“…In Figure f, the Fe 2p 3/2 spectrum is shown with a peak deconvolution based on the Gupta and Sen (GS) method, which considers various possible factors related to the peak broadening of Fe 2p (e.g., electrostatic interactions, L-S coupling between unpaired 3d electrons in a photoionized Fe cation and its 2p hole, and crystal field interactions). , Using the GS method, six independent componentsprepeak, surface peak, and multiple peaks 1 to 4 ranging from 708.5 to 713.8 eVwere required to fit an Fe 2p 3/2 spectrum of a single phase of Fe-oxide, and each phase of Fe-oxide had unique ratios among peaks 1–4 . In our Fe-oxide/Al 2 O 3 , ratios among peaks 1–4 were 22, 35.89, 19.14, and 22.97%, respectively, which did not perfectly match any of single phase of Fe-oxide, implying that our Fe-oxide nanoparticles were instead a mixture of various Fe-oxide phases (e.g., FeO, Fe 2 O 3 , Fe 3 O 4 , FeOOH, ...). ,, Raman spectroscopy analysis of the Fe-oxide/Al 2 O 3 could also indicate the coexistence of various phases of Fe-oxide (Figure S4). Although the additional small component at 707 eV can be attributed to either Fe carbide or metallic Fe, its relative composition was less than 10%.…”