l
<br />_ObIES Ii+U=. iEL:aCi'-~ti_-~~5-'
<br />r • ANALYTICPL GH.CM:STRY, t~03. l=J. i, dPr,`IL t t;c+
<br />Tsbiu f. Dependence of tae Siic:r r filter tteaction
<br />rrrieiency on Filler ?nro 91.e. Sample Flea Oat?, ~nri
<br />:• liter Thlcknese
<br /> % anah~te Ag tUter [h)cknesa
<br /> reacted with Ag (6µm pore size)
<br />Iluw rote? 5-µm O.a6•µm %rec. 9o rec, 96 rec,
<br />mLjmitr pore size pore sloe 40 µm 80 µm 120 µm
<br />6.9a Gl 91 61 66 67
<br />3.80 83 8^ 83 88 se
<br />3.02 88 93 B8 91 90
<br />2.47 97 99 97 y9 99
<br />1.75 101 102 l0l 99 101
<br />1.23 1Gl loo
<br />0.76 100 101
<br />'Analyse 1 mg/L CN', pH = 12.
<br />timize lira AA spectrometer by dimctly aspirating silvor standard
<br />solutions and adjustiug the wavelength, air•toacetylene ratio, slit
<br />width, and burner position to obtain maximum signal. Check all
<br />snmDlea for silver content by direct cotnparisan of signets from
<br />filtered (.0.45-µm) sample solutinna with silver standards, using
<br />distilled water as n blank. With the AA spectrometer optimized,
<br />place the Ag f (ter, photocell, and pump in•]ino as shown in Figtrre
<br />2. With rho pump rate sot at 10 mLjmin, rinse rho titter with
<br />pH = 12 prefillered distilled water until a zero (no silver) baseline
<br />is achieved. Next, bypaeaing the photocell, introduce several
<br />cyanide standards at 2.5 mLjmin, while rinsing with blank so-
<br />lution in between samples, until repxxiucible signals ere obtained,
<br />Always stop the pump when switching solutions to avoid intros
<br />ducitlg air 6uhbles into the system. With the photocell still
<br />bypassed, introduce sample until a steady signal is obtained.
<br />Comparison of the eemple signals with the signals ahtained from
<br />the cyanide standards will yield the cyanide mncentratiott of the
<br />eemple.
<br />Next, with eemple continuing to ba pumped, place the photocell
<br />in-line. Pump until s atendy baseline signal is obtained. Stop
<br />the satnltle flow and allow rho sample W be irradiated ter 31 min.
<br />liestart the pump and analyze the irsadiated sample. Follow thin
<br />sample with on aliquot of cyanide standard of an equivalent
<br />volume as the irradiated sample (the volume of the photocell),
<br />Comparison of the sample signals with aignala obtained from the
<br />cyanide standards determines the total cyanide conconvation.
<br />Determination of Cyanide Nat Amaeablo toChlor/nation.
<br />Place 40 mL of the preserved sample to he analyzed in a beaker
<br />on a magnetic stirrer and stir with a tritlttoroethyylane (TFEI
<br />coated stir bar, Ensure the pH is wall above 12 by diseohdng 0.2
<br />g of potassium hydroxide its the sample. Chlorinate fine sample
<br />by adding calcium hypoehlorite solution until potassium iodide
<br />starch paper sltuws azt excess o(chiorine. Maintain thin extese
<br />of chlorine and pH for ] h. Keep solutions protected from light.
<br />$liminote residual chlorine by adding 50-mg incremonta of as-
<br />corbic ocid, testing with Ki stardt paper after each addition. Allow
<br />solution to age ar least 1 h, as m+ ascorbic ocid-KOId precipitate
<br />forms. Filter this precipitate from the solution. Add 5 nrL of
<br />sndirw hypophosphite solution and miz well. Analyze the sample
<br />ac above for total cyanide.
<br />RESULTS AND DISCUSSION
<br />Detarminntlon of Frees Cyanide. 1'he ohility crt the de-
<br />scribed siher AAS system to detect cyanide is directly related
<br />to the efficiency of raactiott 2. Initially, free cyanide deter-
<br />minntiuns relied on the reaction of cyanide with narrow gauge
<br />silvor wire; however, this required lengthy convect times. Tha
<br />efficiency of reaction '? teas greatl? increased by providing a
<br />iar.er silver surface area to the c?snide. Passing the pnalvte
<br />^ ^i'?° :r,!1:13Gr. ti:tuugc s Filler eotn posed of pure silver
<br />..... .,.,~~, ,,-,e n•emrle to an enormous px~ess of silver. The
<br />.._ t :ne .^~tr .. r-,r . tvi:.`::he sik'er is A
<br />, •, .I,:• ~-.r.. ~ ... .i .4~ `~i:rr anri the flow rate
<br />-~ - ~~ . ~':nxs that filters
<br />..:,a Clore reaction of
<br />_ -, .,. ._~, .v,w er, at tluw rAl e9 of
<br />
<br />•
<br />..o,~ ;+~ a~.vo r.e yd, Dour°I~atio:.
<br />~., C:r ar.caa, mg, L amt t:V- ro•W:dP ~~ ~'-
<br />0.000 0.006 t 0.001
<br />D.O7G 0.015 t 0.0(u
<br />0.050 0.051 3 0.(Ul
<br />O.il6 0.111 t 0.002
<br />0.230 0'?30 t 0.017
<br />0.423 0.439 f O.Ol2
<br />0.820 0.819 t 0.029
<br />'n = 4 asmplos.
<br />Table III. Kinetic Equilibrium 1£f[eet: Stability and Timc
<br />Factors on the 911ver Probe Itesotlon
<br />4o cyanide
<br />converted m Ag(CN)=
<br />arability 5 2.5 0.5
<br />cnmpd cones mL(mip mL/min mL/min
<br />CN- 61 97 102
<br />2niCN)rr" 10" a8 90 100
<br />Cr(CNIa"- 10'1° 2 9 18
<br />FeICN)a ' lOrt 0 0 0
<br />FaICN)s°' 30°= 0 0 0
<br />2.5 mL/mht and below, the recovery for both idlers was nearly
<br />identical, producing complete silver-cyanide tromplexation.
<br />Uniika pure size, the thickness of t}Le number of stacked silvor
<br />filters has little to no effect on reec~tton efficiency, as the data
<br />in Tflblo I illusvate. The small, 0: ti-µm, pore siu fVtors have
<br />the advantage of near complete conversion of cyanide ion to
<br />the soluble dicyanosilver complex+even aL higher flow rotas.
<br />This advantage is, however, somhwltat offset by the high
<br />bubble pressure of these 151ters, The bubble pressure, o
<br />function of pore aize and lifter tlticknoss, is greater for the
<br />0.45-µm filters than the preaaures provided by CS'pical peris-
<br />taltic or syringe pumps, M a result, any gas bubble invoduced
<br />to Lhe filter cannot be passed, eattJting partial or even total
<br />lslockage of tits system. lluo to this pperational disadvantage,
<br />it was found that the 5•µm Clter (3p-50 µtn thick), operated
<br />at a flow rate of 2.5 mL/min, provided excellent cyanide
<br />recovery, while allowing for simplified system operation. The
<br />results ubtnhteci with these cohditimta tot free Cyanide samples
<br />ere shown in fable Il. These results exemplify the linear
<br />reeponso this system provides, even aver a wide range of
<br />cyenida concentrations.
<br />By using the optimized silver ftliler method, studies were
<br />carried out to investigate rho anon{cterietics of reaction 2.
<br />While this reaction is often cited wt h oxygen a9 rho driving
<br />mechanism of rho reaction, other reducible chemical species
<br />fmmd in aqueous systems may also fagilitate the sdver~yenido
<br />tenction. By spiking samples with cykrtide and then sparging
<br />the solution with either nitrogen, air1 or oxygen, it was found
<br />that reaction 2 p[uceoda equally wet! with air or oxygen sat-
<br />uration. However, satutatirnl with nitrogen completely re•
<br />pressed Lhe reactimt. This indicated Lhat oxygen does drive
<br />reaction 2 in these samples; theroforq, samples should eat be
<br />excluded from ambient air-orygen asttaation prior to analysis.
<br />The Ability of silvor to rntnpeta f¢r cyanide ions already
<br />bound in metal complexes is another portent characteristic
<br />relative m reaction 2. While the dicyanosilver Complex has
<br />a large fnrmatian constant, it is much less then that of many
<br />of the ether taetal cyanide complexes, Therefore, the ability
<br />of the sil•:eY.., lcquire cyanides tilrendy bound to other met,ds
<br />was investigyled. The large excess o silver provided by the
<br />silver iiiter ;.. sr~ •~ido a chemical environment :her is Ca-
<br />p~,ide ~:~r Cn ri• ,ne [he dicyrutu=_ilver product. By ir.trridueing
<br />du. rr:::r .:, ~' ,. ~::ide cc,u.plesaa to rho filter at various rates,
<br />[lie .~bili~,.:~.:i.e~silve; :U tumpala (oi Cho eyeltlde ligarids cmr
<br />
|