Why scipy curve_fit in a log log scale gives a wrong fit?

Question

I am trying to fit a power law to some data following a power law with noise, displayed in a log-log scale:

The fit with scipy curve_fit is the o