**4. Discussion**

Figure 5 shows the nominal electron concentrations for all investigated samples versus the applied Si source temperature together with a fit (black curve) which shows this dependence in a broader range of concentrations and temperatures. Red open circles denote determined (by second more accurate approach) electron concentrations for samples A–C together with respective fit (red curve). The comparison of the nominal dependency and the corrected experimental results shows that the respective adjustments become important for larger concentrations (higher source temperatures).

**Figure 5.** Nominal carrier concentrations (black squares) as a function of applied Si source temperature together with fit (black curve). Open red circles denotes corrected (in the way of second approach) values of the carrier concentration in a function of the applied Si source temperature together with respective fit (red curve).

Therefore, in a range of concentrations presented in this paper, a slight temperature increase seems to be appropriate. For instance, to obtain a nominal concentration for sample A, Si source temperature should be increased by ~15 K from 1300 ◦C to 1315 ◦C to o ffset the observed di fferences. On the other hand, in the range of typical doping (~1016–1017 cm<sup>−</sup>3) of the lasing active areas of quantum cascade lasers, little temperature compensation would be required.
