Dissolved Gases in Sea Water

All of the atmospheric gases are found in solution in sea water. In addition to nitrogen and oxygen, the most abundant gases in the air, carbon dioxide is present in large quantities in sea water, chiefly combined as carbonates and bicarbonates. Of the rarer gases, ammonia, argon, helium, and neon have been reported in sea water, and hydrogen is undoubtedly present in minute quantities. In the absence of dissolved oxygen, hydrogen sulphide may be present, and it is possible that in stagnating water other products of putrefactive decomposition, such as methane, may occur.

Because of its importance in biological processes the dissolved oxygen distribution in the oceans has been examined intensively. Besides being an index to the biological history of the water, the general character of the distribution of oxygen in the deeper water is helpful in studies of currents and of mixing processes. The carbon dioxide distribution is of equal biological importance; its discussion begins on p. 192. Nitrogen has not been studied very widely, as it is apparently chemically inert. Argon is also inert, and is sometimes included with the nitrogen when the

---

dissolved gases are determined gasometrically. The presence of helium and neon has been confirmed by Rakestraw, Herrick, and Urry (1939).

Determination of Dissolved Gases. The content of dissolved oxygen is usually determined by the Winkler method, which depends upon the oxidation of manganous hydroxide by the dissolved oxygen. When acid is added, the oxidized manganese reacts with potassium iodide and sets free iodine, in amounts equivalent to the original dissolved oxygen content, which is determined by titration with sodium thiosulphate. The Winkler method is simple and extremely accurate if certain precautions are observed in handling the water samples and reagents (Thompson and Robinson, 1939).

Problems relating to the determination of carbon dioxide are discussed on p. 192.

Dissolved nitrogen cannot be determined by direct chemical methods, and hence gasometric techniques must be used. In general, the sea-water sample is acidified and all the gases are driven off by boiling or by applying a vacuum. The carbon dioxide is then absorbed in alkali, and the oxygen is absorbed in alkaline pyrogallol. The residual gas is sometimes considered as “atmospheric nitrogen,” although actually there are other gases, principally argon, mixed with it. Rakestraw and Emmel (1937) developed a method for determining the dissolved oxygen and nitrogen content of sea water by first extracting the gases and removing the carbon dioxide, then absorbing the oxygen on phosphorus and the nitrogen on molten lithium. The oxygen contents determined in this way agreed with direct Winkler analyses. The nitrogen determinations on saturated water samples showed results consistently lower than the saturation values according to Fox (1907); further studies (Rakestraw and Emmel, 1938b) indicate that Fox's tables are slightly in error. The gases remaining after the extraction of nitrogen are considered as “argon.”

The presence of hydrogen sulphide can be detected by its characteristic odor. A method for its determination has been described by Gaarder (1916). Although commonly referred to as hydrogen sulphide, a part, at least, will not be present as free gas but as sulphide or bisulphide of some base. A hydrogen sulphide system somewhat comparable to the carbon dioxide system must exist, but it has not yet been investigated.

The determination of ammonia is discussed in the section dealing with nitrogen compounds.

The units to be used in reporting the concentrations of dissolved gases are mg-atoms/L or (ml of gas at NTP)/L.

In some cases it is of interest to know the excess or deficiency of the concentration with respect to water of the same temperature and salinity in equilibrium with the normal dry atmosphere. The saturation values for oxygen and nitrogen are given in tables 38 and 39. If the saturation

---

values are known, the percentage saturation may be calculated. In certain problems it is desirable to know the partial pressures of the gases dissolved in a given water sample. The factors for computing these values are discussed on p. 190.

The dissolved oxygen in the sea varies between zero and 0.75 mg-atoms/L (about 8.5 ml/L), although in areas of low temperature and intense photosynthesis the content may exceed this upper limit. Nitrogen, which is apparently unaffected by biological processes, varies between 0.75 and 1.3 mg-atoms/L (8.4 and 14.5 ml/L). The total

---

carbon dioxide in oceanic waters varies between about 1.5 and 2.5 mg-atoms of C/L (34 and 56 ml/L). “Argon” varies between 0.2 and 0.4 ml/L, and the content of helium and neon in sea water is about 1.7 × 10⁻⁴ ml/L. The latter values apparently represent the saturation values. Hydrogen sulphide, which is present in the water under exceptional conditions, may occur in amounts of more than 1.0 mg-atom of S/L (22 ml/L) (Ström, 1936).

Factors Controlling the Distribution of Dissolved Gases. The following general factors control the distribution of dissolved gases in the oceans: (1) temperature and salinity, which determine the concentrations when the water is at the surface and in equilibrium with the atmosphere, (2) biological activity, which markedly affects the concentrations of oxygen and carbon dioxide, (3) currents and mixing processes, which tend to modify the effects of biological activity through mass movement and eddy diffusion.

Water in contact with the atmosphere will tend to reach equilibrium either by giving up or absorbing the individual gases until the water is just saturated. Although the zone of contact is a thin one, convective movements due to cooling, evaporation, or wind action may bring a layer of considerable thickness into equilibrium with the atmosphere. According to Henry's law the concentration, m, of a gas in a liquid is related to the partial pressure, p, of the gas and to the character of the gas and the liquid: m = c_sp. The numerical value of c_s, the coefficient of saturation (absorption), depends upon the units for expressing the concentration of the gas in the solution and its pressure, and upon the chemical character of the gas and the temperature and salinity of the water.

With the exception of water vapor the relative composition of the atmosphere can be considered for practical purposes as constant (table 40). This does not strictly apply to carbon dioxide, relatively slight changes in the partial pressure of which have a pronounced effect upon

---

the amount in solution, and hence upon the hydrogen ion concentration and other properties (p. 202). Because of the variability in the water vapor pressure, the saturation is always assumed to take place from a dry atmosphere at standard pressure, namely, 760 Torr. The natural fluctuations and regional differences in the atmospheric pressure are neglected.

The solubilities of those gases, such as oxygen and nitrogen, which do not react chemically with the water or its dissolved salts decrease with increasing temperature and salinity. The solubilities of oxygen and nitrogen in sea water of different salinities over the normal range of temperature were investigated by Fox (1907, 1909). Fox's values for oxygen are still the accepted standards, but his data for nitrogen have been superseded by those of Rakestraw and Emmel (1938b). The solubility of carbon dioxide is greater than that of oxygen and nitrogen because it reacts with the water. Part of the carbon dioxide is present as free CO₂ and H₂CO₃, but in sea water by far the greater part is present as carbonates and bicarbonate, and for the same partial pressure the total CO₂ content of sea water is much greater than that of distilled water or neutral salt solutions. The content of free CO₂ and H₂CO₃ decreases with increasing temperature and salinity. Argon is sometimes included with the “atmospheric nitrogen,” and, because its solubility differs from that of nitrogen, the values of the saturation coefficients will be slightly modified. Little is known concerning the other gases in sea water; however, both hydrogen sulphide and ammonia are very soluble gases and their saturation values can play no important part in their distribution.

In table 41 are given values of the saturation coefficients (absorption coefficients) for oxygen, nitrogen, and carbon dioxide in fresh and sea water at different temperatures. The values for oxygen are from Fox (1909), as are also the values for nitrogen in distilled water. The other nitrogen values are from Rakestraw and Emmel(1938b). The values for carbon dioxide (Buch et al, 1932) correspond to the total CO₂ in water of zero alkalinity or to the free CO₂ and H₂CO₃ in sea water. It is seen that carbon dioxide is much more soluble than the other two gases and that oxygen is about twice as soluble as nitrogen.

From table 41 it is seen that within the range of chlorinity normally encountered in the oceans the temperature is the most important property influencing the solubility (see also tables 38, 39).

In studies of the distribution of dissolved gases in the sea it is generally assumed that, whatever the location of a water particle, at some time it has been at the surface and in equilibrium with the air. In their studies of the dissolved nitrogen content Rakestraw and Emmel (1938a) have found that the water is virtually saturated (referred to a normal atmosphere), regardless of depth; therefore this assumption appears valid and also indicates that biological activity involving either fixation or production of nitrogen cannot be sufficient to affect significantly the concentration of this gas in the water. As the waters of the oceans appear to have been saturated with oxygen and carbon dioxide at some stage in their history when they were at the surface, the differences between the saturation values (computed from the temperatures and salinities) and the observed contents are measures of the changes which have been effected by biological agencies. The factors influencing the distribution of carbon dioxide are discussed in the following sections, and the distribution of dissolved oxygen will be considered in many places in the ensuing chapters.

---