A new method for the determination of nitrogen peroxide

The ordinary spectrograpbic method of estimating a substance from its absorption spectrum when photographed under standard conditions has the merit of providing permanent records, but suffers from necessitating the use of an expensive instrument as well as being laborious and as a rule very slow. A method will now be described by means of which nitrogen peroxide, a substance often determined spectrographically, may be estimated in concentrations of 1 : 100,000 and upwards with high accuracy and in a few seconds when once a simple calibration has been made. The method utilises the fact that the spectral region in the visible whereover nitrogen peroxide absorbs most strongly, is close to that at which a potassium photoelectric cell is most sensitive and where it can be used to record, with high accuracy, the light transmitted by the gas under consideration (fig. 1). The optical basis of the method may first briefly be discussed. If a beam of monochromatic light be passed through a column of absorbent medium, the Beer-Lambert law gives log10 (I0 /I) = ε .c .d , (1) where I0 = intensity of light transmitted at zero absorption, I = intensity of light transmitted at measured absorption, ε = the molecular extinction coefficient, which is a constant for the absorbent medium at any given wave-length of the light concerned,c = concentration of absorbent medium,d = length of absorbent column in the direction of the path of the light. Since for a vacuum photocell with an applied voltage in excess of the saturation value (i.e .,ca . 40 volts), the photoelectric current varies directly, over a wide range, as the intensity of the light incident upon the cell, we may re-write equation (1) as logi = logi 0 — ε .c .d , (2) wherei 0 = photoelectric current corresponding to I0 , andi = photoelectric current corresponding to I.