Atmospheric Ozone

For many years traces of oxidising substances have been known to exist in the atmosphere, and have been variously characterised as ozone, hydrogen peroxide, and organic peioxides. Unfortunately early investigators failed to appreciate the fact that it is extremely difficult to distinguish between and severally estimate such minute traces of these substances as occur in the air, although but little difficulty occurs when they are present in larger amounts. It is also now known that oxides of nitrogen would vitiate the earlier tests, and as traces of these gases are likewise frequently present in the atmosphere, no little uncertainty has arisen as to the correct interpretation to be placed upon the majority of the results obtained by early workers.

This confusion is further enhanced by the fact, which has only recently been ascertained with certainty, that hydrogen peroxide sometimes decomposes, yielding ozone and water (vide infra).

In the following table are given a few of the more important results of the so-called determinations of ozone in the atmosphere. The data prior to 1917 should be interpreted as representing the amounts of oxidising substances expressed as milligrams of ozone per cubic metre of air.

Hayhurst and Pring were careful to introduce a correction for oxides of nitrogen, but did not distinguish between ozone and hydrogen peroxide. Usher and Rao appear to have eliminated all oxidisers except ozone, and it is remarkable that they should have discovered no ozone whatever in their series of experiments. In the neighbourhood of London, Reynolds finds one volume of ozone in about 20 million of air. This amount is greatly increased after a thunderstorm, from a study of the absorption of ultra-violet light by ozone, combined with measurements of the amount of the sun's light transmitted by the atmosphere, the conclusion has been reached that, if the ozone were equally distributed throughout the air, its amount would equal 0.6 c.c. per cubic metre, or 6 parts of ozone per 10 million of air.

Hydrogen peroxide is produced in nature in a variety of ways. According to Dixon it occurs as a product of evaporation of water, and it has recently been shown that moist oxygen, when exposed to ultra-violet light, yields distinct traces of hydrogen peroxide in the course of seven or eight days, according to the temperature.

In 1909 Kernbaum produced hydrogen peroxide by exposing water to the action of penetrating rays from radium salts, and two years later Kailan confirmed this result. How much of the atmospheric peroxide is due to the above causes it is difficult to say. Undoubtedly most of the peroxide, and certainly the bulk of the organic peroxides of the air, originate from the direct action of air, moisture, and sunlight upon the essential oils and other organic emanations of plants.

Ozone may be produced in nature in a variety of ways. When water evaporates into the air, particularly when thrown up in the form of spray, traces of ozone are produced, and this accounts for its presence in the fresh sea breeze and in the neighbourhood of waterfalls. The refreshing odour after a shower of rain or the passing of a water-cart over the road is probably due to ozone. The gas is also produced by silent electric discharges from thunder-clouds and accompanies the flash discharge of lightning. In 1886 Wurster called attention to the fact that ozone may result from the action of sunlight upon the clouds, and since then the results of numerous researches have pointed to the fact that the bulk of the atmospheric ozone is yielded by the action of ultra-violet rays from the sun's light upon the oxygen in the upper reaches of the air. This would account for the observation of Thierry and of W. Hayhurst and Pring that the amount of oxidising material increases with the altitude. In clear weather ozone is probably produced also by the direct action of the sun's rays upon the lower layers of the atmosphere. These observations probably explain why R. Lespieau found that in his experiments the proportion of atmospheric ozone was independent of the altitude.

Ozone has recently been shown to result from the action of rays from radio-active substances upon oxygen, but to what extent atmospheric ozone is attributable to this cause is uncertain. Probably a small quantity of ozone is produced by the slow oxidation of the essential oils and other organic exhalations of plants, and Duphil has recently observed an excess of ozone in the air of the maritime pine forests in the neighbourhood of Bordeaux.

From the point of view of ventilation, ozone and peroxides are of interest inasmuch as they impart a crispness or freshness to the air, and the fact that they are readily decomposed by heat is probably one of the causes of the " flatness " of heated air. The well-known Sanitas preparations are essentially solutions of hydrogen peroxide and of different organic peroxides. There can be no doubt that in nature the presence of hydrogen peroxide is an important factor in removing foetid and putrid matter from the atmosphere. Bosisto has calculated that 96,877,440,000 gallons of eucalyptus oil "are held continually at one and the same moment in the leaves of trees massed together and occupying a belt of country over which the hot winds blow " in New South Wales and South Australia alone. Kingzett concludes that this amount of eucalyptus oil "can and must produce in the atmosphere surrounding the forests no less than 92,785,023 tons of peroxide of hydrogen, and about 507,587,945 tons of the soluble camphor, not to mention the other products of oxidation."

Schonbein's method for Detection of Atmospheric Ozone hinges on the fact that since ozone liberates iodine from potassium iodide, starch- iodide papers are readily turned blue in its presence. But inasmuch as peroxides of hydrogen and nitrogen have a like effect, the test is valueless unless the absence of these other substances can be proved. Houzeau therefore recommended a litmus-iodide paper. Nitrogen peroxide liberates iodine but, unlike ozone, will not simultaneously liberate free alkali, so that the litmus either remains unchanged or is slightly reddened. Ozone, on the other hand, not only liberates iodine but alkali which turns the litmus blue. The main disadvantage of this method lies in the tendency of the iodine to mask the colour of the litmus, and thus render any change in the latter difficult to detect. It is true that hydrogen peroxide likewise liberates iodine, but by first passing the air over chromic acid, hydrogen peroxide may be effectively removed whilst the ozone is not affected. Filter paper soaked in an alcoholic solution of tetramethyl base has been recommended, since the paper becomes violet with ozone, yellow with nitrous fumes, but remains unaltered in contact with hydrogen peroxide. The reaction, however, does not appear to be sufficiently sensitive for the tests under discussion.

The majority of investigators in the past have relied upon the liberation of iodine from potassium iodide solution as a convenient method of Estimation of Atmospheric Ozone. Thus Hatcher and Arny aspirated air through potassium iodide solution, whereby first iodine and subsequently potassium iodate are formed, as indicated in the following equations:

2KI + H₂O + O₃ = 2KOH + O₂ + I₂;
2KOH + I₂ = KIO + KI + H₂O;
3KIO = KIO₃ + 2KI.

On acidifying, free iodine is again liberated according to the equation

KIO₃ + 5KI + 3H₂SO₄ = 3K₂SO₄ + 3I₂ + 3H₂O,

and is estimated by titration with thiosulphate. This method does not distinguish between ozone and hydrogen peroxide, neither does it guard against the disturbing influence of nitrogen peroxide. This difficulty was surmounted by Hayhurst and Pring, who estimated the free alkali in addition to the iodine and were thus able to introduce a correction for the peroxide of nitrogen, although they were unable to distinguish between hydrogen peroxide and ozone.

If, however, the air is first freed from peroxides of nitrogen and hydrogen, the liberated iodine owes its presence entirely to ozone. Keiser and M'Master, therefore, recommend the passage of air through a solution of potassium permanganate prior to testing for ozone, as this gas is not affected by the permanganate whereas the two peroxides are destroyed.

Potassium arsenite is oxidised by ozone to arsenate, and may be substituted for potassium iodide in the previous method; although here again it is necessary to first remove the peroxides of hydrogen and nitrogen from the air prior to testing.

A method that appears capable of greater accuracy and less open to criticism than any of the foregoing is that devised by Usher and Rao. It hinges on the fact that ozone oxidises aqueous solutions of alkali nitrites to nitrates, the reaction proceeding quantitatively according to the equation

NaNO₂ + O₃ = NaNO₃ + O₂.

Two samples of air are taken and collected in large stoppered bottles of some 7-litres capacity. One sample is admitted through two tubes containing respectively chromic anhydride and powdered manganese dioxide, the second sample through a tube containing chromic anhydride only. They are then shaken with a dilute standard solution of sodium nitrite rendered slightly alkaline, and the nitrite content subsequently determined colorimetrically.

The first sample of air contains only nitrogen peroxide, the hydrogen peroxide and ozone having been destroyed by the chromic acid and manganese dioxide respectively. The increase in the amount of nitrite in the bottle thus gives the measure of the nitrogen peroxide absorbed.

The second sample contains ozone and nitrogen peroxide and the difference between the nitrite contents of the two bottles is equivalent to the ozone present. The presence in the air of ammonia, sulphur dioxide, and hydrogen sulphide does not interfere with the Estimation of Atmospheric Ozone and nitrogen peroxide by this method as all three gases are completely absorbed during passage through the chromic anhydride tube.