From Current Literature

Absorption of Light in Bromine Vapor and in Chlorine

(Ribaud, M. G. Contribution à l’étude de l’absorption de la lumière par les gaz. Annales de Physique 12, pp. 107–226, 1919.)

Bromine and chlorine vapors possess, along with an unusually complex absorption spectrum consisting of an enormous number of fine lines, also a continuous absorption spectrum. The line absorption spectrum of bromine extends approximately from 500 μμ to the infrared region; the limits of the continuous spectrum lie approximately between 340 and 520 μμ. In chlorine, the region of the line spectrum lies within the limits from 480 to 535 μμ; the continuous spectrum is found approximately between 280 and 450 μμ.

The author has studied with particular care the discontinuous spectra of both elements. The method of investigation is spectrographic. To eliminate various complications and errors connected with the law of blackening of photographic plates, Ribaud applies a new procedure. From a source giving a continuous spectrum (a Nernst filament), two spectra are obtained side by side on one and the same photographic plate. One of the spectra corresponds to the passage of light through the absorbing vapor. In the path of the beam of light forming the second spectrum there is no absorbing medium, but there is a device for attenuating the light to any desired degree. This is accomplished by a fixed Nicol and a rotating Foucault prism. Thus the first spectrum is attenuated in different ways in different parts, according to the selective absorption of the vapor under investigation; the second spectrum is attenuated uniformly throughout, according to the rotation of the Foucault prism. By attenuating the light of the second beam to a sufficient degree, one can always find a place in the two spectra placed side by side where the blackening is the same. The wavelength is then determined with the aid of a third, standard spectrum, obtained on the same plate, for example from a mercury lamp. Knowing the angle of rotation of the Foucault prism, one can thus, for that wavelength where the blackening is the same, directly determine the coefficient of absorption of light. By changing the angle of rotation of the prism, one can in this way study the entire absorption spectrum. The accuracy of measurement in this case varies in different parts of the spectrum from 1 to 5%.

To measure the absorption of bromine vapor at high temperatures, Ribaud used an all-quartz vessel with fused-in quartz plane-parallel walls (vessel diameter 3 cm, length 5 cm).

Table 1 gives the results of Ribaud’s measurements on bromine vapor for temperatures of 16°, 320°, and 620°; here the quantity \(nk\), connected with the absorption coefficient \(K\) for a given wavelength \(\lambda\), is given as follows:

\[ nk = \frac{K\lambda}{4\pi}. \]

In the experiments at 16° the pressure of the bromine vapor was 66 mm. For the high temperatures the values of \(k\) have been reduced to the same pressure.

TABLE 1.

Introducing into absorbing vapors neutral, nonabsorbing gases does not in practice exert almost any influence either on the form of the absorption curve or on its dimensions. In these experiments the pressure of carbon dioxide was brought up to 56 atm., that of oxygen to 115 atm., and that of hydrogen to 80 atm.

The absence of any substantial influence of temperature and pressure on the continuous absorption band of bromine compels one to renounce the attempt to apply the Lorentz theory of absorption to the explanation of the broad absorption bands of the visible spectrum. As is known, in this theory the cause of absorption is assumed to be, chiefly, molecular collisions.

On the other hand, the observed form of the bromine absorption curve is sharply inconsistent with the theoretical form following from all possible variants of the classical theory of absorption. If the values of the damping coefficient for bromine molecules are calculated from the measurement data for various waves, one obtains, for example, the following figures: for $\lambda = 336$ μμ the damping coefficient is $4{,}3 \cdot 10^9$, for 402 μμ $1{,}7 \cdot 10^8$, etc., whereas theoretically this quantity should be constant. Some anomaly of the bromine absorption curve in the long waves should be attributed to the still-superposed absorption of the line spectrum here—an amendment, however, that is insignificant.

For the absorption of chlorine at 16° and a pressure of 10 cm, P and B obtained the following figures (Table 2).

TABLE 2.

Applying the electronic forms of the classical theory of absorption to the absorption data, one can calculate the values \(p \cdot \dfrac{e}{m}\), where \(p\) is an integer, and \(e, m\) are the charge and mass of the electron. The ratio of charge to mass in the case of a purely electronic vibrator is \(1.77 \cdot 10^7\). From his data, Ribot finds for the continuous absorption spectrum of chlorine \(p \dfrac{e}{m} = 0.0024 \cdot 10^7\), and for bromine \(0.012 \cdot 10^7\). From this point of view only an insignificant fraction of the chlorine and bromine molecules participates simultaneously in continuous absorption.

Ribot’s measurements in the line absorption spectrum of bromine have only a preliminary character. The width of the lines themselves is negligible; under normal conditions it is of the order of thousandths of a \(\mu\mu\). However, in contrast to the continuous spectrum, the lines broaden sharply when nonabsorbing gases, \(CO_2\) and \(H_2\), are added to bromine vapor, which Ribot makes use of. Assuming that the fine absorption lines of bromine correspond to theoretical Lorentz curves, Ribot uses the following method for measuring line widths, i.e., the distance in wavelengths between two such ordinates of the absorption curve where the absorption is half that at the maximum. Two neighboring absorption lines of approximately equal intensity are chosen, lying at a distance from one another of several hundredths of a \(\mu\mu\). By gradually adding a neutral gas to the bromine, the bands can be broadened and, consequently, brought closer together. At the moment when the lines have completely merged, it may be assumed approximately that the distance between the maxima of the two lines is equal to the width of each line. This method, in any case, is not free from objections. Thus, for bromine lines lying between 5460.39 Å and 5461.16 Å, Ribot finds a line width of 0.026 Å at 1 atm pressure of \(CO_2\) and 66 mm pressure of bromine, and 0.062 Å for 1 atm pressure of hydrogen. The line width increases approximately proportionally to the pressure of the neutral gas, in accordance with Lorentz’s theory. Absolute measurements of the absorption coefficient at the maximum of the line gave the following values for the line 5460.93 Å: upon addition of air, with the addition of 1 atm, \(nk = 3.6 \cdot 10^{-6}\); upon addition of 1 atm of hydrogen, \(nk = 1.55 \cdot 10^{-6}\); the pressure of the bromine vapor in this case was 1 cm. Hence the value \(p \cdot \dfrac{e}{m} = 5 \cdot 10^7\).

The last part of Ribot’s work contains a brief account of magneto-optical measurements on the line spectrum of bromine. The principal result of this part of the work, which we shall not set forth in detail, is the indirect discovery (rotation of the plane of polarization in a magnetic field) of the presence of a very small Zeeman effect for the absorption lines of bromine. The splitting in a field of somewhat more than 20,000 gauss is less than 0.01 Å.

Ribot’s work contains numerous critical reviews of research in the field of continuous absorption of gases.

S. Vavilov.