A recent study maps the limitations of today’s lithium-ion batteries and outlines several promising alternatives, including lithium-sulfur, lithium-metal, lithium-air, zinc-air, sodium-ion, and redox flow batteries. The authors argue that breakthroughs such as solid-state electrolytes, self-healing components, and flexible energy-storage architectures will be essential to meet future demands for greater safety, better performance, and stronger sustainability goals. They also emphasize the need for a chemistry-neutral battery roadmap beyond 2030, one in which artificial intelligence and advanced materials-discovery tools accelerate the shift toward safer, more reliable, and climate-neutral energy-storage technologies.

Ammonia gas, a popular industrial chemical, is dangerous to human health. A new study by Hanbat National University researchers presents a wearable ammonia gas sensor that detects harmful ammonia levels visually and electronically. The sensor is flexible, stretchable, and works reliably when attached to human skin and exposed to high humidity. By combining two sensing methods in one device, the platform remains accurate even if one sensing mode fails, making it suitable for real-world use.

Next-generation sodium- and potassium-ion batteries offer resource-unconstrained, cost-effective, and sustainable energy storage systems. In a recent review, researchers from Japan redefine the electrode-electrolyte interphase (SEI and CEI) to improve battery stability and performance. By systematically analyzing these overlooked layers, the team demonstrates how controlling interfacial reactions can influence electrochemical performance and safety. Their findings could accelerate the development of the next-generation battery systems for grid storage, electric vehicles, and other energy applications.

Finely dispersed particulate matter with a diameter of ≤2.5 μm (PM_2.5) poses a significant health- and climate-risk, yet tracking its chemical composition remains a challenge. Now, researchers have developed a deep-learning model that accurately estimates hourly concentrations of five key PM_2.5 chemical components, without chemical analysis. Using air-quality and meteorological data, the model achieved high accuracy outperforming existing methods, and may strengthen air-pollution monitoring, fill data gaps, and support targeted emission control strategies worldwide.