Single-Exciton Photoluminescence in a GaN Monolayer inside an AlN Nanocolumn.

GaN/AlN heterostructures with thicknesses of one monolayer (ML) are currently considered to be the most promising material for creating UVC light-emitting devices. A unique functional property of these atomically thin quantum wells (QWs) is their ability to maintain stable excitons, resulting in a particularly high radiation yield at room temperature. However, the intrinsic properties of these excitons are substantially masked by the inhomogeneous broadening caused, in particular, by fluctuations in the QWs' thicknesses. In this work, to reduce this effect, we fabricated cylindrical nanocolumns of 50 to 5000 nm in diameter using GaN/AlN single QW heterostructures grown via molecular beam epitaxy while using photolithography with a combination of wet and reactive ion etching. Photoluminescence measurements in an ultrasmall QW region enclosed in a nanocolumn revealed that narrow lines of individual excitons were localized on potential fluctuations attributed to 2-3-monolayer-high GaN clusters, which appear in QWs with an average thickness of 1 ML. The kinetics of luminescence with increasing temperature is determined via the change in the population of localized exciton states. At low temperatures, spin-forbidden dark excitons with lifetimes of ~40 ns predominate, while at temperatures elevated above 120 K, the overlying bright exciton states with much faster recombination dynamics determine the emission.

2D-GaN/AlN Multiple Quantum Disks/Quantum Well Heterostructures for High-Power Electron-Beam Pumped UVC Emitters.

This article describes GaN/AlN heterostructures for ultraviolet-C (UVC) emitters with multiple (up to 400 periods) two-dimensional (2D)-quantum disk/quantum well structures with the same GaN nominal thicknesses of 1.5 and 16 ML-thick AlN barrier layers, which were grown by plasma-assisted molecular-beam epitaxy in a wide range of gallium and activated nitrogen flux ratios (Ga/N2*) on c-sapphire substrates. An increase in the Ga/N2* ratio from 1.1 to 2.2 made it possible to change the 2D-topography of the structures due to a transition from the mixed spiral and 2D-nucleation growth to a purely spiral growth. As a result, the emission energy (wavelength) could be varied from 5.21 eV (238 nm) to 4.68 eV (265 nm) owing to the correspondingly increased carrier localization energy. Using electron-beam pumping with a maximum pulse current of 2 A at an electron energy of 12.5 keV, a maximum output optical power of 50 W was achieved for the 265 nm structure, while the structure emitting at 238 nm demonstrated a power of 10 W.

Embeddings from protein language models predict conservation and variant effects.

The emergence of SARS-CoV-2 variants stressed the demand for tools allowing to interpret the effect of single amino acid variants (SAVs) on protein function. While Deep Mutational Scanning (DMS) sets continue to expand our understanding of the mutational landscape of single proteins, the results continue to challenge analyses. Protein Language Models (pLMs) use the latest deep learning (DL) algorithms to leverage growing databases of protein sequences. These methods learn to predict missing or masked amino acids from the context of entire sequence regions. Here, we used pLM representations (embeddings) to predict sequence conservation and SAV effects without multiple sequence alignments (MSAs). Embeddings alone predicted residue conservation almost as accurately from single sequences as ConSeq using MSAs (two-state Matthews Correlation Coefficient-MCC-for ProtT5 embeddings of 0.596 ± 0.006 vs. 0.608 ± 0.006 for ConSeq). Inputting the conservation prediction along with BLOSUM62 substitution scores and pLM mask reconstruction probabilities into a simplistic logistic regression (LR) ensemble for Variant Effect Score Prediction without Alignments (VESPA) predicted SAV effect magnitude without any optimization on DMS data. Comparing predictions for a standard set of 39 DMS experiments to other methods (incl. ESM-1v, DeepSequence, and GEMME) revealed our approach as competitive with the state-of-the-art (SOTA) methods using MSA input. No method outperformed all others, neither consistently nor statistically significantly, independently of the performance measure applied (Spearman and Pearson correlation). Finally, we investigated binary effect predictions on DMS experiments for four human proteins. Overall, embedding-based methods have become competitive with methods relying on MSAs for SAV effect prediction at a fraction of the costs in computing/energy. Our method predicted SAV effects for the entire human proteome (~ 20 k proteins) within 40 min on one Nvidia Quadro RTX 8000. All methods and data sets are freely available for local and online execution through bioembeddings.com, https://github.com/Rostlab/VESPA , and PredictProtein.

Monolayer-Scale GaN/AlN Multiple Quantum Wells for High Power e-Beam Pumped UV-Emitters in the 240-270 nm Spectral Range.

Monolayer (ML)-scale GaN/AlN multiple quantum well (MQW) structures for electron-beam-pumped ultraviolet (UV) emitters are grown on c-sapphire substrates by using plasma-assisted molecular beam epitaxy under controllable metal-rich conditions, which provides the spiral growth of densely packed atomically smooth hillocks without metal droplets. These structures have ML-stepped terrace-like surface topology in the entire QW thickness range from 0.75-7 ML and absence of stress at the well thickness below 2 ML. Satisfactory quantum confinement and mitigating the quantum-confined Stark effect in the stress-free MQW structures enable one to achieve the relatively bright UV cathodoluminescence with a narrow-line (~15 nm) in the sub-250-nm spectral range. The structures with many QWs (up to 400) exhibit the output optical power of ~1 W at 240 nm, when pumped by a standard thermionic-cathode (LaB6) electron gun at an electron energy of 20 keV and a current of 65 mA. This power is increased up to 11.8 W at an average excitation energy of 5 µJ per pulse, generated by the electron gun with a ferroelectric plasma cathode at an electron-beam energy of 12.5 keV and a current of 450 mA.

The Effect of Interface Diffusion on Raman Spectra of Wurtzite Short-Period GaN/AlN Superlattices.

We present an extensive theoretical and experimental study to identify the effect on the Raman spectrum due to interface interdiffusion between GaN and AlN layers in short-period GaN/AlN superlattices (SLs). The Raman spectra for SLs with sharp interfaces and with different degree of interface diffusion are simulated by ab initio calculations and within the framework of the random-element isodisplacement model. The comparison of the results of theoretical calculations and experimental data obtained on PA MBE and MOVPE grown SLs, showed that the bands related to A1(LO) confined phonons are very sensitive to the degree of interface diffusion. As a result, a correlation between the Raman spectra in the range of A1(LO) confined phonons and the interface quality in SLs is obtained. This opens up new possibilities for the analysis of the structural characteristics of short-period GaN/AlN SLs using Raman spectroscopy.

Phonons in Short-Period GaN/AlN Superlattices: Group-Theoretical Analysis, Ab initio Calculations, and Raman Spectra.

We report the results of experimental and theoretical studies of phonon modes in GaN/AlN superlattices (SLs) with a period of several atomic layers, grown by submonolayer digital plasma-assisted molecular-beam epitaxy, which have a great potential for use in quantum and stress engineering. Using detailed group-theoretical analysis, the genesis of the SL vibrational modes from the modes of bulk AlN and GaN crystals is established. Ab initio calculations in the framework of the density functional theory, aimed at studying the phonon states, are performed for SLs with both equal and unequal layer thicknesses. The frequencies of the vibrational modes are calculated, and atomic displacement patterns are obtained. Raman spectra are calculated and compared with the experimental ones. The results of the ab initio calculations are in good agreement with the experimental Raman spectra and the results of the group-theoretical analysis. As a result of comprehensive studies, the correlations between the parameters of acoustic and optical phonons and the structure of SLs are obtained. This opens up new possibilities for the analysis of the structural characteristics of short-period GaN/AlN SLs using Raman spectroscopy. The results obtained can be used to optimize the growth technologies aimed to form structurally perfect short-period GaN/AlN SLs.

Protein-protein and protein-nucleic acid binding residues important for common and rare sequence variants in human.

BACKGROUND: Any two unrelated people differ by about 20,000 missense mutations (also referred to as SAVs: Single Amino acid Variants or missense SNV). Many SAVs have been predicted to strongly affect molecular protein function. Common SAVs (> 5% of population) were predicted to have, on average, more effect on molecular protein function than rare SAVs (< 1% of population). We hypothesized that the prevalence of effect in common over rare SAVs might partially be caused by common SAVs more often occurring at interfaces of proteins with other proteins, DNA, or RNA, thereby creating subgroup-specific phenotypes. We analyzed SAVs from 60,706 people through the lens of two prediction methods, one (SNAP2) predicting the effects of SAVs on molecular protein function, the other (ProNA2020) predicting residues in DNA-, RNA- and protein-binding interfaces. RESULTS: Three results stood out. Firstly, SAVs predicted to occur at binding interfaces were predicted to more likely affect molecular function than those predicted as not binding (p value < 2.2 × 10-16). Secondly, for SAVs predicted to occur at binding interfaces, common SAVs were predicted more strongly with effect on protein function than rare SAVs (p value < 2.2 × 10-16). Restriction to SAVs with experimental annotations confirmed all results, although the resulting subsets were too small to establish statistical significance for any result. Thirdly, the fraction of SAVs predicted at binding interfaces differed significantly between tissues, e.g. urinary bladder tissue was found abundant in SAVs predicted at protein-binding interfaces, and reproductive tissues (ovary, testis, vagina, seminal vesicle and endometrium) in SAVs predicted at DNA-binding interfaces. CONCLUSIONS: Overall, the results suggested that residues at protein-, DNA-, and RNA-binding interfaces contributed toward predicting that common SAVs more likely affect molecular function than rare SAVs.

Modeling aspects of the language of life through transfer-learning protein sequences.

BACKGROUND: Predicting protein function and structure from sequence is one important challenge for computational biology. For 26 years, most state-of-the-art approaches combined machine learning and evolutionary information. However, for some applications retrieving related proteins is becoming too time-consuming. Additionally, evolutionary information is less powerful for small families, e.g. for proteins from the Dark Proteome. Both these problems are addressed by the new methodology introduced here. RESULTS: We introduced a novel way to represent protein sequences as continuous vectors (embeddings) by using the language model ELMo taken from natural language processing. By modeling protein sequences, ELMo effectively captured the biophysical properties of the language of life from unlabeled big data (UniRef50). We refer to these new embeddings as SeqVec (Sequence-to-Vector) and demonstrate their effectiveness by training simple neural networks for two different tasks. At the per-residue level, secondary structure (Q3 = 79% ± 1, Q8 = 68% ± 1) and regions with intrinsic disorder (MCC = 0.59 ± 0.03) were predicted significantly better than through one-hot encoding or through Word2vec-like approaches. At the per-protein level, subcellular localization was predicted in ten classes (Q10 = 68% ± 1) and membrane-bound were distinguished from water-soluble proteins (Q2 = 87% ± 1). Although SeqVec embeddings generated the best predictions from single sequences, no solution improved over the best existing method using evolutionary information. Nevertheless, our approach improved over some popular methods using evolutionary information and for some proteins even did beat the best. Thus, they prove to condense the underlying principles of protein sequences. Overall, the important novelty is speed: where the lightning-fast HHblits needed on average about two minutes to generate the evolutionary information for a target protein, SeqVec created embeddings on average in 0.03 s. As this speed-up is independent of the size of growing sequence databases, SeqVec provides a highly scalable approach for the analysis of big data in proteomics, i.e. microbiome or metaproteome analysis. CONCLUSION: Transfer-learning succeeded to extract information from unlabeled sequence databases relevant for various protein prediction tasks. SeqVec modeled the language of life, namely the principles underlying protein sequences better than any features suggested by textbooks and prediction methods. The exception is evolutionary information, however, that information is not available on the level of a single sequence.