Page images
PDF
EPUB

pH electrode. The long-term stability of ammonium ion will be discussed below. Chloride was also determined potentiometrically using a chloride ion selective electrode. However, the concentrations of chloride in the simulated rainwater are well below the linear range of response. Thus they are prone to a large uncertainty. The chloride measurements serve only as an indication of concentration and were not used in the final statistical analysis. This method and its values are presented for information only.

The acidity of SRM 2694-II was determined by coulemetric reduction of hydrogen ion in a weighed sample, from which CO2 had been removed by purging with argon [3,4]. Titrations were carried out to the neutrality point, determined potentiometrically.

Thermal ionization isotope dilution mass spectrometry was used to determine the total sulfur in the simulated rainwater. This value was then converted to sulfate concentration. Briefly, the method involves the reduction of sulfate to sulfide, distillation of sulfide and collection in a basic arsenic (III) solution, followed by mass spectrometric identification and quantitation of AsS+ isotopes (7,8).

Spectrophotometry was used for the determination of nitrate. The procedure was based on the color reaction produced by interaction of nitrate ion and the organic reagent, brucine. Careful attention to experimental procedures and frequent calibration with nitrate standards were required to obtain satisfactory results by this method.

Laser-enhanced ionization flame spectrometry, a new analytical tool in the Inorganic Analytical Research Division, was utilized for the determination of sodium, potassium, calcium, and magnesium. Wavelength scans near the analysis lines were performed to check for spectral background and interfering lines. Minor corrections were applied to the magnesium determination to correct for sodium interference. Magnesium concentration was also determined by flame atomic absorption spectrometry.

Inductively coupled plasma spectrometry was used to measure the calcium concentration in the samples, and flame emission spectrometry was used to determine sodium and potassium. The recovery of each analyte was checked by the single standard addition method.

In addition, the density of the solutions was determined to be 0.997g/ml at 23 °C, essentially identical to pure water at this temperature.

standard deviation of a single measurement. Values in braces are the number of analyses performed. We have used established techniques to calculate weighted averages for the SRM certificate values. The statistical weighting is based on the observed variablities of the various data sets for each analyte. The procedures for deriving the weighted averages and their uncertainties have been described in a previous article (9). Table 7 contains the certified values and uncertainties for SRM 2694.

The uncertainties associated with fluoride, nitrate, sodium, potassium, calcium, and magnesium are two standard deviations of the certified values. The uncertainties in the certified values for pH, acidity, and specific conductance are based on scientific judgment and experience, rather than on true statistical evaluations because there were no practical second methods of analysis for these components.

The uncertainties tabulated for sulfate are also based on scientific judgment even though there were two independent methods of analysis. For this analyte, the agreements both within and between the methods of analysis were so close that the statistical evaluations of the uncertainties were not believed to be realistic. The uncertainties for these latter four components (pH, acidity, specific conductance, and sulfate) are believed to be roughly equivalent to two standard deviations of the certified values.

The values for chloride and ammonium are listed for information only. Chloride was not certified because the potentiometric method had insufficient precision and accuracy at these levels to corroborate the ion chromatographic data. Until a second independent technique verifies the IC measurements, chloride will not be certified. Ammonium ion was not certified because of very real concerns about the stability of this ion in these solutions.

5. Discussion

The problem with the stability of ammonium was first noticed upon reanalysis of RM 8409 eight months after its preparation. The concentration of ammonium in RM 8409-1 had decreased from 0.085 mg/L to 0.025 mg/L. Reanalysis of other samples of simulated rainwater, which were part of a long-term stability study of pH and conductivity, indicated that similar decreases in ammonium ion had occurred. The decrease was significant when the initial concentration of ammonium was below 0.2 mg/L and the pH was above 4.0 (see table 8). The cause of this decrease is not known at this time, but it is suspected to be biological activity. The loss of ammonium does not appear to have significantly affected any

4. Results and Statistical Analysis

The average values from each measurement technique are shown in table 6. Values in parentheses represent the

[blocks in formation]

12. Magnesium, mg/L A. LEIS

0.0231 0.0003) {14}

0.0492 0.0005) {13} B. FAAS

0.025 0.001) {6}

0.052 (.006) {6} lValues in pareacheses represent the standard deviation of a single measurement. Valuca in braces are the number of determinations other components. There have been isolated incidences us that a sample of RM 8409 which had been opened, of visible fungal growth in a few bottles of RM 8409. In recapped, and stored in a refrigerator lost virtually all of such cases the values for nitrate, pH, acidity, and spe- its nitrate content but gained a significant amount of cific conductance, have changed. A solution to this nitrite. problem involving the sterilization of the "simulated SRM 2694 should be stored in an area free from acid rainwater" is currently under investigation.

and/or ammonia vapors. These vapors can permeate the It must be noted that the solutions of SRM 2649 are polyethylene bottles and contaminate the samples. A set very dilute, unbuffered solutions, and, as such, are very of samples placed in our laboratory refrigerator, which susceptible to contamination causing gross changes in also contained a polyethylene bottle of concentrated the certified values. Therefore, the solutions should be ammonium hydroxide, showed a substantial increase in used immediately upon opening. No assurance can be ammonium ion concentration. Refrigeration of SRM made as to the composition or stability of the solutions 2694 is not necessary. However, the solutions should not after being opened and recapped. It has been reported to be exposed to extreme heat (i.e., temperatures above

[blocks in formation]

The contributions of the following individuals at the National Bureau of Standards are gratefully acknowledged: D. G. Friend, D. E. Swearingen, and F. Smithers for bottle cleaning, bottling and packaging; T. C. Rains, R. W. Burke, M. Knoerdel, R. W. Kelly, G. C. Turk, Mo De-Ming, M. S. Epstein, T. A. Rush, T. A. Butler, Han Kai, and M. V. Smith for the chemical measurements; and T. E. Gills and L. J. Powell for coordinating the technical and support aspects from the Office of Standard Reference Materials.

[blocks in formation]

References

Sodium, mg/L Potassium, mg/L Ammonium, mg/L Calcium, mg/L Magnesium, mg/L

0.419+0.015 0.106 +0.008

(1.0) 0.049+0.011 0.051+0.003

0.014 +0.003 0.024 +0.002

The estimated uncertainties are two standard deviations of the certified values, except in the cases of pH, acidity, conductance, and sulfate which are based on scientific judgment and are roughly equivalent to two standard deviations of the certified value.

2 The values for chloride and ammonium are not certified, but are listed for information only.

a

35 °C) as this will accelerate transpiration of water vapor from the bottles. The bottles, sealed in aluminized bags to retard transpiration, should remain within the sealed bags until just before use. The search for a better container has been initiated.

If conductance and pH are to be measured on the same sample, conductance should be measured first. Otherwise, leakage of concentrated KCl from the pH reference electrode will affect the conductance reading. The measurement of pH should be performed according to the guidelines set forth in an attachment to the Certificate of Analysis of SRM 2694 and appended to this report (appendix 1). Adherence to this procedure will minimize the bias caused by residual liquid junction potentials. Acidimetric titrations should be performed on samples that have been purged of dissolved carbon dioxide to prevent drifting endpoints and high results.

[1] Koch, W. F.; G. Marinenko and Y. C. Wu, The Development of

Reference Materials for Acid Rain Research, Environment

International 10, 117-121 (1984). [2] Koch, W. F., and G. Marinenko, Simulated Precipitation Refer

ence Materials: Measurement of pH and Acidity, Sampling and Analysis of Rain, ASTM STP 823, S. A. Campbell, ed.,

American Society for Testing and Materials, 10–17 (1983). [3] Marinenko, G., and W. F. Koch, A Critical Review of Mea

surement Practices for the Determination of pH and Acidity of Atmospheric Precipitation, Environment International

10, 315-319 (1984). [4] Marinenko, G., and W. F. Koch, Evaluation of Methods Used

for the Determination of Acidity in Acid Rain Samples, Natl.

Bur. Stand. (U.S.), NBSIR 85-3114, 16 pp. (March 1985). [5] NBS Standard Reference Materials Catalog 1984-1985, C. H.

Hudson, ed., Natl. Bur. Stand. (U.S.), Special Publication

260, p. 2, (February 1984). [6] Moody, J. R., and E. S. Beary, Purified Reagents for Trace

Metal Analysis, Talanta 29, 1003-1010 (1982). [7] Paulsen, P. J., and W. R. Kelly, Determination of Sulfur by

Isotope Dilution Thermal Ionization Mass Spectrometry as

AsS+ Ions, Anal. Chem. 56, 707-713 (1984). [8] Kelly, W. R., and P. J. Paulsen, Determination of Sulfur in NBS

Coals by Isotope Dilution Thermal Ionization Mass Spec-
trometry, in Methods and Procedures used at the National
Bureau of Standards to Certify Sulfur in Coal SRM's for Sulfur
Content, Calorific Value, Ash Content, T. E. Gills, ed., Natl.
Bur. Stand. (U.S.), Special Publication 260-94, pp. 7-13 (De-

cember 1984). [9] Paule, R. C., and J. Mandel, Consensus Values and Weighting

Factors, J. Res. Natl. Bur. Stand. 87, 377–385 (1982).

Table 8. Stability of ammonium ion in simulated rainwater.

Ammonium lon, mg/l Initial

Feb. 20, 1985

Sample

pH

RM 8409-1 RM 8409-11

Preparation Date

June 1984
June 1984

[blocks in formation]

4.3
3.6

0.085
1.07

0.025
1.07

4.5 3.5 4.0

0.14
1.11
0.42

0.09
1.08
0.42

Appendix

Guidelines for the Measurement of pH in Acidic Rainwater

This report presents a recommended procedure for the measurement of pH in acidic rainwater. The intent of this guideline is to improve the accuracy and precision of the pH measurement with special emphasis on reducing the effect of the residual liquid junction potential. It consists of three major parts: Calibration Sequence, Control Sequence, and Rainwater Measurement Sequence. The purposes of the Calibration Sequence are to accurately calibrate the pH measurement system with robust buffer solutions, to accurately set the slope, and to verify that the measurement system is functioning properly. The purpose of the Control Sequence is to quantitatively determine the magnitude of the residual liquid junction potential bias for a particular set of electrodes which must be applied in the rainwater measurement sequence to obtain more reliable and intercomparable results. Each sequence should be executed in stepwise order with strict adherence to detail.

Note: This guideline is applicable only to the measurement of pH in acidic rainwater and acidic low ionic strength aqueous solutions. It should not be used for any other applications as inaccuracies may ensue.

General Directions

Control Sequence

Record the solution temperature to within 1 °C. Record all pH values to at least 0.01 pH unit. Make all measurements in a quiescent solution. Fully document all calibration and control standards.

a

Calibration Sequence

3) Rinse the electrodes thoroughly with distilled water (ASTM Type II or better). Remove drops of water on the electrode by blotting gently (Do Not Rub!) with a clean lab tissue. 4) Insert the electrodes into a clean beaker (10-20 mL capacity) containing a portion (10-20 mL) of the rainwater control standard (e.g., SRM 2694-I).'Be certain that the reference junction and glass bulb are completely immersed. Do not insert the electrodes directly into the polyethylene bottles. 5) Stir or swirl the solution to ensure homogeneity and contact with the electrodes. 6) Allow the solution to settle to a quiescent state (approx. 30 seconds). Record the pH after the reading has stabilized." 7) Discard this portion of the control standard. Do not use for subsequent control checks or for other analytical determinations such as specific conductance, anions, cations, and acidity. 8) Repeat steps 3 and 7 with a second rainwater control standard (e.g., SRM 2694-II).” 9) Calculate the differences between the true pH values of the rainwater control standard and the values as determined by the pH measurement system. Average the differences and apply this bias correction to subsequent rainwater measurements. (For example, if the pH measurement system displays the pH of the control

1) Standardize the pH electrodes and meter using SRM 185f, Potassium Hydrogen Phthalate (pH(S) 4.006 at 25.0 °C, 0.05 molal], or equivalent.' Refer to ASTM D1293, “Standard Test Methods for pH of Water" for guidance. Record the value. Rinse the electrodes with distilled water (ASTM Type II or better). 2) With the slope adjustment of the meter set at 100 percent, and the temperature adjustment set at the temperature of the buffer solution, check the Nernstian response of the pH measurement system with a second buffer, SRM 1861c/18611c, Potassium Dihydrogen Phosphate/Disodium Hydrogen Phosphate pH(S) 6.863 at 25 °C), or equivalent.' Refer to ASTM D1293 for guidance. If the reading for the second buffer is not within 0.03 pH units of the prescribed value, recheck the calibration of the system.? DO NOT CONTINUE until the conditions for calibration and Nernstian response have been satisfied. If the reading for the second buffer is within 0.03 pH units of the prescribed value, record the value and continue.

4

Storage of Electrodes

0.13 pH units lower than the true value, add 0.13 pH units to the subsequent pH measurements of rainwater.)

a

Rainwater Measurement Sequence

18) When not in use, soak the electrodes in a solution which is 0.1 mol/L potassium chloride and 1x 10-4 mol/L hydrochloric acid. Do not store the electrodes in buffers, concentrated acids, concentrated potassium chloride, basic solutions, or distilled water. Do not allow the electrodes to dry out. 19) Use these electrodes exclusively for rainwater measurements.

1

2

10) Rinse the electrodes thoroughly with distilled water (ASTM Type II or better). Remove drops of water on the electrode by blotting gently (Do Not Rub!) with a clean lab tissue. 11) Insert the electrodes into a clean beaker containing a portion (10-20 mL) of the rainwater sample. Be certain that the reference junction and glass bulb are completely immersed. 12) Stir or swirl the solution to ensure homogeneity and contact with the electrodes. 13) Allow the solution to settle to a quiescent state (approx. 30 seconds). Record the pH after the reading has stabilized." 14) Apply the bias correction as determined in step 9 and report this corrected value as the pH of the rainwater sample. 15) Discard this portion of the rainwater sample. Do not use it in other analytical tests. 16) Repeat steps 10 through 15 for subsequent rainwater samples. 17) Repeat the Control Sequence at regular intervals, based upon quality control guidelines, performance history of the measurement system, frequency of measurements, and required accuracy.

For the highest accuracy and the most direct traceability to the National Bureau of Standards, it is recommended that freshly prepared solutions of Standard Reference Materials (SRM's) be used. ? Possible causes for deviation from Nernstian response include:

i) improper calibration ii) old or contaminated buffer solutions, iii) insufficient rinsing of the electrodes between solutions, iv) plugged reference junction, v) defective electrodes.

vi) defective meter. It may be necessary to replace the electrodes with a new pair if the non-Nernstian behavior persists even with accurately and freshly prepared buffer standards.

3 Standard Reference Materials 2694, Simulated Rainwater, is issued by the National Bureau of Standards, Office of Standard Reference Materials.

Some systems may require five minutes or more to stabilize. If drifting persists, record the reading after 10 minutes and annotate the data accordingly. With such severe drift, it would be advisable to acquire a different type of pH electrodes which do not exhibit this adverse characteristic.

S The difference should agree to within 0.05 pH units. If not, repeat the Control Sequence.

4

« PreviousContinue »