Programmable Pore Environments in Multivariate ZIF Membranes for Ultra-Selective Helium Recovery from Natural Gas

Helium is an essential yet finite resource with critical applications in medical imaging and semiconductor manufacturing, whose production currently relies almost exclusively on energy-intensive cryogenic separation of trace helium from natural gas. Membrane-based separations offer an attractive alternative, but existing materials lack the selectivity required for industrial deployment. Here, we introduce a strategy for pore microenvironment programming in multivariate zeolitic imidazolate framework (MTV-ZIF) membranes, enabling ultraselective helium recovery under realistic feed gas conditions. By precisely combining Zn²⁺, 2-methylimidazole, and halogen-substituted benzimidazole linkers, we create synergistic combinations of steric constraints and enhanced CH₄-framework interactions, which collectively suppress CH₄ transport while preserving rapid He permeation. Under simulated industrial feed conditions (0.6% He/99.4% CH₄ by volume), the best-performing membrane delivered a record He/CH₄ selectivity of 3174, with stable operation over 960 h. Process simulations further show that a two-stage membrane cascade can deliver >99.95% He purity with an 83% reduction in energy demand compared to cryogenic distillation. These results highlight multivariate pore programming in MOFs as a powerful platform for efficient, low-energy He recovery from natural gas.