Tc

La

b

c

a

ARRAA

KHESAU

1

ciaitee2as

of

0d

Journal of Hazardous Materials 163 (2009) 1360–1368

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journa l homepage: www.e lsev ier .com/ locate / jhazmat

he determination of hexavalent chromium (Cr6+) in electronic and electricalomponents and products to comply with RoHS regulations

. Huaa,b,∗, Y.C. Chanb, Y.P. Wuc, B.Y. Wub

Department of Chemistry & Life Science, Hubei University of Education, Wuhan, Hubei 430205, ChinaDepartment of Electronic Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon Tong, Hong KongDepartment of Material Science, Huazhong University of Science and Technology (HUST), Wuhan, Hubei 430074, China

r t i c l e i n f o

rticle history:eceived 12 March 2007eceived in revised form 6 July 2008ccepted 25 July 2008vailable online 9 August 2008

eywords:exavalent chromium (Cr6+)DXRF screeningpot colorimetric testlkali digestionV–vis spectrophotometry

a b s t r a c t

Toxicity of hexavalent chromium (Cr6+) was focused on with a publication of EU RoHS directive, a novelmethod to determine hexavalent chromium is developed. It is a combination of energy dispersion X-rayfluorescence spectrometry (EDXRF), spot test, alkali digestion and UV–vis spectrophotometric analysis.First, by EDXRF screening, the presence or absence of element Cr was established. Spot test was followedto identify the valent state of chromium because Cr6+ and Cr3+ normally coexist. After alkali digestion,Cr(VI) was separated without an undersired Cr(VI)–Cr(III) interconversions. With a color reagent (DPC) tochelated with Cr(VI), the solution was finally detected by a UV–vis spectrophotometer at a wavelength of540 nm which is the basis of analyzing Cr(VI) quantitatively.

Some parameters affecting analyses were studied. It was found that when pH in the final solutionwas 2.0, the extraction time was 60 min, the extraction temperature was 90 ◦C, pH during the extractionprocess was 7.5–8.5, and a mixed buffer solution (0.5 M K2HPO4/0.5 M KH2PO4) was added up to 1 ml,

colorimetric reagent was added to 2 ml, it is optimal for extraction. Under this condition, interferencesfrom Fe3+, Pb2+, Ag+, etc., were overcome. It was also found that the curves are rectilinear in the range of0–500 �g l−1, the correlation coefficient is up to 0.999924, and the recovery rates are more than 85%, theCr(III)–DPCO complex can be kept stable for 24 h with a relative humidity (RH) range of 60–90%, and atemperature range of 5–40 ◦C. So it can be concluded that the proposed method has a good sensitivity andhigh precision. It is a more convincing and reliable method due to its relative standard deviation (R.S.D.)

ermin

tfaii

iiC

<1% after six replicate det

. Introduction

It is a major goal that economic development and environmentalonservation become compatible and indeed mutually reinforc-ng. The responsibility of the industrial sectors for environmentalnd sustainable development has received prominent recognitionn recent years due to the need for regulatory compliance. Withhe European Union deadline of full compliance with two majornvironmental regulations relating to the imported electronic and

lectrical components and products coming into force from 1 July006 [1], there is a growing pressure on manufacturers in electronicnd electrical (EE) industries to reduce the use of such hazardousubstances. The need to severely restrict the presence of Cr6+ (the

∗ Corresponding author at: Department of Electronic Engineering, City Universityf Hong Kong, Tat Chee Avenue, Kowloon Tong, Hong Kong. Tel.: +852 31758383;ax: +852 31758333.

E-mail addresses: txuehua@cityu.edu.hk, txuehua@163.com (L. Hua).

iafmcreita

304-3894/$ – see front matter © 2008 Published by Elsevier B.V.oi:10.1016/j.jhazmat.2008.07.150

ations of Cr(VI) in an Fe–Ni alloy sample.© 2008 Published by Elsevier B.V.

hreshold limit is 1000 ppm in a matrix), there is a great challengeor thousands of companies such as those producing stainless andlloy steels, pigments, tanning agents, and galvanizing, electroplat-ng and corrosion-resistant industries, etc. where chromium is usedn large amounts [2–4].

Normally, chromium complexes in different valent states coex-st in the same product, both trivalent and hexavalent chromiumn most cases, but the toxicity of these species differ significantly.r3+ is an essential material for humans and animals, and plays an

mportant role as a glucose-tolerance factor (GTF) in insulin, lipin,nd protein metabolism [5,6]. On the other hand, Cr6+ is very toxicor humans and living organisms [7,8]. It can diffuse through cell

embranes after absorption and has been classified as an A1 car-inogen [9]. That is the reason that Cr6+ not Cr3+ is fallen under the

egulatory scope of the RoHS directive. The determination of Cr(VI),specially in complex solid samples such as a PCBs, resistors, etc.,s regarded as one of the most challenging specification tasks dueo interconversions between Cr(VI) and Cr(III) in the course of annalytical process. In recent years, experimental procedures have

ous M

bmaa(sssCt(pmcaebau[cmetdsiXa

2

2

lUa1ao1a(wst

2

t(s1cfdKaD5i0

vip3atMaw4wptba

2

hufnmsmo(ifm[Cspb

2

phiiemrt5assfid11vs

L. Hua et al. / Journal of Hazard

een significantly improved with developments of spectrophoto-etric and chromatographic technologies such as on-line atomic

bsorption spectrometry (AAS) [10], inductively coupled plasma-tomic emission spectrometry (ICP-AES), ICP-mass spectrometryICP-MS) and ion chromatography (IC), etc. applied to determineuch a hazardous substance [11–13]. However, these procedurestill suffer interferences from other ions and coexisting organic sub-tances. In particular, they are not sensitive and selective with highr(III)/Cr(VI) interconversion ratios. Some literature has reportedhat a combination of high-performance liquid chromatographyHPLC) and ion chromatography (IC) can remove interfering com-ounds before Cr(VI) determination [14]. However, the combinedethod is expensive and complex, furthermore, two and three

hromium species are all retained on the separation column andre not affected by a solvent washing. Electrochemical methods,specially, adsorptive stripping voltammetry (AdSV) [15] is said toe simple, inexpensive, and sensitive to determine Cr(VI), however,nother paper attacks this standpoint and reports that the methodnderestimates Cr(VI) due to the presence of insoluble chromate16]. In view of the above difficulties and the pressure of RoHSompliance, this paper presents a novel, cheap, easy, and sensitiveethod to determine Cr6+ in materials of electronic and electrical

quipment, where Cr6+ is used as a so-called “chromic acid” solu-ion for both hard and decorative chrome plating [17]. The methodeveloped can selectively determine the special soluble, sparinglyoluble and insoluble Cr(VI) in the presence of Cr(III) and othernterfering ions. It is a combined application of energy dispersion-ray fluorescence (EDXRF), a spot colorimetric test, alkali digestionnd UV–vis spectrometry.

. Experimental

.1. Apparatus

A NITON XLt 797Z Energy Dispersion X-ray Fluorescence Ana-yzer (supplied by Thermo Electron Corporation in Massachusetts,SA) is used. And a Lambda 25 UV–vis spectrophotometer for uset 540 nm (PerkinElmer, USA), which can provide a light path ofnm was employed. The solutions were placed in quartz cells withpath length equal to 1 cm. A calibrated balance with an accuracyf 0.1 mg, and a heater which can heat up to 250 ◦C, 10, 20, 50, 100,000 ml volumetric flask with acceptable precision and accuracy,ssorted calibrated autopipettes, and cellulosic filter membranes0.42/0.45 �m) were used. A Thermo Orion pH Benchtop meterith a combined electrode was calibrated using standard buffer

olutions of pH being 4, 7, 10 (Hamilton) and was employed to detecthe pH values of solutions.

.2. Reagents and materials for Cr6+

All chemicals were of analytical reagent grade. All aqueous solu-ions were prepared with distilled water (DI Water). A chromiumCr6+) stock solution was prepared by dissolving 2.829 g of potas-ium dichromate (AR Grade 99.9%) in DI water and diluting tol mark, put in a LDPE bottle and kept in a freezer, the con-entration then was 1000 ppm. Working standard solutions werereshly obtained by diluting the stock solution with DI water toifferent concentrations. A phosphate buffer (0.5 M K2HPO4/0.5 MH2PO4) at pH 7 was prepared by dissolving 87.09 g K2HPO4

nd 68.04 g KH2PO4 in a 1 l beaker and diluting to mark withI water. A 0.5 M phthalate buffer was prepared by dissolving.1085 g of potassium hydrogen phthalate in DI water and dilut-

ng to 50 ml. The color chelating agent was prepared by dissolving.25 g 1,5-diphenylcarbazide (DPC, Special Grade, Merck) in a 50 ml

3

s(

aterials 163 (2009) 1360–1368 1361

olumetric flask and diluting to mark with actone, which was storedn an amber LDPE bottle in a freezer. The digestion solution wasrepared by adding 20.0 ± 0.05 g sodium hydroxide (NaOH) and0.0 ± 0.05 g sodium carbonate (Na2CO3) in a 1 l volumetric flasknd diluting to mark with DI water when it cooled down to roomemperature, the pH of this should be 11.5 or greater before use.

agnesium chloride solution was prepared by dissolving of 4.886 gnhydrous MgCl2 in water and diluting to 100 ml, and the solutionas prepared monthly. 0.5 M Nitric acid was prepared by adding5 ml nitric acid (70%) in a 100 ml volumetric flask, marked with DIater and put in an amber BOD bottle. 30% sulfuric acid was pre-ared by adding 30 ml concentration sulfuric acid (98%), marked upo level with DI water and stored in a wheaton glass bottle. All glass,eakers and plastic bottles, etc. were soaked for 24 h in a 5 M nitriccid solution and rinsed at least five times with DI water before use.

.3. The procedure

Fig. 1 shows the procedure to determine sensitive and specialexavalent chromium (Cr6+) for RoHS compliance. First, EDXRF wassed to determine the existence or absence of chromium. If it wasound, a qualitative analysis was necessary because Cr6+ and Cr3+

ormally coexist in the same component. But the direct measure-ent and specification of Cr6+ in the presence of other chromium

pecies are more of a problem. A spot test as an easy and cheapethod was subsequently used to identify the hexavalent state

f chromium compounds. A mixed solution of sodium hydroxideNaOH) and sodium carbonate (Na2CO3) were effective in extract-ng Cr6+ selectively from trivalent chromium specials. In order tourther purify and retain stability during the time of analysis, a pH

eter was applied to monitor the pH of solutions in the range of 1–318,19]. After that, a DPC, a color agent was added to chelated withr6+, and the violet resultant was finally introduced into a UV–vispectrophotometer. At a wavelength of 540 nm, a sharp absorptioneak of the Cr(III)–DPCO complex was displayed, which was theasis for analyzing Cr6+ quantitatively.

.4. The steps and methods

With reference to US EPA 3060A, and 7196A, analyses wereerformed by a UV–vis spectrophotometer. A series of 2.0 ± 0.01 gomogeneous samples, e.g. materials of EE products were weighed

n a 300 ml beaker and 50 ml of the digestion solution was added,ndividually. Then 1 ml of various buffer solutions were added inach case and heated up to 80–100 ◦C with a speed of 45 rpmagnetic stirring, respectively. They were then cooled down to

oom temperature and filtered with the 0.42/0.45 �m cellulose fil-er membranes. The pH of solutions were adjusted to 7–9 with the.0 M nitric acid, diluted to 100 ml and then 95 ml drawn out todjust the pH to 0.5–4 with 30% sulfuric acid. A series of K2Cr2O7tandard solution were processed using the same procedure. Someamples with the addition of standard solutions but only extractedor 15, 30, 45, 60, 75, 90 min and kept at 6–10 of pH levels dur-ng the process of digestion were also studied. After filtering andiluting with DI water, 95 ml filtrates were adjusted to pH 0.8, 1.2,.6, 2.0, 2.4, 2.8, 3.2, 3.6, and heated at temperatures of 30, 60, 90,20, 150, 180, 210, 250 ◦C, individually. Finally, color reagents inarious volumes were added into every solution for the UV–visiblepectrometric tests.

. Results

Taking a Ni–Fe alloy as an example, the EDXRF spectrum ishown in Fig. 2, there are two peaks for Cr at energies of 5.41K� line) and 5.95 (K� line) keV. The peak heights are 30.36 and

1362 L. Hua et al. / Journal of Hazardous Materials 163 (2009) 1360–1368

Fig. 1. The procedure to determine the special Cr6+.

ometr

4Btftd1

c

Fig. 2. The EDXRF spectr

.22, respectively, and the total concentration of Cr 261,500 ppm.ecause the same sample was weighed and analyzed six times,he final values are the means. To consider the presence of Cr,

urther study was necessary. Fig. 3 is the result of UV–vis spec-rophotometric analysis. In case of exceeding the limits of UV–visetection (maximum to 500 �g l−1), the final solutions was diluted000 times. The concentration then was 18,200 ppm.

Fig. 3. The UV–vis spectra of Cr6+ in Ni–Fe alloy matrix.

w(s

lstcms

4

4

fds

y of Ni–Fe alloy matrix.

Cr6+ concentrations (mg/kg) in the samples were calculated:

oncentration (ppm) = A × D × F

S

herein A, concentration detected by UV–vis spectrometrymg l−1); D, dilution factor; F, final digest volume (ml); S, initialample weight (g)

Because concentration detected by the EDXRF technique wasarger than that detected by UV–vis spectrometric analysis, it can beeen that other valence Cr (e.g. Cr3+ or Cr2+) existed. In most cases,his was true particularly when trivalent and divalent chromiumoexisted with hexavalent chromium, so combination of variousethods is beneficial to improve analytical sensitivity of chromium

pecies.

. Discussions

.1. The relative standard deviation (R.S.D.)

The relative standard deviation (R.S.D.) was calculated by theormula that the standard deviation is divided by the mean of theata and multiplied by 100 to give a % value which reflects the preci-ion of the method developed. The R.S.D. values of EDXRF screening

L. Hua et al. / Journal of Hazardous Materials 163 (2009) 1360–1368 1363

tra of

wssp1is

4

e(itwCtfttretpbitobg

TT

B

12345

o

A

C

B

P

S

4

4D

rC[tbs(cpwintensities of the resultant (Cr(III)–DPCO complex). It can be seenthat when the pH is 2.0, there is a maximum absorbed intensity.Thus, this pH is preferred.

Fig. 4. The UV–vis spec

ere 5.9%, 2.3%, 1.4%, 2.0%, 8.5%, 9.4% after six analyses of the sameample. The mean value of the UV–vis spectrophotometric analy-es was 0.8%. Although, both the techniques can determine Cr, therecisions to detect were distinctly different. EDXRF has a R.S.D. of–10%, whereas the UV–vis technique has a R.S.D. less than 1%. Thus

t can be concluded that the later method has higher precision andatisfactory reliability.

.2. Validation of extraction of Cr(VI)

In order to validate the extraction of Cr(VI), several spiked recov-ry experiments were performed. The recovery rate (%) is equal tothe measured concentration/the expected concentration) × 100%,t is an important parameter in estimating the validity of the extrac-ion method developed. Some reagents such as CaCl2, AgNO3, etc.ere added into a series of samples containing 0.5, 1, 2, 4, 5 ppmr(VI), respectively, and extracted with NaOH and Na2CO3 solu-ions. After about half an hour, some solid powder was seen to haveallen to the bottom of the beaker. After an hour, the solutions wereaken out, cooled down upto the room temperature and filtered, andhe filtrate was studied by UV–vis spectrophotometry. The spikeecoveries are shown in Table 1. Where, it can be seen that thexpected values are in good agreement with the certified values andherefore the applied digestion reagents were validated. Some solidowder was precipitate because Ca2+, Ag+, etc. can be precipitatedy OH− as shown in formulae (1), (2). In this way, some other metalons were removed away from the analytes. On the other hand, in

he literature [20], it is reported that Na2CO3 can transform someriginally insoluble and sparingly soluble chromates into the solu-le forms as shown in equations (3–5). Thus, NaOH and Na2CO3 areood digestion reagents which can help to improve the validation

able 1he recovery rates of Cr(VI)

atchs Expectedconcentration (mg l−1)

Certified concentration(mg l−1)

Recovery (%)

0.5 0.4998 99.961 0.9867 98.672 2.0024 100.124 3.8086 95.215 5.1031 102.06

t

TR

p

01122233

Cr(VI) at different pH.

f extracting Cr(VI).

g+ + OH− → AgOH ↓ (1)

a2+ + CO32− → CaCO3↓ (2)

aCrO4 + CO32− → BaCO3↓ + CrO4

2− (3)

bCrO4 + CO32− → PbCO3↓ + CrO4

2− (4)

rCrO4 + CO32− → SrCO3↓ + CrO4

2− (5)

.3. Optimizing the conditions

.3.1. Effect of pH in final solutions on the reaction action ofPC–Cr(VI) and on the recovery rates

The literatures on Cr speciation diagram showed that in the pHange of 0.5–4, the coexisting Cr6+ and Cr3+ can be separated, sincer6+ exists mainly in the anionic forms of (HCrO4

−) and (CrO42−)

21,20]. So in the determination of Cr(VI), pH was monitored inhis range of acidity. In order to study effect of pH on the reactionetween the colorant and analyte, a series of 0.2 mg l−1 K2Cr2O7tandard solutions were added into a batch of iron pin samples8 solutions, with the same content and the same weight in eachase) and digested. 95 ml filtrates were finally adjusted to differentH values according to the proposed procedure, then 2 ml colorantere added in each solution. Fig. 4A and B shows the peak absorbed

The recovery of Cr(VI) was affected substantially by the pH, andhe interrelation was investigated. Table 2 gives the data on the

able 2ecovery as a function of pH

H Ordinate concentration(mg l−1)

Measuredconcentration (mg l−1)

Recovery (%)

.8 0.3458 0.3611 90.275

.2 0.3735 0.3901 97.525

.6 0.3839 0.4009 100.2250.3898 0.407 101.75

.4 0.3864 0.4035 100.875

.8 0.3819 0.3988 99.7

.2 0.3719 0.3884 97.1

.6 0.3415 0.3566 89.15

1364 L. Hua et al. / Journal of Hazardous Materials 163 (2009) 1360–1368

saFrdap2uo

4

daearsao

4

odblttFtc

ivttqsop

Ft

towtaapodtt

adte9

4Taking into account a knowledge of the kinetics of most impor-

tant redox reactions of chromium species, the pH during theextraction process must be carefully adjusted in case of somereductants and oxidants such as sulphides, sulfites or manganese

Fig. 5. The change of absorbed intensities with extraction time.

piked recovery. The results verified that under the condition ofpH 2.0, the recovery rate of the method can reach a maximum.

urthermore, all of the recovery rates exceed 85%, which is in theange of 80–120% [22]. So it can also be concluded that in this con-ition, there are no gross losses during the processes of separationnd purification. The procedure used to select and specify Cr6+ isrecise and reliable. In all subsequent studies, the pH was kept at. In the above cases, 0.4 mg l−1 K2Cr2O7 standard solutions weresed and the absorbed intensities were measured at a wavelengthf 540 nm.

.3.2. Influence of digestion timeA series of e-samples with addition of 0.1 mg l−1 K2Cr2O7 stan-

ard solutions were processed using the recommended procedurebove. 50 ml of alkaline reagent was added and the solutions werextracted for 15, 30, 45, 60, 75, 90 min, respectively. After this, allcidities were adjusted to the optimal condition of pH 2.0. Theesults are shown in Fig. 5. From here, a maximum absorbed inten-ity of Cr6+ was presented when extracted for 60 min. After that, thebsorbed intensity decreased due to volatilization and separationut with a longer heating time.

.3.3. Effect of extract temperatureAs was known, temperature greatly affected the extraction rate

f Cr6+ from other compounds, this interrelation was studied. Theseata were obtained by processing an electronic watch chain (sevenatches, every one of which was 0.5 g) with a yellowish coating

ayer according to the proposed procedure at different tempera-ures. All other parameters were kept constant, that is, pH is equalo 2.0, and the digestion time is 60 min.The results are shown inig. 6. From here, it can be seen that when the temperature is upo 90 ◦C, the efficiency of extraction for Cr6+ from it’s other valentompounds is optimum.

However, in some other samples, especially in leather materials,n the process of tannaging, Cr can present in at least three differentalences, i.e. Cr6+, Cr3+ and Cr2+.When the environment changes,hese species can transform into each other. So in these materials,

6+

he temperature affecting on the specification of Cr complex isuite different. Fig. 7 gives the data obtained from a white leatherample. From here, it can be noted that the absorbed intensity graphf Cr(VI) went up when heated to a higher temperature (the criticaloint is 170 ◦C), and a possible reason is that at higher temperatures,

ig. 6. Relationship between the temperature and absorbed intensity of Cr6+, detec-ion was performed at pH 2, and the digestion time was 60 min.

he oxidation ability to the objective species (Cr6+) is enhanced. Inther words, a higher temperature provokes oxidation of Cr3+, Cr2+

hich increases the concentration of Cr6+. However, about 212 ◦C,he magenta color initially developed slowly disappeared and thebsorbed intensities of the Cr(III)–DPC complex also decreased, andt the same time, a yellow color appeared. The results of this andrevious studies [23], proved that the yellow color is the productf a higher oxidation product diphenylcarbadiazone (DPCDO) oxi-ized by Cr(VI) [24]. DPCDO does not form a complex with any ofhe Cr species, so there is no absorbed intensity at 540 nm whenhe reaction resultant is DPCDO.

Those studies, have proved that the temperature can actuallyffect the resultant of Cr(VI) chelated with DPC. However, theegree influenced in the complex formation is different. In ordero determine Cr(VI) contained in most electronic and electricalquipments, a too high or low digestion temperature is unsuitable,0–100 ◦C is recommended.

.3.4. Effect of pH on the digestion process

Fig. 7. The temperature affect on the determination of Cr(VI) in a leather.

L. Hua et al. / Journal of Hazardous Materials 163 (2009) 1360–1368 1365

dbphatcsfds

4

rbdApnsm

4

aaespc

TE

C

0000001

FCm2

4

trppbtCasdpdto0sai

Fig. 8. Effect of pH of digestion solution on the absorbed intensities.

ioxide, hydrogen peroxide, etc. inducing to the interconversionsetween Cr(VI) and Cr(III). In order to study the effect of theH on the redox contributions, a sample of a PCB board whichas complex contents was digested using a buffered extractionnd heated at 90–100 ◦C. Alkaline digestions with an option forotal Cr(VI), includes the extraction of Cr(VI) with a mixed sodiumarbonate–sodium hydroxide solution for 60 min. The result washown in Fig. 8. It can be seen that when pH was maintained at dif-erent values, the homogeneous samples of the same weight hadifferent absorbed intensity at a wavelength of 540 nm. This alsohows that the optimum pH range of extraction process is 7.5–8.5.

.3.5. Effect of concentration of digestion reagentsThe influence of the concentration of digestion solutions in the

ecovery of Cr(VI) was studied. In Table 3, the results were obtainedy screening a series of 500 mg kg−1 K2Cr2O7 standard solutionsigested at conditions of different concentrations of alkali reagents.ll experiments were performed with a pH. This shows that inde-endent of pH, a change of the alkali solution concentration doesot significantly affected the recovery of Cr(VI). For subsequenttudies, a 0.5 M NaOH/0.28 MNa2CO3 solution was preferred toaintain agreement with a previously published paper [25].

.3.6. Effect of volume of colorimetric reagentThe effect of the volume of colorimetric reagent (DPC) on the lig-

nd selectivity of Cr6+ (0.4 mg l−1 K2Cr2O7 standard solutions were

nalyzed) was studied. Various volumes of the DPC were added intoxtraction solutions before they were introduced into the UV–vispectrometer. All experiments were performed with conditions ofH 2.0, the digestion time 60 min, the temperature 90 ◦C, and theoncentration of DPC 2 × 10−2 mol l−1. The results are shown in

able 3ffect of concentration of digestion solutions on the recoveries of Cr(VI)

onc. of digestion solutions (mol l−1) Recovery of Cr(VI) (%)

.05 96.44

.1 98.71

.2 99.01

.4 98.45

.5 97.38

.75 97.6798.23

At

4

CsiTwiCtoi

Fig. 9. Effect of volume of color reagent on the recovery of Cr6+.

ig. 9. It can be seen that when the volume was 2 ml, the recovery ofr6+ is at the highest, after that, the recovery does not change whenore DPC reagent was used. So for all subsequent studies, 2 ml of× 10−2 mol l−1 DPC solution was added to select Cr(VI).

.3.7. Influences of the buffer solutionBuffer solutions play an important role in preventing Cr6+

urning into Cr3+ and in extracting Cr6+ from chromates. Someesearch has been reported by exploiting an ammonium sul-hate/ammonium hydroxide buffer [26], however, a more recentaper showed that insoluble Cr(VI) was not extracted using thisuffer [25]. In view of this, in this study, another four buffer solu-ions were introduced which can extract the soluble and insolubler(VI). The effects of different types of buffer solutions on thebsorbed intensity in the UV–vis spectrometry were studied. Aeries of steel board with addition of 0.2 mg l−1 K2Cr2O7 stan-ard solutions were used and prepared as in the above proposedrocedure. Usually, the volume of buffer solutions added into theigestion solution is a trace, e.g. 0.5–1 ml [6]. For mode 1, 1 ml mix-ure of 0.5 M K2HPO4 and 0.5 M KH2PO4 was added. For mode 2, 1 mlf 0.5 M phthalate buffer solution was added. For mode 3, 1 ml of.5 M KH2PO4 was added. For mode 4, 1 ml of 0.5 M K2HPO4 bufferolution was added. It can be seen from Fig. 10 that in mode 1 thebsorbed intensity reaches a maximum, and the recovery rate fornsoluble Cr(VI) is up to 98.5%, so this buffer solution is preferred.ll modes were done with a pH 2.0, a digestion time of 60 min, at a

emperature of 90 ◦C.

.4. Interferences of coexisting ions

Interferences of some coexisting ions in the determination ofr(VI) were investigated. In these experiments, 0.5 mg l−1 Cr6+

tandard solutions containing the added interfering ions were spec-fied according to the above recommended optimum conditions.he results are shown in Table 4. This shows the Cr(VI) analysesere not affected markedly by Na+, K+, PO4

3−, Mg2+, etc. Trivalent

ron ions disturbed more markedly the spectrometric analysis ofr(VI), the tolerance limit regarded as the largest amount makinghe spike recovery of Cr6+ less than 85% was 0.05 g l−1. Inferencesf other ions such as Mo(VI), V(V), mercury, etc. in determin-ng Cr6+ special was reported in literature [27,28]. In order to

1366 L. Hua et al. / Journal of Hazardous Materials 163 (2009) 1360–1368

Fig. 10. The absorbed intensity changed with types of the buffer solutions.

Table 4Effect of coexisting ions on absorbed intensities

Coexisting ions Coexisting concentrationlimit (g l−1)

Change in absorbedintensity (%)

Na+ 5.0 1.2K+ 5.0 0.3Cl− 5.0 0.5NO3

− 2.5 1.1SO4

2− 2.5 1.6PO4

3− 2.5 0.35Mg2+ 2.0 0.01Al3+ 2.0 0.4Cu2+ 2.0 1.1Hg2+ 2.0 1.5ZPF

kwiociswab

nwe

4

aFai

tdcu0taOcp

4

Tt[p

onCtw

n2+ 2.0 0.6b2+ 0.1 1.3e3+ 0.05 3.1

nowledge about how to act of iron ions, a 2.5 g iron nail matrixas weighed to do experiment. The spectrum from that matrix

s shown in Fig. 11, where the absorbed peak at a wavelengthf 556 nm showed that it is a co-resultant of both Cr6+and Fe3+

helateding with DPC. In general, iron atoms have no capabil-ty to chelated with DPC, only iron ions may do so. To overcome

uch a problem, a phosphate buffer was added, Fe3+ first chelatedith the PO4

3−, and the resultant is colorless which does notffect the absorbed intensity of the objective Cr(III)–DPCO complexecause of 99.93% recovery rate. For other ions, 20 mg l−1 mag-

Fig. 11. The spectrum from Cr6+ with interference from Fe3+.

(mf(2tlbd

4

t

Scheme 1. The theory to determination of Cr6+.

esium chloride (MgCl2) was added in the process of extraction,hich suppressed the oxidation and minimized other ion interfer-

nce.

.5. A study of oxidation of Cr(III) to Cr(VI)

In order to investigate the potential oxidation of Cr(III) to Cr(VI),n extraction was carried out according to the following procedure.irst, a 400 mg kg−1 K2Cr2O7 standard solution was reduced bydding a sufficient volume of 0.93 M Na2SO3 solution, the reactions shown in equation (6).

2K+ + Cr2O72− + 8H+ + 6Na+ + 3SO3

2−

→ 2K+ + 6Na+ + 3SO42− + 2Cr3+ + 4H2O (6)

After Cr(VI) was turned into Cr(III) completely, the solution washen digested according to the proposed alkali extraction proce-ure as shown in Section 2.4. The results obtained show that theoncentration of Cr(VI) was below the detection limit of the methodsed. This meant that the percentage of oxidation of Cr(III) is below.005% which was in agreement with a reported result in litera-ure [29] where it was stated that the presence of alkali extractiongents minimized the possibility of the oxidation of Cr(III) to Cr(VI).n the other hand, another experiment showed that the possibleonversion of Cr(VI) to Cr(III) is small only if the pH in the extractionrocess is kept at 7–9.

.6. Stability of the Cr(III)–DPCO complex

DPC is a good colorant to identify chromium(VI) quantitatively.he Reactions A and B are the theoretical basis of the formation ofhe red–violet color complex [30]. The cationic chelated resultantCrDPCO(3−n)+] is not known because of an unknown number ofrotons (n) [31].

However, the reactions are very sensitive and selective becausether metal ions and chromium ions in other valent states can-ot form magenta with DPC. In order to study the stability of ther(III)–DPCO complex during working days, five samples (i.e. elec-roplating layer of steel (solution (1)), leather coat (solution (2)),hite paint (solution (3)), catalyst agent (solution (4)), resistor

solution (5))) were processed as shown in the proposed extractionethod, and the Cr(III)–DPCO magenta solutions formed with dif-

erent concentrations were kept in conditions of a relative humidityRH) in the range of 60–90%, and a temperature range of 5–40 ◦C for4 h. The results are shown in Fig. 12. From this graph, it can be seenhat the Cr(III)–DPCO complex has good stability during this ana-ytical time. Thus it is feasible to use DPC chelated with Cr(VI) as theasis of the most common spectrophotometric method for Cr(VI)etermination for EE products and components.

.7. Linearity and sensitivity

Under the optimum conditions described above, the calibra-ion curve of UV–Vis spectrometry in the concentration range of

Scheme 2. The magenta chelate complex.

L. Hua et al. / Journal of Hazardous M

Fig. 12. The absorbed intensities of the Cr(III)–DPCO complex as a function of time.

0c0ltCwrmts

5

oasitgral5

itimFactrhaIt

A

Wat(

R

[

[

[[

[[[[

[[

[[

[

[23] J.-F. Jen, G.-L. Ou-Yang, C.-S. Chen, S.-M. Yang, Analyst 118 (1993) 1281.

Fig. 13. The calibration curve of Cr(VI).

–0.5 mg l−1 is linear (see Fig. 13). The curve was obtained by pro-essing a series of K2Cr2O7 standard solutions containing 0.05,.1, 0.2, 0.4 mg l−1 Cr6+ with the proposed procedure. From the

ine, calibration constants were obtained based on the exponen-ial regression equation for y = Ae−bx where x, concentration ofr6+. The zero point was obtained by screening ultrapure distilledater. The correlation coefficient or the slope of line (m = Ae−b)

eflects the sensitivity of the chelating efficiency in the color deter-ination of Cr(VI). A correlation coefficient of 0.999924 showed

hat the analysis procedure has reliable sensitivity and preci-ion.

. Conclusions

Pointing to the pressure in complying with the regulationf hazardous substances in electronic and electrical equipment,multiple technique analytical method for the separation and

pecification of Cr(VI) restricted by RoHS directive is displayedn this paper. Because of the combined application of multipleechniques, the selectivity and validity of extracting Cr6+ werereatly improved, the interconversion between Cr6+ and Cr3+ was

educed greatly, the interference of other ions was minimized,nd the resultant Cr(III)–DPCO complex can be kept stable for aong time in the RH range of 60–90%, and temperature range of–40 ◦C.

[

[

[

aterials 163 (2009) 1360–1368 1367

Several parameters which influence the sensitivity and selectiv-ty of Cr6+ were studied, the result showed that when the pH is equalo 2.0, the extraction time is 60 min, the temperature of extractions 90 ◦C, the pH of extraction solution is 7.5–8.5, and the 1 ml of a

ixed phosphoric buffer solution is added, this is optimal. Althoughe3+ interferes in the specification of Cr6+, NaOH and MgCl2 weredded during the extraction, and the problem was solved. The Crompound was determined according to the proposed procedure,he R.S.D. was less than 1%, the calibration curve was linear, the cor-elation coefficient was high, so it can be concluded that the methodas high selectivity and reliable precision. It is also inexpensivend not complex, comparing to other determination techniques.t is expected to have a promising future for RoHS compliantests.

cknowledgements

The authors would like to acknowledge the Institute of RoHS andEEE Test Technology for providing the experimental conditions

nd the relative equipment, and the financial support provided byhe FSE funded project on Green Chemistry and RoHS applicationsproject no. 9610020) in City University of Hong Kong.

eferences

[1] Directive 2002/95/EC of the European Parliament and of the Council on therestriction of the use of certain hazardous substances in electrical and electronicequipment. Off. J. Eur. Union 1, 27 (2003) 19–37.

[2] J. Hathaway, N.H. Proctor, J.P. Hughes, et al., Proctor and Hughes Chemical Haz-ards of the Workplace, 3rd ed., Van Nostrand Reinhold, New York, 1991, p.178.

[3] P.J. Sheehan, R. Ricks, et al., Am. Ind. Hyg. Assoc. J. 53 (1992) 57.[4] D.J. Paustenbach, P.J. Sheehan, et al., Toxicol. Ind. Health (1991) 159.[5] S. Langard, T. Norseth, in: L. Friberg, et al. (Eds.), Handbook on the Toxicol-

ogy of Metals, vol. II, Specific Metals, 2nd ed., Elsevier, Amsterdam, 1990,pp. 185–210.

[6] T.M. Florence, G.E. Batley, CRC Crit. Rev. Anal. Chem. (1980) 219–267.[7] E. Nieboer, S.L. Shaw, In chromium in the natural and human environments, in:

J.O. Nriagu, E. Nieboer (Eds.), Advances in Environmental Science and Technol-ogy, vol. 20, Wiley, New York, 1998, pp. 399–442.

[8] E.R. Weiner, A Practical Guide for Environmental Professionals on Cr, CAS#7440-47-3, CRC Press LLC, Lewis Publishers, 2000, 216–217.

[9] P. Russo, A. Catassi, A. Cesario, A. Imperatori, N. Rotolo, M. Fini, P. Granone,L. Dominioni, Molecular mechanisms of hexavalent chromium-induced apop-tosis in human bronchoalveolar cells, Am. J. Respir. Cell Mol. Biol. 33 (2005)589–600.

10] R. Ganeshjeevan, R. Chandrasekar, S. Yuvaraj, G. Radhakrishnan, J. Chromatogr.A 988 (2003) 151.

11] E. Skrzydlewska, M. Bdlcerzak, F. Vanhaecke, Anal. Chim. Acta 479 (2003)191.

12] K. Ohta, H. Uegomori, S. Itoh, T. Mizuno, Microchem. J. 56 (1997).13] M.V. Balarama Krishna, K. Chandrasekaran, et al., Speciation of Cr(III) and Cr(VI)

in waters using immobilized moss and determination by ICP-MS and FAAS,Talanta 65 (2005) 135–143.

14] G.-L. Ou-Yang, J.-F. Jen, Anal. Chim. Acta 279 (1993) 329.15] Z. Gao, K.S. Siow, Electroanalysis 8 (1996) 602.16] D.W. King, Environ. Sci. Technol. 32 (1998) 2997.17] L. Bidstrup, R. Wagg, in: L. Parmeggiani (Ed.), Encyclopedia of Occupational

Health and Safety, 3rd ed., International Labor office, Geneva, Switzerland, 1989,p. 470.

18] A.I. Zouboulis, K.A. Kydros, K.A. Matis, Water Resour. 29 (7) (1995) 1755.19] S. Jambunathan, P.K. Dasgrpta, Determination of hexavalent chromium in

leather extracts by the diphenylcarbazide (IU C-18) procedure: pitfalls andrefinements. JSL TC, 84, 63.

20] J. Kotas, Z. Stasicka, Environ. Pollut. 107 (2000) 263.21] L.V. Mulaudzi, J.F. van Staden, R.I. Stefan, Determination of chromium(III) and

chromium(VI) by use of a spectrophotometric sequential injection system, Anal.Chim. Acta 467 (2002) 51–60.

22] Environmental standardization for electrical and electronic products and sys-tems, International Electro-technical Commission. IEC 62321, 1st ed., Italy, pp.43–45 (Committee Draft).

24] G.J. Willems, N.M. Bla ton, O.M. Peeters, C.J. de Ranter, Anal. Chim. Acta 88 (1997)345.

25] K. Ashley, A.M. Howe, M. Demange, O. Nygren, J. Environ. Monit. 5 (2003)707.

26] C.P. Karwas, J. Environ. Sci. Health A 30 (1995) 1223.

1 ous M

[

[

368 L. Hua et al. / Journal of Hazard

27] M. Pettine, S. Capri, Digestion treatments and risks of Cr(III)–Cr(VI) intercon-versions during Cr(VI) determination in soils and sediments – a review, Anal.Chim. Acta 540 (2005) 231–238.

28] S. Yalcin, R. Apak, Chromium (III, VI) speciation analysis with preconcentra-tion on a maleic acid-functionalized XAD sorbent, Anal. Chim. Acta 505 (2004)25–35.

[

[

[

aterials 163 (2009) 1360–1368

29] M. Korolczuk, M. Grabarczyk, Evaluation of ammonia buffer containingEDTA as an extractant for Cr(VI) from solid samples, Talanta 66 (2005)1320–1325.

30] J. Foit, M. Cuadros, M.R. Reyes, Influence of various factors on chromium (VI)formation by photo, JSLTC 83 (1999) 300–306.

31] E. Castillo, et al., Anal. Chim. Acta 464 (2002) 15–23.