In conclusion, Lu, Li, Zeng, and co-workers innovatively presented a universal strategy that computationally predicted the C₃H₈ selectivity of Zn-siloxane clusters and then incorporating them as building blocks into various MOF structures. This unique design effectively endows Zn₈(CH₃COO)₈-siloxane cluster-based MOFs with excellent separation potential for C₃H₈/C₃H₆ mixtures. Among these MOFs, JNU-66-II exhibits superior performance in adsorption kinetics, desorption kinetics, Δq, Q_st, and cycling efficiency compared with previously reported C₃H₈‑selective MOF materials. Moreover, JNU-66-II achieves the direct cylinder-based collection of polymer-grade C₃H₆ with a purity of ≥ 99.5% for the first time, establishing a new platform for highly efficient C₃H₈/C₃H₆ separation in industry. This work not only provides a new strategy for guiding the design of metal clusters through computation to enhance the selectivity of alkanes rather than that of alkenes, but also significantly promotes the application transformation of MOF-based adsorptive separation technology in actual industrial settings.