The SEM results of fresh and recovered catalysts are shown in Fig. 9. The as-received Ni2O3 grains have irregular shapes and different sizes. After recovered from the first catalytic run (B(1), B(2)),lots of fine particles appear and the shape becomes more regular. No obvious coke deposition is observed. While for the catalyst recovered from the fifth time of catalytic run (C(1), C(2)), a lot of carbon nanofibers with filamentous lines and helical structures are observed and they deposited on the surface of the Ni2O3 particles,agreeing with the results analyzed by XRD characterization. As discussed above, the more amount of cupric bromide and acetylene in the feedstock, the less the selectivity to acrylic acid would be obtained, which is proved again by the SEM results (Fig. 9(D andE)). It can be inferred that the activity would decline if the catalyst grains are completely covered by these carbon nanofibers.