Anal. Chem. 1983, 55, 667-670 (15) Osseo-Azare, K.; Keeney, M. E. ”International Solvent Extraction Conference, [Proceedings], Liege”; 1980; Vol. I , Paper 80-121. (16) Osseo-Azare, K.; Keeney, M. E. Sep. Scl. Techno/. 1980, 15 (4), 999. (17) Robinson, B. H.; James, A. D. J. Chem. Sm., Faraday Trans 1 1973, 74, 10. (18) Roblnson, B. H.; Whlte, N. C. J. Chem. Soc., Faraday Trans 11978, 74, 2625. (19) Relnsborough, V. C.; Roblnson, 8. H. J. Chem. Soc., Faf8daY Trans. 11979, 75, 2395. (20) Roblnson, B. H.; Steytler, D. C.; Tack, R. D. J. Chem. Soc., Faraday Trans. 11979, 75, 481. (21) Chin-Kwang, W.; Hung-Chang, K.; Tien, C.; Seng-Chung, L.; Tlen-Chu, K.; Kwang-Hsien, H. “Internatlonal Solvent Extractlon Conference, [Proceedings], Liege”; 1980; Vol. 1, Paper 80-23. (22) Komornicki, J. Thesis, Paris, France, 1981. (23) Bauer, D.; Komornlcki, J.; Telller, J. French Patent Appl. No. 81.24.287. (24) Bauer, D.; Komornlckl, J.; Telller, J. French Patent Appl. No. 82.03.231.
667
(25) Helgorsky. J.; Leveque, A. “CIM Speclal Volume”; Canadian Institute of Mlnlng and Metallurgy: Montreal, 1979; No. 21, Vol. 2, p 439. (28) Leveque, A.; Helgorsky, J. European Patent No. 29.70 Rh6ne-Poulenc co. (27) Fourre, P.; Bauer, D. C . R . Mbd. Seances Acad. Scl. I I 1981, 292, 1077. (28) Bauer, D.; Fourre, P.; Lemerle, J. C . R. Hebd. Seances Acad. Scl. I I 1981, 292, 1019. (29) Fourre, P.; Bauer, D., unpublished work, Paris, France. (30) Buddle, W. M., Jr.; Hartlage, J. A. U.S. Patent 3637711, 1972. (31) Bowcott, J. E.; Schulman, J. H. 2.€lektrochem. 1955, 59, 283. (32) Dvolaitzky, M.; Guyot, M.; Lagues, M.; Le Pesant, J. P.; Ober, R.; Sauterey, C.; Taupin, C. J. Chem. Phys. 1978, 69 (7), 3279. (33) Hwan, R. N.; Mlller, C. A.; Fort, T., Jr. J. Colloid Interface Sci. 1979, 68 (2), 221. (34) Jorpes-Friman, M. Flnn. Chem. Left. 1979, 7-8, 240.
RECEIVED for review May 24,1982. Accepted December 14, 1982.
Immobilized Morin as Fluorescence Sensor for Determination of Aluminum( I I I) Linda A. Saari and W. Rudolf Seltz” Department of Chemlstty, Unlversl@ of New Hampshlre, Durham, New Hampshlre 03824
We have prepared a sensor for AI3+ based on fluorescence. Morln Is lmmoblllred on cellulose powder and attached to the end of a bifurcated fiber optic. When the lmmoblllred morln Is placed In a solution containing Ai3+, fluorescence Is observed from the A13+-morln complex. Response Is linear from 1 X I O d to 1 X IO-“ M Ala+ at pH 4.8. At this pH the Mndlng constant for the A13+-hnmoblllred morln complex Is 1.7 X IO‘. Response time is 1-2 mln. The detection limit Is 1 X I O d M. The detection llmlt Is established by variations In background signal which Is comprised of detector dark current and background fluorescence. Under conditions of this study, the sensor extracts about 1% of the aluminum In the sample when operating In the range of linear response.
The development of sensors based on immobilized fluorogenic reagents is a subject of growing interest. We have previously reported a pH sensor based on immobilized fluoresceinamine (I). Other research in this area has been described in the Focus section of ANALYTICAL CHEMISTRY (2). These sensors offer two important advantages relative to conventional solution fluorescence measurements: they can measure concentrations without significantly perturbing the sample, and they can be used for continuous sensing. In this report we describe the characteristics of a sensor for AP+ based on immobilized morin (3,5,7,2’,4’-pentahydroxyflavone). Morin is only weakly fluorescent by itself but forms highly fluorescent complexes with A P . It has been widely used as a reagent for fluorometric analysis of aluminum and other metals as well as for spectrophotometric analysis of metals that do not form fluorescent complexes (3-5).
THEORY This section develops theory to describe the relative response of the immobilized morin sensor as a function of aluminum ion concentration. It is assumed that the total number of immobilized morin molecules, C, is much less than the number of aluminum ions in solution. Under these conditions 0003-2700/83/0355-0667$01.50/0
the insertion of the sensor will not significantly affect the aluminum ion concentration in solution. The equilibrium constant for aluminum binding to immobilized morin can be represented (assuming a 1:l complex)
K = MM/MaM
(1)
where M is the number of immobilized morin ligands not associated with aluminum, M M is the number of immobilized morin ligands associated with aluminum ion, aM is the aluminum metal ion activity in solution, and K is the equilibrium constant. Because the total number of immobilized morin molecules, C, is fixed:
C=M+MU (2) Since morin is essentially nonfluorescent by itself and the morin-aluminum complex is fluorescent, the fluorescence signal will depend on the amount of aluminum bound to the morin
I = kMM
(3)
where I is the fluorescence intensity and k is a proportionality constant relating fluorescence intensity to the amount of aluminum bound to morin. It is assumed that the conditions are such that intensity is proportional to the number of sites (i.e., no inner filter effects). By substituting eq 2 into eq 1 and rearranging, it is possible to express MM in terms of C, a ~and , K
MA, = u M K C / ( l + aMK) (4) By substituting eq 4 into eq 3, an expression for fluorescence intensity as a function of aluminum ion activity is obtained I =I kKCUAl/(1 UAlK) (5)
+
To determine K for the sensor a linear form of eq 5 can be used
a M / I = uM/kC+ 1/kCK
(6)
A plot of U M / Ivs. a h will be linear with a slope of l / k C and an intercept of l / k C K . 0 1983 American Chemlcal Soclety
668
ANALYTICAL CHEMISTRY, VOL. 55, NO. 4, APRIL I983
The experimental equilibrium constant, K , is a conditional constant that depends on pH because M3+displaces hydrogen ion when binding to morin as shown below.
OH OH
0
/
OH
+ H+ 00...*,+3
Table I. Experimental Values of the Conditional Equilibrium Constant, K , for Aluminum Binding to Immobilized Morin PH
K
PH
K
3.8 4.1 4.6
4 x 103 1.2 x 104
5.1 5.5 5.9 6.5
3.3 x 104 2.8 x 104 3 x 104 1.8 X lo4
4.8
1.4 x 104 1.7 x 104
In addition, hydrolysis of A13+ will affect K. EXPERIMENTAL SECTION Apparatus. The filter fluorometer used to measure the fluorescence of the immobilized A13+-morin complex has been described previously ( I ) . The excitation wavelength was selected with a 2 in. X 2 in. interference filter from Oriel with maximum transmittance at 420 nm and a 10-nm bandwidth at half maximum transmittance. The emission wavelength was selected by a 1in. X 1/2 in. interference fiiter from Edmund Scientificwith maximum transmittance at 488 nm and a bandwidth of 7 nm at half maximum transmittance. The bifurcated fiber optic from Fiber Optics Technology is glass with a diameter of 3/ls in. Excitation spectra of the immobilized morin and its aluminum complex on the optic surface were measured with an SLM 8000 spectrofluorometer equipped with an MC 300 manual monochromator. Measurements were made by attaching the excitation and emission arms of the bifurcated fiber optic to the source and detector lens housings in the sample chamber by means of light-tight aluminum fittings. pH was measured with an Orion Digital Ionalyzer/SOl. Fluorescence excitation and emission solution spectra were measured on a Perkin-Elmer 204 spectrofluorometer. UV-VIS absorbance measurements were made on a Bausch and Lomb Spectronic 200 UV spectrometer. Reagents. Morin and cyanuric chloride were purchased from Aldrich. Powdered cellulose (microcrystalline for TLC, through 60 sieve, Baker) was used as the cellulose support. Immobilization Procedure. Morin was immobilized on powdered cellulose, using cyanuric chloride according to a previous procedure ( I ) . The concentration of the cyanuric chloride solution was 0.14 M in acetone and the concentration of morin was 0.0076 M in acetone. The product was washed with acetone, dried by suction, and stored at room temperature. A single batch of immobilized morin was used for all characterization studies. Morin was also immobilized as its A13+complex, because this should prevent the functional groups involved in complexation from reacting with cyanuric chloride. The soaking solution consisted of acetone-water (3:2) 0.0021 M in morin and 0.014 M in A13+. For comparison a batch of uncomplexed morin was immobilized under the same conditions, but omitting the aluminum. The morin bound uncomplexed gave slightly higher intensity then the morin bound complexed when both were saturated with aluminum, and consequently, morin bound in the uncomplexed form was used for all experiments. The amount of morin bound per gram of cellulose was found to be slightly greater when morin was bound as the complex. Thus, it is not clear why this preparation failed to respond more sensitively to A13+. This is an interesting question but was not pursued. Determination of Amount Bound. The amount of morin bound/gram of cellulose was found by stripping the morin from the cellulose in 1M KOH and measuring the morin concentration by spectrophotometry. Weighed amounts (0.0311 g and 0.0356 g) of the immobilized morin were soaked in 1M KOH solutions (10.00 mL) until the cellulose lost color (4 h). The absorbance of 1 : l O dilutions of the soaking solutions was measured at 405 nm along with morin standards in 1M KOH. Absorbance spectra were measured for the standards and soaking solutions from 350 to 600 nm. The amount of active aluminum binding sites contained/gram of cellulose was determined by two methods. The first method consisted of soaking weighed amounts (0.0077 g and 0.0115 g) of the immobilized morin in solutions of aluminum sulfate (5.45 X lod M, 0.400 mL) at pH 5.1 for 15 min. The decrease in aluminum concentration of the soaking solutions was then determined by taking 0.100 mL of each soaking solution, adding morin (0.00683
M, 0.175 mL), diluting to 25 mL with pH 5.1 acetate buffer, and measuring the fluorescence intensity using the SLM 8000 spectrofluorometer. The excitation wavelength was set at 420 nm and the emission wavelength was selected by using the 488 nm interference filter. Standards of varying aluminum concentration, of the same pH, acetate concentration, and morin concentration were used to determine aluminum concentration. Since this method involves the morin-A13+ system at equilibrium, the equilibrium constant for binding at pH 5.1 (Table I) was used in calculating the number of active sites from the decrease in aluminum concentration. In the second method used to determine the amount of active aluminum binding sites contained/gram of cellulose, weighed amounts (0.0215 g and 0.0197 g) of the immobilized morin were M) each soaked in a solution of aluminum sulfate (7.56 X at pH 4.8 for 0.5 h. This concentration of aluminum is sufficient to saturate all binding sites. The excess solution was washed away with four 10-mL portions of distilled, deionized water on a sintered glass funnel. In deionized water the kinetics of complex dissociation are so slow that the washing step may be completed without significant dissociation of the complex. The cellulose was then soaked in 10 mL of 20% acetic acid for 1h. The cellulose was fluorescent under UV light after the aluminum soaking and washing, but nonfluorescent after the acetic acid treatment indicating that all the aluminum was stripped off. The solution was poured off and the cellulose was rinsed with 5 mL of distilled water, making 15 mL of solution. The pH was adjusted to 4.7 with KOH and the solutions were analyzed for aluminum by adding morin (0.1 mL of 0.00777 M) and measuring the fluorescenceintensity with the fiber optic fluorometer. Standards of varying aluminum concentration, of the same pH, acetate concentration, and morin concentration were used to determine the aluminum concentration. Fluorescence Measurements. To determine the wavelengths for excitation and emission, fluorescence spectra were measured on the morin-A13+ complex in solution by use of the Perkin-Elmer 204 spectrofluorometer. Fluorescence measurements on the immobilized morin (i.e., the sensor) were made by attaching the immobilized morin to the end of the fiber optic and immersing the optic in acetate buffer solution. A thin layer of the cellulose support was spread on a piece of cellophane tape (Scotch brand) and the tape was held in place on the end of the optic by a piece of Tygon tubing. This has been described previously ( I ) . The tape is not entirely satisfactory because it becomes cloudy when wet with reduced transmittance and significant scattering. Because of the limitations of the tape, collodion was also tried. A small amount of the collodion was spread on the end of the optic and allowed to become sticky and then the powder was applied as a thin layer. This was not found to be satisfactory because it was difficult to form a thin, uniform layer. The collodion also loses its adhesive properties with time. Aluminum concentration was varied both by adding known amounts of aqueous standard solutions of aluminum sulfate to the same solution and by immersing the sensor in buffered solutions of varying aluminum concentrations. The solution was stirred as measurements were taken. The effect of pH on the response to aluminum was studied by using acetate buffers of varying pH in the range 2.9 to 6.5. Interferences were studied by adding various ions separately to the immobilized morin sensor when it was responding to aluminum. Metal ions tested were iron(III), cobalt, copper(II), beryllium, calcium, and magnesium. The reversibility of the sensor response was determined by adding aluminum ion to the sensor in buffer solution until the response was constant and then immersing the sensor into a buffer
ANALYTICAL CHEMISTRY, VOL. 55, NO. 4, APRIL 1983
I
669
0
f W V
z W
V
v)
'!
A O(B0
1.60
2.140
3:20
4.b0
AI+^ ] 1 0 - 4
Flgure 1. Relative fluorescence intensity of the immoblllzed morin sensor as a function of AI3+ concentration at pH 3.8 (A),pH 4.8 (O), and pH 5.9 (0).
solution without aluminum. The sensor was then submerged into a fresh buffer solution and more aluminum was added. The aluminum was also stripped by immersing the sensor in a solution of EDTA (0.0255 M at pH 7) and washing with water. Excitation spectra of the immobilized morin and its aluminum complex were measured at a fixed emission wavelength of 488 nm at a pH of 4.8 using the SLM 8000 spectrofluorometer. Measurements were made on morin in solution for comparison to the immobilized morin. The fluorescence of the morin-A13+ complex, the effect of pH, and the effect of interferences on the morin-A13+ complex were studied. The excitation spectra of morin and its aluminum complex in solution were measured with the SLM 8000 spectrofluorometer. The concentration of morin used M. in the solution study was 8.00 X
RESULTS AND DISCUSSION Response to Aluminum. The fluorescence of the sensor based on immobilized morin increases with aluminum concentration as shown in Figure 1. This is expected since aluminum forms a fluorescent complex with morin. The response is linear from lo4 M to M aluminum concentration. Above 10"' M there is a change in slope and the response levels off. This leveling off is due to saturation of the morin with aluminum. The same type of response is observed for morin in solution. The fluorescence intensities and aluminum concentrations were plotted in the form of the theoretical linear equation modeling the system (see eq 6). The linear model is adequate for the system at the 95% confidence level as shown by the appropriate regression analysis. From the slope and intercept, values for the conditional binding equilibrium constant, K , were determined by the immobilized morin in solutions of varying pH. The results are shown in Table I. These values reflect changes in protonation of the immobilized morin and formation of hydroxide complexes of aluminum. At low pH the formation constant decreases because protons tend to displace A13+. At high pH the formation constant decreases because more of the aluminum is in the form of hydroxide complexes. The binding constant for aluminum ion to morin in solution has been reported as 2.96 X lo6,but the pH is not specified (3, 5). By use of the slope and intercept of the linear plot a t pH 4.8, the theoretical equation to describe the response of the sensor (eq (5) is plotted in Figure 2. Since the data are represented adequately by the theory for a 1:l morin-aluminum complex, this suggests that the complex is primarily in this form.
I 0
I
I
I
I
I
5
IO
15
20
25
AI+^] 1 0 - 4 Flgure 2. A plot of the theoretical response of the sensor (curve) represented by eq 5 and experimental points (0)at pH 4.8.
Determination of Amount Bound. The amount of morin bound/gram of cellulose is 4.2 X mol as determined by measuring the amount of morin stripped off the cellulose in strong base by spectrophotometry. The absorbance spectra are identical for the morin in base and the solutions containing the stripped morin. This indicates that treatment with base yields free morin rather than a morin-cyanuric chloride conjugate. The number of moles of active complexing sites/gram is 26% of the number of moles of morin bound/gram (1.1 X lo4 mol of A13+/gramof immobilized morin as determined by the two methods which are in agreement). That the number of active sites is less than the total shows that the group responsible for complexation is probably tied up in the immobilization procedure. It is not known which of the morin phenolic sites the cyanuric chloride binds to, but the product is most likely a mixture with the cyanuric chloride bound to different sites. Operating Characteristics. The optimum pH for the sensor is 4.8. The response to aluminum at various pH values is shown in Figure 1. At low pH the kinetics of the response are very slow and the intensity is low. At higher pH values the fluorescence intensity is low due in part to the hydrolysis of aluminum ion. Fluorescence intensity vs. time data at different pH values are shown in Figure 3. The response time for the sensor is 1-2 min to a change in aluminum concentration. When the sensor is saturated and placed in a buffer solution without aluminum, the fall time is about 3 min. The response time when immobilized reagent is inserted into solution for the first time is 3-4 min. Measurements were made with 1 mg of derivatized support on the optic surface. This corresponds to 4.2 X mol of morin on the optic surface as determined by absorbance measurements. With the following relationship which is valid at any aluminum level: no. of AI ions bound/total no. of sites = intensity /intensity at saturation it is possible to show that only 1% of the aluminum ion is extracted from solution when operating in the region of linear response to aluminum. This supports the assumption in the Theory section that insertion of the sensor does not significantly affect concentration. The excitation spectra of the immobilized morin (i.e., background), its aluminum complex, and the morin-AP+
670
ANALYTICAL CHEMISTRY, VOL. 55, NO. 4, APRIL 1983
y
Table 11. Effect of Interferences on the Al" Response of the Immobilized Morin Sensor
100.-
w
IQ
1 W
5
interference
75.-
>.
Ca2+ Mg2+ BeZ+
IH
cn
+
50.
Co2+
L
H
W 0
L
w
0
v, W L1!
Cua+ 25.
AA
Fe3+
4
2 -.0 1 0.0 LL
a
1.0
0.5
1.5
2.0
2.5
TIME
CMINUTESI Flgure 3. Response to added AI3' as a function of time for the lmmobillzed morln sensor at pH 3.5 (A), pH 4.8 (O), and pH 6.5 (*). M. The AI3+ concentration was 7.2 X lOOr
v -
340
~~
360
380
400
420
440
460
480
WAVELENGTH ( n m )
Flgure 4. Excitation spectra (uncorrected) of the morin-Ai3' complex in solution (A), the Immobilized morln-AI3' complex (0),and the Imat pH 4.8. The spectra for mobilized morln without aluminum (0) Immobilized morin wlth and without AI3+ are drawn to the same scale. The spectrum for the complex in solution is normalized to have the same maximum value as the spectrum for the immoblllzed complex. The Ai3' concentration was 1.0 X IO-' M for the spectra of the complexes.
complex in solution are shown in Figure 4. The spectra of the immobilized complex and the complex in solution are similar with the spectrum of immobilized complex showing a slight broadening. The excitation spectrum of the immobilized morin shows a small peak similar in shape to the spectrum of the complex superimposed on a flat continuous background. This indicates that the background includes both a contribution from weak morin fluorescence plus a signal due to reflected and scattered excitation radiation. The signal to background ratio is 2 5 1 at aluminum conM. Dark current represents 50% of centrations of 1 X
concn, M
signal
3.32 X lo-' 2.83 x 3.19 X 4.73 X 3.75 X 2.93 X 7.64 X 7.59 X 3.27 x lo-' 4.87 X
0.088
[A13+]= 3.59 X
0.108 0.084 0.084 0.097 0.097 0.134 0.134 0.162 0.162 M.
signal with interference 0.090 0.118 0.103 0.153 0.097 0.080
0.108 0.039 0.138 0.020
[A13+]= 7.18 X
%
change +2' +ga +23a +82' +Oa
-18' -19& -71b -15& -88&
M.
the background with the remainder due to scattered and reflected light plus residual morin fluorescence. The signal to background ratio can be improved to 7:l at 1 X M aluminum, using the SLM spectrofluorometer as source and detector with dark current representing 25% of the background. The detection limit for the sensor, as the concentration equivalent to 2 times the standard deviation of the background signal, is 1 X lo4 M AP+ (0.027 ppm). Because the background is quite stable, we are able to measure small signals superimposed on a relatively large background. Interferences. Because beryllium, calcium, and magnesium form fluorescent complexes with morin (3-5)they were tested as interferences. The optimum pH for formation of these complexes, however, is above 7 (3, 4). Iron(III), copper(II), and cobalt are all heavy metal ions and could quench fluorescence by complexation or by the heavy atom effect in solution. The effect of these ions in the aluminum response is shown in Table 11. The major interferences are beryllium which increases the signal and iron(II1) and copper(I1) which quench fluorescence to an appreciable extent. Copper(I1) causes the signal due to aluminum to decrease, but the sensor still responds to further increases in aluminum concentration. The quenching by copper(I1) may be due to ions in solution or to weak complexation, rather than due to strong complex formation as in the case of iron(II1). After being quenched by the addition of iron, the sensor no longer responds to added aluminum. To use the sensor in the presence of iron, it would be necessary to reduce iron to the ferrous form and tie it up as a complex. Morin in solution behaves in a similar manner when subjected to the ions tested. Registry No. Aluminum,7429-90-5;morin, 480-16-0; cellulose, 9004-34-6.
LITERATURE CITED (1) Saarl, L. A.; Seltz, W. R. AnalChern. 1982, 54, 821-823. (2) Borman, S. A. AndChern. 1981, 53, 1616A-1618A. (3) Katyal, M.;Prakash, S. Talanfa 1077, 24,367-375. (4) Katyal, M. T8/8f?t81968, 15, 95-106. (5) Nevskaya, E. M.; Nazarenko, V. A. J . A n d . Chem. USSR (Engl. Trensl.) 1972, 27, 1544-1561.
RECEIVED for review September 13,1982. Accepted December 20, 1982.
