In Reply Refer to: September 19, 1980 EGS-Mail Stop 412 QUALITY OF WATER BRANCH TECHNICAL MEMORANDUM NO. 80.26 Subject: WATER QUALITY--Preservation of nutrient samples by addition of mercuric chloride. Beginning October 1, 1980, water samples submitted to the National Water Quality Laboratories for analysis of nitrogen and phosphorus compounds in filtered and unfiltered samples must contain 40 mg/L mercuric ion (Hg2+) added as mercuric chloride (HgC12) and be chilled to 4 degrees C. Samples should not be frozen. Tablets containing about 10 mg Hg2+ as HgC12 in a NaCl base have been prepared and will be distributed to the District Offices. Instructions for obtaining the HgC12 tablets and amber bottles for nutrient samples are given in Central Laboratory Newsletter No. 80.02 which was published in late August or early September. One (l) tablet is to be added to each FC and RC bottle submitted for nutrient analysis. Sample volume must be at least 200 mL and, ideally, 250 mL in order for the NaCl concentration to be within critical limits for analysis. After October 1, 1980, the nitrogen and phosphorus analyses at the Central Laboratories will be calibrated using standard solutions containing 40 mg/L Hg added as HgC12. The analysis of a sample submitted without HgC12 added or with an Hg2+ concentration outside the range 20-60 mg/L will have an indeterminate error. The lab should be informed in advance of samples requiring special handling; if not, the analyses may be delayed. A surcharge will be applied to the analytical cost and billed to a district account to cover special handling for projects that need to arrange "special" sample handling without HgC12 tablets. "Specials" must be arranged prior to sampling. Explanation The rationale behind the decision to use HgC12 and chilling for nutrient samples preservation is summarized below. We invite you to read the briefing paper and annotated bibliography for further detail. Comments and new information are welcomed. On the average, water samples are in-transit and awaiting analysis for at least 3-5 days, and often longer, from time of collection. In order for the analytical results to fairly represent concentrations of nitrogen and phosphorus compounds in the water at the time of collection, chemical and biological activity in the sample must be stopped for the period between time of collection and time of analysis. Studies performed over the past 10 years by both Survey and academic researchers suggest that chilling alone, the only preservation techniques now used by WRD, is not reliable. That is to say the method preserves some samples adequately but not all. Moreover, it is not possible to determine under what circumstances the method is adequate and under what circumstances it is not. These same studies also suggest that there are no completely reliable techniques for preserving nitrogen and phosphorus concentrations in water samples for 3-5 days. But reviews of the state-of-the-art in 1972 and 1976 both concluded that chilling plus addition of a biocide is a more reliable preservation techniques than is chilling alone. The best biocide among nearly 10 is HgC12, which appears to be effective in most samples studied at concentrations of 25-40 mg/L Hg2+ ion. Some studies indicate that 40 mg/L Hg2+ is not completely effective in samples having organic carbon concentrations of 20 mg/L or more, a concentration encountered in only 7 percent of NASQAN sample analyses since 1972. Freezing and addition of HgC12 was found to be a more reliable method than chilling and addition of HgC12 in some studies but, in freezing, there exists a risk that ortho-phosphates will coprecipitate with calcite or amorphous silica causing irreversible losses of ortho-phosphate from solution. This risk is judged to be unacceptable. For this reason, freezing is discouraged as a routine preservation method. Potential for Cross-Contamination Concern has been expressed by WRD personnel that routine use of HgC12 will inevitably lead to dispersion of mercury in field vehicles, filtration equipment, and shipping containers, thereby greatly increasing the potential for contamination of the FA and RA samples with mercury. The possibility definitely exists but can be minimized by the following procedures which should be adopted by all field personnel: 1. The tablets, which are hard and individually packaged, should be kept in a closed, metal container where they cannot be crushed or abraded. The package of a crushed or noticeably abraded tablet should be discarded unopened in a safe place outside of the field vehicle or service laboratory. 2. When not in use, tablets should be kept in a cool place. Mercuric chloride has a slight vapor pressure at room temperature. Keeping the tablets cool will reduce the vapor pressure and extend their effective life. 3. The bare tablets should not be handled. The package can be opened and the tablet dropped into the sample bottles without being touched. 4. When FA or RA and FC or RC samples are being prepared at the same site, nitric acid should be added to the FA or RA sample first, hands thoroughly washed with clean water to remove traces of nitrlc acid, then add a HgC12 tablet added to the FC or RC sample. Finally the hands should be thoroughly washed again. 5. Caution: A sample to which HgC12 has been added should never be filtered. Doing so may contaminate the filtration equipment so completely that it cannot be cleaned satisfactorily and must be discarded. Clearly great care must be exercised in the field to avoid cross- contamination. Each HgC12 tablet contains only 10 mg of mercury. But ambient mercury concentrations in water are usually less than 10 ug/L, and contamination at the 0.1 ug/L level would substantially change the mercury concentration of most samples. Recent literature has suggested that metallic mercury vapor can pass through the walls of polycarbonate bottles and contaminate samples. This finding raises the question of contamination of samples stored in field vehicles or gage houses from mercury used to service manometers or spilled from broken themometers. The Quality of Water Branch is continuing to study the issue and will report findings at a later date. The Branch will also do periodically check the mercury concentration reported at NASQAN stations for signs of mercury contamination. Safety Mercuric chloride is extremely toxic to animals. The Merck Index reports the LD50 dose to rats as 37 mg/kg body weight. One or two grams ingested orally--equivalent to about 75 to 150 tablets--is often fatal to humans. Other mercuric compounds cause inflamation of skin and mucous membranes. There is also evidence that mercury accumulates in the body. The mercuric chloride tablets must be treated with respect. The best safety measure to follow is to avoid contact. Following the practices and precautions listed above will help prevent contamination of acidified samples and will also ensure a good margin of safety for all who use the tablets. The health hazard to workers using the tablets is so small that they meet or exceed all OSHA and pharmaceutical standards. The comfortable margin of safety is, however, no license for sloppy technique. R. J. Pickering Chief, Quality of Water Branch Attachment: Briefing Paper Memoranda superseded: QW Branch Technical Memos 69.07, 70.02, 72.09 Keywords: Water quality, sampling, nutrients, preservation WRD Distribution: A, B, S, FO, PO