810
Anal. Chem. 1984, 56,810-813
surface activity (ai') and the concentration gradient within the adhering layer become the same, independently from the initial a? values. Thus in the introductory period (0 < t < t? the slope of the transient signals should be proportional to AE = E(t? - E(a:) = log aim/a:, i.e., to log K , as after t > t'the transient signals run on the same path. At bp for an activity decrease, when the activity steps directed to more diluted solutions K -
c *
80
cn L
W
+
60
L
c
W
> c c
40
a
20.0 0.0
~
I
I
I
I
I
I
400. 420. 440. 460. 480. 500. 520. WAVELENGTH Cnml pH=3.63
2tU
LD
z
W t-
z
Y
W
> n
t-
a
-1
W
tz
2om 0 .o400. 420. 440. 460. 480. 500. 520.
WAVELENGTH Cnml pH=5.20
100.0-
> t(-3
Lr,
80.0-
z
W
t-
z
60.0-
(-3
w >
40.0
n
F
a
J
20.0
W
c!
, tF-
Lr,
80.0-
z
W I-
z
60.0-
I
Y
W
>
40.04
c
I-
a
J
20
.o
W CL
0 .o
400. 420. 440. 460. 480. 500. 520, WRVELENGTH C n m l Figure 1. Fluorescence excitation spectra for dissolved (0)and immobilized (A)calcein at pH 1.26, pH 3.63, pH 5.20, and pH 7.16.
812
ANALYTICAL CHEMISTRY, VOL. 56, NO. 4, APRIL 1984
.
^ ^
W
0
z
w
0
L
" W
cn
U
J 3
W
50.j
! l
W
> H
+
a _I
W
25.
i
!
LL
loo*k I 75.
50.
P
14:
I
J
0. 0.0
W
1.0
2.0
3.0
4.0
14:
5.0
-LOG P
- . 0
0.00 0 . 5 0 1 . 0 0 1 - 5 0 2 . 0 0 2 . 5 0 3 . 0 0
Flgure 2. Relative fluorescence intensity from immobilized calceln in the presence of Co(I1) at pH 7 as a function of the fraction of free metal. As IDA increases a larger fraction of the metal is complexed and decreases. The functional relation between /3 and IDA concentrations is given by eq 4.
are similar. The primary effect of immobilization is to broaden the spectra and cause a small shift to longer wavelength. In addition as the excitation wavelength approaches the emission wavelength, there is an increase in signal due to scattering. The observation that excitation spectra for dissolved and immobilized calcein are similar at all four pH values indicates that immobilization does not significantly affect the acid-base behavior of the immobilized ligand at neutral and acidic pHs. This is expected since cyanuric chloride will couple to one of calcein's aromatic OH groups (12)which would only ionize at higher pHs. Binding Constants. Equilibrium constants were determined from measurements of fluorescence intensity as a function of added iminodiacetic acid (IDA) in the presence of a small amount of quenching metal ion. As the IDA concentration increases, it will complex strongly enough to pull the metal ion away from the calcein. This is accompanied by an increase in fluorescence (Figure 2). The point at which fluorescence has increased to half its limiting value is the point where half of the metal ion has been removed from the calcein. At this point
ccU+2l
10-5 (bl
100.9 W 0
z
W 0
75.
cn W
cd
u 3 _1
50.
LL W
>
W cd
0.00 0 . 5 0
1 .OO
1 .50
C C ~ + ~ I
2.00
2 . 5 0 3.00
10-5
IC1
lO0.K
e = ML where t is the number of unassociated immobilized ligands and is the number of ligands associated with metal. For one-to-one complexes, the binding equilibrium may be represented as
14: 0
2 LL
so.
+ -=
W
>
Y
I
a
J
W
where Kf is the equilibrium constant for binding and [MI is the concentration of free metal in solution. Thus at the point where fluorescence intensity is half its limiting value
(3) where CMis the total concentration of metal in solution and p is the fraction of free metal. The value of p is readily calculated from the relation
where Kfl and Kfz are formation constants for metal-IDA complexes, aL-is the fraction of IDA in the deprotonated form, and CL is the total concentration of IDA. The value of aL-
cd
0. 0.00 0 . 5 0
1 .OO
CNi+21
1 .50 2 . 0 0
2.50
3.00
x
Flgure 3. Fluorescence intenslty vs. added metal ion for (a) Cu(II), (b) Co(II), and (c) Ni(I1) at pH 5.0 (0)and pH 7.0 (A).The remaining slgnal in the presence of excess metal Ion is due to scatter rather than to residual fluorescence from the complex.
is calculated from the acid ionization constants of IDA. Since all the necessary values are readily available (13),it is a simple matter to calculate values for and Kf. It should be noted that eq 4 is valid only if the amount of IDA in the form of metal complexes is small relative to the total amounts of IDA. In practice, this was the case. Table I lists binding constants determined for Co2+,Cu2+, and Ni2+ at pH 5.2 and 7.0. These values are conditional,
813
Anal. Chem. 1984, 56.813-814
Table I. Values for the Conditional Binding Constants for Immobilized and Dissolved Calcein (Expressed as log, K ) immobilized calcein dissolved calcein metal pH 5.15 pH 6.95 F 5 . 1 5 pH 6.95 Cu(I1) 9.4 12.4 9.2 12.3 5.5a 8.2a 8.0 Co(I1) 5.9 6.2 9.5 6.6 11.1 Ni(I1) a
pH 5.05 and 7.05, respectively.
depending on the p H of the measurement. The values are larger than the binding contants for Chelex-100 (14) where the binding group is an immobilized iminodiacetate group. The reason calcein binds more strongly than Chelex-100 is that the hydroxy group ortho to the iminodiacetic acid is also involved in complexation. Calcein is a tetradentate ligand while iminodiacetate is only a tridentate ligand. For comparison, Table I includes the values of binding constants for dissolved calcein determined in the same way. They are similar to those for immobilized calcein indicating that immobilization does not interfere with complexation. Although one of the hydroxy groups ortho to the iminodiacetate group must have reacted with cyanuric chloride in the immobilization step, the other is still available. Because the binding constants are so large, immobilized calcein would be an effective reagent for preconcentrating metal ions for subsequent analysis. It has the additional convenient property that one can tell if it is saturated with transition metal ions by observing whether or not it fluoresces. For use as a preconcentrating reagent, it would be desirable to increase the amount of immobilized calcein per gram of substrate. Use of Immobilized Calcein as a Sensor. We originally immobilized calcein for use as the reagent in an optical sensor that would respond to transition-metal ions. When immobilized calcein is placed on the end of a bifurcated fiber optic and placed in solution, fluorescence decreases as a function of added metal ion (Figure 3). However, because the binding is so strong, the response is not reversible. The reagent is completely extracting metal ion from solution a t these pHs. To use immobilized calcein in a reversible sensor, it will be necessary to work a t lower pHs where the conditional formation constant is smaller or to use calcein in a complexing
medium where it will tend to extract a smaller percentage of metal ion. Another approach to using immobilized calcein as a sensor is to form a nonfluorescent complex and then to add a nonquenching metal ion that displaces the metal in from the nonfluorescent complex. For example, when Zn(I1) is added to a sensor in which immobilized calcein has been reacted to form the Co(I1) complex, an increase in fluorescence is observed as Zn(I1) displaces Co(I1). Immobilized Calcein as a Reusable End Point Indicator. The sensor based on calcein can be used to determine end points of complexometric titrations. For example, Cu2+ was titrated with EDTA at pH 7.0. At the equivalence point in this titration a large increase in calcein fluorescence is observed, since a very slight excess of EDTA is sufficient to pull the Cu2+away from the immobilized indicator. Registry No. Calcein, 1461-15-0;cyanuric chloride, 108-77-0; cellulose, 9004-34-6; nickel, 7440-02-0; copper, 7440-50-8; cobalt, 7440-48-4.
LITERATURE CITED (1) Saari, L. A.; Seitz, W. R. Anal. Chem. 1983, 55, 667-670. (2) Dltzler, M. A.; Doherty, G.; Sieber, S.; Allston, R. Anal. Chim. Acta 1982, 142, 305-311. (3) Diehl, H.; Ellingboe, J. L. Anal. Chem. 1983, 35, 882-884. (4) Wallach, D. F. H.; Surgenor, D. M.; Soderberg, J.; Delano, E. Anal. Chem. 1959, 3 7 , 456-460. (5) Kepner. B. L.; Hercules, D. M. Anal. Chem. 1963. 35, 1238-1240. (6) Hill, J. B. Clin. Chem. (Winston-Salem, N . C . ) 1965, 1 1 , 122-130. (7) Moser, G. B.; Gerarde, H. W. Clin. Chem. (Winston-Salem, N . C . ) 1969, 15, 376-380. (8) Bandrowski, J. F.; Benson, C. L. Clin. Chem. (Winston-Salem, N . C . ) 1972, 78, 1411-1414. (9) Hefley, A. J.; Jaselkls, B. Anal. Chem. 1974, 4 6 , 2036-2038. (10) Wallach, D. F. H.; Steck, T. L. Anal. Chem. 1963, 35, 1035-1044. (11) Saari, L. A.; Seitz, W. R. Anal. Chem. 1982, 5 4 , 821-823. (12) Kay, G.; Crook, E. M. Nature (London) 1967, 216, 514-515. (13) Smith, R. M.; Martell, A. E. “Critical Stability Constants”: Plenum: New York, 1974; Vol. I . (14) Eger, c.; Anspach, W. M.; Marinsky, J. Inorg. N U C ~ Chmn. . 1968, 30, 1899-1909.
L. A. Saari W. R. Seitz* Department of Chemistry University of New Hampshire Durham, New Hampshire 03824
RECEIVED for review October 31,1983. Accepted January 12, 1984. Partial support for this research was provided by NSF Grant CHE82-06131.
Reaction of Gibbs Reagent with Para-Substituted Phenols Sir: Gibbs reagent (2,6-dichloro-p-benzoquinone4chloroimine) is used as a reagent for the detection of phenol derivatives (1). The reagent adds to the para position of the phenol ring to give an indophenol. Indophenols are brightly colored and undergo a dramatic color change with p H due to the ionization of the phenolic proton. Remarkably, many phenol derivatives bearing substituents at the para position react readily with Gibbs reagent. Dacre has summarized the literature on this “nonspecific” Gibbs reaction (2). A variety of phenol derivatives bearing alkoxy, aldehyde, halogen, or other groups in the para position were found to give colored products with Gibbs reagent. For each product, A- was listed and ,,e was calculated, based on the assumption of quantitative conversion of the phenol derivative to an indophenol. The structures of the resulting products were not assigned 0003-2700/84/0356-0813$01.50/0
by Dacre (2). Here, we show that the reaction of para-substituted phenols with Gibbs reagent occurs via loss of the para substituent. The resulting colored product is thus identical with that formed by phenol itself. The wide variations in emu noted by Dacre (2)represent varying yields of product rather than differences in molar absorptivity.
EXPERIMENTAL SECTION Gibbs reagent and 2,6-dichloroindophenolwere obtained from Fisher Scientific and phenol derivatives from Aldrich Chemical Co. or other commercial sources and were used without further purification. Reagents were prepared as stock solutions (5 mM) in ethanol. Aliquots (0.25 mL) of Gibbs reagent and phenol derivative were mixed in borate buffer, pH 9.2,24.5 mL, and the optical spectrum of the blue indophenolate anion was recorded after 2 h. An authentic sample of 2,6-dichloroindophenol was 0 1984 American Chemical Society