Partial density of states representation for accurate deep neural network predictions of X-ray spectra.

The performance of a machine learning (ML) algorithm for chemistry is highly contingent upon the architect's choice of input representation. This work introduces the partial density of states (p-DOS) descriptor: a novel, quantum-inspired structural representation which encodes relevant electronic information for machine learning models seeking to simulate X-ray spectroscopy. p-DOS uses a minimal basis set in conjunction with a guess (non-optimised) electronic configuration to extract and then discretise the density of states (DOS) of the absorbing atom to form the input vector. We demonstrate that while the electronically-focused p-DOS performs well in isolation, optimal performance is achieved when supplemented with nuclear structural information imparted via a geometric representation. p-DOS provides a description of the key electronic properties of a system which is not only concise and computationally efficient, but also independent of molecular size or choice of basis set. It can be rapidly generated, facilitating its application with large training sets. Its performance is demonstrated using a wide variety of examples at the sulphur K-edge, including the prediction of ultrafast X-ray spectroscopic signal associated with photoexcited 2(5H)-thiophenone. These results highlight the potential for ML models developed using p-DOS to contribute to the interpretation and prediction of experimental results e.g. in operando measurements of batteries and/or catalysts and femtosecond time-resolved studies, especially those made possible by emergent cutting-edge technologies, especially X-ray free electron lasers.

Conformational Control of Donor-Acceptor Molecules Using Non-covalent Interactions.

Controlling the architecture of organic molecules is an important aspect in tuning the functional properties of components in organic electronics. For purely organic thermally activated delayed fluorescence (TADF) molecules, design is focused upon orthogonality orientated donor and acceptor units. In these systems, the rotational dynamics around the donor and acceptor bond has been shown to be critical for activating TADF; however, too much conformational freedom can increase the non-radiative rate, leading to a large energy dispersion of the emitting states and conformers, which do not exhibit TADF. To date, control of the motion around the D-A bond has focused upon steric hindrance. In this work, we computationally investigate eight proposed donor-acceptor molecules, exhibiting a B-N bond between the donor and acceptor. We compare the effect of steric hindrance and noncovalent interactions, achieved using oxygen (sulfur) boron heteroatom interactions, in exerting fine conformational control of the excited state dynamics. This work reveals the potential for judiciously chosen noncovalent interactions to strongly influence the functional properties of TADF emitters, including the accessible conformers and the energy dispersion associated with the charge transfer states.

Mechanochemical Polyoxometalate Super-Reduction with Lithium Metal.

In this first systematic investigation of mechanochemical polyoxometalate (POM) reduction, (TBA)3[PMo12O40] was reacted with n equiv of lithium metal (n = 1-24) to generate PMo12/n products which were shown to be mixtures of electron-rich PMo12Lix species. FTIR analysis revealed the lengthening/weakening of terminal Mo═O bonds with increasing levels of reduction, while EXAFS spectra indicated the onset of Mo-Mo bond formation at n ∼ 8 and a significant structural change at n > 12. Successive MoVI reductions were monitored by XANES and XPS, and at n = 24, results were consistent with the formation of at least one MoIV-MoIV bonded {MoIV3} triad together with MoV. Upon dissolution, the PMo12Lix species present in the solid PMo12/n products undergo electron exchange and single-peak 31P NMR spectra were observed for n = 1-12. For n ≥ 16, changes in solid state and solution 31P NMR spectra coincided with the emergence of features in the UV-vis spectra associated with MoV-MoV and {MoIV3} bonding in an ε-Keggin structure. Bonding between {Li(NCMe)}+ and 2-electron-reduced PMo12 in (TBA)4[PMo12O40{Li(NCMe)}] suggests that super-reduction gives rise to more extensive Li-O bonding that ultimately causes lithium-oxide-promoted TBA cation decomposition and POM degradation, which might explain the appearance of XPS peaks for Mo2C at n ≥ 16. This work has revealed some of the complex, unexplored chemistry of super-reduced POMs and establishes a new, solvent-free approach in the search for a better fundamental understanding of the electronic properties and reactivity of electron-rich nanoscale metal oxides.

Attosecond impulsive stimulated X-ray Raman scattering in liquid water.

We report the measurement of impulsive stimulated x-ray Raman scattering in neutral liquid water. An attosecond pulse drives the excitations of an electronic wavepacket in water molecules. The process comprises two steps: a transition to core-excited states near the oxygen atoms accompanied by transition to valence-excited states. Thus, the wavepacket is impulsively created at a specific atomic site within a few hundred attoseconds through a nonlinear interaction between the water and the x-ray pulse. We observe this nonlinear signature in an intensity-dependent Stokes Raman sideband at 526 eV. Our measurements are supported by our state-of-the-art calculations based on the polarization response of water dimers in bulk solvation and propagation of attosecond x-ray pulses at liquid density.

Web-CONEXS: an inroad to theoretical X-ray absorption spectroscopy.

Accurate analysis of the rich information contained within X-ray spectra usually calls for detailed electronic structure theory simulations. However, density functional theory (DFT), time-dependent DFT and many-body perturbation theory calculations increasingly require the use of advanced codes running on high-performance computing (HPC) facilities. Consequently, many researchers who would like to augment their experimental work with such simulations are hampered by the compounding of nontrivial knowledge requirements, specialist training and significant time investment. To this end, we present Web-CONEXS, an intuitive graphical web application for democratizing electronic structure theory simulations. Web-CONEXS generates and submits simulation workflows for theoretical X-ray absorption and X-ray emission spectroscopy to a remote computing cluster. In the present form, Web-CONEXS interfaces with three software packages: ORCA, FDMNES and Quantum ESPRESSO, and an extensive materials database courtesy of the Materials Project API. These software packages have been selected to model diverse materials and properties. Web-CONEXS has been conceived with the novice user in mind; job submission is limited to a subset of simulation parameters. This ensures that much of the simulation complexity is lifted and preliminary theoretical results are generated faster. Web-CONEXS can be leveraged to support beam time proposals and serve as a platform for preliminary analysis of experimental data.

Tracking nuclear motion in single-molecule magnets using femtosecond X-ray absorption spectroscopy.

The development of new data storage solutions is crucial for emerging digital technologies. Recently, all-optical magnetic switching has been achieved in dielectrics, proving to be faster than traditional methods. Despite this, single-molecule magnets (SMMs), which are an important class of magnetic materials due to their nanometre size, remain underexplored for ultrafast photomagnetic switching. Herein, we report femtosecond time-resolved K-edge X-ray absorption spectroscopy (TR-XAS) on a Mn(III)-based trinuclear SMM. Exploiting the elemental specificity of XAS, we directly track nuclear dynamics around the metal ions and show that the ultrafast dynamics upon excitation of a crystal-field transition are dominated by a magnetically active Jahn-Teller mode. Our results, supported by simulations, reveal minute bond length changes from 0.01 to 0.05 Å demonstrating the sensitivity of the method. These geometrical changes are discussed in terms of magneto-structural relationships and consequently our results illustrate the importance of TR-XAS for the emerging area of ultrafast molecular magnetism.

Spin-Vibronic Intersystem Crossing and Molecular Packing Effects in Heavy Atom Free Organic Phosphor.

We present a detailed investigation into the excited state properties of a planar D3h symmetric azatriangulenetrione, HTANGO, which has received significant interest due to its high solid-state phosphorescence quantum yield and therefore potential as an organic room temperature phosphorescent (ORTP) dye. Using a model linear vibronic coupling Hamiltonian in combination with quantum dynamics simulations, we observe that intersystem crossing (ISC) in HTANGO occurs with a rate of ∼1010 s-1, comparable to benzophenone, an archetypal molecule for fast ISC in heavy metal free molecules. Our simulations demonstrate that the mechanism for fast ISC is associated with the high density of excited triplet states which lie in close proximity to the lowest singlet states, offering multiple channels into the triplet manifold facilitating rapid population transfer. Finally, to rationalize the solid-state emission properties, we use quantum chemistry to investigate the excited state surfaces of the HTANGO dimer, highlighting the influence and importance of the rotational alignment between the two HTANGO molecules in the solid state and how this contributes to high phosphorescence quantum yield.

Improving the Conductivity of Amide-Based Small Molecules through Enhanced Molecular Packing and Their Application as Hole Transport Mediators in Perovskite Solar Cells.

Organic-inorganic hybrid halide perovskite solar cells (PSCs) have attracted substantial attention from the photovoltaic research community, with the power conversion efficiency (PCE) already exceeding 26%. Current state-of-the-art devices rely on Spiro-OMeTAD as the hole-transporting material (HTM); however, Spiro-OMeTAD is costly due to its complicated synthesis and expensive product purification, while its low conductivity ultimately limits the achievable device efficiency. In this work, we build upon our recently introduced family of low-cost amide-based small molecules and introduce a molecule (termed TPABT) that results in high conductivity values (∼10-5 S cm-1 upon addition of standard ionic additives), outperforming our previous amide-based material (EDOT-Amide-TPA, ∼10-6 S cm-1) while only costing an estimated $5/g. We ascribe the increased optoelectronic properties to favorable molecular packing, as shown by single-crystal X-ray diffraction, which results in close spacing between the triphenylamine blocks. This, in turn, results in a short hole-hopping distance between molecules and therefore good mobility and conductivity. In addition, TPABT exhibits a higher bandgap and is as a result more transparent in the visible range of the solar spectrum, leading to lower parasitic absorption losses than Spiro-OMeTAD, and has increased moisture stability. We applied the molecule in perovskite solar cells and obtained good efficiency values in the ∼15% range. Our approach shows that engineering better molecular packing may be the key to developing high-efficiency, low-cost HTMs for perovskite solar cells.

A Δ-learning strategy for interpretation of spectroscopic observables.

Accurate computations of experimental observables are essential for interpreting the high information content held within x-ray spectra. However, for complicated systems this can be difficult, a challenge compounded when dynamics becomes important owing to the large number of calculations required to capture the time-evolving observable. While machine learning architectures have been shown to represent a promising approach for rapidly predicting spectral lineshapes, achieving simultaneously accurate and sufficiently comprehensive training data is challenging. Herein, we introduce Δ-learning for x-ray spectroscopy. Instead of directly learning the structure-spectrum relationship, the Δ-model learns the structure dependent difference between a higher and lower level of theory. Consequently, once developed these models can be used to translate spectral shapes obtained from lower levels of theory to mimic those corresponding to higher levels of theory. Ultimately, this achieves accurate simulations with a much reduced computational burden as only the lower level of theory is computed, while the model can instantaneously transform this to a spectrum equivalent to a higher level of theory. Our present model, demonstrated herein, learns the difference between TDDFT(BLYP) and TDDFT(B3LYP) spectra. Its effectiveness is illustrated using simulations of Rh L3-edge spectra tracking the C-H activation of octane by a cyclopentadienyl rhodium carbonyl complex.

Disentangling the evolution of electrons and holes in photoexcited ZnO nanoparticles.

The evolution of charge carriers in photoexcited room temperature ZnO nanoparticles in solution is investigated using ultrafast ultraviolet photoluminescence spectroscopy, ultrafast Zn K-edge absorption spectroscopy, and ab initio molecular dynamics (MD) simulations. The photoluminescence is excited at 4.66 eV, well above the band edge, and shows that electron cooling in the conduction band and exciton formation occur in <500 fs, in excellent agreement with theoretical predictions. The x-ray absorption measurements, obtained upon excitation close to the band edge at 3.49 eV, are sensitive to the migration and trapping of holes. They reveal that the 2 ps transient largely reproduces the previously reported transient obtained at 100 ps time delay in synchrotron studies. In addition, the x-ray absorption signal is found to rise in ∼1.4 ps, which we attribute to the diffusion of holes through the lattice prior to their trapping at singly charged oxygen vacancies. Indeed, the MD simulations show that impulsive trapping of holes induces an ultrafast expansion of the cage of Zn atoms in <200 fs, followed by an oscillatory response at a frequency of ∼100 cm-1, which corresponds to a phonon mode of the system involving the Zn sub-lattice.

An on-the-fly deep neural network for simulating time-resolved spectroscopy: predicting the ultrafast ring opening dynamics of 1,2-dithiane.

Revolutionary developments in ultrafast light source technology are enabling experimental spectroscopists to probe the structural dynamics of molecules and materials on the femtosecond timescale. The capacity to investigate ultrafast processes afforded by these resources accordingly inspires theoreticians to carry out high-level simulations which facilitate the interpretation of the underlying dynamics probed during these ultrafast experiments. In this Article, we implement a deep neural network (DNN) to convert excited-state molecular dynamics simulations into time-resolved spectroscopic signals. Our DNN is trained on-the-fly from first-principles theoretical data obtained from a set of time-evolving molecular dynamics. The train-test process iterates for each time-step of the dynamics data until the network can predict spectra with sufficient accuracy to replace the computationally intensive quantum chemistry calculations required to produce them, at which point it simulates the time-resolved spectra for longer timescales. The potential of this approach is demonstrated by probing dynamics of the ring opening of 1,2-dithiane using sulphur K-edge X-ray absorption spectroscopy. The benefits of this strategy will be more markedly apparent for simulations of larger systems which will exhibit a more notable computational burden, making this approach applicable to the study of a diverse range of complex chemical dynamics.

Ionic Accumulation as a Diagnostic Tool in Perovskite Solar Cells: Characterizing Band Alignment with Rapid Voltage Pulses.

Despite record-breaking devices, interfaces in perovskite solar cells are still poorly understood, inhibiting further progress. Their mixed ionic-electronic nature results in compositional variations at the interfaces, depending on the history of externally applied biases. This makes it difficult to measure the band energy alignment of charge extraction layers accurately. As a result, the field often resorts to a trial-and-error process to optimize these interfaces. Current approaches are typically carried out in a vacuum and on incomplete cells, hence values may not reflect those found in working devices. To address this, a pulsed measurement technique characterizing the electrostatic potential energy drop across the perovskite layer in a functioning device is developed. This method reconstructs the current-voltage (JV) curve for a range of stabilization biases, holding the ion distribution "static" during subsequent rapid voltage pulses. Two different regimes are observed: at low biases, the reconstructed JV curve is "s-shaped", whereas, at high biases, typical diode-shaped curves are returned. Using drift-diffusion simulations, it is demonstrated that the intersection of the two regimes reflects the band offsets at the interfaces. This approach effectively allows measurements of interfacial energy level alignment in a complete device under illumination and without the need for expensive vacuum equipment.

Uncertainty quantification of spectral predictions using deep neural networks.

We investigate the performance of uncertainty quantification methods, namely deep ensembles and bootstrap resampling, for deep neural network (DNN) predictions of transition metal K-edge X-ray absorption near-edge structure (XANES) spectra. Bootstrap resampling combined with our multi-layer perceptron (MLP) model provides an accurate assessment of uncertainty with >90% of all predicted spectral intensities falling within ±3σ of the true values for held-out data across the nine first-row transition metal K-edge XANES spectra.

Atomic-scale observation of solvent reorganization influencing photoinduced structural dynamics in a copper complex photosensitizer.

Photochemical reactions in solution are governed by a complex interplay between transient intramolecular electronic and nuclear structural changes and accompanying solvent rearrangements. State-of-the-art time-resolved X-ray solution scattering has emerged in the last decade as a powerful technique to observe solute and solvent motions in real time. However, disentangling solute and solvent dynamics and how they mutually influence each other remains challenging. Here, we simultaneously measure femtosecond X-ray emission and scattering to track both the intramolecular and solvation structural dynamics following photoexcitation of a solvated copper photosensitizer. Quantitative analysis assisted by molecular dynamics simulations reveals a two-step ligand flattening strongly coupled to the solvent reorganization, which conventional optical methods could not discern. First, a ballistic flattening triggers coherent motions of surrounding acetonitrile molecules. In turn, the approach of acetonitrile molecules to the copper atom mediates the decay of intramolecular coherent vibrations and induces a further ligand flattening. These direct structural insights reveal that photoinduced solute and solvent motions can be intimately intertwined, explaining how the key initial steps of light harvesting are affected by the solvent on the atomic time and length scale. Ultimately, this work takes a step forward in understanding the microscopic mechanisms of the bidirectional influence between transient solvent reorganization and photoinduced solute structural dynamics.

Mind the GAP: quantifying the breakdown of the linear vibronic coupling Hamiltonian.

Excited state dynamics play a critical role across a broad range of scientific fields. Importantly, the highly non-equilibrium nature of the states generated by photoexcitation means that excited state simulations should usually include an accurate description of the coupled electronic-nuclear motion, which often requires solving the time-dependent Schrödinger equation (TDSE). One of the biggest challenges for these simulations is the requirement to calculate the PES over which the nuclei evolve. An effective approach for addressing this challenge is to use the approximate linear vibronic coupling (LVC) Hamiltonian, which enables a model potential to be parameterised using relatively few quantum chemistry calculations. However, this approach is only valid provided there are no large amplitude motions in the excited state dynamics. In this paper we introduce and deploy a metric, the global anharmonicity parameter (GAP), which can be used to assess the accuracy of an LVC potential. Following its derivation, we illustrate its utility by applying it to three molecules exhibiting different rigidity in their excited states.

Excited-State Lifetime Modulation by Twisted and Tilted Molecular Design in Carbene-Metal-Amide Photoemitters.

Carbene-metal-amides (CMAs) are an emerging class of photoemitters based on a linear donor-linker-acceptor arrangement. They exhibit high flexibility about the carbene-metal and metal-amide bonds, leading to a conformational freedom which has a strong influence on their photophysical properties. Herein we report CMA complexes with (1) nearly coplanar, (2) twisted, (3) tilted, and (4) tilt-twisted orientations between donor and acceptor ligands and illustrate the influence of preferred ground-state conformations on both the luminescence quantum yields and excited-state lifetimes. The performance is found to be optimum for structures with partially twisted and/or tilted conformations, resulting in radiative rates exceeding 1 × 106 s-1. Although the metal atoms make only small contributions to HOMOs and LUMOs, they provide sufficient spin-orbit coupling between the low-lying excited states to reduce the excited-state lifetimes down to 500 ns. At the same time, high photoluminescence quantum yields are maintained for a strongly tilted emitter in a host matrix. Proof-of-concept organic light-emitting diodes (OLEDs) based on these new emitter designs were fabricated, with a maximum external quantum efficiency (EQE) of 19.1% with low device roll-off efficiency. Transient electroluminescence studies indicate that molecular design concepts for new CMA emitters can be successfully translated into the OLED device.

A Computational-Experimental Approach to Unravel the Excited State Landscape in Heavy-Atom Free BODIPY-Related Dyes.

We performed a time-gated laser-spectroscopy study in a set of heavy-atom free single BODIPY fluorophores, supported by accurate, excited-state computational simulations of the key low-lying excited states in these chromophores. Despite the strong fluorescence of these emitters, we observed a significant fraction of time-delayed (microseconds scale) emission associated with processes that involved passage through the triplet manifold. The accuracy of the predictions of the energy arrangement and electronic nature of the low-lying singlet and triplet excited states meant that an unambiguous assignment of the main deactivation pathways, including thermally activated delayed fluorescence and/or room temperature phosphorescence, was possible. The observation of triplet state formation indicates a breakthrough in the "classic" interpretation of the photophysical properties of the renowned BODIPY and its derivatives.

Circularly polarised luminescence in an RNA-based homochiral, self-repairing, coordination polymer hydrogel.

The aqueous equimolar reaction of Ag(i) ions with the thionucleoside enantiomer (-)6-thioguanosine, ((-)6tGH), yields a one-dimensional coordination polymer {Ag(-)tG} n , the self-assembly of which generates left-handed helical chains. The resulting helicity induces an enhanced chiro-optical response compared to the parent ligand. DFT calculations indicate that this enhancement is due to delocalisation of the excited state along the helical chains, with 7 units being required to converge the calculated CD spectra. At concentrations ≥15 mmol l-1 reactions form a sample-spanning hydrogel which shows self-repair capabilities with instantaneous recovery in which the dynamic reversibility of the coordination chains appears to play a role. The resulting gel exhibits circularly polarised luminescence (CPL) with a large dissymmetry factor of -0.07 ± 0.01 at 735 nm, a phenomenon not previously observed for this class of coordination polymer.

Efficient Adversarial Generation of Thermally Activated Delayed Fluorescence Molecules.

Adversarial generative models are becoming an essential tool in molecular design and discovery due to their efficiency in exploring the desired chemical space with the assistance of deep learning. In this article, we introduce an integrated framework by combining the modules of algorithmic synthesis, deep prediction, adversarial generation, and fine screening for the purpose of effective design of the thermally activated delayed fluorescence (TADF) molecules that can be used in the organic light-emitting diode devices. The retrosynthetic rules are employed to algorithmically synthesize the D-A complex based on the empirically defined donor and acceptor moieties, which is followed by the high-throughput labeling and prediction with the deep neural network. The new D-A molecules are subsequently generated via the adversarial autoencoder, with the excited-state property distributions perfectly matching those of the original samples. Fine screening of the generated molecules, including the spin-orbital coupling calculation and the excited-state optimization, is eventually implemented to select the qualified TADF candidates within the novel chemical space. Further investigation shows that the created structures fully mimic the original D-A samples by maintaining a significant charge transfer characteristic, a minimal adiabatic singlet-triplet gap, and a moderate spin-orbital coupling that are desirable for the delayed fluorescence.