Anal. Chem. 1984, 56, 755-757
755
Kinetics of the Complexation of Iron(I I) with Ferrozine James C. Thompsen and Horacio A. Mottola*
Department of Chemistry, Oklahoma State University, Stillwater, Oklahoma 74078
The klnetlcs of Iron( I I ) complexatlon by ferrorlne (benzenesulfonlc acid, 4,4'-[3-( 2-pyrldlnyl)-1,2,4-trlazlne-5,6-dlyl]bls-, monosodium salt dihydrate) has been studled by applylng stopped-flow mlxing and photometrlc monltorlng at 562 nm. The study was conducted at pH 2.80-5.50, an lonlc strength of 0.10 M, and temperatures from 18 to 40 "C. The reaction was found to be flrst order wlth respect to Fe(I1) and thlrd order with respect to the ligand. The pH dependence of the reactlon Is Interpreted to mean that the unprotonated form of the ligand Is the reactlve specles. The pK, of ferrozlne was found to be 3.13 f 0.02 by klnetlc lnterpretatlon of experlmental data and 3.27 f 0.05 by dlrect potentlometrictitratlon. The complex formatlon and dlssoclatlon rate coefflclents were (3.08 f 0.03) X 10'' W3 8- (25 "C) and (4.25 f 0.04) X I O 4 s-' (25 "C), respectively. The actlvatlon energy for the formation of the complex Is 9.87 f 0.17 kJ/mol. A mechanlsm Is proposed to satlsfy the experlmental observatlons and rate equatlon. Also Included Is a comparlson wlth Fe( I I ) complexation by two other common reagents for Iron, 1,lOphenanthrollne, and 2,2'-blyprldlne.
'
Ferrozine is frequently used for the determination of iron in aqueous solutions because its water solubility and molar absorptivity are higher than those of other iron reagents. The structure of ferrozine is shown in Figure 1. Equilibrium studies and some properties of the iron(I1)-ferrozine complex have been reported since the introduction of the ligand as a reagent for iron (1,2). These studies have shown that the complex exists in a 3:l ratio of ferrozine to iron (1) and that it is stable in the p H range of 3 to 6 (2). A value of 27900 M-l cm-l, at ,A, = 562 nm, has been reported for the molar absorptivity of the complex ( I ) , which compares favorably in terms of sensitivity with other complexes commonly used for iron determination [e.g., 1,lO-phenanthroline (11100 M-l cm-l at A,, = 510 nm ( 1 ) )and 2,2'-bipyridine (8650 M-l cm-l at A, = 522 nm (3))]. Ferrozine, as a reagent for iron, has been used in a variety of samples including blood serum (4-6), lake and tap water (3,and plant nutrient solutions (8). Both batch and continuous-flow sample processing have been used in these applications. No report has been found describing the kinetics of this complexation reaction. Besides the basic interest in this, rate information is of relevance in situations of continuous-flow sample processing such as flow injection analysis (9-11) and is an aid in the development of reaction rate determinations. With this relevance in mind, kinetic information was obtained with the help of a stopped-flow mixing device for the Fe(11)-ferrozine complexation reaction and is reported here. From such information a mechanism compatible with the experimental data is proposed and the results are compared with those of 1,lO-phenanthroline and 2,2'-bipyridine.
EXPERIMENTAL SECTION Apparatus. Monitoring of all reactions (except those performed in hydrochloric acid media for the calculation of the dissociation rate coefficient) was performed with the aid of a stopped-flow unit consisting of a GCA/McPherson stopped-flow mixing module EU-730-11 (Acton, MA) with a dead time of ap-
proximately 5 ms. An Osram 64460 (24 V, 100 W) tungsten lamp (Osram,Berlin, Germany) operated by a 24-V, 5-A power supply (Adtech Power, Anaheim, CA), an H-10 Jobin-Yvon concave holographic grating monochromator (Instrument SA, Inc., Metuchen, NJ), and an EG&G (Salem, MA) UV-040 BG photodiode detector, which delivers a voltage proportional to absorbance, were used to monitor the reaction in the stopped-flow observation cell. The stopped-flow module was thermostated with water from a Lauda/Brinkman (Westbury, NY) K-2/R water circulator that maintained the temperature within f0.05 OC. The signal from the detector was fed to a data acquisition system consisting of a 12-bit analog-to-digitalconverter, a sample/hold amplifier, and a 16-bit timer. The data acquisition system has been described previously (12). It was controlled by an Apple 11+ (Cupertino, CA) microcomputer (48K) equipped with a monitor, disk drive, and dot matrix printer. The reactions performed in hydrochloric acid solutions for the determination of the dissociation rate coefficient were monitored with a Beckman 25 UV/VIS spectrophotometer equipped with a Beckman recorder/controller unit as readout. An Orion Research (Cambridge, MA) Model 601A digital pH meter was used for potentiometric titrations and for adjusting the pH of buffer solutions. Reagents and Solutions. All chemicals used were of AR grade. The water used for solution preparation was deionized water that was further purified by distillation in an all-borosilicate glass still with a quartz immersion heater. For the pH study a phthalate buffer system was used in which 0.010 M potassium acid phthalate was mixed with appropriate volumes of either 0.010 M NaOH or 0,010 M HCIOl to obtain the desired pH. All solutions were adjusted to a final ionic strength of 0.10 M with NaC104. Iron(I1) solutions were prepared daily from FeS04.7H20 since about 20% of the iron in such a solution is oxidized to Fe(II1) in a 24-h period. Procedure. The stopped-flow spectrophotometric unit was standardized daily by setting the zero absorbance with water in the cell and the maximum absorbance with a standard solution of iron(I1)-ferrozine complex of the same concentration as that expected at the completion of the reaction. All solutions were thermostated at least 15 min before use. At least 15 250-~L injections were made before starting the collection of data to ensure a reliable run (after about 10 injections all reaction profiles, for a given set of conditions, exhibited excellent replication). For the study of the dissociation reaction, 1.50 mL of 9.00 x lo4 M complex solution was placed in a 1-cm cuvette in the Beckman 25 spectrophotometer and the reaction was initiated by injecting 1.5 mL of 0.20 M HC1. All absorptiometric measurements were made at a wavelength of 562 nm. Potentiometric titrations of ferrozine solutions were performed by placing a 0.010 M ferrozine solution (0.10 M in NaC104) into a thermostated titration vessel (25 "C) and titrating with a 0.115 M NaOH solution standardized against potassium acid phthalate.
RESULTS AND DISCUSSION Order with Respect to Iron(I1). The order of the reaction with respect t~ iron(I1) was studied in reactions in which the ferrozine concentration was about 100 times the iron concentration. The order with respect to iron was found to be one. A nonlinear least-squares method proposed by Zuberbuhler and Kaden (13) was used to compute the rate coefficients and their standard deviations from first-order curves. This method does not require either an initial or final concentration to be known. This is an advantage when analyzing stopped-flow data, where the first value is collected
0003-2700/84/0356-0755$01.50/0 0 1984 American Chemical Society
756
ANALYTICAL CHEMISTRY, VOL. 56, NO. 4, APRIL 1984 ,503H
Y Flgure 1.
Structure of ferrozine.
Table I. Experimental Order with Respect to Ferrozine (Temperature, 25.0 "C; Ionic Strength, 0.10 M ) exptl conditions
ordera
phthalate buffer, pH 5.00
2.98 2.94 2.84 phthalate buffer, pH 2.80 2.80 0.185 M H,SO, The overall average order ( 9 replicas) was found to be 2.95 I0.14 a
3
5
4
PtFlgure 2. Effect of H on the observed forward rate coefficient kobd ([Fe] = 8.00 X 10- M; [Fz] = 8.00 X M): ( 0 )data points; (-) fit according to eq 11.
B
See text for calculation
a few milliseconds after the reaction has started and there is no true time zero value. Order with Respect to Ferrozine. The order with resped to ferrozine was determined, again with the ferrozine concentration about 100 times that of iron (pseudo-first-order rate). First-order rate coefficients obtained under these conditions were used to compute the experimental order with respect to ferrozine, according to
where L is the ferrozine ligand, K1 and K 2 are equilibrium constants, and k3 and k-, are rate coefficients. Steps 2 and 3 are preequilibrium ones and step 4 is the rate-determining step. These three steps lead to the following rate expression in accordance with the experimental orders reported above:
d [FeL3]/dt = k3K1K2[Fe(I1)1[LI3 d[FeL3]/dt = k3[FeL2][L] - k-,[FeL3]
with kobd = k[ferrozineIoY. A log kobd vs. log [ferro~ine]~ plot allows one to obtain from the slope the value of y, the order with respect to the ligand. Values of y under different conditions are tabulated in Table I and indicate that the experimental order with respect to ferrozine is 3. Rate Expression and Proposed Mechanism. Because of the statistical improbability of a three- or four-body collision, the proposed mechanism (in accordance with what has been reported for similar iron ligands (14,15)) is composed of the following steps: K
+ L & FeL K FeL + L 2FeL,
Fe(I1)
(2)
(3)
k
FeL,
+L& FeL, k-3
(4)
_--_____
Both FeLz and FeL are present under steady-state conditions so that their concentrations may be replaced by their equilibrium constant expressions and be substituted into eq 7 to obtain eq 5. Rate Dependence on pH. The effect of pH on the reaction rate was studied in the pH range 2.80 to 5.50 fixed by phthalate buffers. The concentrations of iron(I1) and ferrozine were 8.00 x lo4 M and 8.00 X M, respectively, and the ionic strength was adjusted to 0.10 M with NaC104. Figure 2 illustrates the effect of the hydrogen ion concentration on the observed rate coefficient, kobsd. This effect can be explained by recognizing the basic characteristics of the two nitrogens responsible for coordination to the iron(I1) central
__
_.__
molar absorptivity, M - ' cm-'
11100 ( 1 )
8650 ( 3 )
27900 ( I )
solubility in water, g / l O O mL, 25 "C
0.3 (17 )
0.5 ( 1 7 )
7 (2)
PK,
4.96 (18)
4.33 ( 1 9 )
formation rate coefficient, k f , M - 3 s-'
( 2 . 2 5 0.22) X 10" I = 0.50 25 "C (15)
(1.4
3.13 t 0.02a ( c ) 3.27 i 0.05* ( c ) (3.08 i 0.03) x 10" I = 0.10 (4.25 25 "C L (0.05) c) x
dissociation rate coefficient, k,, S-'
7.5 x
formation ( k f / h d ) constant, M - 3 25 "C (direct) activation energy of formation, kJ/mol a
By kinetic interpretation. -
-
10-5
I = 0.50 25 "C (15) 3 x lo2' (15) 2 x 10'' ( 1 8 )
By potentiometric titration.
(6)
Since the reaction goes to completion with an insignificant dissociation of the complex in the pH range studied, the second term of eq 6 may be ignored leaving d[FeL,] / d t = k3[FeLz][L] (7)
Table 11. Summary of Data for the Complexation of Iron(11) b y 1,lO-Phenanthroline, 2,2'-Bipyridine, and Ferrozine 1,lO-phenanthroline 2,2'-bipyridine ferrozine
i_
(5)
which is the result of
I
0.2) x
1013
I = 0.025 1 7 "C ( 1 4 ) 1.2 x 10-4 I = 0.025 25 "C ( 1 4 ) 1 . 2 1 x 1017 ( 1 4 ) 1.17 x 1 0 1 7 ( 1 4 ) 0 (14) This work.
I = 0.10
25 "C ( c ) (7.2 i 0.3) X l O I 5 ( c ) (3.6 i 1.6) X l o t 5( 2 ) 9.87 i. 0.17 (c)
ANALYTICAL CHEMISTRY, VOL. 56, NO. 4, APRIL 1984
ion and the proton competition for those basic sites. Such consideration leads to a postulation that the rate is directly proportional to the concentration of the unprotonated form of the ligand, which, from equilibrium considerations,is given by [L] (Ka[HL+]1/ [H+] (8) By taking into consideration the mass balance equation for the ligand, [LIT = [HL+] + [L], eq 8 can be rewritten as [L] = ([LlTKa)/(Ka
+ [H'])
(9)
and substituting this into eq 5 results in d[FeL31/ d t = kf[Fe(II)l(([Ll~Ka)/(Ka+ [H'l)13
(10)
Equation 10 was used to fit the data in Figure 2 for k f and K , by assuming kobsd
= kf(([L]~Ka)/(Ka+ [H'])13
(11)
and using an activity coefficient of 0.81 (16) to convert the pH to [H+]. The solid line in Figure 2 represents the best fit with a pKa value of 3.13 f 0.02 and a forward rate coefficient of (3.08 f 0.03) X lo1' M-3 5-l. To test thc assumption t h a t the rate is affected by the ratio of the unprotonated to the protonated form of the ligand, the pK, of the ligand was independently determined by potentiometric titration. Ten titration values around the half equivalence point were used to determine the pK, from the equation
Ka = [H+I/((CL/(CB+ [H'I - [OH-])) - 1) (12) where CL is the total concentration of ligand and C B is the total concentration of added strong base. Equation 12 is derived, without simplifying assumptions, from the equations for K,, K,, and the charge and mass balance equations for the titration reaction. A pKa value of 3.27 f 0.05 was obtained which agrees closely with the value extracted from the kinetic interpretation. Dissociation Reaction. The complex dissociation reaction was studied in 0.10 M hydrochloric acid and allowed to react overnight. The dissociation was found to be first order in complex with a dissociation rate coefficient of (4.25 f 0.04) X s-'. This value in conjunction with the value of the complex formation rate coefficient reported earlier in this paper yields a formation constant for FeL, equal to (7.2 f 0.3) X 1015 M-3, which is of the same order of magnitude as the value of (3.6 f 1.6) X 1015 M-3 obtained by a study under equilibrium conditions (2). Temperature Dependence of the Formation Reaction Rate. Rate coefficients were experimentally determined at
757
temperatures of 18.0, 24.6, 30.0, 35.2, and 40.0 "C. The Arrhenius plot for the corresponding values yielded an activation energy for the complex formation of 9.87 f 0.17 kJ/mol. Low values for the activation energy of formation seem to be characteristic of this type of ligand since for the iron(I1) complexation by 2,2'-bipyridine a value of zero activation energy has been reported (14). Comparison with 1,lO-Phenanthroline and 2,2'-Bipyridine. Some information is available regarding ligands for iron(I1) commonly used in analytical chemistry and of similar structure to ferrozine. In Table I1 a summary of data regarding these ligands and ferrozine is presented. Both the rate coefficients and equilibrium constant values are lower for ferrozine than for the other two reagents; still these are substantially large values in the pH range of analytical use so as to classify the complex formation as favored and fast. The lower pK, for ferrozine, however, allows one to work at lower pH values with considerably higher rates than with the other two ligands. Registry No. Iron, 7439-89-6;ferrozine, 69898-45-9.
LITERATURE CITED (1) Stookey, L. L. Anal. Chem. 1970, 42, 779-781. (2) Glbbs, C. R. Anal. Chem. 1976, 48, 1197-1201. (3) Schilt, A. A. "Analytical Applications of 1,lO-Phenanthroline and Related Compounds"; Pergamon Press: New York, 1969. (4) Carter, P. Anal. Biochem. 1971, 4 0 , 450-458. (5) White, J. M.; Flashka, H. A. Clin. Chem. (Winston-Salem, N . C . ) 1973, 19, 526-528. (6) Yee, H. Y.; Goodwln, J. F. Clin. Chem. (Winston-Salem, N.C.)1974, 20, 188-191. (7) Eswara Dutt, V. V. S.; Hanna, A,; Mottola, H. A. Anal. Chem. 1976, 48, 1207-1211. (8) Brown, J. C. Agron. J . 1972, 6 4 , 240-243. (9) Palnton, C. C.; Mottola, H. A. Anal. Chem. 1981, 51, 1713-1715. (10) de Andrade, J. C.; Rocha, J. C.; Pasqulnl, C.; Baccan, N. Analyst (London) 1983, 108, 621-625. (11) Betterldge, D.; Sly, T. J.; Wade, A. P.; Tlllman, J. E. W. Anal. Chem. 1983, 55, 1292-1299. (12) Thompsen, J. C.; Mottola, H. A. Chem., Siol., Environ. Instrum., in press. (13) Zuberbuhler, A. D.; Kaden, T. A. Chimia 1977, 3 1 , 442-444. (14) Baxendale, J. M.; George, P. Trans. Faraday SOC. 1950, 46, 736-744. (15) Lee, T. S.; Kolthoff, I.M.; Leussing, D. L. J . Am. Chem. SOC.1948, 70, 3596-3600. (18) Krumholz, P. J . Am. Chem. SOC.1949, 71, 3654-3656. (17) Wlndholz, M. "The Merck Index"; Merck and Co., Inc.: Rahway, NJ, 1976. (18) Lee, T . S.; Kolthoff, I.M.;Leussing, D. L. J . Am. Chem. SOC.1948, 70. 2348-2352. (19) Baxendale, J. H.; George, P. Trans. Faraday SOC. 1950, 46, 55-63.
RECEIVED for review September 26,1983. Accepted January 9, 1984. This work was supported by a grant from the National Science Foundation (CHE-7923956).
