THE SOLUBILITY OF CALCITE IN SEA WATER

Lament-Doherty Geological Observatory, Colimibia University, Palisades, New York 

The mineral calcite is one of the most prominent phases in deep sea sediments. Its distribution with water depth on the flanks of the oceanic ridges everywhere in the world oceans has the same basic character. Sediments with uniformly high calcite content extend from the crest down the ridge flank to what Berger (1968) has teimed the lysocline. Here a decrease in calcite content with water depth commences. This decrease continues imtil sediments nearly free of calcite are encountered. Below this horizon (often referred to as the calcite compensation depth) the sediments are free of calcite. Although the pattern is everywhere the same the depth of the lysocline and the width of the transition zone (i.e. distance between lysocline and compensation depth) vary from basin to basin (see Fig. l). 

It is generally agreed that this pattern is generated by calcite dissolution. The sediments above the lysocline have lost little calcite to dissolution while those below compensation depth have lost almost all their calcite to dissolution. However, there is little agreement with regard to the relationship between the lysocline (or compensation depth) and the sat iration horizon (i.e. that depth at which sea water goes from calcite supersaturation to undersaturation). Some workers have concluded that it is more or less coincident with the lysocline while others conclude that it lies roughly a kilometer above the lysocline. 

Two uncertainties permit such a range of opinion (l) the solubility of calcite in equilibrium with sea water cannot be 

Fraser (ed.J, Thermodynamics in Geology, 365-379. All Rights Reserved. Copyright 1911 by D. Reidel Publishing Company, Dordrecht-Holland. 

---

In defence of his solubility results Berner makes two arguments with which we would like to take issue. First he states that since he calculated the solubility of calc it e in sea water from measurements of the solubility of aragonite in sea water and the ratio of the solubility of calcite to aragonite as measured in fresh water that he avoids the serious kinetic effects associated with calcite in sea water. Whereas we do not challenge the legitimacy of Berner s approach, we do challenge his contention that direct measurements of the solubility of calcite in sea water cannot be made because of kinetic effects which are not shared by aragonite. In this connection we make four observations. 

The activities of Mg++ and Ca++ obtained from the model of sea water proposed by Garrels and Thompson have recently been confirmed by use of specific Ca++ and Mg++ ion electrodes, and for Mg++ by solubility techniques and ultrasonic absorption studies of synthetic and natural sea water. The importance of ion activities to the chemistry of sea water is amply demonstrated by consideration of CaC03 (calcite) in sea water. The total molality of Ca++ in surface sea water is about 10 and that of COf is 3.7 x 1C-4 therefore the ion product is 3.7 x 10 . This value is nearly 600 times greater than the equilibrium ion activity product of CaCO of 4.6 x 10-g at 25°C and one atmosphere total pressure. However, the activities of the free 10ns Ca++ and COj = in surface sea water are about 2.3 x 10-3 and 7.4 x 10-S, respectively thus the ion activity product is 17 x 10 which is only 3,7 rimes greater than the equilibrium ion activity product of calcite. Thus, by considering activities of sea water constituents rather than concentrations, we are better able to evaluate chemical equilibria in sea water an obvious restatement of simple chemical theory but an often neglected concept in sea water chemistry. 

One other factor must be mentioned. The solubility of calcite increases with pressure. Based on the molax volume of calcite (36.9 cm3 mol l) and the partial molax volumes of Ca + and coitions in sea water (-2 1.6 and 22.1 cm3 mol" respectively at 25 C), Millero and Berner (1972) estimated the AV for this dissolution reaction to be -39. cm3 mol"l. Direct estimates based on solubility meas irements made by Pytkowicz and his associates (Pytkowicz and Connors, 196U Pytkowicz and Fowler, I967 Pytkowicz et al., 1967 Hawley and Pytkowicz, 1969 Culberson, 1972), at various pressixres up to 1000 atm yield a value of -31.0 cm3 mol l at 25 C and -35 6 cm3 mol l at 2 C. The more recent study of Ingle (1975) yields a value of -3. U 0.1 cm3 mol l at 25 C and -U2.3 l.k cm3 mol l at 20C. Thus, the AV value estimated from the partial molar voliames of various ions in sea water at 25 C (-39. cm3 mol l) is significantly larger than those obtained from the solubility experiments under high pressure (-31.0 and -3. cm3 mol ). Were partial molar volumes of Ca and COo ion available at 2 C, a similar discrepancy would likely exist. 

Manganese appears to be capable of altering the apparent solubility behavior of calcite in deep sea sediments. This element has long been noted to be associated with calcium carbonate in deep sea sediments (e.g., Wangersky and Joensuu, 1964), and extensive experimental evidence exists for coprecipitation of Mn2+ with calcite (see Chapter 3). Both Pedersen and Price (1982) and Boyle (1983) have noted the close association of Mn with carbonate material in Panama Basin sediments. In fact, some of the pore waters approach equilibrium with MnC03,and mixed carbonate... 

Sjoberg, E.L. A fundamental equation for calcite dissolution kinetics. Geochim. Cosmochim. Acta 40, 441-447 (1976). Weiss, R.F. Carbon dioxide in water and sea water the solubility of a non-ideal gas. Mar. Chem. 2, 203-215 (1974). Lyman, J. Buffer mechanism of sea water. Ph. D. Thesis,... 

Fig. 4. The apparent solubility of calcite as indicated by in situ saturometry (x ), by the oceanic lysocline ( ), and by laboratory studies at one atmosphere pressure (l for Ingle et al., 1973 and B for Berner, 1976). The numbers associated with the oceanic measurements are the phosphate content of the water (no phosphate was added to the artificial sea water used in the laboratory experiments). The equation fit through the points has the thermodynamic form ...

Calcium carbonate solubility is also temperature and pressure dependent. Pressure is a 6r more important fector than temperature in influencing solubility. As illustrated in Table 15.1, a 20°C drop in temperature boosts solubility 4%, whereas the pressure increase associated with a 4-km increase in water depth increases solubility 200-fold. The large pressure effect arises from the susceptibility of the fully hydrated divalent Ca and CO ions to electrostriction. Calcite and aragonite are examples of minerals whose solubility increases with decreasing temperature. This unusual behavior is referred to as retrograde solubility. Because of the pressure and temperature effects, calcium carbonate is fer more soluble in the deep sea than in the surfece waters (See the online appendix on the companion website). 

Our calculation will be based on solubility data, mainly TK values, valid for 25 °C. and 1 atm. In a few cases stoichiometric constants for different media are available, but only two (for calcite and SrC03) refer to sea water. Furthermore, many of the given values are uncertain, and there are only a few systems where a complete and reliable set of solubility constants enables one to construct a solubility diagram. 

Table 9.5 Model calculation using the computer program PHREEQC (Parkhurst 1995) on deep-sea waters of the ocean. The constant of the solubility product for calcite is accordingly corrected for temperature and pressure. A comparable decomposition of organic matter as contained in Table 9.3 was excluded in this example.

Neither group included phosphate. It is possible that the same calcite (or aragonite) exposed to different amounts of sea water, or the same amount of sea waters with different trace contaminant levels could yield different apparent solubilities due to different states induced into the outermost several molecular layers of the crystal surface. It must be noted that the crystal/sea water volume ratio employed in Berner s experiments is considerably smaller than that employed by Ingle et al. (1973) and for the saturometer experiments. This points out the need for experiments involving greatly different ratios of sea water volume to calcite siurface area and experiments comparing real and artificial sea... 

---