cerium Oxide), scientists can create quantum devices that hold information for much longer, bringing us closer to real-world quantum technology.

In short: They built a perfectly silent room (Cerium Oxide) and tested two different guests. One guest (Erbium) stayed calm and quiet for a long time, while the other (Thulium) got too excited and left quickly because it didn't fit the room's atmosphere as well. Now we know which guest to invite for the party!

Technical Summary

1. Problem Statement

Quantum emitters in solid-state hosts are often limited by spin bath decoherence caused by hyperfine interactions with lattice nuclei possessing non-zero nuclear magnetic moments (e.g., $^{13}$C in diamond or $^{29}$Si in SiC). While isotopic purification can mitigate this, it is costly and complex.
The authors propose Cerium Oxide (CeO$_2$) as an ideal host because:

• All stable isotopes of Cerium have zero nuclear spin.
• The only spin-active isotope of Oxygen ($^{17}$O) has a negligible natural abundance (0.04%).
• This makes CeO$_2$ effectively a magnetically purified host, theoretically capable of preserving intrinsic quantum coherence without isotopic enrichment.

However, the performance of rare-earth (RE) dopants in CeO$_2$ depends heavily on the specific dopant-host electronic interaction. Previous studies have shown varying lifetimes, and the underlying mechanisms for non-radiative recombination pathways in different RE-doped CeO$_2$ systems were not fully understood.

2. Methodology

The study employed a combination of advanced thin-film growth, structural characterization, optical spectroscopy, and first-principles calculations:

• Film Growth: High-quality epitaxial thin films of CeO$_2$ doped with Thulium (Tm$^{3+}$) (1% and 10% concentrations) and Erbium (Er$^{3+}$) (1% concentration) were grown on YSZ (001) and Si (001) substrates using Pulsed Laser Deposition (PLD).
  • Growth conditions included optimized substrate temperatures (~820°C), oxygen partial pressures, and the use of a buffer layer for Si substrates to manage interfacial oxides.
• Structural Characterization:
  • XRD/XRR: Confirmed single-crystalline fluorite structure, out-of-plane orientation, film thickness, and surface roughness.
  • RBS/C (Rutherford Backscattering Spectrometry): Measured elemental concentration and ion channeling to determine the substitutional fraction of dopants.
  • STEM/EDS: Atomic-resolution imaging and elemental mapping to verify homogeneous doping and lattice incorporation.
• Optical Characterization:
  • Photoluminescence (PL): Measured emission spectra at room temperature and cryogenic temperatures (4 K) under various excitation wavelengths (UV and NIR).
  • Time-Resolved PL: Used time-correlated single-photon counting (TCSPC) to measure excited-state lifetimes.
  • Up-conversion: Investigated NIR-to-visible/NIR up-conversion processes.
• Computational Modeling: Density Functional Theory (DFT+U) calculations were performed to analyze the electronic density of states (DOS), specifically investigating the hybridization between RE 4f orbitals and O 2p orbitals.

3. Key Contributions

• Demonstration of High-Quality Epitaxy: Successfully grew atomically smooth, single-crystalline Tm:CeO$_2$ and Er:CeO$_2$ films on both YSZ and Si substrates, achieving near-perfect substitutional doping (~98% for Tm).
• Comparative Lifetime Analysis: Provided a direct comparison of photoluminescence lifetimes between Tm and Er dopants in the same host material grown under identical conditions.
• Mechanistic Insight via DFT: Correlated experimental lifetime differences with electronic structure calculations, revealing that the hybridization of dopant 4f states with host O 2p states dictates non-radiative recombination rates.
• Up-conversion Discovery: Observed and characterized efficient up-conversion emission in Tm:CeO$_2$ driven by NIR excitation.

4. Key Results

Structural Quality

• YSZ Substrates: Films exhibited superior crystallinity with a sharp rocking curve (FWHM ~0.04° for the film peak) and compressive strain due to lattice mismatch.
• Si Substrates: Films grew with a (111) orientation (lower surface energy) and tensile strain. While crystalline quality was slightly lower than on YSZ, they were suitable for optical verification without rare-earth background noise from the substrate.
• Dopant Incorporation: RBS channeling confirmed a minimum yield ($\chi_{min}$) of ~3.09% for Tm and ~5.49% for Ce, indicating excellent crystallinity. The calculated substitutional fraction for Tm was 98.22%, confirming Tm atoms occupy Ce lattice sites. STEM/EDS maps showed homogeneous distribution.

Optical Properties & Lifetimes

• Tm:CeO$_2$ (Thulium):
  • Showed strong NIR emission at ~793 nm ($^3$H$_4 \to ^3$H$_6$) and ~808 nm.
  • Lifetimes: Significantly short. The 793/808 nm transitions showed bi-exponential decay with components of ~3.8 $\mu$s and ~14.8 $\mu$s. The 1210 nm transition ($^3$H$_5 \to ^3$H$_6$) had a longer lifetime of ~68.8 $\mu$s.
  • Observed up-conversion emission when excited at ~1210 nm.
• Er:CeO$_2$ (Erbium):
  • Showed emission at ~1530 nm and ~1535 nm ($^4$I$_{13/2} \to ^4$I$_{15/2}$).
  • Lifetimes: Significantly longer. The 1534 nm transition exhibited a lifetime of ~5.32 ms, and the 1530 nm transition showed ~2.94 ms.
  • These lifetimes are comparable to or exceed those reported for MBE-grown films with much lower doping concentrations, suggesting low defect densities in the PLD films.

Theoretical Findings (DFT)

• Er-Doped CeO$_2$: The Er 4f states are deep in the valence band (−8 to −4 eV) and well-separated from the O 2p band. The 4f electrons remain localized and shielded, resulting in narrow, atomic-like peaks and minimal hybridization. This isolation prevents non-radiative pathways, leading to long lifetimes.
• Tm-Doped CeO$_2$: The Tm 4f states are higher in energy, overlapping significantly with the O 2p valence band. This leads to strong covalent hybridization (Tm 4f + O 2p).
  • Charge density analysis showed ~1 $\mu_B$ of spin density delocalized onto neighboring oxygen atoms.
  • This hybridization creates non-radiative recombination channels, explaining the drastically reduced lifetimes in Tm-doped films compared to Er-doped films.

5. Significance

• Host Selection Strategy: The study establishes that while CeO$_2$ is an excellent magnetically purified host, the choice of dopant is critical. Er$^{3+}$ is identified as a superior candidate for long-coherence quantum applications in CeO$_2$ due to its electronic isolation, whereas Tm$^{3+}$ suffers from rapid non-radiative decay due to host-dopant hybridization.
• Scalability: Demonstrates that high-quality, quantum-grade RE-doped oxide films can be grown on standard substrates (including Silicon) using PLD, a scalable technique, rather than just Molecular Beam Epitaxy (MBE).
• Fundamental Physics: Provides a clear link between electronic structure (hybridization) and quantum coherence times, offering a predictive framework for selecting dopants in wide-bandgap oxide hosts.
• Future Applications: The long lifetimes of Er:CeO$_2$ (up to ~5 ms) make it a promising platform for quantum memories, single-photon sources, and quantum networks operating in the telecommunication band, leveraging the magnetic purity of the CeO$_2$ host.