CLIN. CHEM.38/10, 1995-2001 (1992)

CLINICAL CHEMISTRY, Vol. 38, No. 10, 1992 1995

Direct Determination of Cadmium and Lead in Whole Blood by PotentiometricStripping AnalysisPeter Ostapczuk

The commercially available equipment for potentiometricstripping analysis (PSA) was tested for routine lead andcadmium determination in whole-blood samples. In con-trast to anodic stripping voltammetry, PSA is not subjectto background interferences from organic electroactiveconstituents in the sample or to the presence of dissolvedoxygen (i.e., oxygen removal is not necessary). To deter-mine lead and cadmium by PSA, it is sufficient to dilutethe blood sample with an appropriate supporting electro-lyte (0.5 mol/L HCI). The detection limit changes withdeposition time and volume of blood sample used. For 1mL of blood and a 1-mm deposition time, the detectionlimit is 1 zg/L for both elements. If the deposition timeincreases to 10 mm, cadmium can be determined at itsnormal concentration in blood (the detection limit is im-proved to <0.1 g/L). Procedures for routine determina-tion of lead and cadmium in whole blood are presented.

Additional K.yphrues: electrochemical detection anodicstripping voltammetiy compared . potentiometi’y ele-ments

For routine determinations of cadmium and lead inbody fluids, only a few analytical methods are suitable,given the low concentration of these elements (especial-ly of cadmium). In interlaboratory comparisons in re-cent years, the analytical technique applied by themajority of the participating laboratories was electro-thermal atomic absorption spectrometry (ETAAS) withvarious pretreatment techniques, such as chemical ma-trix modifiers and Zeeman background correction.’ An-odic stripping voltammetry (ASV), because of its sensi-tivity, was the only alternative method used (1).

Electroanalytical methods based on the strippingtechnique have become widely used for trace metalanalysis in biological fluids (2). The remarkable sensi-tivity of stripping analysis is attributed to the precon-centration (deposition) step that takes place before theactual measurement (stripping) step. For trace metaldetermination in body fluids, ASV and potentiometricstripping analysis (PSA) have received attention. ASV,especially in the differential pulse mode, is a verysensitive low-cost method that often allows the simulta-neous determination of several metals (e.g., Cd and Pb).However, because small amounts of organic impurities

Institute of Applied Physical Chemistry, Research Center (KFA)JUhich, P.O. Box 1913, D-5170 Jillich, FEC.

1Nonsd abbreviations: ETAAS, electrothermal atomic

absorption spectrometry; ASV, anodic stripping voltaininetry;PSA, potentiometric stripping analysis; and CEM, Certified Ref-erence Material.

Received October 30, 1991; accepted April 2, 1992.

in the solution may interfere with the determination bydifferential pulse ASV, complete digestion of the sampleis necessary (3).

The stripping analyzer from ESA (Bedford, MA),which is widely used for clinical purposes (4-7), utilizesthe staircase stripping mode. The sensitivity of thisequipment is not sufficient to determine the low concen-trations of cadmium ordinarily present in whole-bloodsamples (5). However, lead can be quantified in bloodwith this instrument even at concentrations <100 .tg/L,but only if one has a human blood standard with lead at20-30 pg/L (6). A polymer-coated thin mercury filmelectrode can be used to directly determine lead in wholeblood, urine, and sweat, but the sample must be deaer-ated (8). PSA is more suitable for trace metal analysis inblood and serum samples (9, 10).

In this work I tested the commercially availablesystems of computerized PSA for routine lead and cad-mium determinations in whole-blood samples after asimple pretreatment of the sample.

Materials and Methods

Instrumentation

I used a TraceLabT” and SAM2O sample station (Ra-diometer, Copenhagen, Denmark) in connection with aVectra 286/12 computer (Hewlett-Packard, Palo Alto,CA). For possible automation I tested a SAC8O SampleChanger and an ABU93 Triburette, both from Radiom-eter. The Tap2 software package was used to carry outthe analysis. For ASV after high-pressure digestionwith nitric acid in a high-pressure ashing system (HPA;Hans KUrner, Rosenheim, FRG) (11), I used the Model3MB Polarographic Analyzer from EG&G (PAR, Prin-ceton, NJ). All glassware, pipette tips, and vessels werecleaned several times with 10-fold-diluted nitric acid.All operations of sample preparation and determinationwere done in a laminar-flow hood with a high-efficiencyparticle-accumulator filter (Karl Blaymehi, Julich,FRG).

Chemicals

A mercury plating solution with Hg(ll), 800 mg/L, in1.3 molIL hydrochloric acid was prepared from Hg2C12(p.A. grade; Merck, Darmstadt, FRG) and concentratedhydrochloric acid (30%, Suprapur; Merck); alterna-tively, I used the plating solution supplied by Radiom-eter (no. S2201).

Samples

Certified Reference Materials (CRMs) BCR 194, BCR195, and BCR 196 (Community Bureau of Reference,Brussels, Belgium) and two standards prepared in our

Peak area ta/Vi

/

/,

//

/

400

000

200

100

0

/L’Cadmh.m

000 700 000 000 1000 1100 1200 1300 1400

- E ImVI

FIg. 1. Effect of deposition potential on the peak areas for cadmiumand leadSample:0.1 ml of BCR 196 + 9.3 mLof l1O + 0.5 mLof concd. HCI + 0.2mL of plating solution(Hg2’ 800 mg/I). DepositiontIme 3 mm

1996 CLINICAL CHEMISTRY, Vol. 38, No. 10, 1992

laboratories (stored frozen at -150 #{176}C)were used fortesting the analytical method. All remaining sampleswere analyzed as part of international comparison pro-grams or were taken from apparently normal people inour medical institute. These blood samples were stabi-lized with K2 EDTA (Monovette; Sarstedt Inc., Newton,NC).

ProceduresMercury plating procedure. Every morning the glassy-

carbon electrode was polished with aluminum oxidepowder (0.3-un particles) and rinsed with water. Aftermixing 10 mL of water and 1 mL of plating solution inthe 5-20-mL beaker (Radiometer, code no. 904-488), Iset the potential to -3000 mV (vs a saturated standardcalomel electrode) and the stirrer to “stir 5.” The evolu-tion of hydrogen bubbles on the electrode surface marksthe cleaning of electrode surface by this procedure. After

a few seconds I turned off the potential, rinsed theelectrode with water, and then started the platingprocedure. After being plated for 6 mm at differentpotentials (starting at -300 mV, then decreasing at1-mm intervals by 100 mV, to end at -800 mV), theelectrode surface is covered with a mercury ifim. Thisfilm can be used for several analyses (all day or forseveral days if the potential is maintained over anextended period). Before determining cadmium and leadby ASV, the whole-blood samples are digested withnitric acid in the high-pressure ashing system.

Lead determination. I mixed 9.3 mL of water with0.1-0.25 mL of whole blood in the standard 5-20-mLbeaker (no. 904-488), 0.5 mL of concentrated hydrochlo-ric acid, and 0.2 mL of plating solution. I placed thebeaker at the SAM2O sample station and started theprogram for lead determination. The deposition time forlead with stirring was 1 mm at -800 mV and the curvewas recorded after an equilibrium time of 30 s. Tocalibrate the sample with only one standard addition,one should have the standard addition increase theconcentration of lead by fivefold. If two or more standardadditions are used, then each can be approximatelytwice the analyte concentration in the sample solutionused. To precondition a freshly prepared mercury thin-film electrode, I performed one depositionlstripping cy-cle in the sample before using the electrode for tracemetal determination.

Combined cadmium and lead determination. I mixed9.3 mL of water with 0.5 or 1 mL of blood in the standardbeaker, then added 0.5 mL of concentrated hydrochloricacid and 0.2 mL of plating solution. Cadmium and leadwere deposited at -1150 mV. The deposition time de-pended on the cadmium concentration. For cadmiumconcentrations <0.1 .tg/L, the deposition required 10mm; for concentrations of-i 1.gfL, 1 mm was sufficient.All other steps are similar to those described for the leaddetermination procedure.

Results and DIscussionOptimization of the Procedure

In stripping techniques three conditions are mostimportant: deposition potential, deposition time, and

sample volume. Figure 1 demonstrates the influence of

the deposition potential on the peak area of cadmiumand lead in diluted CRM BCR 196 at constant depositiontime. The optimal deposition potentials for lead andcadmium determinations, under the described condi-tions, ranged from -1100 to -1300 mV (vs a saturatedstandard calomel electrode). At constant deposition po-tential (-1150 my), the peak area changes with theacid concentration (see Figure 2), for two reasons: theliberation of cadmium and lead ions and the evolution ofhydrogen. At pH <2, all cadmium and lead ions areelectrochemically available. An increase of H concen-tration increases the amount that can be reduced toform H2 gas. The small hydrogen bubbles destroy theabsorption layer of organic substances at the electrode,so the peak areas for cadmium and lead increase. Forvery high H concentrations or at more negative depo-sition potentials, the evolution of hydrogen gas on theelectrode surface (large H2 bubbles) decreases the elec-trode surface, so the peak areas for lead and cadmiumdecrease also.

The lead and cadmium peak areas also change lin-early with the blood volume used (Figure 3). This is themost important factor for cadmium determinations inwhole-blood samples at “normal” concentrations. “Nor-mal” lead concentration in whole-blood samples is <30g/L (12). For cadmium the “normal” concentration canbe <1/300 of the lead concentration (13). Given thesedifferences, it is very difficult to determine both ele-ments at the same time. For lead determination asample size of 50 L is adequate; however, handlingsuch a small sample needs an experienced technician,given the risks for contamination or volume errors whentransferring the whole-blood samples (the first volumepipetted must be discarded). For cadmium quantifica-tion, �0.5 mL of whole blood is necessary.

The third variable that influences peak area is thedeposition time. As expected, peak areas of cadmiumand lead change linearly with the deposition time (Fig-

No. of measurements

Sample volume ImLI

Lead - Cadmium

* #{149}09.9 .1- 3.1 (4.4%)

1 2 3 4 6No. of measurement.

2 2.5 3Time fmlnl

4 CadmIum -*- Lead

CLINICALCHEMISTRY, Vol. 38, No. 10, 1992 1997

BCR 194 tCdl 0.5 ugh.

z - 34.89 ..‘- 1.00 (3.0%)

x - t07 .1- 0.48 (7.0%)

tO mL blood, I mm.

mL blood. 1 mm.

BCR 194 IPbl . us uig/l.

Fig. 2. Effect of pH and time on the cadmium and lead concentrationsin BCR196solutionSample and deposition time as in Fig. 1; deposition potential -1150 mV

ure 3). The deposition time determines the total analy-sis time. if the total analysis time for both elements withtwo standard additions is to be <10 mm, then �i-mLsamples of whole blood must be used for the cadmiumdetermination at normal concentrations. if only verysmall sample sizes are available for the determination,a longer deposition time is needed (especially for cadmi-um). Moreover, because of the very low cadmium con-centration in whole blood, all operations with this sam-ple must be done (if possible) under clean-benchconditions.

Figure 4 demonstrates the influence of the samplevolume and the deposition time on the reproducibility ofthe peak area. If the peak area is -5 s/V, the CV variesbetween 5% and 10%. At normal cadmium concentra-tions in blood, sufficiently good results can be obtainedby using 1 mL of sample and a i-mm deposition time (or0.1 mL of sample and a 10-mm deposition time). Fornormal lead concentrations, 0.1 mL of sample and adeposition time of 1 mm are adequate.

The mercury(ll) concentration in the solution influ-ences the sensitivity of the determinations, if the Hg2concentration in the solution is increased, the timeneeded for oxidation of lead and cadmium from the

Fig. 3. Change of peak area with sample volume (BCR 196) anddeposition timeDepositionpotential -1150 my; deposition tIme 3 mm. Conditions as InFig. I

amalgam is reduced and the sensitivity decreases. Tominimize environmental pollution by mercury (wastewater and the laboratory environment), the concentra-tion of Hg2 in the solution should be kept low. I foundthat for low concentrations of Hg2 (<4 mg/L) thereproducibility of the peak area is poor (CV 10-15%).For Hg2 concentrations >40 mgfL, the CV of the peakarea is -2%. I chose an Hg concentration between 8and 20 mg/L as a compromise between peak reproduc-ibility and environmental pollution.

The concentration of hydrochloric acid is also impor-tant in these lead and cadmium determinations (Figure2). At greater pH values (pH >3), the complexing agentsin the blood sample (EDTA, proteins, etc.) form com-plexes with lead and cadmium, so that no oxidationpeaks for these elements appear. At low pH values (pH<0.5), the proteins in the sample coagulate; however,this does not substantially influence the results ob-tained.

As Figure 2 shows, the cadmium and lead peak areasdepend very strongly on the pH of the whole-bloodsample. At pH <0.6, the cadmium peak area changedwith time, probably because of an exchange betweenfree and bound cadmium ions. At pH L5, after 1 h only-80% of the cadmium ions are electrochemically activeat the deposition potential of -1150 mV. For lead, theexchange between bound and free ions is very quick and

o 4 8 1216 2024283236404448Time Iminj

400

300

200

100

0 -_,--_- ---

0 4 8 12 16 20 24 28 32 36 40 44 48Time fminJ

FIg. 4. Effect of sample volume and deposition time on the repro-duCibilityof cadmium and lead peak areas (BCR 194 sample)DepositionpotentIal -1150 mV (other conditions as in Fig. 1)

Peak area Iv/sl Lead

1998 CLINICALCHEMISTRY, Vol. 38, No. 10, 1992

100

..

--

II I I-0.5 0-5 -4.6 -4 -3.6 -3 -2.6 -2 -1.6 -1

log (C ISI Triton X-i0O)

- with Tr iton X-100 - without Triton X-lOo

Peak area Iv/.I Cadmium

Table 1. CadmIum and Lead DetermInatIons byPotentlometric StrippIng AnalysIs

Mean (SD) concn, tig/L

Cadmium Lead

Sample PSA Certified value PSA Certified value

8CR 194 0.36(0.08) 0.5 (0.1) 123 (5) 126(4)BCR 195 5.28 (0.21)8CR 196 12.7 (0.4)

5.39 (0.24)12.4 (0.5)

422 (22)785(31)

416 (9)772(11)

n = 10 each for PSA determinations. Cd, 1 moVL = 112.4 g/L; Pb,1 pmol/L 207.2 zg/L.

in all cases (pH <2) the total amount of lead is deter-mined (see data in Table 1). From these experimentaldata it can be concluded that “speciation” of elementscould also be carried out by PSA.

The observed change of the peak area with pH affectsthe sensitivity of the determination for lead. The peakarea for lead at pH 0.6 is -=75% greater than at pH 1.5.For cadmium this change is even more significant. Theprobable explanation for this phenomenon is the self-cleaning of the electrode surface, which removes ad-sorbed surfactant.s by hydrogen evolution.

\log (C Isi Triton X-100)

-a-- with Triton X-100 - without Triton X-100

Fig. 5. Effect of Triton X-100 on the peak areas of lead and cadmium

Conditionsas in Fig. 2

Influence of Triton X-100

Jagner et al. (10) used high concentrations of TritonX-iOO in the supporting electrolyte for PSA determina-tion of lead in blood. They developed a new type ofelectrode for very small sample volumes without astirrer. In my system, stirring is a condition that can beused to increase the sensitivity. At greater rotationalspeeds, the amount of reduced elements on the electrodesurface increases (the peak area is larger). The optimalstirrer rotation speed for a volume of 10 mL, determinedexperimentally, is 5-7. At higher stirrer rotations thesolution becomes turbulent.

Figure 5 demonstrates the influence of Triton X-100on the peak areas for lead and cadmium determinationsin blood. For lead, increasing the Triton X-100 concen-tration makes no substantial differences in the peakarea. For cadmium, a very low concentration of TritonX-iOO greatly reduces the peak area. At higher TritonX-100 concentrations (>10 mLIL), the stirred solution issaturated with air bubbles, which can be adsorbed onthe electrode surface. For cadmium determinations instirred blood samples, the addition of Triton X-i00 is notuseful.

Determinations in CRMs

Lead and cadmium concentrations are greater inCRMs than in normal blood samples. Thus the determi-nation of lead in CRMs by PSA is a very rapid and easilyperformed analytical procedure. The total time needed

Cadmium concentration lug/LI

* #{149}3.43 .1- 0.12 (3.4%)

4. x #{149}2.61 .1- 0.11 (4.4%).4.

4-

* #{149}1.91 #{149}/-0.16 (8.4%)2 -- - -- - ---------------------- ------ -

0 20 40 60 80 100 120 140

Time thourel

- Sample 9 + 8ampio 7 * Sample 12

Lead concentration lug/LI

0 20 40 60 80 100 120 140Time Ihoursi

CLINICAL CHEMISTRY, Vol.38, No. 10, 1992 1999

for one determination is -5 mm (with two standardadditions). For lead concentrations >0.5 moI/L (1001zgfL), the shorter deposition time can be used, whichshortens the determination time considerably. Cad-mium concentrations in CRMs are 3- to 100-fold greaterthan those found in normal blood. Thus the cadmiumdetermination in these samples is easier and less timeconsuming than in real whole-blood samples. Table 1shows lead and cadmium results obtained by PSA insome CRMs. The agreement between certified valuesand measured values is very good.

Determinations in Fresh Blood Samples

Lead and cadmium concentrations in fresh bloodsamples (not frozen, but analyzed as quickly as possibleafter sample collection) are lower than have been ob-served in the available CRMs. The lead values in thesefresh samples varied between 40 and 60 1g/L, similar tothe concentrations found by Novak et al. (12). Thecorresponding cadmium value was <1 gfL; in only afew samples were higher values found. All fresh bloodsamples are deep red after dilution with water. Addinghydrochloric acid to these solutions changes the color tobrown or black (depending on the proportion of blood inthe diluted sample) and eliminates the differences be-tween fresh samples and CRMs. If the fresh bloodsamples are stored by freezing at -20 #{176}C,the color of thestored samples changes with time to brown or black(similar to the CRMs).

Figure 6 demonstrates the effect of storage time onthe cadmium and lead determinations in three bloodsamples. The increased cadmium concentration in thesesamples could indicate that the blood was taken fromsmokers or was contaminated during sampling. Giventhe lack of effect of the time from the collection of thesample to its analysis, lead and cadmium can be deter-mined in fresh whole-blood samples directly after sam-pling.

Method Evaluation

Results of interlaboratory comparison. Part of theevaluation for the described method involved participa-tion in interlaboratory comparison studies. Table 2presents the results of an external quality-control pro-gram organized by the Danish National Institute ofOccupational Health. In only one case was the cadmiumconcentration measured lower than the values found bythe reference laboratories. The second round of theinterlaboratory comparison was started in spring 1991for the determination of lead only. The results of thisexternal quality assessment are reported elsewhere (1).

Internal quality control. To test the quality of the leadand cadmium determinations by PSA in whole-bloodsamples, I compared this method with other analyticalmethods: ETAAS, the method most widely used for traceelement determination in whole-blood samples, andstripping voltammetry, the latter requiring the diges-tion of blood samples by high-pressure ashing (14). Asshown in Table 3, in all concentration ranges the values

6C. .1- 3.8 (6.3%)

Sc, x #{149}53 .1- 1.1(2.2%).i-* x #{149}41,8 #{149}/-2.2 (6.4%)

-*--------

-

-. --- -

-------

4C,

3C

.c

iC

* Sample 12 + 8ampie 7 . Sampie 9

Fig. 6. Change of cadmium and lead concentrations in fresh bloodsamples with time after sample collectionSample: 9.3 mLof water + 0.25 mLof blood + 0.5 mLof concd. HCI+ 0.2 mLof plating solution; deposition tIme 3 mm

found by PSA agree well with values found by otheranalytical methods.

Day-to-day precision. Table 4 presents the results ofday-to-day precision studies for lead and cadmium de-terminations in four different blood samples. The preci-sion depended strongly on the quality of sample prepa-ration and standard addition. Preliminary experimentswith the autosampler have demonstrated a significantlybetter precision for some samples if automated standardaddition was used (data not shown).

Advantages of PSA for Cadmium and Lead Analysis

PSA, used here for direct determination of lead andcadmium in whole blood, may also be suitable for thedirect determination of these elements in serum (9) andurine (15). In comparison with ASV, PSA has severaladvantages. Sample preparation for PSA consists ofsimple dilution with an appropriate supporting electro-lyte. The contamination or loss of elements by sampledigestion before an ASV determination is eliminated.Sample deoxygenation also is not necessary in PSA. Thedetermination sensitivities of PSA and ASV are verysimilar, but the time needed to determine lead andcadmium in whole-blood samples by PSA is consider-ably less than by ASV (also without digestion). The ESAModel 3010A stripping analyzer and the TraceLab an-alyzer determine lead within similar times, but theTraceLab can also determine cadmium.

CadmIum, nmol/L

Mean ± SD

Lead, gimol/L

‘Ave determinations of each sample.b 23 partIcipants (for n <23, not all laboratories presented usable results).C 29 partIcipants (forn <29, not all laboratories presented usable results).d Listed In parentheses is n = results used for mean calculation.

Direct determInation

CadmIum Lead CadmIum Lead

After HPA digestion

Sample PSA ETAAS PSA ETAAS PSA DPASV PSA DPASV

Blood 1 3.41 3.74 30.1 32.0 3.60 3.82 34.6 34.8

Blood 2 3.03 3.11 80.2 75.0 2.90 2.96 76.6 78.7

Blood 3 2.71 2.78 139 134 2.86 2.78 146 150Blood 4 1.88 1.44 229 221 1.45 1.55 215 225

Blood 5 0.85 0.90 322 300 1.17 1.01 335 330SM-A 0.71 49.2 49.5 0.82 0.85 50.2 50.8SM-B 0.77 0.84 90.1 90.6 0.80 0.78 88.9 85.5

Sample

Lead, Mg/L1#{149}2aCadmium, g/L3b40

Day

1 2

22.6 (1 .7)d 20.2 (1.3)279 (19) 285 (15)

3 4 5

23.0 (2.1) 20.6 (0.9) 21.7 (1.1)262 (10) 253 (12) 271 (14)

Overallmean (SD) CV, %

0.56(0.15) 0.61 (0.11) 0.72 (0.18) 0.59(0.09) 0.63 (0.12)

3.06(0.21) 3.38(0.19) 3.02(0.14) 3.17(0.20) 3.29 (0.15)

21 .6 (1 .2)

270 (13)

0.62 (0.06)3.18 (0.15)

5.64.8

9.84.8

Table 2. Interlaboratory Survey Results of CadmIum and Lead DetermInations

2000 CLINICAL CHEMISTRY, Vol. 38, No. 10, 1992

5-

DEQAS1/1DEQAS1/2DEOAS1/3DEQAS1/4DEQAS1/5

PSA33 ±228.7 ± 1.416.5 ± 3.512.1 ± 1.19.0 ± 0.7

Other Iabsb

32.6 ± 5.5 (19)c28.2 ± 4.8 (20)23.2 ± 4.7 (20)12.6 ± 2.5(17)

8.5± 1.6(16)

PSA

0.172 ± 0.009

0.356 ± 0.0170.705 ± 0.0261.038 ± 0.0531.606 ± 0.067

Other IabsC

0.173 ± 0.029(24)0.372 ± 0.050 (25)0.707 ± 0.091 (26)1.228 ± 0.134(25)1.745 ± 0.150 (25)

Table 3. Cadmium and Lead ConcentratIons (g.*g/L)In Whole-Blood Samples Measured by VarIous Methods

SM-Aand SM-B:our intemal control materials, prepared in German Environmental Specimen Bank. HPA, high-pressureashing. DPASV, differentialpulse ASV.

Table 4. Day-to-Day Precision of Lead and CadmIum DetermInatIons In Blood by PSA

#{149}Samples from Danish National Institute of Occupational Health, AMI:ICL-round 5.b Sample from nonsmolcer.C Sample from smolcer.d Mean (and SD) of 3 determinations each day.

The main advantage of PSA over atomic absorptiontechniques is that it allows cadmium and lead to bedetermined simultaneously. The PSA instrumentationis small and does not need any servicing or specialinstallation requirements (e.g., cooling, gas supply). Forcadmium and lead determination, ETAAS needs samplevolumes of whole blood similar to those in PSA.

An important advantage of ETAAS over PSA, how-ever, is automation, although it may also be possible toautomate PSA (16). A preliminary test with the sam-pler changer (SAC8O) and automated standard additionwith ABU93 (from Radiometer) demonstrated that thesample volume needed for automated PSA determina-tion must be larger [only standard beaker 12-45-mL(code no. 904-677) can be used with the minimumsolution volume of 20 mU. The sensitivity attributable

to constant stirring speed is lower than that observedwith the manual system. My experience indicates thatthis automated system works very well for the determi-nation of lead. In this case the total volume of wholeblood needed for the determination is 0.1 mL (or lower,depending on the concentration in the sample or thedeposition time used). For the cadmium determinationthe smallest volume that can be used is 0.5 mL. Thelonger deposition time needed leads to a longer analysistime. If a large number of samples are to be analyzed,the determination time can be reduced by the use of acalibration plot instead of the standard addition tech-nique. The cost of this automated system for PSA, basedon commercially available equipment, is about half thatof an ETAAS system with comparable sensitivity.

The future development of an automated PSA system

CLINICAL CHEMISTRY, Vol.38, No. 10, 1992 2001

for determination of elements in body fluids can, like theuse of flow cells with microelectrodes, reduce the re-quired sample volume and determination time (17).

References

1. Anglov T, Christensen JM. A comparative study of certifiedreference materials and control materials for the quality assur-ance of blood lead determination. Analyst (in press).2. Wang J. Stripping analysis of trace metals in human bodyfluids. J Electroanal Chem 1982;139:225-32.3. Ohms M, Lund W. Comparison of digestion procedures for thedetermination of heavy metals (Cd, Cu, Pb) in blood by anodicatripping voltainmetry. Fresenius Z Anal Chem 1979;298:260-8.4. Morrel G, Giridhar G. Rapid micromethod for blood lead anal-ysis by anodic stripping voltammetry. Cliii Chem 1976;22:221-3.5. Stoeppler M, Mohi C, Ostapczuk P, Goedde M, Roth M, Waid-mann E. Rapid and reliable determination of elevated blood leadlevels. Fresenius Z Anal Chem 1984;317:486-90.6. Roda SM, Greenland RD, Bornscheun RL, Hammond PB. An-odic stripping voltamnietry procedure modified for improved accu-racy of blood lead analysis. Cliii Chem 1988;34:563-7.7. Osterloh JD, Sharp DS, Hata B. Quality control data for lowblood lead concentration by three methods used in clinical studies.J Anal Toxicol 1990;14:8-.11.8. Hoyer B, Florence TM. Application of polymer-coated glassycarbon electrodes to the direct determination of trace metals inbody fluids by anodic stripping voltainmetry. Anal Chem 1987;59:2839-42.9. Jagner D, Joaefson M, Westerlund S, Aren K. Simultaneous

determination of cadmium and lead in whole blood and in serumby computerized potentiometric stripping analysis. Anal Chem1981;53:1406-10.10. Jagner D, Renman L, Wang Y. Simple procedure for thedetermination of lead in microliter amounts of whole blood using aportable potentiometric stripping analyser. Analyst (in press).11. Knapp G. Routes to powerful methods of elemental traceanalysis in environmental samples. Fi-esenius Z Anal Chem 1984;317:213-9.12. Novak L, Ostapczuk P, Stoeppler M. Richtigkeit und Re-produzierbarkeit der AAS-Direktbestimmung niedriger Blut-bleispiegel von normalexponierten Personen. In: Welz B, ed. Col-loquium Atomspektroskopische Spurenanalytik, Vol. 5.Bodenseewerk, FRG: Perkin-Elmer GmbH, 1989:845-54.13. Stoeppler M. General analytical aspects of the determinationof lead, cadmium and nickel in biological fluids. In: Facchetti S, ed.Analytical techniques of heavy metals in biological fluids. Amster-dam: Elsevier Science Publishers B.V., 1981:133-54.14. Ostapczuk P, Stoeppler M, Dttrbeck HW. Present potential ofelectrochemical methods for metal determination in referencematerials. Fresenius Z Anal Chem 1988;332:662-5.15. Jagner D, Danielsson LG, Aren K. Potentiometric strippinganalysis of lead in urine. Anal Chim Acta 1979;106:15-21.16. Cladera A, Estela J, Cedra V. Potentiometric stripping anal-ysis automation. Talanta 1990;37:689-93.17 Huiliang H, Jagner D, Renman L. Automated determinationof lead in urine by means of computerized flow potentiometricstripping analysis with a carbon-fibre electrode. Talents 1987;34:539-42.