Talanta 52 (2000) 539–544

Extraction spectrophotometric determination of vanadium innatural waters and aluminium alloys using pyridyl azo

resorcinol (PAR) and iodo-nitro-tetrazolium chloride (INT)

K. Gavazov *, Zh. Simeonova, A. AlexandrovDepartment of General and Inorganic Chemistry, Plo6di6 Uni6ersity, 24 Tsar Assen, 4000 Plo6di6, Bulgaria

Received 21 September 1999; received in revised form 21 March 2000; accepted 3 April 2000

Abstract

Extraction-spectrophotometric methods are developed for the determination of vanadium content in natural watersand aluminium alloys. They are based on the formation and subsequent extraction into chloroform of the ternary ionassociation complex of V(V) with PAR and INT in the presence of CDTA and NH4F as masking agents. OptimumpH range of the reaction is 5.5–7.5. Maximum absorbance of the extracted complex is at 560 nm. The method fordetermination of V(V) in drinking waters can be successfully applied at a concentration level of 3 ppb and higherwithout additional pre-concentration. Among studied more than 30 foreign ions potentially present in natural watersonly Ca(II) can interfere. It is removed by precipitation as CaF2 and filtration. A 40-fold excess of V(IV) does notinterfere with determination of V(V) and can also be determined indirectly (after oxidation to V(V)). The proposedmethod is applied to analysis of model mixtures as well as to the analysis of tap and mineral waters. Beer’s law isobeyed for up to 15 mg of V(V) in 40 ml aqueous phase. The accuracy and precision are reasonable. The RSD is inthe range 6.5–23.2% for determination of 6.3 ppb V(V). The procedure for analysis of aluminium alloys differs fromthe procedure for analysis of waters by the order of introduction of the reagents. The macrocomponent does notinterfere and is not separated. Mg, Mn, Cu, Zn, Fe, Cr, Ti and Zr do not interfere. A 25-fold excess of Ni interferes.The method is tested in the analysis of reference standards containing 0.005 and 0.007% V, respectively. The RSD is1.4%. © 2000 Elsevier Science B.V. All rights reserved.

Keywords: Vanadium-PAR complex; Extraction-spectrophotometry; Water analysis; Aluminium base alloys analysis

www.elsevier.com/locate/talanta

1. Introduction

Vanadium is of importance for living organ-isms. Its role in physiological systems includes an

insulin like effect, inhibitory effect on some en-zymes and cholesterol synthesis, catalytic effect inoxidation of various amines [1,2]. Vanadium af-fects favourably the diuretic kidney function, thecardiac muscle and cell growth. Antidiabetic func-tion of vanadium is a subject of intense studies [2].It has been established that drink water contain-

* Corresponding author. Tel.: +395-32-261420.E-mail address: kgavazov@ulcc.pu.acad.bg (K. Gavazov)

0039-9140/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.

PII: S0 039 -9140 (00 )00405 -7

K. Ga6azo6 et al. / Talanta 52 (2000) 539–544540

ing 25–75 ppb vanadium normalises the sugarlevel of rats suffering from diabetes [3]. Traces ofvanadium in drinking water are suggested to af-fect favourably the human health, while higherlevel of vanadium content causes toxification. Sev-eral levels of tolerance have been set in differentcountries. In this country the tolerable level ofvanadium in drinking water is 100 mg l−1. It isknown that the toxicity of vanadium compoundsdepends on the oxidation states of the metal form[1].

PAR is known to be a suitable organic reagentfor spectrophotometric determination of V(V),particularly in the presence of CDTA as a mask-ing agent. The methods using these reagents [4–13] are very advantageous for the determinationof vanadium in complex matrices, but in many ofthese methods difficulties arise due to: (i) the lowselectivity towards U(VI), Ti(IV), Zr(IV), Sn(IV),Ti(III), Nb(V), etc.; (ii) the delay in the colourdevelopment; (iii) significant absorbance for thereagent blank solution particularly when concen-trated solutions of the reagents are used.

Reference [14] reports the formation and char-acteristics of the ternary ion-association complexbetween V(V) anionic chelate with PAR and INT.Masking of interfering ions with CDTA and withCDTA+NH4F is studied as well and it is estab-lished that in the presence of additional maskingagent the reaction runs selectively, the colourbeing immediately developed and the absorbanceof the blank kept low. The coloured complex hasa high molar absorptivity and allows a reliabledetermination of traces of V(V). A 40-fold excessof V(IV) does not interfere, thus making thereaction favourably competing with other meth-ods using PAR and CDTA [1,13].

The present paper aims at developing of com-petitive methods for the determination of vana-dium in drinking waters and in aluminium basealloys.

2. Experimental

2.1. Reagents and apparatus

� Ammonium vanadate of analytical grade

(Apolda), 5×10−3 mol l−1 aqueous solution.Working solutions (1.5×10−5 mol l−1) wereprepared by dilution.

� 4-(2-Pyridylazo)resorcinol disodium salt(Sigma), 2×10−3 mol l−1 aqueous solution.

� 2-(4-Iodophenyl-3- (4-nitrophenyl) -5-phenyl-tetrazolium chloride (Fluka), 3×10−3 mol l−1

aqueous solution.� 1,2-Diaminocyclohexane-N,N,N%,N%-tetraacetic

acid (CDTA) (BDH), 0.125 mol l−1 solutionprepared as follows: hot solution of 7.5 g ofKOH (Fluka) in about 50 ml water were addedto 21.65 g of CDTA. The solution was thentransferred into a 500-ml calibrated flask, di-luted to volume with water and allowed to stayfor a week.

� Ammonium fluoride (NH4F) (pure, Reachim),3.5 mol l−1 aqueous solution.

� Sodium bromate (NaBrO3) (Carlo Erba), 0.15mol l−1 aqueous solution.

� Aluminium sulfate (Al2(SO4)3.18H2O) of ana-lytical grade (Khimsnab). 30.87 g of the saltwere diluted to volume in a 250-ml calibratedflask. 1 ml of the resulting solution contained10 mg Al(III).

� Acetate buffer solution with pH 6.75 preparedby mixing of 0.1 mol l−1 solutions ofCH3COOH and CH3COONa (freshly pre-pared) in a 1:99 v/v ratio.

� Chloroform of analytical grade was addition-ally distilled.

� Specol spectrophotometer (Carl Zeiss, Ger-many) with 10 mm cells was used through-out.

2.2. Procedures

2.2.1. Procedure for the determination of6anadium(V) in waters

An aliquot of the analysed water (up to 27 ml;0.075–15 mg V(V)) (pH was adjusted to 2–3 atthe sampling site) was placed in a 125-ml separa-tory funnel and 3 ml of the NH4F solution wereadded. If the Ca content in the aliquot exceeds 1.5mg (60 mg l−1) (see below) the mixture wasfiltered through a filter paper with wide pores.Then 2 ml of the CDTA solution, 5 ml of acetate

K. Ga6azo6 et al. / Talanta 52 (2000) 539–544 541

buffer, 1 ml of the PAR solution and 3 ml of theINT solution were introduced in the mentionedorder. The aqueous phase was diluted with waterto a total volume of 40 ml, 10 ml of chloroformwere added and the phases shaken for 2 min. Aportion of the organic layer was transferred intothe cell through a paper filter, and absorption wasmeasured at 560 nm against reagent blank. Theamount of V(V) was read from a calibrationgraph plotted under similar conditions.

2.2.2. Procedure for the determination of6anadium in aluminium alloys

0.5 g of the aluminium alloy (0.001–0.023% V)were placed in a 100-ml beaker and dissolved withabout 30 ml of 40% H2SO4 on heating. When thereaction has ceased and all aluminium was dis-solved few drops of conc.HNO3 were added, thewalls of the beaker were rinsed with distilledwater and the heating was continued for 10–15min. If the solution was not clear it was filteredand the residue on the filter was carefully washed.The filtrate and the washings were collected in a50-ml calibrated flask and the resulting solutiondiluted to volume. An aliquot of the final solution(5 ml; 0.5–11.5 mg V) was placed in a separatoryfunnel and pH carefully adjusted to 2–3 by addi-tion of NH4OH. The solution was cooled and 2ml of NaBrO3 solution was added and mixed well.Then 5 ml of acetate buffer, 1 ml of the PARsolution, 2 ml of the CDTA solution, 3 ml of theINT solution and finally 3 ml of the NH4F solu-tion were added in the above cited order. Theaqueous phase (total volume 30 ml) was extractedwith 10 ml of chloroform and absorbance of theorganic layer immediately measured at 560 nmagainst blank solution containing the sameamount of Al(III) as the analysed aliquot. Cali-bration curve was prepared in the presence of thesame amount of Al(III) (e.g. 5 ml of the Al2(SO4)3

solution).

3. Results and discussion

Most of the available methods for determina-tion of vanadium in waters [4,8,9,15–27] and inaluminium alloys [28–36] require a preliminary

separation of vanadium, or are inapplicable toanalysis of low levels of vanadium. The proposedprocedure in this work is sensitive, simple and fastand allows direct analysis of vanadium content.

3.1. Determination of 6anadium in waters

Absorbance of the organic extract is constantfor 15–20 min. Beer’s law is obeyed in the range3–600 mg V(V) l−1 analysed sample (i.e. 0.075–15mg V(V) in 40 ml aqueous phase). Apparent molarabsorptivity is equal to 2.4×104 l mol−1 cm−1.The optimum pH range for formation of the ionassociate is 5.5–7.5. Absorbance value of 0.001was measured for 0.84 mg V(V) l−1 and a 25-mlaliquot. Absorbance of the blank was 0.025, stan-dard deviation SD=5×10−4 (five replicates).

The recommended relatively large volume ofthe aqueous phase allows a larger aliquot to beanalysed and ensures an enhanced sensitivityagainst some ions, e.g. Cl−, NO3

−, Cr(III), Cr(VI),Mn(II), Mo(VI) and Co(II).

The presence of larger amount of Ca than thosementioned above prevents the separation ofphases and causes a decreases in the light ab-sorbance. The observed impeded separation ofphases is an indication that the aqueous layershould be preliminarily filtered. Another way toeliminate the interference of calcium is to workwith smaller aliquots, since the correct determina-tion of V(V) in the presence of Ca does notdepend on the V(V) to Ca ratio, but on calciumcontent (see Table 1). This approach can be ap-plied only in the case the vanadium(V) content ishigh enough.

The proposed procedure allows as well the de-termination of V(IV) by the difference betweenthe total vanadium content and vanadium(V) con-tent. The determination of the sum V(V)+V(IV)requires addition of NaBrO3 (2 ml) for oxidationof V(IV) to V(V) [37].

Table 2 summarises the results of analysis ofsynthetic mixtures. The relative error does notexceed 10% even with high concentrations of di-verse ions which are rarely present in potablewaters. Analytical blank is very important in de-termination of low levels of vanadium content. Itwas evaluated through analysis of model solutions

K. Ga6azo6 et al. / Talanta 52 (2000) 539–544542

Table 1Effect of diverse ions on the extraction of 2.5 mg V(V)

Recovery (%)mg Ion:V(V)Foreign ion

400:11000 103.3Ag+

96.72000:15000Al3+

102.5Ba2+ 25 000 10 000:1a

97.110 000:1Br− 25 000100.0BrO3

− 50 000 20 000:1a

100.01875 750:1Ca2+

95.01500 1500:1b

95.02800:1c1400

100.0Cd2+ 5000 2000:195.04000:110 000

104.2Cl− 12 500 5000:195.52000:15000CO3

2−

Co2+ 1000:1 97.5250095.51500:13750

97.0Cr3+ 625 250:195.0500:11250

100.0Cr6+ 625 250:195.5400:11000

98.2Cu2+ 3750 1500:1100.020 000:1a50 000F−

Fe3+ 400:1 95.01000100.0100:1250

100.0HCO3− 18 750 7500:1

95.010 000:125 000

HPO42− 5000:1 100.012 500

100.020 000:1a50 000K+

20 000:1a50 000 104.0Mg2+

5000 2000:1 95.5Mn2+

104.03000:17500Mo6+

100.0Nb5+ 25 10:190.050:1125

100.0NH4+ 50 000 20 000:1a

95.07500:1NO3− 18 750

1500:13750 97.0Ni2+

100.07500:1Pb2+ 18 75015 000:1 96.037 500

95.41250 500:1Re7+

100.0Sn4+ 250 100:1a

100.020 000:1aSO42− 50 000

20 000:1a50 000 99.0Tl3+

95.475:1187.5Ti4+

750:11875 97.5U6+

100.0V4+ 25 10:1105.040:1100

99.5W6+ 7.5 3:1

97.55000 2000:1Zn2+

95.06250 2500:1

a Higher ion to V(V) ratios not studied.b Extracted 1 mg V(V).c Extracted 0.5 mg V(V).

of the same composition as synthetic mixturesnumbers 2, 3, 8 and 9 from Table 2 in the absenceof V(V). It was established that the light ab-sorbance values of model solutions were indistin-guishable from the blank absorbance.

Table 3 lists data about vanadium content inmineral and tap waters. They are compared withdata obtained by ICP-MS. Vanadium content ob-tained by two methods do not show statisticallysignificant difference. It is assumed that in suchwaters vanadium is present mainly as V(V) species[4,38], the latter being confirmed by the presentstudies: data obtained in the presence and in theabsence of NaBrO3 are statistically indis-tinguishable.

3.2. Determination of 6anadium in aluminiumalloys

The order of addition of the reagents that isvalid for determination of vanadium in waters isnot suitable for the analysis of aluminium alloyssince a precipitation appears prior to the colourdevelopment. In the procedure for aluminium al-loys the order of addition of the reagents ischanged. The described procedure (see Section 2)can be applied in the analysis of aluminium alloyscontaining 0.001–0.023% m/m V. For V contentsout of the mentioned range a correction of theused aliquot or of the sample weight is required.Table 4 lists data about V content in aluminiumalloys analysed according to the proposed proce-dure. Fig. 1 gives an impression about the selec-tivity of the method towards elements (E) usuallypresent in aluminium alloys. It is plotted in thefollowing way: ternary synthetic mixtures wereprepared and analysed according to the developedprocedure. The volume of each mixture was 5 ml.It contained a constant amount of V-2.5 mg (i.e. Vcontent in the mixture is 0.005%) and varyingAl:E ratios but a constant sum of Al+E=50 mg.It can be concluded that the proposed procedureis not suitable for analysis of alloys containingconsiderable amounts of Ni (allowed Ni:V ratio is25). The content of remaining elements in Alalloys is in the range that does not affect thecorrect determination of vanadium.

K. Ga6azo6 et al. / Talanta 52 (2000) 539–544 543

Table 2Determination of vanadium in synthetic mixtures, containing 40 mg l−1 V(V)

Mixture Vanadium foundComposition (mg l−1)(mg l−1 (n=4))

40.390.8Ca2+(50), Mg2+(320), Na+(90), NH4+(80), HCO3

−(240), Cl−(160), SO42−(1400)1

Ca2+(50), Mg2+(320), Na+(90), NH4+(80), HCO3

−(240), CO32−(100), Cl−(160), SO4

2−(1400)2 39.890.8Ca2+(70), Mg2+(80), Na+(30), NH4

+(40), HCO3−(80),Cl−(80), SO4

2−(350)3 3892b

4391aCa2+(150), Mg2+(100), Na+(152), HCO3−(200),CO3

2−(100), Cl−(266), SO42−(400)4

5 4093aCa2+(150), Na+(152), Cl−(266), HCO3−(200), CO3

2−(100)4091Ca2+(50), Mg2+(50), Na+(152), SO4

2−(200), Cl−(266), HCO3−(200), CO3

2−(100)6Ca2+(50), Mg2+(150), Na+(175), K+(20), Zn2+(3), Ti4+(1), Co2+(1), Al3+(1), Cd2+(0.1), 40917Ni2+(0.4), Mo6+(0.5), Ag+(0.06), Pb2+(0.8), Mn2+(0.3), Fe3+(0.3), Cr6+(0.2), U6+(0.6),HCO3

−(200), CO32−(100), Cl−(89), NO3

−(60), HPO42−(0.5), SO4

2−(855), Br−(41)8 36.090.8Ca2+(70), Mg2+(70), Na+(175), K+(20), Zn2+(3), Ti4+(1), Co2+(1), Al3+(1), Cd2+(0.1),

Ni2+(0.4), Mo6+(0.5), Ag+(0.06), Pb2+(0.8), Mn2+(0.3), Fe3+(0.3), Cr6+(0.2), U6+(0.6),HCO3

−(200), CO32−(100), Cl−(89), NO3

−(60), HPO42−(0.5), SO4

2−(855), Br−(41)Ca2+(150), Mg2+(150), Na+(175), K+(20), Zn2+(3), Ti4+(1), Co2+(1), Al3+(1), Cd2+(0.1),9 44.090.9a

Ni2+(0.4), Mo6+(0.5), Ag+(0.06), Pb2+(0.8), Mn2+(0.3), Fe3+(0.3), Cr6+(0.2), U6+(0.6),HCO3

−(200), CO32−(100), Cl−(266), NO3

−(60), HPO42−(0.5), SO4

2−(855), Br−(41)

a Filtered.b 20 ml aliquot.

Table 3Determination of vanadium in watersa

Sample Vanadium found (mg l−1)Settlement

Present method RSD (%) ICP-MS RSD(%)

1b Bratsigovo 6.3 7.5 5.9 3.84.0 11.92c 3.7Plovdiv 4.33.8 19.5Gorna Banya 3.13b 4.9

Devin4b,d 2.3 6.5 2.5 7.9Markovo5c,d 2.2 15.0 2.0 9.5

4.2 23.2 3.7Narechenski bani 5.76b,d

a n=5.b Mineral water.c Tap water.d Pre-concentrated through evaporation.

4. Conclusion

Methods for determination of vanadium con-tent in waters and aluminium alloys are devel-oped. They allowed a direct, low cost, rapid andprecise analysis of vanadium to be performed.

Acknowledgements

The authors are indebted to senior assistantV.M. Stefanova for the help in evaluation of theaccuracy of the method and to senior assistantV.I. Kmetov for the assistance.

K. Ga6azo6 et al. / Talanta 52 (2000) 539–544544

Table 4Determination of vanadium in aluminium alloysa

Vanadium content (%)Sample

Found RSD (%)Certified value

1b 0.005 0.005090.0001 1.40.0072c 0.007190.0001 1.4

a n=5.b Other components (%): Cr(0.002), Cu(0.012), Fe(0.115),

Mg(0.060), Mn(\0.1), Zn(0.22).c Other components (%):Cr(0.006), Cu(0.024), Fe(0.206),

Mg(0.112), Mn(\0.1), Zn(0.012).

[12] A. Yerramilli, C.S. Kavipurapu, R.R. Manda, C.M. Pil-lutla, Anal. Chem. 58 (1986) 1451.

[13] K. Gavazov, Zh. Simeonova, A. Alexandrov, Anal. Lab.7 (1998) 127.

[14] Zh. Simeonova, K. Gavazov, A. Alexandrov, Zh. Neorg.Khim. 44 (1999) 676.

[15] M. Ishibashi, T. Shigematsu, T. Nakagava, Y. Ishibashi,Bull. Int. Chem. Res. Kyoto Univ. 24 (1951) 68.

[16] K. Sugawara, M. Tanaka, H. Naito, Bull. Chem. Soc.Jpn. 26 (1953) 417.

[17] H. Naito, K. Sugawara, Bull. Chem. Soc. Jpn. 30 (1957)799.

[18] E.A. Morgen, L.M. Dimova, Zh. Anal. Khim. 38 (1983)2181.

[19] M.A. Shriadah Mohamed, O. Kunio, Analyst 110 (1985)677.

[20] E.J. Marchevsky, R.A. Olsina, E. Kamachi, C.B.Marone, An. Quim. R. Soc. Esp. Quim. B81 (1985)342.

[21] Y. Yang, Y. Chen, Yankuang Ceshi 7 (1988) 171.[22] S. Nakano, M. Tago, T. Kawashima, Anal. Sci. 5 (1989)

69.[23] S. Nakano, S. Kinoshita, M. Ikuta, T. Kawashima, Anal.

Sci. 6 (1990) 435.[24] K. Hirayama, S. Kageyama, N. Unohara, Analyst 117

(1992) 13.[25] F.B. Serrat, G.B. Morell, Fresenius J. Anal. Chem. 349

(1994) 717.[26] A. Abebaw, B.S. Chandravanshi, Chem. Anal. (Warsaw)

43 (1998) 33.[27] J.M. Bosque-Sendra, M.C. Valencia, S. Boudra, Frese-

nius, J. Anal. Chem. 360 (1998) 31.[28] L. Erdey, F. Szabadvary, Fresenius Z. Anal. Chem. 159

(1957) 429.[29] D.G. Biechler, D.E. Jordan, W.D. Leslie, Anal. Chem. 35

(1963) 1685.[30] V.G. Antonievich, Glasnik Khem. Droushtva (Belgrade)

31 (1966) 305.[31] E.M. Donaldson, Talanta 17 (1970) 583.[32] L.I. Ganago, L.N. Boukhteeva, Zh. Anal. Khim. 32

(1977) 1537.[33] L. Ping, J. Pan, L. Jianyan, Huaxue Fence, 24 (1988) 159;

Chem. Abstr. 113 (1990) 51658e.[34] J.H. Wang, R.H. He, Anal. Chim. Acta 276 (1993)

419.[35] M. Jarosz, S. Oszwaldowski, Chem. Anal. 38 (1993)

109.[36] C. Zucchi, M. Forneris, L. Martinez, R. Olsina, E.

Marchevsky, Fresenius, J. Anal. Chem. 360 (1998) 128.[37] W.F. Hillebrand, G.E.F. Lundell, Prakticheskoe

rukovodstvo po neorganicheskomu analizu, in: Yu.Yu.Lur’e (Ed.), Gosudarstvennoe nauchno-tekhnicheskoe iz-datel’stvo khimicheskoi literatury, Moscow, 1960, p. 472.

[38] V.M. Goldsmith, in: A. Muir (Ed.), Geochemistry, Ox-ford, 1954, p. 491.

Fig. 1. Relationship between the relative error in determina-tion of 0.005% V and the content of some diverse elements(E,%).

.

References

[1] M.J.C. Taylor, J.F. Staden, Analyst 119 (1994) 1263.[2] Vanadium Symposium 1994, 29–31 July, Montreal,

Canada, Can. J. Physiol. Pharm. 72 (1994) No. 3.[3] A. Chakraborty, M. Chatterjee, Can. J. Physiol. Pharm-

col. 72 (1994) 21.[4] G. Chakrapani, D.S. Murty, B.K. Balaji, R. Ran-

gaswamy, Talanta 40 (1993) 541.[5] O. Budevsky, L. Johnova, Talanta 12 (1965) 291.[6] T. Yotsuyanagi, J. Ito, K. Aomura, Talanta 16 (1969)

1611.[7] T. Yotsuyanagi, J. Itoh, K. Aomura, Mem. Fac. Eng.

Hokkaido Univ. 13 (1971) 39.[8] T. Kiriama, R. Kuroda, Anal. Chim. Acta 62 (1972) 464.[9] M. Nishimura, K. Matsunaga, T. Kudo, F. Obara, Anal.

Chim. Acta 65 (1973) 466.[10] N. Grigorova, L. Krasteva, I. Rasheva, Metalurgiya

(Sofia) 6 (1983) 22.[11] J. Chwastowska, E. Kosiarska, Chem. Anal. (Warsaw) 30

(1985) 395.