-
Talanta 50 (1999) 247252
A flow system with an electrochemical reduction cell
forspectrophotometric determinations of major constituents in
Fe:V alloys
Antonio O. Jacintho a, Luiz F.R. Machado b, Amauri A. Menegario
a,Iolanda A. Rufini a, Maria Fernanda Gine a,*
a Centro de Energia Nuclear na Agricultura, Uni6ersidade de Sao
Paulo, A6. Centenario 303, C. Postal 96,13400-970 Piracicaba,
Brazil
b Instituto de Qumica de Sao Carlos, Uni6ersidade de Sao Paulo,
Sao Paulo, Brazil
Received 22 July 1998; received in revised form 29 September
1998; accepted 2 October 1998
Abstract
Insertion of an electrochemical cell in a flow injection system
to evaluate the on-line reduction of ionic species ispresented. The
cell comprised Pt electrodes installed in two sections separated by
a Nafion membrane. The samplewas injected into an acidic carrier
stream and passed through the cathode compartment of the
electrolytic chamberwhere the species were reduced as consequence
of an applied DC voltage. The sample solution leaving the
cellreceived a confluent reagent stream (1,10-phenanthroline
buffered at pH 4.7) and the reacted products were droppedoff in an
open tube for gas:liquid separation. Efficiency of the Fe3 to Fe2
reduction in acidic medium wasevaluated in the presence of strongly
reducing species of V and Mo by monitoring the Fe(II) colored
complex.Interferences from Pb2, Co2, Ni2, Zn2, Cu2, V5 and Mo6 were
evaluated. Production of stronglyreducing species of V at the
electrolytic cell presented higher efficiency for Fe reduction than
the electrolytic chamberitself. Total reduction of Fe3 in solutions
containing up to 10 mg l1 Fe plus 100 mg l1 V or 100 mg l1 of Mowas
achieved by the electrolytic process at 2 A. The quantitative
determination of Fe and V in low silicon Fe:V alloyswas achieved.
Accuracy was assessed with the certified Euro-standard 577-1
ferrovanadium alloy produced by theBureau of Analysed Samples
Limited and no difference at the 95% confident level was found.
1999 Elsevier ScienceB.V. All rights reserved.
Keywords: Electrochemical cell; Iron reduction; Flow analysis;
Iron alloys
1. Introduction
In flow injection analysis reducing columnshave been used for
different applications. Acolumn with copperized cadmium filings was
em-
* Corresponding author. Tel.: 55-194-292-4636; fax:
55-194-294-610.
E-mail address: [email protected] (M.F. Gine)
0039-9140:99:$ - see front matter 1999 Elsevier Science B.V. All
rights reserved.
PII: S0039 -9140 (99 )00023 -5
-
A.O. Jacintho et al. : Talanta 50 (1999) 247252248
ployed for reduction of nitrate to nitrite [1]. Sev-eral
applications used the Jones reductor of amal-gamated Zn packed in
an on-line column toreduce Cr3 to Cr2 [2,3], V4 and V5 to V2
[2] and U6 to U3 [4]. In these papers, thegeneration of strongly
reducing agents in flowsystems which are unstable under
atmosphericconditions was demonstrated. The determinationof total
Fe and Fe2 was also attained afterreduction in a column with the
Jones reductor ina flow system [5].
Electrochemical reduction of ionic species wasreported for the
hydride forming elements [6]. Theapproach was accomplished by using
a flow-through electrolytic cell [7,8]. Papers published onthe last
years described the electrochemical hy-dride generation (EcHG) of
selenides, arsines,stibines and others coupled to different
spectrom-eters, atomic absorption with a T-tube atomizers(ETAAS)
[7,8] or with graphite furnaces(GFAAS) [9] or atomic emission
microwave in-duced plasmas (MIP-AES) [10]. Electrodes of dif-ferent
materials: vitreous carbon [7], platinum[7,8,10], lead and
pyrolytic graphite [9] weredescribed.
In the present work the electrochemical ap-proach using a
flow-through cell is proposed toreduce Fe3 to Fe2. The
spectrophotometricdetection of the ferroine complex was used
toestimate the reduction process. The on line pro-duction of
reduced species of V and Mo and theeffect on reducing Fe3 was
evaluated. The deter-mination of Fe in ferrovanadium alloys and
thepossibility of the indirect determination of V ispresented.
2. Experimental
2.1. Apparatus
The flow system comprised a peristaltic pump(mp-13R Ismatec,
Zurich, Switzerland) providedwith Tygon pumping tubing, an
automatic injec-tor (Model 352, Micronal, Sao Paulo, Brazil)
aflow-through electrolytic cell described below, aspectrophotometer
(Model 432, Femto, SaoPaulo, Brazil) furnished with a flow cell (10
mm
optical path, 80 ml inner volume) and a strip-chartrecorder (REC
61, Radiometer, Copenhagen,Denmark). Accessories such as Y-shaped
connec-tors, and mixing coils of polyethylene tubing (0.8mm i.d.)
were used.
The electrochemical cell was constructed byassembling layers of
Perspex of different thick-ness, with two compartments, (anode 1.6
cm3 andcathode 0.4 cm3) separated by a Nafion mem-brane and
electrodes of platinum sheets (3.2 cm2)connected to a home made
electrical source,which could be varied from 0.2 to 3.0 A.
TheNafion membrane was conditioned by immersingit in hot water 20
min before installing at the cellprotected between two rubber
layers.
The gas phase separator (PS) was constructedfrom a 10 cm5 mm
i.d. glass tube, with twolateral exits at opposite sides as
described earlier[11].
2.2. Reagents and solutions
All reagents were of analytical grade and dis-tilled, de-ionized
water was always used. The elec-trolyte solution was 2.0 M H2SO4.
The1,10-phenanthroline solution (0.25% w:v) wasprepared weekly by
dissolving 0.6250 g of 1,10-phenanthroline hydrochloride in 250 ml
of 0.1 MHCl. Acetate buffer solution (pH 4.7) was pre-pared with
1.0 M ammonium acetate plus 1.0 Macetic acid.
Stock solution containing 1000 mg l1 Fe3
was prepared from Fe(NO)3 9H2O (Merck,Darmstadt, Germany).
Solutions of the individualconcomitants containing 1000 mg l1
Zn2,Co2 and Cr3 from the metals and Ni2 fromNiSO4 6H2O (Johnson and
Matthey ChemicalsJMC), Pb2 from Pb(NO3)2, V5 from NH4VO3and Mo6
from (NH4)6MO7O24 4H2O (Merck,Darmstadt, Germany) were prepared.
Solutionsused to test interference containing 10.0 mg l1
Fe3 plus 10.0 and 100 mg l1 of each elementwere used. A freshly
prepared solution containing10.0 mg l1 Fe2 was employed to estimate
thereduction efficiency.
Working standard solutions, 0.00, 2.50, 5.00and 10.0 mg l1 Fe3
plus 100 mg l1 of V5,were used to determine Fe in the diluted
sample.
-
A.O. Jacintho et al. : Talanta 50 (1999) 247252 249
Standard solutions, 0.00, 2.00, 5.00 and 10.0 mgl1 V or Mo plus
10.0, 20.0 and 30.0 mg l1 ofFe were prepared. Three synthetic
solutions con-taining Fe and V concentrations in different
pro-portions (50:50; 60:40 and 40:60 w:w) in 0.25 MHCl, simulating
alloy samples with low siliconcontents were analyzed together with
the standardreference material Euro-standard 577-1 producedby the
Bureau of Analysed Samples Limited(Middlesborough, UK). About 100
mg of thereference alloy were accurately weighted, received10 ml of
acqua-regia and were heated in a hotplate until complete
dissolution. After cooling toroom temperature 5 ml of perchloric
acid (70%v:v) were added and heated until evolution ofwhite fumes.
The residual solution was diluted to100 ml [12]. Aliquots of 5.0 ml
of this solutionreceived 50 ml of 2.5 M HCl and 50 ml of 1000mg l1
of V before diluting to 500 ml.
2.3. The flow system
The electrochemical reduction of Fe3 was per-formed by using the
flow set up depicted in Fig. 1.The electrolyte solution was
continuously recycled
through the anode compartment of the elec-trolytic cell (EC). In
the situation specified in Fig.1, the sample solution at loop L was
carried by Cthrough the EC cathode compartment and themixing coil
B1. At the confluent point X, thesample solution received a
buffered reagentstream (BR). After passing reaction coil B2
thesolutions was dropped off in an open tube PS toseparate gaseous
components. Part of this solu-tion was aspirated from the bottom of
the tubetowards the detector D with wavelength adjustedto 512 nm.
Moving down the central part of theinjector, L was filled with
another sample solutionwhile part of solution at the PS was
drained.After 30 s, the injector returned to the positionshown in
Fig. 1 and simultaneously the electricalpower at the EC was
switched on during 30 s. Theinjector rested in this position during
60 s and thespectrophotometric detection occurred.
Speciation of Fe2 and Fe3 in acidified watersamples (0.25 M HCl)
could be accomplished withthe flow system in Fig. 1. The injection
of thesample allowed the direct Fe2 determination andin a second
injected sample volume simulta-neously the EC is switched on and
the reduction
Fig. 1. A. Flow diagram for electrochemical reduction of Fe3 to
Fe2. The three rectangles at the left represent the injector inthe
injection position with sample in loop L. The arrow below represent
the displacement of the central part to the alternativeposition.
Lines indicate the tubing used for flowing solutions of electrolyte
E, carrier C, sample S, reagent R, buffer B, and theresidual W.
Numbers in brackets indicate the flow rates in ml min1, and arrows
indicate the pumping direction. Theelectrochemical cell (EC) is
schematized with the Nafion membrane N and the electrodes
connections. Other devices are the mixingand reaction coils B1 and
B2 of 25 cm tubing each, the connector X, the phase separator (PS)
and the detector D. Dashed line afterthe EC, indicates the optional
inlet to introduce the Fe3 solutions.
-
A.O. Jacintho et al. : Talanta 50 (1999) 247252250
Table 1Regression data obtained with solutions from 0.00 to 10.0
mgl1 Fe3 at electrolytic currents of 1.0, 2.0 and 3.0 A
Current (A) r2yaxb R.S.D.10Fea
0.99901.0 0.980.012x0.0022.0 0.018x0.001 0.9999 2.04
0.026x0.0043.0 0.9990 2.93
a Relative standard deviations (R.S.D.) for the 10 mg l1
Festandard based on three replications.
ber. Thus, further experiments were carried out at2 A yielding
results characterized by good linear-ity and precision.
Solutions containing 10.0 mg l1 of Fe3 plus10.0 and 100 mg l1 of
Pb2, Co2, Ni2,Zn2, Cu2, V5 and Mo6 were injected usingthe
electrolytic current switched on and off toverify the possibility
of being interfering either onreduction or on the complexation
reaction. Theeffects of concomitant additions are presented inTable
2. Cobalt solutions were colored and raisedthe background at 512
nm. Using a EC with cleanPt electrodes a signal increase of 10% due
to Pbaddition was observed. After passing repeatedlythe Pb solution
(total mass of 0.1 mg Pb) thesignal stabilized. When the EC was
opened, Pbdeposition on the Pt cathode could be observedby naked
eyes.
Addition of V5 and Mo6 increased the sig-nal proportionally to
the applied current empha-sizing the formation of strongly reducing
species[2].
The efficiency of reduction was evaluated atdifferent
electrolytic currents by comparing ab-sorbances of a 10 mg l1 Fe3
after reductionwith that corresponding to a 10 mg l1 Fe2
of Fe3 occurred permitting the determination oftotal iron.
The flow system in Fig. 1 presents an alterna-tive configuration
to introduce the Fe3 solutions(dashed line after the cell), to
study the effect ofthe reduced species of V and Mo produced at
EC.
3. Results and discussion
The efficiency of the proposed electrolytic re-duction of Fe3 to
Fe2 is dependent on severalparameters such as flow-rates, cell
compartmentdimensions, type and electrode surface, sampleacidity
and electrolytic voltage. One liter of theelectrolyte solution (2 M
H2SO4) was continu-ously pumped at 2.5 ml min1, through the an-ode
compartment. This solution was replacedperiodically to maintain its
characteristics. UnderDC current the electrolyte solution yielded
H3O
which crossed the Nafion membrane towards thecathode
compartment. The temperature at the ECincreased when the current
was raised just to 3 A,sample flow rate was lower than 1.0 ml
min1
and the sample acidity was B0.05 M HCl. Underthese conditions a
damage of the Nafion mem-brane was observed. On the other hand,
afterincreasing the sample acidity beyond 0.25 M HClthe large
quantity of gases evolved impaired theFe3 reduction. Therefore,
sample solutions at0.25 M HCl flowing at 2.5 ml min1 were em-ployed
throughout.
Effect of raising the electrolytic current at theEC to reduce
solutions of Fe3 produced data inTable 1. Despite higher
absorbances being at-tained with 3 A data presented poor
precisionprobably due to higher gas evolution at the cham-
Table 2Results after addition of different ionic species to 10
mg l1
Fe3a
Added amount (mg l1)Ions Absorbance
10 0.18890.003Cu2
100 0.19990.010Ni2 0.18990.00110
100 0.19190.0010.18990.00310Zn2
0.18390.0031000.17790.00110Mn2
100 0.17990.001Co2 0.32090.00310
100 0.40790.008Cr3 10 0.18990.006
100 0.19690.009Mo6 10 0.31590.005
100 0.47290.004V5 10 0.35390.005
100 0.47190.009
a The analytical signal related to 10 mg l1 Fe2 was0.18090.004 A
(n3).
-
A.O. Jacintho et al. : Talanta 50 (1999) 247252 251
Fig. 2. Effect of the electrolytic current and the presence ofV5
on Fe3 reduction (a) 10 mg l1 Fe3; (b) 10 mg l1
Fe3 10 mg l1 V5; and (c) 10 mg l1 Fe3 100 mgl1 V5. A is
absorbance. Dashed line indicates the signalobtained by injecting a
solution of 10 mg l1 of Fe2.
The expected probable competition of Fe2
and V2 for 1,10-phenanthroline [5] was not ob-served even though
a diluted solution of thereagent (0.1% w:v) and a concentrated
solution ofV5 (100 mg l1) reduced at 2 A was used.
The quantification of iron in dissolved fer-rovanadium alloys
(dilution factor of 105) wasachieved by adding 100 mg l1 V to the
samplesand standards, assuming complete reduction ofFe3 (Fig.
2).
The analytical curves in Fig. 3 were used for Vquantification.
The effect of increasing the V con-centrations up to 10 mg l1 in
solutions with 10,20 and 30 mg l1 of Fe was characterized bylinear
coefficients and slopes depending on theFe3 concentrations
following the relationshipA0.018 CFe0.0008 CFeCV. For the
dissolvedsamples, the Fe concentration previously deter-mined was
used to calculate the V concentration,but this is only valid for V
concentrations lowerthan 10 mg l1. When CV10CFe maximumabsorbance
was attained (Fig. 2). Thus, the pro-posal is to add Fe to the
diluted samples to ensurethe determination of V in the linear
analyticalrange (Fig. 3) for samples with a CFeBCV. Theeffect of
adding up to 10 mg l1 Mo to solutionsin three different iron
concentrations presentedgood correlation (Fig. 3 dashed lines). In
thissituation, the slope of the curves were not depen-
solution injected in the same way. Results arepresented in Fig.
2. No signal was observed forFe3 without current at the EC.
Increasing thecurrent up to 3 A the reduction of Fe3 aloneincreased
linearly (r0.9968) attaining 55% ofreduction. The effect of
reducing 10 mg l1 Fe3
alone or together with 0.00, 10.0 or 100 mg l1 ofV5 could be
appreciated by comparing curves atFig. 2. The addition of 100 mg l1
V5 to Fe3
provided the best reduction efficiency, attainingtotal reduction
after a current of 2 A (curve c,Fig. 2).
To elucidate the influence of V and Mo reducedspecies on the
reduction of iron, an experimentinvolving addition of a Fe3
solution after theEC was performed. Solutions of V5 and Mo6
were injected instead of the sample and whileflowing through the
EC (2 A) produced reducedspecies which merged with the Fe3
solution. Thereduction of 10 mg l1 Fe3 occurred in anextent of 80%
with 100 mg l1 V5 and 45% with100 mg l1 Mo6. Therefore, the reduced
Vspecies which formed from 100 mg l1 V5 at theEC cell at 2 A
allowed the attainment of 80%efficiency on Fe3 reduction, while the
efficiencyof the EC alone was 38% (curve a, Fig. 2). Nosignal
corresponding to the reduced species of Vand Mo was observed
probably due to their oxi-dation until reaching the detector
unity.
Fig. 3. Effect of adding different V (solid lines) and Mo(dashed
lines) concentrations to solutions containing 10.0, 20.0and 30.0 mg
l1 Fe3. A is absorbance and C is the concen-tration of V or Mo
added. Regression curves in bold corre-spond to V and those in
italic to Mo.
-
A.O. Jacintho et al. : Talanta 50 (1999) 247252252
Table 3Results of a reference standard ferrovanadium alloy and
Vresults in three synthetic FeV solutions containing 50, 40 and60%
w:w V
Fe (% w:w)Sample V (% w:w)
Certified valueEURO-ST5771 47.21 50.1690.1347.290.5Found valuea
50.191.8EURO-ST5771
1 52.191.738.990.9260.590.83
a Mean9S.D., n3.
ing both elements, once these elements are deter-mined
indirectly by its effect on iron reduction.
Acknowledgements
Financial support from FAPESP (Fundacao deAmparo a` Pesquisa do
Estado de Sao Paulo) andCNPq (Conselho Nacional de
DesenvolvimentoCientfico e Tecnologico) was greatly
appreciated.
References
[1] M.F. Gine, H.Fo6 . Bergamin, E.A.G. Zagatto, B.F. Reis,Anal.
Chim. Acta 114 (1980) 191.
[2] R.C. Schothorst, J.M. Reijn, H. Poppe, G. den Boef,Anal.
Chim. Acta 145 (1983) 197.
[3] R.C. Schothorst, G. den Boef, Anal. Chim. Acta 153(1983)
133.
[4] R.C. Schothorst, M. van Son, G. den Boef, Anal. Chim.Acta
162 (1984) 1.
[5] A.T. Faizullah, A. Townshend, Anal. Chim. Acta 167(1985)
225.
[6] L. Tomlinson, J. Electrochem. Soc. 111 (1964) 593.[7] Y.
Lin, X. Wang, D. Yuan, P. Yang, B. Huang, Z.
Zhuang, J. Anal. At. Spectrom. 7 (1992) 287.[8] A. Brockmann, C.
Nonn, A. Golloch, J. Anal. At. Spec-
trom. 8 (1993) 397.[9] W.W. Ding, R.E. Sturgeon, J. Anal. At.
Spectrom. 11
(1996) 225.[10] C. Schickling, J. Yang, J. Broekaert, J. Anal.
At. Spec-
trom. 11 (1996) 739.[11] T. Blanco, N. Maniasso, M.F. Gine, A.O.
Jacintho, Ana-
lyst 123 (1998) 191.[12] Annual Book of ASTM Standards, Section
3. Metals Test
Methods and Analytical Procedures E31, Philadelphia,USA, 1991,
p. 29.
dent on the Fe concentration as found for V.Results for three
ferrovanadium synthetic solu-tions and one certified reference
material Euro-standard 577-1 are shown in Table 3. The certifiedand
found results were not statistically differentfrom each other at
the 95% confidence level.
4. Conclusions
The proposed electrochemical reduction ap-proach is efficient
for on-line reduction of Fe3
to Fe2 in acid solutions. Addition of V and Moincreased
significantly the Fe3 reduction toFe2.
The proposal suits well for the quantitativedetermination of Fe2
and total Fe, the determi-nation of Fe and V or Fe and Mo in the
respec-tive solubilized iron alloys by using theelectrolytic
chamber reduction and 1,10-phenan-throline. Application of this
approach could notbe used to determine V or Mo in samples
contain-
..
