Login / Signup

Electronic Structures of Small Stoichiometric Zn x O x Clusters.

Shivangi VaishAbigail O GyamfiCaleb D HuizengaHrant P HratchianCaroline Chick Jarrold
Published in: The journal of physical chemistry. A (2024)
Stoichiometric Zn x O x clusters in the subnanometer size regime have been the topic of several computational and mass spectrometry studies that showed the particular stability of the stoichiometric species relative to nonstoichiometric species (Zn x O y , x ≠ y ). In the current study, we present the angle-resolved anion photoelectron (PE) spectra of stoichiometric Zn x O x - clusters (2 ≤ x ≤ 5), which are interpreted with supporting computational studies that include natural orbital ionization calculations on detachment transition cross sections. All spectra show evidence of D x h ring structures, which had been predicted to be the most stable structures in previous computational studies. However, a new lowest energy isomer is reported for the Zn 2 O 2 anion and neutral, a bent chain, which is readily reconciled with the most intense feature in the Zn 2 O 2 - PE spectrum. The computed PE angular distributions (PADs) associated with the lowest energy cluster structures identified computationally agree with the experimental results, with the exception of Zn 5 O 5 - , the experimental PAD of which suggests that strong vibronic coupling may be introducing anomalies. While the lowest lying electronic state of the Zn 2 O 2 chain structure is a triplet state, all neutral ring structures (including Zn 2 O 2 , the anion of which also populates the ion beam), favor a singlet electronic state. The computed singlet-triplet splitting of the D x h structures increases monotonically with x . Overall, we find that the properties of the ring structures evolve smoothly, rather than in the punctuated manner typically seen in the small cluster size regime.
Keyphrases
  • heavy metals
  • high resolution
  • magnetic resonance imaging
  • risk assessment
  • density functional theory
  • machine learning
  • deep learning