nodule, though the results are applicable to other modes of diagenetic car- bonate precipitation. The model derived is based on the following equation for radial ...
CAN DIAGENETIC PRECIPITATION OF CARBONATE NODULES AFFECT PORE-WATER OXYGEN ISOTOPE RATIOS? STEPHEN J. BURNS Geologisches Institut, Universita¨t Bern, Baltzerstrasse 1, Bern CH-3012, Switzerland
ABSTRACT: The idea that carbonate mineral precipitation, by preferentially removing 18O, can cause isotopic depletion in pore waters and subsequently precipitating minerals has been in the literature for 20 years (Irwin et al. 1977). A numerical model of the process shows that for calcite and dolomite precipitation such a mechanism is not viable for reasonable rates of precipitation. Siderite precipitation in marine environments is also not likely to be able to occur rapidly enough to deplete pore waters in 18O.
INTRODUCTION
Early-formed, diagenetic carbonates are common in marine sediments of the continental margins. Calcite, siderite, and dolomite can occur as cements in sandstones, as discrete layers or more commonly as concretions. Diagenetic carbonates may grow over a considerable time period and depth range. For that reason, geochemical studies of concretions can often be used to interpret the evolution of pore-water chemistry in a sedimentary sequence. One of the most useful parameters for examining the growth conditions and history of diagenetic carbonate formation are their oxygen isotope ratios, which can yield information on either the isotopic composition of pore water, if temperature can be estimated or, alternatively, on temperature of formation, if the isotopic composition of pore waters can be estimated. Interpreted temperatures of formation are often combined with estimated geothermal gradients to estimate depths of formation. Thus, oxygen isotope analyses are often combined with measurements of other geochemical parameters to trace changes in pore-water chemistry over time (e.g. Astin and Scotchman 1988; Thyne and Boles 1989; Scotchman 1991). Because of the usefulness of oxygen isotope ratios of carbonates, it is important to understand what factors might influence the oxygen isotopic ratios of diagenetic carbonates. A number of authors have noted that the oxygen isotope ratios of diagenetic carbonates, particularly calcite and siderite, are often several per mil lower than expected for formation during shallow burial in marine environments (see Mozley and Burns 1993 and Morad et al. 1996 for discussion). These anomalously low d18O values indicate either elevated temperatures or lowered pore-water d18O. A number of explanations for this phenomenon have been put forth, including later recrystallization (Burns and Baker 1987; Dix and Mullins 1987; Morad and Eshete 1990; Malone et al. 1994), diagenetic alteration of volcaniclastic ash to smectite (Pirrie and Marshall 1991; Morad and De Ros 1994), influx of meteoric water (Hudson 1978; Carpenter et al. 1988), interaction with organic matter (Sass et al. 1991), increased geothermal temperatures due to overpressuring (Scotchman 1991), and carbonate mineral precipitation itself (Irwin et al. 1977; Lohmann and Walker 1989; Mozley and Carothers 1992; Moore et al. 1992; Mozley and Burns 1993; Morad and De Ros 1994). The original source of the latter idea is Irwin et al.’s (1977) groundbreaking paper on stable isotopic composition of diagenetic carbonates. Therein, they state, ‘‘The major isotopic effect of diagenesis on pore water is depletion of 18O associated with diagenetic minerals enriched in that isotope’’. This explanation has propagated through the literature in the subsequent twenty years without a careful numerical evaluation of its merits. Here, I use a steadystate model of concretion formation to investigate whether mineral precipitation can be rapid enough to develop and maintain pore-water isotopic gradients in the face of diffusive transport, which acts to remove any concentration gradients. JOURNAL OF SEDIMENTARY RESEARCH, VOL. 68, NO. 1, JANUARY, 1998, P. 100–103 Copyright q 1998, SEPM (Society for Sedimentary Geology) 1073-130X/98/068-100/$03.00
MODEL
The effect of authigenic carbonate precipitation on pore-fluid oxygen isotope ratios can be estimated from simple steady-state systems. Two processes need to be considered. First, mineral precipitation, by preferentially removing 18O from the pore water, establishes an 18O concentration gradient, with the pore waters near the growing nodule depleted in 18O relative to waters some distance from the nodule. The faster the nodule grows, the greater is the effect of precipitation. Second, the effect of precipitation is counteracted by the self-diffusion of water, which acts to eliminate the isotopic differences created by precipitation. If the precipitation rate is constant, a steady state is achieved where the removal of 18O is balanced by the diffusive supply. It is not necessary to include advective transport into the model because any fluid flow would transport so much additional water to the site of precipitation that no isotope concentration gradient resulting from precipitation could be sustained. Below, these two processes are modeled for precipitation of a calcite nodule, though the results are applicable to other modes of diagenetic carbonate precipitation. The model derived is based on the following equation for radial diffusion (Crank 1975, or for an alternative treatment of radial diffusion, see Berner 1980):
1b 2 a2
4p Dt Q5
ab
(1)
Cb 2 Ca
in which Q is the quantity of a substance that diffuses through a spherical surface of radius a in time t, D is the sediment diffusion coefficient, Cb and Ca are the concentrations of the substance at a, the surface of the sphere, and b, a distance from the surface, where b . a. We can define a to be the surface of a growing nodule, and the equation can then be used to calculate a flux per unit time of, for example, Ca21 ions or H218O molecules to the nodule. The equation can be simplified by assuming that b k a, so that b/(b-a) will be close to one. The equation then reduces to Q 5 4pDta(Cb 2 Ca)
(2)
Multiplying both sides of Equation 2 by the molar volume v of the material being considered, Qv 5 4pvDta(Cb 2 Ca)
(3) 3
Qv is simply the volume of the sphere of radius a, or 4/3(a ). Substituting this into the above equation and rearranging, t5
a2 3vD(Cb 2 Ca )
(4)
where t is the time required to grow a nodule of radius a for a given concentration gradient. Equation 4 is perfectly adequate if we calculate the growth rate based on a concentration gradient of Ca21, because for every mole of calcium fluxed to the nodule, one mole of carbonate is precipitated. For an H218O concentration gradient, however, further corrections must be made, because the nodule grows much faster than the rate of incorporation of 18O alone, thereby greatly reducing calculated nodule growth times. To calculate the nodule growth rate, the time must be divided by the 18O/(18O 1 16O) ratio of oxygen being incorporated into the nodule. Because 18O K 16O, the latter is essentially the 18O/16O ratio of the nodule, which is related to the
EFFECT OF MINERAL PRECIPITATION ON PORE-WATER 18O
FIG. 1.—Calculated growth times necessary to cause isotopic depletion of porewater d18O for calcite concretions of various size. Numbers alongside curves indicate degree of isotopic depletion in per mil. Also shown is Ca21-diffusion-limited growth rate for conditions outlined in text. 18
O/16O ratio of the water at the nodule surface by the calcite–water fractionation factor . Also, the 18O/16O ratio of the water at the nodule surface is very closely approximated by Ca. When these corrections are applied, Equation 4 becomes t5
Caaa2 3vD(Cb 2 Ca )
(5)
Equation 5 was applied as follows: Cb was defined as the concentration of 18O in SMOW, or water with d18O 5 0‰ (SMOW), temperature was assumed to be 108C for calculating a, a value of 4 3 1026 cm2/s was used for D, which is based on measured self-diffusion of water (Simpson and Carr 1958) adjusted for porosity effects (Li and Gregory 1974), and v for calcite was taken to be 36.9 cm3/mole. Growth times were calculated for nodules of various radii and for Ca values that varied from the 18O concentration of waters with d18O values of from 21 to 25‰ SMOW. Equation 4 was also used to calculate growth rates based on diffusion of Ca21 to the nodule. The case considered assumes that bicarbonate is locally produced in great excess by oxidation of sedimentary organic matter. A value of 5 x 1026 was used for D for Ca21 (Li and Gregory 1974). The concentration of Ca21 in seawater, 10 mM, was used for Cb, and a Ca21 concentration of 1 mM was assumed for Ca. Thus, the calcium concentration gradient assumes an approximately ten times supersaturation with respect to calcite, close to that observed in sediments rich in organic matter (e.g., Sholkovitz 1973). Further decreasing the calcium concentration at the nodule surface results in very little increase in the calcium flux. RESULTS AND DISCUSSION
Figure 1 shows the model results. For example, a nodule of radius 20 cm would have to grow in less than 2000 years to generate and maintain a pore-water isotopic depletion of 1 per mil compared to the initial isotopic composition of the water. The same nodule would have to form in less than 1000 years if the isotopic depletion is 2 per mil. Unfortunately, there is very little solid data on concretion growth rates in natural systems that might be compared to the model results. Although sedimentological evidence and geochemical arguments often support very early concretion growth, even within the upper 1–10 m of sediments (e.g., Seibold 1962; Burns et al. 1988), it is difficult to translate this information into a specific growth rate. In many other cases, co-varying d13C and d18O values for concretions suggest growth over several different diagenetic zones, including from sulfate reduction to methanogenic, or methanogenic to decarboxylation (Irwin et al. 1977; Hudson 1978; Hennessey and Knauth 1985;
101
Burns and Baker 1987; Dix and Mullins 1987; Scotchman 1991). In the continental shelf/slope settings representative of these studies, the diagenetic zones typically span a sediment depth range of several hundreds of meters or more. Thus nodule growth over periods of hundreds of thousands of years to several million years can be inferred, a much longer time than that necessary to produce isotopic alteration of pore waters. Another method of gaining an appropriate perspective on the calculated growth rates is by comparing the growth rate limited by Ca21 to the extent of pore-water isotopic depletion. Growth limited by Ca21 is often considered to be a maximum growth rate for sedimentary concretions in the absence of fluid flow (Berner 1968). Figure 1 shows that in every case, even for a one per mil depletion, the growth rate the Ca21 diffusion-limited case is slower than what is required to achieve a change in the isotopic composition of the pore fluids. A nodule growing as fast as calcium can diffuse to the nodule surface would cause only about a 0.5 per mil depletion in the isotopic composition of the water at the surface of the nodule. The calculated growth rate based on Ca21 diffusion shown in Figure 1 is probably the maximum possible for marine sediments because the concentration of Ca21 goes to 1 mM at the nodule surface. More importantly, both experimental studies (see Morse 1983 for review) and field studies (Raiswell 1988) indicate that Ca21 diffusion is not the rate-limiting factor in nodule growth. Rather than transport-controlled growth, chemical reaction rates or adsorption of foreign ions at the surface of growing crystals in the nodule limits growth rate. Surface-reaction-controlled growth rates are likely to be orders of magnitude slower than diffusion-controlled growth. For example, both Mucci and Morse (1983) and Walter (1986) found that calcite precipitation rates in solution with seawater Mg/Ca ratios were up to two orders of magnitude slower than precipitation from Mgfree solutions. The presence of dissolved sulfate further reduced growth rates (Walter 1986). It could be argued that a ‘‘precipitation effect’’ on oxygen isotope ratios would still be possible in a closed system. Leaving aside the fact that truly closed systems where diffusion of water is not taking place are likely to be very rare in nature, it is problematic to cause rapid and extensive carbonate precipitation in such a system because Ca21 must be internally supplied at a rapid rate. One might envision a sediment system where both calcium and bicarbonate can be internally supplied, with carbonate precipitation then causing depletion of the pore fluids in 18O. The pore-water reservoir of Ca21 in such as system would be rapidly used, however, requiring supply from a solid source. If the source of Ca21 and carbonate is dissolving detrital carbonate the likely change in oxygen isotope ration of the pore fluids will be to become enriched, rather than depleted, in 18O. During burial, temperatures will increase, thereby decreasing the water– carbonate oxygen isotope fractionation such that the oxygen isotope ratio of the dissolving carbonate will be greater than that of the precipitating carbonate, leaving the pore fluids enriched in 18O (Lawrence et al. 1975; Lawrence 1989). The result for calcite concretions also applies to dolomite concretions. The calcium for dolomite growth must also be supplied by diffusion and/or dissolution of detrital calcite. In the former case, growth rates for dolomite concretions would be approximately one half those calculated for calcite for diffusion-limited growth, since the molar volume of dolomite (64.3 cm3/ mole) is about twice that of calcite. Dolomite growing at a calcium-diffusion limited rate would still generate less than a 1‰ depletion in porewater d18O. In the case where calcium is supplied by detrital calcite, dissolution of calcite would buffer the oxygen isotope ratios changes caused by precipitation. Again, a more important factor is likely to be that growth rates are determined by surface reaction rates rather than transport. Dolomite precipitation is known to be strongly kinetically inhibited at low temperatures (see, e.g., Morrow 1982). Oxygen stable isotope ratios of concretions could also be interpreted to argue against precipitation-caused low d18O values. In studies where multiple isotopic analyses of individual calcite and dolomite concretions have
102
S.J. BURNS
been made, there is almost always a trend toward lower d18O values from inner to outer layers of concretions, when trends are present (e.g., Hudson 1978; Kushnir and Kastner 1984; Hennessey and Knauth 1985; Astin and Scotchman 1988; Burns et al. 1988; Scotchman 1991). A number of possible interpretations of such trends are possible, including a precipitation effect (Irwin et al. 1977). If we accept, however, that concretions are likely to grow initially at a faster rate with decreasing growth rate through time, then a precipitation-rate influence on d18O values would dictate a trend of lower to higher d18O values, the opposite of that observed, because the faster the growth rate, the greater the d18O depletion in the pore waters (Fig. 1). The case of siderite precipitation is less straightforward because both Fe21 and bicarbonate may be supplied by diagenetic reactions within the sediment, e.g., CH2O 1 7CO2 1 4Fe(OH)3 → 4Fe12 1 8HCO32 1 3H2O
(6)
and siderite precipitation might proceed as quickly as iron oxyhydroxide reduction. Rapid reduction of iron oxyhydroxides might be expected in sediments rich in both reactive iron minerals and organic matter. Yet, in marine sediments where rapid remineralization of organic matter is taking place, available Fe21 is generally precipitated as iron sulfide minerals. In fact, much of the reduction of iron oxyhydroxides appears to occur by reaction with dissolved hydrogen sulfide either as the iron oxyhydroxides are buried into the zone of sulfate reduction or as hydrogen sulfide produced within the sulfate reduction zone diffuses into overlying sediment (Canfield et al. 1992). Because of the low solubility of iron sulfide minerals, little Fe21 is available for siderite formation, and then only below the zone of sulfate reduction. Thus, very rapid formation of siderite in marine sediments seems unlikely, nor has it ever been reported. It is, therefore, also unlikely that siderite could form quickly enough to generate any depletion in pore-water d18O. In freshwater or mixed marine–freshwater sediments, however, dissolved sulfate concentrations are lower, and rapid precipitation of diagenetic siderite and/or mixed dolomite–calcite–siderite nodules has been observed. Pye et al. (1990) found concretions up to 40 cm in diameter that had grown since World War II in salt marshes in England. Concretion d18O values were as low as -6.4‰. They attributed the low d18O values to precipitation during periods when meteoric water infiltrated the sediments. At the time of their study, however, pore-water samples were of near seawater composition (Pye et al. 1990). Moore et al. (1992) noted the ‘‘overall abundance of carbonates’’, 5–10% by volume, in X-radiographs of Mississippi delta sediments less than 1500 years old. Carbonate d18O values were as low as -8.9‰, which they suggested was caused by rapid siderite precipitation (Moore et al. 1992). At present, not enough information about carbonate formation in such environments is available to determine whether low d18O values are the result of influx of meteoric water or a direct result of carbonate precipitation. In conclusion, numerical modeling of the effect of diagenetic carbonate precipitation on pore-water oxygen isotope ratios demonstrates that precipitation of diagenetic calcite or dolomite cannot occur rapidly enough to cause depletion of 18O. Diffusive mixing of water is rapid in comparison to reasonable precipitation rates, and maintains a nearly constant 18O distribution in the pore water. The same is also likely to be true of siderite formation in marine sediments, where most available Fe21 is precipitated as iron sulfides. ACKNOWLEDGEMENTS
I would like to thank Peter Mozley for numerous discussions; J. Banner, S. Morad and J. Lawrence for helpful, critical comments; and J. Southard for his editorial stewardship.
REFERENCES
ASTIN, M.R., AND SCOTCHMAN, I.C., 1988, The diagenetic history of some septarian concretions from the Kimmeridge Clay, England: Sedimentology, v. 35, p. 349–368. BERNER, R.A., 1968, Rate of concretion growth: Geochimica et Cosmochimica Acta, v. 32, p. 477–483. BERNER, R.A., 1980, Early Diagenesis, A Theoretical Approach: Princeton University Press, Princeton, New Jersey, 241 p. BURNS, S.J., AND BAKER P.A., 1987, A geochemical study of dolomite in the Monterey Formation, California: Journal of Sedimentary Petrology, v. 57, p. 128–139. BURNS, S.J., BAKER, P.A., AND SHOWERS, W.J., 1988, The factors controlling the formation and chemistry of dolomite in organic-rich sediments, in Shukla, V., and Baker, P., eds., Sedimentology and Geochemistry of Dolostones: SEPM, Special Publication 43, p. 41–52. CANFIELD, D.E., RAISWELL, R., AND BOTTRELL, S., 1992, The reactivity of sedimentary iron minerals towards sulfide: American Journal of Science, v. 292, p. 659–683. CARPENTER, S.J., ERICKSON, J.M., LOHMANN, K.C, AND OWEN, M.R., 1988, Diagenesis of fossiliferous concretions from the Upper Cretaceous Fox Hills Formation, North Dakota: Journal of Sedimentary Petrology, v. 58, p. 706–723. COLEMAN M.L., AND RAISWELL, R., 1981, Carbon oxygen and sulphur isotope variation in concretions from the Upper Lias of NE England: Geochimica et Cosmochimica Acta, v. 45, p. 329–340. CRANK, J., 1975, The Mathematics of Diffusion, 2nd Edition: New York, Oxford University Press, 414 p. DIX, G.R., AND MULLINS, H.T., 1987, Shallow subsurface growth and burial of Middle Devonian calcite concretions: Journal of Sedimentary Petrology, v. 57, p. 140–152. HENNESSEY, J., AND KNAUTH, L.P., 1985, Isotopic variation in dolomite concretions from the Monterey Formation, California: Journal of Sedimentary Petrology, v. 55, p. 120–130. HUDSON, J.D., 1978, Concretions, isotopes and the diagenetic history of the Oxford Clay (Jurassic) of central England: Sedimentology, v. 25, p. 339–370. IRWIN, H., CURTIS, C., AND COLEMAN, M., 1977, Isotopic evidence for source of diagenetic carbonates formed during burial of organic-rich sediments: Nature, v. 269, p. 209–213. KUSHNIR, J., AND KASTNER, M., 1984, Two forms of dolomite occurrences in the Monterey Formation, California: Concretions and layers—a comparative mineralogical, geochemical, and isotopic study, in Garrison, R.E., Kastner, M., and Zenger, D.H., eds., Dolomites of the Monterey Formation and other Organic-rich Units: SEPM, Pacific Section, Special Publication 41, p. 171–183. LAWRENCE, J.R., GIESKES, J.M., AND BROECKER, W.S., 1975, Oxygen isotope and cation composition of DSDP pore waters and the alteration of Layer II basalts: Earth and Planetary Science Letters, v. 27, p. 1–10. LAWRENCE, J.R., 1989, The stable isotope chemistry of deep-sea pore waters, in Fritz, P., and Fontes, J.C., eds., Handbook of Environmental Isotope Geochemistry, Vol. B, The Marine Environment: Amsterdam, Elsevier, p. 317–356. LI, Y.-H., AND GREGORY, S., 1974, Diffusion of ions in seawater and deep-sea sediments: Geochimica et Cosmochimica Acta, v. 38, p. 703–714. LOHMANN, K.C, AND WALKER, J.C.G., 1989, The d18O record of Phanerozoic abiotic marine calcite cements: Geophysical Research Letters, v. 16, p. 319–322. MALONE, M.J., BAKER, P.A., AND BURNS, S.J., 1994, Recrystallization of dolomite: evidence from the Monterey Formation (Miocene), California: Sedimentology, v. 41, p. 1223–1239. MOORE, S.E., FERRELL, R.E., JR., AND AHARON, P., 1992, Diagenetic siderite and other ferroan carbonates in a modern subsiding marsh sequence: Journal of Sedimentary Petrology, v. 62, p. 357–366. MORAD, S., AND ESHETE, M., 1990, Petrology, chemistry and diagenesis of calcite concretions in Silurian shales from central Sweden: Sedimentary Geology, v. 66, p. 113–134. MORAD, S., AND DE ROS, L.F., 1994, Geochemistry and diagenesis of stratabound calcite cement layers within the Rannoch Formation of the Brent Group, Murchison Field, North Viking Graben (northern North Sea)—comment: Sedimentary Geology, v. 93, p. 135–141. MORAD, S., DE ROS, L.F., AND AL-AASM, I.S., 1996, Origin of low d18O, pre-compactional ferroan carbonates in the marine Stø Formation (Middle Jurassic), offshore NW Norway: Marine and Petroleum Geology, v. 13, p. 263–276. MORROW, D.W., 1982, Diagenesis 1. Dolomite—Part 1: The chemistry of dolomitization and dolomite precipitation: Geoscience Canada, v. 9, p. 5–13. MORSE, J.W., 1983, The kinetics of calcium carbonate dissolution and precipitation, in Reeder, R.J., ed., Carbonates: Mineralogy and Chemistry: Mineralogical Society of America, Reviews in Mineralogy, v. 11, p. 227–264. MOZLEY, P.S., AND BURNS, S.J., 1993, Oxygen and carbon isotopic composition of marine carbonate concretions: an overview: Journal of Sedimentary Petrology, v. 63, p. 73–83. MOZLEY, P.S., AND CAROTHERS, W.W., 1992, Elemental and isotopic composition of siderite in the Kuparuk Formation, Alaska, effect of microbial activity and water/sediment interaction on early pore-water chemistry: Journal of Sedimentary Petrology, v. 62, p. 681–692. MUCCI, A., AND MORSE, J.W., 1983, The incorporation of Mg12 and Sr12 into calcite overgrowths: Influences of growth rate and solution composition: Geochimica et Cosmochimica Acta, v. 47, p. 217–233. PIRRIE, D., AND MARSHALL, J.D., 1991, Field relationships and stable isotope geochemistry of concretions from James Ross Island, Antarctica: Sedimentary Geology, v. 71, p. 137–150. PYE, K., DICKSON, J.A.D., SCHIAVON, N., COLEMAN, M.L., AND COX, M., 1990, Formation of siderite Mg-calcite iron sulfide concretions in intertidal marsh and sandflat sediments, north Norfolk, England: Sedimentology, v. 37, p. 325–343. RAISWELL, R., 1988, Evidence for surface reaction-controlled growth of carbonate concretions in shales: Sedimentology, v. 35, p. 571–575. SASS, E., BEIN, A., AND ALMOGI-LABIN, A., 1991, Oxygen isotope composition of diagenetic calcite in organic-rich rocks: evidence for 18O depletion in marine anaerobic pore water: Geology, v. 19, p. 839–842.
EFFECT OF MINERAL PRECIPITATION ON PORE-WATER 18O SCOTCHMAN, I.C., 1991, The geochemistry of concretions from the Kimmeridge Clay Formation of southern and eastern England: Sedimentology, v. 38, p. 798–106. SEIBOLD, E., 1962, Kalk-Konkretion und karbonatatisch gebundenes Magnesium: Geochimica et Cosmochimica Acta, v. 26, p. 899–909. SHOLKOVITZ, E., 1973, Interstitial water chemistry of the Santa Barbara Basin sediments: Geochimica et Cosmochimica Acta, v. 37, p. 2043–2073. SIMPSON, J.H., AND CARR, H.Y., 1958, Diffusion and nuclear spin relaxation in water: Physical Review, 2nd Series, v. 256, p. 1201–1208.
103
THYNE, G.D., AND BOLES, J.R., 1989, Isotopic evidence for the origin of the Moeraki septarian concretions, New Zealand: Journal of Sedimentary Petrology, v. 59, p. 272–279. WALTER, L.M., 1986, Relative efficiency of carbonate dissolution and precipitation during diagenesis: A progress report on the role of solution chemistry, in Gautier, D.L., ed., Roles of Organic Matter in Sediment Diagenesis: SEPM, Special Publication 38, p. 1–11. Received 1 August 1996; accepted 29 April 1997.
