Sur une cause d'erreur dans le dosage par voie humide de l'azote dans les aciers speciaux

The estimation of nitrogen in normal or special steels by the Kjeldahl method can give results that are too high if the acid used to dissolve the steel (even “pro analysi” acid) contains impurities such as nitrates, and if the metal itself contains even negligible amounts of molybdenum (e g a few ten-thousandths). In this case the blank test can nevertheless give a very weak or negative result.In order to ascertain from this point of view the quality of the acid used, it is necessary to compare the nitrogen contents obtained on a sample with a very low nitrogen content, e g electrolytic iron, in the absence or in the presence, respectively, of a minute addition (about 0.1%) of molybdenum. The difference between these determinations should be negligible. If it is not, then the acid tested should be rejected, or distilled before use.