Grantee Research Project Results
Final Report: Understanding Mechanisms of Green Oxidation Catalysis by Iron-TAML Peroxide Activators
EPA Grant Number: R832245Title: Understanding Mechanisms of Green Oxidation Catalysis by Iron-TAML Peroxide Activators
Investigators: Collins, Terrence J. , Ryabov, Alexander D.
Institution: Carnegie Mellon University
EPA Project Officer: Hahn, Intaek
Project Period: September 1, 2004 through August 31, 2007
Project Amount: $242,300
RFA: Targeted Research Grant (2004) Recipients Lists
Research Category: Targeted Research
Objective:
The project was aimed at understanding the basic mechanistic behavior of the oxidation catalysis exhibited by iron(III)-Tetra-Amidato Macrocyclic Ligand (TAML) catalysts (1) that are important green chemistry catalysts. TAML® catalysts utilize hydrogen peroxide as their most important source of oxidizing equivalents. The primary objective was to learn the underlying chemical reasons for their superior performance in a variety of environmentally important processes such as the bleaching of dyes, wood pulp, and colored effluents, the degradation of myriad persistent organic pollutants, and the decontamination of chemical and biological warfare agents. Iron-TAML catalysis reaction rates compare with what is typical for related enzymatic reactions and they are capable of at least 103–104 turnovers. We planned to explore the hypothesis that TAML activators function in a mechanistically parallel fashion to peroxidase and catalase enzymes. Once the chemical origin of the unique reactivity for synthetic peroxide activators had been identified, the information was used for further catalyst improvement. Understanding the first-order mechanistic features had already multiple positive impacts. Prior mechanistic studies had suggested ways for trying to advance the already unprecedented catalytic performance of FeIII-TAMLs. One example involved learning how to replace hydrogen peroxide by dioxygen as the primary oxidizing agent. Another focused upon illuminating how to make superior activators. At the same time, the work was identifying novel tasks essential for further characterization the catalytic features of the activators. While this work was being pursuing this work, scholarship was also carried to characterize what is known about the health and environmental implications of pharmaceuticals in the environment. This led to an article referencing EPA support in the Green Chemistry special issue of Chemical Reviews entitled, Human Pharmaceuticals in the Aquatic Environment: A Challenge to Green Chemistry.[Publication 1] And in the same period, a Scientific American article on TAML activators entitled, Little Green Molecules, was produced.[Publication 11] Both these contributions were seen as important contributions to the broader understanding of major problems and developing solutions associated with the public health and environmental impacts of the chemical enterprise. In this context of our general studies on TAML activators and the ongoing scholarship involved with building the field of green chemistry, this specific project was undertaken along the following lines.
Specific Aims of the Project (as formulated)
New H2O2-activating catalysts could substantially green oxidation chemistry by providing economic alternatives to chlorine- and metal-based technologies. In the past, the rapid degradation of candidate catalysts by oxidation and/or hydrolysis has thwarted most development efforts. Iron-TAML activators are now the most significant exceptions. FeIII-TAMLs activate, inter alia, hydrogen peroxide for numerous applications as described in the Background and Significance section of the proposal. We proposed to map the catalytic cycles by determining the mechanistic character of fundamental processes that are significant to the catalysis and by characterizing the reactive intermediate(s). The original specific goals of the proposed research were as follows.
First year:
- The speciation of the solid FeIII-TAML complexes that represent the physical source of the catalysts in water will be determined by various spectroscopic techniques. The resistance of the catalysts to extreme conditions such as high/low pH will be measured to parameterize their stability over a range of possible operating conditions.
- A detailed mechanistic description will be obtained for each process that is a part of the catalytic cycle through the steady-state investigation of the kinetics of oxidation of model simple inorganic/organometallic 1e- donors such as ferrocene, ruthena(II)cycles, etc., and more complicated molecules such as phenols, organic dyes, etc. The processes that are currently thought to be critical on the basis of preliminary results are: peroxidase-like activity in which a substrate is oxidized by H2O2, catalase-like activity of FeIII-TAMLs in which oxygen is produced from H2O2, catalase versus peroxidase activities of FeIII-TAMLs, intramolecular oxidative degradation of FeIII-TAMLs leading to catalyst inactivation, hydrolytic degradation of FeIII-TAMLs leading to catalyst inactivation, the nature of the catalytically active reaction intermediate formed from FeIII-TAMLs and H2O2—in our working hypothesis, it is an Fe-oxo complex.
- The mechanistic conclusions reached under the steady-state will be further investigated in single turnover experiments using a stopped-flow technique.
- Spectroscopic characterization (and possibly isolation and X-ray study) will be attempted for the purported reactive intermediate by electronic, EPR, Mössbauer, and possibly resonance Raman spectroscopy after its generation by different oxidants in water.
Second year:
- The knowledge of basic mechanisms of catalysis by FeIII-TAMLs will be applied to the systematic studies of the catalyzed degradation of surrogates of chemical warfare agents where FeIII-TAMLs display mixed oxidative and hydrolytic activity.
- Learn how to control general acid catalysis of the demetalation of FeIII-TAML activators in neutral and acidic media.
- Explore micellar tuning of FeIII-TAML activation of principal green oxidizing agents.
- Develop mechanistic understanding of the very high rate of activation of benzoyl peroxide by FeIII-TAML activators.
- Quantify the reactivity of the transient oxidized form/s of Fe-TAML activators toward various electron donors (substrates) by steady state and transient kinetic measurements.
- Use hexacloroiridate(IV) as a probe for studying the proprieties of oxidized Fe-TAML activators.
- Move dioxygen chemistry of FeIII-TAML activators into water: activation of O2 in reverse micelles and photochemical catalytic cycles.
- Design and develop FeIII-TAML oxidation processes mediated by electron carriers.
- Design and synthesize new “ideal” FeIII-TAML catalysts suggested by the mechanistic studies.
Chart 1. Iron(III)-TAML activators and other key chemicals mentioned in this proposal.
Summary/Accomplishments (Outputs/Outcomes):
Objectives 1-1 and 1-3: Catalysts at work: High Oxidative Efficacy and Rapid Kinetics of Oxidation. The Catalase Activity: [To be submitted manuscript 1]
The exceptionally high peroxidase-like and catalase-like activities of iron(III)-TAML® activators 1 have been established at pH 6-12.4 and 25-45 °C. While complexity precludes unambiguous data interpretation for the kinetics of Pinacyanol chloride oxidation, oxidation of “ruthenium dye”, the cyclometalated 2-phenylpyridine derivative 2 occurs via the stoichiometric equation
[RuII] + ½H2O2 + H+ [RuIII] + H2O (1)
The kinetics of reaction 1 using different TAML activators 1 has been investigated over a wide pH range by following a decrease in absorbance at 480 nm due to the oxidation of RuII to RuIII. Enzyme-like kinetic traces without induction periods have had significant steady regions and displayed curvature at degrees of conversion around 85%, suggesting zero-order dependence on the concentration of 2. Absorbance changes in the absence of 1 were found to be negligible under these conditions. The steady-state rates have been measured at different concentrations of 1 (3.20×10-8-4.80×10-7 M), H2O2 (8.25×10-5-1.11×10-3 M), and 2 (2.58×10-5-1.96×10-4 M). As anticipated, the rates are directly proportional to the concentrations of 1 implying first order dependences on 1. First order dependence has been also observed in H2O2 concentration. The steady-state rate is independent of the 2 concentration. The zero-order in the RuII complex and the first order in H2O2 hold over the entire pH range (6-11.5). Thus, the experimental rate law given by eq 2 suggests that the rate-limiting step involves the interaction between 1 and H2O2 and followed by fast oxidation of the substrates.
rate = [1][H2O2] (2)
Free radical oxidation chemistry1 is not involved in these FeIII-TAML catalyzed oxidations to any detectable degree. The efficient hydroxyl radical scavenger, mannitol,2,3 has no effect on the rate of reaction 1 catalyzed by 1 at pH 9.0. Similar results were obtained when urea was used. Based on these results, we conclude that this peroxide oxidation, catalyzed by 1, is not mediated by free radical processes.
Figure 1. pH profiles of the rate constants k2,obs for the oxidation of 2 by H2O2 catalyzed by 1a at different temperatures: [1c] 1.28×10-7 M, [2] 5.81×10-5 M, [H2O2] 3.3×10-4 M, 0.01 M phosphate. | Scheme 1. Proposed mechanism of catalyzed by 1 oxidation of “ruthenium dye” 2 by hydrogen peroxide. |
The dependencies of on pH for 1c at 25-45 °C are shown in Figure 1. The rate constant increases dramatically at pH > 9 reaching a maximum around 10.5 and declines at pH > 11. The pH profiles in Figure 2 are rationalized quantitatively using the known pKa’s of the catalysts4 and H2O2.5 Iron(III)-TAML activators are six-coordinated in water with two axial aqua ligands, the first pKa1’s being in the range 9-10.4 The diaqua and aqua/hydroxo FeIII complexes are likely reactive species that interact pairwise with either H2O2 or its conjugate base (pKa2 ~ 11.2-11.65) with the rate constants k1-k4 to give oxidized TAML, the nature of which is discussed below. Oxidized TAML reacts with the ruthenium(II) dye 2 very fast (Scheme 1) and therefore the reactions between iron(III)-TAML and hydrogen peroxide are treated as irreversible. Scheme 1 leads to the rate expression 3 with kinetically indistinguishable k2 and k3 pathways. The experimental data have been fitted to eq 3 neglecting either k2 or k3 pathways.
(3)
This allowed calculation of the rate constants k2 and k3, respectively, as well as k1 and k4. The best-fit values for the rate and equilibrium constants (Table 1) have been used for the emulation of the solid lines in Figure 1. The rate constants k1 and k4 are by a factor of ca. 100 and 10 lower compared to k2, respectively, and this accounts for the sharp maxima in Figure 1. A trend in variation of the intrinsic rate constants k1-k4 is anticipated in view that the interaction between 1 and hydrogen peroxide is a redox reaction. The deprotonated species [FeL(OH)(H2O)]2- is more electron-rich than [FeL(H2O)2]- and therefore H2O2 oxidizes the former much faster than the latter (k2>>k1). In turn, deprotonated hydrogen peroxide HO2- is more electron rich than H2O2 and therefore it reacts slower with [FeL(OH)(H2O)]2- (k2>k4) though a coulombic factor may also affect this step.
Table 1. Equilibrium and rate constants (in M-1 s-1), activation parameters for the FeIII-TAML-catalyzed oxidation of 2 (eq 1) by hydrogen peroxide (0.01 M phosphate).
FeIII- |
T/Activation parameter |
10-2×k1 |
10-4×k2 |
10-6×k3 |
10-3×k4 |
pKa1 |
pKa2 |
1a |
25 °C |
1.3±1.0 |
1.2±0.5 |
0.4±0.2 |
a) |
9.7±1.0 |
10.8±2.0 |
1g |
25 °C |
4.0±1.8 |
1.8±0.7 |
0.8±0.1 |
1.5±1.0 |
9.5±0.4 |
10.9±0.4 |
32 °C |
8.5±11.6 |
3.3±0.2 |
1.3±0.1 |
4.0±2.4 |
9.3±0.1 |
10.4±0.1 |
|
38 °C |
9.0±8.8 |
3.5±0.2 |
1.2±0.1 |
3.8±1.7 |
9.6±0.7 |
10.4±0.1 |
|
45 °C |
12.5±0.7 |
3.6±0.1 |
1.8±0.1 |
5.1±2.3 |
9.2±0.1 |
10.9±0.1 |
|
ΔH≠ kJ mol-1 |
46±12 |
28±13 |
30±10 |
49±19 |
|||
ΔS≠ J mol-1 K-1 |
-51±13 |
-77±25 |
-37±7 |
-32±11 |
|||
a) could not be reliably estimated. |
In the absence of electron donors, complexes 1 display a catalase-like activity (eq 4). Dioxygen evolution is visual at [H2O2] > 0.01 M. This catalytic feature has been studied kinetically by monitoring the initial rates of O2 formation with a Clark electrode.
H2O2 H2O + ½ O2 (4)
The reaction stoichiometry is given in eq 4. Reaction 4 follows first-order kinetics in 1 ((0.6-9)×10-6 M for 1a-δ) and H2O2 (2.3×10-4-0.157 M) in the absence of an electron donor. Thus, rate expression 5 holds for the catalase-like activity, which is identical to eq 2 for the peroxidase-like activity.
rate = [FeIII-TAML][H2O2] (5)
The rate constants are again pH-dependent and the pH profiles for reactions 2 (Figure 1) and 4, both catalyzed by 1g, are similar. Equation 3 has therefore also been applied for fitting the catalase data. The best-fit rate and equilibrium constants are summarized in Table 2. The rate constants in Tables 1 and 2 indicate that the "catalase-like" rate constants are somewhat lower. To relate the activation parameters, the temperature dependence of has been investigated at 25-45 °C and pH 10.1. At this pH, the k2 pathway dominates—its contribution into the overall rate is more than 99%. Therefore, the effective enthalpy of activation obtained from the temperature dependence of should be associated with ΔH≠ for the k2 pathway. The corresponding ΔS≠ should be calculated using k2 from Table 2. Thus calculated ΔH≠ and ΔS≠ are shown in a footnote to Table 2. The similar rate laws (eqs 2 and 5), pH profiles, and the values of the observed second-order rate constants ( and ) suggest a common reaction intermediate in reactions 1 and 4.
Table 2. Rate (in M-1 s-1) and equilibrium constants, activation parameters for iron(III)-TAML catalyzed disproportionation of hydrogen peroxide (eq 4) at 25 °C and 0.1 M phosphate.
FeIII-TAML |
10-2×k1 |
10-4×k2 |
10-3×k4 |
pKa1 |
pKa2 |
1a |
1.4±1.2 |
0.51±0.03 a) |
b) |
11.0±0.2 |
12.1±0.4 |
1i |
4.28±1.8 |
0.90±0.06 |
0.87±0.6 |
11.0±0.2 |
12.15±0.35 |
1g |
1.3±0.7 |
0.64±0.04 c) |
1.6±0.2 |
9.75±0.11 |
11.3±0.3 |
1c |
6±3 |
1.68±0.08 δ) |
3.2±0.2 |
10.5±0.2 |
11.6±0.2 |
a) ΔH≠ 23 ± 3 kJ mol-1; ΔS≠ -95 ± 20 J mol-1 K-1; b) could not be reliably estimated; c)ΔH≠ 14.6 ± 0.7 kJ mol-1; ΔS≠ -123 ± 13 J mol-1 K-1; d)ΔH≠ 8.1±3.3 kJ mol-1; ΔS≠ -137±63 J mol-1 K-1. |
A plausible stoichiometric mechanism accounting both catalytic features of FeIII-TAMLs is shown in Scheme 2. Taking all three steps into consideration the rate of the oxidative (peroxidatic) activity of 1 is given by eq 6.
(6)
Here ED is an appropriate electron donor. The catalase-like activity is described by eq 7.
(7)
The oxidation of ruthenium dye 2 is a zero-order reaction in 2. This implies that kII[ED] >> {kI + [H2O2](kI+kIII)}. Equation 6 becomes very simple, i.e.
(8)
The first order dependence in hydrogen peroxide for the catalase-like activity holds when [H2O2](kI+kIII) >> k-I + kII[ED]. Then, in the absence of added ED eq 7 becomes
(9)
Thus, both peroxidatic and catalase-like reactions have similar rate laws but the effective second order rate constant should be equal (when kIII>kI) or lower (when otherwise) than provided the latter is zero-order in ED. The conditional rate constants kI and kIII are pH dependent. The data in Figures 2 allow us to estimate the dependence of the rate constant kIII on pH using eqs 8 and 9. Thus the calculated plot of kIII versus pH in the range 8-11, where the accuracy in measuring and is the highest, is shown as a dashed line in Figure 2. The rate constant kIII is somewhat lower than kI. This agrees with the fact that observation of product/s of the interaction between complexes 1 and H2O2 by spectral techniques is feasible. The shape of the pH profile for kIII in Figure 2 is explicable. Hydrogen peroxide becomes a better reducing agent upon deprotonation and therefore the rate constant kIII starts to rise at pH > 10. In fact, the reduction potential for the half-reaction O2 + 2 H2O + 2e → H2O2 + 2 OH- equals -0.146 V.6
Scheme 2. Mechanism of catalase and peroxidase activity of 1.
In conclusion, the catalytic activation of H2O2 by iron(III)-TAML activators is characterized by peroxidase-like and catalase-like activity. The rate determining step of the peroxidase-like activity is the interaction between 1 and H2O2. The step is strongly pH-dependent and the maximal activities are observed at pH around 10. The rate constants for the activation of H2O2 under the optimal conditions (>104 M-1 s-1) approach a lower limit of the corresponding enzymatic peroxidase-catalyzed reactions. The activation of H2O2 results in the formation of the reactive intermediate formulated as the oxoiron(V) species which may react with substrate or conproportionate with more iron(III) to give the oxoiron(IV) intermediate (see next section of this report). The further oxidation of electron donors occurs in subsequent fast steps. In the absence of electron donors, the iron(III)-TAML activators display catalase-like activity with the production of dioxygen. The rate of this step is significantly slower than the oxidative bleaching of the dye Safranine O. In addition to underpinning the significant technological importance of the iron(III)-TAML activators,7,8 these results signify that the activators are prominent functional models of the catalase-peroxidase enzymes. Iron(III)-TAML activators display both high catalase-like and peroxidase-like activities, but the latter dominates in the presence of many electron donor substrates other than H2O2 itself, thus eliminating unproductive consumption of hydrogen peroxide. This is a significant finding for industrially relevant oxidation processes.
In terms of publishing these extremely detailed studies, we have held back while we continue to carry out additional tests to continue to establish that this enticing mechanistic analysis does not suffer from any mechanistic oversight on our part. Thus, we have always been worried that the above analysis assumes that the Fe-TAML activators return always to a ferric resting state upon each revolution of the catalytic cycle. However, as described in the next section, the 1 activators in the presence of peroxide evolve to iron(IV)-containing solutions essentially instantly and in the presence of excess peroxide, continue to evolve oxygen by a catalase-like process. We have been concerned that the variable pH behavior could be explained, at least in part, by a partitioning of the total Fe-TAML in solution among various Fe(IV) species controlled by pH with each species having a differing rate of catalytic activity. In order to check on this possibility, it was essential to be able to identify and understand the catalytic role of any species that might be present in the steady state kinetics medium, and to this end, among other things, we conducted work at low temperature to characterize the speciation of the Fe-TAML under differing pHs in the presence of peroxide using Mössbauer spectroscopy. We found that there is a tightly pH dependent equilibrium between the iron(IV)-oxo monomer, [Fe(TAML)O]2-, and bis-iron(IV)-oxo dimer, [{Fe(TAML)}2O]2-, with the former favored at higher pHs and the latter at lower pH’s. We are now studying the kinetics of the reactions of these species with 2. If the reactions of these iron(IV) species with 2 are fast wrt to the rate-determining process in our above scheme (a highly likely contingency which is implied by the results described above), then the conclusion that each revolution of the catalytic cycle returns the catalyst to the iron(III) resting state would be consistent and what we think is a final step toward publication of this very complex mechanistic story will have been made. In comparing the variation of the catalase-like behavior of the Fe-TAML catalysts with pH with the peroxidase-like behavior with 2, the same overall shape of the vs pH graph is observed, but the rates are slightly different and this may reflect that the Fe-TAML is not returning to the ferric state with peroxide. The manuscript for this work (available on request), will be held back until the one final test above is completed.
Objective 1-2: Catalysts at Work: The Nature of The Reactive Intermediate [Publication 2 and To Be Submitted Manuscript 2]
Iron(V)-oxo species have been proposed as key reactive intermediates in the catalysis of oxygen activating enzymes and synthetic catalysts. In Science this year, we reported the synthesis of [Fe(TAML)(O)]- in nearly quantitative yield. Mass spectrometry, Mössbauer, electron paramagnetic resonance and X-ray absorption spectroscopies, reactivity studies, as well as density functional theory calculations, show that this long-lived (hours at -60°C) intermediate is an S = 1/2 iron(V)-oxo complex. Iron-TAML systems have proven to be efficient catalysts in the decomposition of numerous pollutants by hydrogen peroxide and the newly characterized species is a likely reactive intermediate in these reactions. This is the first case of a clearly characterized Fe(V)oxo complex, but thus far, we have only been able to isolate it in organic solvents where we can work at -60 to -80 °C.
The situation in water is quite different. We do not see evidence for the iron(V)-oxo, but we cannot work at low temperatures required for its isolation. Addition of H2O2 to aqueous solutions of iron(III)-TAML complexes 1 produces brownish-green colors. The spectral changes can be measured by uv-vis spectroscopy (Figure 2). Less than a stoichiometric amount of H2O2 causes a major increase in the absorbance in the range 350-550 nm. It requires several minutes to obtain invariable spectra at pH below 8.5, but the reaction becomes much faster at pH > 9. Two isosbestic points hold at 298 and 324 nm for 1e. At pH 11, the changes are fast and the final spectrum remains practically unchanged within 30 minutes. The inset in Figure 2 shows absorbance changes at different wavelengths as a function of added [H2O2], which allow for an estimate of the stoichiometry of the interaction. Matching data obtained at three wavelengths indicate that two iron(III) centers interact with one H2O2 molecule suggesting that the absorbing species are raised one oxidation equivalent above the resting state. Similar results were obtained with 1a.
Figure 2 (left). Spectral changes of 1e (2.5×10-5 M) in the presence of 0, 2×10-6, 4×10-6, 6×10-6, 8×10-6, 12×10-6, and 16×10-6 M H2O2 at pH 11.4, 25 °C. Inset: Titration curves at 242 (), 280 (), and 420 () nm showing a 2:1 (FeIII:H2O2) stoichiometry.
Figure 3 (right). Titration of the intermediate generated from 1e (X1 = X2 = Cl, R = Me) and H2O2 by [Fe(CN)6]4- showing a 1:1 stoichiometry at pH 11.35, 0.1 M KPF6, 25 °C.
The same conclusion was reached using [FeII(CN)6]4- for the back reduction of the species derived from 1 and H2O2. The oxidized TAML was derived from 1e and a half equivalent of hydrogen peroxide (larger amount of H2O2 does not increase the absorbance). The data in Figure 3 shows that 1 equivalent of [FeII(CN)6]4- reduces the oxidized species. The optical density does not reach the initial level exactly due to absorbance of the oxidized titrant, i.e. [FeIII(CN)6]3-, at 420 nm.
When an aqueous solution of tert-BuOOH is added to a solution of 1a at pH 14 (0.01 M phosphate) or 1 M KOH, a red color, stable at room temperature for hours, appears promptly. The same species can be obtained at any pH higher than 12. A titration of 1a by tert-BuOOH followed by the uv-vis spectroscopy revealed that ca. 0.5 equiv of the oxidant per FeIII is needed to obtain the maximum change. Similar results have been obtained with H2O2 under identical conditions all suggesting the formation of an FeIV species.
The zero-field Mössbauer spectra of a frozen aqueous solution of the red compound (generated from 1×10-3 M 57Fe-1a and 0.5 eq t-BuOOH) at 4.2 K consists of one doublet with a quadrapole splitting, ΔEQ = 4.2 mm s-1 and isomer shift (versus Fe metal at 298 K) δ = -0.19 mm s-1 (Figure 4). This highly negative isomer shift is indeed indicative of a high valent FeIV species and agrees with the titration studies. Furthermore, variable high field (1.5, 6.5, and 8 T) and high temperature (150 K) studies indicated the S = 1 species. Such intermediate spin FeIV species with a similar ligand and with the axial cyanides have been reported earlier from our group. Remarkably, the Mössbauer data accounts for ~100 and 80% of the total iron is this oxidized species at pH 14 and 12, respectively. Minor instability of the oxidized species has been detected at pH < 12.
Figure 4. The zero-field Mössbauer spectra of a frozen aqueous solution of the mixture of 57Fe-1a (1×10-3 M) and 0.5 eq t-BuOOH at 4.2 K. See text for details.
The uv-vis, Mössbauer, and EPR spectroscopic data obtained and reported previously9 suggest a rather detailed and complicated picture regarding the speciation of oxidized TAML species derived from 1 and various oxidants in aqueous solution (Scheme 3). If peroxides ROOH are primary two-electron oxidants, the oxidation of 1 gives presumably the oxoiron(V) intermediate 4. Such TAML-based species has never been detected in water but its existence in non-aqueous media has reliably been proved as part of this project and published earlier this year in Science.10 Similar to the sterically hindered anionic porphyrin chemistry,11 the highly oxidized intermediate 4 is rapidly quenched by 1 into less oxidized iron(IV) species, the speciation of which is significantly pH dependent. At pH 14, all iron (ca. 95%) is monomeric oxoiron(IV) species 5', though theoretical DFT simulations (see below) do not exclude its aquated formulation as 5". The latter seems reasonable as the monomerization of the FeIV-O-FeIV dimer is induced by hydroxide ions.
The nature of the reactive intermediates as 5' and 5" 9 has been analyzed by DFT using Gaussian 03.12 We have previously shown that the B3LYP functional and the 6-31G level of basis function predicts reliably electronic and structural properties of complexes 1.13 The same method was used for characterization of the structural and spectroscopic features of 5. The Cs symmetry is maintained in all cases to expedite the calculation time and full geometry optimization. The Mössbauer parameters in Table 3 are calculated using 6-311G with the established calibration curve. The parameters ΔEQ and δ are known to be most accurately predicted by DFT. The results in Table 3 indicate that the formulation of 5 as a 5'/5" pair gives the best match with the experimental values of ΔEQ and δ. An oxo-bridged dimer or FeIII-radical cation (ligand-based oxidation)13 were ruled out at pH > 12 on the basis of either spin or Mössbauer parameter considerations.Scheme 3. Speciation of oxidized TAML species derived from 1 in aqueous solution. Axial aqua ligands are not shown for clarity.
Our calculations also show that the S = 1 spin state is energetically more favorable than both high or low spin states. Time dependent (TD-DFT) analysis has been also applied to the geometry optimized structure of 5' (Figure 5) to ensure they indeed represent structures of the lowest energy. It is interesting to note that the theory predicts 5- and 6-coordinated iron(IV) in 5' and 5", respectively. The central metal in 5' does not hold axial water. This is obviously due to a strong trans effect of the axial oxo ligand of charge -2. It should however be added that the six coordinated iron with weakly bound water cannot be excluded.
Figure 5. DFT-predicted lowest energy structures of oxidized species 5'.
Table 3. Calculated by DFT anticipated parameters of the Mössbauer spectra of products generated from 1 and tert-BuOOH in basic (pH > 12) aqueous solution.
Model |
|
|
Experimentally observed |
|||
ΔEQ / mm s-1 |
4.4 |
3.6 |
3.5 |
2.5 |
1.8 |
3.9 |
δ / mm s-1 |
-0.07 |
-0.15 |
-0.12 |
0.16 |
-0.07 |
-0.19 |
In contrast with organic solvents, the aqueous environment puts severe limitations on our ability to observe the Fe(V) state. All oxidizing agents tested so far increase the oxidation state of iron in 1 by one equivalent to produce iron(IV). Iron(IV) is generated by dioxygen in non-polar organic solvents,9 probably via an iron(V) species and indeed this very species is produced from 1 and organic peroxides in tert-butyronitrile.10 The speciation of these iron(IV) TAML derivatives is crucially pH dependent. The monomeric oxoiron(IV) complexes 5 dominate at pH > 12, whereas at pH 8-10 the major species is the μ-oxo dimer. Though it is just iron(IV), its reduction potential is similar to that of the [IrCl6]2-/3- couple, i.e. peroxides generate very powerful oxidizing agents from 1 in water (see below). Interestingly, the reduction potential is comparable to those of both to hexaaqua FeII/III 6 and Compound II/resting state of horseradish peroxidase14,15 couples though the oxidative chemistry of 1 is predominantly non-radical, i.e. biological like.
Objective 1-4: New Approaches For Evaluating The Catalyst Lifetimes Intermediate [Publication 12]
Kinetic measurements of the FeIII-TAML-catalyzed oxidation of conventional dyes such as Pinacyanol chloride, Safranine-O and Orange II (Chart 1) by H2O2 have revealed that the oxidation does not follow monoexponential kinetics under pseudo-first-order conditions, i.e. under at least ten-fold excess of H2O2 over all dyes used. The data shown in Figure 6 illustrate the non-exponential kinetics. The 1a-catalyzed bleaching of Safranine-O was initiated by addition of an aliquot of H2O2. At low catalysts concentrations the bleaching does not go to completion even at [H2O2] >> [dye]. Addition of the second aliquot of H2O2 at time a does not restart catalytic bleaching (Figure 1) suggesting the catalyst has been deactivated. Accordingly, addition of a new aliquot of 1a resumes the reaction. The bleaching of Safranine-O can be completed by successive additions of the 1a (Figure 6 shows incomplete bleaching). The robustness of Safranine-O toward oxidation by FeIII-TAMLs and H2O2 makes it convenient for studying FeIII-TAML catalyst half-lives.
Figure 6 (left). Kinetics of 1a-catalyzed bleaching of Safranine-O (4.3×10-5 M) by 0.012 M H2O2. Initial concentration of 1a (7.5×10-8 M); aliquots of the same amount of 1a were added after complete inactivation of the catalyst giving rise to the stepped dependence. The dashed line shows that addition of 0.012 M H2O2 does not resume the catalytic bleaching. Inset shows linearization of the data obtained after each addition of 1a in terms of eq 13. Conditions: pH 11, 25 °C.
Figure 7 (right). Dependence of the amount of bleached dye expressed as Ao on the total amount of dye expressed as Ao for 1a-catalyzed (7.5×10-8 M) bleaching of Safranine-O (initial concentration 4.3×10-5 M) by H2O2 (0.012 M) in terms of eq 15, see text for details. Conditions: 25 °C, pH 11.0.
A methodology for quantifying the catalytic performance is based on Scheme 4. The intermolecular inactivation is assumed to be kinetically insignificant at this point, i.e. k2i ≈ 0, because the majority of the dye bleaching experiments were run at very low catalyst concentrations (10-6-10-8 M). Therefore, second-order pathways should not play a substantial role. The medium-induced hydrolytic degradation is known to be kinetically insignificant. Complexes 1 are subjected to the H+-catalyzed demetalation for which the pseudo-first-order rate constant is given by kobs = kα[H+] + kγ[H+]3.4 The demetalation becomes significant with decreasing pH at pH ~ 4 for 1a, a methyl-tailed TAML catalyst, but only at significantly lower pH for 1c, a fluorine-tailed TAML catalyst. The bleaching experiments reported in our paper were performed in the pH range 10-12 and therefore kδ is negligible. There is an additional general-acid catalyzed demetalation of 1a at pH 5-7 which is also unimportant because its time scale is much longer than that of the bleaching. No general-acid-catalyzed demetalation is observed for the fluorine-tailed 1c.
Scheme 4. General reaction scheme showing the major catalytic steps and the steps leading to inactivation of the resting state and the active form of a catalyst involved in a two-substrate process.
The stoichiometric mechanism of catalysis in Scheme 4 is based on eq 6 when kIII is negligible. The rate constants kI, k-I, and kII are effective, pH-dependent parameters. A straightforward evaluation of the rate constants ki and kII is feasible when k-I is negligible and the interaction between 1 and H2O2 is significantly faster than the oxidation of substrate itself, i.e., when the relation kI[H2O2] > kII[Substrate] holds. This condition does not usually hold for low-molecular-weight catalysts of hydrogen peroxide such as iron porphyrins.16 The interaction of synthetic catalysts with H2O2 is usually the slowest step in the oxidation process. In general, this is also true for 1, although the second-order rate constants kI are as high as 104 M-1 s-1 at pH around 10 and the relation kI[H2O2] > kII[Substrate] can be secured by using difficult-to-oxidize dyes with low values of kII. This opens a broader pH range for characterizing the operational stability of the FeIII-TAML catalysts. Safranine-O belongs to the class of difficult-to-oxidize dyes by FeIII-TAMLs. The relation kI[H2O2] > kII[Substrate] holds and the initial rate of its bleaching is directly proportional to [Safranine-O] and independent of [H2O2] (data not shown). Thus, eq 6 simplifies into eq 10.
kII[Substrate][FeIII] (10)
According to Scheme 4, the intramolecular inactivation of active forms of 1 is monoexponential. The corresponding rate constant is ki. Therefore the related differential equation for dye bleaching is given by eq 11.
= kII(Dt - x)([FeIII]t) (11)
Here Dt and [FeIII]t are total concentrations of the dye and an FeIII-TAML catalyst, respectively, and x is the concentration of a bleached dye at time t. Integration of eq 11 under the boundary conditions x = x∞ (x∞ is the concentration of bleached dye obtained with single catalyst aliquot) when t = ∞ affords
(12)
or
(13)
Equation 13 involves easy to measure absorbances At and A∞ at time t and ∞, respectively, instead of concentrations as in eq 12. Equation 13 suggests a simple procedure for analysis of data such as in Figure 6. If the double logarithm of the ratio At/A∞ is plotted against time, the slope of the straight line will equal -ki. The rate constant kII can then be calculated from the intercept which equals ln(kII[FeIII]t/ki). A linearization of the kinetic traces of Figure 6 in terms of eq 13 is demonstrated in the Inset. All four steps in Figure 6 have been analyzed. The values of ki and kII are nearly identical for each trace in Figure 6. The value of ki indicates that the half-life of 1a is 3.4 min under the operating conditions (25 °C, pH 11). The rate constant kII equals approximately 104 M-1 s-1.
Careful inspection of the data in Figure 6 indicates that the total amount of the bleached dye (x∞) after adding a new aliquot of 1a is not constant and gradually decreases after each new addition. This phenomenon is understood in terms of the general kinetic model proposed. The solution of differential equation 11 using the boundary condition x = 0 at t = 0 leads to eq 14.
(14)
Here A0 and At are optical densities at time t = 0 and t, respectively. Since At = A0 – x, x → x∞ when t → ∞. Substitution and rearrangement result in eq 15.
(15)
Eq 15 implies that x∞ should be directly proportional to the starting concentration of dye or to the absorbance A0. The straight line passing through the origin in Figure 7 confirms the proposed model. Moreover, the predicted slope of this line is 0.25±0.04 based on the known values of the rate constants kII and ki and the concentration of 1a (7.5×10-8 M). The slope determined from the experimental data is 0.24±0.02 which agrees well with the calculated value.
Figure 8. A. Kinetics of oxidation of Safranine-O (4.3×10-5 M) by H2O2 (0.012 M) by catalysts 1f, 1a, 1b, and 1c (7.8×10-5 M). B. Linearization of the data in terms of eq 13. Conditions: 25 °C, pH 11.0.
Catalysts 1a-f display variable performance (Figure 8). The bleaching by 1c is both the deepest and the fastest. Equation 13 was applied to all kinetic traces in Figure 8A affording satisfactorily straight lines shown in Figure 8B. The rate constants ki and kII summarized in Table 4 were calculated from the slopes and intercepts, as above. Similar experiments were performed at different temperatures covering the 13-60 °C range. The inactivation (ki) and bleaching (kII) rate constants in Table 4 have been used for calculating the corresponding enthalpies (ΔH≠) and entropies (ΔS≠) of activation, using the Eyring equation.17
Table 4. Rate constants (ki and kII) calculated for different FeIII-TAML catalysts at pH 11 using eq 13 from the data such as in Figure 8.
Catalyst |
T / °C |
103×ki / s-1 |
10-4×kII / M-1 s-1 |
1f |
13 |
0.44±0.05 |
0.24±0.03 |
25 |
1.5±0.1 |
0.7±0.1 |
|
42 |
5.7±0.6 |
1.2±0.1 |
|
60 |
33±3 |
3.6±0.4 |
|
1a |
13 |
1.9±0.2 |
0.9±0.1 |
25 |
3.4±0.3 |
1.1±0.1 |
|
42 |
5.2±0.5 |
1.5±0.2 |
|
60 |
12±1 |
3.1±0.3 |
|
1b |
13 |
2.5±0.3 |
0.99±0.1 |
25 |
6.2±0.4 |
1.9±0.2 |
|
42 |
11±1 |
3.6±0.7 |
|
60 |
18±2 |
5.6±0.6 |
|
1c |
13 |
4.8±0.5 |
6.3±0.6 |
25 |
13±1 |
10±2 |
|
42 |
20±2 |
14±1 |
|
60 |
45±2 |
23±3 |
The rate constants in Table 4 reveal interesting trends. (i) The catalytic activity of 1 in bleaching Safranine-O (kII) increases more than ten-fold at 25 °C when the tail ethyl groups of 1f are replaced by fluorine atoms in 1c. The fluorine-tailed compound 1c is the most active. The same level of reactivity is typical of horseradish peroxidase Compound II toward anilines and phenols.18 This emphasizes that FeIII-TAMLs are really efficient synthetic low-molecular weight peroxidase mimics. (ii) The inactivation rate constants ki also increase on going from 1f to 1c and a similar 10-fold gap holds. Previously, we have reported qualitative data showing a better performance of the methyl-tailed catalyst 1a versus its ethyl-counterpart 1f in bleaching of Pinacyanol chloride.19 Despite the fact that different inactivation processes are operating here, the quantitative kinetic data reported here also conform to the previous qualitative observations concerning comparative catalyst performances.
Our mechanistic studies of the FeIII-TAML catalysts convincingly show that the introduction of fluorine atoms onto the tail changes the reactivity in a dramatic way, allowing activation of H2O2 in acidic media (in addition to the basic range for these catalysts) while also producing significantly more aggressive oxidizing systems. The modification of the "head", in contrast, induces considerably more subtle changes in FeIII-TAML catalyst reactivity through aromatic ring mediated electronic effects, the structural features of the catalysts being otherwise the same.4 We also examined how electron-donating or electron-withdrawing substituents at the aromatic ring affect the rate constants ki and kII. Eight structurally similar, but electronically different FeIII-TAML catalysts 1, all with R = Me, have been investigated by applying eq 13. Practically useful, but rather curious results are found in the linear free energy relationship (LFER). There is a linear dependence between lnkII and lnki with a slope of –1. Electron-withdrawing substituents (NO2, NMe3+, COOEt, Cl) increase the oxidizing power of the FeIII-TAML catalysts with respect to Safranine-O (kII), but retard the intramolecular inactivation (ki). The most inactivation resistant catalyst is the nitro-substituted FeIII-TAML, 1d. It is more stable than 1a by a factor of six. This observation suggests that one of the vulnerable fragments of FeIII-TAML catalysts is the aromatic component. Presumably the active catalyst may destroy itself beginning via oxidative damage at this group. Electron-withdrawing substitutients should protect the ligand system from this type of damage. The slope of -1.0±0.3 is unusual because the slope of +1 might be anticipated with more reactive catalysts decomposing more rapidly.
Equation 13 can be rearranged to a form (eq 16) that describes the relative conversion of substrate, i.e. At/A0, as a function of time. Equation 16 provides convenient means for simulating the catalyst performance using the current model.
(16)
Thus, theoretical predictions can be compared with the experimental data. Additional support for the mechanistic concepts developed here would consist of a good match between the experimental and predicted kinetic data calculated using the obtained values of ki and kII at known concentrations of FeIII-TAML catalysts. Provided the corresponding concentration regime is chosen, the calculated and measured absorbance versus time plots should be similar. Such a comparison is shown in Figure 9, were the dynamics of 1a-catalyzed bleaching of Safranine-O by hydrogen peroxide performed at different 1a concentrations is plotted together with the calculated traces. There is excellent agreement between two experimental and calculated curves. The comparison proves that quantitative bleaching of the dye is in fact achieved just by increasing the 1a concentration. But even after the increase, the concentration is very low, specifically, 10-6 M when the difficult-to-oxidize Safranine-O is bleached. The approach described and justified herein is principally applicable to any two-substrate catalytic system provided the activation of the catalyst is faster than a target chemical reaction.
Figure 9. Normalized experimental and simulated bleaching of Safranine-O (4.3×10-5 M) by H2O2 (0.012 M) catalyzed by 1a at pH 11 and 25 °C. Experimental data are shown as and . The simulations, shown as solid lines, were made using the rate constants from Table 4.
In conclusion, these kinetic studies show that green catalytic oxidation technologies can be designed that are fully capable of supporting sustainable and economic chemical products and processes. Fe-TAML catalysts perform comparably to, if not better than, peroxidase enzymes, but they are much easier to make, study and control. FeIII-TAML catalysts begin to meet the great promise of green chemistry (the design of chemical products and processes that reduce or eliminate the use and generation of hazardous substances) in ways that matter to environmental integrity.8,20,21 The results of this work makes them even more attractive for use in technologies that send effluent streams to environmental media, because there is now a clear scenario of how the catalytic activity of the catalysts affects their half-lives so that process conditions can be set to ensure that no undegraded FeIII-TAML catalyst remains at the point of exit of the effluent stream from the plant. A substantial gain in reactivity is brought about by fluorination of the tail component of FeIII-TAMLs, but this is accompanied by a drop in the half-lives of these more reactive catalysts. Overall, the important process comparison is that the fluorinated catalysts work more rapidly. Suicide inactivation is most efficiently retarded by introducing electron-withdrawing groups at the aromatic ring in the head part of FeIII-TAMLs. Thus, knowing the rate constants ki and kII provides a shortcut for developing green oxidation technologies where one can ensure the catalysts will die in the plant when the intended work is done.
Objective 1-5: Catalysts at Work: Mixed Oxidative and Hydrolytic Activity [Publication 3]
We have studied degradation of three OP pesticides, namely, fenitrothion, parathion and chlorpyriphos methyl, by 1-H2O2. Fe-TAML (1a, 40 μM or 1c, 10 μM)-H2O2 (2.0 M) treatment of fenitrothion (1 mM) in water with 10% tBuOH as co-solvent in buffered pH (pH 8.0, 0.1 M KH2PO4/KOH) led to the immediate formation of a yellow color, λmax=396 nm in UV-Vis spectroscopy, characteristic of 3-methyl-4-nitrophenolate. This absorption band was seen diminishing soon after its formation, eventually nearly disappearing (>95% degradation, within 2 h. Independent treatment of 3-methyl-4-nitrophenolate with 1-H2O2 under the same conditions led to similar UV-Vis observations. The reaction kinetics followed by UV-Vis spectroscopy revealed that the rate of formation of 3-methyl-4-nitrophenolate from fenitrothion is considerably faster than its subsequent oxidative degradation. HPLC analysis confirmed the generation of 3-methyl-4-nitrophenolate as the major initial product (~90%) and its concomitant degradation. The formation of fenitrooxon as a minor product (~10%) was also observed. However, the presence of even small amounts of fenitrooxon was unacceptable due to its high mammalian toxicity (Oral LD50 rat 3.3 mg/kg). Moreover, while fenitrothion and 3-methyl-4-nitrophenolate were easily destroyed by 1-H2O2 under the reaction conditions, it had no observable effect on fenitrooxon even after 2 h.
Phosphate esters undergo more facile base hydrolysis than their thiophosphate analogs. The strongly nucleophilic hydroperoxide ion (HO2-), which forms at elevated pH (pKa of H2O2 ~11.5), can lead to accelerated hydrolysis in a process called perhydrolysis.22 We observed that the hydrolysis rate of fenitrooxon increased by almost 2-orders of magnitude upon raising the pH from 8.0 through 12.0. Separate kinetic studies revealed that the 1/H2O2 system attains maximum oxidative efficiency at around pH 10.0. Therefore, we chose to raise the pH to 10.0 for further study. The treatment of fenitrothion with 1/H2O2 at pH 10.0 under ambient conditions initially led to complete conversion to 3-methyl-4-nitrophenolate and fenitrooxon followed by 95-98% degradation of 3-methyl-4-nitrophenolate with no trace left of fenitrooxon after 2 h (Scheme 5) as determined by HPLC and GC-MS. In a control experiment, at pH 10.0, H2O2 alone led to 85-90% hydrolysis of fenitrothion, producing 3-methyl-4-nitrophenolate in 2 h with no formation of fenitrooxon. However, in this case, the reaction did not proceed to oxidative degradation of 3-methyl-4-nitrophenolate. The inability of H2O2 to oxidize 3-methyl-4-nitrophenolate clearly underscored that activation of H2O2 by 1 was critical in achieving the total degradation of fenitrothion. By raising the pH to 12.0, the H2O2 concentration could be reduced to 0.3 M (compared to 2.0 M earlier), as the effective HO2- ion concentration became significantly higher, accelerating the hydrolysis rate of fenitrooxon.
Scheme 5. Degradation of fenitrothion catalyzed by 1.
The degradation products of fenitrothion after treatment with 1c/H2O2 were identified and quantified using a variety of analytical techniques achieving ~85% mass balance. Four small aliphatic acids - formic, oxalic, maleic, and methyl maleic, were identified as major reaction products and TOC analysis showed 10±2% mineralization (CO2+CO). An aquatic toxicity study of the degradation mixture of fenitrothion was conducted by Microtox® assay. There was a 10-fold reduction in toxicity of the reaction mixture as compared to untreated fenitrothion in solution. Microtox® assays of different components of the reaction mixture showed that residual toxicity in the degradation of fenitrothion could be attributed solely to the tBuOH co-solvent. Treatment with 1-H2O2 also led to facile degradation of other widely used pesticides - parathion and chlorpyriphos methyl, including their corresponding hydrolysates - 4-nitrophenol and 3,5,6-trichloropyridin-2-ol. Degradation of parathion and chlorpyriphos methyl yielded maleic acid and chloromaleic acid, respectively. Oxalic acid and formic acid were also found as degradation products in both cases.
These studies establish that complete degradation of a series of important pesticides can be achieved using 1-H2O2 in a controlled, rapid, versatile and, based on aquatic toxicity assays, in an environmentally friendly manner. This catalytic oxidation approach could potentially be adopted for the safe and environmentally benign disposal/remediation of pesticides and their hydrolysates, including environmentally problematic nitrophenols and other hydroxyaromatics. One could envision the use of mobile reactors for use at obsolete pesticide storage sites reducing the capital costs associated with permanent installations, and the related costs and safety concerns associated with packaging, handling, and transportation of obsolete pesticides.
Objective 2-1. Buffer-ion Induced Demetalation of FeIII-TAML Activators in Neutral and Slightly Acidic Media [Submitted Publication 3]
Iron(III) complexes 1 that are known to be demetalated through specific acid catalysis (the proton) at pH < 3,4 are also a subject to demetalation at higher pHs acidic and basic pH in the range 4-9. The release of iron(III) from solutions of 1a is noticeable by uv-vis spectroscopy and visually as the loss of orange color of solution of 1a at pH ~7 and ambient temperature when phosphate concentrations >0.1 M are used for adjusting pH. The principal uv-vis band of 1a at 368 nm diminishes gradually. The spectral changes indicate demetalation of 1a to form the free ligand. This was confirmed by the 1H NMR analysis of the precipitate formed after running the demetalation at 0.08 M concentration of 1a in 0.5 M phosphate at pH 7 for 12 h. The spectrum of diamagnetic material corresponded to that of free ligand (TAML). There are aromatic (AA'BB' system at δ 7.50 and 7.35) and methyl signals (at δ 1.593 and 1.586).
The reactivity of an FeIII-TAML complex is controlled by the tail portion of the macrocyclic ligand. Compared were the dimethyl- (1a), cyclopropyl- (1b), and difluoro-tailed (1c) complexes under identical reaction conditions. Compounds 1b and 1c undergo essentially no decomposition in 30 min while 1a is demetalated in less than 15 min. Data do not allow for reliable measurement of kobs for much less reactive 1b and 1c without moving to the conditions, at which the specific acid catalysis dominates.4 Therefore 1a was used for collecting kinetic data at 35-65 °C.
Pseudo-first-order rate constants for the decay (kobs) are directly proportional to the phosphate concentration without noticeable intercept. Phosphate-independent demetalation is thus negligible here. The effective second-order rate constants k1,eff (subscript 1 denotes first order in acid) depend on pH. The general acid catalysis for demetalation is consistent with kobs increasing with increasing [phosphate] at constant pH and the effect is stronger at lower pH.17,23 A sigmoid profile of k1,eff as a function of pH with an inflection point ~6.5 suggest dihydrogen phosphate as a major species involved.
(17)
The data was quantified in terms of eq 17 that agrees with a mechanism shown in Scheme 6 (a1 = , b1 = Ka, the acidity constant for H2PO4-). The best-fit values for and Ka obtained at four temperatures are summarized in Table 5.
Scheme 6. Tentative mechanism for the phosphate-induced demetalation of FeIII-TAMLs.
The rate-limiting step of the mechanism in Scheme 6 is an intramolecular attack by phosphate after its binding in the first coordination sphere of 1.24 The pKa value found in this work at 45 °C agrees with that reported elsewhere for free dihydrogen phosphate (Table 5).25 A number of factors support weak phosphate binding. (i) The kobs does not level off at high phosphate concentrations. (ii) The λmax in the uv-vis spectrum of 1b does not change its position in the presence of 0.01, 0.3, and 0.5 M phosphate at pH 7.5, the absorbance intensity varying only negligibly. (iii) The EPR spectra of 1a in water and 0.01 M phosphate buffer show identical features.
Sulfite, acetic acid, pyridine, and tris(hydroxymethyl)aminomethane (TRIS) are other acids and bases with one ionogenic group have been tested. The effect of sulfite toward 1a was studied in the range 0.1-0.5 M at pH 5.25-6.75 and 65 °C. A fit of the kinetic data to eq 17 gave parameters a1 and b1 (Table 5), which could be assigned to and Ka, respectively. Similar behavior was found for acetic and malonic acids; the parameters a1 and b1 are shown in Table 5. There is a Brønsted correlation between loga1 and pKa (or -logb1) with the slope ca. -0.3, i.e. the activity of acids increases with increasing the acidity in accordance with the general acid catalysis. Pyridine and TRIS do not demetalate 1a at pH 4.5 and 65 °C in a matter of 1 h. This is not surprising for TRIS, since the pKa of the amino group is high, viz. 7.83.26
The chemistry of pyridine-X-carboxylic acids {PXC, X = 2 (picolinic), 3 (nicotinic), and 4 (iso-nicotinic)} in demetalation of 1 is intriguing. Similar to pyridine, P3C and P4C by themselves are inactive under a wide range of reaction conditions. In contrast, P2C demetalates complex 1a at 25 °C showing a behavior that is described for phosphate and sulfite, i.e. kobs increase linearly with increasing the concentration of P2C and the slopes (k1,eff) depend on pH; the data are in Table 5. Complex 1b is demetalated by picolinic acid at 65 °C but kobs does now depend on the P2C concentration power two.
Table 5. Parameters of eq 17, values of pKa, activation, and thermodynamic parameters for the demetalation of 1.
1 |
Catalyst |
T / ºC |
a1 / M-1 s-1 |
b1 (Ka) / M |
pKa |
Lit pKa25 a) |
1a |
H2PO4- |
35 |
(2.68±0.03)×10-3 |
(5.2±0.2)×10-7 |
6.29±0.02 |
6.57±0.05 |
45 |
(4.5±0.3)×10-3 b) |
(3.8±0.7)×10-7 c) |
6.42±0.09 |
|||
55 |
(7.8±0.3)×10-3 |
(12±1)×10-7 |
5.92±0.05 |
|||
65 |
(14.9±0.5)×10-3 |
(11.0±0.8)×10-7 |
5.96±0.03 |
|||
1a |
HSO3- |
65 |
(6±2)×10-2 |
(8±5)×10-6 |
5.1±0.6 |
6.34 |
1a |
H3CCOOH |
65 |
0.13±0.02 |
(6±1)×10-4 |
3.25±0.08 |
4.56±0.03 |
1a |
H2C(COOH)2 |
65 |
0.19±0.02 |
(4±1)×10-5 |
4.5±0.2 |
5.07±0.06 |
1a |
Me3CCOOH |
65 |
No catalysis |
|
|
4.83±0.01 |
1a |
Pyridine |
65 |
No catalysis |
|
|
5.229 |
1a |
P2C |
25 |
0.46±0.07 |
(3±1)×10-5 |
4.5± 0.2 |
5.17±0.01 |
1e |
P2C |
25 |
0.34±0.04 |
(2.6±0.8)×10-5 |
4.6±0.1 |
|
1f |
P2C |
25 |
0.20±0.01 |
(1.5±0.3)×10-5 |
4.83±0.06 |
|
1b |
P2C |
65 |
0.116±0.007 δ) |
(1.5±0.3)×10-5 |
4.9±0.1 |
|
1a |
P3C |
65 |
No catalysis |
|
|
4.81±0.03 |
1a |
P4C |
65 |
No catalysis |
|
|
4.87±0.04 |
1a |
(HOCH2)3CNH2 (Tris) |
65 |
No catalysis |
|
|
8.075 |
1a |
HN3 |
25 |
No catalysis |
|
|
4.65±0.02 |
kobs = k0,eff + k2,eff [P2C]2 (18)
Addition of the first order term k1,eff[P2C] to eq 18 did not improve the fit. The third-order rate constants k2,eff depend on pH. Upon fitting the experimental data to eq 17 the parameters a1 = 0.116±0.007 M-2 s-1 and and b1 = (1.5±0.3)×10-5 M were calculated. As it will be shown below, the parameter b1 corresponds to the Ka of picolinic acid. The pKa of 4.8±0.1 is comparable with the value of 5.17±0.02 reported elsewhere.25
Figure 10. Effect of nicotinic acid on demetalation of 1b in the presence of picolinic acid (0.01, 0.05, and 0.1 M) at pH 4.5 and 65 °C. Graphs in boxes on the right show anticipated dependencies of parameters A, B, and C of eq 22 on [P2C]m (m = 2 for A and 1 for B and C). See text for more details.
The reaction order in P2C of one for 1a and two for 1b deserves attention. If P2C acts without kinetically meaningful binding to iron(III) of 1, the second order could be rationalized by a self-association of P2C to produce a more acidic dimer. The acidity of one P2C is sufficient to break fragile 1a but is insufficient for breaking 1b. If one P2Cs just activates the other (or 1b, see below), accelerations of the catalysis by P2C should be observed in the presence of py, P3C, or P4C. This hypothesis finds experimental support. Nicotinic acid (Figure 10) or pyridine (data not shown) do speed up demetalation of 1b in the presence of P2C. A hyperbolic dependence suggests a mechanism that involves a pre-equilibrium binding of two pyridine carboxylates to FeIII of 1b followed by the intramolecular proton transfer from the coordinated acid (Scheme 7). This option was verified by measuring the binding constants for py and other ligands.
Scheme 7. General mechanism of demetalation of 1 by picolinic acid accounting for first (1a, in the box) and second (1b) orders in the acid. Charge of deprotonated carboxylate is only shown for clarity.
Scheme 7 contains several plausible intermediates in demetalation of 1a and 1b. Picolinic acid binds to the axial site of FeIII and then delivers the carboxylic group proton onto a Fe-N bond. This is more difficult in the case of 1b. Therefore the second py ligand increases the electron density at the Fe-N bonds. Particularly the α-carboxylic group is not essential; both py and P3C accelerate the deinsertion (Figure 10). This data is quantifiable in terms of the mechanism in Scheme 8. The reactive intermediates for robust 1b is the diaxially coordinated species, one of the two ligands being picolinic acid (L).
Scheme 8. Stoichiometric mechanism of demetalation of robust FeIII-TAMLs such as 1b by picolinic acid (L) in the presence of other pyridines (P). Axial aqua ligands are omitted for clarity.
Scheme 8 leads to eq 19 for the pseudo-first-order constant for demetalation of 1b at a given pH when the concentration of total iron (Mt) is much less than that of L and P.
kobs = (19)
In the absence of added P the experimentally observed second order in L (eq 18) implies that 1 >> KL[L] + KLK2L[L]2 and k2,eff = kLKLK2L. Therefore in the presence of P eq 19 appears in the form
kobs = (20)
The dependence of kobs on [P] is hyperbolic (Figure 10) and hence the term KPK2P[P]2 in the denominator is negligible. Consequently
kobs = (21)
At constant L the dependence of kobs on [P] is hyperbolic
kobs = (A + B[P])/(1 + C[P]). (22)
Here A = k2LKLK2L[L]2, B = kLPKLKLP[L], and C = KP + KLK2L[L]. Parameters A, B, and C of eq 22 were obtained by fitting the data in Figure 6 to eq 22 at five concentrations of P2C. Both B and C should be a linear function of [L], whereas A should be proportional to [L]2. Graphs in small boxes in Figure 10 show that such dependencies hold. The plots A and B versus [L] and [L]2, respectively, go through the origin. An intercept of the plot C against [L], which should equal KP, is close to 8 M-1 and falls into the range of K1 found.
Thus, acid catalyzed demetalation of FeIII-TAMLs 1 begins when a Brønsted acid coordinates to the iron(III) and delivers its proton in an intramolecular fashion via a favored six-membered ring to the proximity of an Fe-N bond resulting in its cleavage. This mechanism is consistent with the fact that picolinic acid catalyzes the demetalation, but nicotinic and isonicotinic acids do not. Acids such as H2PO4-, HSO3-, or CH3COOH have suitable structures to coordinate to the iron of FeIII-TAMLs via their O-atoms and deliver their protons intramolecularly to the vicinity of an Fe-N bond to promote protonolysis. The mechanism of demetalation discovered in this work shows how buffer components can be chosen based on their structures and acidities so that they will be inactive in the demetalation in the vicinity of neutral pH. Preferably, buffers chosen to avoid demetalation should not have pKa’s in the range 4-8, but if there is a particular advantage for any such buffer, the proton-bearing buffer ion should not have a structure that enables it to bind to the iron(III) of 1 and deliver a proton in an intramolecular fashion to a coordinated amidato-N atom, i.e., avoid 6- and probably also 5-membered rings for proton delivery, Scheme 1. The knowledge developed in this study is especially valuable for avoiding unfavorable catalyst degradation processes in functioning FeIII-TAML activator systems as well as for storing FeIII-TAML catalyst solutions. It also indicates another viable approach for deactivating certain FeIII-TAML activators to add to the well-parameterized oxidative deactivation routes, if such is needed, after they have been used in a process and before in an effluent stream would be released to the environment.
Objective 2-2. Micellar Tuning of the Reactivity of FeIII-TAML Activators [Manuscript in Preparation 5]
Three different types of surfactants – cationic (cetyltrimethylammonium bromide, CTAB), non-ionic (Triton X-100) and anionic (sodium dodecyl sulfate, SDS) – significantly and variably affect the reactivity of the Fe-TAML activators in the presence of hydrogen peroxide or tert-butylperoxide as primary oxidants. Significant accelerations of the catalysis have been achieved in micellar solutions CTAB. Micelles of Triton X-100 retard oxidative processes catalyzed by 1. Micelles of SDS usually affect the reactivity insignificantly, though a case have been disclosed when SDS eliminates completely the oxidative power of 1. All data obtained were interpreted in terms of eq 23, which describes the intrinsic reactivity of 1, and Scheme 9, which reflects interactions of the reagents and key intermediates with the micellar pseudo-phase. Correspondingly, the Berezin pseudo-phase model of micellar effects has been applied.27 Attempts have been made to analyze independently micellar effects on the rate constant kI, which describes the activation of 1 by peroxides, and kII, which corresponds to the bleaching itself, i.e. degradation of a dye by an oxidized Fe-TAML. Therefore the reaction conditions and dyes have been selected such as to make the kI or kII step the rate limiting, i.e. the bleaching has been run under two different kinetic regimes (Scheme 9).
Scheme 9. Mechanism of Fe-TAML catalysis in the presence of micelles.
Significant accelerations of the catalysis by FeIII-TAMLs have been achieved in micellar solutions of CTAB. The observed 2nd-order rate constant kI for the bleaching of Orange II by tBuOOH is plotted against concentrations of CTAB micelles in Figure 11. The rate constant increases rapidly to reach a maximum value and then declines slowly. This is due to the fact that the participants of bimolecular reaction (1a and tBuOOH) are strongly attracted to the surfactant micelles. The binding of both reactants to the CTAB micelles increases their local concentration in the micellar pseudophase. This accounts for an initial increase in kI. Further addition of CTAB dilutes the reactants in the micellar pseudophase, thus slowing down the reaction. Since FeIII-TAML activators are negatively charged at pH ≥ 10, they have an extra electrostatic affinity toward the cationic micelles of CTAB. The organic hydroperoxide tBuOOH is hydrophobic and therefore it is also attracted to the CTAB micelles.
Figure 11. Catalyzed by 1a bleaching of Orange II by tBuOOH: rate constant kI as a function of [CTAB]. Conditions: [1a] 4.8×10-7 M, [Orange II] 5.3×10-5 M, [tBuOOH] 3.5×10-4 M, pH 10 (0.1 M phosphate), 25 °C.
The rate expression for kI (kinetic regime kI in Scheme 9) is given by eq 23.
(23)
Here, V is the molar volume of CTAB in the micelle; CV and (1 - CV) are the fractions by volume of the micellar pseudo-phase and the aqueous phase respectively, kIm and kIaq are the respective bimolecular rate constants. The binding constants KFe(III) and KROOH are related to their partition coefficients (PFe(III) and PROOH, respectively) as: KFe(III) = (PFe(III) - 1)V and KROOH = (PROOH - 1)V. In dilute CTAB solutions ([CTAB] ≤ 0.01 M), whereby the volume fraction of the micellar pseudophase is small (i.e. CV << 1, as V 0.3 27), eq 23 becomes eq 24:
(24)
Here . The values for KFe, KROOH, kI,m, and kI,w in Table 6 were estimated by fitting the data in Figure 11 to eq 24.
Table 6. Equilibrium and rate constants in CTAB and Triton X-100 micelles with Orange II (kI regime) and Sudan III (kII regime) at pH 10 and 25 °C.
TAML/Dye |
Surfactant |
KFe(III)a (M-1) |
KFe(III)b (M-1) |
KBuOOHb (M-1) |
kIaqb (M-1s-1) |
kIm(b) (M-1s-1) |
||||||||
1a/Orange II |
CTAB |
2,500 ± 700 |
(10±7)×103 |
115 ± 7 |
240 ± 68.9 |
6.7 ± 0.9 |
||||||||
1a/Orange II |
Triton X-100 |
360±160 |
256±9 |
|||||||||||
1a/Sudan IIIc |
CTAB |
2,800 ± 700d |
190±50/170±30e |
(11.2±0.5)×104 |
1000±700 |
|||||||||
1g/Orange II |
CTAB |
6,900 ± 600 |
(18±5)×103 |
132 ± 46 |
1350 ± 200 |
8.2 ± 2.9 |
||||||||
1g/Orange II |
Triton X-100 |
500 ± 170 |
1310 ± 200 |
The data in Figure 12 demonstrate that the rate constant kI, decreases with increasing the concentrations of Triton X-100 throughout suggesting that the micellar pseudo-phase separates the reagents.27 There is only the partitioning of 1 between water and the micellar pseudo-phase, whereas the peroxides experience minimal or no affinity to the micelles. It is also seen from Figure 12 that kI approaches zero at higher concentrations of Triton X-100, suggesting that kIm ~ 0. Equation 24 transforms to eq 25 to show a simple descending hyperbolic dependence on concentration of Triton X-100.
(25)
The results in this study show that micelles of cationic surfactants, e.g. CTAB, are excellent reaction media for oxidation of water-insoluble substrates by the Fe-TAML/peroxide system. Cationic micelles accelerate both the formation of the reactive intermediate (determined by the rate constant kI) and oxidation of a substrate, e.g. Sudan III (determined by the rate constant kII) to an appreciable extent. Although no such acceleration is observed with non-ionic or anionic surfactants, the study provides a fair understanding of the nature of surfactant to choose for oxidation with the Fe-TAML/peroxide system.
Figure 12. Dependence of kI for 1a-catalyzed bleaching of Orange II with tBuOOH.
Objective 2-3. Mechanistic Understanding of the Very High Reactivity found for Dibenzoyl Peroxide Toward FeIII-TAML Activators [Publication 13]
The kinetics of oxidation of Orange II by hydrogen peroxide, benzoyl peroxide, tert-butyl and cumyl hydroperoxides catalyzed by 1 in aqueous solutions at pH 9-11 have been compared. The reaction leads to CO2, CO, phthalic acid and smaller aliphatic carboxylic acids as major mineralization products. The products are non-toxic according to the Daphnia magna test. Several organic intermediates have been identified by HPLC and GC-MS that allowed the detailed description of Orange II degradation. For all oxidants the catalyzed oxidation of Orange II follows the rate expression 6 with kIII ~ 0. The rate constant kI equals (74±3)×103, (1.4±0.1) ×103, 24±2, and 11±1 M-1 s-1 for benzoyl peroxide, H2O2, t-BuOOH, and cumyl hydroperoxide at pH 9 and 25 °C, respectively. An average value of kII equals (3.1±0.9)×104 M-1 s-1 under the same conditions.
An unexpected aspect of this study is that the highest rate constant kI is found for dibenzoyl peroxide. This molecule is significantly larger than H2O2, but the steric effect does not seem to be a factor. Assuming that the activation of hydroperoxides by FeIII-TAMLs 1 occurs as it has been suggested for peroxidases, i.e. involving the formation of the FeIII-O-OR intermediate followed by the heterolytic cleavage of the O-O bond,28,29 the reaction of benzoyl peroxide differs (Scheme 10).
Scheme 10. Plausible mechanism for activation of benzoyl peroxide by 1. For details, see text.
Benzoyl peroxide does not have an H-O fragment and therefore its coordination to iron(III) should involve either carbonyl or peroxo oxygen. The DFT calculations suggest that both oxygen atoms may coordinate to iron(III) because their effective negative charges are similar, i.e. -0.38 and -0.30 for carbonyl and peroxo oxygen, respectively. The peroxo oxygen (intermediate 6 in Scheme 10) seems to be a better candidate because the σ* orbital of the O-O fragment is thus much closer to the reducing FeIII center. The DFT calculations (in vacuum) support this mechanistic hypothesis. These have been done by the example of acetyl peroxide instead of benzoyl peroxide. If the carbonyl oxygen is a donor center, the complexation between 1 and benzoyl peroxide is a dead-end pathway, which does not result in the O-O bond cleavage. In contrast, the binding to form intermediate 6 is on the reaction coordinate and the following energy minimum is found for the system consisting of free benzoate and benzoate coordinated to formally FeV (or •+FeIV). Benzoate coordination to iron of TAML does not appear to occur in the aqueous solution because axial ligands are known to undergo ready hydrolysis.4 Ligation of iron of TAMLs by anionic ligands is difficult to achieve in water and therefore benzoate coordination is improbable. If formed, it will be rapidly hydrolyzed. A more realistic mechanism should involve two free benzoates and oxidized iron (Scheme 10). The formation of two benzoate anions requires two electrons. Two electrons could be moved concertedly or stepwise. The net result is the oxidation of FeIII-TAML and generation of two benzoate anions.
Objective 2-4. Quantifying the Reactivity of Transient Oxidized Form/s of Fe-TAML Activators Toward Various Electron Donors (Substrates) by Steady-State and Transient Kinetic Studies [Manuscript in preparation 4]
Stopped-flow kinetic studies of the oxidation of FeIII-TAML activators by tert-butyl hydroperoxide that affords a key intermediate of the FeIII-TAML-catalyzed oxidative cycle have helped to consolidate the mechanistic picture of the first oxidative step of the process. The data collected for this usually rate-limiting step using a series of substituted FeIII-TAMLs at pH 6.0-13.8 and 17-45 oC confirmed that it is first order in both participants. Bell-shaped pH-profiles of the effective second-order rate constants kI have maxima in the pH range 10.5-12.5 depending on the nature of the FeIII-TAML activator. Their "acidic" part is governed by the mono deprotonation of the axially diaqua form of FeIII-TAML and therefore an attachment of electron-withdrawing groups at both the "head" and "tail" of the catalyst broadens and moves to the neutral region the pH profile of the catalytic activity. Representative graphs are demonstrated in Figure 13 by the example of three complexes 1. These bell-shaped profiles are slightly shifted to the basic region as compared to similar graphs obtained for 1a and H2O2. This is an anticipated observation in terms of the reaction mechanism in Scheme 1 because pKa of tBuOOH is larger than that of H2O2. The right branch of the "bell" is controlled by the pKa of peroxide and this accounts for the shift. The left part is controlled by the acidity of 1 and this suggests a tool for widening the pH range where 1 show the highest reactivity in terms of kI. Particularly important is moving the highest reactivity to the neutral or slightly basic media. This is achieved by using complexes 1 with the lowest pKa such as 1d, the ring NO2 and tail F2 moieties of which increase its acidity and lower pKa1. The data in Figure 13 indicate that (i) the right branches of the curves are noticeably close; (ii) the left branches move progressively to acidic media as pKa's of 1 decrease; (iii) a broader pH profile for 1d is shifted most to the acidic conditions such that its reactivity is relatively large even at pH 9. An increase in electrophilicity of FeIII-TAMLs does not enhance the rate constants kI, which is rather unusual taking into account that iron(III) is getting oxidized during this step. Graphs such as in Figures 13 provide extra support for the mechanism of the oxidation of FeIII-TAMLs 1 by peroxides (Scheme 1). The data in Figure 13, as well as similar data for other complexes 1, have been fitted to eq 3 to estimate the corresponding equilibrium and rate constants, which are summarized in Table 7. The solid lines in Figure 13 are calculated curves using the equilibrium and rate constants from Table 7. There is an acceptable agreement between the experimental and calculated sets of data.
Figure 13 (left). pH profiles for the rate constants kI for the reactions of 1d (), 1h (), and 1a () with tBuOOH at 25 °C, 0.01 M phosphate.
Figure 14 (right). The Hammett plots for the rate constants k2 and k3. Effective Hammett constant 2(m+p) reflects two channels, by which the electronic effect from the head-ring substituents is delivered to the central metal.
The pathways driven by the rate constants k2 and k3 are kinetically indistinguishable. We argue that the k2 pathway is more realistic than k3 in view of the extremely high rate constants k3 observed for H2O2 (ca. 106 M-1 s-1). The rate constants k2 and k3 in Table 7 are calculated on assumptions that one or other of the pathways is negligible using one of the products, i.e. k2Ka1 or k3Ka2, in the numerator of eq 3. Formally this may mean that the properties of 1 have been considered twice or once, respectively, because Ka1 is also a characteristic of 1. Perhaps therefore the rate constants k3 are to a certain extent more independent and characterize better the nature of chemical interaction between 1 and tBuOOH. These considerations evolved after looking at the Hammett plots for the rate constants k2 and k3, which appear to be rather controversial (Figure 14). The two have different slopes. The plot is is positive for k3 (r = +0.7) versus very small and negative for k2 (r = -0.08). In other words, electrophilic properties of 1 are decisive in the k3 pathway whereas there is no clear trend for the k2 pathway. Presumably, these features arise because the kI step combines two conflicting phenomena. On one hand, the interaction between 1 and tBuOOH involves a Lewis acid-base pair that ends up in electron transfer, on the other.
Table 7. Rate constants (in M-1 s-1) and pKa values for the processes in Scheme 1 estimated in 0.01 M phosphate buffer by fitting experimental data such as in Figure 13 to eq 3.
Complex |
T / °C |
k1 |
103×k2 |
103×k3 |
k4 |
pKa1 |
pKa2 |
1a |
17 |
2 |
0.72 |
18 |
10 |
10.8 |
12.2 |
25 |
10 |
1.3 |
83 |
15 |
10.8 |
12.6 |
|
35 |
13 |
2.35 |
190 |
38 |
10.7 |
12.6 |
|
25 a) |
45 |
1.8 |
135 |
20 |
10.6 |
12.1 |
|
1h |
25 |
30 |
0.95 |
380 |
35 |
9.9 |
12.5 |
1g |
25 |
37 |
1.4 |
775 |
20 |
9.7 |
12.0 |
If this is a clear-cut acid-base interaction, an increase in electrophilicity of 1 should favor the reaction (cf. the trend in k3). If the electron transfer dominates, an increase in electrophilicity should slow down the electron transfer, since the reaction driving force decreases (cf. the trend in k2). Interestingly, the steps in Scheme 2 chemically agree with the observations obtained through the Hammett approach. In particular, the k3 step is the attack of negatively charged nucleophilic ligand at the iron(III) complex, which though negatively charged is an electrophilic species. Therefore, the value of ρ3 appears a correct one. The k2 step involves neutral tBuOOH and the iron(III) complex of charge –2 and hence its electrophilicity is significantly decreased. The classical metal-ligand interaction is compensated by the electron transfer in the transition state resulting in a lower value of ρ2. This analysis suggests that the modest discrimination between the k2 and k3 pathways on the basis of the absolute values of rate constants and comparing them with those for peroxidase enzymes may not be adequate; the rate constant for activation of H2O2 by 1 could be as high as 106 M-1 s-1.
Objective 2-5. Hexacloroiridate(IV) as a Probe for Studying Proprieties of Oxidized Fe-TAML Activators
The kinetic features of the oxidative catalysis by FeIII-TAML activators have much in common with those of the oxidative enzymes such as peroxidases. The similarities evolve after confirming the fact that oxidized TAMLs can also be generated from 1 and [IrIVCl6]2+, i.e. a strong oxidizing agent, which is used for making reactive Compounds I and II from the resting state of horseradish peroxidase. We have confirmed that addition of [IrIVCl6]2+ to an aqueous solution of 1a at pH 11.0 causes similar spectral changes as in the case of H2O2. There is an isosbestic point suggesting just two absorbing species in solution, 1a and its oxidized derivative (Figure 15). This observation opens the Job’s routine of continuous variation for exploring the stoichiometry of oxidation of 1.
Figure 15 (left). Spectra of complex 1a (1.2×10-4 M) in the absence (a) and in the presence of (1.2, 3, 6, 9, 12, and 15)×10-4 M [IrIVCl6]2-; pH 11, 25 °C.
Figure 16 (right). Job’s plot illustrating a 1:1 molar ratio on interaction between complex 1a and [IrIVCl6]2-; pH 11, 25 °C.
The Job’s plot in Figure 16 shows that 1a and IrIV react in a 1:1 molar ratio. All data obtained suggest that TAMLs oxidized in water are by 1 equiv above the resting state. As shown above, it is an FeIV derivative. Its exact formulation became possible after investigations using the Mössbauer spectroscopy. The evidence was found that at pH > 11 the oxidized TAML is a monomeric iron(IV) species. This allows to view the interaction between 1 and [IrIVCl6]2- as eq 26, estimate the equilibrium constant K1 from the data reported in Figure 15, and then, based on the known reduction potential for the half-reaction [IrIVCl6]2- + e- → [IrIIICl6]3-, calculate the reduction potential for the half-reaction [FeIV] + e- → [FeIII].
FeIII + IrIV FeIV + IrIII (26)
(27)
The data in Figure 15 was collected at comparable concentrations of 1a and IrIV. The mass balance with respect to the both participants was taken into accout, i.e. Fet = [FeIII] + [FeIV] and Irt = [IrIII] + [IrIV]. Only FeIV absorbs light at 777 nm. Therefore A775 = εFeIV[FeIV] and [FeIV] = A775/εFeIV. Concentrations of FeIV and IrIII formed are identical, i.e. [FeIV] = [IrIII]. Therefore [IrIV] = Irt - [FeIV] and [FeIII] = Fet - [FeIV]. Let [FeIV] = A775/εFeIV = x. Substitution of the concentrations into eq 27 as functions of x leads to the square equation 28.
(28)
Its solution affords the dependence of A775 as functions of Fet and Irt
(29)
Fitting the data in Figure 15 to eq 29 gives KI and εFeIV. The former is related to the reaction driving force ΔE (RT lnK1 = -nFΔE). Thus calculated reduction potential is ~0.9 V (NHE) and this accounts for huge oxidative power of FeIII-TAML activators of hydrogen peroxide.Objective 2-6. Moving Dioxygen Chemistry of FeIII-TAML Activators Into Water
A reverse micelle microreactor made of the AOT surfactant (bis(2-ethylhexyl) sodium sulfosuccinate) in n-heptane (Figure 17) has been shown to be efficient for the oxidation of Fe-TAML 1a by dioxygen. This is a remarkable finding because the activator is surrounded by water molecules and therefore should display the catalytic activity toward substrates, which are oxidized in aqueous solutions, not in organic media. Uv-vis spectrum of the material generated from 1a and O2 in the reverse micelles (Figure 18) is identical to that in Figure 2. This is evidence for the fact that water molecules surround the oxidized 1a.
Figure 17 (left). Schematic drawing of a reverse micelle incorporating FeIII-TAML.
Figure 18 (right). Uv-vis spectra of 1a (1.5×10-4 M) in 0.01 M, pH 10 carbonate buffer (a) and in the system of reverse micelles of Aerosol OT in n-heptane (b): 0.05 mL the same carbonate buffer with 1a, 8.25×10-4 mole AOT, 0.95 mL heptane.
Attempts were made to oxidize Orange II and ferrocene by dioxygen in the system of reverse micelles. A limitation was the low solubility of Orange II in n-heptane. The presence of AOT and water could make the dye more soluble in the system. However, in multiple experiments it was confirmed that Orange II precipitated out of solution at larger concentrations. The precipitate was not observed at lower concentrations. However, the registered spectral changes were irreproducible and did not allow us to conclude with confidence that O2 does oxidize Orange II in this system. This was a bit surprising because addition of a tiny amount of hydrogen peroxide into the microreactor initiated rapid and complete bleaching of Orange II. We also could not convert ferrocene into ferricenium cation by O2. The formation of the characteristic band around 600 nm typical of the cation was not detected. A tentative reason for inability of O2 to bleach the dyes is its low concentration in the microreactor. The problem can be solved we believe by in running the bleaching at elevated pressures of O2.
Objective 2-7. Design of the Mediated FeIII-TAML Based Oxidative Processes
A remarkable turnover in using mediators for the Fe-TAML/H2O2 reactions is found. Phenolic compounds such as 4-phenolsulfonic acid, 4-hydroxycinnamic acid, ferulic acid, and sinapinic acid, as well as the dye tartrazine (the major intermediate of which in the 1a-H2O2 decomposition is 4-phenolsulfonic acid) induce a Baeyer-Villiger oxidation of cyclobutanol. Cyclobutanol is completely inert to Fe-TAML/H2O2 alone but undergoes an oxidative ring-expansion reaction yielding δ-butyrolactone in the presence of phenolic mediators (eq 30).
(30)
Objective 2-8. Concluding Step: Mechanistically Inspired Design of New Catalysts [To be submitted manuscript 6]
The new TAML 1d has been prepared with electron-withdrawing groups in the "head” and "tail” parts of the molecule. The results described in Sections 1-1 and 1-4 suggested that it should be beneficial in terms of the catalytic activity (kI), resistance to the self-inactivation under the operating conditions (ki), the H-induced demetalation, and the optimal of its catalytic activity is most significantly shifted to the neutral region. The results shown in Figure 13 show that we in fact succeeded in preparation of the Fe-TAML catalyst, the catalytic activity of which is shifted to a neutral region. In addition to this, 1d is the most active and the resistant to suicidal inactivation made so far.
Additional Study: Destruction of estrogens using Fe-TAML/peroxide catalysis [Manuscript in press 17]
In this study, we demonstrated that Fe-TAML technology has outstanding ability to decompose the two major natural estrogens and their estrogenic decomposition products as well as 17α-ethinylestradiol (EE2), the active ingredient in the birth control pill Endocrine disrupting chemicals (EDCs) impair living organisms by interfering with hormonal processes controlling cellular development. Reduction of EDCs in water by an environmentally benign method is an important green chemistry goal. One EDC, 17α-ethinylestradiol (EE2), is excreted by humans to produce a major source of artificial environmental estrogenicity, which is incompletely removed by current technologies used by municipal wastewater treatment plants (MWTPs). Natural estrogens found in animal waste from concentrated animal feeding operations (CAFOs) can also increase estrogenic activity of surface waters. An iron-tetraamidomacrocyclic ligand (Fe-TAML) activator in trace concentrations activates hydrogen peroxide and was shown to rapidly degrade these natural and synthetic reproductive hormones found in agricultural and municipal effluent streams. Based on liquid chromatography tandem mass spectrometry, apparent half-lives for 17 α- and 17 β-estradiol, estriol, estrone and EE2 in the presence of Fe-TAML and hydrogen peroxide were approximately five minutes, and included a concomitant loss of estrogenic activity as established by E-Screen assay.
References
- Walling, C. Acc. Chem. Res. 1975, 8, 125-131.
- Halliwell, B.; Gutteridge, J. M. Arch. Biochem. Biophys. 1986, 246, 501-514.
- Halliwell, B.; Gutteridge, J. M. Methods Enzymol. 1990, 186, 1-85.
- Ghosh, A.; Ryabov, A. D.; Mayer, S. M.; Horner, D. C.; Prasuhn, D. E., Jr.; Sen Gupta, S.; Vuocolo, L.; Culver, C.; Hendrich, M. P.; Rickard, C. E. F.; Norman, R. E.; Horwitz, C. P.; Collins, T. J. J. Am. Chem. Soc. 2003, 125, 12378-12378.
- Jones, C. W. Applications of hydrogen peroxide and derivatives; The Royal Society of Chemistry: Cambridge, 1999.
- Lide, D. R., Ed. CRC Handbook of Chemistry and Physics; 87th ed.; CRC Taylor & Francis: New York, 2006-2007.
- Sen Gupta, S.; Stadler, M.; Noser, C. A.; Ghosh, A.; Steinhoff, B.; Lenoir, D.; Horwitz, C. P.; Schramm, K.-W.; Collins, T. J. Science 2002, 296, 326-328.
- Collins, T. J. Acc. Chem. Res. 2002, 35, 782-790.
- Ghosh, A.; Tiago de Oliveria, F.; Toshihiro Yano, T.; Nishioka, T.; Beach, E. S.; Kinoshita, I.; Münck, E.; Ryabov, A. D.; Horwits, C. P.; Collins, T. J. J. Am. Chem. Soc. 2005, 127, 2505-2513.
- Tiago de Oliveria, F.; Chanda, A.; Banerjee, D.; Shan, X.; Mondal, S.; Que, L., Jr.; Münck, E.; Bominaar, E. L.; Collins, T. J. Science 2007.
- Hodges, G. R.; Lindsay Smith, J. R.; Oakes, J. J. Chem. Soc., Perkin Trans. 2 1998, 617-628.
- Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J.; Vreven, J. A. T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; O. Kitao; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.; Gaussian 03 (release B.05). Gaussian, Inc., Wallingford CT, 2004. ed.: Wallingford CT, 2004.
- Chanda, A.; Popescu, D.-L.; Tiago de Oliveira, F.; Bominaar, E. L.; Ryabov, A. D.; Münck, E.; Collins, T. J. J. Inorg. Biochem. 2006, 100, 606-619.
- Hayashi, Y.; Yamazaki, I. J. Biol. Chem. 1979, 254, 9101-9106.
- Farhangrazi, Z. S.; Fosset, M. E.; Powers, L. S.; Ellins, W. R. Biochemistry 1995, 34, 2866-2871.
- Zipplies, M. F.; Lee, W. A.; Bruice, T. C. J. Am. Chem. Soc. 1986, 108, 4433-4445.
- Espenson, J. H. Chemical Kinetics and Reaction Mechanisms; 2 ed.; McGraw-Hill, Inc.: New York, 1995.
- Job, D.; Dunford, H. B. Eur. J. Biochem. 1976, 66, 607-614.
- Horwitz, C. P.; Fooksman, D. R.; Vuocolo, L. D.; Gordon-Wylie, S. W.; Cox, N. J.; Collins, T. J. J. Am. Chem. Soc. 1998, 120, 4867-4868.
- Ghosh, A.; Gupta, S. S.; Bartos, M. J.; Hangun, Y.; Vuocolo, L. δ. D.; Steinhoff, B. A.; Noser, C. A.; Horner, D.; Mayer, S.; Inderhees, K.; Horwitz, C. P.; Spatz, J.; Ryabov, A. D.; Mondal, S.; Collins, T. J. Pure Appl. Chem. 2001, 73, 113-118.
- Hangun, Y.; Alexandrova, L.; Khetan, S.; Horwitz, C.; Cugini, A.; Link, D. D.; Howard, B.; Collins, T. J. In Prepr. Am. Chem. Soc., Div. Petr. Chem., 2002; Vol. 47, pp 42-44.
- Wagner, G. W.; Yang, Y.-C. Ind. Eng. Chem. Res. 2002, 41, 1925-1928.
- Jencks, W. P. Catalysis in Chemistry and Enzymology; McGraw-Hill Book Company: New York, 1969.
- Gazzaz, H. A.; El-Guindi, N. M.; El-Awady, A. A. J. Chem. Soc., Dalton Trans. 1993, 2313-2320.
- Smith, R. M.; Martell, A. E. Critical Stability Constants, Vol. 2: Amines; Plenum Press: NY and London, 1975.
- Lewis, J. C. Anal. Biochem. 1966, 14, 495-496.
- Martinek, K.; Yatsimirsky, A. K.; Levashov, A. V.; Berezin, I. V. In Micellization, Solubilization, and Microemulsions; Mittal, K. L., Ed.; Plenum Press: New York, London, 1977; Vol. 2, pp 489-507.
- Dunford, H. B. Adv. Inorg. Biochem. 1982, 4, 41-80.
- Dunford, H. B. Heme Peroxidases; Wiley-VCH: New York, Chichester, Weinheim, 1999.
Journal Articles on this Report : 14 Displayed | Download in RIS Format
Other project views: | All 21 publications | 14 publications in selected types | All 14 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Banerjee D, Markley AL, Yano T, Ghosh A, Berget PB, Minkley Jr EG, Khetan SK, Collins TJ. "Green" oxidation catalysis for rapid deactivation of bacterial spores. Angewandte Chemie International Edition 2006;45(24):3974-3977. |
R832245 (Final) |
Exit Exit |
|
Chahbane N, Popescu D-L, Mitchell DA, Chanda A, Lenoir D, Ryabov AD, Schramm K-W, Collins TJ. FeIII–TAML-catalyzed green oxidative degradation of the azo dye Orange II by H2O2 and organic peroxides: products, toxicity, kinetics, and mechanisms. Green Chemistry 2007;9(1):49-57. |
R832245 (Final) |
Exit |
|
Chanda A, Popescu D-L, Tiago de Oliveira F, Bominaar EL, Ryabov AD, Munck E, Collins TJ. High-valent iron complexes with tetraamido macrocyclic ligands: structures, Mössbauer spectroscopy, and DFT calculations. Journal of Inorganic Biochemistry 2006;100(4):606-619. |
R832245 (Final) |
Exit |
|
Chanda A, Ryabov AD, Mondal S, Alexandrova L, Ghosh A, Hangun-Balkir Y, Horwitz CP, Collins TJ. Activity-stability parameterization of homogeneous green oxidation catalysts. Chemistry-A European Journal 2006;12(36):9336-9345. |
R832245 (Final) |
Exit |
|
Chanda A, Khetan SK, Banerjee D, Ghosh A, Collins TJ. Total degradation of fenitrothion and other organophosphorus pesticides by catalytic oxidation employing Fe-TAML peroxide activators. Journal of the American Chemical Society 2006;128(37):12058-12059. |
R832245 (Final) |
Exit |
|
Chanda A, Shan X, Chakrabarti M, Ellis WC, Popescu DL, Tiago de Oliveira F, Wang D, Que Jr. L, Collins TJ, Munck E, Bominaar EL. (TAML)FeIV=O complex in aqueous solution: synthesis and spectroscopic and computational characterization. Inorganic Chemistry 2008;47(9):3669-3678. |
R832245 (Final) |
Exit Exit Exit |
|
Collins TJ, Walter C. Little green molecules. Scientific American 2006;294(3):82-90. |
R832245 (Final) |
Exit Exit |
|
Ghosh A, Mitchell DA, Chanda A, Ryabov AD, Popescu DL, Upham EC, Collins GJ, Collins TJ. Catalase-peroxidase activity of iron(III)-TAML activators of hydrogen peroxide. Journal of the American Chemical Society 2008;130(45):15116-15126. |
R832245 (Final) |
Exit |
|
Khetan SK, Collins TJ. Human pharmaceuticals in the aquatic environment:a challenge to Green Chemistry. Chemical Reviews 2007;107(6):2319-2364. |
R832245 (Final) |
Exit Exit Exit |
|
Polshin V, Popescu D-L, Fischer A, Chanda A, Horner DC, Beach ES, Henry J, Qian Y-L, Horwitz CP, Lente G, Fabian I, Munck E, Bominaar EL, Ryabov AD, Collins TJ. Attaining control by design over the hydrolytic stability of Fe-TAML oxidation catalysts. Journal of the American Chemical Society 2008;130(13):4497-4506. |
R832245 (Final) |
Exit Exit Exit |
|
Popescu DL, Vrabel M, Brausam A, Madsen P, Lente G, Fabian I, Ryabov AD, van Eldik R, Collins TJ. Thermodynamic, electrochemical, high-pressure kinetic, and mechanistic studies of the formation of oxo FeIV-TAML species in water. Inorganic Chemistry 2010;49(24):11439-11448. |
R832245 (Final) |
Exit |
|
Popescu D-L, Chanda A, Stadler M, Tiago de Oliveira F, Ryabov AD, Munck E, Bominaar EL, Collins TJ. High-valent first-row transition-metal complexes of tetraamido (4N) and diamidodialkoxido or diamidophenolato (2N/2O) ligands: synthesis, structure, and magnetochemistry. Coordination Chemistry Reviews 2008;252(18-20):2050-2071. |
R832245 (Final) |
Exit |
|
Shappell NW, Vrabel MA, Madsen PJ, Harrington G, Billey LO, Hakk H, Larsen GL, Beach ES, Horwitz CP, Ro K, Hunt PG, Collins TJ. Destruction of estrogens using Fe-TAML/peroxide catalysis. Environmental Science & Technology 2008;42(4):1296-1300. |
R832245 (Final) |
Exit |
|
Tiago de Oliveira F, Chanda A, Banerjee D, Shan X, Mondal S, Que Jr. L, Bominaar EL, Munck E, Collins TJ. Chemical and spectroscopic evidence for an FeV-oxo complex. Science 2007;315(5813):835-838. |
R832245 (Final) |
Exit |
Supplemental Keywords:
Media: water, drinking water. Risk Assessment: risk, human health, bioavailability. Chemicals, Toxics, Toxic Substances: chemicals, toxics, metals, heavy metals, solvents, oxidants, organics, effluent, discharge, intermediates. Ecosystem Protection: aquatic. Risk Management: alternatives, sustainable development, clean technologies, innovative technology, remediation, cleanup, oxidation. Public Policy: public good. Scientific Disciplines: environmental chemistry,, Scientific Discipline, Water, Analytical Chemistry, Engineering, Chemistry, & Physics, homeland security, decontamination, TAML catalysts, catalytic oxidation, biological warfare agents, catalyst formulations, environmental chemistry, peroxide activators, green chemistryRelevant Websites:
http://www.chem.cmu.edu/groups/Collins/ Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.