Joachim Sauer

We assess the accuracy of electronic structure methods for modeling the dispersion-dominated adsorption of methane, ethane, and propane, and the adsorption of hydrogen-bonded water in the Brønsted acidic zeolite chabazite (H–CHA) using density functional theory (DFT) with periodic boundary conditions and a hybrid QM:QM (QM = quantum mechanics) approach. In addition to coupled-cluster (CCSD(T)) reference energies, which agree with available experimental values within chemical accuracy limits of ±4 kJ mol–1, we also provide reference adsorption structures obtained with Møller–Plesset perturbation theory (MP2). DFT mostly overestimates binding compared to CCSD(T), with root-mean-square deviations (RMSDs) ranging from 6–29 kJ mol–1 for alkanes and 5–33 kJ mol–1 for water. QM:QM calculations with hybrid functionals as high level can reduce RMSDs from the CCSD(T) reference. With B3LYP+D4 as a high-level method, the QM:QM energies deviate by 7–12 and 11–21 kJ mol–1 for alkanes and water, respectively, depending on the low-level method. With ωB97M-V as high-level method, the RMSDs are 4–10 and 4–15 kJ mol–1, and with MP2, they are 1–3 and 2–8 kJ mol–1, respectively. When ωB97M-V is used as a high-level method or r2SCAN+D4 as a low-level method, predictions for the protonation state of water can be qualitatively wrong. For DFT modeling of periodic systems, when hybrid functionals are not affordable, we recommend revPBE+D3. For QM:QM with hybrid DFT as high-level method, ωB97M-V:revPBE+D3 is the best combination. Only CCSD(T) single points on hybrid MP2 structures reliably deliver chemically accurate adsorption energies and qualitatively correct adsorption structures.

Multicenter transition metal clusters often exhibit a dense manifold of low-lying electronic states arising from the interplay of ligand-field splitting and spin coupling. Together with structural and valence isomerism, these factors make vibrational spectral assignments challenging. The infrared photodissociation (IRPD) spectra of AlM ₂ O ₄ ⁺ (M = Fe, Co) have resisted interpretation because the predicted spectra are sensitive to the isomer and its electronic state. Here we settle the assignment by employing multireference methods in a localized active space framework to identify the low-energy structures and their ground-state electronic configurations, followed by vibrational analysis using Kohn--Sham density functional theory. We show that the experimental IRPD spectra arise from the coexistence of two structural isomers, each in a low-spin ground state. Vibrational spectra calculated for the identified ground states reproduce the experimental IRPD features, whereas those of alternative states do not. This assignment is consistent with reported experimental observations, namely the IRPD spectra, the ion mobility distributions, and the absence of methane reactivity. Analysis of the electronic structure reveals that ligand-field effects dominate over spin coupling in shaping both energetics and vibrational spectra of these clusters. These results show the power of multireference methods capable of treating large active spaces, which are often required for binuclear complexes.

The design of improved metal-organic frameworks (MOFs) for water harvesting requires the reliable prediction of adsorption isotherms, i.e., Gibbs free energies of adsorption, with no other input than the atomic positions. We employ density functional theory (DFT) and show that, in Mg-MOF-74, well-defined adsorption structures exist for water loadings of n = 1, 2, 3, 4, and 5 molecules per Mg²⁺ ion. The first water molecule attaches to the open metal site, while on adsorption of subsequent molecules, structures with an increasing number of hydrogen bonds per molecule form: dimers (n = 2), chains in pore direction (n = 3), and a monolayer on the pore wall (n = 4). For n = 5, a tube-like stack of water trimers connected to the monolayer fills the pore completely, and all water molecules are 4-fold coordinated. For isotherm predictions, we use a Multisite Langmuir model with Gibbs free energies of -33, -19, -13, -10, and -21 kJ/mol for the steps leading to adsorption states with n = 1, 2, 3, 4, and 5 molecules, respectively. The close agreement of the predicted total isotherm with experiment corresponds to an accuracy of ±2 kJ/mol for Gibbs free energies. This is achieved only after adding "high-level" Coupled Cluster corrections (0, 3, 9, 8, and 11 kJ/mol for n = 1, 2, 3, 4, and 5, respectively) to the DFT energies. We show that the large variations observed between different experimental isotherms can be explained by sample imperfections or incomplete evacuation of samples before isotherm measurements.

Zeolites play a pivotal role as catalysts in biomass conversion, and understanding how alcohols interact with these materials is essential for designing next-generation catalysts that reduce dependence on nonrenewable fuels and enable a more sustainable energy landscape. Achieving this understanding requires accurate free-energy calculations. In this study, we present the anharmonic enthalpy, entropy, and vibrational free energy associated with ethanol adsorption on H-ZSM-5, computed at a hybrid DFT level of theory. We critically assess the need to go beyond both GGA functionals and the harmonic approximation to achieve chemically accurate vibrational free energies. The vibrational contributions to the free energy are obtained using a newly implemented strategy that combines the DOS-P method with a QM/QM molecular-dynamics scheme. This approach yields significant improvement in accuracy compared to the commonly used PBE + D2 functional. By combining the anharmonic vibrational free energies computed at the B3LYP + D2 level with electronic energies refined at the CCSD(T) level, we obtain an anharmonic adsorption enthalpy, ΔH, of -101.9 kJ mol^-1, the entropic term, -TΔS, equal to 61.0 kJ mol^-1, and a free energy of adsorption, ΔG, of -40.8 kJ mol^-1 at 313 K. We performed a critical comparison between these values and the experimental data. Notably, the entropic term agrees well with the experimental value of 62.0 kJ mol^-1. We also examine the significance of anharmonicity, finding a correction of approximately 1 kJ mol^-1 for both the vibrational entropy and enthalpy, and about 3.7 kJ mol^-1 for the ZPE.

We employ a chemically accurate (±4 kJ mol ⁻¹ ) embedding approach to calculate enthalpies for CO ₂ , N ₂ , and H ₂ adsorption in the Zn ²⁺ containing zeolites FAU and CHA and for CO ₂ and H ₂ in the metal–organic frameworks (MOFs) Zn-MOF-74 and CALF-20.

The metal‐support interaction (MSI) effect of different supports, namely, N‐doped graphene (N‐graphene) and CeO ₂ (111), on the properties of the planar 2D and 3D isomers of copper clusters of atomicity five was computationally studied at the PBE level (+U for Ce). For graphene‐based systems, both pristine and defective sheets were considered. The stability, geometry and electronic properties of the supported Cu ₅ clusters are discussed, and the dissociation of O ₂ is studied on the most stable systems for both isomers. As expected, the interaction between Cu ₅ and clean graphene or N‐graphene is weak, and although adding nitrogen to graphene enhances the interaction, proper chemical adsorption is only achieved on the defects explored. In contrast, both Cu ₅ isomers interact strongly with CeO ₂ (111). The weaker MSI with N‐graphene materials allows the deformation of the Cu ₅ cluster upon O ₂ adsorption, whereas CeO ₂ (111) supported structures are more robust. O ₂ dissociation on N‐graphene‐supported Cu ₅ is consistent with our previous results on gas‐phase clusters, i.e. the 2D isomer is more resistant to oxidation than the 3D one but the latter is more easily reducible, while on CeO ₂ (111) this trend is reversed due to the strong MSI.

We model water adsorption at pairs of proximate Brønsted acid sites (BASs) in zeolites H-MFI, H-FAU, and H-CHA. We use machine-learning potentials to explore the potential energy surface, combined with quantum mechanical methods for chemically accurate energies of selected structures. We identify BAS pairs that adsorb water cooperatively, forming an H-bonded chain that connects the two BASs and provides additional stabilization. The formation of such a water bridge requires at least two molecules, making the adsorption of the second water molecule stronger than the first, e.g., by 20 and 44 kJ mol^-1 for an Al9-Al10 and an Al4-Al6 pair, respectively, in H-MFI, and by 11 kJ mol^-1 for H-FAU. The magnitude of this extra-stabilization depends on the relative alignment of the BASs. Both Al pairs separated by just one SiO₄ tetrahedron (next-nearest neighbor sites) and pairs across a 10-membered ring are included. The increase of the heat of adsorption with the water loading per BAS contrasts with the decrease obtained for isolated BASs and aligns with observations in some experiments.

The discovery of crystalline reticular materials remains largely trial-and-error despite their societal importance. We introduce our algorithmic iterative reticular synthesis (AIRES) cycle, which integrates automated synthesis, image recognition, single-crystal X-ray diffraction and, crucially, customized algorithmic decision-making, to maximize distinct crystal discoveries rather than optimizing single targets. Demonstrated on zeolitic imidazolate frameworks (ZIFs), AIRES achieves twice the discovery rate of random exploration, crystallizing 10 new linkers into diverse ZIF topologies and expanding the single-linker Zn-ZIF library by one-third. By transforming reticular synthesis from an empirical process to a systematic exploration, AIRES provides a scalable and efficient blueprint for accelerating materials discovery.