The vapor pressure is an increasing function of temperature. The vapor pressure changes continuously with temperature whatever the element is, solid or liquid. The boiling point is the value of the temperature at which the vapor pressure reaches 1 bar. It is one point of the vapor pressure curve. For arsenic the "boiling point" is lower than the melting point.

You can find relatively easily Honigs's curves on internet. I attached some of them.

I do not understand what you call by "sublimation transition". The sublimation pressure increases with vapor pressure thus with temperature. Thus the evaporation (condensation) speed increases with temperature.

A long time ago I distilled cadmium from a much oxidized cadmium powder in a quartz glass tube under vacuum. I obtained very pure cadmium.

The problem is to avoid that the powder goes into the vacuum pomp. I used a sintered metal filter to avoid this.

Basic thermodynamics calculated by HSC give boundary conditition:

at 1000 C deltaG=43.154 kJ

1500 C dG=0.015 kJ

1600 C dg=-8.351 kJ etc.

Evaporation (equilibrium) thermodynamically start at 1500 C. It is necessary to take into consider, that this say nothing about kinetics of evaporation and and activity of Ca is equal to 1 in this calculation. It is possible to calculate more realistic values in "real" material context.

Thank you very much for your valued answers