A sample taken for chemical analysis is supposed to be a snapshot of the quantity of analyte present at the time the sample was taken. Reactions of the analyte with other constituents in the sample matrix cannot be allowed to proceed, nor can loss of the analyte by evaporation, precipitation, or oxidation. The act of adding a chemical and refrigeration is intended to preserve the analyte concentration. Unfortunately, it is possible that the very act of attempting to preserve an analyte concentration in the sample bottle actually makes matters worse. Depending on conditions and the sample matrix itself our attempts at stabilizing what we are looking for generates more of it. Sometimes we cause it to disappear. But even worse, sometimes we are killing it and creating it all at the same time within that one little bottle. It would not be so bad except for decisions are made on the measurements finally made. These decisions could result in fines for analyte concentrations that weren’t really there, or a false assurance that analyte was absent when it really wasn’t. Other problems arise on questions of compliance. In the regulated community not following sampling and preservation protocol means sample collection was not valid. Changing the sample collection procedure is only allowed if the laboratory has data to support the change, however, in most cases laboratories are unable to fully characterize the matrix of a sample prior to sample collection. We will look into the potential interferences that impact cyanide analysis and what can be done to minimize these impacts.
Purpose of Sample Preservation
Cyanide methods are developed as an attempt to measure various cyanide species. These species range from the most toxic free cyanide to the ultra conservative estimate of cyanide toxicity we know as total cyanide. Total cyanide measurements include free cyanide, available cyanide, and non toxic strong metal complexes. Also included in the definition of total cyanide are insoluble particulate or colloidal cyanide complexes. The method that the samples are being collected for needs to be known at the time of sample collection. In most instances total cyanide will be analyzed. We will be discussing protocols for sampling and preservation of cyanide defined in Part 136, however, since these potential interferences apply to all samples this discussion should be applicable to all intended uses of data and to all cyanide methods.
Oxidizers must be removed immediately since they rapidly react with CN decreasing its concentration. The presence of oxidizers is determined by starch iodide test strips, or using field portable DPD kits. Oxidizers must be removed prior to any pH adjustment or they will rapidly oxidize any free and most available cyanide present. Do not add a reducing agent unless oxidizers are detected, or known to be present. A literature search on the web of laboratory SOP%u2019s reveals that ascorbic acid and pH adjustment to 12 – 13 is the most commonly practiced preservation technique for cyanide samples. This is unfortunate because:
Ascorbic acid is a carbon source that can actually be a precursor for CN generation during storage.
Multiple holding time studies have demonstrated that samples containing CN and ascorbic acid rapidly lose CN upon storage at high pH.
Use of ascorbic acid combined with hydroxide both destroys cyanide and creates it.
If ascorbic acid is used for dechlorination the holding time is reduced to about 24 hours. For example, a synthetic sample prepared at 200 ppb CN with ascorbic acid added and the pH adjusted to 12 recovered less than 25 % of the original cyanide present after storage for 3 days. Again, not only does ascorbic acid cause cyanide to be lost, but it can cause it to be generated as well. Basically, the use of ascorbic acid to dechlorinate samples unless analysis is possible within 24 hours.
Sodium thiosulfate may be used to dechlorinate samples, however, it must not be added in excess. There are no spot tests available to estimate amount of thiosulfate present in a sample. Boiling hot sulfuric acid solution (cyanide distillation) containing thiosulfate generates colloidal sulfur and sulfur dioxide. Sulfur dioxide distills into the absorber solution. If the absorber solution is analyzed immediately, and chloramine T is doubled 80% recovery is possible. However, as solutions sit the SO2, now Sulfite, reacts in the basic solution with the NaCN oxidizing it to cyanate and lowering recovery. Therefore, if thiosulfate is suspected to be present, samples need to be analyzed as soon as possible after distillation, and the analyst needs to verify that the amount of chloramine T added is enough to guarantee a chlorine residual. This means that automated methods that use colorimetry (335.3 and Kelada 01) should not be used because there is no way the analyst can verify that enough chloramine T was added.
Sodium arsenite has been demonstrated as an effective preservative in most cases, however, a few studies have found slight false positives when combining sodium arsenite with distillation methods. Since sodium arsenite is an arsenic compound no one really wants to carry it around in the field, or be adding it to sample bottles.
Sodium borohydride is mentioned in the Kelada 01 method. There are legitimate concerns with its use since it generates hydrogen gas upon acidification. Anyone familiar with analyzing arsenic and selenium by hydride generation are familiar with this. Since distillations are taking place near a heat source the hydrogen generation could result in an explosion hazard. Also, rapid generation of hydrogen gas within a digestion vessel could result in exploding vessels. This is especially likely with a closed vessel such as the Lachat Microdist.
Samples that contain sulfide at concentrations above 50 ppm lose significant amounts of cyanide within 24 hours. Once Sulfide is reduced below 50 ppm the holding time can be extended. Even so, samples should be analyzed as soon as possible and preferably with a method that uses on-line sulfide abatement such as ASTM D6888-04 or OIA 1678. In fact, with slight reagent modification, ASTM D6888-04 and OIA 1678 can handle sulfide concentrations up to 200 ppm. However, remember that cyanide concentrations are rapidly depleting as cyanide remains in contact with sulfide. Basically, all methods listed in the CFR for sulfide removal just don%u2019t work. Headspace expelling and dynamic stripping leave residual sulfide behind, which interferes with distillation and analysis. The headspace and stripping methods are difficult to use and essentially require a mobile laboratory. pH adjustments, as well as flow rates, must be precise. Since the methods are volatilizing high levels of sulfide these procedures must be done under a hood, or with plenty of ventilation. Recall that these procedures are removing sulfide by generating hydrogen sulfide gas. Precipitation with cadmium in the presence of iron cyanide complexes forms a very stable and insoluble cadmium iron cyanide complex. When the cadmium sulfide is filtered off so is the iron cyanide. Mercury cyanide (a WAD, CATC, or Available cyanide species) is also lost by precipitation with cadmium. The precipitation with cadmium was put in place to replace precipitation with lead. This was because lead sulfide rapidly reacts with cyanide to form thiocyanate lowering the results. Other precipitants, such as Bismuth, also result in lower recoveries.
The only sulfide removal procedure that recovers CN quantitatively is dilution of the sample till sulfide is no longer detected by the lead acetate test strips. An argument against dilution is the increase in detection limit, however, most automated methods are sufficiently sensitive so that dilutions of up to 10 X still allow detection at about 5 ppb. Also, besides diluting sulfide other interferences are being diluted as well. Again, the only way to remove sulfide is to dilute the sample till sulfide is no longer detected on the lead acetate test strips. Then analyze the sample ASAP by a method that utilizes on-line sulfide abatement. If samples must be distilled, ASTM D7284 was developed specifically to handle samples containing sulfide. This method utilizes gas-diffusion amperometry as the measurement step after samples are distilled.
40 CFR Part 136 specifically states that if sulfite, thiosulfate, or thiocyanate are thought to be present to use a UV digestion method, or a non distillation gas-diffusion method. Many people have problems with this statement because at present there are no commercial suppliers of the Kelada, and OIA 1677 is not a total method. The intent of this statement was that since distilled colorimetrically determined cyanide results from samples that contain these substances cannot be trusted, an available cyanide result by method 1677 is a more accurate estimate of toxic cyanide than what is possible by distillation/colorimetry. The problem with naming the Kelada method is that it is a distillation/colorimetric method, and does not eliminate, or even significantly minimize, the interferences experienced because of these compounds. More so, since the Kelada is an automated colorimetric method any sulfur dioxide that distills into the absorber solution can react with cyanide forming cyanate and react with chloramine T increasing the chlorine demand. In effect, any samples that contain sulfite cannot be determined by the Kelada method. This is evidenced by a similar methods that state that sulfite concentrations higher than 1 mg/l interfere. Remember again that thiosulfate reacts under heated acid conditions (distillation) to elemental sulfur and sulfur dioxide. Since the Kelada method is a distillation method, this means that samples that contain thiosulfate cannot be analyzed. In fact the Kelada method says that thiosulfate was evaluated for oxidant removal and caused an interference with the method.
Thiocyanate in the presence of nitrate or nitrite reacts to generate cyanide. Sulfamic acid has been used to minimize this effect. Thiocyanate alone reacts with the small amounts of oxidant that form due to irradiation and generate cyanides as well. The Kelada method suggests an alkaline digest be used in the presence of thiocyanate to minimize degradation of thiocyanate to cyanide. However, if sulfite is also present in the sample, contact of sulfite and cyanide in alkaline solution rapidly oxidizes cyanide to cyanate.
Solutions and Conclusions
ASTM D 7284 analyzes total cyanide after distillation by gas-diffusion amperometry. The method has been validated by extensive single laboratory studies and has been evaluated for performance in the presence of multiple interferences. The method was developed specifically to overcome sulfide interferences with colorimetry, but in the process of evaluating interferences it was found to overcome sulfite and thiosulfate interferences as well. A modification of the currently published procedure eliminates, or at least significantly minimizes thiocyanate plus nitrate interferences. OIA 1678 is a UV irradiation gas-diffusion amperometry method. It differs from the Kelada and EPA 335.3 methods because no heat is necessary to separate CN from the acidified matrix. Earlier literature documents that automated distillation alone only liberates free cyanide and that in automated methods UV irradiation is needed to analyze total cyanide. OIA 1678 relies on gas diffusion instead of distillation to separate cyanide from the acidified matrix. Since OIA 1678 does not need heat the interferences are minimized.