Rational Functionalization of UiO-66 with Pd Nanoparticles: Synthesis and In Situ Fourier-Transform Infrared Monitoring.

Functionalization of metal-organic frameworks (MOFs) with noble metal nanoparticles (NPs) is a challenging task. Conventional impregnation by metals often leads to agglomerates on the surface of MOF crystals. Functional groups on linkers interact with metal precursors and promote the homogeneous distribution of NPs in the pores of MOFs, but their uncontrolled localization can block channels and thus hinder mass transport. To overcome this problem, we created nucleation centers only in the defective pores of the UiO-66 MOF via the postsynthesis exchange. First, we have introduced defects into UiO-66 using benzoic acid as a modulator. Second, the modulator was exchanged for amino-benzoic acid. As a result, amino groups have decorated mainly the defective pores and attracted the Pd precursor after impregnation. The interaction of the metal precursor with amino groups and the growth of NPs were monitored by in situ infrared spectroscopy. Three processes were distinguished: the gaseous HCl release, NH2 reactivation, and growth of extended Pd surfaces. Uniform Pd NPs were located in the pores because of the homogeneous distribution of the precursor and pore diffusion-limited nucleation rate. Our work demonstrates an alternative approach of controlled Pd incorporation into UiO-66 that is of great importance for the rational design of heterogeneous catalysts.

Adsorption Sites on Pd Nanoparticles Unraveled by Machine-Learning Potential with Adaptive Sampling.

Catalytic properties of noble-metal nanoparticles (NPs) are largely determined by their surface morphology. The latter is probed by surface-sensitive spectroscopic techniques in different spectra regions. A fast and precise computational approach enabling the prediction of surface-adsorbate interaction would help the reliable description and interpretation of experimental data. In this work, we applied Machine Learning (ML) algorithms for the task of adsorption-energy approximation for CO on Pd nanoclusters. Due to a high dependency of binding energy from the nature of the adsorbing site and its local coordination, we tested several structural descriptors for the ML algorithm, including mean Pd-C distances, coordination numbers (CN) and generalized coordination numbers (GCN), radial distribution functions (RDF), and angular distribution functions (ADF). To avoid overtraining and to probe the most relevant positions above the metal surface, we utilized the adaptive sampling methodology for guiding the ab initio Density Functional Theory (DFT) calculations. The support vector machines (SVM) and Extra Trees algorithms provided the best approximation quality and mean absolute error in energy prediction up to 0.12 eV. Based on the developed potential, we constructed an energy-surface 3D map for the whole Pd55 nanocluster and extended it to new geometries, Pd79, and Pd85, not implemented in the training sample. The methodology can be easily extended to adsorption energies onto mono- and bimetallic NPs at an affordable computational cost and accuracy.

Pd nanoparticle growth monitored by DRIFT spectroscopy of adsorbed CO.

Synchrotron-based X-ray absorption spectroscopy and scattering are known in situ probes of metal nanoparticles (NPs). A limited number of laboratory techniques allow post-synthesis diagnostics of the active metal surface area. This work demonstrates the high potential of infrared spectroscopy as an in situ laboratory probe for the growth of metal NPs on a substrate. We introduce a small fraction of CO molecules into the reaction mixture as a probe to monitor the reduction kinetics of the Pd2+ precursor on ceria in hydrogen.

Machine Learning for Quantitative Structural Information from Infrared Spectra: The Case of Palladium Hydride.

Infrared spectroscopy (IR) is a widely used technique enabling to identify specific functional groups in the molecule of interest based on their characteristic vibrational modes or the presence of a specific adsorption site based on the characteristic vibrational mode of an adsorbed probe molecule. The interpretation of an IR spectrum is generally carried out within a fingerprint paradigm by comparing the observed spectral features with the features of known references or theoretical calculations. This work demonstrates a method for extracting quantitative structural information beyond this approach by application of machine learning (ML) algorithms. Taking palladium hydride formation as an example, Pd-H pressure-composition isotherms are reconstructed using IR data collected in situ in diffuse reflectance using CO molecule as a probe. To the best of the knowledge, this is the first example of the determination of continuous structural descriptors (such as interatomic distance and stoichiometric coefficient) from the fine structure of vibrational spectra, which opens new possibilities of using IR spectra for structural analysis.

Optimal Dynamic Regimes for CO Oxidation Discovered by Reinforcement Learning.

Metal nanoparticles are widely used as heterogeneous catalysts to activate adsorbed molecules and reduce the energy barrier of the reaction. Reaction product yield depends on the interplay between elementary processes: adsorption, activation, desorption, and reaction. These processes, in turn, depend on the inlet gas composition, temperature, and pressure. At a steady state, the active surface sites may be inaccessible due to adsorbed reagents. Periodic regime may thus improve the yield, but the appropriate period and waveform are not known in advance. Dynamic control should account for surface and atmospheric modifications and adjust reaction parameters according to the current state of the system and its history. In this work, we applied a reinforcement learning algorithm to control CO oxidation on a palladium catalyst. The policy gradient algorithm was trained in the theoretical environment, parametrized from experimental data. The algorithm learned to maximize the CO2 formation rate based on CO and O2 partial pressures for several successive time steps. Within a unified approach, we found optimal stationary, periodic, and nonperiodic regimes for different problem formulations and gained insight into why the dynamic regime can be preferential. In general, this work contributes to the task of popularizing the reinforcement learning approach in the field of catalytic science.

In Situ FTIR Spectroscopy for Scanning Accessible Active Sites in Defect-Engineered UiO-66.

Three UiO-66 samples were prepared by solvothermal synthesis using the defect engineering approach with benzoic acid as a modulator. They were characterized by different techniques and their acidic properties were assessed by FTIR spectroscopy of adsorbed CO and CD3CN. All samples evacuated at room temperature contained bridging µ3-OH groups that interacted with both probe molecules. Evacuation at 250 °C leads to the dehydroxylation and disappearance of the µ3-OH groups. Modulator-free synthesis resulted in a material with open Zr sites. They were detected by low-temperature CO adsorption on a sample evacuated at 200 °C and by CD3CN even on a sample evacuated at RT. However, these sites were lacking in the two samples obtained with a modulator. IR and Raman spectra revealed that in these cases, the Zr4+ defect sites were saturated by benzoates, which prevented their interaction with probe molecules. Finally, the dehydroxylation of all samples produced another kind of bare Zr sites that did not interact with CO but formed complexes with acetonitrile, probably due to structural rearrangement. The results showed that FTIR spectroscopy is a powerful tool for investigating the presence and availability of acid sites in UiO-66, which is crucial for its application in adsorption and catalysis.

Photoswitchable Zirconium MOF for Light-Driven Hydrogen Storage.

Here, we report a new photosensitive metal-organic framework (MOF) that was constructed via the modification of UiO-66-NH2 with diarylethene molecules (DAE, 4-(5-Methoxy-1,2-dimethyl-1H-indol-3-yl)-3-(2,5-dimethylthiophen-3-yl)-4-furan-2,5-dione). The material that was obtained was a highly crystalline porous compound. The photoresponse of the modified MOF was observed via UV-Vis and IR spectroscopy. Most of the DAE molecules inside of the UiO-66-pores had an open conformation after synthesis. However, the equilibrium was able to be shifted further toward an open conformation using visible light irradiation with a wavelength of 520 nm. Conversely, UV-light with a wavelength of 450 nm initiated the transformation of the photoresponsive moieties inside of the pores to a closed modification. We have shown that this transformation could be used to stimulate hydrogen adsorption-desorption processes. Specifically, visible light irradiation increased the H2 capacity of modified MOF, while UV-light decreased it. A similar hybrid material with DAE moieties in the UiO-66 scaffold was applied for hydrogen storage for the first time. Additionally, the obtained results are promising for smart H2 storage that is able to be managed via light stimuli.

Dehydrogenation of Ethylene on Supported Palladium Nanoparticles: A Double View from Metal and Hydrocarbon Sides.

Adsorption of ethylene on palladium, a key step in various catalytic reactions, may result in a variety of surface-adsorbed species and formation of palladium carbides, especially under industrially relevant pressures and temperatures. Therefore, the application of both surface and bulk sensitive techniques under reaction conditions is important for a comprehensive understanding of ethylene interaction with Pd-catalyst. In this work, we apply in situ X-ray absorption spectroscopy, X-ray diffraction and infrared spectroscopy to follow the evolution of the bulk and surface structure of an industrial catalysts consisting of 2.6 nm supported palladium nanoparticles upon exposure to ethylene under atmospheric pressure at 50 °C. Experimental results were complemented by ab initio simulations of atomic structure, X-ray absorption spectra and vibrational spectra. The adsorbed ethylene was shown to dehydrogenate to C2H3, C2H2 and C2H species, and to finally decompose to palladium carbide. Thus, this study reveals the evolution pathway of ethylene on industrial Pd-catalyst under atmospheric pressure at moderate temperatures, and provides a conceptual framework for the experimental and theoretical investigation of palladium-based systems, in which both surface and bulk structures exhibit a dynamic nature under reaction conditions.

Taking a snapshot of the triplet excited state of an OLED organometallic luminophore using X-rays.

OLED technology beyond small or expensive devices requires light-emitters, luminophores, based on earth-abundant elements. Understanding and experimental verification of charge transfer in luminophores are needed for this development. An organometallic multicore Cu complex comprising Cu-C and Cu-P bonds represents an underexplored type of luminophore. To investigate the charge transfer and structural rearrangements in this material, we apply complementary pump-probe X-ray techniques: absorption, emission, and scattering including pump-probe measurements at the X-ray free-electron laser SwissFEL. We find that the excitation leads to charge movement from C- and P- coordinated Cu sites and from the phosphorus atoms to phenyl rings; the Cu core slightly rearranges with 0.05 Å increase of the shortest Cu-Cu distance. The use of a Cu cluster bonded to the ligands through C and P atoms is an efficient way to keep structural rigidity of luminophores. Obtained data can be used to verify computational methods for the development of luminophores.