By Dan Kroll

Summary: Accurate measurement of trace metals in drinking water has always posed serious analytical problems. Current methods rely on expensive apparatus and dangerous reagents. Increased interest in analysis of silver for its use as a treatment chemical is due to the need for accurate measurement to ensure proper disinfection levels. Recommendations vary from 10-40 mg/L. Development of a field testing kit for silver at these levels is explored here.

Silver has a long history of use as a disinfection agent. The ancient Greeks and Egyptians were known to store water supplies in silver vessels to limit contamination. Many early medical treatments for disease (some still in use today) incorporated silver salts. The Soviet space program utilized a silver-based disinfection system to sanitize drinking water supplies.

Benefits and risks
In low concentrations, silver’s antibiotic properties make it desirable for use as a fungicide and for drinking water disinfection purposes. Many studies have shown its effectiveness as a secondary disinfection agent for at-risk populations, such as in hospitals, nursing homes and clinics. Its effectiveness against Legionella has made it especially popular in these applications. Due to the move away from chlorine-based products, silver has also been gaining in popularity as a pool and spa biocide.

According to the World Health Organization (WHO), however, continuous exposure to silver in drinking water—0.4 milligrams (mg) or more—in humans causes argyria, an irreversible condition that produces a bluish-gray discoloration of the skin, hair, nails and eyes.4 Long-term, continuous exposure to silver has also been implicated in liver damage and enzyme inactivation in humans.4 The WHO hasn’t yet set limits for safe silver concentrations in drinking water.5 The USEPA adopted the Public Health Service (PHS) standard where silver in domestic water not exceed 50 micrograms per liter (mg/L),3 which was set to protect aquatic life and human health. Canada adopted a similar 50 mg/L standard while the European Union standard is 10 mg/L.5

Assessing silver
There are many current methods to measure silver in the mg/L ranges. These include atomic adsorption (AA) by flame or electrothermal techniques, inductively coupled plasma (ICP) or colorimetry. Each method requires a complicated and expensive apparatus, hazardous chemicals and/or a large investment in time and equipment. Each also has its drawbacks. AA is accurate at moderate concentrations, but displays sensitivity to ion interference. ICP techniques have higher minimum detection limits and are sensitive to refractory elements. Colorimetry loses sensitivity at these ranges and uses potentially hazardous chemicals.6

There’s a need for a portable, rapid, easy and safe screening method for low levels of silver. The method described in this article meets all of these criteria and requires less time than conventional methods. Users obtain an on-site reading that only takes minutes to gather. The procedure features a simple visual comparison for obtaining semi-quantitative measurements for silver between 0-50 mg/L. Single-reagent addition and quick results make this method ideal for field testing.

Method and chemistry
This test makes use of the fact that, under certain specified conditions, metals form brightly colored precipitates with certain dyes or mixtures of dyes. In the case of silver, the complex formation is with 1-10 Phenanthroline and Bromopyrogallol Red. This chemistry is an adaptation of a reaction described in 1964.1 In this reaction, each silver (Ag) ion first reacts with two 1-10 Phenan-throline molecules to form a colorless complex. Two molecules of this complex then react with a molecule of the Bromo-pyrogallol Red to form a blue precipitate (see Figure 1).

Researchers in the 1964 report made use of this system for the spectrophotometric detection of silver with a lower detection limit of 20 micrograms per liter (µg/L), or parts per billion (ppb). The system utilized in this new kit follows this chemistry until the method of detection. The blue complex is an insoluble precipitate in an aqueous sample.

A reagent powder is prepared that contains a sodium citrate-citric acid buffer system with a pH of 7. The powder also contains Tetrasodium EDTA as a masking reagent to remove interference from other metals. An excess of 1-10 Phenanthroline is added to make sure there’s more than enough present to complex the silver along with any iron that may be present. Finally, the reagent powder contains the Bromopyrogallol Red. The final reagent powder is then packaged in unit dose form at the amount needed to react with a 100 milliliter (ml) sample.

After the initial chemical reaction is carried out, 100 ml of the reacted sample is filtered and forced with a syringe through a nitrocellulose microporous filter. The blue precipitate is trapped on the filter. The intensity and hue of the filter are dependent on a quantitative manner upon the original concentration of silver present in the sample. The colored filter is then compared to a color-matching chart with different shades of blue corresponding to different concentrations of silver.

By manipulation of the dye concentrations, filter size, pore size and sample volume, the levels to which the test can be made effective can be altered. Using a 100-ml sample, 5 millimeter (mm) diameter filter size (a Gelman #4317 13-mm plastic filter holder is modified with washers to expose a 5-mm surface area on the filter) and 12-µm pore size, visual levels of detection down to 5 ppb Ag can be achieved. The final color-coding chart for silver has gradations of 0, 5, 10, 25 and 50 ppb Ag (see Figure 2). This method has been validated on a number of different water matrices using NIST standard spikes (SRM3151) and compared to AA (Varian SPECTRAA 20-plus) for verification. This method allows visual detection of silver in ranges comparable to, and in some cases lower than, detection levels possible with more costly and time consuming methods such as AA.

Interference studies were conducted by preparing a known silver solution (about 10 µg/L) and the potential interfering ion. Positive interference was tested by running blanks of deionized water that contained the potential interfering ion. The ion was said to interfere when the resulting change threw off the color match by more than a half step on the chart. The following substances, at the stated levels, show no interference on a 10-ppb Ag standard test. There should be no problem with color matching or bad blanks at these levels. These tests were run at the levels indicated, although greater concentrations may be tolerated.

Turbid samples should be pre-filtered through a glass fiber filter. Oils and surfactants may interfere by preventing formation of the insoluble complex; however, appropriate digestion may eradicate some of these interferences.
Materials that weren’t tested, and according to the literature also interfere, are uranium (VI), thorium (IV) and niobium (V). They all form blue-colored complexes with Bromopyrogallol Red. They can be masked at a tenfold excess over silver by the addition of excess fluoride for uranium and thorium, and of hydrogen peroxide for niobium.1

In a single laboratory, using a standard solution of 10 µg/L silver and three representative lots of reagents, a single operator performing 50 tests per lot obtained no results that weren’t properly matched to the 10 µg/L color dot within a half step. In other words, all results were within a range of 7.5 to 17.5 µg/L. These results were compared to results obtained on an AA (Varian SPECTRAA 20-plus). The results for the AA on 30 repetitions using the same standard gave an average of 6 µg/L with a standard deviation of 4.98 µg/L. The recommended lower level of detection for the AA was 20 µg/L.

Assuming a worst case scenario where the standard deviation for the visual test method is 7.5 µg/L, statistical analysis shows that in the case where amounts of silver to be detected are below the recommended level of detection for the AA, the visual method outperforms the AA.

Performance on environmental samples
Samples from a hot tub utilizing bromine as a disinfectant, tap water as well as pool water from Ames, Iowa, were spiked at a level of 10 µg/L silver. Tests run in the field using the visual method resulted in 100 percent recovery on these spiked samples. All samples matched the 10 µg/L spot.

Samples run on sewage effluent from the Ames, Iowa, sewer plant and surface water from the Skunk River near Ames resulted in poor or no recovery. After a mild digestion, 100 percent recovery was achieved on these samples. Moreover, samples that contain large quantities of organic material, thiosulfate or cyanide should be digested before testing.

The method described was utilized to test various water samples for the presence of Ag. All of the matrices that were examined gave acceptable results. The results were comparable to those obtained with atomic adsorption (AA) methods; and, in the case of levels of 10 ppb and below, the visual method was found to be superior to AA. This method should be useful as a field method for determining trace amounts of silver in a variety of environmental samples. Use of this method should save time and money without compromising analysis accuracy.


  1. Dagnall, R.M., and T.S. West, “A selective and sensitive colour reaction for silver,” Talanta, Vol 11, pp. 1533-1541, 1964.
  2. Digestion and Analysis of Wastewater, Solids, and Sludges, Hach Company. Loveland, Colo., 1987.
  3. Guidelines for Drinking Water Quality, Vol. 1: Recommendations, World Health Organization, 1984.
  4. Guidelines for Drinking Water Quality, Vol. 2: Health Criteria and Supporting Information, WHO, 1984.
  5. Hach Water Analysis Handbook, 2nd edition, Hach Company, Loveland, Colo., 1992.
  6. Standard Methods for Examination of Water and Wastewater, 18th edition, American Public Health Association, 1992.
  7. Title 40 CFR Section 261.24, Code of Federal Regulations, July 1, 1996 edition.

About the author
Dan Kroll is a senior research and development chemist for Hach Company, a leader in development of analytical systems for water and wastewater. With 13 years at Hach, Kroll has bachelor’s degrees in microbiology and genetics and a master’s degree in water resource management and environmental engineering from Iowa State University. He can be reached at (970) 663-1377 ext. 2637, or email:


Comments are closed.