Publications

@article{Franceschi2022,

title = {Reconstruction changes drive surface diffusion and determine the flatness of oxide surfaces},

author = {Giada Franceschi and Michael Schmid and Ulrike Diebold and Michele Riva},

doi = {10.1116/6.0001704},

year  = {2022},

date = {2022-02-22},

urldate = {2022-02-22},

journal = {Journal of Vacuum Science & Technology A},

volume = {40},

number = {2},

pages = {023206},

publisher = {American Vacuum Society},

abstract = {Surface diffusion on metal oxides is key in many areas of materials technology, yet it has been scarcely explored at the atomic scale. This work provides phenomenological insights from scanning tunneling microscopy on the link between surface diffusion, surface atomic structure, and oxygen chemical potential based on three model oxide surfaces: Fe}2}O_{3}(1-102), La_{1−x}Sr_{x}MnO_{3}(110), and In_{2}O_{3}(111). In all instances, changing the oxygen chemical potential used for annealing stabilizes reconstructions of different compositions while promoting the flattening of the surface morphology—a sign of enhanced surface diffusion. It is argued that thermodynamics, rather than kinetics, rules surface diffusion under these conditions: the composition change of the surface reconstructions formed at differently oxidizing conditions drives mass transport across the surface.},

keywords = {P02},

pubstate = {published},

tppubtype = {article}

}

@article{ACSMEASURE2022,

title = {Why and How Savitzky–Golay Filters Should Be Replaced},

author = {Michael Schmid and David Rath and Ulrike Diebold},

url = {https://pubs.acs.org/doi/10.1021/acsmeasuresciau.1c00054},

doi = {10.1021/acsmeasuresciau.1c00054},

year  = {2022},

date = {2022-02-17},

urldate = {2022-02-17},

journal = {ACS Measurement Science Au},

volume = {2},

number = {2},

pages = {185--196},

abstract = {Savitzky–Golay (SG) filtering, based on local least-squares fitting of the data by polynomials, is a popular method for smoothing data and calculations of derivatives of noisy data. At frequencies above the cutoff, SG filters have poor noise suppression; this unnecessarily reduces the signal-to-noise ratio, especially when calculating derivatives of the data. In addition, SG filtering near the boundaries of the data range is prone to artifacts, which are especially strong when using SG filters for calculating derivatives of the data. We show how these disadvantages can be avoided while keeping the advantageous properties of SG filters. We present two classes of finite impulse response (FIR) filters with substantially improved frequency response: (i) SG filters with fitting weights in the shape of a window function and (ii) convolution kernels based on the sinc function with a Gaussian-like window function and additional corrections for improving the frequency response in the passband (modified sinc kernel). Compared with standard SG filters, the only price to pay for the improvement is a moderate increase in the kernel size. Smoothing at the boundaries of the data can be improved with a non-FIR method, the Whittaker–Henderson smoother, or by linear extrapolation of the data, followed by convolution with a modified sinc kernel, and we show that the latter is preferable in most cases. We provide computer programs and equations for the smoothing parameters of these smoothers when used as plug-in replacements for SG filters and describe how to choose smoothing parameters to preserve peak heights in spectra.},

keywords = {P02},

pubstate = {published},

tppubtype = {article}

}

@article{Liu2022,

title = {Phase transitions of zirconia: Machine-learned force fields beyond density functional theory},

author = {Peitao Liu and Carla Verdi and Ferenc Karsai and Georg Kresse},

doi = {10.1103/physrevb.105.l060102},

year  = {2022},

date = {2022-02-16},

journal = {Physical Review B},

volume = {105},

number = {6},

pages = {L060102},

publisher = {American Physical Society (APS)},

abstract = {Machine-learned force fields (MLFFs) are increasingly used to accelerate first-principles simulations of many materials properties. However, MLFFs are generally trained from density functional theory (DFT) data and thus suffer from the same limitations as DFT. To achieve more predictive accuracy, MLFFs based on higher levels of theory are required, but the training becomes exceptionally arduous. Here, we present an approach to generate MLFFs with beyond DFT accuracy which combines an efficient on-the-fly active learning method and Δ-machine learning. Using this approach, we generate an MLFF for zirconia based on the random phase approximation (RPA). Specifically, an MLFF trained on the fly during DFT-based molecular dynamics simulations is corrected by another MLFF that is trained on the differences between RPA and DFT calculated energies, forces, and stress tensors. We show that owing to the relatively smooth nature of these differences, the expensive RPA calculations can be performed only on a small number of representative structures of small unit cells selected by rank compression of the kernel matrix. This dramatically reduces the computational cost and allows one to generate an MLFF fully capable of reproducing high-level quantum-mechanical calculations beyond DFT. We carefully validate our approach and demonstrate its success in studying the phase transitions of zirconia. These results open the way to many-body calculations of finite-temperature properties of materials.},

keywords = {P03},

pubstate = {published},

tppubtype = {article}

}

@article{Tiyatha2022,

title = {Oxidative coupling of methane—comparisons of MnTiO_{3}–Na_{2}WO_{4} and MnO_{x}–TiO_{2}–Na_{2}WO_{4} catalysts on different silica supports},

author = {Worapinit Tiyatha and Thanaphat Chukeaw and Sarannuch Sringam and Thongthai Witoon and Metta Chareonpanich and Günther Rupprechter and Anusorn Seubsai},

url = {https://www.nature.com/articles/s41598-022-06598-6#citeas},

doi = {10.1038/s41598-022-06598-6},

year  = {2022},

date = {2022-02-16},

urldate = {2022-02-16},

journal = {Scientific Reports},

volume = {12},

pages = {2595},

publisher = {Springer Science and Business Media LLC},

abstract = {The oxidative coupling of methane (OCM) converts CH_{4} to value-added chemicals (C_{2+}), such as olefins and paraffin. For a series of MnTiO_{3}-Na_{2}WO_{4} (MnTiO_{3}-NW) and MnO_{x}-TiO_{2}-Na_{2}WO_{4} (Mn-Ti-NW), the effect of loading of MnTiO_{3} or MnO_{x}-TiO_{2}, respectively, on two different supports (sol–gel SiO_{2} (SG) and commercial fumed SiO_{2} (CS)) was examined. The catalyst with the highest C_{2+} yield (21.6% with 60.8% C_{2}+ selectivity and 35.6% CH_{4} conversion) was 10 wt% MnTiO_{3}-NW/SG with an olefins/paraffin ratio of 2.2. The catalyst surfaces with low oxygen-binding energies were associated with high CH_{4} conversion. Stability tests conducted for over 24 h revealed that SG-supported catalysts were more durable than those on CS because the active phase (especially Na_{2}WO_{4}) was more stable in SG than in CS. With the use of SG, the activity of MnTiO_{3}-NW was not substantially different from that of Mn-Ti-NW, especially at high metal loading.},

keywords = {P08, TACO-associated},

pubstate = {published},

tppubtype = {article}

}

@article{Shi2022,

title = {High-performance water gas shift induced by asymmetric oxygen vacancies: Gold clusters supported by ceria-praseodymia mixed oxides},

author = {Junjie Shi and Hailian Li and Alexander Genest and Weixuan Zhao and Pengfei Qi and Tao Wang and Günther Rupprechter},

doi = {10.1016/j.apcatb.2021.120789},

year  = {2022},

date = {2022-02-01},

urldate = {2022-02-01},

journal = {Applied Catalysis B: Environmental},

volume = {301},

pages = {120789},

publisher = {Elsevier BV},

abstract = {Modifying and controlling sites at the metal/oxide interface is an effective way of tuning catalytic activity, beneficial for bifunctional catalysis by reducible oxide supported \underline{metal nanoparticles}. We employed mixed ceria-praseodymia supported Au clusters for the \underline{water gas shift} reaction (WGSR). Varying the Ce:Pr ratio (4:1, 2:1, 1:4) not only allows to control the number of oxygen vacancies but, even more important, their local coordination, with asymmetrically coordinated O# being most active for water activation. These effects have been examined by X-ray absorption near edge structure (XANES), extended X-ray absorption fine structure (EXAFS), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy, temperature programmed desorption/reduction (TPD/TPR), and density functional theory (DFT). Using the WGSR performance of Au/CeOx as reference, Au/Ce_{4}Pr_{1}O_{x} was identified to exhibit the highest activity, with a CO conversion of 75% at 300 °C, which is about 5-times that of Au/CeO_{x}. Au/Ce_{4}Pr_{1}O_{x} also showed excellent stability, with the conversion still being 70% after 50 h time-on-stream at 300 °C. Although a higher Pr content leads to more O vacancies, the catalytic activity showed a “volcano behavior”. Based on DFT, this was rationalized via the formation energy of oxygen vacancies, the binding energy of water, and the asymmetry of the O# site. The presented route of creating active vacancy sites should also be relevant for other heterogeneous catalytic systems.},

keywords = {P08},

pubstate = {published},

tppubtype = {article}

}

@article{Kiatsaengthong2022,

title = {Effects of Mg, Ca, Sr, and Ba Dopants on the Performance of La_{2}O_{3} Catalysts for the Oxidative Coupling of Methane},

author = {Danusorn Kiatsaengthong and Kanticha Jaroenpanon and Pooripong Somchuea and Thanaphat Chukeaw and Metta Chareonpanich and Kajornsak Faungnawakij and Hiesang Sohn and Günther Rupprechter and Anusorn Seubsai},

doi = {10.1021/acsomega.1c04738},

year  = {2022},

date = {2022-01-04},

urldate = {2022-01-04},

journal = {ACS Omega},

volume = {7},

number = {2},

pages = {1785--1793},

publisher = {American Chemical Society (ACS)},

abstract = {Oxidative coupling of methane (OCM) is a reaction to directly convert methane into high value-added hydrocarbons (C_{2+}) such as ethylene and ethane using molecular oxygen and a catalyst. This work investigated lanthanum oxide catalysts for OCM, which were promoted with alkaline-earth metal oxides (Mg, Ca, Sr, and Ba) and prepared by the solution-mixing method. The synthesized catalysts were characterized using X-ray powder diffraction, CO_{2}-programmed desorption, and X-ray photoelectron spectroscopy. The comparative performance of each promoter showed that promising lanthanum-loaded alkaline-earth metal oxide catalysts were La-Sr and La-Ba. In contrast, the combination of La with Ca or Mg did not lead to a clear improvement of C_{2+} yield. The most promising LaSr50 catalyst exhibited the highest C_{2+} yield of 17.2%, with a 56.0% C_{2+} selectivity and a 30.9% CH_{4} conversion. Catalyst characterization indicated that their activity was strongly associated with moderate basic sites and surface-adsorbed oxygen species of O_{2}^{–}. Moreover, the catalyst was stable over 25 h at a reactor temperature of 700 °C.},

keywords = {P08},

pubstate = {published},

tppubtype = {article}

}

@article{MontesCampos2021,

title = {A Differentiable Neural-Network Force Field for Ionic Liquids},

author = {Hadrián Montes-Campos and Jesús Carrete and Sebastian Bichelmaier and Luis M Varela and Georg K. H. Madsen},

doi = {10.1021/acs.jcim.1c01380},

year  = {2022},

date = {2022-01-03},

urldate = {2022-01-03},

journal = {Journal of Chemical Information and Modeling},

volume = {62},

number = {1},

pages = {88--101},

abstract = {We present NeuralIL, a model for the potential energy of an ionic liquid that accurately reproduces first-principles results with orders-of-magnitude savings in computational cost. Based on a multilayer perceptron and spherical Bessel descriptors of the atomic environments, NeuralIL is implemented in such a way as to be fully automatically differentiable. It can thus be trained on ab-initio forces instead of just energies, to make the most out of the available data, and can efficiently predict arbitrary derivatives of the potential energy. We parametrize the model for the case of ethylammonium nitrate. We discuss the best way to include chemical information in the atom-centered descriptors for a many-component system. Furthermore, we demonstrate an ensemble-learning approach to the detection of extrapolation. With out-of-sample accuracies better than 0.1 kcal/mol in the energies and 100 meV/Å in the forces, our potential model considerably outperforms molecular-mechanics force fields and opens the door to large-scale thermodynamical calculations with ab-initio-like accuracy for ionic liquids. Including the forces does away with the idea that vast amounts of atomic configurations are required to train a neural network force field based on atom-centered descriptors. We also find that a separate treatment of long-range interactions is not required to achieve a high-quality representation of the potential 

energy surface of these dense ionic systems.},

keywords = {P09},

pubstate = {published},

tppubtype = {article}

}

@article{Haunold2021,

title = {LiOx-modification of Ni and Co_{3}O_{4} surfaces: An XPS, LEIS and LEED study},

author = {Thomas Haunold and Günther Rupprechter},

doi = {10.1016/j.susc.2021.121915},

year  = {2021},

date = {2021-11-01},

urldate = {2021-11-01},

journal = {Surface Science},

volume = {713},

pages = {121915},

publisher = {Elsevier BV},

abstract = {LiO_{x} was deposited at room temperature by physical vapor deposition (PVD) on polycrystalline Ni foil and Co_{3}O_{4}(111) thin film, creating uniform model systems well-suited for surface-sensitive characterization by X-ray photoelectron spectroscopy (XPS), low energy ion scattering (LEIS) or low energy electron diffraction (LEED). In the case of Ni, about 15 layers of LiO_{x} film were grown under the current conditions either stepwise or continuously, with XPS analysis indicating a deposition rate of 0.16 and 0.24 ML/min, respectively. Li 1s and O 1s spectra revealed that Li_{2}O and to a lesser extent LiOH were preferentially formed. The stability of the LiO_{x}films was examined in UHV, upon annealing at 573 K and upon hydrogen reduction at 723 K. On the more reactive Co_{3}O_{4}(111) film grown on Ir(100), the Li accommodation rate was about twice as high, at least within the first minutes of deposition. Post-deposition LEED showed an obscured cobalt oxide diffraction pattern, not unexpected in light of the LiO_{x} deposited. On both substrates, LEIS characterization of Li (≈ 103 eV) was prevented by the high background in this kinetic energy region, due to surface roughness and unspecific scattering. Still, LiO_{x} deposition was evident from the vanished LEIS signals of Ni or Co. The prepared LiO_{x}-modified surfaces may serve as starting point for the future growth of epitaxial Li_{x}CoO_{2} model systems.},

keywords = {P08},

pubstate = {published},

tppubtype = {article}

}

@article{Verdi2021,

title = {Thermal transport and phase transitions of zirconia by on-the-fly machine-learned interatomic potentials},

author = {Carla Verdi and Ferenc Karsai and Peitao Liu and Ryosuke Jinnouchi and Georg Kresse},

doi = {10.1038/s41524-021-00630-5},

year  = {2021},

date = {2021-09-30},

urldate = {2021-09-30},

journal = {npj Computational Materials},

volume = {7},

pages = {156},

publisher = {Springer Science and Business Media LLC},

abstract = {Machine-learned interatomic potentials enable realistic finite temperature calculations of complex materials properties with first-principles accuracy. It is not yet clear, however, how accurately they describe anharmonic properties, which are crucial for predicting the lattice thermal conductivity and phase transitions in solids and, thus, shape their technological applications. Here we employ a recently developed on-the-fly learning technique based on molecular dynamics and Bayesian inference in order to generate an interatomic potential capable to describe the thermodynamic properties of zirconia, an important transition metal oxide. This machine-learned potential accurately captures the temperature-induced phase transitions below the melting point. We further showcase the predictive power of the potential by calculating the heat transport on the basis of Green–Kubo theory, which allows to account for anharmonic effects to all orders. This study indicates that machine-learned potentials trained on the fly offer a routine solution for accurate and efficient simulations of the thermodynamic properties of a vast class of anharmonic materials.},

keywords = {P03},

pubstate = {published},

tppubtype = {article}

}