The emission could arise from the electric dipole of very rapidly spinning (10–60 GHz) extremely small (nanometer) dust grains as suggested by Bruce T. Draine and Alex Lazarian in 1998, most likely polycyclic aromatic hydrocarbons.
Kogut et al. had correlated COBE Differential Microwave Radiometer observations at centimeter wavelengths with DIRBE dust emission at 140 μm, while Leitch et al. had correlated Owens Valley Radio Observatory ring observations at 14.5 and 32 GHz with IRAS 100 μm.
The suggestion at the time was the correlation was due to free-free or Bremsstrahlung emission from ionized gas caused by young hot stars which are formed in these dusty regions.
Interactions with the ambient gas and radiation field cause the grains to spin.
Denoting by ω their (vector) angular velocity, and assuming the grains to be spherical for simplicity (so they are not wobbling), Larmor's formula gives the power radiated by such a rotating dipole:
The distribution function f(ω) depends on the ambient gas temperature, density and ionization fraction and on the local radiation field.