Abstract The unexpected and sporadic detection of cyanide in sour gas plant wastewaters and stack test scrubber solutions is investigated. It is shown that hydrogen sulphide and other reduced sulphur compounds are capable of producing a false positive in the standard method for cyanide analysis. The interference is unusual in that the response declines with increasing concentration of interferant. Introduction The pyridine-barbituric acid method for cyanide analysis in industrial wastewaters(1) is used widely, and is often a method recommended by statutory authorities where the wastewater is discharged to a natural watercourse. The limit of detection is commonly cited as 1 μg.dm−3. We have frequently seen reports of positive tests for cyanide by this method in sour gas plant wastewaters or stack test scrubber solutions. Such results are usually difficult to explain in terms of the processes involved, especially in the case of incinerator stack gases, where high temperatures (of the order of 500?C) prevail in the presence of excess oxygen. In the following, data are presented which suggest that the method may be limited to determining concentrations of CN-well above 40 μg.dm−3 in the presence of certain sulphur compounds, even when steps to remove interference from inorganic sulphides have been taken(1). A schematic of the reaction sequence is shown in Figure 1; accordingly, no positive interference is expected unless the interfering substance is capable of producing a compound analogous to the quaternized pyridine ring [A], which must also cleave to give the dye [C] with barbituric acid. In spite of this, the apparent detection of cyanidein sour gas plant incinerator stack gases led to the suspicion hat some sulphur compounds might give rise to an interference. Experimental In the present experiments, solutions of 5% and 10% w/v NaOH (ACS Reagent Grade, American Scientific & Chemical Co.) were prepared in deionized water and were spiked in two or more replicates with COS, H2S, CS, and ethyl methyl sulphide. In the case of gases, the material was added slowly by means of a syringe, such that no visible gas bubbles escaped the solution. The liquids were added to the agitated aqueous solution by means of a microsyringe. No phase separation was to be expected at the concentrations used, and none was observed. The composition of CO2, H2S (Matheson Gas Products, 99.5%) CS2 (J.T. Baker Analyzed Reagent, 99+ %) and methyl ethyl sulphide (Canadian Western Natural Gas Co. Ltd., 98%) were known to be essentially free of HCN, such that a 1 μg.dm−3 concentration of CN− in solution could not be reached in this experiment. All solutions were stored in acid washed, deionized water rinsed amber glass bottles, and together with NaOH blanks, numbered at random. The numbering sequence was withheld from the analyst, who followed the "Standard Methods"(1) pyridine-barbituric acid procedure for cyanide; the method was amended in accordance with EPA method 335.3(2), to permit the use of a Technicon auto-analyzer. The V.V. lamp was shut off, and a distillation device described by Conetta et al. (3) was incorporated into the circuit.
Read full abstract