1992-05-04_REVISION - M1988112 (2)

Home Browse Search

~i,6IE5: IhJ~. TEL: ~~_~--~ _•_--~°,-- ~: ~.. ~;a ';~ 507 hJo .~:~t=~4 F.07 • _~.•~. `//)L 99. 40. 7, AP.4 ]! '99t aei rfl.,In l'. Coha•:1nr nr Fii v,.r dnr:ng Ana:;; ale een:ple Ag conch, pg/L unr.: nteu dampie tiico:ec stabui~ru' tree cyanide CNATC total cyanide AgNOr 60 tit 0 0 0 ~ 0 AgNO~ 60 Ag(CN)•i 60 l07 46 61 ti 49 °Filwrcd with 0.46•pm pore size filter, alahili:ed at pH = 12. Tahlo Vl . Sample Composition of Analyte Mixtures composition, ug/L [defined cetogorlcal cnmpaaitlnnJ chro.nium zinc silver cyanide ferrous fnaic cobalt cyanide cyanide cyanide [weak 50%j eyanido q'anide Cyanide thio- sample [Free 100%j [free 100%j [week 100%j [CNATC 6D%] [CNATC 100°!0] [CNATC 100%] [CNATC 100%J cyanate ] 40 200 0 0 400 100 0 0 2 40 40 40 10 40 0 40 40 3 0 0 20 0 20 100 20 0 4 200 400 0 0 0 0 200 400 6 d0 40 20 0 40 100 0 0 a o o D o zoo loo 200 0 7 40 20 0 20 40 0 40 0 8 100 0 0 0 100 100 0 400 quired for the CNATC and total cyanide procedures, einco silver liberated during those procedures is eliminated through the inherent precipitation and filtration steps already in• corporated in those procedures. Table V illustrates these points. Analysis of 8ampie Miaturos. In order to study the recovery of various cyanide species using the silver probe reaction procedures, several synthetic samples were prapeted. Those samples, containing both free end motel complexes of cyanide, were prepared in a distilled water matrix. The composition of the samples ie listed In Table VI. 1'or each sample, a calculated concentration for free, CNATC, end total tyanido has been determined with the categorical definitions. Tahlo VII lisle these values, along with the concentrations measured by the throe categorical procedures. The results indicate thgt, even in samples of varied eompoaition, the cnncontretion of each categorical division can be determined with excellent agreement to defined concentrations. Of the three categories, the best accuracy w•ae found with the total cyanide division. which has a relative standard deviation (RSD) of 0.04. This ecatracy iq similar to that of direct atomic absorption analyses, indicating only minimal error is intra• doted by the indirect analysis. Both the free cyanide and CNATC dotormirtatione have RSll's of apprnrimetoly 0.10. The froo cyanide measurement is slightly elevated. while the CNATC measurement has a slight negative deviation from the defined concentration The bias found in the CNATC measuremanty can be at least partially attributed to the presence of free cyanide impurities in the metal Cyanide so- lotions. These ate present both ae reagent Impurities qnd as pltotodiasociatinn products from laboratory lighting. Such fl conclusion is supported by rho reaulte. Such impurities would, of course. elevate the froo c}'anide concentration while reducing the CNATC measurement by allowing additional cyanide to be removed through oxidation. Such errors are, however, only minor, as indicated by the overall RSD of OA82 for all three cntegories. Limi[s oC Detection. Given the al:propriate conditions, the Ax fi'.ter provides complete reaction .af cyanide snalytes v:ui: via ~ih•er metal. This is ,nee fur all three cat.egarieel .~m~~caures. As the silverc}'anide reaction indieatre, there s over e 2:1 mass increae when the _iller atom, rather thou Table VII, CateQOrical Cyaalde De rmiaatlon of Aaalyte Ml:tare. free cyanide Weal cyanide eonen° CNATC ncn° mncn° sample defined meaed define defined mewed 1 0.240 0.263 0.600 .473 0.740 0.749 2 0.080 0.078 0.100 .09d Q210 0.22A 9 0.000 0.002 0.110 .127 0.180 0.149 4 0.600 0.561 0.200 .149 0.800 0.800 6 0.080 0.082 0.140 .144 0.240 0.231 8 0.000 0.029 O.SOO .443 0.600 0.482 7 0,060 0.006 0.090 .081 0.180 0.108 8 0.100 0.119 0.200 .170 0.300 0.306 •Alt concentrations io ug/mL. ~ - cyanide ion, becomes the analytic species. This is coupled with the fact that silver ie one element{ which has an ertremoly low 1[Init o[ detection by atomic abso pt[on spectroscopy. AB a result, the detection limit for the sr ver AAS system, for all species of cyanide, using flame atom' ion AAS u 0.002 mg/L $y using carbon rod atomization, t at limit is reduced to 0.00006 mg/L. Both of these levels are well below those of standard cyanide measurement methods and below the EPA water quality criterion. CONCLU3101'f8 In tltie paper, what might be considered classical reaction chemistry has been utilized in a unigyyse system to provide the basis of en ultratrace cyanide detection method. 3traight- forwprd reaction kinetics and well-Qe[ined photochemical ligend dissociation reactions have be<n utilized to provide a means of species differentiation. The result ie a sitnPle analytical system that can ~nteriaco to any atomic absorption epectromnter to provide nuantitativo date ou concentrations of cyanide ln'v ~-_c~'•, •.,~~ ~~: t:: ef"' 1'he c rrent ;yytem. costing less ~n :..--'- . -.- , ,_. __ ~ .-eGc~re? '.. _ _ _ single cyanirie i ..:.. .. .... . ._ ^::ntle :upem,;w: and can eatiii': ~,. _ .. ._ ....., .. , ..f yuirkiv ~- ptu.: ~ . - ~ .. ...