Water sample losses

In order to avoid loss of oxygen from the water sample it is fixed by its reaction with manganese)II) hydroxide which is converted rapidly and quantitatively to manganese)III) hydroxide ... 

The TIC trace from the LC-MS analysis of an extracted river water sample, spiked with 3 p.g dm of atrazine and three of its degradation products, is shown in Figure 3.30. The presence of significant levels of background makes confirmation of the presence of any materials related to atrazine very difficult. The TIC traces from the constant-neutral-loss scan for 42 Da and the precursor-ion scan for m/z 68 are shown in Figure 3.31 and allow the signals from the target compounds to be located readily. 

Another significant difference between large- and small-scale processing is dilution of the product samples with water. Food processing equipment for fluids often needs to be started with water, and the food needs to be flushed from the system with water before the process is shut down. When making small batch samples in this type of equipment, care must be taken to obtain a representative sample with a minimum of dilution. There will also be a difference in the weights of the sample into and out of the process due to water addition or sample loss to minimize dilution. 

The method using GC/MS with selected ion monitoring (SIM) in the electron ionization (El) mode can determine concentrations of alachlor, acetochlor, and metolachlor and other major corn herbicides in raw and finished surface water and groundwater samples. This GC/MS method eliminates interferences and provides similar sensitivity and superior specificity compared with conventional methods such as GC/ECD or GC/NPD, eliminating the need for a confirmatory method by collection of data on numerous ions simultaneously. If there are interferences with the quantitation ion, a confirmation ion is substituted for quantitation purposes. Deuterated analogs of each analyte may be used as internal standards, which compensate for matrix effects and allow for the correction of losses that occur during the analytical procedure. A known amount of the deuterium-labeled compound, which is an ideal internal standard because its chemical and physical properties are essentially identical with those of the unlabeled compound, is carried through the analytical procedure. SPE is required to concentrate the water samples before analysis to determine concentrations reliably at or below 0.05 qg (ppb) and to recover/extract the various analytes from the water samples into a suitable solvent for GC analysis. 

Analytical accuracy. The mixture of all deuterium-labeled internal standards is added to each water sample before extraction. This does not prevent the loss of the unlabeled herbicides from the sample in subsequent processing steps, but a proportional loss of the deuterated internal standard precludes the need to correct for recovery. Although referring to recovery in this type of analysis is inappropriate, the accuracy of this method should be monitored. 

Samples for mercury analysis should preferably be taken in pre-cleaned flasks. If, as required for the other ecotoxic heavy metals, polyethylene flasks are commonly used for sampling, then an aliquot of the collected water sample for the mercury determination has to be transferred as soon as possible into glass bottles, because mercury losses with time are to be expected in polyethylene bottles. 

A potentially more sensitive and selective approach involves reaction of formic acid with a reagent to form a chromophore or fluorophore, followed by chromatographic analysis. A wide variety of alkylating and silylating reagents have been used for this purpose. Two serious drawbacks to this approach are that inorganic salts and/or water interfere with the derivatisation reaction, and these reactions are generally not specific for formic acid or other carboxylic acids. These techniques are prone to errors from adsorption losses, contamination, and decomposition of the components of interest. Enzymic techniques, in contrast, are ideal for the analysis of non-saline water samples, since they are compatible with aqueous media and involve little or no chemical or physical alterations of the sample (e.g., pH, temperature). 

Sample preconcentration was performed by means of an automated on-line SPE sample processor Prospekt-2 (Spark Holland, Emmen, The Netherlands). Oasis HLB cartridges (Waters, Barcelona, Spain) were used to preconcentrate cannabi-noids present in the water samples whereas isolation of the rest of the compounds was done in PLRPs cartridges (Spark Holland). Before extraction, influent samples were diluted with HPLC water (1 9, v/v) to reduce matrix interferences and to fit some analyte concentrations, e.g., cocaine (CO) and benzoylecgonine (BE), within the linear calibration range. A sample volume of 5 mL was spiked with the internal standard mixture (at 20 ng/L) in order to correct for potential losses during the analytical procedure, as well as for matrix effects. Elution of the analytes to the LC system was done with the chromatographic mobile phase. 

To determine absorption rate, another group of animals was immersed in a 5 mg/1 solution of PNA in Instant OceanR (closed chamber) and the water sampled at intervals to measure the loss of radioactivity compared to controls (Fig. 3). 

When an aqueous solution of aldrin (0.07 M) in natural water samples collected from California and Hawaii were irradiated (7, <220 nm) for 36 h, 25% was photooxidized to dieldrin. By comparison, no loss was reported when aldrin in deionized water was subject to UV light for 10 h. 

Samples were prepared and analyzed as reported previously (18). Because of slow concentration decreases with time, low volatilization rates relative to hydrolysis rates in some cases, and small artificial losses of pesticide due to repeated water sampling, the most accurate method of determining volatilization rate constants was to divide the average pesticide concentration for that day into the average volatilization rate over the same period (Equation 1). Rate constants for the seven days were averaged. The entire experiment was performed in triplicate. 

The collection and preparation of water samples requires individual approaches for different analytical tasks. If heavy metals or long-lived radionuclides at the trace and ultratrace concentration range are to be determined in water samples by ICP-MS, especially careful sampling is necessary to avoid possible contamination (using clean bottles and containers washed and cleaned before use, for example, with 2 % nitric acid and high purity water to stabilize traces in the samples), and the loss of analyte by adsorption effects or precipitation should be also considered. 

Several investigations on the stability of chloramphenicol during thermal treatment have been also carried out. Chloramphenicol has been found quite stable to heating when added to water or milk. Even after 2 hours of boiling, no more than 8% loss of its amount could be observed by physicochemical methods of analysis (16). Application of microbiological methods has shown, however, a 33.3-35% reduction in the activity of chloramphenicol after heating the milk or water samples at 100 C for 30 min (3). 

Two water samples obtained at 5.7 m were filtered through membrane filters. Filtration through either 0.45- or 0.2- xm filters completely inhibited H202 decay over a 24-h period. At every sampling point along the transect an unfiltered and filtered (0.22- xm filter) sample was obtained. At all five locations the sample that had been passed through a 0.22- xm filter showed no loss of H202 over a 24-h period. 

Trace analysis has its special hazards for the unwary. The most important of these are loss of material in the analytical process and contamination by outside sources. Everyone realizes that trace constituents can be lost from samples, but few are aware of the many ways in which this can occur. For example, phosphate has been observed to disappear mysteriously from water samples in polyethylene bottles (10). Nitric acid, used to clean plastic vials, has been observed to convert these surfaces to ion exchangers, which readily take up as much as 10 12 moles per sq. cm. of trace metals (16). Lead nitrate solutions unless made distinctly acidic, plate out much of the lead on the walls of glass bottles. While everyone realizes that formation of a precipitate is liable to carry out trace constituents either by adsorption or occlusion, it is not as well-known that vanishingly small amounts of precipitates—amounts likely to be overlooked on casual observation—may also do this. The fly-ash and soot, which seem to be inescapable components of city air,... 

---