Anal Bioanal Chem (2004) 380: 135–140

DOI 10.1007/s00216-004-2733-8

O R I GI N A L P A P E R

Marcelo Firmino de Oliveira Æ Adelir Aparecida Saczk

Leonardo Luiz Okumura Æ Andréa Pires Fernandes
Mercedes de Moraes Æ Nelson Ramos Stradiotto

Simultaneous determination of zinc, copper, lead, and cadmium

in fuel ethanol by anodic stripping voltammetry using a glassy
carbon–mercury-film electrode
Received: 15 April 2004 / Revised: 15 June 2004 / Accepted: 16 June 2004 / Published online: 20 July 2004

Springer-Verlag 2004

Abstract A new, versatile, and simple method for of energy in various countries. As a result, ethyl
quantitative analysis of zinc, copper, lead, and cad- alcohol obtained from sugar cane has been successfully
mium in fuel ethanol by anodic stripping voltammetry used as an alternative fuel for national automobiles in
is described. These metals can be quantified by direct Brazil since 1975. It was initially used as an additive in
dissolution of fuel ethanol in water and subsequent gasoline and subsequently as a new fuel option with
voltammetric measurement after the accumulation economic and ecological advantages [1, 2].
step. A maximum limit of 20% (v/v) ethanol in water Despite generating less pollution than petroleum
solution was obtained for voltammetric measurements derivatives [3–6], ethyl alcohol does contain contami-
without loss of sensitivity for metal species. Chemical nants, for example organic compounds (methanol,
and operational optimum conditions were analyzed in aldehydes, esters, etc.) and inorganic species, for
this study; the values obtained were pH 2.9, a 4.7-lm example metals (zinc, copper, cadmium, iron, chro-
thickness mercury film, a 1,000-rpm rotation frequency mium, nickel, cobalt, lead, etc.). These contaminants
of the working electrode, and a 600-s pre-concentra- can arise from production process, transport, or fuel
tion time. Voltammetric measurements were obtained storage [7, 8]. In this context, analysis of metallic
using linear scan (LSV), differential pulse (DPV), and contaminants in fuel ethanol is a relevant research field
square wave (SWV) modes and detection limits were in in Brazil; conventional techniques such as chromatog-
the range 10
9–10
8 mol L
1 for these metal species. raphy and spectrometry are used [7–9].
The proposed method was compared with a traditional
analytical technique, flame atomic absorption spec-
trometry (FAAS), for quantification of these metal Electroanalytical applications
species in commercial fuel ethanol samples.
Electrochemical methods of analysis have been exten-
Keywords Alternative fuel Æ Ethanol Æ Heavy metals Æ sively employed for determination of trace metals in a
Mercury film electrode Æ Anodic stripping voltammetry wide range of matrixes of human interest, for example
food [10, 11], beverages [12, 13], biological fluids [14],
and others; the practicability, precision, and selectivity
of these methods are comparable with those of tradi-
Introduction tional techniques of analysis such as spectrometric and
chromatographic methods, besides utilizing more
Fuel ethanol accessible equipment.
A frequently used electroanalytical procedure for
Several petroleum crises from 1970–1980 resulted in a metal trace determination is anodic stripping voltam-
substantial increase in the price of petroleum per metry; reported detection limits (DL) are in the ppb
barrel and stimulated research into alternative sources range for various metals. This experimental procedure
consists in accumulation of reduced metallic species on
M. F. de Oliveira (&) Æ A. A. Saczk Æ L. L. Okumura the working electrode surface by applying a reduction
A. P. Fernandes Æ M. de Moraes Æ N. R. Stradiotto potential for a pre-established interval of time:
Institute of Chemistry, UNESP,
14801-970 Araraquara, SP, Brazil Mxþ þ x:e
! M0 ðaccumulation stepÞ ð1Þ
E-mail: [email protected]
E-mail: [email protected] and subsequent potential sweep at oxidation range,
136

being obtained a resulting anodic current with linear electrolyte solutions described above; the solutions were
dependence to metal concentration in a broad range of standardized by complexometric titration according to
analysis. the literature [8]. Two groups of 1.0·10
3 mol L
1
solutions containing each metal were prepared, one in
M0 ! Mxþ þ x:e
ðreoxidation stepÞ ð2Þ water and the other in alcohol. A 1.0·10
2 mol L
1
mercury nitrate stock solution in 1.0·10
2 mol L
1
aqueous lithium chloride solution was used for mercury
film preparation.
Mercury film electrodes (MFEs)

Chemical modification of electrode surfaces by Equipment

mechanical or electrochemical inclusion of analytes has
several advantages in terms of selectivity, sensitivity, and Cell and electrodes
efficiency in the determination of metallic species by
electroanalytical techniques [15–17], including metal Voltammetric measurements were obtained in a con-
analyses of petroleum-derived fossil fuels [18, 19]. ventional electrochemical cell with a 2-mm diameter
In this context, a successful electrode for metal glassy carbon electrode (working electrode), a 0.25-cm2
analysis is an MFE comprising a thin layer of metallic platinum plate (auxiliary), and a saturated calomel ref-
mercury electrodeposited on a conventional solid and erence electrode (SCE). Oxygen was removed from the
inert surface such as glassy carbon. This kind of elec- solutions of interest by nitrogen gas flow, through a
trode has experimental advantages such as a high sur- coupled Teflon tube in the cell.
face area/volume ratio, resulting in a higher
concentration of amalgam during the deposition step,
Apparatus
which increases the sensitivity. Additionally, MFEs have
high mechanical resistance, being stable to vigorously A potentiostat from PAR EG&G, model 273A, coupled
stirring or coupled with flow systems [20, 21]; this to a microcomputer was used to obtain the voltammo-
increases their applicability. grams. During the deposition step electrode rotation was
Despite the importance of quality control of fuel achieved by using a rotation generator from PAR
ethanol, the literature contains no reports of studies of EG&G, model 616RDE.
voltammetric determination of these metallic species in pH was measured by using a pH meter from Corning,
this fuel. So, the aim of this work was to develop an model 455, with a coupled temperature corrector.
electrochemical method for simultaneous determination Flame atomic absorption spectroscopy (FAAS) ex-
of zinc, copper, lead, and cadmium in fuel ethanol by periments were carried out by using a spectrometer from
anodic stripping voltammetry using a glassy carbon– Perkin–Elmer, model A Analyst 100, coupled to a
mercury-film electrode. printer. Hollow cathode lamps from Perkin–Elmer were
utilized for each metallic species.

Experimental
Procedure
Reagents
Preparation of the MFE
Supporting-electrolytes
Electrochemical deposition of a mercury film on the
Hydrochloric acid from Merck, lithium chloride from glassy carbon electrode surface was achieved according
Mallinckrodt, and lithium hydroxide from Vetec were to a conventional procedure reported in the literature
used to prepare lithium chloride solutions of different pH [22]. The electrochemical cell contained 30 lL
to act as supporting electrolytes. All reagents were of 1.0·10
2 mol L
1 mercury nitrate solution diluted to a
analytical grade purity. Lithium chloride was chosen be- final volume of 15 mL with an aqueous solution of
cause of its high solubility in ethyl alcohol. Two groups of 1.0·10
2 mol L
1 lithium chloride, resulting to a
stock solutions of 1.0·10
2 mol L
1 of these reagents 2.0·10
5 mol L
1 solution in mercury ions. After dea-
were prepared separately, one in deionized water and the erating this solution with nitrogen flow for 15 min, a
other in HPLC-grade ethyl alcohol from Aldrich. potential of
0.9 V vs. SCE was applied to the system
for 20 min, with a rotation value of 200 rpm for the
Metal standard solutions working electrode, and the resulting electrical current
was recorded as a function of the time. After film for-
Standard solutions of zinc, copper, lead, and cadmium mation the electrode was rinsed with deionized water
were prepared by dilution of corresponding analytical and stored in an aqueous supporting-electrolyte solu-
purity-grade sulfate and chloride salts in the supporting- tion.
 137

Optimization of the composition of the reaction medium gas stations in Araraquara city were analyzed. Each
3 mL sample was initially diluted to a final volume of
The best reaction medium for analysis was investigated 15 mL with supporting electrolyte. Further 100-fold
by using different alcohol/water compositions from 0 to dilution to furnish second and third samples was also
100% w/w, prepared by mixing aqueous supporting necessary. Quantitative analysis of these metals by flame
electrolyte solution, as previously described, with ethyl atomic absorption spectrometry (FAAS) was also per-
alcohol. To each 15 mL of studied solution was added formed to enable comparison between a conventional
30 lL of each metal standard solution, to give metal technique and the new proposed methodology; FAAS
solutions of concentration 2.0·10
6 mol L
1. Voltam- was performed in accordance with experimental condi-
metric measurement of the different solutions was per- tions reported in the literature [8] and the concentration
formed by linear sweep voltammetry (LSV) with a scan of the metallic species was obtained by use of a cali-
rate of 100 mV s
1, a potential interval from
1.4 to bration curve in the interval from 1.0·10
7 to
0 V versus SCE, a
1.4 V versus SCE potential value for 1.5·10
5 mol L
1.
the stripping step at a 10-min time interval, and a 200-
rpm rotation for the MFE.
Results and discussion
Optimization of experimental conditions Optimization of the composition of the reaction medium
The optimum pH for analysis was investigated. Ten 15- and other experimental conditions
mL aliquots of the previously optimized reaction med-
ium were prepared at pH ranging from 2 to 11 and fixed Voltammograms obtained for zinc, copper, lead, and
ionic strength. Standard solutions of the metal species cadmium at different water/alcohol compositions were
were added to these solutions in the same way as de- analyzed in terms of anodic peak current (iap) and
scribed above and LSV was again used for voltammetric anodic peak potential (Eap) for each metallic species; the
measurements, under the same conditions as described results are reported in Table 1.
above. Film thickness was also investigated. Mercury According to these results iap values decrease when
films of different thickness were obtained by the proce- the concentration of alcohol is increased, indicating
dure described above for different deposition times from more difficult accumulation of metal on the mercury
2 to 30 min. The optimum value of the rotation fre- film. An explanation for these results might be 13 times
quency for MFE was obtained by variation of the fre- higher cell resistance [23] for ethanol and lower dielectric
quency from 0 to 1,000 rpm. Finally, the best value for constant. Table 1 indicates that 80% water and 20%
deposition time stripping step was studied from 1 to ethanol (w/w) gives the best results, i.e. this is the max-
30 min. imum percentage of ethanol which does not lead to loss
of sensitivity. For solutions with alcohol concentrations
above 80%, amperometric response is observed for
Zinc, copper, lead, and cadmium analysis cadmium only. It was also shown that voltammetric
in commercial samples of fuel ethanol speciation of these metals is not possible for ethanol
concentrations higher than 80%.
After optimization of experimental and voltammetric Optimized experimental conditions, i.e. pH, film
conditions for the three different modes of voltammetry, thickness, rotation, and deposition time, for voltam-
three commercial samples of fuel ethanol from different metric analysis are reported at Table 2.

Table 1 Influence of the composition of the reaction medium on iap and Eap for analysis of zinc, copper, cadmium, and lead by anodic
stripping voltammetry in LS mode

Composition Zinc Cadmium Lead Copper

(% water)
iap (lA) Eap (V) vs. SCE iap (lA) Eap (V) vs. SCE iap (lA) Eap (V) vs. SCE iap (lA) Eap (V) vs. SCE

100 7.2
1.10 61.6
0.54 18.4
0.41 21.4
0.06

90 1.7
1.12 37.0
0.57 10.6
0.43 23.0
0.16
80 1.6
1.05 36.2
0.49 16.4
0.34 19.6
0.16
70 1.2
1.06 28.8
0.56 11.5
0.42 12.5
0.16
60 1.3
1.11 20.9
0.60 13.8
0.49 15.9
0.17
50 1.3
1.10 11.2
0.63 8.7
0.52 16.1
0.20
40 1.3
1.11 5.6
0.66 7.2
0.53 16.5
0.22
30 1.2
1.10 4.3
0.70 6.7
0.57 17.2
0.27
20 – – 3.7
0.78 – – – –
10 – – – – – – – –
0 – – – – – – – –

v 100 mV s
1, tdep 10 min, Eapl -1.4 V vs. SCE

138

Table 2 Optimized experimental conditions for voltammetric

analysis of metallic species in fuel ethanol 0

1
1
1
pH L lm r rpm tdep s
a A
2.9 4.7 1,000 600
10 b
c Zn
d
Optimization of conditions for voltammetry
e B
An important condition in LSV is the scan rate. In this 20 f
study it was observed that Eap values were disturbed at

i (µA)
g 40 Pb
high positive values of the scan rate and that the linear h
dependence of iap on the square root of scan rate (v) i
Cd
30
enables important information to be obtained about 30
j Cu
mass transport and about the reversibility of the chem-

i ap (µA)
ical reaction [24, 25]. An initial value of the scan rate of k 20 Cd

100 mV s
1 was established for voltammetric mea- l Cu

surements. 40 10
After optimization of the voltammetric conditions, Zn
concentrations of zinc, copper, lead, and cadmium from 0
8.0·10
9 to 5.0·10
7 mol L
1 were investigated; the 0 5 10 15 20 25 30

voltammograms obtained are reported in Fig. 1A and 50 Pb -8 -1

C M (x 10 mol. L )
the respective calibration curves for these metals are
reported in Fig. 1B. 0,0 -0,2 -0,4 -0,6 -0,8 -1,0 -1,2 -1,4 -1,6

E (V) vs. SCE

Fig. 2 A Voltammograms obtained in DPV mode for different

concentrations of the metallic species. tdep 10 min, r 1,000 rpm.
-15 Concentrations: (a) 0.10 (b) 0.20 (c) 0.50 (d) 0.83 (e) 1.60 ( f ) 3.33
(g) 6.66 (h) 10.0 (i) 16.6 (j) 25.0 (k) 33.3 (L) 50.0·10
8 mol L
1.
-10 B Respective calibration curves for the metallic species

-5

0 A Use of differential pulse voltammetric (DPV) mode

Zn B
resulted in optimized potential scan rate and pulse
5 35
Pb magnitude of 5 mV s
1 and 100 mV, respectively.
30
Concentrations of the metallic species from 8.0·10
9 to
10
i (µA)

a
b 25
Cd
5.0·10
7 mol L
1 were investigated using this mode of
c
15 d
20
voltammetry. The voltammograms obtained are re-
iap (µA)

e
f
ported at Fig. 2A and respective calibration curves for
20 g 15
these metals are reported in Fig. 2B. Similar behavior in
h
i Cd 10 Cu terms of affinity was obtained for this modality, i.e.
Zn
25 5
Pb>Cd>Cu>Zn.
For square wave voltammetry (SWV) the optimized
30
0
0 10 20 30 40 50 60 pulse magnitude (Esw) and pulse frequency (f) were
Cu -8 -1
established as being 100 mV and 75 Hz, respectively.
35 C M (10 mol.L ) The range of concentrations studied was from 8.0·10
9
Pb
to 5.0·10
7 mol L
1. The voltammograms obtained are
40 reported in Fig. 3A and respective calibration curves are
0,2 0,0 -0,2 -0,4 -0,6 -0,8 -1,0 -1,2 -1,4 -1,6 shown in Fig. 3B.
E (V) vs.SCE The calibration curves obtained for the metallic spe-
cies using the three modes of voltammetry enabled
Fig. 1 A Voltammograms obtained in LSV mode for different determination of important electrochemical properties
concentrations of zinc, copper, lead, and cadmium in a solution of such as amperometric sensitivity (a), linear correlation
0.01 mol L
1 lithium chloride in 80/20% water/ethanol at pH 2.9 coefficient (r), standard deviation (SD), and detection
as supporting electrolyte; tdep 10 min, r 1,000 rpm. Concentrations:
(a) 0.83 (b) 1.66 (c) 3.33 (d) 6.66 (e) 10.0 (f) 16.6 (g) 25.0 (h) 33.3 (i) limit (DL). The values obtained are reported at Table 3.
50.0·10
8 mol L
1. B Respective calibration curves for the For DL calculation a theoretical procedure was followed
metallic species according to the literature [22].
 139

Three different samples of conventional fuel ethanol

20 were studied, each one in triplicate. Examples of the
voltammograms obtained for each sample are reported
in Fig. 4.
A Traces of zinc, copper, and lead were found in the
40
commercial samples studied. Despite use of one of the
highest amperometric sensitivities, no traces of cadmium
were found in these samples. Mean values and standard
Zn deviations for these trace metals are reported in Table 4.
60 B
i (µA)

a
Pb
b 70
c
Spectroscopic analysis
60
d
80 e 50
Cd For determination of metal concentrations in the com-
f
mercial samples by FAAS, sensitivity for cadmium was
iap (µA)

g 40
h Cu highest, with an a value of 3.17·104 mol
1 L, an r-value
i 30
10 0 j
of 0.9968, and an SD value of 0.0099. These results re-
k
20
sulted in a DL value of 8.98·10
6 mol L
1.
l 10
Zn
Cd
0
0 2 4 6 8 10 12
12 0 0
Cu Pb -8
CM (10 mol.L )
-1

c
2
0,2 0,0 -0,2 -0,4 -0,6 -0,8 -1,0 -1,2 -1,4 -1,6
4 Pb Zn
E (V) vs. SCE

Fig. 3 A Voltammograms obtained in SWV mode for different 6

concentrations of the metallic species. tdep=10 min, r 1,000 rpm.
Concentrations: (a) 0.10 (b) 0.20 (c) 0.50 (d) 0.83 (e) 1.60 (f) 3.33 (g) 8 Cu
6.66 (h) 10.0 (i) 16.6 (j) 25.0 (k) 33.3 (L) 50.0·10
8 mol L
1.
B Respective calibration curves for the metallic species

Zinc, copper, lead, and cadmium analysis in commercial 0

samples of fuel ethanol
2 b
Voltammetric analysis Pb
i (µA)

4 Zn
Cu
Because of its good DL values and better working
concentration range, DPV mode was employed for 6
voltammetric analysis of zinc, copper, lead, and cad-
mium in commercial fuel ethanol samples. 8

Table 3 Voltammetric data obtained from analysis of zinc, cad-

mium, lead, and copper by use of LSV, DPV and SWV modes
0
a r SD DL
(107 lA mol
1 L) (lA) (10
9 mol L
1) 2 a
Pb Zn
Zn 4
LSV 1.45 0.9957 0.40 25.9
DPV 0.87 0.9964 0.16 19.8 6
SWV 6.3 0.9981 0.16 8.06
Cd
LSV 4.63 0.9974 0.40 25.9 8 Cu
DPV 6.68 0.9977 0.44 19.7
SWV 34.5 0.9972 1.12 9.70 0,0 -0,2 -0,4 -0,6 -0,8 -1,0 -1,2 -1,4
Pb
LSV 5.35 0.9975 0.68 38.0
DPV 19.6 0.9971 0.94 14.5
E (V) vs. SCE
SWV 33.7 0.9961 1.31 16.0
Cu Fig. 4 Voltammetric analysis in DPV mode of zinc, cooper, lead,
LSV 1.92 0.9991 0.15 23.4 and cadmium in commercial fuel ethanol samples. tdep 10 min, r
DPV 5.37 0.9980 0.31 17.3 1,000 rpm, LiCl 0.01 mol L
1 as supporting electrolyte. a Sample
SWV 12.2 0.9978 0.35 8.58 1: 3.0 mL of added sample; b Sample 2: 300 lL of added sample
c Sample 3: 300 lL of added sample
140

Table 4 Concentrations of
trace metals in commercial fuel Sample CZn (·10
7 mol L
1) CCu (·10
7 mol L
1) CPb (·10
7 mol L
1)
ethanol samples measured by
DPV and by FAAS DPV FAAS DPV FAAS DPV FAAS

1 ND ND 4.78±0.20 ND 12.1±0.57 13.4±0.73

2 7.40±0.28 8.23±0.53 8.31±0.39 8.71±0.48 24.1±1.10 26.3±1.30
3 1.04±0.04 ND 6.42±0.31 ND 8.10±0.39 7.4±0.42
ND not detected

Zinc resulted in the second highest absorptive sensi- 2. Sandelim K, Backman R (1999) Environ Sci Technol 33:4508–
tivity, 2.67·104 mol
1 L, with an r-value of 0.9983, an 4513
3. Jorgensen AD, Stetter JR (1982) Anal Chem 54:381–385
SD value of 0.002 and a DL of 2.36·10
7 mol L
1. 4. Skerfving S, Gerhardsson L, Schutz A, Stromberg U (1998)
Copper and lead resulted in the lowest of sensitivity, J Trace Elem Exp Med 11:289–301
7.37·103 and 1.55·103 mol
1 L, respectively. The r-val- 5. Bettinelli M, Spezia S, Baroni U, Bizzarri G (1996) Mikrochim
ues for the calibration curves for these metals were 0.9992 Acta 123:217–230
6. Brüning IMRA, Malm EB (1982) Bio Technol Petrobrás
and 0.9732, respectively, and the SD values 0.0019 and 25:217–228
0.0003. DL values for these metals were 8.78·10
7 and 7. Carbonel V, Salvador A, de la Gardia M (1992) Fresenius J
6.57·10
7 mol L
1, respectively. According to the re- Anal Chem 355:529–532
sults obtained by use of this technique, without an 8. Padilha PM, Padilha CCF, Rocha JC (1999) Quim Anal
accumulation step, DL values for the metals studied were 18:299–303
9. Padilha PM, Gomes LA, Moreira JC, Filho NLD, Gushiken Y
between 10
7 and 10
6 mol L
1, so the technique was (1998) J Bras Chem Soc 9:494–498
less sensitive than voltammetry. Results obtained from 10. Jagner D, Danielsson L, Josefson M, Westerlund S (1981) Anal
analysis of metals in commercial samples by FAAS are Chim Acta 127:147–156
also reported at Table 4; this enables comparison of the 11. Ahmed R, Valenta P, Nürberg HW (1981) Mikrochim Acta
1:171–184
spectroscopic and voltammetric techniques. 12. Jagner D, Westerlund S (1980) Anal Chim Acta 117:159–164
13. Stradiotto NR, Barbeira PJS, Mazo LH (1995) Analyst
120:1647–1650
Conclusions 14. Djane N, Armalis S, Ndung’u K, Johansson G, Mathiasson L
(1998) Analyst 123:393–396
15. Monterroso SCC, Carapuca HM, Duarte AC (2003) Electro-
According to the results obtained it is possible to con- analysis 15:1878–1883
clude that electrochemical deposition of a mercury film 16. Farghaly OA (2003) Microchem J 75:119–131
on glassy carbon electrodes enables use of the electrode 17. Armalis S, Kubiliene E (2001) Chem Anal Warsaw 46:715–723
18. Ahmad N, Khan LA, Sattar A (1991) J Chem Soc Pakistan
as a transducer for determination of zinc, copper, lead, 13:74–78
and cadmium in fuel ethanol samples by anodic strip- 19. Ceccon L, Coco FL, Monotti P, Rizzotti S (1999) Anal Chim
ping. The optimum pH was 2.9 and the best supporting Acta 386:41–46
electrolyte was 20/80 ethanol/water. 20. Munoz E, Palmero S, Garcia MA (2002) Talanta 57:985–992
Furthermore, the voltammetric modes LSV, DPV, 21. Korolczuk M (2000) Electroanalysis 12:1502–1504
22. Rajeshwar K, Ibanez J (1997) Environmental electrochemis-
and SWV when employed with an accumulation step try—fundamentals and applications in pollution abatement.
enabled simultaneous analysis of zinc, copper, and lead Academic, California, p 233
species in commercial fuel ethanol samples with DL 23. Bruss DB, Vries T (1956) J Am Chem Soc 78:733–736
values from 10
9 to 10
8 mol L
1, approximately ten to 24. Bard AJ, Faulkner LR (1980) Electrochemical methods: fun-
damentals and applications. Wiley, New York, pp 213–248
100 times more sensitive than FAAS. Voltammetric 25. Bond AM (ed) (1980) Modern polarographic methods in ana-
techniques also have experimental advantages, for lytical chemistry. Marcel dekker, New York, pp 178–197
example no need for an atomization step, use of specific 26. Mannino S (1983) Analyst 108:1257–1260
reagents for the accumulation step, and low cost of cell
and equipment.

References

1. Proálcool. http://www.tvcultura.com.br/resguia/historia/cenas/
proalc.htm