Isotopic investigations of carbonate growth on concrete structures R.V. Krishnamurthy a, *, D. Schmitt a , E.A. Atekwana b and M. Baskaran c a Department of Geosciences, Western Michigan University, Rood Hall, Kalamazoo, MI 49008, USA b Department of Geology, Indiana University-Purdue University, Indianapolis, IN 46202, USA c Department of Geology, Wayne State University, Detroit, MI 48202, USA Received 7 August 2001; accepted 10 March 2002 Editorial handling by R. Harmon Abstract Stable C and O isotope ratios were measured in carbonate minerals, growing under concrete structures from two locations in the United States. These locations were under a bridge in Michigan and under an overpass in New York. The d 13 C of the carbonate samples ranged from 21.6 to 31.4% (with respect to V-PDB) and clearly indicated pre- cipitation under non-equilibrium conditions. Indeed, the values in some cases were more negative than could be accounted for by existing models that invoke 4 stages of kinetic fractionation. There have been suggestions that microbial activity involving C from gasoline and other fossil fuel sources might be responsible for the relatively low C isotope ratios measured in these carbonates. To explore this possibility, 14 C measurements were made in some of the samples. All samples measured for 14 C contained bomb C. The range of 14 C concentrations suggested a non-uniform growth rate, although possible fossil fuel-derived carbon in the system needs future investigation. The d 18 O values of the carbonates analyzed from Michigan range from 12.5 to 15.7% (with respect to V-SMOW), with a mean value of 13.7%. The d 18 O values of the NY samples range from 11.8 to 15.2%, with a mean value of 13.1%. The nearly iden- tical mean values at both locations favors incorporation of O from atmospheric CO 2 in carbonate precipitation. Additionally, the 210 Pb radiometric technique was also attempted to explore the applicability of this technique in dating concrete derived carbonates as well as recent carbonates forming in a wide variety of environments. The results gave ages between 64 and 3.8 a and are consistent when compared with the date the bridge was constructed. # 2002 Elsevier Science Ltd. All rights reserved. 1. Introduction Concrete is the most widely used material for con- struction in the world. Per capita, concrete consumption may be used as an index of industrial development. Degradation of concrete is a worldwide concern. Great economic and safety implications are associated with the premature weakening of concrete infrastructures. Prevention of this degradation would be beneficial, by lengthening the life spans of infrastructures, and requiring less revenue to repair or rebuild dilapidated structures (Ropke, 1982). Secondary carbonate mineral growths occur on con- crete structures in most parts of the world, and are believed to initiate and hasten degradation of these structures. These secondary carbonate minerals occur primarily in the form of stalactites, stalagmites, flow stones and crust. Growth initiation of carbonates often is along cracks in the structures and can extend to sev- eral centimeters in length and diameter. If indeed geo- chemical reactions that produce these secondary carbonates contribute to structural damages, the eco- nomic implications are enormous. The settings where these deposits form differ slightly from typical cave and sub soil environments in that there is no soil cover to provide a source of C (in the 0883-2927/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S0883-2927(02)00089-6 Applied Geochemistry 18 (2003) 435–444 www.elsevier.com/locate/apgeochem * Corresponding author. Tel.: +1-616-387-5505; fax: +1- 616-387-5513. E-mail address: [email protected](R.V. Krishnamurthy).
Transcript
Isotopic investigations of carbonate growthon concrete structures
R.V. Krishnamurthya,*, D. Schmitta, E.A. Atekwanab and M. Baskaranc
aDepartment of Geosciences, Western Michigan University, Rood Hall, Kalamazoo, MI 49008, USAbDepartment of Geology, Indiana University-Purdue University, Indianapolis, IN 46202, USA
cDepartment of Geology, Wayne State University, Detroit, MI 48202, USA
Received 7 August 2001; accepted 10 March 2002
Editorial handling by R. Harmon
Abstract
Stable C and O isotope ratios were measured in carbonate minerals, growing under concrete structures from two
locations in the United States. These locations were under a bridge in Michigan and under an overpass in New York.The d13C of the carbonate samples ranged from �21.6 to �31.4% (with respect to V-PDB) and clearly indicated pre-cipitation under non-equilibrium conditions. Indeed, the values in some cases were more negative than could be
accounted for by existing models that invoke 4 stages of kinetic fractionation. There have been suggestions thatmicrobial activity involving C from gasoline and other fossil fuel sources might be responsible for the relatively low Cisotope ratios measured in these carbonates. To explore this possibility, 14C measurements were made in some of the
samples. All samples measured for 14C contained bomb C. The range of 14C concentrations suggested a non-uniformgrowth rate, although possible fossil fuel-derived carbon in the system needs future investigation. The d18O values ofthe carbonates analyzed from Michigan range from 12.5 to 15.7% (with respect to V-SMOW), with a mean value of13.7%. The d18O values of the NY samples range from 11.8 to 15.2%, with a mean value of 13.1%. The nearly iden-
tical mean values at both locations favors incorporation of O from atmospheric CO2 in carbonate precipitation.Additionally, the 210Pb radiometric technique was also attempted to explore the applicability of this technique in datingconcrete derived carbonates as well as recent carbonates forming in a wide variety of environments. The results gave
ages between 64 and 3.8 a and are consistent when compared with the date the bridge was constructed.# 2002 ElsevierScience Ltd. All rights reserved.
1. Introduction
Concrete is the most widely used material for con-
struction in the world. Per capita, concrete consumptionmay be used as an index of industrial development.Degradation of concrete is a worldwide concern. Great
economic and safety implications are associated with thepremature weakening of concrete infrastructures.Prevention of this degradation would be beneficial,
by lengthening the life spans of infrastructures, and
requiring less revenue to repair or rebuild dilapidatedstructures (Ropke, 1982).Secondary carbonate mineral growths occur on con-
crete structures in most parts of the world, and arebelieved to initiate and hasten degradation of thesestructures. These secondary carbonate minerals occur
primarily in the form of stalactites, stalagmites, flowstones and crust. Growth initiation of carbonates oftenis along cracks in the structures and can extend to sev-
eral centimeters in length and diameter. If indeed geo-chemical reactions that produce these secondarycarbonates contribute to structural damages, the eco-nomic implications are enormous.
The settings where these deposits form differ slightlyfrom typical cave and sub soil environments in thatthere is no soil cover to provide a source of C (in the
0883-2927/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
form of CO2) for percolating water nor is there any pre-existing CaCO3 providing the Ca (Harmon et. al, 1975.,Hendy, 1971). The bulk of the calcium in the carbonatesis provided by the concrete itself (Troxell et al. 1968;
Lea, 1970). Carbon from the original limestone used inthe manufacturing process is lost during ‘‘clinkering’’according to the reaction
CaCo3 þ heat 900�Cð Þ ! CaO þ CO2 "
The overall chemical reaction involved in the forma-tion of carbonates from the concrete degrading envir-onment is believed to be the following:
Ca OHð Þ2þCO2 ! CaCO3 þH2O
Few field and some laboratory studies of concrete
carbonates have been carried out using stable C and Oisotopes (e.g. Van Strydonck et al., 1989; Macleod et al.,1990, 1991; Rafai et al., 1992). These studies on the one
hand were designed to address the question of bacte-rially mediated processes involving fossil fuel (oil spillfrom automobiles, for instance) and other general che-
mical means of carbonate mineral formation on con-crete structures. This study describes the use of multipleisotopic techniques to understand the processes respon-
sible for the formation of carbonate mineralization, thesource of CO2, as well as the ages of these growths. Suchmultiple isotope studies have relevance in carbonatedeposition in environments which lack an external
source of C.
2. Sample location
Samples utilized in this study were collected from
Nassau Community College, Garden City, New Yorkand from a bridge west of Flint in Michigan (Fig. 1).The New York samples were collected from a walkingoverpass, walls supporting the overpass, as well as from
beneath the overpass, which also served as a sidewalkfor pedestrians. The Michigan site was a bridge over the
Shiawassee River. Samples were collected from underthe bridge and from the walls beneath the bridge. Thisparticular bridge was extremely degraded. Both thestructures were essentially made up of pure concrete
except for the metal beams that were used for additionalsupport of the structures.The secondary calcite minerals exist in several forms
at these two locations; stalactites, crusts (on the surface,underside, and sides of the structures), and flowstones.The present study utilized stalactite and flowstone sam-
ples. The stalactites analyzed ranged in length from fewcm to approximately 50 cm. The flowstones are largedome-like features that form beneath the stalactites, as
water drips from the stalactites to the flat surfacesbelow. These features range in width from approxi-mately 6 cm to approximately 20 cm. Their heights maybe 6 cm or greater.
Ordinary straight edge razor blades served as themain excavation tool during sample collection. The sta-lactites were removed by cutting between the concrete
structures itself and the bottom of the stalactites. Smallstalactites were placed in glass tubes to protect themduring transport. Flowstone samples were also removed
with the aid of a straight edge razor blade, in a mannersimilar to the stalactites. Flowstones and large sta-lactites were wrapped in tissue paper, placed in plastic
sampling bags and then placed in ordinary cardboardboxes for protection during transport.
3. Experimental
Radiocarbon analyses were performed via Accel-
erator Mass-spectrometry (AMS) at Purdue Uni-versity’s PRIME Lab, Lafayette, IN. For this mgquantities of the samples were digested with H3PO4
(100%) in evacuated septum tubes and the evolved CO2
was purified and sealed in 9-mm glass tubes. Typically,each AMS run required 50–70 micro moles of CO2 fortarget preparation. All stable isotopes analyses were
performed at Western Michigan University’s IsotopeLaboratory according to previously discussed methods(Krishnamurthy et al., 1997). From the yield obtained
after reaction with H3PO4, it was estimated that thesamples analyzed were nearly purely CaCO3 althoughthe mineralogy was not determined. The O and C iso-
tope data are reported with respect to VSMOW andVPDB, respectively. While the reproducibility of d18Oand d13C are better than 0.1% for pure carbonates, for
the filed samples it was 0.2% for both the isotopes.For the purpose of 210Pb chronology, the calcite
sample was cut into segments and weighed. Samplesfrom the segments were powdered using an agate mortar
and the 226Ra concentrations were determined using ahigh purity Ge-Well gamma ray spectrometer (Baskaranand Naidu, 1995). The gamma ray spectrometer was
Fig. 1. Map of the United States showing the two locations
from where the concrete derived carbonates were sampled.
calibrated using IAEA sediment standards. After assay-ing 226Ra concentrations, a known amount of 209Pospike was added to the powdered carbonate as a yieldmonitor and the sample was subsequently dissolved in 6
M HCl. The solution was centrifuged and the super-natant was used for Po plating onto Ag planchets (Bas-karan and Iliffe, 1993). In most samples, very little
residue remained after acid treatment, testifying to thepurity of carbonate. The planchets were assayed for210Po activity using a solid-state surface barrier detectorcoupled to an Octete PC multichannel analyzer.
4. Results and discussion
All the isotopic data including radiocarbon resultsand 210Pb ages are given in Table 1. The data for the
samples from Michigan are also depicted graphically inFigs. 2 and 3, respectively . The significance of thesedata is discussed below.
4.1. 210Pb ages
The 210Pb chronology is based on the argument that
when surface water containing Ca and 226Ra passesthrough the cracks in the bridge, Ra gets incorporatedinto the carbonate that precipitates from the flowing
water. Since 210Pb is highly particle-reactive, the dis-solved 210Pb concentration in the water is likely negli-gible. The source of 210Pb in the carbonate precipitate is
likely to have derived from the decay of its parent,
Fig. 2. Cross sectional views of the flowstones and stalactite
specimens from the Michigan site along with their 210Pb dates.
The arrows indicate the spots from where sub samples were
extracted for analysis.
Fig. 3. Cross sectional views of the flowstones and stalactite specimens from the Michigan site along with their C and O isotope ratios.
The arrows indicate the spots from where sub samples were extracted for analysis. Corresponding sample numbers are shown in
226Ra. Thus, the disequilibrium between 210Pb-226Racan be utilized to date the age of the concrete derivedcarbonate mineral collected under the Michigan bridge.The activity of 210Pb at any time is given by the rela-
tionship:
[210Pb]m=210Pbi e
�lt+226Ra [ 1- e�lt]
where [210Pb]m is the measured activity, [210Pb]i is theinitial activity (when the carbonate mineral precipitated),l is the decay constant of 210Pb (=0.03108 a�1), and ‘t’is the time elapsed from the time of precipitation to
counting of the samples. Assuming that the initial 210Pbconcentration is negligible, from the measured value of210Pb and 226Ra, the age can be obtained.
Table 1
Isotopic results for concrete derived carbonate samples from Michigan ( s1-s21) and New York (ncc1-nc9)
210Pb ages determined for five samples from theShiawassee River Bridge in Michigan are shown inFig. 2. The samples analyzed include a composite oflayers from a stalactite and individual layers of a flow-
stone. The purpose of the 210Pb dating was two-fold: (1)to determine whether this technique is applicable todating concrete- derived carbonates of minerals, and (2)
to determine the timing of the process of initiation ofsecondary mineral growth and presumably degradationof the bridge.
Results from composite layers of the stalactite gaveages of 64 and 18 a for the longer and the shorter sta-lactites, respectively (Fig. 2). Since these dates were a
composite age of several layers, the date obtained is anaverage, or composite, of the stalactite, and thus can beconsidered a minimum age. Based on the nature of thegrowth of stalactites, the age of the longer stalactite
suggests that growth of this stalactite and presumably,degradation of the bridge started at least 64 years ago.This is consistent with the fact that the bridge was con-
structed in 1931 and it was 67 years old at the time ofsample collection and analysis in 1998. The smaller sta-lactite records an age of 18 a. This young age is con-
sistent with the manner suggested for the formation ofthese stalactites, as they are expected to grow in breadthand length with time (Macleod et al., 1991).
The 210Pb dates may provide other clues to the timingof the onset of degradation of the bridge. If severalsamples are dated with respect to their position below abridge, then information on the onset and progression
of carbonates and hence the likely degradation spatiallymay be obtained. 210Pb dates for the flowstone from thesame bridge are also shown in Fig. 2. The ages ranged
from 4.8 a at the feature’s top to 5.9 at the base. Thedistribution of the ages suggests that the formationprocess associated with flowstones is rather rapid. The
data represent 2 cm growth in just 2 a. The fact that theages obtained for the flowstone are much younger thanthose of the stalactites suggests that the flowstone wasformed much later in the degradation process. The
results from these analyses also demonstrate the poten-tial of this method as a useful chronological tool inconcrete-degrading environments.
4.2. Stable carbon isotope ratios
A distinctive mass dependent fractionation accom-panies carbonate dissolution-reprecipitation cycle innature (Deines et al., 1974). The dissolution can occur
either in an open system or a closed system, although innature the actual mechanism might be a combination ofboth. In open systems, where there is a continual supplyof CO2, the d13C of the dissolved inorganic C (DIC) is
controlled by the d13C value of the gaseous CO2 source.The CO2may be atmospheric or from the soil zone derivedfrom root respiration or organic matter decomposition.
Further, in the latter case the CO2 dissolving in perco-lating water will often carry an imprint of the vegetationtype i.e. the so-called C3 plants (average isotope value�25%) or C4 plants (average isotope value �12%).
As opposed to open system dissolution, under closedsystem conditions the d13C values of DIC is not con-trolled exclusively by the gaseous CO2 source. The H
+
ions provided by the H2O+CO2 mixture which origi-nate in the soil zone, are utilized for the dissolution ofaquifer carbonates. The C supplied from the dissolution
of aquifer carbonates has a value of approximately 0% .Dissolution of this parent carbonate and re-precipita-tion under isotope equilibrium conditions and the
accompanying fractionations is reasonably well under-stood. The predicted isotopic ratios of carbonates pre-cipitating under equilibrium conditions for varioussystems, using the above end member values and a tem-
perature of 25 �C are summarized in Table 2. It may benoted that the above discussion is meant to summarizethe normal geochemical pathway of carbonate genera-
tion in nature and not strictly to suggest similarities withthe secondary carbonate mineralzation studied here.Values of d13CCaCO3 observed in the present study
(Table 1; Fig. 3 ) are all more negative than would beexpected if precipitation occurred under isotopic equili-brium conditions with atmospheric or soil respired CO2
as the source CO2 gas (Mook et al., 1974). The d13Cvalues of the New York stalactites ranged from �21.6 to�31.4%, with a mean value of �28.4%. The range ford13C of the Michigan samples analyzed was �22.3 to
�28.9%, with a mean value of �24.6%. Hence, pre-cipitation of these carbonates under isotopic equili-brium conditions can be ruled out.
If the precipitation did not occur under isotopicequilibrium conditions, the isotopic values observedmust be the result of some kinetic process(es). A kinetic
fractionation, ranging from �11.3 to �10%, accom-panies the production of freshwater travertine by inter-action of CO2 with hyper-alkaline waters (O’Neil andBarnes, 1971). This fractionation results as CO2 mole-
cules cross the gas–liquid interface. It has been sug-gested that the fractionations associated with this hyper-alkaline system are comparable with those associated
with carbonate precipitation in concrete environments(Letolle et al., 1988, 1990a, 1992; Macleod et al., 1991).Adopting a post anthropogenic atmospheric CO2 d13Cvalue of �8%, the d13C of the carbonates should bebetween �19.3 and �18%. Additional fractionation isrequired to explain the values obtained in this study.
Other studies suggest this fractionation is a result of 4different kinetic fractionation steps, namely (1) theinvasion of the liquid phase by gaseous CO2, (2) thediffusion of CO2 in the liquid phase, (3) the reaction of
CO2(aq) with OH� ions, and (4) the kinetic effectsresulting from CaCO3 precipitation (Letolle et al.,1990b). The total maximum fractionation suggested by
this model is �15.5�1.5%. This would result in the
d13C values of the carbonates studied to be in the regionof �23%. Observed d13C values of several carbonatesdeposited in other concrete environments (Macleod et
al., 1991), as well as some in the present study, are morenegative than can be explained by a combination of theabove fractionations (Letolle et al., 1990a). An addi-tional fractionation process must occur to account for
the more negative values observed.It has been suggested that the additional fractionation
may be a consequence of a secondary diffusive process
that results from the passing of CO2(aq) moleculesthrough calcitic crusts that rapidly precipitate at thebase of concrete structures upon contact with atmo-
spheric CO2 (Macleod et al., 1991). This fractionation issuggested to be pH dependent (Siegenthaler and Mun-nich, 1981). The proposed fractionation resulting from
the diffusion through calcitic crusts is �7%, at a pH of10. Further, it has also been suggested that the diffusivefractionation processes may be enhanced at pH>12.5.Carbonates precipitated from solutions with such pH
values are suggested to be more depleted in 13C thanthose precipitated from solutions with lower pH (Letolleet al., 1990b).
The model attributing the additional fractionation ofC isotopes as 12C preferentially permeates through cal-citic crusts is based on d13C values observed for sta-
lactites. The majority of those stalactites had d13Cvalues of �26%. The majority d13C values observed forthe stalactites in the present study are more negative(�29%), at least from one of the sites, than can be
explained by this model. This is illustrated by plottingthe d13C values of the carbonates in the form of a his-togram, which shows two strong groupings (Fig. 4). The
�29% grouping belongs to the stalactites obtained fromthe New York site, and the �24% grouping belongs tothe flowstone samples obtained from the Michigan site.
Consideration of the enhanced fractionation abovepH 12.5 is intriguing. The values observed in the presentstudy become more negative as one measures from the
top of a stalactite towards its bottom. This can be seenfor sample ‘‘ncc200, where the d13C values of the sta-lactite’s top, middle, and bottom are �27.3, �27.8, and�29.2%, respectively (Table 2). If the additional frac-
tionation is pH controlled, the data suggest that the pHof the solution must vary from one portion of the sta-lactite to another. Another interesting observation is
that the flowstones are consistently heavier than the
stalactites. The flowstones form a strong grouping near�24%, whereas the stalactites form the grouping near�29%. If the fractionation is pH dependent, this trend
implies that the solutions responsible for flowstone for-mation have lower pH than those responsible for sta-lactite formation. Indeed, it is assumed that thefractionations that apply to the stalactites also apply to
the flowstones. This assumption can be field tested byanalyzing the two types from the same locality.Irrespective of the number of fractionation steps, it is
evident from the stable C isotope ratios that the princi-pal source of C in these secondary minerals must beatmospheric CO2. Further support for this argument is
available from the radiocarbon measurements in someof these samples.
4.3. Radiocarbon
The 14C activity range of the samples analyzed in thepresent study is 102.9 to 111.4% modern C (pmc). Fig. 5
shows the atmospheric 14C activity time series (Genty,personal communication) for the lifespan of the Shia-wassee River Bridge in Michigan. Radiocarbon data in
young cave stalagmites have been recently modeled(Genty et al., 1997, 1998; Genty and Massault, 1999).These authors used cave stalagmites to construct a 14C
activity time series. The reconstructed radiocarbonactivities were less than those of the measured atmo-spheric 14C activities of the corresponding years. Alsoobserved was a lag in the 14C record, which can vary
between 10–40 a, with the stalagmite 14C data laggingbehind atmospheric 14C data.Three C sources were suggested for the formation of
the speleothems: (1) limestone (dead C), (2) organic Cwith a fast turnover rate, and (3) organic C with a slowturnover rate. Organic C that has a fast turnover rate
(decays rapidly) will have a 14C activity equal to that ofthe atmospheric activity of when it was alive. Organic Cwith a slow turnover rate will have a composite 14C
activity of organic C of different ages (and different 14Cactivities). The lag time and dampening of 14C activitieswas primarily attributed to the contribution of the Cfrom the organic C with a slow turnover rate. This effect
(up to 80%) was more pronounced in caves that were inforested areas. A small contribution to the dampeningmay be supplied by dead C (limestone).
Table 2
Theoretical d13CCaCO3 with respect to various d13C CO2
d13C of Source CO2 Open d13CDIC System d13CCaCO3 Closed d13CDIC System d13CCaCO3
Our 14C data (Tables 1 and 2) can be interpreted in asimilar manner. Firstly, it is assumed that the contribu-tion from dead C is zero. Secondly, it is assumed that
no, or very little, organic is expected to contribute tothis system, due to the lack of a soil zone above bridges.Thirdly, it is assumed that atmospheric CO2 is the pri-
mary source of C for the formation of these stalactitesand flowstones. The stalactite and flowstone samplesanalyzed in this study are composites of several layers of
carbonates that were formed at different times. Theportions of the samples analyzed for 14C were not indi-vidual layers, and therefore provide minimum ages. For
Fig. 4. Histogram showing distribution of C isotope ratios of the carbonates from both New York and Michigan locations.
Fig. 5. Time series of atmospheric radiocarbon concentration (after Genty and Massault, 1999) showing the input of bomb produced
example, the 14C activity of sample ‘‘s5’’ dated 64 a (seeTable 1) by the 210Pb method is actually given by:
a 14C ¼
P64
i¼1
a 14Catmi
64
The total 14C activity will thus depend upon a weigh-ted contribution from the activities of the years duringwhich the stalactite was deposited.
Based on the measured atmospheric 14C data (Fig. 5)and assuming that sample s5 grew uniformly during thehistory of the bridge, the expected 14C activity is 127
pmc. It is interesting that such a value was not observedfor any of the samples analyzed in the present study.Two tentative explanations are provided for this: (1) thegrowth rates are not uniform; periods of rapid carbon-
ate growth may exist. Variations in the growth rates ofthese minerals may have implications in terms of therate of concrete degradation. (2) at least for samples
whose d13C values are more negative than model pre-dictions coupled with the fact that the radiocarbonconcentrations are also somewhat lower than estimated,
more studies are warranted to rule out microbiallyintroduced C into the system.
4.4. Stable oxygen isotopes
The d18O values of the carbonates analyzed from theShiawassee River Bridge range from 12.5 to 15.7%, with
a mean value of 13.7%. The d18O values of the NYsamples range from 11.8 to 15.2%, with a mean value of
13.1%. Adopting a mean annual d18O value of �8% forthe local meteoric water in Michigan (Machavaram andKrishnamurthy, 1994), and assuming that it is involvedin the carbonate precipitation under equilibrium condi-
tions, temperatures of carbonate formation can beobtained by incorporating the calcite water fractiona-tion equation of Friedman and O’Neil (1977). The esti-
mated temperatures range from 63.7 to 67.8 �C, with amean temperature value of 65.7 �C. These temperaturesare unreasonably high and would be even more unrea-
listic if only the isotope ratio of summer or winter pre-cipitation alone is considered. This is more evidence that(1) non-equilibrium processes are involved and (2) the O
in the carbonates can not be provided exclusively bymeteoric water.Plotting d13C vs. d18O values of the carbonates pre-
cipitated at the sites in the present study shows a posi-
tive correlation (Fig. 6). Such a correlation is againconsistent with kinetically controlled precipitation ofthese carbonates, resulting in a co-variation of d13C and
d18O.It has been suggested that the d18O of freshwater tra-
vertine is determined by both the waters which formed
the precipitate and atmospheric CO2 (O’Neil andBarnes, 1971). Two thirds of the O atoms in the calciumcarbonates may be derived from atmospheric CO2, with
the remaining third coming from the formation waters.Taking an atmospheric CO2 d18O value of +41% (Bot-tinga and Craig, 1969), the expected d18O of the carbo-nates would therefore be 24.7%. The fact that the
samples in the present study do not possess such valuesimplies that some re-equilibration process involving the
Fig. 6. Plot showing d13C–d18O correlation between all the carbonate samples analyzed. A strong correlation is suggestive of kinetic
fractionation driving the precipitation of these carbonates.
CO2 and formation waters must take place during theprecipitation of these secondary minerals.d18O values of meteoric waters vary for different
locations. d18O values of atmospheric CO2, however, are
not dependent on location. Thus, the similarity betweenthe d18O values of the New York and Michigan samplessuggests that the imprint of the waters which formed the
CaCO3 have been erased by re-equilibration and thatthe O isotope ratio is influenced by atmospheric CO2 .
5. Conclusions
Multiple isotope investigations were carried out onsecondary carbonate minerals that grew under concretestructures in two localities in the USA. Application ofthe 210Pb dating technique shows promise in that this
technique may be useful for dating similar deposits.Stable C and O isotope ratios indicate that these carbo-nates were deposited under non-equilibrium conditions.
While the d13C values fall close in several cases to thatexpected by a 4-stage kinetic precipitation model, thereare some deviations. Taken together, the d13C values
and radiocarbon concentrations point to atmosphere asthe predominant source of C in the system although thepossibility of microbially derived C playing some role
warrants further investigation. The O isotopes in thesecarbonates appear to be mostly derived from atmo-spheric CO2 .
Acknowledgements
D. Schmitt is grateful to the Graduate College, Wes-tern Michigan University, for a graduate research grant.We are also grateful to Dr. D. Genty for providing us
atmospheric radiocarbon data and the AMS facility atPurdue University’s PRIME lab for providing a seedgrant to carry out the radiocarbon measurements. Par-tial support from the Faculty Research and Creative
support, WMU, is acknowledged. Constructive reviewsby Drs. Genty and Fallick are greatly appreciated.
References
Baskaran, M., Iliffe, T.M., 1993. Determination of recent cave
deposits using excess 210Pb-a new technique. Geophys. Res.
Lett. 20, 603–606.
Baskaran, M., Naidu, A.S., 1995. 210Pb-derived chronology
and the fluxes of 210Pb and 137Cs isotopes into continental
shelf sediments, East Chukchi Sea, Alaskan Arctic. Geochim.
