Rex E. Shepherd

Chromatographic and related electrophoretic methods in the separation of transition metal complexes or their ligands

Abstract It is difficult for most inorganic chemists to be aware of the availability of the many new techniques, described largely in analytical chemistry, environmental chemistry, and biomedical journals, which are increasingly of value in the separation and characterization of complexes of transition metals. Additionally, the design and use of supported transition metal complexes as sites for molecular recognition in the separation of biomolecules and organic materials has the potential to dominate multi-billion dollar industries in proteins and chiral drugs. This review serves as a roadmap for inorganic chemists to the recent chromatographic literature. The review covers advances in separation methods that involve transition metal chemistry which have occurred in the decade of 1992 through early 2003 with 360 references. The review is intended to assist readers in finding key papers that illustrate techniques of chromatography that might be applicable to purposes in the reader’s laboratory. Covered topics include the standard separation of inorganic ions and metal complexes, capillary electrophoresis methods (CE, CZE, MEKC), electrochemical detection in flow methods by [Ru(bpy) 3 ] 3+/2+ cycling in response to analytes, separations of metal complexes of interest to environmental and biomedical disciplines via size exclusion chromatography (SEC), and detection methods with electrospray mass spectrometry (ESI-MS). Advances in affinity chromatography in the separation of peptide, proteins and biopolymers (IMAC) and of organic substrates (IMCOS) are discussed. Recent advances in understanding the mechanisms of chromatographic separations, and of the technique of polymer imprinting to produce selective recognition sites for metal ions and metal complexes are described.

Reconsidered mechanism in RuIII(hedta)-catalyzed epoxidation of stilbenes

Abstract The cis -stilbene oxide/ trans -stilbene oxide isomer assignment given formerly in Inorg. Chim. Acta, 193 (1992) 217 should be reversed. This assignment was based on the 1 H NMR spectra of the stilbene oxides given in The Varian High Resolution NMR Spectral Catalog , Vol. 2, which is in error. The large amount of trans -oxide versus cis -oxide (5.5:1) from the cis -stilbene epoxidation by [Ru III (hedta)]/t-BuOOH, and 100% trans -oxide from trans - stilbene, is reinterpreted as indicative of a radicaloid intermediate in ⩾85% of the reaction channels. Comparisons are made with 15 ruthenium oxo catalysts which epoxidize stilbenes. Two categories are observed which are (A) stereoretentive Ru IV O catalysts or sterically hindered Ru VI O porphyrin oxidants, and (B) those which give isomer mixtures for Z -olefins and which are less hindered Ru V O and Ru IV O catalysts. The mechanistics aspects of these epoxidations are discussed.

[Ru2II(ttha)(H2O)2]2− is a rapid No scavenger (ttha6− = triethylenetetraminehexaacetate)

Abstract The reaction of NO(aq.) with [Ru 2 II (ttha)(H 2 O) 2 ] 2− = (A) and [Ru 2 II (ttha) (bpy) (H 2 O)] 2− = (B), (ttha 6− = triethylenetetraminehexaacetate; bpy = 2.2′-bipyridine) was monitored by electrochemical methods (cyclic voltammetry, differential pulse polarography). Each of two sites of [Ru 2 II (ttha)] may be represented by [LRu] for convenience. Waves for the [LRu II (NO·)] → −1e− [LRu II (NO + )], 2[LRu II (NO 2 − )] → −2e− [LRu III (NO 3 − )] + [LRu III (NO + )], and [LRu II (NO + )] → −1e− [LRu III (NO + )] oxidations were identified at −0.06, +0.98, and 1.14 V vs normal hydrogen electrode, respectively. The protonated form of [LRu II (NO·)], [LRu II (NOH)], was detected as a separate Ru II/III wave at +0.10 V. The pK a of [LRu II (NOH)] is 1.80. The rate of substitution of NO on (A) is 22.7 M −1 s −1 at 22°C indicating a normal rate of neutral ligand substitution on Ru II -polyaminopolycarboxylates via dissociative intermediates. The identical waves for the nitrosylated (B) indicate that the nitrosyls of (A), [(Ru II (NO)·)) 2 (ttha)] 2− = (C), are terminally coordinated rather than a single bridged nitrosyl. The nitrosyl (C) does not react with H 2 , precluding a catalytic scavenging of NO by (A) followed by H 2 reduction for environmental control purposes. However, the nitrosyl (C) is robust and dissociates very slowly under Ar purging. Thus the parent complex [Ru 2 (ttha)(H 2 O) 2 ] 2− and related mononuclear Ru II -polyaminopolycarboxylates such as [Ru(hedta)(H 2 O)] − have several features that lend them toward uses as antisepsis agents for the control of septic shock. [Ru 2 III (ttha) (H 2 O) 2 ] also reacts directly with NO, forming [(Ru II (NO + )) 2 (ttha)] which exhibits the same waves as [Ru 2 II (ttha)(NO) 2 ] 2− since the Ru II (NO + ) units readily reduce electrochemically at glassy carbon to the Ru II (NO·) complex (C) below −0.06 V. It was observed that the [LRu III (NO + )] catalyzes the electrochemical oxidation of NO to NO + and, hence HNO 2 at 1.14 V at glassy carbon.

Design of Affinity Tags for One‐Step Protein Purification from Immobilized Zinc Columns

Affinity tags are often used to accomplish recombinant protein purification using immobilized metal affinity chromatography. Success of the tag depends on the chelated metal used and the elution profile of the host cell proteins. Zn(II)‐iminodiacetic acid (Zn(II)‐IDA) may prove to be superior to either immobilized copper or nickel as a result of its relatively low binding affinity for cellular proteins. For example, almost all Escherichia coli proteins elute from Zn(II)‐IDA columns between pH 7.5 and 7.0 with very little cellular protein emerging at pH values lower than 7.0. Thus, a large portion of the Zn(II)‐IDA elution profile may be free of contaminant proteins, which can be exploited for one‐step purification of a target protein from raw cell extract. In this paper we have identified several fusion tags that can direct the elution of the target protein to the low background region of the Zn(II)‐IDA elution profile. These tags allow targeting of proteins to different regions of the elution profile, facilitating purification under mild conditions.

Synthesis and characterization of meso-[Ru(NO)Cl(dioxocyclam)] and the 1H NMR comparison with [MII(dioxocyclam)] complexes (MII=NiII and PdII) (dioxocyclam=1,4,8,11-tetraazacyclotetradecane-5,7-dione)

Abstract trans -[Ru(NO)X(dioxocyclam)], X=Cl − and OH − , dioxocyclam=(1,4,8,11-tetraazacyclotetradecane-5,7-dione) have been prepared and characterized by NMR, IR and ESI-MS techniques. The trans -[Ru(NO)Cl(dioxocyclam)] shows the nitrosyl stretch at 1846 versus 1875 cm −1 in the cyclam analogue, indicative of strong π-donation from the deprotonated dioxocyclam ligand to the Ru II center and, in turn, to the NO + group. Upon coordination, the dioxocyclam ligand no longer undergoes facile H/D exchange of the NH and C-6 protons. The methylene protons exist in symmetric patterns indicative of the meso isomer with both C-12 and C-14 methylene carbons projecting below the RuN 4 plane toward the side of the axial Cl − for the major product. A lesser amount of the rac isomer, wherein the C-12 and C-14 methylene carbons reside on opposite sides of the RuN 4 coordination plane, is detected. mmff 94 calculations determined that the meso form is more stable than rac by 3.33 kcal mol −1 . The Ru II –N(amide) bonds were calculated as having normal lengths (2.07 A), but the calculated Ru II –N(amine) bonds are elongated to 2.35 and 2.36 A. (It is known that molecular mechanics can overestimate bond lengths for second and third row metal centers by 0.10–0.15 A). Even after correcting for the over-calculated bond distance factor, it is seen that the amine–ruthenium distances are longer, and hence the bonds are weaker, than for ‘normal’ Ru II –amines. The presence of the axial Cl − is established by the ESI-MS ion fragments at m / z =394 for {d 2 -[Ru(NO)Cl(dioxocyclam)]H} + . Ions for {[Ru(NO)(H 2 O)(dioxocyclam)]} + ( m / z =377) and {[Ru(NO)(dioxocyclam)]} + ( m / z =359), the latter from loss of HCl or H 2 O from the 394 and 377 ions, are detected. 1 H and 13 C NMR data for meso -[Ru(NO)Cl(dioxocyclam)] are compared to the meso -[Pd II (dioxocyclam)] complex, and to the Ni II derivative that was synthesized at lower temperature than in previous literature reports. The meso -[Ni II (dioxocylam)] complex was identified previously. In the present work, it is shown that below 50°C the product for the Ni II system is rac -[Ni II (dioxocyclam)]. The assignments for the {Ru(NO)} 3+ , Ni II and Pd II dioxocyclam complexes as having the dominant isomeric forms as meso for {Ru(NO)} 3+ and Pd II , and rac for Ni II are supported by the H–H COSY spectrum for {RuNO)} 3+ , H–H and C–H COSY spectra for Ni II and the H–H COSY spectrum for Pd II derivatives.

Ligand field factors in promoting S = 32 {FeNO}7 nitrosyls

The EPR spectra of {FeNO}7 iron nitrosyls are of interest as models for nitrosylated nonheme proteins which exhibit the unusual S = 32 spin state. Few models of such S = 32 species are known, although [Fe(edta)NO]2− adopts this spin behavior. The present study examines [FeIIL] and [FeIIL(NO)] complexes of polyaminopolycarboxylates and pyridylmethylamines which are derived from nta3− and edta4− as models of such systems. The series of L = nta3− (nitrilotriacetate), uedda2− (N,N′-ethylenediaminediacetate), pida2− 2-pyridylmethyliminodiacetate), tpa (tris(2-pyridylmethyl)amine), tpen (N,N,N′,N′-tetrakis(2-pyridylmethyl)ethylenediamine), and edta4− (ethylenediaminetetraacetate) was chosen to provide ligands of increasing ligand field strength for both six-coordinate and seven-coordinate [FeL(NO)] complexes in order to investigate the effect on the net spin states of the nitrosyls. Electrochemical methods (CV, DPP) show that the dominant complexes near 1:1 ligand/FeII ratios are [Fe(pida)(H2O)2] and [Fe(uedda)(H2O)2] having E12 values of 0.39 V and 0.26 V vs NHE (μ = 0.1, T = 22°C). [Fe(tpa)(H2O)Cl]+ and [Fe(tpen)]2+ have waves at 0.547 V and 0.84 V. Bis complexes occur at 20:1 ratios for pida2− and uedda2−. [FeIIL(NO)] complexes are spontaneously formed by admitting NO to Ar purged FeIIL solutions. EPR spectra of frozen samples (77–105K) show that [Fe(nta)NO]− (g = 4.36, 402, 2.00), [Fe(pida)(H2O)(NO)] (g = 4.02 (axial), 2.00), [Fe(tpa)Cl(NO)]+ (g = 4.00, 2.00) are S = 32 complexes. [Fe(tpen)NO]2+ (g = 2.03, 1.97, 1.96) is a low-spin S = 12 complex with no N-shf coupling, suggestive of a seven-coordinate structural analogue of the [Fe(edta)NO]2−, S = 32 complex. MO diagrams for six- and seven-coordinate [FeL(NO)] complexes as a function of increasing ligand field strength, which incorporate the spin-polarization effect for weak fields, are presented to explain the change from S = 32 to S = 12 of the nonheme protein model complexes. The strong-field limit yields the Enemark-Feltham order. Cyclic voltammetry and differential pulse polarography studies support the prior Rhodes-Barley-Meyer conclusion for [Fe(edta)NO]2− that, for the [FeL(NO)] complexes, the oxidation [FeIIL(NO)] −1e→ [FeIIL(NO+)] occurs coincidentally at the same, or nearly the same, potential as their [FeIIL(H2O)]−1e[−→[FeIIIL(H2O)] complex. The [FeIIL(NO)] complexes of nta3−, pida2−, and edta4− are electrochemically silent at glassy carbon, whereas those of tpa and tpen are electrochemically reversible. Consistent with the MO predictions, six-coordinate S = 12 {FeNO}7 complexes have both metal dxz and ligand NO N character in the HOMO, and exhibit N-shf in the EPR spectrum. Seven-coordinate S = 12 {FeNO}7 complexes have purely metal-based dxy HOMOs and exhibit no N-shf coupling. The absence of stable CO adducts for many nonheme proteins and their model complexes is explained by the MO orders and the reduction of spin polarization as a stabilizing factor.

RuIII(hedta) as an oxygen atom transfer catalyst in the epoxidation of stilbenes

Abstract RuIII(hedta) and RuIII((CH3)2edda)+ (hedta3−=N-hydroxyethylethylenediaminetriacetate; (CH3)2edda2−=N, N− dimethethylethylenediamine-N, N−diacetate) catalyze the epoxidation of cis-stilbene and trans-stilbene using tert- butylhydroperoxide (t-BuOOH) as the oxygen source. Prior spin-trapping studies have documented the existence of LRuIIIO ↔ LRuIVO·− ↔ LRuVO2− character in the species obtained from RuIIIL and to-BuOOH (L=polyam- inopolycarboxylate ligands related to edta4−). The O-atom complex, LRuIIIO, appears responsible for the epoxidation of stilbenes. Yields as high as 63.5% cis-stilbene oxide plus 11.0% trans-stilbene oxide from cis-stilbene and 65.1% cis-stilbene oxide from trans-stilbene (with no trans-stilbene oxide) are formed in the epoxidation reactions. Secondary oxidations of the epoxide products produce between 4 to 8% benzaldehyde depending on conditions. The product distribution using the RuIIIL/t-BuOOH catalyst requires at least three epoxidation pathways: (i) concerted transfer of the oxenoid oxygen to the stilbene nucleophile; this process is favored for cis-stilbene; (ii) an outer-sphere electron transfer from the stilbene to LRuIIIO forming a carbon-centered cation radical adjacent to LRuIIIO·−; this radical pair may couple directly for cis-stilbene or after a rapid isomerization of the trans- stilbene radical; (iii) an acyclic pathway which has both free radical and carbocation resonant character; this allows for isomerism of cis-stilbene to trans-stilbene oxide products. RuIIIO(hedta) is also observed to cleanly oxidize benzaldehyde to benzoic acid, sec-phenetyl alcohol to acetophenone, and benzyl alcohol to benzaldehyde and benzoic acid. Cyclohexene is hydroxylated and further oxidized to 2-cyclohexene-1-one.

[RuII(hedta)]− complexes of 2,2′-dipyridylamine (dpaH) and a bifunctional tethered analog, N,N,N′,N′-tetrakis(2-pyridyl)adipamide (tpada)

Abstract [RuII(hedta)L]− complexes (hedta3−=N-hydroxyethylethylenediamine-N,N,N′-triacetate); L=dpaH (2,2′-dipyridylamine) and tpada (N,N,N′,N′-tetrakis(2-pyridyl)adipamide)) have been studied by 1H NMR and electrochemical methods in aqueous solution. The bidentate rings of dpaH and tpada are differentiated as shown by NMR upon coordination to RuII due to differences in the local environment. The dpa–R headgroup of each ligand binds ‘in-plane’ with the en backbone of hedta3− and with one pyridyl ring being nearer the amine of hedta3− having the pendant glycinato group (matching the known arrangement with bpy (2,2′-bipyridine)). RuII/III E1/2 values follow the order dpaH (0.32 V)

One-electron reduction potentials of coenzyme B12 and alkylcobalamins

Abstract The one-electron reduction potentials for the alkylcobalamins, R = CH 3 , CH 2 CH 3 , n-propyl, isobutyl, neopentyl and deoxyadenosyl, were examined by differential pulse polarography in 1:1 DMF:H 2 O (μ = 0.10 LiClO 4 ) at 24.6 °C. The E 1 2 values for the couple RCo + e − ⇄ RCo · (RCo = alkylcobalamin) were found to be −1.60 V, R = CH 3 ; −1.54 V, R = CH 2 CH 3 ; 1.55 V, R = n-propyl; −1.48 V, R = isobutyl; −1.38 V, R = neopentyl; −1.35 V, R = deoxyadenosyl versus SCE. E 1 2 correlates linearly with the Taft steric parameter, E s ; a new E s value for the deoxyadenosyl functional unit is estimated to be −2.03 from this relationship. A moderate solvent influence was observed by replacing H 2 O with D 2 O for methylcobalamin ( E 1 2 = −1.68 V versus SCE) and for deoxyadenosylcobalamin (coenzyme B 12 ) (−1.43 V). This suggests that solvation effects are about the same for alkylcobalamins compared to methylcobalamin and therefore do not account for the 0.22 V more negative reduction potential of methylcobalamin. The gradation in E 1 2 starting with methylcobalamin and continuing through the alkylcobalamin series may reflect changes in the axial ligand distances which modulate the energy of the lowest σ type MO (LUMO) of these complexes.

Electron spin resonance studies of the solution structure of vanadyl amino acid complexes and mixed ligand complexes of oxalate.

Esr and electronic spectra of complexes of the general composition VO(AA)2 and VO(ox)(AA) have been characterized; AA = gly, his, cys, pro, val, met, asp amino acids. Spectra of the formulation VO(ox)(LL) (with LL = imidazole plus monodentate oxalate, histamine plus monodentate oxalate, histidine, cysteine, 4-imidazolepropionic acid, mercaptopropionic acid, ethylenediamine and ethanolamine) have been used to deduce a self-consistent assignment of AL, a ligand donor additivity constant contribution to the observed hyperfine splitting, Aiso. Values of AL are sensitive to inductive effects in the ligand structure. The solution structures and likely coordination geometries of VO(his)2, and VO(cys)22-- are discussed. The role of the imidazole moiety as a sigma donor and sulfhydryl sulfur as a pi acceptor is observed in VO(AA)2 and VO(ox)(AA) complexes.