Flotation separation of cobalt and copper from fresh waters and their determination by electrothermal atomic absorption spectrometry

Ž .Microchemical Journal 65 2000 165]175

Flotation separation of cobalt and copper from freshwaters and their determination by electrothermal

atomic absorption spectrometry

ˇ UKatarina Cundeva, Trajce Stafilov , Gorica PavlovskaˇFaculty of Natural Sciences and Mathematics, Institute of Chemistry, St. Cyril and Methodius Uni ersity, Skopje, Macedonia

Received 28 January 2000; received in revised form 16 May 2000; accepted 17 May 2000

Abstract

A rapid flotation method for cobalt and copper separation and enrichment from fresh waters is established. At pH6.0, ionic strength of 0.02 molrl, using sodium dodecylsulfate as a foaming reagent, cobalt and copper were separated

Ž . y4 Ž y.simultaneously with 10 mg Fe III and 3=10 mol hexamethylenedithiocarbamate HMDTC added to 1 l ofŽ . Ž .aqueous solution. Iron III hexamethylenedithiocarbamate, Fe HMDTC , is used as a new additional flotation3

collector. The proposed procedure of preconcentration is applied before electrothermal atomic absorption spectro-Ž .metric ETAAS determination of these two analytes. The method is verified by comparison of the ETAAS results

Ž .with inductively coupled plasma-atomic emission spectrometric measurements ICP-AES . The limit of detection forcobalt is 0.0012 mgrl, while for copper it is 0.0120 mgrl. Q 2000 Elsevier Science B.V. All rights reserved.

Keywords: Cobalt; Copper; Electrothermal atomic absorption spectrometry; Inductively coupled plasma-atomic emission spec-trometry

1. Introduction

The role of trace heavy metals in animal andplant biological systems is very important, butvery complex. A lack of these microelements inan organism can cause many diseases and ill-

U Corresponding author. Tel.: q389-91-117-055; fax: q389-91-226-865.

ŽE-mail address: [email protected] T..Stafilov .

nesses, however, their sufficient quantities caninduce many harmful consequences, also. Themost frequently present microelements cobalt andcopper are introduced into human and animalbodies by food, water, air, etc. On the other hand,rain, snow, fertilizer and water of irrigation arethe most common routes for heavy metals intro-duced into plants. Because human beings andanimals cannot exist without drinking waterŽ .spring, well, tap, etc. , as well as plants withoutwater for irrigation, natural water has particular

0026-265Xr00r$ - see front matter Q 2000 Elsevier Science B.V. All rights reserved.Ž .PII: S 0 0 2 6 - 2 6 5 X 0 0 0 0 0 5 0 - 3

ˇ ( )K. Cunde¨a et al. r Microchemical Journal 65 2000 165]175166

importance. Therefore, the development of accu-rate and rapid determination methods for moni-toring the level of cobalt and copper concentra-tion in natural waters is necessary and indis-pensable.

Ž .Atomic absorption spectrometry AAS pro-vides accurate and rapid determination of manyheavy metals in natural waters. Nevertheless, veryfrequently for the extremely low concentration ofcobalt and copper in waters, a direct determina-tion cannot be applied without their previouspreconcentration and separation. Today, in ana-lytical practice very modern expensive separationinstrumentation is used, but many labs aroundthe word which cannot provide them, can applysome of the classical preconcentration methodssuch as solvent]solvent extraction, ion exchange,

w xevaporation and coprecipitation 1,2 . One veryfrequently used method for these purposes is themethod of coprecipitation by means of tetra- or

w xhexamethilenedithiocarbamate 1]13 . During thisenrichment method, analytes are incorporatedinto the structure of the collector precipitatewhose nature is colloidal. Next, the bulky colloidprecipitate is separated from the mother liquor byfiltration or centrifugation. After the dissolutionof the precipitate, analytes could be tested in thesolution by AAS. Nevertheless, there are manyinconveniences. The first is the separation of thecolloidal collector precipitate from the motherliquor by filtration or centrifugation which takesconsiderable time. The second disadvantage is thelimitation of the volume of the sample solutioninvestigated at approximately 0.5 l and conse-quently an achievement of a lower enrichmentfactor. These inconveniences can be overcome byreplacement of filtration or centrifugation byflotation. The major advantages of the flotationpreconcentration are the rapidity of the proce-dure and the excellent recoveries of the trace

w xelements investigated 1,2,14,15 . The necessaryequipment for flotation preconcentration is sim-ple and inexpensive. The use of a little amount ofsurfactant and tiny air bubbles necessary to per-form the proper flotation cannot permit seriouscontamination risks, which could be manifestedby the high blank values. One of the most impor-tant advantages of the flotation method is the

ability to analyze a larger volume of sample solu-tions and to obtain a greater preconcentrationfactor than the preconcentration as a conventio-nal carrier precipitation technique.

Traces of cobalt and copper in aqueous solu-tions can be collected and enriched by the methodof precipitate flotation using hydrated metal ox-

Ž . Ž .ides as hydrated iron III oxide Fe O ?xH O2 3 2w x Ž . Ž16]18 , hydrated aluminium III oxide Al O ?2 3

. w x Ž . ŽxH O 19 , hydrated zirconium IV oxide ZrO ?2 2

. w x Ž . ŽxH O 20 and hydrated indium III oxide In O2 2 3. w x?xH O 21 , were among the collectors first used2

for this kind of separation of cobalt and copper.w xIn our previous report 22 , cobalt and copper

Ž .traces where floated by hydrated iron III oxideŽ . Ž .Fe O ?xH O in a combination with iron III2 3 2

Ž .tetramethylenedithiocarbamate Fe TMDTC . In3Žthis work, the first collector is the same Fe O ?2 3

. Ž .xH O , but the second collector is Fe HMDTC .2 3w xThis well known chelating collector 5]13Ž .is formed in the tested solutions containing Fe III

ions by addition of hexamethyleneammonium]Žhexamethylenedithiocarbamate HMA]HMD-

.TC . The investigations have shown thatŽ .Fe HMDTC is a more appropriate reagent for3

Ž .cobalt and copper flotation, than Fe TMDTC 3w x23]25 .

2. Materials and methods

2.1. Apparatus

A Perkin-Elmer 1100 B atomic absorptionspectrometer was used for all AAS determina-tions. The instrument was equipped with agraphite-tube furnace HGA-700. Cobalt, copper,calcium and magnesium Perkin-Elmer hallowcathode lamps served as radiation sources. Theoptimal instrumental parameters for AAS de-termination of the analytes are given in Table 1.The introduction of the sample solutions into thegraphite furnace was performed by an Eppen-dorff micropipette. The volume of the introducedsolutions was between 20 and 50 ml. Inductivelycoupled plasma-atomic emission spectrometryŽ .ICP-AES was carried out by a Varian spec-

Ž .trometer Model Liberty 110 Table 2 . The elec-

ˇ ( )K. Cunde¨a et al. r Microchemical Journal 65 2000 165]175 167

Ž .trokinetic z potential of the surface particleswas determined electrophoretically by means of adevice analogous to Chaikovskii’s equipmentw x25]28 . The flotation was performed in a glass

Ž .cylinder 4=105 cm with a sintered glass discŽ .porosity No. 4 at its bottom to generate air

Žbubbles. A combined glass electrode Iskra, M. Ž .0101 and pH meter Iskra, M 5705 were used to

monitor the pH of the solutions during the copre-cipitation.

2.2. Reagents and standards

All chemicals used for the preparation of solu-Ž .tions were of an analytical-reagent grade Merck

excluding the surfactants sodium dodecylsulfateŽ . Ž .NaDDS , sodium oleate NaOL , sodium palmi-

Ž . Ž .tate NaPL , sodium stearate NaST , ben-Ž .zethonium chloride BTC , cetyltrimethylam-

Ž . Žmonium bromide CTAB and triton X-100 TX-.100 . The aqueous solutions were prepared by

deionized redistilled water. The stock solution ofŽ . Ž .cobalt 1 mgrml was made from Co NO . The3 2

Ž .stock solution of copper 1 mgrml was preparedŽ .by dissolving copper metal 0.5 g in concentrated

Ž .HNO 12 ml and diluting to 0.5 l by water.3

Table 1Optimal instrumental parameters for AAS

ETAAS Co Cu

Wavelength 240.7 nm 324.8 nmSpectral width slit 0.7 nm 0.7 nmLamp current 30 mA 15 mABackground correction D D2 2Drying 1008C, 30 s 1008C, 30 sCharring 13008C, 30 s 9008C, 20 sAtomizing 26008C, 5 s 23008C, 5 sCleaning 26508C, 5 s 26508C, 5 sGas Argon

aFAAS Ca MgWavelength 422.8 nm 285.2 nmSpectral width slit 1.3 nm 0.7 nmLamp current 10 mA 15 mAOxidantrfuel gas mixture Airracetylene

a FAAS, flame atomic absorption spectrometry.

Before each investigation, standard solutions ofthese analytes were freshly prepared by dilutingtheir stock solutions. The stock solution of ironŽ .30 mgrml was prepared by dissolving an ap-propriate amount of high-purity iron metal inconc. HNO . The solution was heated to trans-3

Table 2Instrumentation and operating conditions for cobalt and copper ICP-AES

ICP system Varian, Liberty 110

RF generatorOperating frequency 40.68 MHz

Ž .Coupling Direct Serial Coupling, DISC f70% efficiencyŽR.f. power Auto tune, 1.0 kW 238.892 nm for Co and 324.754 for Cu

SpectrometerOptical arrangement Czerny Turner 0.75 m focal lengthGrating HolographicGroove density 1800 linesrmmSize 100=90 mmSpectral resolution 18 pmBackground corrector Dynamic

Sample introduction areaPlasma Ar flow rate 12 lrminAuxiliary Ar flow rate 0.75 lrminNebulizer Ar flow rate 2 lrminSpray chamber Inert Sturman-MastersSample flow rate 2 mlrminPeristaltic pump 25 rollers, 1 turnrmin incrementIntegration time 3 s, smart integration 5 s, direct reading instrument


Ž . Ž .form all iron to iron III . Diluting this Fe NO3 3stock solution, a series of standards with the ironconcentration ranging from 2.5 to 100 mgrl wereobtained. The solution of HMA]HMDTC wasmade as 0.1 molrl in 96% ethanol. The 0.5%solutions of tensides were prepared by dissolvingthe appropriate amount of surfactant in waterŽ . Ž .TX-100 , in 95% ethanol NaDDS, NaOL and in

Ž .99.7% propan-2-ol NaPL, NaST . The pH ofŽmedia was regulated by a HNO solution 0.13

. Ž .molrl and solutions of KOH 2.5 and 10% . Asaturated solution of KNO was used to adjust3the ionic strength of the system. A solution of

Ž .NH NO 0.1 molrl served to transfer the con-4 3tent of the beaker into the flotation cell.

2.3. Procedure

A combined glass electrode was immersed intoŽ .a sample 1 l of acidified fresh water. Next 6 ml

of saturated KNO solution and 1 ml of the 103Ž .mgrml Fe NO solution were added. The solu-3 3

tion had a pH 2.5]3.0. The pH of the reactionmixture was carefully adjusted to 6.0 by means of

Ž .the 10% KOH solution at the beginning andŽ .2.5% KOH at the end . The yellow]brown pre-

cipitate of Fe O ?xH O growing was stirred for 52 3 2Ž .min a first induction time, t . Then, 3 ml 0.11

molrl solution of HMA]HMDTC were addedŽ .and the black precipitate of Fe HMDTC was3

Žformed. After the second induction time } t 152.min of stirring , 1 ml of the NaDDS solution was

added and the content of the beaker was trans-ferred into the flotation cell by a small portionŽ . Ž3]4 ml of 0.1 molrl NH NO . An air stream 504 3

.mlrmin was kept for 1]2 min to raise the precip-itate flakes to the water surface. There a blackfoamy layer was obtained and the aqueous solu-tion in the cell became completely clear andcleaned of the collector. Then, the glass pipette-tube was immersed into the cell through the foamlayer and the water phase was sucked off. Thesolid phase into the cell was decomposed using a

Ž .hot 65% HNO solution 2.5 ml . When the liquid3in the cell became clear yellow, the solution wassucked off by vacuum through the bottom of the

Ž .Fig. 1. Influence of iron mass on cobalt recoveries R : pHsŽ y.5.5; I s0.02 molrl; c HMDTC s0.0002 molrl; 1 ml 0.5%c

alcoholic solution of NaDDS.

cell and collected in a volumetric flask of 25 ml.The cell and the pipette-tube were washed with 4molrl HNO solution. The flask was filled up to3the mark with the solution of 4 molrl HNO and3the sample was ready for AAS measurements.

3. Results and discussion

3.1. Influence of iron mass on cobalt and copperflotation efficiency

The influence of iron mass on cobalt and cop-per flotation efficiency was studied by changingthe iron mass added to the reaction system, while

Žall other parameters were kept constant pHs5.5,amount of HMDTCy, ns2=10y4 mol, ionicstrength, I s0.02 molrl, 1 ml of NaDDS solu-c

.tion . For this purpose a series of flotations wereperformed by addition of different masses of ironŽ . Ž .2.5]100 mg to the working solutions 1 l con-taining 25 and 50 mg cobalt, i.e. copper. Theexperimental data of these investigations showthat cobalt and copper floatation recoveries can

Žbe maximal 94.9]95.4% for Co and 94.5]95.0%. Ž .for Cu using 20 mg of iron Figs. 1 and 2 .


Ž .Fig. 2. Influence of iron mass on copper recoveries R :Ž y.pHs5.5; I s0.02 molrl; c HMDTC s0.0002 molrl; 1 mlc

0.5% alcoholic solution of NaDDS.

3.2. Influence of medium pH on cobalt and copperflotation efficiency

The effect of medium pH on cobalt and copperflotation recoveries was studied at different pH

Ž .values of solutions 1 l containing 25 and 50 mgŽ .of colligend cobalt and copper . The mass of iron

ŽFig. 3. Influence of pH on cobalt recoveries 20 mg Fe,2=10y4 mol HMDTCy, I s0.02 molrl by KNO , NaDDSc 3

.as surfactant .

Ž . y Ž y4 .20 mg , amount of HMDTC 2=10 mol ,Ž .ionic strength 0.02 molrl and amount of NaDDS

Ž .1 ml 0.5% alcoholic solution were kept constant.The pH was adjusted within the range of 3.0]6.5.Since at pH values higher than 6.5 the collector

Ž .Fe HMDTC begins to hydrolyze, the investiga-3tions on the pH values higher than 6.5 were notcarried out. The results of these investigations are

Ž .presented as RrpH curves Figs. 3 and 4 . It isobvious that within the pH range of 3.5]5 therecoveries of both colligends are lower due to theprotonation of the dodecylsulfate anion in acidicmedia. The optimal pH range for cobalt flotationis 6.0]6.5, where the recoveries are in the rangeof 98.0]100.0%. Within the same pH intervalcopper recoveries are 95.9]100.0%. The pH 6 waschosen as appropriate for the further investiga-tions.

3.3. Influence of amount of HMDTCyon cobaltand copper flotation efficiency

This influence was studied by varying they Ž .HMDTC amount, while iron mass, pH 6.0 and

Ž .ionic strength I s0.02 molrl were kept con-cstant. Four series of standard solutions, contain-ing 25 mg of cobalt, i.e. copper per 1 l, were

ŽFig. 4. Influence of pH on copper recoveries 20 mg Fe,2=10y4 mol HMDTCy, I s0.02 molrl by KNO , NaDDSc 3

.as surfactant .


y Žfloated by different amounts of HMDTC 1.3=y4 y4 .10 y6.0=10 mol . The first series of solu-

tions, contained 5 mg, the second 10 mg, the third20 mg and the fourth 30 mg of iron. The data ofthese investigations are presented in Figs. 5 and6.

As can be seen, the recoveries of two colligendsarise by addition of larger amounts of HMDTCy.

Ž .Quantitative separations of cobalt R)95.0%can be obtained by using an iron mass of 5]30 mgin combination with 3=10y4 or 6=10y4 mol ofHMDTCy added to 1 l of the solution tested.Copper can be quantitatively separated by addi-tion of 10]30 mg of iron together with 3=10y4

or 6=10y4 mol of HMDTCy to 1 l. For simulta-neous collection of these two colligends, a massof 10 mg Fe and 3=10y4 mol HMDTCy werechosen as the most appropriate.

3.4. Ionic strength

The coagulation of the system tested dependsŽ .on the ionic strength I of the media. The effectc

of I on cobalt and copper flotability was studiedcby flotation of three standards of the analytesinvestigating at pH 6 with 10 mg iron and 3=10y4

mol HMDTCy without adding any ionic strength

Ž y.Fig. 5. Cobalt flotation dependence on n HMDTC at con-stant pHs6.0, I s0.02 molrl and with constant mass of ironcŽ .5, 10, 20 and 30 mg .

Ž y.Fig. 6. Copper flotation dependence on n HMDTC at con-stant pHs6.0, I s0.02 molrl and with constant mass of ironcŽ .5, 10, 20 and 30 mg .

adjuster. The volumes of standards were 250, 500and 1000 ml. Each standard contained 25 mgcobalt, i.e. copper, so that the final solution con-

Ž .centrated by flotation 25 ml had a concentrationof 1 mgrml. To each standard, 10 mg Fe wereadded by 1 ml of 0.1791 molrl solution of

Ž .Fe NO . The ionic strength of the first, second3 3and third solution, respectively, were 0.0043,0.0022 and 0.0011 molrl. After flotations, cobalt,i.e. copper, was determined by AAS. The resultsof these investigations are given in Table 3. Thevalues of cobalt flotation recoveries of the firstsolution were 100%, while of the second and

Ž .third were not quantitative 88.6 and 77.2% . Therecovery of the first, second and third flotation ofcopper were 96.6, 94.2 and 89.9.7%, respectively.These data proved that it is necessary to adjustthe ionic strength of the system higher than 0.0011molrl to obtain the proper coagulation. There-fore, an ionic strength of 0.02 molrl was chosenfor further investigations.

3.5. z potential of the collector particle surfaces

The electrokinetic potential of the collectorsused is an important parameter conditioning theselection of surfactants. Consequently, it is neces-sary to know the z potentials of the collectorparticle surfaces. The experimental data showed


Table 3Ž .Dependence of the cobalt and copper flotation recoveries on the ionic strength I of the solutionsc

Ž . Ž .m Fe rV R %

a bŽ . Ž .10 mgr250 ml g Co g Cu3qŽ .c Fe s0.0007162 molrl I rmolrl 1 mgrml 1 mgrmlc

yŽ .c NO s0.0021487 molrl 0.0043 100.0 96.63

Ž . Ž .10 mgr500 ml g Co g Cu3qŽ .c Fe s0.0003581 molrl I rmolrl 1 mgrml 1 mgrmlc

yŽ .c NO s0.0010743 molrl 0.0022 88.6 94.23

Ž . Ž .10 mgr1000 ml g Co g Cu3qŽ .c Fe s0.0001791 molrl I rmolrl 1 mgrml 1 mgrmlc

yŽ .c NO s0.0005372 molrl 0.0011 77.2 89.93

a Ž .g Co , mass concentration of Co in the final solution concentrated by flotation.b Ž .g Cu , mass concentration of Cu in the final solution concentrated by flotation.

that the z potentials of Fe O ?xH O particles2 3 2were 25.0]28.0 mV, while the second collector

Ž .Fe HMDTC flocs had more positive values of368.4]69.0 mV. The positive signs of the z poten-tials of two collectors hinted that anionic surfac-tants should be proper for flotations.

3.6. Selection of surfactant

The investigations performed within the pHŽ .range of 4.5]6.5 Table 4 confirmed that the hint

of the previous section was correct. Cationic sur-Ž .factants BTC and CTAB foamed very well over

the whole pH range investigated and a copiouswhite scum was formed at the top of the aqueousphase, but the black precipitate of coagulated

Ž .Fe HMDTC was not separated by flotation and3remained in water. The flotations by use of the

Ž .non-ionic tenside TX-100 were also ineffective.The anionic surfactants were more effective.

They were investigated singly and in pairsŽNaDDSrNaOL, NaPLrNaOL and NaSTr

.NaOL . The cobalt flotation recoveries obtainedŽ .by NaDDS within the optimal pH range 6.0]6.5

were 97.8%, while the copper recoveries by thesame surfactant were 100.0%. Copper can also be

Ž .floated successfully 97.5% by NaOL. Among thepairs, the combination NaPLrNaOL was the most

Ž .effective for cobalt flotations 98.3% . For copperflotations the highest recoveries were achieved by

Ž .the combinations NaDDSrNaOL 100.0% and

Ž .NaPLrNaOL 95.1]100.0% . Considering our ex-perience that the application of more than onesurfactant gives an unnecessarily too copious scamwhich is very difficult to destroy by conc. HNO ,3

Table 4Applicability of divers surfactants for colloid precipitate

Žflotation of cobalt and copper I s0.02 molrl, 10 mg iron,cy4 y.3=10 mol HMDTC

Ž .Co R %pH 4.5 5.0 5.5 6.0 6.5

BTC Foam, no flotationCTAB Foam, no flotationNaDDS 64.2 94.2 97.8 97.8 97.8NaOL 63.7 74.2 90.7 90.7 100.0NaPL No foam 48.9 63.1 71.4 76.9NaST No foam 65.8 71.9 71.4 76.9TX-100 Foam, no flotationNaDDSrNaOL 62.3 68.1 83.8 87.8 96.0NaSTrNaOL 32.6 48.9 68.6 72.8 94.1NaPLrNaOL 54.5 82.8 91.4 98.3 98.3

Ž .Cu R %PH 4.5 5.0 5.5 6.0 6.5BTC Foam, no flotationCTAB Foam, no flotationNaDDS 77.1 94.8 97.5 100.0 100.0NaOL 78.3 92.3 92.3 97.5 97.5NaPL No foam 61.6 90.3 90.3 95.1NaST No foam 83.1 83.1 83.1 87.7TX-100 Foam, no flotationNaDDSrNaOL 22.5 85.3 94.5 100.0 100.0NaSTrNaOL 24.8 63.9 68.8 92.6 92.6NaPLrNaOL 68.8 87.7 90.3 95.1 100.0


NaDDS was chosen as the most appropriatereagent for simultaneously flotation of cobalt andcopper.

3.7. Flotability of macroelements present in freshwaters

For more effectual flotation separation, it isvery useful to know flotability of macroelementspresent in the natural waters. Because calciumand magnesium are the most usual constituent offresh waters with higher hardness, their flotabilitywas investigated. The AAS determinations of cal-cium and magnesium concentrations in the watersamples before flotation and in the final concen-trated solutions served to estimate their flotation

Ž .recoveries Table 5 . It was obvious that calciumŽ .recovery of 0.33]1.07% and magnesiumŽ .0.21]0.38% could not float under recommendedconditions for cobalt and copper. Their recoveriesevidenced that they were left in the water phaseas they were not present in the solution tested byAAS. In this way their interferences on cobaltand copper during ETAAS are excluded.

3.8. Detection limit

To determine the standard deviation of the

Table 5Flotability of macroelements

Sample of Before flotation After flotation Ry1 y1Ž . Ž . Ž .water g Ca rmg? l g Ca rmg? l %

Pantelejmon 70.5 0.23 0.33Sreden izvor 89.8 0.21 0.23Kapistec 62.6 0.17 0.27ˇRadusa 2.8 0.03 1.07ˇKavadarci 30.7 0.15 0.45Rasce 81.2 0.27 0.33ˇ

Sample of Before flotation After flotation Ry1 y1Ž . Ž . Ž .water g Mg rmg? l g Mg rmg? l %

Pantelejmon 8.63 0.02 0.23Sreden izvor 9.2 0.02 0.22Kapistec 19.5 0.04 0.21ˇRadusa 48.1 0.14 0.29ˇKavadarci 0.8 0.003 0.38Rasce 8.2 0.02 0.24ˇ

Table 6Ž . Ž .Standard deviation s , relative standard deviation s andr

Ž .detection limit L of cobalt and copper determined bydETAAS

y1 y1Ž .Element srmg ? l s % L rmg ? lr d

Co 0.004 1.9 0.012Cu 0.005 4.5 0.015

method, 10 blanks were floated by the recom-mended procedure and then the concentration ofcobalt, i.e. copper was determined by ETAAS.

Ž .The detection limit of the method L was esti-dmated as three values of the standard deviationŽ .s of the blank. The precision of the method wasexpressed by means of the relative standard devi-

Ž .ation s . The values of L , s and s for cobaltr d rand copper are given in Table 6.

3.9. Application of the method

The proposed flotation method was used foranalyses of cobalt and copper in six fresh watersamples with diverse water hardness. Immediatelyafter the sapling samples were treated by a fewmilliliters of conc. HNO to prevent the possible3hydrolytic precipitation of some mineral salts. ThepH of the conserved sample had to be 2.5]3.

After the flotation, the metals investigated weredetermined by ETAAS using calibration curves,as well as the method of standard additions. Forthis purpose 1-l aliquots of water samples werespiked by known amounts of cobalt and copper,

Žrespectively. Both ETAAS measurements by the.calibration curve and by standard addition seem

to be equally valid. The recoveries of 95.0]104.0%Ž .for cobalt Table 7 and 92.0]104.8% for copper

Ž .Table 8 show that the preconcentration andseparation of these two colligends by the recom-mended procedure is satisfactory. The results ob-tained by ETAAS were compared to the ICP-AESresults. For ICP-AES measurements acidified wa-ter samples were 40-fold concentrated by evap-

Ž .oration from 1000 to 25 ml .

4. Conclusion

The present paper has proved that cobalt and


Table 7ETAAS determination of cobalt in natural water samples by the method of standard additions and compared with ICP-AES results

aSample ETAAS ICP-AESof water Added Estimated Found Recovery Found

Ž .water mgrl Co mgrl Co mgrl Co % mgrl Co

Pantelejmon 0.00 ] 0.12 ] -0.125o b15.05 DH 1.25 1.37 1.40 102.4

pHs7.84 2.50 2.62 2.70 103.2

Sreden Izvor 0.00 ] 0.25 ] 0.23"0.01o17.65 DH 1.25 1.50 1.45 96.0

pHs7.36 2.50 2.75 2.85 104.0

Radusa 0.00 ] 0.125 ] -0.125ˇo25.57 DH 1.25 1.375 1.400 102.0

pHs8.50 2.50 2.625 2.687 102.5

Kavadarci 0.00 ] 0.125 ] 0.12"0.01o5.71 DH 1.25 1.375 1.40 102.0

pHs7.58 2.50 2.625 2.50 95.0

Rasce 0.00 ] 0.25 ] ]ˇo16.49 DH 1.25 1.50 1.45 96.0

pHs7.18 2.50 2.75 2.85 104.0

Kapistec 0.00 ] 0.25 ] 0.19"0.01ˇ23.36 DHo 1.25 1.50 1.45 96.0pHs7.50 2.50 2.75 2.85 104.0

a Results of comparative ICP-AES determination of cobalt. Samples were enriched by evaporation.b Ž .DH Deutsche Harte , German degree of water hardness¨

copper can be separated and enriched success-fully by precipitate flotation by means of Fe O ?2 3

Ž .xH O and Fe HMDTC as a collector mixture2 3

before ETAAS determination. The investigationŽ .evidenced that Fe HMDTC is a more suitable3

Ž . w xflotation collector than Fe TMDTC 22 . When3the collector combination Fe O ? x H Or2 3 2

Ž .Fe TMDTC is used, the flotation has to be3Žperformed by two surfactants NaDDS and

.NaOL , while the combination Fe O ?xH Or2 3 2Ž .Fe HMDTC can be carried by one foaming3

Ž .reagent NaDDS . The effective flotation usingonly NaDDS indicates that the addition ofHMDTCy makes the collector mixture suffi-ciently hydrophobic and does not need two sur-factants. The more positive z potential value of

Ž . Ž .Fe HMDTC 69.0 mV than the value of3Ž . Ž .Fe TMDTC 46.0 mV evidences its higher hy-3

drofobility. The use of only one foaming agentŽneeds a less amount of conc. HNO 2.5 ml rather3

y.than 10 ml conc. HNO with TMDTC and3makes the matrix of the final solutions tested byETAAS less complex. The detection limits of

Ž . Ž .cobalt 0.012 mgrl and copper 0.015 mgrl ob-tained by the new proposed method are lower

Ž .than those obtained for cobalt 0.15 mgrl andŽ . ycopper 0.03 mgrl by the method with TMDTC

w x y22 . The precision of the method with HMDTC ,as a relative standard deviation, is 1.9% for cobaltand 4.5% for copper, while with TMDTCy is

w x5.55% for cobalt and 6.53% for copper 22 . Allthese parameters show the advantage of theprocedure with HMDTCy in relation to themethod with TMDTCy. The recommendedmethod extends the range of conventional atomicabsorption determination of cobalt and copper


Table 8ETAAS determination of copper in natural water samples by the method of standard additions and compared with ICP-AES results

aSample 0ETAAS ICP-AESof water Added Estimated Found Recovery Found

Ž .mgrl Cu mgrl Cu mgrl Cu % mgrl Cu

Pantelejmon 0.00 ] 0.54 ] 0.60"0.02o15.05 DH 1.25 1.88 1.80 100.8

pHs7.84 2.50 3.04 2.95 96.4

Sreden Izvor 0.00 ] 0.43 ] 0.42"0.01o17.65 DH 1.25 1.68 1.63 96.0

pHs7.36 2.50 2.93 2.79 94.4

Radusa 0.00 ] 3.98 ] 3.90"0.03ˇo25.57 DH 1.25 5.23 5.13 92.0

pHs8.50 2.50 6.48 6.58 104.0

Kavadarci 0.00 ] 1.52 ] 1.52"0.02o5.71 DH 1.25 2.77 2.67 92.0

pHs7.58 2.50 4.02 3.99 98.8

Rasce 0.00 ] 4.14 ] 4.12"0.03ˇo16.49 DH 1.25 5.39 5.45 104.8

pHs7.18 2.50 6.64 6.50 94.4

Kapistec 0.00 ] 0.29 ] 0.27"0.01ˇ23.36 DHo 1.25 1.54 1.46 93.6pHs7.50 2.50 2.79 2.72 97.2

a Samples enriched by evaporation.

and can be applied for analyses of traces of thesecolligends in large volumes of dilute aqueoussolutions.

References

w x1 A. Mizuike, Enrichment Techniques for Inorganic TraceAnalysis, Springer-Verlag, Heidelberg, 1983.

w x2 N.M. Kuzmin, Y.A. Zolotov, Koncentrirovanie SledovElementov, Moskva, Nauka, 1988.

w x Ž .3 V.M. Byr’ko, Ditiokarbamaty. Moskva: Nauka, 1988 .w x Ž .4 K.V. Krishnamurty, M.M. Reddy, Anal. Chem. 49 1977

222]226.w x5 A.I. Busev, V.M. Byrko, A.P. Tereshchenko, N.N.

Novikova, V.P. Naidina, P.B. Terentiev, Zh. Anal. Khim.Ž .25 1970 665]669.

w x6 A.I. Busev, V.M. Byrko, A.P. Tereshchenko, N.B. Kro-Ž .tova, V.P. Naidina, Zh. Anal. Khim. 28 1973 649]862.

w x7 A.I. Busev, A.P. Tereshchenko, N.B. Krotova, T.V.Nolde, V.M. Byrko, O.G. Puzanova, Zh. Anal. Khim. 28Ž .1973 858]862.

w x8 H. Berndt, E. Jackwerth, Fresenius Z. Anal. Chem. 290Ž .1978 369]371.

w x9 A. Dornemann, H. Kleist, Fresenius Z. Anal. Chem. 291Ž .1978 349]353.

w x10 M. Betz, S. Gucer, F. Fuchs, Fresenius Z. Anal. Chem.¨Ž .303 1980 4]9.

w x11 S. Bruggerhoff, E. Jackwerth, Fresenius Z. Anal. Chem.¨Ž .326 1987 528]535.

w x12 R. Eidecker, E. Jackwerth, Fresenius Z. Anal. Chem.Ž .328 1987 469]474.

w x13 R. Eidecker, E. Jackwerth, Fresenius Z. Anal. Chem.Ž .331 1988 401]406.

w x14 M. Caballero, R. Cela, J.A. Perez-Bustamante, Talanta´Ž .37 1990 275]300.

w x Ž .15 A. Mizuike, M. Hiraide, Pure Appl. Chem. 54 19821556]1563.

w x16 L.M. Cabezon, M. Caballero, R. Cela, J.A. Perez-´Ž .Bustamante, Talanta 31 1984 597]602.

ˇw x Ž .17 T. Stafilov, K. Cundeva, S. Atanasov, Anal. Lab. 5 1996255]259.

ˇw x Ž .18 K. Cundeva, T. Stafilov, S. Atanasov, Analusis 24 1996371]374.

w x19 M. Hiraide, Y. Yoshida, A. Mizuike, Anal. Chim. ActaŽ .81 1976 185]189.

w x Ž .20 S. Nakashima, M. Yagi, Anal. Lett. 17 1984 1693]1703.


w x21 M. Hiraide, T. Ito, M. Baba, H. Kawaguchi, A. Mizuike,Ž .Anal. Chem. 52 1980 804]807.

ˇw x Ž .22 K. Cundeva, T. Stafilov, Anal. Lett. 30 1997 833]845.ˇw x23 G. Pavlovska, T. Stafilov, K. Cundeva, Fresenius J. Anal.

Ž .Chem. 361 1998 213]216.ˇw x24 G. Pavlovska, K. Cundeva, T. Stafilov, Bull. Chem. Tech-Ž .nol. Macedonia 16 1997 131]138.

ˇw x25 T. Stafilov, G. Pavlovska, K. Cundeva, Microchem. J. 60Ž .1998 32]41.

w x26 O.N. Grigorov, I.F. Karpova, Z.P. Kozmina et al.,Rukovodstvo k Prakticheskim Rabotam Po KolloidnoiKhimii, Khimiya, Moskva, 1964.

w x27 L.D. Skrylev, A.N. Purich, S.K. Babinetz, Zh. Prikl.Ž .Khim. 54 1981 1832]1835.

w x28 J. Eggert, L. Hock, G.-M. Schwab, Lehrbuch derPhysikalischen Chemie, Hirsel Verlag, Stuttgart, 1960.

Flotation separation of cobalt and copper from fresh waters and their determination by electrothermal atomic absorption spectrometry

Documents