Dear Mikhail, thank you for the questions!
I will first answer question #3:
3. Presently, fullerene oxidation under light exposure remains a strong hypothesis for description of cluster growth in low-polar fullerene solutions ( see
https://www.sciencedirect.com/science/article/pii/S0021979715000326 and Refs. therein). In our experiments we also clearly see, that if the solution is kept sealed and in the dark, we exclude cluster growth for long periods of time.
1. The kinetics of clusters growth depend on solvent properties, e.g. viscosity, dielectric constant, etc. Also (see answer #3), the oxide formation should depend on some of the solvent properties. One of our present goals is to construct a theoretical model of kinetics of cluster growth which will account for differences of these solvents.
2. There's not much points at this part of the graph, and the slope is not that clear. ~0.7 +- 0.2 may be the best estimate. This is an initial regime, which is observed in several theories.