In a paper with David Walton C50 was predicted to be decavalent and that addition of 10 monovalent atoms such as H or Cl atoms should result in a stable compound.

Capture c50

Abstract

Recent developments in experiment and theory have enhanced our understanding of fullerene stability and reactivity to the point where predictions may be made concerning the probable existence of specific fullerenes and their derivatives. One approach, considered here for small fullerenes (C(n), n < 60) – which cannot have isomers in which all pentagons are isolated – invokes the concepts of localised surface stability domains and angle strain release.

http://65.54.113.26/Publication/54492684/stable-derivatives-of-small-fullerenes

This was shown to be correct in the paper below in which, oddly at the time it seemed to me, our paper was not referenced.

C50Cl10_(2004 Science)

As I was a bit miffed by not referencing our nice prediction I wrote to the main author who wrote back that they knew about our work but there was not enough room to reference our paper! [A colleague has since pointed out that an advisor is referenced on their manuscript who may have made the “helpful suggestion” that the paper should be shortened by deleting our reference.]

Valence of D5h C50 Fullerene
Abstract

The energetic and electronic properties of D5h C50 before and after passivation by H or Cl are investigated using first-principle computational method of density functuional theory with generalized gradient approximation and local density approximation functionals. The results show that H or Cl addition can lead to energetic stabilization. Additions also increase the highest occupied molecular orbit−lowest unoccupied molecular orbital (HOMO−LUMO) gaps of C50 fullerides and make them chemically more stable. In the series of C50H2m (m = 0∼7), the Saturn-shaped D5h C50H10 has the largest HOMO−LUMO gap, which suggests that such a structure of C50H10 is a “magic-number” stable one of C50 adducts, and ten is a pseudovalence or effective valence of C50 fullerene pseudoatom. This point also is supported by the energetic properties of C50H2m series such as binding energies, etc. A minimal energy reaction pathway is constructed to get C50H10 and C50H14. Some useful experience for determining the favorable addition sites was summarized. A simple steric method is developed to predict the effective valences of classical fullerenes.

http://onlinelibrary.wiley.com/doi/10.1002/chem.201102330/abstract

http://onlinelibrary.wiley.com/doi/10.1002/chem.201102330/abstractC50H10 chinese paper jp064500b