Determination of Total CO₂ Concentration in Seawater

Reprinted from Takahashi et al. (1993)

The TCO₂ concentration in 1000 seawater samples was determined during the TUNES-2 Cruise (Fig.3). The coulometric analysis system which was used to measure the TCO₂ concentration in seawater samples during the expedition is similar to the one described by Johnson et al. (1985) and is briefly summarized below. This system consists of a coulometer (Model 5011) manufactured by the UIC Inc. (Jollies, IL) and a sample introduction/CO₂ extraction system of the LDEO design, which differs from the Single Operator Multiparameter Metabolic Analyzer (SOMMA) system used by most of the other participants of the DOE/CO₂ program. In the LDEO system (Chipman et al., 1993), a precisely known volume of seawater sample is introduced manually into an CO₂ extraction vessel using a calibrated syringe, instead of an automated pipette used by the SOMMA system.

Samples for TCO₂ analysis were drawn from the Niskin samplers of the rosette casts directly into 250 ml glass reagent bottles with ground standard-taper stoppers, sealed with silicone vacuum grease. All samples were poisoned with 200 p1 of 50% saturated mercuric chloride solution to prevent biological alteration of the TCO₂, and were analyzed within 24 hours of collection. For analysis, a calibrated volume (ranging between 19 and 20 ml) of water sample was introduced into a CO₂ extraction chamber through a Hamilton valve. The mass of the seawater sample delivered was calculated using the density of the seawater at the temperature of injection using the International equation of State of Seawater (Millero et al., 1980). Prior to the expedition, the volume of each sampling syringe between the two reference stops was determined by repeatedly weighing it "empty" and "filled" with distilled water. The measurements were corrected for the buoyancy due to the air displaced by the water, which amounted to approximately 0.1% of the weight of the water, and the volume was computed using the density of pure water at the temperature of the measurement. Repeated measurements yielded a precision of ± 0.03%. The seawater sample in the extraction vessel was acidified with 1 ml of 8.5% phosphoric acid introduced through the Hamilton valve into an extraction chamber where the evolved CO₂ was removed from the sample and transferred into the electrochemical cell of the CO₂ coulometer by a stream of CO₂-free air generated by a pure air generator. In the coulometer cell, the CO₂ was quantitatively absorbed by a solution of ethanolamine in dimethylsulfoxide (DMSO). Reaction between the CO₂ and the ethanolamine formed the weak hydroxyethylcarbamic acid. The pH change of the solution associated with the formation of this acid resulted in a color change of thymolphthalein pH indicator in the solution. The color change, from deep blue to colorless, was detected by a photodiode, which continually monitored the transmissivity of the solution. The electronic circuitry of the coulometer, on detecting the change in the color of the pH indicator, caused a current to be passed through the cell, electro-generating hydroxyl (OH^-) ions formed a small amount of water in the solution. The OH^- generated titrated the acid, returning the solution to its original pH (and hence color), at which point the circuitry interrupted the current flow. The product of current passed through the cell and time was related by the Faraday constant to the number of moles of OH^- generated to titrate the acid and hence to the number of moles of CO₂ absorbed to form the acid. A thermostated foot-bath mounted on the base of the titration cell was used to extract the heat generated in the cell during titration, to eliminate the shifting of the endpoint of the titration due to change in temperature of the cell solutions.

The coulometer was calibrated using research grade CO₂ gas (99.998% pure) introduced into the carrier gas line upstream of the extraction tube, using alternately two fixed-volume sample loops on a gas sampling valve and measuring the pressure of the gas in the loops by venting them to the atmosphere and determining the barometric pressure using the same electronic barometer used with the pCO₂ system; the loop temperatures were measured to ± 0.05^oC with a thermometer calibrated against one traceable to the NBS, and the non-ideality of CO₂ was incorporated in the computation of the loop contents. Prior to the expedition, the volume of the loops was determined by the weight difference between the loop/injection valve assembly empty and that filled with water. Repeated measurements indicated that the volumes of those loops were precise to ± 0.02%. During the expedition, the coulometer was calibrated several times daily using the gas sampling system described above. The calibration factor, which represents the ratio between the number of moles of CO₂ in the loop and the reading of coulometer, changed during the use of a titration cell above or below the ideal ratio of unity by a few tenths of percent depending upon the condition of the coulometric solution in the titration cell. Figure 4 shows the typical variation of the calibration factor as a function of the cumulative amount of CO₂ titrated by a cell and indicates that it may be represented by a quadratic form. Accordingly, the CO₂ concentration in each seawater sample was determined using a calibration factor estimated using an equation fitted to the calibration data obtained several times for each titration cell. Generally, a titration cell had to be cleaned and filled with new solution after about 40 samples were analyzed. Beyond this number of analyses, the cell began to behave erratically yielding unreliable analytical results. The working equation used for computing the coulometer/cell calibration factor (CF) is as follows.

CF = ( 12.011 x 10⁶) x PA x [LPV x ( 1 + 3 x a) x (TK(calib)-TK(lp))] / {MV(CO₂) x [RD-(TM x BL)]}

where

The following relationships were used for the computation of the total CO₂ concentration in seawater samples using the coulometer;

TCO₂ (µ mol/kg) = CF x DF x [RD - AB - (TM x BL)1 / ( 12.011 x VL x RHO)

where

The precision of the total CO₂ measurements was further tested by analyzing the Certified Reference Solutions each day along with the seawater samples. These Reference Solutions were provided by Dr. A. G. Dickson of SIO and their total CO₂ concentration was determined by the staff of C. D. Keeling, SIO, using his manometric method. Figure 5 shows a comparison between the results of ship board measurements of these Reference Solutions during the expedition. Our shipboard measurements yielded a mean value of 2303.2 ± 1.5 µ mol/kg (N = 156), which compares with 2304.6 ± 1.6 µ mol/kg (N = 9) determined manometrically by the staff of C. D. Keeling. Although our value is smaller than the Keeling value by 1.4 µ mol/kg, these mean values overlap within one standard deviation. Thus, the total CO₂ concentration values reported in this report are consistent with the manometric values within ± 1.5 µ mol/kg. The total CO₂ values listed in this report are not corrected for the systematic difference of 1.4 µ mol/kg.

The water samples for total CO₂ analyses were collected from Niskin samplers after 1 to 2 liters of headspace was formed by the withdrawal of other water samples. Since the marine air introduced into the headspace had lower pCO₂ values than those for the seawater in the samplers, especially after several degrees of warming occurred during hoisting, it is likely that CO₂ was lost from the sample waters to the headspace. During the South Atlantic Ventilation Experiment (SAVE) program, the loss has been estimated by comparing the total CO₂ concentrations determined for those from the 10-liter Niskin samplers with that drawn from the 280-liter Gerard samplers. Because of its large mass, the water temperature in a Gerard sampler changed little during hoisting from the sampling depth. In addition, because of the height (about 1.5 meters), an water sample drawn from the base of the sampler was thought to be unaffected by the air introduced into the headspace. For these reasons, it was considered that little or no CO₂ had been lost f rom the water samples withdrawn f rom Gerard samples. The results of nearly 40 pairs of comparison indicate that the deep water samples collected from the Niskin samplers appear to have lost no more than 2 µ mol/kg of CO₂ before the seawater samples were transferred into the sample bottles for analysis (Takahashi et al., 1993)