pH Measurements
Discrete pH measurements were not collected on P16S. Two different groups - University of Miami (UM) and University of South Florida (USF) - measured pH using slightly different spectrophotometric techniques on P16N leg 1. Only the USF group measured pH on P16N leg 2. The pH measurements from both groups are reported on the total scale at 25°C.
Note: In the master data file p16n_2006a_hy.csv at CDIAC and CCHDO, the pH measurements from UM group reported for leg 1 (stations 1 - 43) and from USF for leg 2 (stations 44 - 84). The separate file p16n_2006a_all_ph.csv with all pH measurements from both groups is posted at CDIAC at: http://cdiac.ornl.gov/ftp/ndp090/
Section P16N_2006 UM pH Measurements
The pH measurements of seawater were made by Dr. Frank Millero's group from the University of Miami on the leg 1 of Section P16N_2006, using the spectrophotometric techniques of Clayton and Byrne (1993). The pH of the samples using m-Cresol Purple (mCP) is determined from
pH = pKind + log [(R - 0.0069)/(2.222 - 0.133 R)] (1)
where Kind is the dissociation constant for the indicator and R (A578/A434) is the ratio of the absorbance of the acidic and basic forms of the indicator corrected for baseline absorbance at 730 nm. The pH of the samples is perturbed by the addition of the indicator. The magnitude of this perturbation is a function of the difference between the seawater acidity and indicator acidity; therefore, this correction was quantified for each batch of dye solution. To a sample of seawater (~5mL), a volume of 2 milli-molar mCP (0.008 mL) was added and the absorbance ratio was measured. From a second addition of mCP and absorbance ratio measurement, the change in pH per mL of added indicator (ΔpH) was calculated. From a series of such measurements over a range of seawater pH, the first addition of indicator used to calculate pH was described as a linear function of the pH measured with the subsequent addition of indicator (i.e., standard addition correction due to the indicator as a function of pH). In the course of routine seawater pH analyses, this correction was applied to every measured pH; i.e., the corrected pH is calculated as
pH = 1.0013(pHi) - 0.0083 (2)
This equation was applied twice for double addition indicator runs and once for single addition indicator runs to yield pseudo replicate runs for every pH sample when a second addition of indicator was added.
Clayton and Byrne (1993) calibrated the mCP indicator using tris (hydroxymethyl) aminomethane (TRIS) buffers (Ramette et al. 1977) and the pH equations of Dickson (1993). They found that
pKind = 1245.69/T + 3.8275 + (2.11 * 10-3) (35 - S) (3)
where T is temperature in degrees Kelvin and is valid from 293.15 to 303.15 K (20 to 30°C) and S = 30 to 37. The values of pH calculated from equations (1) and (3) are on the total scale in units of mol/(kg-soln). The total proton scale (Hansson 1973) defines pH in terms of the sum of the concentrations of free hydrogen ion, [H+], and bisulfate, [HSO4-]
pHT = -log[H+]T = -log{[H+] + [HSO4-] }= -log{[H+](1 + [SO42-]/KHSO4)} (4)
where the concentration of total sulfate, [SO42-] = 0.0282 × S/35 and KHSO4 is the dissociation constant for the bisulfate in seawater (Dickson 1990a).
Lee et al. (1996) have redetermined the value of pKind from 273.15 to 313.15 K (0 to 40 °C) using a 0.04 m TRIS buffer (Ramette et al. 1977). The pH of the TRIS buffer was determined from the emf measurements made with the H2, Pt/AgCl, Ag electrode system (Millero et al. 1993a). At 273.15 K (25°C) the buffer had a pH of 8.076 and yielded spectrophotometric values of pH that were in excellent agreement (~ 0.0001) with those found using equations (1) and (3). These results from 273.15 to 313.15 K (0 to 40°C) were fitted to the equation (S = 35)
pKind = 35.913 - 216.404/T - 10.9913 log (T) (5)
with the standard error of 0.001 in pKind where the constant is on the total scale in {mol/(kg-H2O)}
The values of pH calculated from equations (1) and (5) are on the total scale in units of {mol/(kg-H2O)}. The conversion from the total scale (pHT) {mol/(kg-H2O)} to the seawater scale (pHSWS) in {mol/(kg-soln)} can be made using (Dickson and Riley 1979; Dickson and Millero 1987):
pHSWS = pHT - log{(1 + [SO42-]/KHSO4 + [F-]/KHF )/(1 + [SO4-2]/KHSO4])} - log (1 - 1.005 x 10-3 S) (6)
where the total concentration of fluoride, [F-] = 0.000067 * 35/S, and KHF is the dissociation constant for hydrogen fluoride (Dickson and Riley 1979). The seawater pH scale (pHSWS) was used in further calculations of the internal consistency (Millero et al. 1993b) of the four parameters since the carbonate constants used are on this scale (Dickson and Millero 1987).
The pH system is automated and makes measurements of discrete pH approximately every 12 min on a sample volume of 25 cm3. A microprocessor-controlled syringe pump (Kloehn 50300) and sampling valve aspirates and injects the seawater sample into the 10 cm optical cell at a precisely controlled rate. The syringe rinses and primes the optical cell with 20 cm3 of sample and the software permits 5 min for temperature stabilization. A refrigerated circulating temperature bath (Neslab RTE-17) regulates the temperature of the sample at 25 ± 0.01 °C. An Agilent 8453 UV/VIS spectrophotometer measures the background absorbance of the sample. The automated syringe and sampling valves aspirates 4.90 cm3 seawater and 0.008 cm3 of indicator and injects the mixture into the cell. After the software permits 5 min for temperature stabilization, a Guildline 9540 digital platinum resistance thermometer measures the temperature and the spectrophotometer acquires the absorbance at 434, 578, and 730 nm. During Leg 2 of the cruise, the pH system was converted to an underway mode, in which a Seabird thermosalinograph was inserted on a flowing line from which the syringe pump could draw a sample every 10 min. The measurement process was the same as the procedure above, with the exception of the input salinity coming from the Seabird. The water jacket enclosing the 10 cm optical cell was thermostated with the same underway seawater to yield true in situ measurements totaling 1250 runs. Eight stations of discrete measurements were made at the temperature of the surface waters relative to when the measurement was made and were later normalized to 25°C
Section P16N_2006 USF pH Measurements
University of South Florida (USF) personnel measured seawater pH on the Section P16N legs 1 and 2 cruise using the procedures outlined in SOP 7 (DOE 1994) and in Clayton and Byrne (1993). The pHT on the total scale is calculated using the following equation:
pHT = 1245.69/T + 3.8275 - 0.00211(35 - S) + log[(R - 0.00691) / (2.222 - 0.1331R)]
where T is the measurement temperature (T = 273.15 + t) and S is salinity. The overall precision of pH measurements from duplicate samples was better than 0.001.
On leg 1, twenty-eight of 43 stations were sampled. Discrete USF pH measurements were made on all water samples for which discrete TCO2 measurements were made on leg 2.
USF personnel participating in P16N_2006 cruise are listed below.
| Leg 1: | Dr. Xuewu Liu
Dr. Renate Bernstein |
|
| Leg 2. | Dr. Robert H. Byrne
Dr. Zhaohui Aleck Wang Dr. Johan Schijf Mr. Ryan Bell |
Measurements of seawater pH were obtained using m-Cresol Purple (mCP) as an indicator. Seawater pH, on the total hydrogen ion concentration ([H+]T) scale, was calculated from the equation
pHT = -logTKI + log[(R - e1)/(e2 - Re3)] (1)
where e1 = 0.00691; e2=2.222; and e3=0.1331. The temperature (T) and salinity (S) dependence of the mCP equilibrium constant (TKI) is given as:
-logTKI = 1245.69/T + 3.8275 + 0.00211(35 - S) (2)
and pHT is related to pH on the free hydrogen ion concentration scale (pH = -log[H+]) as follows:
pHT = -log [H+]T = -log[H+] - log[1 + (ST/KHSO4)] (3)
where ST is the total sulfate concentration and KHSO4 is the HSO4- dissociation constant.
A stock solution of m-Cresol purple (mCP) (10 mM) was prepared with mCP sodium salt (Aldrich, catalogue number 211761) in MilliQ water. The R ratio (absorbance of the base form [I2-] divided by the absorbance of the acid form (HI-) of the stock solution was adjusted to 1.5 with an NaOH solution. The R ratio was checked periodically during the cruise and was shown to be stabilized at 1.44. The dye solution was stored in an aluminum-sandwiched plastic bag to exclude air exchange and light from the indicator.
pH samples were fed directly to 10 cm cylindrical glass cells via a 20 cm section of flexible silicone tubing. After the cell was flushed for 20 s, it was sealed with poly-tetrafluoroethene (PTFE) caps, ensuring that there was no trapped air. After the sealed cell was rinsed with tap water and dried with Kimwipes, samples were housed in a 36-position cell warmer at 25°C. After the cells had been thermostated for about 30 min, the pH measurements were initiated beginning with surface samples.
The exterior of the cell was carefully cleaned; the cell was then placed in the thermostated sample compartment of the spectrophotometer (Agilent 8453 UV-Vis Spectrophotometer). The baseline was recorded at three wavelengths (434, 578, and 700). One of the cell caps was then removed and indicator dye was added with a Gilmont pipette. The cap was replaced and the cell was briefly shaken to mix the seawater and the dye. The cell was returned to the spectrophotometer and absorbances were recorded at three wavelengths.
The measurements were computer controlled with a macro code for sample information input, data acquisition, and storage. The program also implemented quality controls for baseline stability and measurement precision. The overall precisions of pH data were evaluated with duplicate samples during the cruise. The precisions of leg 1 pH data were 0.0006 (n=38) and the precision of leg 2 pH data were 0.0014 (n=82) (Figure 4).
The indicator perturbation to seawater sample was evaluated empirically. A pair of additions of dye was made to each of a series seawater samples that had been adjusted to pH before 8.1 and 7.0. The stock indicator concentration was 10 mM. During the cruise, three different volumes of indicator were used. Table 6 presents a summary of indicator addition.
The perturbation term can be adequately expressed by following equations:
For 10 µL addition:
ΔpH = 0.006574pH(measured) - 0.0508, r2 = 0.992 (4a)
For 20 µL addition:
ΔpH = 0.01324pH(measured) - 0.1023, r2 =0.992 (4b)
Thus, sample pH can be calculated using following equations:
For 10 µL addition:
pHcorr = 1.006574 * pH(measured) - 0.0508 (5a)
For 20 µL addition:
pHcorr =1.01324 * pH(measured) - 0.1023 (5b)
The result suggested that perturbation term due to indicator addition is proportional to indicator volume added (Figure 5). A few stations were measured with 15 µL indicator additions. The perturbation term was estimated based on average result of 10 and 20 µL addition.
For 15 µL addition:
pHcorr =1.099 * pH(measured) - 0.0766 (6)
Sample temperature was controlled by circulating the water bath to 25°C. As soon as the sample was measured, its temperature was measured with a platinum temperature probe (traced to NIST standard). Most of the samples were measured at 25±0.1°C. The small temperature difference from 25°C will not add error to measurement due to the inherent properties of mCP and CO2 chemistry. For example, if a sample is measured at 24.9 °C, but t = 25°C was assumed to produce a calculated pH = 8.0000, the pH calculated at 24.9°C would be 7.9985. Using CO2 system thermodynamic relationships, when pH measurements at 24.9 °C are corrected to 25°C, the correction factor is 0.0014, resulting in a corrected value (in the example above) equal to 7.9999. When temperature differs by as much as 0.2°C, the error by assuming t = 25°C is less than 0.0002, which is well within the experimental error of the measurements. Thus no temperature corrections were made to the cruise dataset.
It has been demonstrated that due to the different impurities in indicator batches from different vendors, measured pH can be significantly different. We examined the effect of impurities as function of sample pH. Figure 6 shows the pH discrepancies between Sigma-Aldrich mCP (indicator used during P16N_2006 cruise) and Kodak mCP (indicator used in 1991 cruise) over a range of pH. The pH offset between the two indicators decreases as sample pH decreases. In this case, it is seen that indicator impurities cause smaller pH measurement artifacts at lower pH. As a result, for ocean pH measurements, errors introduced from indicator impurities will be most significant for surface seawater samples. Figure 6 shows pH discrepancies from 7.2 to 8.2. The result shown in Figure 6 can be fitted into the following equation:
pH(Sigma-Aldrich) - pH(Kodak) = 0.0010 + 0.0008 * (pH(Sigma-Aldrich) - 7.2) + 0.0042*(pH(Sigma-Aldrich) - 7.2)2 (7)
Equation 7 provides an empirical approach to correct offsets in pH measurements attributable to different sources of indicator to make measurements directly comparable and consistent. Equation 7 was used to correct all of the data. Further corrections are possible subsequent to characterization of purified indicator dye.
