Pass emcut instead of emcut + particle_def.mass to PROPOSAL as an emcut

This PR combines two fixes concerning the usage of PROPOSAL energy cuts

1. Numerical problems due to v_cut in ionization

As @riehn and @Nikos have reported in the Slack, warnings such as [CrossSection.DNDX] [warning] Negative dNdx value for E = 0.9271 MeV, vbar = 1.0000 detected in interaction type Ioniz. Setting dNdx to zero. appeared as a result of !539 (merged).

Looking into this issue, the problem was actually numerical issues for the ionization cross section when introducing the vcut. After discussing this with @asandrock, the quickest fix for this would be to just don't just the v_cut for ionization losses.

This is introduced with this PR. Unfortunately, this means that we explicitly have to define the parametrizations within CORSIKA. However, this approach is going to be changed anyway if we combine PROPOSAL with our steering plans.

2. Combination of emcut and EnergyCutSettings

At the moment, we pass emcut + particle_def.mass to the EnergyCutSettings of PROPOSAL. However, it doesn't make sense to add the mass of the initial particle here. The parameter we pass to PROPOSAL defines the size of energy losses that we want to treat stochatically. For an emcut of, let's say 0.5 MeV, this should actually be 0.5 MeV (because we are interested in all losses above 0.5MeV). Currently, we would pass (for example for electrons) 0.5 MeV + ME, which is about 1.1MeV. This means that, for example a bremsstrahlungs loss of 1 MeV, would not be sampled by PROPOSAL, although we set the emcut to 0.5MeV. This might very well be the solution to the missing photons I have presented at the workshop.