Previous work on the solubility of oxygen in freshwater has been extended to seawater. Measurements of the Henry coefficient in the ranges 0" < t < 45°C and 0 < S < 50 fit the Setschenow relationship for the variation with salinity. The temperature dependence of the Setschenow coefficient for oxygen is found to be K = 0.0225034 -13.6083/T + 2,565.68/p. The equation for the Henry coefficient as a function of temperature and salinity is used to calculate values for unit standard atmospheric concentrations (USAC) in freshwater and seawater in equilibrium with air at a total pressure of 1 atmosphere. It is estimated that the possible error in the new USAC values is no greater than -+O.l% and probably less. Tables and equations are presented for obtaining accurate USAC values in the ranges 0" < t < 40°C and 0 < S < 40. Simple procedures are given for calculating standard atmospheric concentrations at pressures different from 1 atm. The presence of sea salt has a negligible effect on the fractionation of the oxygen isotopes during solution.Benson and Krause (1980a) presented new values for the concentration of oxygen in freshwater in equilibrium with the atmosphere in the temperature range 0" to 40°C. These were based upon very precise and accurate measurements (Benson et al. 1979) of the Henry coefficient, koO, for oxygen The work to be reported here extends the results on oxygen to saline waters. New measurements of the Henry coefficient for oxygen in waters with varying salinity are used to examine the functional dependence of ko,, on salinity S, and to determine the way salting-out varies with temperature. Values are derived for the concentration of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere as a function of temperature, salinity, and at-

Equations and tables are presented from which accurate values can be obtained, in any of several sets of units, for the concentration of oxygen dissolve11 in freshwater in equilibrium with the atmosphere from 0" to 40°C and 0.5 to 1.1 atm. The y are based on values for the Henry coefficient of oxygen, k,,, which have an estimated accuracy of 0.02%. Equations are derived which relate k, to equilibrium concentrations in natilral waters. The equations include corrections for molecular interactions in the vapor pha.;e. Uncertainty about the best way to correct for these interactions limits the estimated act Iracy of the derived values to about & 0.07% at 0°C and 0.04% at 4O"C, but the new results are much more accurate than values from the UNESCO tables. Within their random errors, previous measurements agree very well with the new results. Under equilibrium conditions, and between 0" and 6O"C, the per mil difference between the 340,:3202 abundance ratio in t le dissolved gas and the air is given by 6 = -0.730 + (427/T), where T is in kelvin and the standard deviation is <0.02%0.

No abstract

No abstract

Analyses of precise measurements on dilute aqueous solutions of seven nonpolar gases have revealed several empirical laws that may suggest new theoretical approaches to the structure of water and act as critical tests for specific models. Relationships have been discovered among the solution parameters, the thermodynamic properties of the gases, and the molecular parameters of the gases. The properties of the solutions within the 0–50 °C temperature range of the measurements are illuminated by extrapolating the results to examine the way the systems would behave if they obeyed the same rules outside the experimental temperature range as they do within it. The empirical results are: (1) ln(1/k) =A3(T1/T−1) +A2(T1/T−1)2, where T1 is the absolute temperature at which the Henry coefficient k hypothetically would be unity. (2) A2=36.855 is a dimensionless constant universal to the seven gases. It has the thermodynamic significance of being equal to the hypothetical value at the temperature T1 of the difference in partial molal heat capacity at constant pressure for the gas between the usual standard states for the liquid and gaseous phases, divided by twice the gas constant R. (3) For the noble gases both T1 and A3 are linear functions of the square root of the force constant ε/kB of the gas, and also approximately linear with the fourth root of the gas polarizability. A3 times R is equal to the hypothetical value of the partial molal entropy of solution at T1. (4) For the noble gases, extrapolated graphs of ΔH̄° vs 1/T intersect at a nearly common temperature not very different from the critical temperature for water. A similar statement applies to ΔS̄° vs 1/T. An especially interesting corollary of these laws is that at any ’’scaled temperature’’ T/T1, ΔC̄°p has the same value for all the gases and it is inversely proportional to the square of the scaled temperature. Graphs of ΔS̄° vs (T1/T)2 are linear, with the same slope for all gases and with intercepts which vary smoothly from helium to xenon for the noble gases. The same statement applies to ΔH̄°/T1 vs T1/T. If Henry’s law and law (1) described the properties of the system, not only in the very dilute solutions where the measurements were made, but also at much lower temperatures and much higher dissolved gas concentrations, T1 would be approximately the normal boiling temperature of the hypothetical pure liquefied gas. The values of T1 from the solubility measurements have been found to be related smoothly to the actual boiling temperatures. Similarly, for the noble gases the calculated value of the partial molal enthalpy of solution at T1 is a smooth function of the actual heat of vaporization of the pure liquid. Frank and Evans found for nonpolar gases at a fixed temperature that plots of ΔS̄° vs ΔH̄° were straight lines. Although graphs of our results suggest a similar trend, the apparent relationship is misleading and the two variables are not in fact linearly related. Instead, the equations for the thermodynamic variables lead to (−ΔS̄°)−(−ΔS̄°)1= 1/2[(ΔC̄p°)−(ΔC̄p°)1]. Values for ΔH̄°, ΔS̄°, and ΔC̄p° have been compared with results obtained by other workers. The new values for ΔC̄p° may be the first reasonably reliable ones for that variable. The parameters in Table XV yield probably the best currently available values for the solubilities of helium, neon, krypton, and xenon in pure water. The precisions of these solubilities are approximately 0.1% to 0.2%. Henry’s law specifies the variation of gas solubility with pressure. The corresponding dependence of gas solubility on temperature is given by law (1). When combined they yield f= x exp[A3(1−T1/T)−A2(1−T1/T)2], which constitutes an equation of state for the dilute aqueous solution of nonpolar gases.

No abstract

No abstract

No abstract