Applied Sciences

In geological exploration and oil and gas drilling engineering, rock brittleness is a key indicator for evaluating formation drillability and reservoir fracability. The tight sandstone reservoir of the Chang 3 Member of the Yanchang Formation in the Weibei Oilfield is shallowly buried and characterized by tight lithology and strong heterogeneity. Therefore, assessments of its brittleness and fracability should comprehensively account for confining-pressure effects and in situ stress conditions. Downhole cores from the Chang 3 Member in six wells were subjected to multidirectional uniaxial and triaxial compression tests at confining pressures of 0, 5, 10, and 15 MPa. Acoustic emission Kaiser-effect tests were also conducted to determine the in situ stresses. Rock mechanical parameters, failure modes, and energy evolution were analyzed. A two-component energy-based brittleness index (BI) was established from pre-peak elastic energy storage and post-peak self-sustained fracturing capacity. Brittleness, peak strength, fracture box-counting dimension, and in situ stress differential were then coupled to construct a comprehensive fracability index (FI). Young’s modulus ranged from 6.41 to 13.87 GPa, and Poisson’s ratio was mainly 0.18–0.33. Peak strength and Young’s modulus generally increased with confining pressure. Failure changed from axial splitting under uniaxial loading to inclined shear failure under triaxial loading. BI ranged from 0.42 to 0.93, indicating strong brittleness to a brittle-plastic transition. BI decreased with confining pressure, particularly between 10 and 15 MPa, and the post-peak self-sustained fracturing component BI₂ was more pressure sensitive. The Kaiser-effect tests showed that the three principal stresses increased with burial depth, while the stress regime changed from reverse-faulting at shallow depths to strike-slip at greater depths. FI generally decreased with confining pressure by 9.7–53.6%. At low confining pressure, Well 37 had the highest FI of 0.842. At high confining pressure, Well 45 had the highest FI of 0.618, whereas Well 49 had the lowest value of 0.300. Confining pressure reduced fracability by suppressing post-peak elastic-energy release, increasing peak strength, and enhancing the in situ stress differential. Greater fracture complexity partly compensated for unfavorable stress conditions. The results provide a scientific basis for geological exploration and evaluation, drill-bit selection, and drilling and completion design for the tight sandstone reservoirs of the Chang 3 Member of the Yanchang Formation in the Weibei Oilfield.

In this paper, we report the progress of our research on adenosine A₁/A₃ receptor antagonists, aimed at identifying new small molecules for therapeutic applications. Starting from two recently published reference compounds, we have synthesized new products with 1,8-naphthyridone scaffold performing targeted chemical modifications at positions 1 or 3. These elaborations led to compounds that maintained affinity for the inhibitory adenosine receptors A₁ and A₃ but lower selectivity compared to the reference compounds. The best balance between affinity and selectivity was found for 4f (pKi A₁R < 4; pKi A₃R = 6.25 ± 0.10) and 4i (pKi A₁R = 4.91 ± 0.02; pKi A₃R = 6.51 ± 0.15), whose antagonist profile and selectivity against A_2AR were determined through NanoBRET binding assay. Overall, these results can be useful in guiding the optimization process of small-molecule heterocyclic A₁R/A₃R receptor antagonists, facilitating their future development into effective therapeutic agents.

The rapid expansion of digital transformation and complex network infrastructures has increased the sophistication and scale of cyber threats, challenging the effectiveness of traditional defense mechanisms. Traditional security operations centers (SOCs) often struggle to meet modern cybersecurity requirements because of their reactive structures, siloed toolkits, and over-reliance on manual processes. Many existing artificial intelligence (AI)-driven SOC (AI-SOC) studies emphasize simulation-based evaluations and seldom analyze the impact of human factors on the operational performance of AI systems. To address these shortcomings, this study proposes a hybrid AI-SOC architecture integrating user and entity behavior analytics, supervised machine learning (random forest), isolation forest, and large language model-based decision-support mechanisms within an Information Security Management Systems (ISMS)-aligned human-in-the-loop framework. The proposed architecture was evaluated using operational security data from the 2023 baseline and 2025 AI-SOC evaluation periods within a public university network. The comparison between evaluation periods revealed an increase in the threat detection rate from 61% to 91% (+30 percentage points), a reduction in the false-positive rate from 14% to 5% (−9 percentage points), a 69.05% reduction in mean time to detect (MTTD), and a 70.83% reduction in mean time to respond (MTTR). These findings highlight the importance of integrating algorithmic automation and adaptive AI capabilities with human expertise to improve security operations in complex institutional environments.