The aim of this study is to investigate the effects of chemically modified hexagonal boron nitride (hBN) particles on the tribological properties of polyethylene glycol (PEG) based lubricant and sedimentation in PEG. Thermally purified and chemically modified hBN particles were compared with commercially available hBN particles of various sizes. Tribological tests were performed on a ball-on-disk optical tribometer under rolling–sliding conditions using a 100Cr6 steel ball and BK7 glass disk. The effects of hBN particle size and surface modification on the frictional behavior across boundary, mixed, and elastohydrodynamic (EHL) lubrication regimes were examined. The most pronounced reduction in the friction was achieved with the largest hBN particles (≈6.6 μm), which unfortunately tend to agglomerate in liquid PEG though. It was observed that agglomeration was strongly suppressed by chemical modification of hBN´s surface. The work also includes a post-experimental analysis which demonstrated good durability of hBN particles in PEG lubricant after tests.
Substantia nigra hyperechogenicity (SN+) detected by transcranial sonography (TCS) has been proposed as a risk marker for synucleinopathies, particularly Parkinson’s disease. Its role in prodromal dementia with Lewy bodies (DLB) remains insufficiently explored. To examine clinical, cognitive, genetic, EEG, and plasma biomarker differences between SN+ and normal echogenicity (SN−) individuals in a well-characterized at-risk-of-DLB cohort, with additional sex-stratified analyses. A total of 121 participants aged 55–70 years from an at-risk-of-DLB cohort underwent TCS, neurological and neuropsychological assessment, nighttime wrist actigraphy, high-density EEG, genetic testing for APOE allelic variants, and plasma biomarker analysis (NfL, GFAP, pTau181). SN status was defined using laboratory-specific cut-offs. Group differences were assessed using non-parametric statistics with relevant corrections for multiple comparisons and adjustment for age. SN+ status as compared to SN− status was associated with poorer verbal memory performance and altered visuo-constructive performance (driven mostly by males), and significant differences in distinct actigraphy-derived sleep parameters. SN+ participants showed higher plasma NfL levels (particularly in males) and a higher prevalence of DLB-specific EEG patterns (driven by females). No significant associations were observed for global cognition, motor parkinsonism, cognitive fluctuations, hallucinations, REM sleep behavior disorder, depressive symptoms, olfactory function, plasma biomarkers of degeneration, neuroinflammation, β-amyloid deposition, or APOE status. In individuals at risk of DLB, SN hyperechogenicity is linked to early markers of neurodegeneration and cholinergic dysfunction, including elevated NfL and DLB-specific EEG abnormalities, as well as selective posterior cortical cognitive impairment and sleep disturbances. The observed sex-specific effects require further investigation.
Audio comprehension—including speech, non-speech sounds, and music—is essential for achieving human-level intelligence. Consequently, AI agents must demonstrate holistic audio understanding to qualify as generally intelligent. However, evaluating auditory intelligence comprehensively remains challenging. To address this gap, we introduce MMAU-Pro, the most comprehensive and rigorously curated benchmark for assessing audio intelligence in AI systems. MMAU-Pro contains 5,305 instances, where each instance has one or more audios paired with human expert-generated question-answer pairs, spanning speech, sound, music, and their combinations. Unlike existing benchmarks, MMAU-Pro evaluates auditory intelligence across 49 unique skills and multiple complex dimensions, including long-form audio comprehension, spatial audio reasoning, multi-audio understanding, among others. All questions are meticulously designed to require deliberate multi-hop reasoning, including both multiple-choice and open-ended response formats. Importantly, audio data is sourced directly ``from the wild" rather than from existing datasets with known distributions. We evaluate 22 leading open-source and proprietary multimodal AI models, revealing significant limitations: even state-of-the-art models such as Gemini 2.5 Flash and Audio Flamingo 3 achieve only 57.33% and 45.9% accuracy, respectively, approaching random performance in multiple categories. Our extensive analysis highlights specific shortcomings and provides novel insights, offering actionable perspectives for the community to enhance future AI systems' progression toward audio general intelligence.
Precise and up-to-date cross-sections with uncertainty propagation data of materials used in control rods are essential for the smooth functioning of nuclear reactors, since fast neutrons interact with control rods that regulate chain reactions by absorbing excess neutrons. The present study aims to measure the neutron-induced reaction cross-section values for the isotopes of silver (Ag) and indium (In). Deuterium–tritium (D–T) fusion neutrons with an energy of 14.96 ± 0.22 MeV were produced and used to irradiate natural samples of Ag and In targets, inducing measurable activation for the reactions 109Ag(n,2n)108Agg, 109Ag(n,p)109Pdm, 107Ag(n,2n)106Agm, 115In(n,p)115Cdg, and 115In(n,α)112Ag at the Neutron and Ion Irradiation Facility, Institute for Plasma Research (NIIF-IPR), Gujarat, India. The radioactive samples were subsequently taken for offline γ-ray counting in a high purity germanium detector (HPGe), with a fine resolution of 2.1 keV at 1.33 MeV γ-ray energy of 60Co connected with GENIE software. Two well-known standard reactions of Aluminium, 27Al(n,p)27 Mg and 27Al(n,α)24Na were employed for neutron flux measurements as per the irradiation period. Appropriate correction factors were applied in cross-section assessment, and uncertainties from all input parameters were rigorously propagated to the final values through covariance analysis. The experimental results were also validated by predictions from the nuclear code TALYS-2.0. Furthermore, prior reported studies from EXFOR and evaluated libraries from ENDF were compared with the presently measured results.
LuFeO3 (LFO) is a perovskite oxide with promise for optical and electroceramic applications. In the present study, LFO and Co-substituted compositions (LFO, LuFe0.95Co0.05O3, and LuFe0.90Co0.10O3) were synthesized by a conventional solid-state route and characterized by SEM, Raman spectroscopy, diffuse reflectance, and broadband dielectric/impedance measurements. Co substitution alters the powder microstructure, yielding more irregular agglomerates composed of finer sub-units than undoped LFO. Dielectric spectra showed that the loss tangent (tanδ), the dissipation factor, of investigated samples was below 1 over the studied temperature–frequency window. It was seen that Co substitution decreased dielectric loss in the mid-to-high frequency region but raised low-frequency loss at advanced temperatures. The real part of impedance Z^' declined with both temperature and frequency, and Nyquist plots displayed depressed arcs, indicating non-Debye behavior with distributed grain and grain-boundary contributions. Arrhenius analysis of relaxation maxima yielded activation energies of ∼ 0.32–0.76 eV, consistent with oxygen-vacancy energies. Raman spectra revealed Co-induced lattice perturbations, including mode broadening/attenuation and red-shifts in the 200–600 cm− 1 range, together with a strengthened stretching feature near 600–650 cm− 1. Kubelka–Munk analysis was utilized to determine the band gaps of the studied samples, 2.19 eV (LFO), 2.29 eV (5
