Establishment of the light-absorption ratio variation approach and application to determination of CO(II) in the ng ml-1 level and novel characterization of Co(II) and Zn(II) complexes with 1,5-di(2-hydroxy-5- sulfophenyl)-3-cyanoformazan

Article abstract of DOI:10.1039/b306826a

The light-absorption ratio variation approach (LARVA) has been established and applied to the determination of Co on the ng ml^-1 level with 1, 5-di(2-hydroxy-5-sulfophenyl)-3-cyanoformazan (DSPCF) at pH 9.33. Competitive replacement complexatio

Full text of DOI:10.1039/b306826a

View Article Online / Journal Homepage / Table of Contents for this issue
Establishment of the light-absorption ratio variation approach
and application to determination of Co(II) in the ng mlꢀ1 level
and novel characterization of Co(II) and Zn(II) complexes with
1,5-di(2-hydroxy-5-sulfophenyl)-3-cyanoformazan
Hong-Wen Gao,*ab Hong-Yan Wang,bc Sheng-Yi Zhangb and Jian-Fu Zhaoa
a
State Key Laboratory of Pollution Control and Resources Reuse, School of Environmental
Science and Engineering, Tongji University, Shanghai 200092, P. R. China.
E-mail: hwgao@mail.tongji.edu.cn
Department of Chemistry, Anhui University, Hefei 230039, P. R. China
b
c
Department of Chemistry and Biology, Suzhou Teachers College, Suzhou 234000, P. R. China
Received (in Montpellier, France) 15th June 2003, Accepted 19th August 2003
First published as an Advance Article on the web 7th October 2003
The light-absorption ratio variation approach (LARVA) has been established and applied to the determination
of Co on the ng mlꢀ1 level with 1, 5-di(2-hydroxy-5-sulfophenyl)-3-cyanoformazan (DSPCF) at pH 9.33.
Competitive replacement complexation (CRC) among DSPCF, Zn(II) and Co(II) as described was used to
improve the analytical selectivity. The determination of the complex formation constants was also improved.
Results have shown that DArꢀ1 (where DAr is the light-absorption ratio variation) is linear in the range of Co(II)
between 10 and 200 ng mlꢀ1. The limit of detection (3d) of Co(II) is only 3.7 ng mlꢀ1. The combination of CRC
and LARVA has been applied to the analysis of water quality with satisfactory results.
reaction, too. For the same purposes as above, many highly
Introduction
sensitive chromophores such as porphyrins, hydrazine and its
derivatives, oximes and sulfonates are increasingly synthesized
and applied.18 In spite of these efforts, the sensitivity up-limit
of spectrophotometric analysis has not yet been substantively
improved.
In trace analysis, especially in control of materials and the
environmental, a number of expensive instruments such as
ICP-AES, ETAAS and XRF are used extensively. Most of
them are used for elemental analysis and often offer a high sen-
sitivity and carry out the simultaneous determination of multi-
ple components. However, they are not appropriate for the
study of the mechanism of a chemical reaction and for mole-
cular morphological analysis. From the periodic reviews of
‘‘ultraviolet and light absorption spectrometry’’ published
from 1978 to 1998,1 we learn that spectrophotometry is an
effective and practical analytical method though its study
trends to weaken in recent years, particularly in advanced
countries. The development of reagents, analytically significant
chromophoric systems and spectrophotometric methodology
provides light absorption spectrometry with a very wide range
of applications for the analysis of food,2 water quality3 and
tropospheric substances4 and in the identification of gems,5
studies of inorganic particle formation,6 solid-phase struc-
tures7 and protein folding and unfolding.8 Nowadays, activity
in the area of spectrophotometric analysis of inorganic species
seems to have remained relatively constant. In the area of
photometric methods for organic substances, activity has not
significantly increased but continues to center around applica-
tions in the clinical, pharmaceutical and biological areas.
Many of the improvements including instrumentation and
components,9 chemometrics10 and optimization techniques11
have been significantly developed. Computers and micropro-
cessors seem to be stimulating a great deal of interest in the
areas of multiple wavelength,12 derivative13 and flow injection
techniques.14 All of these have played an important role in the
improvement of selectivity, accuracy, sensitivity and precision
of trace analysis.
In the present work, the light-absorption ratio variation
approach (LARVA) is proposed. It is established on the basis
that a sensitive variation of the light-absorption ratio of a reac-
tion between a chromophore and a detected component
appears with the component concentration at the ultra-trace
level. In order to enhance selectivity of the reaction, a conven-
tional method is to add a number of masking reagents in the
solution. However, competitive replacement complexation
(CRC)19 is simple and effectively improves the selectivity; it
is described and applied in the present work. The spectral cor-
rection technique (SCT)20 is a very useful method for the
characterization of a chemical reaction such as ligand-metal
complexation or the interaction between small molecules and
biomacromolecules.21 The combination of SCT, CRC and
LARVA not only characterizes clearly the complexation but
also make trace analysis much more selective and highly
sensitive.
In the present work, 1,5-di(2-hydroxyl-5-sulfophenyl-)-3-
cyanoformazan (DSPCF) was selected as the ligand to com-
plex Co(II) and Zn(II) at pH 9.33. DSPCF is a strong colorant
with a high molar light absorptivity (e548nm ¼ 1.47 ꢁ 104
l
molꢀ1 cmꢀ1). Its structure, shown in Fig. 1, belongs to the cya-
noformazan group and is a typical multi-coordinative ligand.
It may complex many metal ions such as Cu(II), Fe(II), Zn(II),
Mn(II), Ni(II), Co(II), Al(III) and Pb(II) to form binary com-
plexes. So, it is not fit for the direct determination of any indi-
vidual metal ion. However, we found that Co(II) may replace
the Zn(II) coordinated in the Zn-DSPCF complex to form a
blue Co-DSPCF complex. The replacement complexation is
sensitive and selective in the presence of a high concentration
The addition of an assistant such as a surfactant,15 a cyclo-
dextrin16 or an association ion17 is often necessary in a color
DOI: 10.1039/b306826a
New J. Chem., 2003, 27, 1649–1656
1649
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
The effective fraction (Z) of L and the complexation ratio (g)
of L to M are determined and calculated by the relations:20
AC2 ꢀ Al2
Z ¼
þ 1
ð1Þ
A0l2
and
CL0
g ¼ Z ꢁ
ð2Þ
ð3Þ
CM0
Fig. 1 Chemical structure of DSPCFH.
where
of Zn(II). The Zn-DSPCF and Co-DSPCF complexes have
been characterized by SCT and replacement complexation
has been used for the determination of Co(II) at the ng mlꢀ1
level in water by LARVA. Certainly, the Zn/Co-DSPCF reac-
tion is only an example. In fact, many similar complexations
such as metal ion-dye ligand, protein-organic molecule or
drug-oxidant can be characterized by SCT and most of these
might be used for the ultra-trace analysis of metal ions, drugs,
biomacromolecules, inorganic and organic compounds by
LARVA in the future.
Al2 ꢀ bAl1
1 ꢀ ab
AC2
¼
and where the symbols CM0 and CL0 are the initial concentra-
tions of M and L. In Fig. 2(1), 2(3) and 2(5) and Fig. 3(A), AC2
indicates the real absorbance of only the ML complex at wave-
length l2against water, which cannot be measured directly. Al02
is the absorbance of the L solution measured at l2 against
water. The symbols Al2 and Al1 are the absorbances of the
M–L solution, respectively measured at l2and l1 against water.
Both b and a are correction constants and they may be calcu-
lated from curves 1 and 2 by means of:
A0l2
A0l1
Principles
b ¼
ð4Þ
Improvement in the determination of the stepwise stability
constants Kn
and
ACl2
ACl1
In the earlier determinations of the stepwise stability constants
(K1, . . ., Kn, . . ., KN) of a ligand (L)–metal (M) complex MLN
(where N is the maximum coordination number), several repe-
titive measurements of only a single solution were used.19–21
Thus, it is possible for Kn to have an assignable error. Here,
a multi-point measurement method was established and the
accurate determination of Kn can be made by linear regression.
The complexation between L and M (M1 or M2) illustrated in
Fig. 2 proceeds via the stepwise reactions:
a ¼
ð5Þ
The M–L solution with a g of L to M varying between n ꢀ 1
and n mostly consists of both MLn ꢀ 1 and MLn complexes and
free L, especially for n ꢀ 0.7 < g < n ꢀ 0.3. Thus, the other
complexes MLn ꢀ 2 and MLn + 1only represent a too small
fraction to affect the determination of Kn . The n-step stability
constant (Kn) of complex MLN is calculated by the relation:
K1
*
L
L
þ
þ
M
)
ML
y ¼ Knx
ð6Þ
K2
*
ML
)
ML2
where
ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ
x ¼ CL0 ꢀ gCM0
Kn
*
L
þ
MLnꢀ1
)
MLn
MLN
ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ ꢂ
KN
*
L
þ
MLNꢀ1
)
Fig. 2 Schematic of the color reaction between L, M1 and M2 : (1) L
solution (purplish red), (2) M1L complex solution (blue) where
CM1 ꢃ CL0 , (3) M1–L reaction solution (violet) consisting of (4) an
excess of L (light red) and (5) M1L complex (light blue), (6) color mix-
ture of M1L (blue) and M2L (green) where M1 is replaced in a number
of M1L complexes by M2 to form the M2L complex.
Fig. 3 (A) Plots of the absorption spectra for the establishment of the
LARVA methodology: (curve 1) solution of L, (curve 2) ML complex
solution without free L, (curve 3) M–L reaction solution, (curve 4)
virtual spectrum of excess of L from curve 3 and (curve 5) virtual
spectrum of ML complex formed from curve 3. B- Plots of DAr vs. CM
.
0
1650
New J. Chem., 2003, 27, 1649–1656

View Article Online
and
in the contribution of the instrument’s noise signal. Therefore,
the higher the light absorptivity of the chromophore and
the more sensitive the spectrometer, the lower the detection
limit of a component in the LARVA method. In the present
case, the addition of the chromophore to produce a peak
absorbance between 0.05 and 0.2 is suggested.
ꢀ
ꢁ
1
y ¼
ꢀ 1
n ꢀ g
(for n ꢀ 0.7 < g < n ꢀ 0.3).
Therefore, each Kn can be determined by varying the molar-
ity of L in a constant molarity M solution. From all Kn , the
cumulative stability constant (K) of MLN is calculated by:
Competitive replacement complexation (CRC)
N
Y
Eqn. (7) is now for the first time derived and applied in this
work.
In order to improve the analytical selectivity, CRC was devel-
oped and it has been applied to a number of complexation pro-
cesses with satisfactory results.19 As shown in Fig. 2(1) and
2(2), in the reaction between L and M1 , all of the L molecules
are complexed by M1 when the M1 molarity is much higher
than the L molarity. Thus, only one colored compound, the
M1L complex, exists in such a solution. However, the addition
of M2 will replace competitively M1 from the M1L complex to
form a mixture of both the colored complexes M1L and M2L,
as shown in Fig. 2(6). Thus, the M1L complex is regarded as a
new chromogenic reagent to use in the determination of trace
amounts of M2 . Because M1 is present in a high enough
amount in the replacement reaction, only the M1L (1 : 1) com-
plex is formed and it will prevent the interference of many
metal ions possibly present at the micro level in a sample. In
most cases, it is not necessary to add any masking reagent
in trace analysis. Moreover, CRC obviously improves the
analytical selectivity. LARVA is still appropriate for the
replacement complexation.
K ¼
Kn
ð7Þ
n¼1
Light-absorption ratio variation approach
The stepwise complex reactions above are merged as follows:
n L
CL0ðA0l1 at
þ
M
Ð
MLn
Initiation
CM0
0
l1 and A0l2 at l2Þ
CL ¼ CL0 ꢀnCM0 CM ! 0
ðAL1 at l1 and
L2 at l2Þ
From the illustration in Figs 2(1), 2(3), 2(4), 2(5) and 3(A),
Al1 ¼ AL1 þAC1 ð7Þ
CM0ðAC1 at
Equilibrium
l1 and AC2 at l2Þ
A
Experimental section
and
Al2 ¼ AL2 þAC2
Apparatus and reagents
Al2 AL2 þAC2 dCLeL þdCM0eMl2L
Absorption spectra were recorded on a TU1901 spectrophoto-
meter (PGeneral, Beijing). The pH of the solution was mea-
sured with a Model delta 320 pH-meter (Mettler Toledo
Group, Shanghai). A Model BS110S electronic balance (Sar-
torius Instruments, Beijing) was used to accurately weight
the standard substances and DSPCF. A Model 110R electric
thermal constant-temperature bath was used to adjust the tem-
perature of the replacement reaction. A Model PLA-SPEC
ICP-AES (Leeman Instruments) was used to examine the
quantitative result of Co(II) obtained by LARVA.
dCLeLl2 þdCM0eML
Ar ¼
¼
¼
;
Al1 AL1 þAC1
l1
l1
A0l2 dCL0eLl2
Ar0
Ar ꢀAr0
¼
¼
¼
A0l1 dCL0eLl1
CLeL þCM0eMl2L elL2
CLeLl2 þCM0eML el1
ꢀ
L
l1
l1
ꢂ
ꢃ
ML
l2
ML
l1
eLl1
e
ꢀeLl2e
CM0
DAr ¼
ꢂ
ꢃ
2
ðCL0 ꢀnCM0Þ eLl1 þCM0elM1LelL1
Stock standard solutions of Zn(II) (200 mg mlꢀ1) and Co(II)
(200 mg mlꢀ1) were prepared by dissolving the metals in spec-
trometric purity in 20% hydrogen chloride and diluting with
deionized water. Standard solutions of 1.00 mg mlꢀ1 Zn(II),
0.100 and 1.00 mg mlꢀ1 Co(II) were prepared daily by diluting
the above stock solution. Standard stock DSPCF (0.800
mmol lꢀ1) solution was prepared by dissolving 243 mg
of disodium 1,5-di(2-hydroxy-5-sulfophenyl)-3-cyanoforma-
zan (content 80%, provided by Changke Reagents Institute
of Shanghai) in 500 ml deionized water. It was used to complex
Co(II) and Zn(II). The 0.200 mmol mlꢀ1 DSPCF solution was
prepared accurately by diluting its standard stock solution.
The ammonium buffer solutions at pH 7.26, 8.05, 8.52, 9.11,
9.33, 9.59, 9.85, 10.11 and 10.52 were prepared with ammonia
water and ammonium chloride or ammonium acetate. The Zn–
DSPCF complex (0.0400 mmol lꢀ1) solution was prepared by
mixing 80 ml of 200 mg mlꢀ1 Zn(II), 25 ml of pH 9.33 buffer
solution and 25.0 ml of 0.200 mmol lꢀ1 DSPCF, then diluting
to 250 ml with deionized water. Trace amounts of Co(II) may
replace Zn from the Zn–DSPCF complex.
ꢂ
ꢃ
ꢂ
ꢃ
2
CL0 eLl1
eLl1
e
ꢀneLl1
ML
l1
DAꢀr 1
¼
ꢂC
þ
e
ꢀ1
M0
ML
l2
ML
l1
ML
l2
ML
e
eLl1 ꢀe eLl2
M0
eLl1 ꢀe eLl2
l1
;DAꢀr 1 ¼ p0C þq0
ð8Þ
ꢀ1
where
and
ꢂ
ꢃ
2
CL0 eLl1
p0 ¼
ML
l2
ML
l1
e
eLl1 ꢀe eLl2
ꢂ
ꢃ
eLl1
e
ꢀneLl1
ML
l1
q0 ¼
:
ML
l2
ML
e
eLl1 ꢀe eLl2
l1
Both p0 and q0 are constants.
A sketch of eqn. (8) is shown in Fig. 3(B). The plot DAr vs.
CM is linear when the molarity of M is much less than that of
L, that is C0M0 . Thus, eqn. (8) can be rewritten as:
0
DAr ¼ pCM þ q
ð9Þ
0
where the symbols p and q are constants. Eqns. (8) and (9) can
both be directly used in quantitative detection of traces of M.
This method is called the light-absorption ratio variation
approach (LARVA). From the equation above, the sensitivity
factor p0ꢀ1 is the inverse ratio of CL0 . Therefore, the analytical
sensitivity increases as less L is added. Nevertheless, a too-low
amount of L will certainly bring an obvious error—an increase
General procedures
Characterization of the Zn–DSPCF and Co–DSPCF com-
plexation reactions. Into ten 10 ml calibrated flasks were added
1.00 mg of Zn(II), 1 ml of pH 9.33 buffer solution and 0.0200
mmol lꢀ1 DSPCF solution from 0.00 to 6.00 ml in steps of
New J. Chem., 2003, 27, 1649–1656
1651

View Article Online
0.50 ml. The mixture was diluted to 10 ml with deionized water
and mixed thoroughly. After 10 min, the absorbances were
measured at 638 and 510 nm against water and AC , CL , Z
and g were calculated by eqn. 1 and 2. In the same way, into
another ten 10 ml calibrated flasks were added 2.00 ml of
1.00 mg mlꢀ1 Zn(II) and the other reagents were added as
above. Finally, both N and Kn can be calculated by a linear
regression of eqn. 6. Thus, the Zn–DSPCF complex was char-
acterized. Similarly, we used Co(II) to replace Zn(II) and
the above procedures were repeated. The absorbances were
measured at 695 and 558 nm against the reagent blank. The
Co–DSPCF complex was characterized as well.
solution. If the solution has a pH less than 9, it is possible that
the buffer capacity of the ammonium acetate is inadequate for
a practical sample. Therefore, a pH 9.33 ammonium chloride–
ammonia buffer solution was selected for this work. From
curve 5 in Fig. 4(A), the peak is located at 695 nm and the
valley at 558 nm. From curve 5 in Fig. 4(B), the peak is located
at 638 nm and the valley at 510 nm. These four wavelengths
were used for the characterization of the Co–DSPCF and
Zn–DSPCF complexes.
Composition of the complexes
From Fig. 5(A), the absorbance ratio A558 nm/A695 nm of the
Co–DSPCF solutions decreases rapidly with increase of Co(II)
molarity. It approaches a minimum and remains constant
when the amount of Co(II) is more than once that of DSPCF.
This indicates that DSPCF is complexed almost completely
with Co(II) and only a colored compound, the Co–DSPCF
complex, exists in the presence of enough excess Co(II). The
absorption spectrum of the Co–DSPCF complex is shown as
curve 2 in Fig. 5(C). In addition, the break point approach22
was applied to estimation of the complexation ratio. From
Fig. 5(A) the break point is located at about a 0.5 : 1 ratio of
Co to DSPCF. So the Co(DSPCF)2 complex is possibly
formed, which will be confirmed below. In the same way, from
Fig. 5(B), the ratio A510 nm/A638 nm of the Zn–DSPCF solu-
tions approaches a minimum and remains constant when the
amount of Zn(II) added is more than 0.8 times the DSPCF
molarity. Only one colored compound, the Zn–DSPCF com-
plex, exists in the presence of enough excess Zn(II) and its
absorption spectrum is shown as curve 3 in Fig. 5(C). In addi-
tion, the break point in Fig. 5(B) is located at about a 0.35 : 1
ratio of Zn to DSPCF. So the Zn(DSPCF)3 complex is possi-
bly formed in this case. From Fig. 5(C), DSPCF is a violet
solution at pH 9.33 and its peak absorbance is located at
Determination of ultra-micro amounts of Co(II) in water. A
water sample (5.00 ml) was added into a 10 ml flask and
1 ml of pH 9.33 buffer solution and 2.00 ml of the 0.040
mmol lꢀ1 Zn-DSPCF complex solution were added. The solu-
tion was diluted to 10 ml with deionized water. After 10 min,
the absorbances were measured against water. In the meantime,
a reagent blank without Co(II) was prepared and measured the
same as that above. Thus, DAr is calculated by the relation:
A679nm A0679nm
DAr ¼
ꢀ
ð10Þ
A0574nm
A574nm
are the absorbances of the Co–
nm
where A679
and A574
nm
Zn(DSPCF) solution measured at 679 and 574 nm against
water. A0679nm and A5074nm are the absorbances of the solution
without Co(II), respectively measured at 679 and 574 nm
against water. From eqns. (8) or (9), CCo in the sample was
calculated.
Results and discussion
Variation of absorption spectra with pH
548 nm with a molar light absorptivity of e ¼ 1.47 ꢁ 104
l
The absorption spectra of the Co–DSPCF and Zn–DSPCF
solutions in various pH buffer mediums are shown in Fig. 4.
From the 9 curves in Fig. 4(A), the complexation between
Co and DSPCF is always sensitive when the pH is less than
9.59. In the same way, from Fig. 4(B), curve 5 shows the
highest absorption peak. In a basic solution with pH over 10,
the dehydrogenation of two –OH and the protonation of
–NH– of DSPCF will occur simultaneously and the mixed
ligand complexes of Zn(II) and Co(II), that is Zn(OH)(DSPCF)
and Co(OH)(DSPCF), will also be possibly formed. Thus, the
coordination of DSPCF with Zn(II) and Co(II) is weakened.
We also found that the replacement complexation between
Co(II) and Zn(DSPCF) became difficult in a strongly basic
molꢀ1 cmꢀ1 from curve 1 in Fig. 5(C). The Co–DSPCF com-
plex is a green solution and its peak absorbance is located at
641 nm in curve 2. The Zn–DSPCF complex is a blue solution
and its peak absorbance is located at 615 nm, as shown in
curve 3. By comparing curve 1 with curves 2 and 3, the red
shift of the peak of the Co–DSPCF complex reaches 93 nm
but that of the Zn–DSPCF complex is only 67 nm. Therefore,
the Co–DSPCF coordination bond is stronger than the
Zn–DSPCF bond. From curves 1, 2 and 3, the correction
¼ 0.451,
DSPCF
638nm=510nm
coefficients were calculated to be
ZnðDSPCFÞ
b
CoðDSPCFÞ
DSPCF
695nm=558nm
b
¼ 0.0667, a
¼ 0.565 and a
¼
510nm=638nm
558nm=695nm
0.616. These will used to calculate Z, g and CL below.
The variation of the g value of DSPCF to Co(II) and Zn(II)
with DSPCF molarity is shown in Fig. 6. g approaches a maxi-
mum constant value of 2 in curve 1 and of 3 in curve 2 when
DSPCF is more than 0.015 mmol lꢀ1. Therefore, the spectral
correction technique confirms that complexes Co(DSPCF)2
and Zn(DSPCF)3 are formed at pH 9.33. The possible struc-
tures of ML complexes (1:1) are given in Fig. 7. The coordina-
tion of both Zn and DSPCF forms a macrocycle but its
tightness is poor. The circle lengthens the conjugate chain con-
sisting of the p-bond and lone pair electrons of oxygen and
nitrogen to result in the spectral red shift of the Zn–DSPCF
complex. However, because Zn is not coordinated directly with
=
=
the –N N–C(CN) N–NH– functional group, the spectral red
shift of the Zn–DSPCF complex is only 67 nm as mentioned
above. Thus, the complexed Zn is easily replaced by other
metal ions, for example Co(II). On the contrary, the O–Co–
N chains divide the initial structure of the Co–DSPCF
complex into three small adjacent circles so as to stabilize the
complex structure. In addition, Co(II) coordinates directly to
Fig. 4 Variation of the absorption spectra with pH. (A) Co (1.00 mg
mlꢀ1)–DSPCF (0.040 mmol mlꢀ1) solution measured against a blank
without Co(II). (B) Zn (1.00 mg mlꢀ1)–DSPCF (0.040 mmol mlꢀ1) solu-
tion measured against a blank without Zn(II). Spectra 1 to 9: pH 7.26,
8.05, 8.52, 9.11, 9.33, 9.59, 9.85, 10.11 and 10.52.
=
=
the head-N and the end-N of the –N N–C(CN) N–NH–
functional group to dispersively extend its conjugation and this
1652
New J. Chem., 2003, 27, 1649–1656

View Article Online
Fig. 5 Formation of the Co–DSPCF and Zn–DSPCF complexes at pH 9.33 where both the solutions no longer contain an excess of DSPCF . (A)
Variation of A558 nm/A695 nm versus the initial molar ratio of Co(II) to DSPCF (fixed at 0.040 mmol mlꢀ1). (B) Variation of A510 nm/A638 nm versus
the initial molar ratio of Zn(II) to DSPCF (fixed at 0.040 mmol mlꢀ1). (C) Curves 1, 2 and 3 are the absorption spectra of DSPCF, Co(DSPCF) and
Zn(DSPCF) complexes, all at 0.040 mmol mlꢀ1 measured against water, while curve 4 shows the absorption spectrum of the solution initially con-
taining 0.040 mmol mlꢀ1 Zn(DSPCF) and 0.1 mmol mlꢀ1 Co(II) against water. The solutions’ real color reactions are shown in the text tubes.
results in an obvious red shift of the absorption spectrum of
the Co–DSPCF complex mentioned above.
two solutions both contain only one identical colored complex,
Co(DSPCF). Therefore, Co(II) replaces all Zn(II) in the Zn-
(DSPCF) complex in the solution containing sufficient Co(II)
Plots of [(n ꢀ g)ꢀ1 ꢀ 1] vs. (CL0 ꢀ gCM0) of the solutions
above are shown in Fig. 8. From curves 1 and 2 in Fig.
8(A), the first-step and second-step stabilities of complex
Co(DSPCF)2 were calculated to be K1 ¼ 1.29 ꢁ 107 and
K2 ¼ 5.54 ꢁ 105 l molꢀ1. With the same method, from curves
1, 2 and 3 in Fig. 8(B), the first-, second- and third-step stabi-
lities of complex Zn(DSPCF)3 were calculated to be K1 ¼
4.44 ꢁ 107, K2 ¼ 5.56 ꢁ 106 and K3 ¼ 1.22 ꢁ 106 l molꢀ1. Thus,
their cumulative stability constants are KCo–DSPCF ¼ 7.15 ꢁ
at pH 9.33. However, from the results above, K1,Co–DSPCF
K
<
1,Zn–DSPCF . Therefore, using only the difference in K1 of
two complexes to judge a replacement reaction is one-sided
and sometimes even incorrect. In addition, the dynamics of
the reaction and the stability of the chelated structure should
also be considered.
Fig. 9 shows the absorption spectrum of the replacement
complexation between Co(II) and Zn(DSPCF) complex at
pH 9.33, measured against a reagent blank without Co(II).
From the obvious peak–valley difference of the solution only
containing 0.100 mg mlꢀ1 Co(II), the replacement complexation
1012 l2 molꢀ2 and KZn–DSPCF ¼ 3.01 ꢁ 1020 l3 molꢀ3
.
Competitive replacement complexation between Zn(DSPCF)
and Co(II)
Analysis of the absorption spectra of the replacement com-
plexation. From curves 2 and 4 in Fig. 5(C), we observe that
two spectra are completely coincident. This means that the
Fig. 6 Variation of g with DSPCF molarity: curve 1 for Co and curve
2 for Zn (both at 0.200 mg mlꢀ1).
Fig. 7 Structural graphs of the Co–(DSPCF) and Zn–(DSPCF)
complexes.
New J. Chem., 2003, 27, 1649–1656
1653

View Article Online
Fig. 8 Determination of the stepwise stability constants at pH 9.33
and 20 ꢄC through plots of [(n ꢀ g)ꢀ1 ꢀ 1] vs. (CL0 ꢀ gCM0). (A) Co–
DSPCF complex and (B) Zn–DSPCF complex. Curve 1 is for the
first-step stability constant (K1), curve 2 for the second-step stability
constant (K2) and curve 3 for the third-step stability constant (K3).
Fig. 10 Effect of temperature (A) and time (B) on DAr of the replace-
ment complexation reaction studied at 20 ꢄC: (A) 0.0040 mmol mlꢀ1
Zn(DSPCF) and 0.100 mg mlꢀ1 Co(II), measured after coloring for
20 min; (B) 0.200 mg mlꢀ1 Co(II) and 0.0080 mmol mlꢀ1 Zn(DSPCF).
is very sensitive. The peak is located at 679 nm and the valley
at 574 nm. These two wavelengths were used in the determina-
tion of DAr below.
replacement complexation after coloring for 12 h is only 1.2
times higher than that after coloring for 0.5 h.
Effect of temperature and time on the replacement complexa-
tion. From curve A in Fig. 10, DAr has a wave-like variation
between 0.19 and 0.23 with temperature between 20 and
80 ꢄC but the peak-valley difference occupies only about
20%. Therefore, 20 ꢄC was selected for the replacement com-
plexation. From curve B, D Ar reaches a maximum constant
value after 20 min. The replacement reaction is rapid though
the Zn(II) molarity is much greater than that of Co(II) in solu-
tion. The light absorption measurement of the replacement
solution was thus carried out after coloring for 20 min.
Calibration graphs and detection limit of Co(II)
Three series of standard Co(II) solutions between 0 and 0.200,
0 and 0.200 and 0 and 0.400 mg mlꢀ1 were prepared and 1.00,
2.00 and 3.00 ml of 0.0400 mmol lꢀ1 Zn(DSPCF) complex
were added, respectively. The absorbances of each were mea-
sured. The DAr of each solution was measured. The calibration
graphs are shown in Fig. 12 and the parameters p’ and q’ of
the regression equation DArꢀ1 ¼ p’CCo( )ꢀ1 + q’ are given in
II
ꢀ1
Table 1. Plots of DArꢀ1 vs. CCO are all linear. The detection
Effect of the Zn–DSPCF complex on the replacement com-
plexation. From the variations shown in Fig. 11 of DAr of
the solutions with a constant initial molar ratio of Co(II) to
Zn(DSPCF) at 25 mg mmolꢀ1, DAr decreases slowly with
increasing Zn(DSPCF) concentration. The primary reason is
that the self-aggregation of the Zn–DSPCF complex will not
occur in dilute enough solution. In addition, the contribution
of instrumental noise will increase, especially when the absor-
bance is less than 0.020. Of course, a high-sensitive and low-
noise spectrometer can overcome such a defect. By comparing
curves 1 and 2 in Fig. 11, one sees that the DAr of the
limits of Co(II), defined as the blank value plus 3 times the
standard deviation of 8 replicated blanks, were calculated
and are also given in Table 1. By comparing them, Line 2
shows the lowest detection limit at only 3.7 ng mlꢀ1. Therefore,
the addition of 2.00 ml of a 0.040 mmol lꢀ1 Zn(DSPCF) com-
plex solution to the replacement reaction solution gave the best
results for the sensitive determination of ultra-micro amounts
of Co(II) in water samples.
Fig. 11 Effect of Zn(DSPCF) concentration on DAr of solutions initi-
ally containing DSPCF (from 0.0020 to 0.016 mmol mlꢀ1) and Co(II)
(from 0.050 to 0.400 mg mlꢀ1), where the initial molar ratio of Co to
Zn(DSPCF) always remain constant at 25 mg mmolꢀ1. Curve 1 mea-
sured 0.5 h after the beginning of the replacement reaction. Curve 2
measured after 12 h.
Fig. 9 Absorption spectrum of the replacement solution containing
0.0040 mmol mlꢀ1 Zn(DSPCF) and 0.100 mg mlꢀ1 Co(II), measured
against a reagent blank without Co(II) at pH 9.33 and 20 ꢄC.
1654
New J. Chem., 2003, 27, 1649–1656

View Article Online
Table 2 Effect of foreign metal ions and organic compounds on the
DAr of a solution containing 2.00 mg of Co(II) and the error
Foreign ion or
compound X
X added/
mg (10 ml)ꢀ1
a
DAr
% Error
None
Ca(II)
Mg(II)
—
100
50
10
10
10
2.0
2.0
10
2.0
10
5.0
10
10
10
5.0
5.0
50
20
20
50
10
10
10
50
10
10
10
10
3
0.1630
0.1561
0.1561
0.1528
0.1599
0.1491
0.138
0
ꢀ4.2
ꢀ4.2
ꢀ6.2
ꢀ1.9
ꢀ8.5
ꢀ15.4
+16.9
+4.4
+13.7
ꢀ8.8
ꢀ4.3
ꢀ4.4
ꢀ2.8
ꢀ6.8
ꢀ11.4
ꢀ3.1
ꢀ4.2
+0.1
+5.0
+4.2
ꢀ2.5
+2.5
+0.1
+9.6
+3.1
+6.9
ꢀ6.8
ꢀ7.4
ꢀ7.9
ꢀ6.9
ꢀ0.67
Al(III)
Fe(II)
Fe(III)
Cu(II)
Mn(II)
0.1621
0.1702
0.1854
0.1487
0.156
Pb(II)
Ni(II)
Cd(II)
Ge(IV)
Sn(II)
0.1559
0.1584
0.1519
0.1444
0.1579
0.1561
0.1631
0.1711
0.1698
0.1589
0.1671
0.1632
0.1786
0.1681
0.1743
0.1519
0.1509
0.1501
0.1517
0.1619
Cr(III)
Ba(II)
Fig. 12 Plots of DArꢀ1 vs. CCOꢀ1 for the determination of Co(II) with
Zn(DSPCF) complex as the replacement reactant at pH 9.33. Line 1:
0.0040 mmol mlꢀ1 Zn(DSPCF) and Co(II) from 0.010 to 0.200 mg
mlꢀ1; line 2: 0.0080 mmol mlꢀ1 Zn(DSPCF) and Co(II) from 0.010 to
0.200 mg mlꢀ1; line 3: 0.0120 mmol mlꢀ1 Zn(DSPCF) and Co(II) from
V(V)
Hg(II)
Acetic acid
Acetone
Formal
Ethanol
Benzoic acid
Lysine
0.010 to 0.400 mg mlꢀ1
.
Effect of foreign ions and compounds
Cystine
Glucose
Citric acid
Salicylic acid
Tartaric acid
Lactic acid
EDTA
Thiourea
SDS
In fact, DSPCF can coordinate many metal ions such as
Pb(II), Cd(II), Hg(II), Al(II), Cu(II), Mn(II), Fe(II), Ni(II). How-
ever, we found that most metal ions, including Pb(II), Cd(II),
Hg(II) and Al(II) in less than 10 mg lꢀ1 levels will not complex
DSPCF in the presence of plenty of Zn(II). Mn(II), Ni(II),
Fe(II) and Cu(II) at the same concentration a Co(II) will not
seriously affect the direct determination of Co(II). Simple
organic compounds such as acetic acid, acetone, formal, etha-
nol, benzoic acid, lysine, cystine, glucose, citric acid, salicylic
acid, tartaric acid, lactic acid and sodium dodecyl sulfate
(SDS) at less than 10 mg lꢀ1 levels did not interfere with the
replacement coordination of Co(II). However, ethylenediamine
tetraacetic acid (EDTA) and thiourea at over 3 mg lꢀ1 levels
will affect the direct determination of Co(II). The effects of six-
teen metal ions and fifteen organic compounds on the replace-
ment complexation are shown in Table 2. Beyond all doubt,
the presence of a dye compound, for example, in a sample will
interfere seriously with the absorption spectrum of the replace-
ment complexation solution. Therefore, dye wastewaters
must be pre-treated through an active carbon filter layer
before coloring.
3
100
a
Error ¼ (DAXr þ Co ꢀ DACr o)/DArCo ꢁ 100
plasma-atomic emission spectrometry (ICP-AES) and the
results are also given in Table 3. The two methods show almost
the same results. Therefore, LARVA is accurate and feasible
for the ultra-trace analysis of a component.
Conclusion
In principle LARVA is different from the traditional absor-
bance ratio23, which is often used to examine the purity of
an organic compound or to identify its structure, such as
drugs,23 polymers24 or proteins.25 It is suggested for LARVA
to use a deeply colored chromophore and a low noise
and highly sensitive spectrophotometer. Thus, it can play a
Analysis of water samples
As a test of the method, Co(II) in three water samples was
determined. The results are listed in Table 3. The recovery
rates of Co(II) added are between 90.1 and 104% and the
RSD less than 11%. The samples were also examined, after
pre-concentration of the samples, by inductively coupled
Table 3 Determination of Co(II) water samples
Sample
from
Co(II) added/ Co(II) found/
%
%
RSD Recovered
mg lꢀ1
0.000
0.050
mg lꢀ1
a
Table 1 Standard regression equation parameters and detection limits
for Co(II) obtained from Fig. 12
Mine well
< 0.0037
(0.0025)b
0.048
96
Zn(DSPCF)
Co(II)/ added/
Line mg mlꢀ1 mmol lꢀ1
a
Huaihe River 0.000
0.070 (ꢅ0.0019)
2.66
(0.075)b
0.160
a
p0
q0
R
sb
DTc /ng mlꢀ1
0.100
90.1
11
a
Lake water
0.000
0.100
0.0041(ꢅ0.00045)
1
2
3
0–0.200 0.004
0–0.200 0.008
0–0.400 0.012
0.9673 ꢀ4.63 0.9973 0.00176 5.1
2.2659 ꢀ5.85 0.9988 0.00055 3.7
2.5747 0.229 0.9968 0.00068 5.3
(0.0037)b
0.104
104
a
Average of 3 replicate determinations. b One determination by ICP-AES
after pre-concentration.
a
c
b
Linear correlation coefficient. Standard deviation of 8 replicated blanks.
Detection limit of Co(II) was calculated by DT ¼ 3sp0.
New J. Chem., 2003, 27, 1649–1656
1655

View Article Online
10 (a) R. G. Brereton, Analyst, 1995, 120, 2313; (b) H. Mark and J.
Workman, Spectroscopy (Eugene, Oreg.), 1994, 9, 26; (c) H.
Mark and J. Workman, Spectroscopy (Eugene, Oreg.), 1994, 9,
30; (d ) D. Perez-Bendito, Analyst, 1990, 115, 748; (e) A. Lorben,
Anal. Chem., 1984, 56, 1004.
11 (a) W. Cao, F. Yongjian and Z. Shifu, Fenxi Huaxue, 1994, 22,
781; (b) L. Costadinnova and T. Nedeltcheva, Analyst, 1995,
120, 2217; (c) J. Zolgharnein, H. Abdollahi, D. Jaefarifar and
G. H. Azimi, Talanta, 2002, 57, 1067; (d ) K. Mutaftchiev and
K. Tzachev, Phytochem. Anal., 2003, 14, 160; (e) M. Bos and E.
Hoogendam, Anal. Chim. Acta, 1992, 267, 73.
12 (a) F. Bosch-Reig, A. Campins-Falcop, A. Sevillano-Cabeza, R.
Herraez-Hernandez and C. Molins-Legua, Anal. Chem., 1991,
63, 2424; (b) V. Ulvi, J. Pharm. Biomed. Anal., 1998, 17, 77; (c)
H. W. Gao, Y. C. Li, P. F. Zhang, M. Tao and L. Wang, J. Anal.
Chem., 2001, 56, 1007.
revolutionary role in trace analysis. Using this principle,
many of the present color reactions may be improved notably.
If applied to fluorophotometric analysis, the improvement
of the sensitivity and selectivity should be obvious, too. The
combination of SCT, CRC and LARVA is suitable for most
of the color change reactions currently in use, including the
quantitative detection of metal ions, inorganic and organic
compounds, polymers and biomolecules in trace levels and
analysis of the reaction mechanism occurring in diluted
solutions, thin membranes and solid powders. Therefore, we
believe that the originality of LARVA will strongly promote
the design and synthesis of highly sensitive and selective
chromophores.
13 (a) K. Kitamura, H. Mano, Y. Shimamoto, Y. Tadokoro, K.
Tsuruta and S. Kitagawa, Fresenius’ J. Anal. Chem., 1997, 358,
509; (b) M. Kazemipour, E. Noroozian, M. S. Tehrani and
M. Mahmoudian, J. Pharm. Biomed. Anal., 2002, 30, 1379; (c) F.
Salinas, J. B. Nevado and A. E. Mansilla, Talanta, 1990, 37, 347;
(d ) C. B. Ojeda, F. S. Rojas and J. M. C. Pavon, Talanta, 1995,
42, 1195.
14 (a) F. R. P. Rocha, P. B. Martelli and B. F. Reis, Talanta, 2001,
55, 861; (b) I. E. Bechmann, L. Norgaard and C. Ridder, Anal.
Chim. Acta, 1995, 304, 229; (c) A. F. Danet and M. Cheregi,
Rev. Chim. (Bucharest), 1994, 45, 604.
15 Q. F. Hu, G. Y. Yang, Z. J. Huang and J. Y. Yin, Talanta, 2002,
58, 467.
16 G. Granero, C. Garnero and M. Longhi, J. Pharm. Biomed. Anal.,
2002, 29, 51.
17 A. S. Amin and Y. M. Issa, Spectrochim. Acta, Part A, 2003,
59, 6630.
References
1
(a) J. A. Howell and L. G. Hargis, Anal. Chem., 1978, 50, 243R;
(b) J. A. Howell and L. G. Hargis, Anal. Chem., 1982, 54, 171R;
(c) J. A. Howell and L. G. Hargis, Anal. Chem., 1986, 58, 108R;
(d ) J. A. Howell and L. G. Hargis, Anal. Chem., 1990, 62,
155R; (e) J. A. Howell and L. G. Hargis, Anal. Chem., 1994, 66,
445R; ( f ) L. G. Hargis and J. A. Howell, Anal. Chem., 1980,
52, 306R; (g) L. G. Hargis and J. A. Howell, Anal. Chem., 1984,
56, 225R; (h) L. G. Hargis and J. A. Howell, Anal. Chem., 1988,
60, 131R; (i) L. G. Hargis and J. A. Howell, Anal. Chem., 1992,
64, 66R; ( j) L. G. Hargis, J. A. Howell and R. E. Sutton, Anal.
Chem., 1996, 68, 169R; (k) J. A. Howell and R. E. Sutton, Anal.
Chem., 1998, 70, 107R.
2
3
W. J. Hurst, in Spectroscopy Techniques in Food Analysis,
ed. R. H. Wilson, VCH, New York, 1994, pp. 221–240.
O. Thomas, F. Theraulaz, C. Agnel and S. Suryani, Trib. Eau,
1995, 48, 47.
U. Platt, Eur. Comm., [Rep.] EUR, 1994, EUR 15609, 664–673.
E. Fritsch, Analusis, 1995, 23, M37.
A. Muramatsu, H. Itoh and T. Sugimoto, Tohoku Daigaku Sozai
Kogaku Kenkyusho Iho, 1993, 49, 101.
S. Matsuoka and K. J. Yoshimura, Flow Injection Anal., 1995,
12, 8.
18 (a) Z. J. Li, J. M. Pan and J. Tang, Anal. Chim. Acta, 2001, 445,
153; (b) H. M. Ma, U. Jarzak and W. Thiemann, New J. Chem.,
2001, 25, 872; (c) M. Su, Y. Liu, H. Ma, Q. Ma, Z. Wang, J. Yang
and M. Wang, Chem. Commun., 2001, 960; (d ) S. Inoue and M.
Uto, Bunseki, 1993, 11, 877.
19 (a) H. W. Gao, J. AOAC Int., 2001, 84, 532; (b) H. W. Gao and
N. L. Hu, J. Chin. Chem. Soc. (Taipei), 2003, 49, 965; (c) H. W.
Gao and J. F. Zhao, Croat. Chem. Acta, 2003, 76, 1.
20 (a) H. W. Gao and P. F. Zhang, Analyst, 1994, 119, 2109; (b)
H. W. Gao, Talanta, 2000, 52, 817; (c) H. W. Gao, J. X. Yang,
Z. Z. Zhou and J. F. Zhao, Phytochem. Anal., 2003, 14, 91.
21 (a) H. W. Gao and J. F. Zhao, Aust. J. Chem., 2002, 55, 767; (b)
H. W. Gao and H. D. Mei, Macromol. Biosci., 2002, 2, 280; (c)
H. W. Gao, J. Jiang and L. Q. Yu, Analyst, 2001, 126, 528; (d )
H. W. Gao, J. X. Yang and J. Jiang, Supramol. Chem., 2002,
14, 315.
22 H. W. Gao, Z. J. Hu and J. F. Zhao, Chem. Phys. Lett., 2003,
376, 251.
23 (a) F. Z. Tamura, Bunseki Kagaku, 1967, 16, 193; (b) J. X. Chen
and M. J. Xin, J. Pharm. Anal., 1991, 11, 225.
24 G. C. Pandey, Ajay Kumar and R. K. Garg, J. Appl. Polym. Sci.,
2001, 82, 2135.
25 J. B. Murphy and M. W. Kies, Biochem. Biophys. Acta, 1960,
45, 382.
4
5
6
7
8
9
H. Mach, D. B. Volkin, C. J. Burke and C. R. Middaugh,
Methods Mol. Biol. (Totowa, N.J.), 1995, 40, 91.
(a) L. Ayllon, M. Silva and D. Perez Bendito, J. Pharm. Sci., 1994,
83, 1135; (b) I. E. Bechmann, L. Norgaard and C. Ridder, Anal.
Chim. Acta, 1995, 304, 229; (c) L. Norgaard and C. Ridder,
Chemom. Intell. Lab. Syst., 1994, 23, 107; (d ) O. Lee, A. P. Wade
and G. A. Dumont, Anal. Chem., 1994, 66, 4507; (e) H. Liu and
P. K. Dasgupta, Anal. Chim. Acta, 1994, 289, 347; ( f ) R. Xiong,
A. Velasco, M. Silva and D. Perez-Bendito, Anal. Chim. Acta,
1991, 251, 313; (g) T. Umemura, Y. Kasuya, T. Odake and K.
Tsunoda, Analyst, 2002, 127, 149; (h) Y. Hirata and S. Katoh,
J. Microcolumn Sep., 1993, 4, 503; (i) V. K. Popov, J. A. Banister,
V. N. Bagratashvii, S. M. Howdle and M. J. Poliakoff, Supercrit.
Fluids, 1994, 7, 69.
1656
New J. Chem., 2003, 27, 1649–1656