{"data":{"id":327,"lang":"en","title":"AI Scans 67,000 Magnetic Compounds, Finds 25 Rare-Earth-Free Alternatives That Could Transform EVs","summary":"UNH researchers used AI to scan 67,573 magnetic compounds, discovering 25 rare-earth-free materials that could cut EV costs and reduce China dependency.","category":"science","category_label":"Science","published_at":"2026-02-22T01:02:46Z","updated_at":"2026-02-22T01:02:46Z","url":"https://techtools.cz/tools/news/article/327-ai-scans-67-000-magnetic-compounds-finds-25-rare-earth-free-alternatives-that","api_url":"https://techtools.cz/news-api/v1/articles/327?lang=en","markdown_url":"https://techtools.cz/news-api/v1/articles/327.md?lang=en","image_url":"https://services.meteored.com/img/article/new-ai-system-accelerates-search-for-sustainable-magnets-powering-clean-energy-1771538867834_1280.jpeg","word_count":150,"reading_time_minutes":1,"sources":[{"url":"https://www.sciencedaily.com/releases/2026/02/260218031611.htm","domain":"sciencedaily.com"},{"url":"https://www.theweather.com/news/science/new-ai-system-accelerates-search-for-sustainable-magnets-powering-clean-energy.html","domain":"theweather.com"}],"translation":{"lang":"cs","id":312,"title":"Průlom: umělá inteligence odhalila 25 nových magnetů bez vzácných zemin pro budoucnost elektromobility","url":"https://techtools.cz/tools/news/clanek/312-prulom-umela-inteligence-odhalila-25-novych-magnetu-bez-vzacnych-zemin-pro","api_url":"https://techtools.cz/news-api/v1/articles/312?lang=cs"},"content":"University of New Hampshire researchers have built an AI-powered database of 67,573 magnetic compounds by training machine learning models to read and extract data from thousands of scientific papers. The system identified 25 previously unrecognized materials that stay magnetic at high temperatures — a critical requirement for electric vehicle motors and wind turbines that currently rely on expensive rare earth elements controlled largely by China.\n\nLead author Suman Itani and physics professor Jiadong Zang developed the AI to predict whether a material is magnetic and calculate the exact temperature at which it loses magnetism. The 25 new candidates could dramatically reduce EV manufacturing costs and supply chain vulnerability, since rare earth prices have surged over 300% in recent years due to export restrictions.\n\nThe study, published in Nature Communications, makes the full database publicly searchable — giving researchers worldwide a tool to accelerate discovery of sustainable magnets without starting from scratch."}}