Talanta 52 (2000) 417 424

www.elsevier.com/locate/talanta

Direct determination of molybdenum in seawater by

adsorption cathodic stripping square-wave voltammetry
Yuh-Chang Sun a,*, Jerzy Mierzwa b, Chi-Ren Lan c
a

Nuclear Science and Technology De6elopment Center, National Tsing-Hua Uni6ersity, 30043 Hsinchu, Taiwan, ROC
b
Department of Chemistry, Central Michigan Uni6ersity, Mount Pleasant, MI 48859, USA
c
Chemistry Laboratory, Power Research Institute, Taiwan Power Company, 23802 Taipei, Taiwan, ROC
Received 7 December 1999; received in revised form 7 February 2000; accepted 22 February 2000

Abstract
A reliable and very sensitive procedure for the determination of trace levels of molybdenum in seawater is
proposed. The complex of molybdenum with 8-hydroxyquinoline (Oxine) is analyzed by cathodic stripping squarewave voltammetry based on the adsorption collection onto a hanging mercury drop electrode (HMDE). This
procedure of molybdenum determination was found to be more favorable than differential pulse cathodic stripping
voltammetry because of inherently faster scan rate and much better linearity obtained through the one-peak (instead
of one-of-two peaks) calibration. The variation of polarographic peak and peak current with a pH, adsorption time,
adsorption potential, and some instrumental parameters such as scan rate and pulse height were optimized. The
alteration of polarographic wave and its likely mechanism are also discussed. The relationship between peak current
and molybdenum concentration is linear up to 150 mg l 1. Under the optimal analytical conditions, the determination
limit of 0.5 mg l 1 Mo was reached after 60 s of the stirred collection. The estimated detection limit is better than
0.1 mg l 1 of Mo. The applicability of this method to analysis of seawater was assessed by the determination of
molybdenum in two certified reference seawater samples (CASS-2 and NASS-2) and the comparison of the analytical
results for real seawater samples (study on a vertical distribution of Mo in the seawater column) with the results
obtained by Zeeman-corrected electrothermal atomization atomic absorption spectrometry (Zeeman ETAAS). A
good agreement between two used methods of molybdenum determination was obtained. 2000 Elsevier Science
B.V. All rights reserved.
Keywords: Cathodic stripping square-wave voltammetry; Oxine; Molybdenum; Seawater

1. Introduction

* Corresponding author. Tel.: + 886-3-5727309; fax: + 8863-5723883.

E-mail address: ycsun@mx.nthu.edu.tw (Y.-C. Sun)

A special attention has been devoted to molybdenum in seawater because of its important role
in enzymatic redox reaction and its specific geochemical behavior [1]. The predominant form of
molybdenum occurrence is oxyanion MoO34 (at

0039-9140/00/$ - see front matter 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 3 9 - 9 1 4 0 ( 0 0 ) 0 0 3 9 1 - X

418

Y.-C. Sun et al. / Talanta 52 (2000) 417424

the typical seawater pH [2]) and the over-all average concentration of dissolved molybdenum in
seawater is indicated about 9 11 mg l 1, although
most of the data has been insufficiently precise to
detect some small variations [1]. Much more precise analytical results are demanded by oceanographers to carry out further marine geochemical
prediction, for example, the cycling properties of
molybdenum. Molybdenum and iron are two
trace elements both required for nitrogen fixation
in seawater. Several analytical methods have been
developed for the determination of molybdenum
in various environmental matrices. Among the
most commonly applied are the spectrophotometric analysis of molybdenum complexes [3,4] and
electrothermal atomization atomic absorption
spectrometry [5,6]. Both of these methods are
usually combined with some analyte preconcentration prior to its determination.
The most remarkable feature of molybdenum is
its chemical versatility. In addition to its multiform coordination behavior, the electrochemical
characteristics of molybdenum is also diverse. The
chemistry of molybdenum involves oxidation
states III, V, and VI. In aqueous solutions, the
Mo(VI) oxidation state cannot be reduced below
III. Usually, polarographic reduction of molybdate ion takes place in two steps, Mo(VI)
Mo(V)
and
Mo(V)Mo(III),
in
many
investigated media [7]. The low solubility of
molybdenum in mercury [8] and the unique redox
properties limits the application of electrochemical methods for the direct determination of
molybdenum in natural waters. For conventional
electroanalytical methods, the detection limit obtained was in the range of sub-ppm level [911].
Whereas the use of the catalytic wave obtained in
nitrate or perchlorate solutions containing Mo
(VI) ions can result in the detection limits in mg
l 1 range [12,13].
Extremely sensitive and versatile, alternative cathodic stripping voltammetric methods for the
determination of Mo have been developed to
determine various trace metal ions achieved via
adsorptive collection of their surface active metal
complexes onto the mercury working electrode.
These methods have been reviewed, e.g. by van
den Berg [14,15], Kalvoda and Kopanica [16], and

Riley and Skirrow [17]. However, some improvements are still desirable, especially, for the analytical case of peak merging at higher concentration
or non-linearity of the relationship between peak
current and concentration in the use of oxine as a
complexing agent.
The application of square-wave voltammetry as
a voltammetric technique for very reliable quantitative analysis has received increasing attention
because it can provide extremely high scan rate,
excellent rejection of background current, and
affects kinetics of charge transfer reactions [18].
According to van den Berg [14,15], very fast potential scanning techniques can be employed due
to the unique accumulation mechanism of adsorption voltammetry.
In the present study, a possibility of the application of square-wave voltammetry for molybdenum determination by cathodic stripping
voltammetry based on the collection of Mo(VI)oxine onto the hanging mercury drop electrode
(HMDE) is investigated. By the use of very fast
scan rate and large pulse height squarewave potential scanning modulation, an unique
one-peak voltammogram instead of two-peak
voltammogram, and both excellent linearity and
detection power were found. The proposed
method was successfully applied to the reliable
determination of molybdenum traces in seawater
samples.

2. Experimental section

2.1. Apparatus
A Princeton Applied Research (PAR, USA)
Model 384B polarographic analyzer system, with
a PAR 303A static mercury drop electrode, a
PAR Model 305 stirrer, and a Houston HIPLOT
DMP-40 digital plotter, was used for all squarewave stripping measurements. All potentials were
given with respect to the Ag/AgCl/saturated KCl
reference electrode. The Mo-Oxine complex was
accumulated at the working electrode with small
size mercury drop. All measurements were performed in a laminar flow hood, or in a Class 100
clean room supplied with the filtered air.

Y.-C. Sun et al. / Talanta 52 (2000) 417424

A Perkin Elmer 5100 PC atomic absorption

spectrometer (Perkin Elmer GmbH, Germany)
with Zeeman background correction system,
equipped with an HGA-600 electrothermal atomizer, and an AS-60 autosampler was used for the
comparative determination of molybdenum. A
molybdenum hollow cathode lamp of Intensitron type (Perkin Elmer Corporation, USA),
pyrolytically coated graphite tubes and Lvov
platforms were also used. A Mo spectral line at
313.3 nm was recorded. Palladium nitrate
modifier was applied and peak areas were
recorded as the analytical signal. The typical temperatures of graphite furnace were as follow: drying 120C, ashing 1400C, atomization 2600C,
clean-up 2700C.

2.2. Reagents, samples and glassware

All reagents used were Suprapur or analyticalreagent grade. Water was purified in Milli-Q system and after then in a double-distillation quartz
apparatus. Mercury was of triple-distilled quality.
An 8-Hydroxyquinoline (E. Merck, Darmstadt,
Germany) solution (0.1 M) was prepared by dissolving the appropriate amount of the compound
in 0.1 M HCl (spectral pure, from E. Merck)
solution and used without purification. Working
standard solutions of molybdenum (VI) were prepared daily from 0.01 M ammoniumheptamolybdate (E. Merck) by serial dilution of the stock
solution.
The following certified reference materials
(CRMs) were obtained from the National Research Council of Canada (NRCC, Ottawa,
Canada): NASS-2 Seawater (Open Ocean) and
CASS-2 Seawater (Coastal). All glassware,
polyethylene bottles and polarographic cells were
soaked in a bath filled with 6 M HNO3 for a few
days, rinsed four to five times with a double
distilled
water,
wrapped
in
pre-cleaned
polyethylene bags and kept in a class 100 cleanbench until required. The open ocean seawater
samples from a vertical profile were collected
from the surface to the depth of 1500 m at a
sampling site in the Philippine Sea (3 45%N, 124
01%E), by GO-Flo sampling system (General
Oceanic, USA) hung on Kevlar line. The seawater

419

samples were filtered on board through 0.45 mm

Millipore filters immediately after sampling and
preserved on site by an addition of 5 ml of
sub-boiling 65% nitric acid (per 1 l of seawater).
The samples were then stored in darkness, in
pre-cleaned polypropylene bottles and the analysis
was complete within 6 weeks after sampling.

2.3. Analytical procedure

A 10-ml of sample was pipetted into the polarographic cell and its pH was adjusted to pH 2.2
using diluted (1:1) hydrochloric acid. Then 10 ml
of 0.07 M oxine (final oxine concentration was
7 10 4 M) was added. The Mo concentration in
bi-distilled water at pH 2.2 and in the presence of
7 10 4 M oxine was below 0.5 mg l 1. The
solution was then deaereated by purging for 15
min with pure nitrogen. Nitrogen was also passed
continuously over the surface of solution during
all measurements. About 1012 s after the purging was stopped (to allow the solution to become
stationary), the potential of the HMDE was set at
0.17 V and a new mercury drop was extruded,
which signifies the beginning of the adsorption
period of 60 s using the square wave modulation
(SWAV) with the frequency of 100 s 1 and pulse
height of 100 mV. The height (in nA) of Mo peak
was recorded. The calibration curve was prepared
by the standard additions technique, adding
molybdenum standard to an aliquot of a Mostripped seawater sample (blank). The concentration of molybdenum in blank was below the limit
of detection. All these measurements were carried
out in a class-100 clean room at the room
temperature.
All described in the next part experiments were
repeated five times.

3. Results and discussion

3.1. Polarographic characteristics of molybdate

It is noteworthy for the study of the polarographic behavior of molybdate in nitric acid solution that the presence of nitrate resulted in an
enhanced reduction current, which was attributed

420

Y.-C. Sun et al. / Talanta 52 (2000) 417424

to the catalytic reduction for Mo(VI) determination. Oxine is well known as complex forming
agent for Mo(VI) determination. The behavior of
Mo(VI) in acid supporting electrolyte was studied
and two close reduction peaks with EP1 = 0.54
V and EP2 = 0.59 V were obtained in oxine
containing samples due to the reduction of
Mo(VI) to Mo(V) and Mo(III) [19]. A possible
course of the reaction on the electrode is proposed
here:

According to the proposed above mechanism,

the catalytic reaction between Mo(III) and nitrate
results in more complex overall reaction of reduction and amplifies the height of peak 2. Thus, the
disproportional amplification of peak 2 makes
two reduction peaks easier to merge and causes
some non-linear changes of peak height, and nonlinear relationship of molybdate concentration

versus peak height. With a view to improve the

analytical figures-of-merits of the molybdenum
determination by adsorption voltammetry using
oxine as the complexing agent, the square-wave
potential scanning modulation was further investigated. The effect of the potential scan rate on the
peak current at pH 2.2 in seawater is shown in
Fig. 1. It was found that peak 2 decreases with the
increased scan rate when the differential pulse
scanning modulation was used, but never could be
totally removed.
Fig. 2 shows the stripping voltammograms obtained for square wave modulation of increasing
pulse height. It was found that peak 2 die away
with the increasing pulse height and was finally
removed at the applied pulse height up to 100
mV. The results obtained show that the voltammogram can be treated as one-peak reaction at
the high enough potential (DE \ E(Mo(V)
Mo(III); 0.085 V [12]) within a moderate pulse
duration.
The calibration graph of peak current versus
molybdate concentration was obtained by the
standard additions method (Fig. 3). A very good
linearity (correlation coefficient r= 0.9999 was

Fig. 1. Effect of the different scan rate on peak shape (A: 2 mV s 1; B: 5 mV s 1; C: 10 mV s 1, differential pulse mode,
adsorption time = 60 s, pH 2.2).

Y.-C. Sun et al. / Talanta 52 (2000) 417424

421

Fig. 2. Effect of the different pulse height on peak shape (A: 20 mV; B: 70 mV; C: 100 mV, square wave mode, adsorption time= 60
s, pH 2.2).

calculated) over a wide concentration range (from

0.1 up to 150 mg l 1) was obtained and this can
confirm that one-peak adsorption voltammetric
method for the determination of molybdenum in
seawater has an excellent analytical potential.
Furthermore, the relationship between Mo concentration and peak current leveled off at the
molybdate concentration higher than approximately 150 160 mg l 1 because of the saturation
of HMDE. The shape of the curve presented in
Fig. 3 is practically the same as the shape of the
adsorption curve derived from the Langmuir type
isotherm of adsorption [20], and this fact can be
regarded as a supporting evidence to the
monomolecular layer adsorption of the complexed species on the surface of mercury
electrode.

It can be seen from the above equation that the

formation yield of MoO2(Ox2) depends not only
on pH but also on the oxine concentration. However, the formation of the surface-active complex
affects significantly the sensitivity of adsorption
voltammetry. Fig. 4 shows the relationship between the acidity of solution and peak current. It
was found that the peak current decreases with
the acidity of solution at pH above 2.3 and less

3.2. Effect of pH
Stary [21] reported earlier that molybdenum is
present predominantly as H2MoO4 at pH 12 and
the interaction of molybdenum and oxine can be
expressed by the following equation:
MoO2(OH)2 +2HOx MoO2(Ox)2 +2H2O

Fig. 3. Effect of Mo(VI) concentration on peak current (adsorption time = 60 s, pH 2.2, square wave mode).

422

Y.-C. Sun et al. / Talanta 52 (2000) 417424

3.3. Effects of the adsorption potential and

adsorption time

Fig. 4. Influence of pH of the analyzed solution on Mo peak

current (adsorption time = 60 s, square wave mode).

Fig. 5. Effect of the adsorption potential on Mo peak current.

than 1.9. In the lower pH range (pHB 1.9) protons can compete against H2MoO4 in the reaction
with oxine, therefore MoO2(Ox)2 complex formation rate is depressed and the yield of complex
formation falls off. Two possible reasons may be
suggested to explain the decrease of peak current
at higher pH (pH \2.3): (1) the occurrence of
MoO24 increases as the pH value is higher than
2.3 and therefore the yield of MoO2(Ox)2 decreases; and (2) molybdate loses the electroactivity
rapidly in the solution more acidic than approximately pH 2. A similar decrease in the electroactivity of molybdate with pH change was also
reported for the nitrate [12] and sulfuric [22]
system.

The charge on the surface of mercury electrode

plays an important role in the adsorption efficiency of the interesting metal complexes [8],
whereas the applied potential and composition of
solution actually affect significantly the charge on
the mercury electrode.
Fig. 5 shows the effect of changes of the adsorption potential on the peak current of seawater
at pH 2.2 and in the presence of 7 10 4 M
oxine. This diagram can be divided into two
individual parts (part I: from 0.1 to 0.2 V,
part II: B 0.2 V). Because the molybdateoxine complex is a neutral molecule and the
optimal adsorption efficiency of neutral
molecule is at the potential of zero charge [23]
(the potential of zero charge of seawater is 0.5/
0.55 V [13]), the higher peak current obtained
in part II as the adsorption potential is approaching the potential of zero charge of seawater.
According to Navratilova and Kopanica [24], the
half-wave potential of molybdate-oxine is
0.2 V. The possible explanation of the decrease
of peak current with the decreasing adsorption
potential in part II of graph is the concomitant
loss (reduction) of adsorbed molybdateoxine complex on the electrode during the accumulation step. From the viewpoint of both the
optimal sensitivity and precision, a potential of
0.17 V was finally selected as the adsorption
potential.
From the experimental results of effect of adsorption time investigation, it can be concluded
that the analytical sensitivity (current of an analytical peak) increases with the increasing of the
adsorption time up to 200 s. For the analysis of
the seawater samples the compromise adsorption
time of 60 s was chosen, which gave an adequate
sensitivity to determine 0.5 mg l 1 of molybdenum in these samples.

3.4. Interferences
The sensitivity of the presented method in media of very different ion concentration that is

Y.-C. Sun et al. / Talanta 52 (2000) 417424

423

Table 1
Analytical results of Mo determination in CRMs (n = 5) and a sample of real seawater
Sample

CASS-2

NASS-2

Real seawater

Found (mg l1)

Certified value (mg l1)

9.37 9 0.18
9.01 90.28; 9.15 90.24a

11.89 0.7
11.5 9 1.9

10.15 90.21
10.05 90.25a

Zeeman ETAAS determination.

doubly distilled water, tap water and tested seawater was almost the same and interferences were
not observed. The interferences of natural organics were investigated in a typical way by the
addition of humic acid and non-ionic surfactant
Triton X-100. It was observed that the elevated
concentrations of these compounds might decrease analytical response of molybdenum. The
total amount of dissolved organic material in the
open seawater usually did not exceed 2 mg l 1
and the concentration of humic acids is less than
0.11 mg l 1 [25], therefore no noticeable interference may be expected in open seawater analysis,
however preliminary UV-digestion of water and
the use of standard additions technique of calibration is still recommendable.
The possible interference of coexistent cations
was also explored by the addition of 10 mg l 1 of
Zn, Cd, Pb and Cu into the all tested samples of
seawater. Fortunately, the analytical results were
found to be free from the aforementioned
interferences.

3.6. Analytical application

The developed method was applied to the analysis (without any pretreatment or preconcentration) of seawater samples collected in the summer
of 1995 from the Philippine Sea. The vertical
distribution of dissolved molybdenum in the water column is shown in Fig. 6. This distribution is
consistent with those reported in the earlier literature. The comparison of the analytical results of
the described above method (Fig. 6(a)) with the
results of the direct Mo determination by Zeeman
ETAAS (Fig. 6(b)) used as a control method, and
the previously reported results [26,27] for the vertical profile of molybdenum concentration in seawater shows good agreement and indicates that
the proposed method can be suitable for the
determination of molybdenum traces in real seawater samples.

3.5. Accuracy and repeatability of method

The accuracy of the proposed adsorption
voltammetric method was tested by the analysis of
NRCC Certified Reference Materials NASS-2
(Open Ocean Seawater) and CASS-2 (Coastal
Seawater). The obtained values, given in Table 1,
are in good agreement with the certified values.
As can be seen from this table, good agreement
between the results of this voltammetric method
and Zeeman ETAAS method was also obtained
for a real seawater sample. The repeatability (expressed as the relative standard deviation in percentage) of five independent analyses of the tested
in this study CRMs was in the range from 1.9 to
6.0%.

Fig. 6. Concentration-depth profile of Mo(VI) in seawater

samples: (A) Mo determined by the proposed voltammetric
method; and (B) Mo determined by Zeeman ETAAS.

424

Y.-C. Sun et al. / Talanta 52 (2000) 417424

Acknowledgements
The authors are thankful to Professor M.H.
Yang for very useful discussions and gratefully
acknowledge financial support from the National
Science Council (Taipei, Taiwan).

References
[1] R.W. Furnace, P.S. Rainbow, Heavy Metals in the
Marine Environment, CRC Press, Boca-Raton, FL, 1990.
[2] G. Bearman, Ocean Chemistry and Deep-Sea Sediments,
Pergamon, London, 1989.
[3] M. Soylak, U. Sahin, L. Elci, Anal. Chim. Acta 322
(1996) 111115.
[4] I. Nukatsuka, A. Nishimura, K. Ohzeki, Anal. Chim.
Acta 304 (1995) 243247.
[5] D. Baralkiewicz, J. Siepak, Anal. Chim. Acta 353 (1997)
85 89.
[6] Y.H. Sung, Z.S. Liu, S.D. Huang, J. Anal. At. Spectrom.
12 (1997) 841847.
[7] A.J. Bard, R. Parsons, J. Jordan, Standard Potentials in
Aqueous Solution, Marcel Dekker, New York, 1985.
[8] F. Vydra, K. Stulik, E. Julakova, Electrochemical Stripping Analysis, Wiley, New York, 1976.
[9] M.G. Johnson, R.T. Robinson, Anal. Chem. 24 (1952)
366 369.
[10] K. Ogura, Y. Enaka, J. Electroanal. Chem. 95 (1979)
169 175.

[11] P. Lagrange, J.P. Schwing, Anal. Chem. 42 (1970) 1844

1845.
[12] A.T. Violanda, W.D. Cooke, Anal. Chem. 36 (1964)
2287 2290.
[13] G.P. Haight Jr, W.F. Sager, J. Am. Chem. Soc. 74 (1952)
6056 6059.
[14] C.M.G. van den Berg, Analyst 114 (1989) 15271530.
[15] C.M.G. van den Berg, Anal. Chim. Acta 250 (1991)
265 276.
[16] R. Kalvoda, M. Kopanica, Pure Appl. Chem. 112 (1989)
98 112.
[17] J.P. Riley, G. Skirrow, Chemical Oceanography, Academic Press, San Diego, CA, 1989.
[18] J.G. Osteryoung, R.A. Osteryoung, Anal. Chem. 57
(1985) 101A 110A.
[19] C.M.G. van den Berg, Anal. Chem. 57 (1985) 15321536.
[20] B. Jirgensons, M.E. Starumans, Colloid Chemistry,
Macmillan, New York, 1962.
[21] J. Stary, Anal. Chim. Acta 28 (1963) 132 149.
[22] M. Codell, J.J. Mikula, G. Norwitz, Anal. Chem. 25
(1953) 1441 1444.
[23] F.C. Anson, Acc. Chem. Res. 8 (1975) 400 407.
[24] Z. Navratilova, M. Kopanica, Anal. Chim. Acta 244
(1991) 193 196.
[25] C.S. Wang, E. Boyle, K.W. Bruland, J.D. Burton, E.D.
Goldberg, Trace Metals in Sea Water, Plenum, London,
1983.
[26] Y. Nozaki, Geochem. J. 26 (1992) 25 39.
[27] E. Merian, Metals and Their Compounds in the Environment/Occurrence, Analysis, and Biological Relevance,
Verlag Chemie, Weinheim, 1991.