Article pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Lewis Acid Assisted Nitrate Reduction with Biomimetic Molybdenum Oxotransferase Complex Lee Taylor Elrod and Eunsuk Kim* Department of Chemistry, Brown University, Providence, Rhode Island 02912, United States S Supporting Information *
ABSTRACT: The reduction of nitrate (NO3−) to nitrite (NO2−) is of significant biological and environmental importance. While MoIV(O) and MoVI(O)2 complexes that mimic the active site structure of nitrate reducing enzymes are prevalent, few of these model complexes can reduce nitrate to nitrite through oxygen atom transfer (OAT) chemistry. We present a novel strategy to induce nitrate reduction chemistry of a previously known catalyst MoIV(O)(SN)2 (2), where SN = bis(4-tert-butylphenyl)-2-pyridylmethanethiolate, that is otherwise incapable of achieving OAT with nitrate. Addition of nitrate with the Lewis acid Sc(OTf)3 (OTf = trifluoromethanesulfonate) to 2 results in an immediate and clean conversion of 2 to MoVI(O)2(SN)2 (1). The Lewis acid additive further reacts with the OAT product, nitrite, to form N2O and O2. This work highlights the ability of Sc3+ additives to expand the reactivity scope of an existing MoIV(O) complex together with which Sc3+ can convert nitrate to stable gaseous molecules.
1. INTRODUCTION Nitrate reduction to nitrite is of significant biological and environmental importance. Due to widespread use of nitrate in agricultural applications and high solubility in water, it is a pervasive contaminant in groundwater and can lead to eutrophication.1 Nitrate consumed can be reduced to nitrite by nitrate reducing bacteria in the mouth and gut, and further reduced to NO in the body, and is linked to negative health effects including the formation of reactive N-nitroso compounds associated with cancer and the blood disorder methemoglobinemia.2,3 Nitrate reductases found in prokaryotic and eukaryotic microorganisms catalyze the reduction of nitrate to nitrite for metabolic processes, and play an important role in denitrification and nitrate assimilation steps within the global nitrogen cycle.2,4 One of the well-studied nitrate reductases is periplasmic nitrate reductase (Nap) from Desulfovibrio desulf uricans that belongs to the dimethyl sulfoxide reductase (DMSOR) family of enzymes. The active site of Nap for the reduction of nitrate to nitrite contains a MoIV/VI metal center bound by two pyranopterin cofactors (Figure 1).5 The two-electron reduction of nitrate (NO3−) to nitrite (NO2−) is coupled with oxygen atom transfer (OAT) from NO3− to MoIV to form the MoVI− oxo species which subsequently releases water upon the addition of two external electrons and protons (eq 1).6−8 Formate dehydrogenase (Fdh), another DMSOR family enzyme which catalyzes the oxidation of formate to CO2 and has an active site structure very similar to that of Nap (Figure 1), has also been shown to be capable of reducing nitrate to nitrite.9 © XXXX American Chemical Society
Figure 1. Active site structures of the oxidized forms of periplasmic nitrate reductase (Nap) from D. desulf uricans5 and formate dehydrogenases (Fdhs) from Escherichia coli (X = SeCys)10 or Rhodobacter capsulatus (X = Cys).11
NO3− + 2H+ + 2e− → NO2− + H 2O
(1)
Artificial systems that remediate nitrate contamination are desirable to combat against groundwater contamination associated with increased use of fertilizers and global industrialization. Bioinspired catalysts could play an important role in the reduction of nitrate to nitrite in the remediation processes. There are a number of biomimetic model complexes that replicate the active site structure and reactivity of DMSOR, utilizing a variety of ligand scaffolds including dithiolenes,12,13 tris(pyrazolyl)borates,14,15 salan,16 pyridylmethanethiolate,17,18 and Schiff base derivatives.19 However, few are known to reduce nitrate to nitrite.20−24 Binuclear (μ-oxo)molybdenum(V) complexes reported by Holm and Young have demonstrated nitrate reduction capability.20−22,25 Catalytic nitrate Received: November 28, 2017
A
DOI: 10.1021/acs.inorgchem.7b02956 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
photosystem II (PSII).40 While there is not complete consensus on the role of Ca2+ in the oxidation of water, its presence is essential to the observed reactivity of OEC.41,42 Interest in PSII and the role of Ca2+ in OEC has led to the investigation of the effects of redox-inactive Lewis acid metal cations on reactivity of metal−oxo species.43,44 Redox-inactive Lewis acid additives have been shown to alter redox potentials,45−48 electron transfer rates,45,49−51 and reactivity47,52−57 of metal oxo complexes. Oxygen atom transfer (OAT) reactivity can be influenced by the addition of Lewis acids. The rate of OAT reactivity of Goldberg’s MnV(O) porphyrinoid complexes is dramatically decreased with the addition of Zn2+.47 Another MnV(O) system reported by Nam demonstrates the ability of Sc3+ to enhance OAT activity by increasing the oxidizing power of MnV(O) through binding to the ligand.48 A biomimetic complex MoIV(O)(SN)2 (2), where SN = bis(4-tert-butylphenyl)-2-pyridylmethanethiolate), reported by the research group of Holm17 is a very well documented system that facilitates efficient OAT from an organic/inorganic substrate to yield MoVI(O)2(SN)2 (1) (Figure 2). Complex 2
reduction to nitrite coupled to triphenylphosphine oxidation using a MoIV dithiolene complex was reported by Sarkar.24 Nitrate reduction by a W(IV) bis(dithiolene) complex containing a sterically encumbered axial sulfido ligand was reported by Holm and co-workers.23 One of the reasons very few complexes out of many structural biomimetic models can replicate the activity of nitrate reductases might be the absence of secondary coordination environment found in the metalloenzyme. Recent studies by Moura and Cerqueira emphasize that the small differences in the amino acid residues in the secondary coordination spheres dictate the type of catalytic reactions between Fdh and Nap who share remarkably similar structures.26,27 A recent study by Leimkühler showed that the highly conserved, positively charged arginine residue in the secondary coordination environment of Fdh and Nap is important in nitrate reduction by facilitating proper binding and stabilization of the substrate during the catalytic cycle (Scheme 1).9 Scheme 1. Nitrate Reduction by R. capsulatus Fdha
Figure 2. MoVI(O)2(SN)2 (1) and MoIV(O)(SN)2 (2) (SN = bis(4tert-butylphenyl)-2-pyridylmethanethiolate).17
a
mediates OAT chemistry with a wide variety of substrates including various N- and S-oxides.17 However, the original report indicated that nitrate reduction to nitrite by 2 was unsuccessful. Inspired by recent reports demonstrating the ability of redox-inactive metal cations to alter the OAT reactivity of MnV(O) systems,47,48 we investigated the effect of Lewis acid additives on the OAT activity of 2. Herein, we report a strategy to induce novel nitrate reduction with 2 through addition of Sc(OTf)3 (OTf = trifluoromethanesulfonate), demonstrating the ability of Lewis acid additives to expand the reactivity scope of previously developed metal−oxo systems.
Adapted from ref 9.
The importance of the secondary coordination environment in enzyme catalysis is not limited to Nap and Fdh. Noncovalent interactions in the secondary coordination sphere of metalloenzymes, such as H-bonding networks present in cytochrome P450,28 horseradish peroxidase,29 and hemoglobin,30 play a critical role in the chemical transformations they take part in. Strategies to incorporate noncovalent interactions into synthetic model complexes through ligand design have been developed in recognition of the importance of the secondary coordination sphere.31−35 Seminal examples include Collman’s picket fence porphyrins36,37 and Borovik’s non-heme iron scaffold bearing an amide microenvironment.38 More recently, the research group of Fout applied this strategy to successfully achieve nitrate reduction with a bioinspired iron catalyst.39 Addition of redox inactive Lewis acid metal cations can offer an alternative approach to influence reaction environment without directly tailoring the ligand backbone. Nature employs redox-inactive metals in conjunction with high valent metal− oxo species in metalloenzymes, such as the Mn4CaO4 cluster found in the oxygen evolving complex (OEC) found in
2. RESULTS AND DISCUSSION 2.1. Preparation of Mo(O)2(SN)2 (1) and Mo(O)(SN)2 (2). Complex Mo(O)2(SN)2 (1) was prepared from Mo(O)2(acac)2 (acac = acetylacetonate) and LiSN (SN = bis(4tert-butylphenyl)-2-pyridylmethanethiolate) following the reported procedure17 with minor modifications. Additional recrystallization from THF/pentane was necessary to obtain the pure yellow microcrystalline solid in 85% yield. The IR (νMoO = 901, 936 cm−1) (Figure S2) and diamagnetic 1H NMR spectral features (Figure S3) are in good agreement with the previously published values.17 Complex Mo(O)(SN)2 (2) was prepared from Mo(O)2(SN)2 (1) using excess triphenylphosphine in place of previously used triethylphosphine17 and was isolated as a dark brown solid in 64% yield. Additional recrystallizations from CH2Cl2/pentane afforded brown/black crystalline material and were found to aid the long-term stability of 2 at −35 °C under N2. The IR spectral features B
DOI: 10.1021/acs.inorgchem.7b02956 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry including a characteristic νMoO frequency at 947 cm−1 (Figure S2) and well resolved diamagnetic 1H NMR features of 2 (Figure S4) matched the previously published values (Figure S2), with minor discrepancies in the chemical shifts probably due to different instruments employed. The UV−vis absorption spectra of 1 (λmax = 370 nm) and 2 (λmax = 328 nm), originally reported in DMF, are nearly identical when dichloromethane is used as the solvent (Figure S1). Elemental analysis was performed on 1 and 2 to ensure purity of the bulk material. 1H NMR samples of 1 and 2 in CDCl3 indicate that 0.6 and 0.5 equiv of dichloromethane from recrystallization are present in crystalline samples of 1 and 2, respectively, and are accounted for in the elemental analysis. 2.2. Lewis Acid Requirement for Nitrate Reduction. Complex Mo(O)(SN)2 (2) carries out an efficient OAT reaction with a record number of substrates.17 However, there is a limit in the substrate scope. Holm and co-workers reported that the reaction between 2 and nitrate does not yield 1 without providing further experimental details and characterization. With fresh insights on the substrate activation mechanisms for Fdh and Nap (section 1), we thought that altering the secondary coordination environment by Lewis acid may induce previously unseen OAT activity with 2. The OAT activity with nitrate can be readily examined by UV−vis spectroscopy from pronounced spectroscopic differences of 1 and 2 (Figure S1). Consistent with the previous report,17 we did not observe the formation of 1 from the reaction of Mo(O)(SN)2 (2) and [Bu4N][NO3] at room temperature in CH2Cl2 up to 24 h (Figure S5, Scheme 2). The reactivity of 2 and nitrate is
reactivity from NO3− to 2 with other Lewis acids such as triphenylborane and Ca(OTf)2 were unsuccessful. No intermediates were observed for the novel OAT reactivity even at low temperatures. Instead, monitoring the reaction by UV−vis spectroscopy at −40 °C shows a clean conversion of Mo(O)(SN)2 (2) to Mo(O)2(SN)2 (1) with an ∼88% conversion yield (Figure 3). Addition of Sc(OTf)3 (2 equiv)
Figure 3. Reaction of Mo(O)(SN)2 (2) (0.8 mM), [Bu4N][NO3] (8.0 mM), and Sc(OTf)3 (1.7 mM) followed by UV−vis spectroscopy at −40 °C in dichloromethane for 1.5 h. The spectral changes correspond to the conversion of Mo(O)(SN)2 (2) (λmax = 328, 430 nm) to Mo(O)2(SN)2 (1) (λmax = 370 nm).
Scheme 2. Reactivity of Mo(O)(SN)2 (2) and [Bu4N][NO3] in the Absence (Top) and Presence (Bottom) of Sc(OTf)3a
a
to the dichloromethane solution of 2 (1 equiv) and [Bu4N][NO3] (10 equiv) results in the loss of the absorption bands 328 and 430 nm and the emergence of a single absorption band at 370 nm with isosbestic points at 340 and 407 nm resulting from the formation of Mo(O)2(SN)2 (1) (Figure 3). The observed isosbestic points match those reported for the OAT from Ph3AsO to 2 forming 1 at room temperature.17 2.3. Nitrate as Oxygen Atom Source: 18O Labeling Experiments. Isotopic labeling studies were conducted to confirm nitrate as the oxygen atom source for the observed conversion of Mo(O)(SN)2 (2) to Mo(O)2(SN)2 (1). Attempts at monitoring the reaction using N18O3− were unsuccessful due to insufficient purity of commercially available 18 O-labeled nitrate. As an alternative approach, we have prepared 18O labeled 2 for the reaction with NO3−. Doubly 18 O labeled Mo(18O)2(SN)2 (118/18) was prepared through the treatment of 1 with H218O by adopting a known procedure.17 Upon 18O substitution, the molybdenum oxygen stretches were shifted from 901 and 936 cm−1 to 887 and 858 cm−1 in the IR spectra, consistent with the known report17 and in good agreement with theoretical values from reduced mass calculations (890 and 856 cm−1) (Figure 4). As reported for the synthesis of 2 (see above), the 18O labeled 2, Mo(18O)(SN)2 (218), was prepared from 118/18 and triphenylphosphine. Treatment of Mo(18O)(SN)2 (218) with [Bu4N][NO3] in the presence of Sc(OTf)3 was accompanied by the brown to yellow color change observed for the analogous unlabeled reaction. The IR spectrum of the recrystallized reaction product shows major molybdenum−oxo stretches at 867 and 924 cm−1 in between the values for 116/16 and 118/18, suggesting the formation of the mixed labeled dioxo species Mo-
SN = Bis(4-t-butylphenyl)-2-pyridylmethanethiolate.
remarkably altered with the addition of Lewis acid (Scheme 2). Addition of Sc(OTf)3 (2 equiv) to a CH2Cl2 solution of 2 (1 equiv) and [Bu4N][NO3] (10 equiv) results in a rapid color change from brown to bright yellow. The UV−vis spectrum of the reaction mixture contains a single absorption band at 370 nm and matches that of authentic Mo(O)2(SN)2 (1) (Figure S6). Complex 1 generated from the Sc3+ assisted OAT from nitrate by 2 was isolated in good yields (80%) following recrystallization from dichloromethane/pentane after the removal of byproducts with MeOH. The formation of 1 was further confirmed by IR (Figure S7) and 1H NMR spectroscopy (Figure S8). The generation of nitrite (NO2−), the other OAT product, was confirmed through positive Griess reagent tests58 on the MeOH soluble material (see section 2.4). Use of 1 equiv of Sc(OTf)3 in the nitrate reduction with 2 resulted in lower conversion yield (∼50%) to 1 without generating any intermediates or byproducts. Use of 0.5 equiv of Sc(OTf)3 resulted in no OAT reaction. Attempts to promote the OAT C
DOI: 10.1021/acs.inorgchem.7b02956 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
[15NO2] were independently prepared (see section 4) due to the insolubility of Na[15NO3] and Na[15NO2] in CD2Cl2. The formation or consumption of nitrate or nitrite can be readily identified by the 15N chemical shifts for [Bu4N][15NO3] at 6.46 ppm and [Bu4N][15NO2] at 243.67 ppm in CD2Cl2. There was no reaction between [Bu4N][15NO3] and Sc(OTf)3 (1:1 ratio), from which only unreacted nitrate signal at 6.46 ppm was observed in the 15N NMR spectrum (data not shown). Likewise, when Mo(O)(SN)2 (2) was reacted with [Bu4N][15NO3] (1−10 equiv) in the absence of Sc(OTf)3, only the unreacted starting reagents, 2 and [Bu4N][15NO3], were observed in 15N and 1H NMR, IR, and UV−vis spectroscopy. When the reaction of 2 (1 equiv) and [Bu4N][15NO3] (10 equiv) was carried out in the presence of Sc(OTf)3 (2 equiv) in a sealed J Young NMR tube, the clean formation of the O atom abstracted metal product, Mo(O)2(SN)2 (1), was detected by 1 H NMR spectroscopy, which was further confirmed by UV− vis and IR spectroscopy. However, the 15N NMR spectrum of the reaction products did not show a signal for 15NO2− nor for any other 15N-containing products beside excess unreacted substrate, 15NO3−. (Figure 5a).
Figure 4. IR spectra (KBr) of Mo(16O)2(SN)2 (1, black dashed) and Mo(18O)2(SN)2 (118/18, black solid), along with Mo(16/18O)2(SN)2 (118/16) generated from 218 with [Bu4N][NO3] in the presence of Sc(OTf)3 (red solid), and Mo(16/18O)2(SN)2 (118/16) generated from 218 with trimethylamine n-oxide (blue dotted).
(18/16O)2(SN)2 (118/16) (Figure 4). Holm and co-workers reported that 218 reacts with Ph2SO to yield the mixed labeled bis-oxo compound 118/16 that was characterized by mass spectrometry,17 but the IR features of 118/16 were not reported. In order to confirm that the product generated from 218/ NO3−1/Sc(OTf)3 is the same O atom transferred product obtained from 218 with other known substrates, the reaction of 218 with trimethylamine n-oxide (TMAO) was carried out. The product obtained from 218/TMAO resulted in the identical IR spectrum as the one from 218/NO3−1/Sc(OTf)3 (Figure 4), indicating that nitrate is the source of the oxygen atom to generate Mo(18/16O)2(SN)2 (118/16) from Mo(18O)(SN)2 (218) (Scheme 3). Scheme 3. Mo(18/16O)2(SN)2 (118/16) Preparation from Mo(18O)(SN)2 (218) via Sc(OTf)3 Assisted Nitrate Reduction (I) or Trimethylamine n-Oxide (II)
Figure 5. Room temperature 15N NMR (in CD2Cl2) spectra of (a) the reaction mixture of Mo(O)(SN)2 (2) (1 equiv), Sc(OTf)3 (2 equiv), and Bu4N(15NO3) (10 equiv) showing excess nitrate signal at 6.46 ppm, (b) authentic [Bu4N][15NO2], and (c) an equimolar mixture of [Bu4N][15NO2] and Sc(OTf)3.
2.4. Fate of Nitrate. The Lewis acid assisted OAT from nitrate by Mo(O)(SN)2 (2) would generate nitrite (NO2−) as a reaction product. The formation of nitrite was first probed by the Griess reagent test,58 the most commonly used method of detection of nitrite. All the reaction products and byproducts excluding 1 were extracted with MeOH, on which the Griess test was conducted. Treatment of the MeOH soluble material with the Griess reagent resulted in the formation of the azo dye with an absorbance band at 548 nm (Figure S9) indicative of the presence of nitrite. However, only a trace amount (
