Short communication

Simultaneous determination of orthophosphate and total

phosphates (inorganic phosphates plus purine nucleotides)
using a bioamperometric ow-injection system made up by a
16-way switching valve
Toshio Yao *, Kazuyoshi Takashima, Youko Nanjyo
Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University, 1-1 Gakuencho, Sakai 599-8531, Japan
Received 4 December 2002; received in revised form 5 February 2003; accepted 5 February 2003
Abstract
Orthophosphate and total phosphates (inorganic phosphates plus purine nucleotides) can be determined
simultaneously in a novel flow-injection system made up by a 16-way switching valve with two sample loops, acid
phosphatase (AcP) immobilized reactor and a delay coil needed to separate two peaks corresponding to two sample
portions injected simultaneously. An orthophosphate enzyme electrode with a hybrid membrane of trienzyme film and
poly(1,2-diaminobenzene) film was used to selectively detect both the endogenous orthophosphate and orthophosphate
generated enzymatically into the AcP immobilized reactor, without any interferences from electroactive species, such as
ascorbate and urate. Because two sample portions passed through the flow line with different residence time, two peaks
were obtained. The first peak corresponded selectively to orthophosphate and the second peak to the total of inorganic
phosphates and purine nucleotides. The maximum currents of both peaks were linearly related to the concentration of
orthophosphate and total phosphates (as orthophosphate) in the range 5/10
7
/8/10
4
M, respectively; 30 samples
per hour can be processed with an R.S.D. B/2.5%.
# 2003 Elsevier Science B.V. All rights reserved.
Keywords: Flow-injection analysis (FIA); Simultaneous analysis; Orthophosphate; Total phosphates; Enzyme electrode; Enzyme
reactor
1. Introduction
The determination of orthophosphate is espe-
cially important in the field of environmental [1]
and food analysis [2] because increased orthopho-
sphate concentration leads to eutrophication of
lakes and rivers and an excess diet of orthopho-
sphate contained in foodstuffs affects human
health. In addition, there are a variety of phos-
phate-containing compounds in the environment
and foods, other than orthophosphate, which are
* Corresponding author. Fax: /81-722-54-9910.
E-mail address: yao@chem.osakafu-u.ac.jp (T. Yao).
Talanta 60 (2003) 845/851
www.elsevier.com/locate/talanta
0039-9140/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
PII: S 0 0 3 9 - 9 1 4 0 ( 0 3 ) 0 0 1 2 8 - 0
present as inorganic phosphates, such as pyropho-
sphate and tripolyphosphate, and organic phos-
phates, such as purine nucleotides. These
compounds gradually produce orthophosphate
by chemical and bacterial degradations. Therefore,
the determination of the total of such inorganic
phosphates and nucleotides is also important as
well as that of orthophosphate.
Many research groups have proposed the en-
zyme sensors for the specific detection of ortho-
phosphate. Engblom [3] has recently reviewed the
characteristics of the biosensors developed for the
determination of orthophosphate. Most of these
enzyme sensors are based either on the inhibitory
effect of orthophosphate on alkaline phosphatase
reaction [4,5] or the use of purine nucleoside
phosphorylase [6,7], maltose phosphorylase [8,9]
and pyruvate oxidase [10,11], which require ortho-
phosphate as a cosubstrate. Mousty et al. [12] used
the trienzyme system composed of maltose phos-
phorylase (MP), mutarotase (Mut) and glucose
oxidase (GOD) entrapped in an inorganic laponite
clay to fabricate the phosphate biosensor. Further-
more, Conrath et al. [13] reported a sensitive
biosensor involving amplification by adding acid
phosphatase (AcP) to the trienzyme system. How-
ever, both biosensors are proposed as only an
orthophosphate biosensor and not as a total
phosphate sensor.
This communication describes a novel flow-
injection system based on the combined use of
the biosensor and bioreactor for the simultaneous
determination of orthophosphate and total phos-
phates (inorganic phosphates plus purine nucleo-
tides). Here, we used the combination of MP, Mut
and GOD to fabricate an orthophosphate biosen-
sor. Furthermore, a poly(1,2-diaminobenzene)
film with a size-exclusion function [14] was hybri-
dized to a trienzyme cross-linked membrane to
improve the selectivity for orthophosphate. By
three successive enzymatic reactions, orthopho-
sphate can be converted to electroactive hydrogen
peroxide as an end product, which is detected
amperometrically. Furthermore, acid phosphatase
(AcP) was covalently immobilized onto the ami-
nopropyl-CPG and was used as the enzyme
reactor in a FIA system. Inorganic phosphates
and purine nucleotides are enzymatically hydro-
lyzed to generate orthophosphate during passage
through the AcP reactor. The produced orthopho-
sphate is detected at a downstream orthopho-
sphate biosensor. Flow injection system is based
on the combined use of a 16-way auto-switching
valve with two sample loops, a delay coil needed to
separate two peaks corresponding to two sample
portions injected simultaneously, an AcP immobi-
lized reactor and a downstream orthophosphate
biosensor. The proposed FIA method was a
simple, selective and rapid technique for the
simultaneous determination of orthophosphate
and total phosphates and, furthermore, worked
without any interference from electroactive spe-
cies.
2. Experimental
2.1. Materials
The enzymes used, AcP (EC 3.1.3.2, 5.1 U mg
1
of solid from potato), GOD (EC 1.1.3.4, 182 U
mg
1
of solid from Aspergillus niger) and Mut
(EC 5.1.3.3, 3300 U mg
1
of solid from porcine
kidney) were obtained from Sigma (St. Louis,
MO) and MP (EC 2.4.1.8, 570 U ml
1
of solution
from bacteria) was from Oriental Yeast (Tokyo).
The 5?-nucleotides (mono, di and triphosphates) of
xanthosine, inosine and guanosine were obtained
from Sigma and 5?-nucleotides (mono, di and
triphosphates) of adenosine from Oriental Yeast.
Bovine serum albumin (BSA), glutaraldehyde
(20% solution), ascorbic acid, uric acid, cysteine
and 1,2-diaminobenzene were obtained from
Wako (Osaka). All other chemicals were of
analytical reagent grade. Controlled pore glass
(aminopropyl CPG; mean pore size 52.7 nm;
particle size 120/200 meshes) was obtained from
CPG (Fairfield, NJ) and was used as a support
material for enzyme immobilization. Acetate buf-
fers (0.1 M) were prepared from sodium acetate.
Phosphate standard solution (10 mM) was pre-
pared from sodium dihydrogen phosphate. Stan-
dard solutions (10 mM each) of other inorganic
phosphates and 5?-nucleotides were freshly pre-
pared before use. Distilled water purified using a
T. Yao et al. / Talanta 60 (2003) 845/851 846
Millipore Milli-Q system (Nippon Millipore, To-
kyo) was used throughout.
2.2. Preparation of orthophosphate biosensor
An Eicom (Kyoto) cross-flow electrochemical
flow-cell was used for the surface modification of
the electrode. This consisted of a platinum disk (3
mm in diameter) as a working electrode, a silver/
silver chloride reference electrode and a stainless-
steel tube as an auxiliary electrode. Prior to the
enzyme coating, the platinum surface was polished
with 1 mm diamond particles (BAS) and then
coated with the poly(1,2-diaminobenzene) film by
the electropolymerization of 1,2-diaminobenzene
according to the previously described procedures
[14]. Then, the platinum disk was modified by
cross-linking enzymes and BSA using glutaralde-
hyde.
The method was similar to that described before
[15] and was as follows: the MP (5.7 U), Mut (660
U), GOD (73 U) and 10 ml of 10% (w/v) aqueous
BSA were added to 30 ml of 0.1 M sodium
phosphate buffer (pH 7.0). A 7 ml portion of a
4% (v/v) solution of glutaraldehyde was added to
the solution and mixed well. A 4 ml aliquot of the
resulting solution was carefully spread over the
poly(1,2-diaminobenzene) film coated platinum
disk and air-dried at room temperature for half a
day. Then the electrode was assembled into the
Eicom flow-cell and washed with 0.1 M glycine
buffer (pH 7.0) to remove the excess of enzymes
and the residual aldehyde groups onto the enzyme
membrane. The completed electrode was stored in
acetate buffer (0.1 M, pH 5.5) at 4/5 8C when not
in use.
2.3. Preparation of AcP immobilized reactor
Aminopropyl CPG was packed into a 10 mm
long glass column (3 mm i.d.) and activated by
glutaraldehyde as previously described [16]. After
washing with 0.1 M sodium phosphate buffer (pH
6.0), AcP (15.3 U) was loaded on the column by
circulating enzyme/sodium phosphate buffer (0.1
M, pH 6.0) solution for 2 h at room temperature.
The excess of enzymes and the residual aldehyde
groups on the support was removed by washing
with glycine buffer (0.1 M, pH 7.0) for 3 h. When
not in use, the enzyme reactor was stored in 0.1 M
sodium acetate buffer (pH 5.5) at :/4 8C to
minimize microbiological degeneration.
2.4. Apparatus and procedures
A Hitachi K-1000 FIA apparatus has been used
throughout; it was composed of a double plunger
pump, a 16-way auto-switching valve with two
sample loops of :/60 ml each, a peristaltic pump
used for delivery of sample solution to sample
loops and a thermostated oven to hold the
temperature of the flow line with enzyme reactor
constant. A 16-way switching valve was used to
design the flow-injection line, by positioning AcP
immobilized reactor (R) and two sample loops
(SL
1
and SL
2
), as shown in Fig. 1. All parts of the
flow-line were held in a column oven thermostated
at 309/0.2 8C. Constant potential (0.6 V versus
Ag/AgCl) was applied to the flow-through ortho-
phosphate biosensor with an Eicom ECD-300
potentiostat and current was recorded on a SIC
Chromatocorder.
In the first stage (ready and charge mode in Fig.
1; for 0.2 min) of the operation, carrier solution
was pumped as shown at the constant flow rate.
Simultaneously, two sample loops were filled with
sample solution by working the peristaltic pump.
In the second stage (injection mode for 2.0 min),
sample solutions which were stored in the two
sample loops were injected into the carrier stream
simultaneously by switching the valve. Switching
of the valve was repeated automatically at a
regular time interval. A delay coil DC was also
inserted in the flow line as shown. Because of the
different residence time of two sample solutions
(SL
1
and SL
2
) to reach the orthophosphate
biosensor, two peaks were obtained. The first
peak corresponded to SL
2
and the second to SL
1
injections. The separation of the two peaks de-
pended on the length of the delay coil; a delay
PTFE coil, 3.0 m/0.5 mm i.d., was needed to
separate the two peaks completely.
T. Yao et al. / Talanta 60 (2003) 845/851 847
3. Results and discussion
3.1. Construction of orthophosphate biosensor
The reactions that occur at the enzyme electrode
prepared in this work are shown schematically in
Fig. 2.
In a preliminary experiment, an orthophosphate
electrode coated with an enzyme membrane alone
was used in a single channel FIA system without
AcP immobilized reactor and 0.1 M sodium
acetate buffer (pH 6.0) containing 10 mM maltose
was pumped at a flow rate of 1.0 ml min
1
. The
injection of orthophosphate into the flow line gave
a well-defined FIA signal because it resulted in the
production of electroactive hydrogen peroxide as
an end product by three successive enzymatic
reactions during diffusion through the composite
trienzyme membrane. However, at an applied
potential of 0.6 V versus Ag/AgCl, the enzyme
electrode gave a fairly large response for each of
ascorbate, urate and cysteine, other than ortho-
phosphate and H
2
O
2
, as shown in Table 1. In
contrast, although the sensitivity of an orthopho-
sphate biosensor hybridized with a poly(1,2-dia-
minobenzene) film was decreased to :/55% of
their previous values for hydrogen peroxide and
Fig. 1. Schematic diagram of two stages (A: ready and charge
mode; B: injection mode) of a ow-injection system with a 16-
way switching valve for simultaneous determination of ortho-
phosphate and total phosphates (inorganic phosphates plus
purine nucleotides). R, acid phosphatase immobilized reactor;
DC, delay coil (PTFE coil, 3.0 m/0.5 mm i.d.); SL
1
and SL
2
,
sample loop (PTFE coil, 30 cm/0.5 mm i.d.; :/60 ml in
volume); detector, orthophosphate biosensor.
Fig. 2. Working principle of orthophosphate biosensor with a
hybrid membrane of trienzyme lm and poly(1,2-diaminoben-
zene) lm. MP, maltose phosphorylase; Mut, mutarotase;
GOD, glucose oxidase.
Table 1
Effect of poly(1,2-diaminobenzene) lm hybridized to trien-
zyme membrane of orthophosphate biosensor
Substrate (0.1 mM each) Response current
a
(nA)
Without film With film
H
2
O
2
960 525
Orthophosphate 822 450
Ascorbic acid 750 B/5
Uric acid 680 B/5
Cysteine 695 B/5
a
Response (peak current) obtained by orthophosphate
biosensor with and without poly(1,2-diaminobenzene) film.
T. Yao et al. / Talanta 60 (2003) 845/851 848
orthophosphate, the currents for ascorbate, urate
and cysteine were decreased up to negligibly small
values (see Table 1). Due to such a size exclusion
function of the poly(1,2-diaminobenzene) film as
described previously [14], the access of such
electroactive compounds to the platinum surface
was blocked. Consequently, orthophosphate bio-
sensor with a hybrid membrane of trienzyme film
and poly(1,2-diaminobenzene) film responded se-
lectively to orthophosphate without any interfer-
ences in the presence of ascorbate, urate and
cysteine below 0.2 mM.
3.2. Analytical characterization of orthophosphate
biosensor
Experiments were carried out to establish the
optimum conditions for the orthophosphate detec-
tion using the trienzyme electrode with poly(1,2-
diaminobenzene) film, especially for the pH and
the maltose concentration of the carrier solution.
Sodium acetate buffers (0.1 M) at various pH
values containing 5 mM maltose, were tested as
the carrier solution, using a single channel FIA
system without AcP immobilized reactor. Max-
imum response to orthophosphate was observed at
pH 5.5.
Without maltose in the carrier solution, no
response to orthophosphate was observed. As the
maltose concentration in the carrier was raised
from 0.5 to 5 mM, higher concentrations, prob-
ably because of oxygen limitation in the carrier
solution. From these results, 0.1 M sodium acetate
buffer at pH 5.5 containing 5 mM maltose was
selected as the carrier solution.
The carrier flow rate is related to the residence
time of the sample zone on the orthophosphate
biosensor. FIA signals for 0.1 mM orthopho-
sphate were recorded at the carrier flow rates in
the range of 0.2/1.0 ml min
1
. The signal
intensity increased with decreasing flow rates, but
the time required to determine became excessive at
low flow rates. Therefore, a flow rate of 1.0 ml
min
1
was recommended.
3.3. Simultaneous determination of orthophosphate
and total phosphates (inorganic phosphates plus
purine nucleotides)
The AcP immobilized reactor was positioned in
the flow line as shown in Fig. 1. Purine nucleotides
and polyphosphates, for example, adenosine nu-
cleotides (ATP, ADP, AMP) and tripolyphosphate
are enzymatically hydrolyzed to generate ortho-
phosphate during passage through the AcP im-
mobilized reactor, as shown in reactions (1) and
(2). The produced orthophosphate is amperome-
trically detected with the downstream orthopho-
sphate biosensor.
In the flow-line in the injection mode (Fig. 1B),
two sample portions (SL
1
and SL
2
) pass through
the flow-line with different residence time before
reaching the orthophosphate biosensor and there-
fore two peaks are obtained for an orthophosphate
injection. The first peak corresponds to SL
2
injection and second to SL
1
. However, enzyme
(1)
(2)
T. Yao et al. / Talanta 60 (2003) 845/851 849
electrode responded selectively to orthophosphate
and not to nucleotides and other inorganic phos-
phates. Therefore, under optimum conditions,
orthophosphate gave two FIA signals separated
completely, while each of nucleotides and other
inorganic phosphates gave only one signal corre-
sponding to the second signal of orthophosphate,
as shown in Table 2. In addition, Table 2 shows
slopes (sensitivities) of calibration plots obtained
for all the purine nucleotides and inorganic
phosphates examined in this work. The ratio of
the sensitivities for nucleoside triphosphate, dipho-
sphate and monophosphate was :/3:2:1 for each
group of the purine nucleotides. Similarly, the
ratio of the sensitivities for tripolyphosphate,
pyrophosphate and orthophosphate was :/3:2:1.
This means that the enzymatic conversions of
purine nucleotides to nucleosides (reaction (1))
and of tripolyphosphate to orthophosphate (reac-
tion (2)) into the AcP immobilized reactor were
almost complete under optimum flow conditions.
As a result, the first peak corresponded to
orthophosphate and the second to total phosphate
(inorganic phosphates plus purine nucleotides).
The proposed FIA system offered high selectivity
for orthophosphate and phosphate containing
compounds. The following species gave zero
relative responses: sulfate, nitrate, chlorate, thio-
sulfate, carbonate, lactate, tartrate, citrate, succi-
nate, oxalate and ammonium. However, glucose
was only an interferent for the measurement of
phosphate because it was responsible on the
biosensor system as a substrate of GOD.
Linear calibration plot for orthophosphate was
obtained over the range 5/10
7
/8/10
4
M for
each of the first and second peaks. The detection
limits were 2.5/10
7
and 4/10
7
M for the first
and second peaks, respectively, based on a signal-
to-noise ratio of three. The slope, y-intercept and
linear correlation coefficient were 4.55 nA mM
1
,
1.15 nA and 0.999 for the first peak and 3.75 nA
mM
1
, 0.98 nA and 0.999 for the second peak of
orthophosphate, respectively. The R.S.D. for se-
ven replicate injections at 1/10
5
M were 1.9
and 2.5% for the first and second peaks, respec-
tively. Also, up to 30 samples per hour could be
analyzed under the optimized flow conditions.
When the AcP immobilized reactor and orthopho-
sphate electrode were used repeatedly to confirm
their stabilities (20 samples per day), their activities
decreased to :/60% of their original values after
60 and 35 days, respectively.
In conclusion, the combination of AcP immo-
bilized enzyme reactor and orthophosphate bio-
sensor into the amperometric flow-injection
system made up by a 16-way switching valve
with two sample loops affords a highly selective,
rapid and simultaneous determination of ortho-
phosphate and other phosphate containing com-
pounds (inorganic phosphates plus purine
nucleotides). The proposed FIA method will be
useful for the determination of phosphate in the
field of environmental and food analysis. Details
of these will be also reported subsequently to-
gether with analytical results of different samples,
e.g. in rain, lake and a variety of foodstuffs.
References
[1] H. Tiessen, Phosphorus in the Global Environment:
Transfers, Cycles and Management, Wiley and Sons,
Chichester, 1995.
Table 2
Substrate specicities and sensitivities for a variety of phos-
phate containing compounds in a simultaneous FIA system
shown in Fig. 1
Substrate Sensitivity (nA mM
1
)
First peak Second peak
Orthophosphate 4.55 3.75
Pyrophosphate B/0.1 7.60
Tripolyphosphate B/0.1 11.4
Adenosine-5?-monophosphate n.d. 3.80
Adenosine-5?-diphosphate n.d. 7.61
Adenosine-5?-triphosphate n.d. 11.3
Inosine-5?-monophosphate n.d. 3.71
Inosine-5?-diphosphate n.d. 7.45
Inosine-5?-triphosphate n.d. 11.0
Guanosine-5?-monophosphate n.d. 3.69
Guanosine-5?-diphosphate n.d. 7.40
Guanosine-5?-triphosphate n.d. 11.1
Xanthosine-5?-monophosphate n.d. 3.62
Xanthosine-5?-diphosphate n.d. 7.25
Xanthosine-5?-triphosphate n.d. 11.0
n.d., Not detected.
T. Yao et al. / Talanta 60 (2003) 845/851 850
[2] H. Kawasaki, K. Sato, J. Ogawa, Y. Hasegawa, H. Yuki,
Anal. Biochem. 182 (1989) 366.
[3] S.O. Engblom, Biosens. Bioelectron. 13 (1998) 981.
[4] Y. Su, M. Mascini, Anal. Lett. 28 (1995) 1359.
[5] S. Cosnier, C. Gondran, J.C. Watelet, W.D. Giovani,
R.P.M. Furriel, F.A. Leone, Anal. Chem. 70 (1998) 3952.
[6] U. Wollenberger, F. Schubert, F.W. Scheller, Sens. Actuat.
B 7 (1992) 412.
[7] H. Kinoshita, D. Yoshida, M. Miki, T. Usui, T. Ikeda,
Anal. Chim. Acta 303 (1995) 301.
[8] S. Huwel, L. Haalck, N. Conrath, F. Spener, Enzyme
Microb. Technol. 21 (1997) 413.
[9] H. Nakamura, M. Hasegawa, Y. Nomura, Y. Arikawa, R.
Matsukawa, K. Ikebukuro, I. Karube, J. Biotechnol. 75
(1999) 127.
[10] H. Nakamura, H. Tanaka, M. Hasegawa, Y. Masuda, Y.
Arikawa, Y. Nomura, K. Ikebukuro, I. Karube, Talanta
50 (1999) 799.
[11] N. Gajovic, K. Habermuller, A. Warsinke, W. Schuh-
mann, F.W. Scheller, Electroanalysis 11 (1999) 1377.
[12] C. Mousty, S. Cosnier, D. Shan, S. Mu, Anal. Chim. Acta
443 (2001) 1.
[13] N. Conrath, B. Grundig, K. Huwel, K. Cammann, Anal.
Chim. Acta 309 (1995) 47.
[14] T. Yao, M. Satomura, T. Nakahara, Anal. Chim. Acta 296
(1994) 271.
[15] T. Yao, Anal. Chim. Acta 148 (1983) 27.
[16] T. Yao, Y. Nanjyo, T. Tanaka, H. Nishino, Electroana-
lysis 13 (2001) 1361.
T. Yao et al. / Talanta 60 (2003) 845/851 851