Initially we tried this standard method with no dilution (or lesser dilution for >100 uM Si samples). Humics interfered with recovery of silicate. A color blank using appropriately diluted sulfuric acid solution without molybdate (APHA Standard methods..., 1981) could not completely correct for the interference: "corrected" recoveries of Na₂SiF₆ added to humic- colored water were low.

Sample Volume:

3 mL

Sample Collection:

Filter through 0.6 um polycarbonate filter (Poretics) in plastic apparatus, used only for similar samples.

Sample Storage:

in polyethylene bottle at 4 ^oC

4.5 M H₂SO₄:

Carefully (slowly and with stirring) add 250 mL concentrated (sp. gr. = 1.84 g cm^-³ H₂SO₄ to 750 mL water. Cool and dilute to 1 L with water (use plastic graduated cylinder rather than volumetric glassware).

Acid Molybdate:

Dissolve 38 g (NH₄)₆Mo₇O₂₄.4H₂O in 300 mL water. Add this to 300 mL 4.5 M H₂SO₄. DO NOT ADD ACID TO MOLYBDATE. Use an all-plastic automatic dispenser. Store in dark.

Saturated Oxalic Acid Solution:

Dissolve 80 g (COOH)₂.2H₂O in 800 mL water. Use an all-plastic automatic dispenser.

Ascorbic Acid:

Dissolve 2.8 g C₆H₈O₆ in 100 mL water. Use an automatic dispenser (may be brown glass). Store in a dark bottle in refrigerator.

10 mM Primary:

Dry Na₂SiF₆ overnight at 105 ^oC. Dissolve 0.9403 g in about 100 mL water in a plastic beaker, warming if necessary, and quickly make up to 500 mL with water in a glass volumetric flask. Store in plastic bottle.

1 mM Secondary:

0.5 mL Primary Standard + 4.5 mL water, or similar convenient amount. Make fresh each day.

Working:

Make fresh each day:

10 uM: 4.950 mL water + 50 uL Secondary Std.

20 uM: 4.900 mL water + 100 uL Secondary Std.

40 uM: 4.800 mL water + 200 uL Secondary Std.

Spectrophotometer.

Plastic test tubes (15 mL centrifuge tubes are fine), beakers, graduated cylinders, and bottles for making and storing reagents and performing analysis.

Koroleff (1983) method but dilute samples considerably before analysis, for several reasons. Silicate in Georgia rivers is very high, often above the range in which Beer's Law applies for this analysis (which is very sensitive). Salinity requires a correction, but this can be avoided by dilution with fresh water. Humic material interferes with the analysis, but this can also be avoided by dilution. Dissolved inorganic silicate is determined.

Make up triplicate blanks and working standards in tubes, 5 mL each. Pipette 0.5 mL (if salinity is < 25 ppt) or 1 mL (if salinity is >= 25 ppt) of each sample into 3 tubes. Make up to 5 mL with water. Add 200 uL of acid molybdate to each tube and vortex. Let stand for 20 min, then add 200 uL oxalic acid, vortex, immediately add 100 uL ascorbic acid, and vortex. (Try to be very consistent in timing the reagent additions to tubes in racks so that standing time between acid molybdate and oxalic acid is the same for EACH TUBE.) Keep tubes in a dark place until read at least 3 hours but not more than 6 hours later.

Read the absorbance in a 1 cm cell at 810 nm (against the absorbance of water), using a sample sipper or cuvette. Settings for the sample sipper for the Shimadzu Model UV160U spectrophotometer are pump speed medium, sip time 4 sec, dwell time 1 sec, purge time 4 sec, 1 rinse. If using a cuvette, rinse once with sample.

Tubes may be simply rinsed 4 times with 18 megohm-cm water and allowed to dry between uses.

Plot the standard concentration vs. absorbance. The slope of this line is the calibration factor F. F should be approximately 47 for a 1 cm cell and is very dependent on timing between reagent additions.

Correct sample calculations for dilution where necessary. Dilution factor = 10 if sample salinity was < 25 ppt (i.e. 1:10 dilution used) or Dilution factor = 5 if sample salinity was >= 25 ppt (i.e. 1:5 dilution used).

Sample concentration (uM) = F × (sample abs - average blank abs) × Dilution factor

With dilution of seawater samples according to this protocol, errors due to salinity are reduced to perhaps 2% at most (Koroleff, 1983) and therefore can be ignored. Color blanks for diluted samples (APHA Standard methods..., 1981), even those of originally high humic content, are zero and need not be run.

Limit of Detection:

With 18 megohm-cm water, blanks are all 0.000 so std. dev. = 0. With a minimum absorbance reading of 0.001, limit of detection = 0.05 uM if sample is not diluted, or 0.5 uM if sample is diluted 1:10.

Range:

100 uM if sample is not diluted (absorbance is near 1.9). Dilute appropriately above this range.

Accuracy:

95% confidence limits for prediction of a concentration near the mean for a typical standard curve (17.5 uM) approx. = ± 2.57 uM for a triplicate determination if sample is not diluted, or ± 25.7 uM for a sample near 175 uM which is diluted 1:10. (Confidence limits for prediction near the ends of a standard curve are, of course, somewhat larger.)

American Public Health Association Standard methods for the examination of

water and wastewater: 15th edition. 1981. American Public Health Association, Washington, D.C.

Koroleff, F. 1983. Determination of silicon. Pages 174-183 in Grasshoff, K., M.

Ehrhardt, and K. Kremling, editors. Methods of seawater analysis: second, revised and extended edition. Verlag Chemie, Weinheim.