Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Aqueous chemistry of vanadium complexes of therapeutic interest Glover, Nicholas 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0035.pdf [ 2.42MB ]
JSON: 831-1.0059504.json
JSON-LD: 831-1.0059504-ld.json
RDF/XML (Pretty): 831-1.0059504-rdf.xml
RDF/JSON: 831-1.0059504-rdf.json
Turtle: 831-1.0059504-turtle.txt
N-Triples: 831-1.0059504-rdf-ntriples.txt
Original Record: 831-1.0059504-source.json
Full Text

Full Text

AQUEOUS CHEMISTRY OFVANADIUM COMPLEXESOF THERAPEUTIC INTERESTbyNICHOLAS GLOVERB.Sc. (Honours), University of East Anglia,Norwich, Norfolk, U.K., 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Chemistry)We accept this thesis as conformingto the required standardThe University of British ColumbiaNovember 1993© Nicholas Glover, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of__________The University of British ColumbiaVancouver, CanadaDateDE-6 (2/88)AbstractThe aqueous chemistry of bis(maltolato)oxovanadium(IV), VO(ma)2bis(kojato)oxovanadium(1V), VO(ka)2 and bis(maltolato)dioxovanadate(V), [V02(ma)]-were studied as part of a research interest regarding their potential as insulin mimetictherapeutic agents.UV-Visible spectroscopy provided information regarding changes in the immediatevanadium environment with variable pH, and determined that oxidation of VO(ma)2andVO(ka)2occurs upon exposure to air. Variable pH cyclic voltammetric electrochemicalinvestigation of the compounds of interest revealed that only VO(ma)2displayed reversibleelectrochemical behaviour. Room temperature and frozen solution ESR spectralparameters for the complexes were determined. Reduction of [V02(ma)]-to VO(ma)2inacidic solution was determined by ESR spectroscopy. 5’V NMR spectroscopy was used toprobe the type, and stoichiometry, of complexes present in aqueous solution as a functionof pH. Oxidation of VO(ma)2and VO(ka)2to vanadium(V) complexes was proven by 5’VNMR spectra. Potentiometric titrations were used to determine ligand protonationconstants, and metal-ligand stability constants.11TABLE OF CONTENTSpageAbstract iiTable of Contents iiiList of Tables vList of Figures viList of Abbreviations viiiAcknowledgements xDedication xiChapter 1 Introduction 1Chapter 2 Experimental 101. Solution Preparation and Chemicals 102. Spectroscopic Studies 112.2.1. 51V NMR Spectroscopy 112.2.2. ESR Spectroscopy 122.2.3. UV-Visible Spectroscopy 132.3. Electrochemical Studies 132.4. Potentiometric Equilibrium Measurements 142.4.1. Instrumentation 142.4.2. Calibration 142.4.3. Potentiometric Tiirations 152.4.4. Computations 162.4.5. Data Collection 17111Chapter 3 ResuJts and Dscuion . 193.1. Absorption Spectroscopy 193.2. Electrochemistry 283.3. ESR Spectroscopy 353.4. 51V NMR Speceroscopy 463.5. Stability Constant Determinations 61Chapter 4 Suggestions for Future Study 73References 75ivLIST OF TABLESTable 3.1. Visible and IR spectra] parameters of VO(ma)2and VO(ka)2 25Table 3.2. Variable pH electrochemistry of VO(rna)2 32Table 3.3. Isotropic ESR spectral parameters of VO(ma)2,VO(ka)2,NH4[V02(ma)]and VO(H2O)5 41Table 3.4. Anisotropic ESR spectral parameters of VO(ma)2,VO(ka)2,NH.4[V02(ma)]and VO(H20)5 42Table 3.5. 5V NMR spectroscopic properties 46Table 3.6. Variable pH 5W NMR chemical shifts of vanadate 49Table 3.7. 51 V NMR chemical shifts of vanadate-maltol complexes 56Table 3.8. Log $3 values for the vanadyl-maltol system 66Table 3.9. Log 13 values for the vanadyl-kojic acid system 67VLIST OF FIGURESFigure 1.1.Figure 1.2.Figure 3.1.Figure 3.2.Figure 3.3.Figure 3.4.Figure 3.5.Figure 3.6.Figure 3.7.Figure 3.8.Figure 3.9.Figure 3.10.Figure 3.11.Figure 3.12.Figure 3.13.Figure 3.14.Figure 3.15.Figure 3.16.Figure 3.17.Figure 3.18.Representation ofVO(H2O)5 3(3-Hydroxy-4-pyrone)-vanadium compounds investigatedin this study 7Variable pH visible spectra of VO(ma)2 21Graph oft versus pH for VO(ma) at X. = 620 nm 22Variation in the visible spectra of VO(ma)2with time followingexposure to air 23Variable pH UV spectra ofNH.4[V02(ma)] 27Reduction potentials for vanadium species as a function of pH.. 28Variable pH electrochemistry of VO(ma)2 31Variable pH electrochemistry of VO(ka)2 33Variable pH electrochemistry ofNH4[V0(ma)] 34Variable pH isotropic ESR spectra of VO(ma)2 37Isotropic ESR spectra of VO(ma)2,VO(ka)2,NH4[V02(ma)]and VO(HO)5 39Representation of the structure ofK[V02(ma)] 40Experimental and simulated anisotropic ESR spectraof VO(ma)2 43Possible ESR spectral lines due to vanadyl-rat transferrin.(VO(ma)2treated STZ-diabetic rat spectrum - non-treatedSTZ-diabetic rat spectrum) 45Variable pH 51V NMR spectra of vanadate 48Variable maltol concentration 51V NMR spectra 50Variable vanadate concentration 51V NMR spectra 52Variable pH ‘v NMR spectra ofNH4[V02(ma)] 55Variable pH5’V NMR spectra of VO(ma)2 57viFigure 3.17. Variable pH 51V NMR spectra of NH.jV02(ma)] 55Figure 3.18. Variable pH 51V NMR specua of VO(ma)2 57Figure 3.19. Variable pH V NMR spectra of VO(ka)2 59Figure 3.20. Vanadyl-maltol potentiomethc titration curves 65Figure 3.21. Vanadyl-kojic acid potentiometic titration curves 67Figure 3.22. NH.4[V02(ma)]potentiometric titration curve 68Figure 3.23. Speciation diagrams for solutions containing a 1:2 ratio ofVO2:pyrone (mM) 72Scheme I Equilibria, 51V NMR chemical shifts and possible routes offormation for vanadium(V) and vanadium(IV) complexeswith the maltolate anion 60vflLIST OF ABBREVIATIONSAbbreviation Meaninga (moles base added - moles acid present)/moles ligandA angstromA electron-nuclear spin coupling constant (lO4cm-’)Ar argonoverall stability constantdegrees Celsiuscm-1 wave number6 chemical shift£ molar absorptivity (M-1cm)E electrochemical potential (volts)e.m.f. constant of cellESR electron spin resonanceg Landé effective electron g-factorG GaussH hydrogen ionHka (ka) kojic acid, 5-hydroxy-2-hydroxymethyl-4-pyroneHma (ma) maltol, 3-hydroxy-2-methyl-4-pyroneHz Hertzic/a anodic/cathodic currentI ionic strengthJR infraredK degrees Kelvinabsorbance wavelength (nm)yinL ligandmL millilitremM millimolarmV millivoltM metal ionM (after number) molarity (mol din-3)nm nanometreNMR nuclear magnetic resonanceuv=o vibrational stretching modepH -Iog[H]pKa ligand protonation constant (log3o,ij)pK ionic product constant of waterppm parts per milliona standard deviations secondsS.H.E. standard hydrogen electrodeSTZ streptozotocinUV-Vis ultraviolet-visiblevanadium-51V voltsV02+ vanadyl ionVO(ka)2 bis(kojato)oxovanadium(W)VO(ma)2 bis(maltolato)oxovanadium(IV)[V02(ma)] bis(maltolato)dioxovanadate(V) anionX-band Ca. 9.5 GHzixACKNOWLEDGEMENTSAcknowledgements are due to those who have helped make this Thesis possible.Firstly, I would like to thank Dr. Chris Orvig, for guidance, understanding and supportthroughout my time at U.B.C.. and in the preparation and proofreading of this thesis. Mysincere appreciation is extended to Dr. Lucio Gelmini for experimental assistance, tediousproofreading and invaluable collaborative efforts. I would also like to express my gratitudeto Dr. Geoff Herring for many hours of assistance with the ESR measurements. The helpgiven by support staff in the Chemistry Department, in particular the NMR spectroscopystaff, was greatly appreciated. Many past and present members of the Orvig group,especially Dr. Martha Kline, have offered invaluable assistance, and I thank them. Specialthanks to Violet Yeun for the rat plasma samples.Financial support from U.B.C. in the form of a Teaching Assistantship is gratefullyacknowledged.A simple thanks is not enough to express my appreciation of the support,encouragement and love that Mum, Dad, Steve, Elaine and the rest of my family havegiven me, and their pride in my achievements. I would dearly like to thank Jim andColette, who have welcomed me so warmly into their family, and have helped Barbara andI in so many ways. Finally, the biggest thanks of all are for my wife Barbara, for hercontinuing wholehearted love and support, and much more.xThis work is dedicated to my Dad, Ronald Glover.The qualities ofperseverance, appreciation and pride in doing yourbest are gifts I will always treasure.xiChapter 1IntroductionVanadium was first recognized by Nils Gabriel Sefstrom in 1831, and named for itsmany multicoloured solutions after Vanadis, the Norse goddess of beauty. Vanadium isfound widely distributed in the world, and has a natural abundance greater than that ofcopper, zinc or lead. It is accumulated in above average concentrations in coal, oil and incertain rock formations, such as sandstone and shale. In addition it is found diversely in theplant and animal kingdoms. In mammals it is an ultratrace element, found particularly inblood plasma. 1,2The discovery by Henze in 1911 of high concentrations of vanadium in the bloodcells of tunicates, a form of marine ascidian, did much to promote early interest in thebiological chemistry of vanadium.3 Since then, the essentiality of vanadium in certainorganisms has been confirmed, and increasing evidence suggests vanadium is an essentialtrace element in mammals.4 Amavadin, a low molecular weight protein found in thetoadstool genus Amanira, was the first vanadium-containing protein to be isolated.5 Abromoperoxidase from the marine algae Ascophyllum nodosum was the first enzyme foundto have a vanadium-containing active site.6’7 The discovery by Cantley et al in 1977 thatvanadate is a potent inhibitor of Na, K-ATPases led to numerous studies on the inhibitoryand stimulatory effects of vanadium.8 Recent findings have established that vanadiumcompounds can exhibit insulin-like properties in biological systems, and this has led to anintensive research effort in this area during the past 10 years.9 There have been severalextensive reviews on the chemistry of vanadium in biological systems.’06The coordination chemistry of V’ species, generally referred to as vanadates, isdominated by the VO and cis-VO2moieties; the latter is often manifested in a complexof distorted octahedral geometry. The molecular geometries of vanadium(V) complexes1can be more varied than this, however, as is evidenced in recent studies showing someunusual stereochemistries for vanadates.17 The similarity of V043-and P043-, in both sizeand geometry, has led many investigators to speculate that this may be one of the factorscontributing to the biological relevance of the metal.11 The inhibitory effects of vanadiumon many phosphohydrolases (vide infra) have been attributed to the vanadium(V) oxidationstate; Gresser and Tracey have published extensively in this area, utilizing 51V NMRspectroscopy as a structural probe.’8 These and other findings have prompted numerousinvestigations of the biological chemistry of vanadates, for example vanadium(V)promoted catalysis of NADH to NADPH,23 and the role of vanadium(V) in halideoxidation by hydrogen peroxide.24 Under physiological conditions, it is believed that aconsiderable proportion of vanadate is reduced to vanadium(IV); glutathione and catecholare among the cellular constituents that have been shown to do this.4’The aqueous chemistry of both vanadium(V) and vanadium(IV) species ischaracterized by a series of complex equilibria and hydrolyses, still the subject of muchstudy and debate.2627 The formation of oligomers is highly favoured in these systems, withmany pH-dependent species formed. The speciation of vanadates is discussed further inChapter 3.4. Owing to the favourable nuclear properties of vanadium, 51V NMR is thespectroscopic method of choice for studying vanadate systems.By far the most common vanadium(IV) compounds contain the V02moiety(oxovanadium(IV) ion or, more commonly, vanadyl ion), which is the most stableoxocation of the first row transition metals. The vanadyl ion is analogous to the divalenttransition and alkaline earth metals, and reacts to form cationic, anionic and neutral squarepyramidal complexes with a variety of ligands,1’7with the oxo-oxygen occupying an axialposition (Figure l.l.). The relatively short vanadyl to oxo-oxygen bond length, —1.6 A, isI Although strictly referring to a bound substrate, ligand as used here refers to both bound and unboundsubstrates.§ A less common vanadyl complex coordination geometry that is occasionally observed is trigonalbipyramidal, as adopted by bis(2-methyl-8-quinolinato)oxovanadium(1V).112indicative of considerable multiple bond character, in most studies considered to be adouble bond. These complexes are reactive, substitution reactions occur fairly rapidly inthe apical position trans to the vanadyl oxygen (i.e. for a solvent molecule), and are oftenfollowed by rearrangement in the plane of the ligands. Vanadyl to trans -ligand bondlengths are typically some 0.2 - 0.3 A longer than those of the corresponding equatorialligands.2.1 AFigure 1.1. Representation of the penta(aquo)oxovanadium(W) ion, VO(H2O)5.9In biology, vanadium(IV) is found in the blood of ascidians (and ubiquitouslythroughout nature) and is thought to form strong complexes with proteins, for example theiron-carrier proteins such as transferrin.16’301 In addition, vanadium(IV) also forms stablecomplexes with small in vivo chelating ligands such as citrate, ATP and amino acids.12 Itappears unlikely therefore that free VO exists in biological systems. Dissolved oxygen inplasma may cause oxidation to vanadium(V), and it is probable that a partition existsphysiologically between vanadium(V) and vanadium(IV).3032 Oxovanadium(IV) has apararnagnetic, 3d’, electronic ground state, and favourable nuclear properties that make it apowerful ESR spin probe (Chapter 3.3.).A2.4A3In highly acidic solution oxovanadium(IV) exists as [VO(H2O)5],but above pH 2air oxidation to form vanadates occurs.27 However, the degree of oxidation can beeffectively retarded by using anionic, chelating ligands, with hard donor atoms such as 0 orN. The speciation of vanadyl as a function of pH is discussed further in Chapters 3.3. and3.4. Only the species found in the acidic regime can be assigned with any certainty.Inherent difficulties have made the elucidation of all equilibria and species existing inneutral and basic solution thus far impossible.26’33Pharmacological effects of vanadium previously noted include the lowering ofcholesterol and glucose levels, a diuretic effect, the contraction of blood vessels, and theenhancement of the oxygen affinity of hemoglobin and myoglobin. In addition, vanadiumhas a direct cardiac effect, similar to digitalis, possibly caused by the inhibition of the Na,K+ATPase. Several vanadium compounds have been studied with respect to theircytostatic activity, with moderate success; vanadyl species seem to be most active. Thesefindings have been summarized elsewhere.16 Vanadium is used extensively in varioustypes of industry (i.e. in alloys, as a catalyst, etc), and exposure to high vanadiumconcentration levels is not uncommon. Thus, the potential for vanadium toxicity is a viableconcern. The current consensus is that there is little risk associated with routine exposureto the metal, provided caution is observed.1Diabetes mellitus encompasses a heterogeneous group of diseases with variousetiologies. These diseases are all characterized by variable and chronic hyperglycemia andother disturbances in carbohydrate and lipid metabolism, as well as a variety of vascularand neurological complications. The prevalence of the disease in the North Americangeneral population is approximately seven million. Diabetes mellitus may be directly orindirectly responsible for some 400,000 deaths per year.34The treatment for diabetes mellitus is dominated by the regular use of insulin, ahormone secreted in the pancreas to decrease blood glucose levels, or stimulate glucose4oxidation. In diabetes sufferers the production of insulin is compromised, leading to acharacteristically elevated blood glucose level. Insulin treatment therefore involves theintroduction of the hormone to attain normal concentration levels of insulin or of glucose.However, the digestive juices of the intestinal tract render insulin inactive. This means thatoral administration is not possible, and therefore the hormone has to be injected directlyinto the patient. It is this inconvenient method of administration, and its incumbentcomplications, that have promoted the search for an orally active therapeutic agent to beused in conjunction with, or to replace, insulin treatment.35The action of vanadium compounds on glucose metabolism was first noted byShechter et al, who reported that vanadates displayed insulin-like properties in vitro in ratadipocytes;36results that were subsequently confirmed by other groups.37 However, thelevels of vanadate needed to stimulate insulin-like response in diabetic rats has beenassociated with severe toxic side effects, even death.38 Vanadyl (as the sulphate salt) hasbeen shown to be just as effective as vanadate at lowering blood glucose levels in vivo, aswell as being considerably less toxic.39 These results have prompted the synthesis, andsubsequent testing, of vanadyl based compounds as potential insulin mimetic agents.One such compound developed in this laboratory is bis(maltolato)oxovanadium(IV)(Figure 1.2(a)), abbreviated as BMOV in the pharmacological literature and as VO(ma)2 inthis study.4° The rationale behind this formulation is the low toxicity of the ligand (thecommon food additive maltol) and the belief that the complex formed would be less toxic.The use of an organic framework around the active vanadyl centre was exploited so as toprovide greater lipophilicity and gastrointestinal absorption, thereby potentially decreasingthe dosage needed to produce a positive effect. Since the development of VO(ma)2,itsinsulin mimetic properties have been shown to be highly encouraging. As well, incomparison to other vanadium treatments, it displays considerable improvement in cardiacdysfunction in diabetic rats and in the prevention of diabetes induced pathoIogy.4135There have been studies that question the assumption that vanadate is inherentlytoxic. It is also noted that vanadyl can be oxidized to vanadate both in vitro and inThis has led to the formulation of a related vanadate-based maltol compoundbis(maltolato)dioxovanadate(V),NH4[V02(ma)],in order to compare the chemistry andphysiological effects of two similar vanadyl and vanadate complexes (Figure 1.2(c)). Thiscompound has been isolated, fully characterized, and a crystal structure determination hasrecently been performed.45 In addition, a further vanadyl compound,bis(kojato)oxovanadium(IV), VO(ka)2 has been synthesized using kojic acid as a maltolanalogue (Figure 1.2(b)). Kojic acid is a biologically relevant ligand precursor, obtainedfrom the action of the Aspergillus mold family on carbohydrates.46 These three compoundswere actively tested for insulin mimetic properties at the onset of this study, and are the fociof this work. This investigation follows a previous M.Sc. thesis from this laboratory,47concerned with the synthesis and characterization of further potential insulin mimeticvanadyl and vanadate compounds, and is a contribution to a number of ongoing studies.6CH3—IhBis(malto)ato)oxovanadium(IV) I1(c) Bis(rnaltolato)dioxovanadate(V)Figure 1.2. (3-Hydroxy-4-pyrone)-vanadium compounds investigated in this study.(a) VO(ma)2, (b) VO(ka)2and (c)NH4[V02(ma)2]•0II20v,I2[(b) Bis(kojato)oxovanadium(IV) I[3I7Recently, there have been a significant number of studies performed concerning thebiological chemistry of vanadium; the number of papers published in this field has risenalmost exponentially over the last 25 years. Despite this burgeoning study, there have beenfew examples of concerted spectroscopic and potentiometric investigations in the literature.The majority of papers published within this field thus far deal with the preparation ofvarious vanadium-ligand complexes, whereas relatively little work has been performed onthe solution chemistry. It is well known that oxovanadium(IV) and vanadium(V) formstable complexes with hard 0-donor ligands, such as carboxylates, hydroxycarboxylates,hydroxamates, salicylates, catecholates, etc.164854 However, there are relatively fewexamples of comprehensive investigations of the solution chemistry of such complexes.Relevant previous studies will be introduced as needed to corroborate and/or illustratepertinent points in this work. Apparently, comprehensive synthetic, spectroscopic and invivo studies are necessary in order to better understand the coordination and physiologicalchemistry of these compounds. The work presented here represents preliminary dataconcerning their aqueous chemistry.In this study UV-Visible spectroscopy has been used to monitor changes in theimmediate vanadium environment as a function of pH. Air oxidation of the vanadylcompounds has been postulated from time dependent spectra (Chapter 3.1.).Electrochemistry provides information regarding the redox behaviour of the compounds.The variation in cyclic voltammetric waves with pH is presented, and possible redoxreactions are discussed (Chapter 3.2.). ESR is a particularly powerful technique for thestudy of pararnagnetic vanadyl compounds. Changes in the ESR signals with pH have beenmonitored, and values for isotropic and anisotropic constants determined. These werecompared to results previously determined for similar compounds. The appearance of ESRsignals in spectra obtained from solutions of vanadium(V) complexes proves that reductionto vanadium(IV) occurs (Chapter 3.3.). 51V NMR spectra were used to monitor the pH8titration ofNH4[V02(ma)],and to provide stoichiomethc information regarding complexformation and speciation in solution (Chapter 3.4.). Potentiomethc titrations were used todetermine ligand protonation constants, and metal-ligand stability constants. Speciationdiagrams were constructed for the metal:ligand systems (Chapter 3.5.). Suggestions forfuture studies are proposed (Chapter 4). It is hoped that this information provides a basisfrom which to study further these potentially therapeutic agents, and to design betteranalogues.9Chapter 2Experimental2.1. Solution Preparation and ChemicalsWater was deionized (Barnstead D8902 and 1)8904 cartridges), then distilled(Corning MP-1 Megapure Still). Solutions of NaOH (—0.15 M) were prepared fromdilutions of a 50% aqueous solution of sodium hydroxide (reagent grade, less than 0.1%Na2CO3) with freshly boiled, distilled and deionized wateT, and standardizedpotentiometrically against potassium hydrogen phthalate (KHP, BDH certified, driedovernight at 180 °C and used without further purification). Hydrochloric acid solutionswere prepared from dilutions of 12 M HCI (reagent grade) with distilled, deionized water,and standardized potentiometrically against the standard NaOH solution. A Gran plot wasused to calculate the concentration of the standard solutions.lMaltol (3-hydroxy-2-methyl-4-pyrone) and kojic acid (2-hydroxymethyl-5-hydroxy-y-pyrone) were obtained from Aldrich and used without further purification. Vanadylsolutions used for potentiometric determinations were obtained from dilutions of avanadium atomic absorption standard (Sigma). The amount of excess acid present in thesolution was determined by a Gran plot. Vanadium sulphate trihydrate, forspectrophotometric titrations, was obtained from Aldrich, analyzed for water content, andused as supplied.Bis(maltolato)oxovanadium(IV), VO(ma)2was prepared as previously described,4°by the addition of two equivalents of maltol to one equivalent of vanadium sulphate. Thecomplex was recrystallized prior to use, and fully characterized by IR, NMR and elementalI The Gran plot. (1) Before the endpoint of the titration, plot Vt vs. Vt*eT (2) After the titiationendpoint plot Vt vs. Vt*eT. The x-intercept yields the equivalence volume, Ve (Vt is the volume oftitrant, E is the emf of the cell, F is the Faraday constant).55’610analysis. Bis(kojato)oxovanadium(IV), VO(ka)2was prepared and purified under an inertatmosphere and characterized in a similar fashion.47Ammonium bis(maltolato)dioxovanadate(V),NH4[V02(ma)1,was prepared by theaddition of maltol to a solution ofNIL1VO3,and stirring overnight in air.” The compoundwas recrystallized and fully characterized by IR, NMR and elemental analysis prior to use.The dioxo- anion can be obtained using a number of countercations: NH, K+ etc.Samples of VO(ma)2and NH4[V02(ma)2] were kindly supplied by Dr. L. Gelmini;VO(ka)2was kindly supplied by Y. Zhou, both from this laboratory.Concentrations of solutions were determined by dilution of accurately weighedsamples in analytical standard, graduated glassware. Where appropriate, vanadiumcontaining solutions were prepared and kept under an inert atmosphere, i.e. Ar or N2.2.2. Spectroscopic Studies2.2.1. 51V NMR SpectroscopyVanadium-51 NMR spectra (78.864 MHz) were recorded on solutions in 5mm o.d.tubes at room temperature, using a Varian XL-300 Spectrometer. Neat VOC13 sealed in a5mm o.d. tube was used as an external standard, and assigned a chemical shift (6) of 0.0ppm.57 The spectra were obtained by judicious choice of number of transients (typically-.5,000), pulse width (8 - 12 spectral window (1 x iO Hz) and line broadening (1 - 3Hz), depending on the system under observation. The error associated with each chemicalshift is estimated to be ± 0.2 ppm.§ To retard air oxidation of vanadyl containing solutions.112.2.2. ESR SpectroscopyX-band ESR spectra, recorded as the first derivative of absorption, were obtained atCa. 9.60 GHz using a Bruker ECS-180 ESR spectrophotometer. A Eurotherm B-Vt 2000flowthrough variable temperature controller was used to provide temperatures of 100-120K at the sample position in the cavity for anisotropic spectra. Calibration of microwavefrequency and magnetic field were performed automatically with Bruker ECS-180software, using an E.LP. 625A CW microwave frequency counter. Samples were containedin either 4 mm o.d. glass tubes, or 20 tL microcapillaries. Typically, instrument settingswere: field set = 3200 - 35000; scan range = 1200 - 1600 0; power =2 - 20 mW; gain =1000 - 10000; time constant = I - 2 s; modulation amplitude = 1 - 5 0; scan time = 1 - 3mm. Frozen solutions were prepared by slow pre-freezing of the sample from the bottomup in a Dewar of liquid nitrogen, before admittance to the spectrophotometer cavity. Thislowers sample tube mortality rates due to freeze fracturing, thus protecting the cavity, andlessens solute aggregation during freezing, and facilitates glass formation. Analyses of thespectral parameters were performed using software routines developed in this department,run on a 486-PC computer. Isotropic parameters were determined by software analysisdirectly from the spectral data. Anisotropic parameters were determined by spectralsimulation in the following manner: (1) Estimates of likely g-factors and A-values weredetermined from the experimental spectra. (2) A spectral simulation was obtained usingthese values.58 (3) The experimental and simulated spectra were compared, by softwareroutine. (4) Estimates of simulated parameters were refined until the comparison could notbe further improved. Estimated errors in the spectral parameters were determined by theminimum change in a particular parameter necessary to cause an observable change in thesimulated spectrum.122.2.3. UV-Visible SpectroscopyUV-Visible spectra of aqueous solutions were recorded in 1 cm quartz cells at 25 Cunder Ar. Spectra were measured between 200 - 450 nm and 500 - 900 nm using aShimadzu UV-2100 Spectrometer linked to a data terminal. A cell containing 0.15 M NaCIwas employed as a solvent blank.2.3 Electrochemica) StudiesCyclic voltammetric data were obtained using a Princeton Applied Research Model264 polarographic analyzer/stripping voltammeter, and a P.A.R. Model RE0089 X-Yrecorder. Electrochemical measurements were carried out under an argon atmosphere atroom temperature. Solution concentrations were i0 M in complex and 0.15 M insupporting electrolyte (sodium chloride for aqueous solutions). Voltammograms wererecorded using a platinum working electrode, a platinum wire counter electrode and aAg/AgCI reference electrode checked periodically relative to a 1.0 x i0 M solution offerrocene in acetonitrile containing 0.1 M TEAP for which the ferrocenium/ferrocenereduction potential was 400 mV and E =72 mV at a scan rate of 100 mV/s.59132.4. Potentiometric Equilibrium Measurements2.4.1. InstrumentationPotentiometric titrations were performed using an automated potentiometric titrationapparatus developed in this laboratory. The emf measurements were made with a FisherAccumet 925 pH meter equipped with Orion Ross research grade glass and referenceelectrodes.The solutions were maintained at a constant temperature of 25 ± 0.1C in waterjacketed beakers by means of a Julabo UC circulating water bath. The beakers weresecurely sealed by means of a rubber bung, with entries for the electrodes, argon gas line,and automatic burette tip. The solutions were gently stirred with a magnetic stir bar on aninsulated magnetic stir plate, and were kept constantly degassed under an argonatmosphere. The ionic strength was adjusted to 0.15 M (isotonic) by the addition of NaCLA Metrohm Dosimat 665 automatic burette was used to deliver the standardized acidand base solutions. The automated potentiometric titration system was controlled by adedicated Acer 710 PC-XT computer running computer programs specifically written forthis application.2.4.2. CalibrationThe electrodes were calibrated before, and occasionally after,l each titration run bythe potentiometric titration of an acidic solution of known concentration with standardNaOH. The titrations were performed using the computer program CTITRA1.60 Theresults of these titrations were analyzed by a non-linear least squares fitting program (CALANAL).6° The program reads calibration titration data (mL titrant, observed mV readings)and refines values for electrode E and slope. Only values in the ranges pH = 2.3 - 2.9 and1 To determine if significant electrode drift had occwTcd during the course of the titration.14pH = 10.8 - 11.3 are refined in the calibration, as these yield the most linear electroderesponse.6’ Outside these regions it was sometimes necessary to calculate both F and theslope directly from calibration data within a specific range, to give best fit in buffer regionsand areas of steeply changing slope, as well as in extremes of acid and base.2.4.3. Potentiometric TitrationsPotentiometric titration apparatus was controlled by use of the computer programTITRA1.6°The values of electrode F and the slope were refined in the calibration andwere used to determine pH valuest from observed millivolt readings using equation 2.1.mV = - log [H] . slope (2.1)The critical factors that determine the accuracy of automated titrations are: (1)allowing equilibrium to be reached, and (2) waiting for pH electrode response to stabilize.The program allows for this by measuring and comparing data sets of observed emf values(and hence pH). If the variance in a measured data set is not acceptable, then the programcontinues to take further readings until a suitable value for the variance is obtained. Onlythen will the computer instruct a further aliquot of titrant to be added. The program thenwaits for a pre-determined period of time, whilst stabilization of emf readings occurs,before resuming data set measurement. The program yields output of mL titrant added andobserved pH. It is prudent to choose a number of experimental points that is not too low (<20) nor high (> 120).63 Typically, —50 - 80 data points are obtained in each titration run.Titration runs were repeated with different concentrations, and on different days to ensurereproducibility.t Because of the widespread use of the term pH in potentiometric studies, it is perhaps necessary tocomment on its definition and measurement. The quantity pH, defined as -logai+. is not explicitlymeasurable; however, under conditions of constant ionic strength maintained by an inert supportingelectrolyte, activity coefficents are essentially constant. Thus the potential of the hydrogen electrodevaries linearly with hydrogen ion concentration (-log[H]), as well as with hydrogen ion activity.62 Inthis study pH refers to -log[Hj, as defined above.152.4.4. ComputationsThe computer program BEST” was used to determine the successive overallstability constants, (3, for each of the metal ligand systems M H L (Metal HydrogenLigand). The program sets up simultaneous mass-balance equations for all componentspresent at each point in the titration curve, and calculates pH values for these pointsaccording to the set of stability constants judiciously chosen by the user, and theconcentrations of each component. The program then minimizes the sum of the squareddifferences between the observed and calculated pH values from variations of the selectedstability constants. A standard deviation, a, is given by the equation 2.2, where N is thenumber of titration points.a = L (pHk. pH)2/ (N-i) (2.2)This represents the fit in log units for the reproducibility of the titration curve.Adjustment is continued until there is no further minimization of fit. The programincorporates a weighting factor (numerically calculated square reciprocal slope) to increasesensitivity of the calculations to buffer regions, and to offset points encountered in the lessaccurate regions caused by a steeply changing slope. The treatment of data using BESTallows for the variation in selected equilibrium constants to minimize the least squares fit ofobserved to calculated data. Of primary importance in obtaining accurate results with thiscomputational method is the judicious choice of viable complexes to be included in thecalculation.The computer program PKAS65 utilizes a simplified version of the algorithm foundin BEST and was used to calculate protonation constants in the simpler H L system.162.4.5. Data CollectionThe equilibrium constants described in this study are defined according toequation 2.3:xVO2 + yIP +zL - (VO2+)x(H+),(.)zX+Y•Z (2.3)CalibrationData collected in the pH ranges 2.3 - 2.9 and 10.8 - 11.3 were used in the calibration.Typically some 20-30 data points were obtained in each run.Vanadyl-maltol system(i) Protonation Constant of Maltol.Data collected from 15 titrations (—800 data points), in the concentration range 0.3mM [HL] 0.4 mM, pH 9 were used to determine the protonation constant of thehydroxyl proton of maltol.(ii) Stability Constants of Vanadyl Maltol Complexes.Data were collected from 15 titrations (—800 data points), in the concentration range0.8 mM [HL] 5.0 mM, 0.4 mM VO 2.0 mM, pH 7. The metal:ligand ratio wasvaried in the order 1:1, 1:2, 1:3, 1:10.Vanadyl-kojate system(i) Protonation Constant of Kojic Acid.Data were collected from 6 titrations (—250 data points), in the concentration range0.3mM [HL]0.4mM,pH8.17(ii) Stability Constants of Vanadyl-Kojic Acid complexes.Data were derived from 11 titrations (—700 data points), in the concentration range0.4 mM [HL] 4.0 mM, 0.4 mM [VO2]0.8 mM, pH 8. The metal:ligand ratio wasvaried in the order 1:1, 1:2, 1:3, 1:10.N114[VOi(ma)i]Data were derived from 5 titrations (—250 data points). (Conditions: 5 mM, 0.15 MNaCI, 25 C, under Ar). Data analysis was performed using the Henderson-Hassleblachequation, 2.4.log [HA]/[A] = log 3 + log [H] (2.4)A plot of log [HA]I[Ai vs log [H] yields an intercept corresponding to log 3, with a slopeof 1.5618Chapter 3Results and Discussion3.1. Absorption SpectroscopyOxovanadium(IV) complexes have visible spectra that are quite distinct from othervanadium(IV) species.” The strong axial perturbation of the vanadium-oxo group has asignificant effect in determining the spectrum.69 Absorption spectra have been used toprobe the variation in the ligand field environment around the metal centre in vanadylcomplexes with small chelating ligands and in proteins.7074 These spectra have been usedto assign electronic transitions, and to help elucidate structural changes in solution.A convenient method for predicting the stoichiometry of a dominant species insolution is the method of continuous variation.75 The technique has been used, amongothers, to study the stoichiometry of the interaction of vanadate with monosaccharides andnucleosides,76 and vanadyl interaction with phytic acid and other phosphorus ligands ofbiological significance.77 The method is simple, and provides a first approximation forfurther studies. A series of solutions with the mole fraction of the ligand X rangingbetween 0 and 1 were prepared and absorbance measurements performed at a selectedwavelength (typically between 625 - 765 nm). As both the metal ion and the complexabsorb appreciably at this wavelength, and the ligand does not, the quantity can bedetermined, defined as: = A - Ath (Ath, is the theoretical absorbance due to themetal alone). A plot of Ai versus mole fraction X (Job plot) should have a maximumcorresponding to the composition of the complex. Three maxima at XL = 0.6 wereobserved in the Job plot for the VO - maltol system, under argon, at three wavelengths inthe visible region (X= 765 nm, 700 nm, 625 nm). These correspond to a solution consisting19of a major component of 1:2 metal:ligand stoichiometry, and a minor component of 1:1stoichiometry.75The visible spectra of a typical pH titration of VO(ma)2 under argon are shown inFigure 3.1. For pH <3 the spectra show a maxima at —760 nm with a shoulder at higherenergy. These are similar to the spectrum of the aqueous vanadyl ion (see Figure 3.1.Inset), which has absorbances at 760 nm and 625 nm.71 For 3 <pH < 8 a band appears at—620 nm, and becomes more defined with increasing basicity. The spectra generallyindicate that there is a change in the ligand field environment around V0 in this region,and the spectra appear significantly different from that of the vanadyl ion.78 The band at—760 nm shifts to lower energy with increasing pH. Above pH >9, the general increase inabsorbance and the observed shift to higher energies in the region 500 nm < < 600 nmsuggest hydrolysis and/or oxidation of vanadyl.78 A plot of £ (620 nm) vs pH (Figure 3.2.)shows that for the region 3 <pH < 8 £ remains relatively constant, indicating that in thisregion the ligand environment around the VO centre remains essentially unchanged, andis presumed to be an 04 or an 05 ligand donor set. Similar inferences regarding suchchanges in ligand environment have been drawn from observations reported in theliterature.70 In very acidic or very basic solutions, a different environment is observed,corresponding to aquo- and hydrolyzed/oxidized species.A study by WUthrich clearly demonstrates the difference in the appearance of spectraof 1:1 and 1:2 vanadyl:ligand complexes.79 The spectra observed in this study areconsistent with those previously assigned for vanadyl:ligand 1:2 complexesY’8°From thework of Kozlowski et al, significant dimerization of vanadyl complexes is manifested bythe appearance of four distinct bands in the absorption spectra;8’hence it is also possible todiscount the significant formation of dimers as major species in these solutions. Thepredominance of a 1:2 complex is indicated from these data.20pH10.5bf:<500 600 700 800 900(rim).Avtg?NFigure 3.1. Variable pH visible spectra of VO(rna)2(1 mM, 0.15 M NaCI, 25 underAr) Ordinate represents variation in £ (arbitrary units).213025.20.w15.10.•)og[H+]Figure 3.2. Plot oft (molar absorbance) versus pH for VO(ma)2at = 620 nm. Dataand conditions from Figure 3.1.In a related experiment, a solution of VO(ma)2 at pH 6 was exposed to air, andmonitored at successive time intervals over a number of hours. The variations in spectra,Figure 3.3., show that there is a change in the ligand environment around the vanadylcentre with time, as the solution changed colour from green to orange. The bands at 620nm and 870 nm decrease in intensity with increasing time, and eventually disappear, whilethe intense band below 550 nm remains. If oxidation of the complex occurs, then it is thed-d bands which are expected to disappear upon removing the lone 3d eIectron. Theintense band below 550 nm is indicative of a charge transfer band,82 and therefore thespectra are strongly suggestive of an oxidation to a vanadium(V) species. The oxidant ispresumed to be dissolved oxygen, as evidenced by the lack of oxidation of solutions thatare stored under argon. Saito has observed a similar oxidation for the conversion of[VWO(nta)] to [VVO2(nta)].l83§ Note, however, that the onset of oxidation in vanadyl complexes is both concentration and pH dependent.I ma refers to nitriloacetic acid.I I I I I1 3 5 7 9 1122(Hours).24\ime101i3I I 1 —1500 600 700 800 900)(nm)Figure 3.3. Visible spectra at various times after the exposure to air of an aqueoussolution of VO(ma)2 (1 mM, pH 6.0,0.15 M NaC1, 25 C).23Similar observations were made in a study by Stewart and Porte,8’twho proposedthat bisligand-vanadyl complexes, such as those investigated, undergo reversible oxidationin aqueous solution to form dioxobis(ligand)vanadate(V) anions. The mechanism by whichthis one electron oxidation with molecular oxygen proceeds is not yet understood. Apossible pathway by which the oxidation of VO(ma)2proceeds is given in an unbalancedform in equation 3.1.OII{OVO(ma)2 - [V02(ma)7J(3.1)In Porte’s study, the sixth coordinated ligand was postulated as being trans to the oxogroup. It appears likely that VO(ma)2 is oxidized and that anionic [V02(ma)]- issubsequently formed in these solutions. A crystal structure determination performed atU.B.C. recently proves that K[V02(ma)]is a cis complex.45 Hence, initial oxygensubstitution might occur in the trans position to V=O, followed by apical to basalrearrangement to form a cis product. Stewart and Porte also note that no evidence wasfound for dimer formation.85In the analogous kojic acid complex similar observations were made. In the variablepH visible spectra of VO(ka)2 the appearance of the spectra are also strongly indicative ofthe dominance of a 1:2 species throughout the neutral pH region. A solution exposed to air,and monitored by visible spectroscopy with time, shows that gradual oxidation also occursin this instance.The assignment of bands in the absorption spectra of vanadyl complexes has been amatter of some controversy.67’Typically three bands are observed in the visible region:bands I (between 910 - 625 nm), II (between 690 - 520 nm), and III (between 475 - 330nm). Molar absorbances of 5 - 150 M-’ cm-1 are typical. The energy level schemedeveloped by Ballhausen and Gray is perhaps most often used to interpretoxovanadium(IV) visible spectra.86 However, some studies have questioned the universal24validity of the Bailhausen-Gray scheme,’87and there is no real agreement on the orderingof the vanadyl energy levels.The V=O bond stretching frequency, UV, is an important characteristic ofoxovanadium(IV) complexes, and is generally observed at 985±50 cm.67 The magnitudeof the stretching frequency is particularly sensitive to the a-donor qualities of the otherligands bound to the vanadyl centre. Donors that increase the electron density of thevanadium centre will reduce its acceptor properties towards the multiply bonded oxygen(decreased pit -+ dit donation). A decrease in the vanadium-oxygen bond order is reflectedin a decrease in uvo. It has been found in general that values of uvo —950 cm4 areindicative of octahedral geometry, and values of —980 - 1000 cm-’ of square-pyramidalgeometry with the V=O axial, and having no ligand trans.67’8Visible and IR spectral parameters for VO(ma)2and VO(ka)2are given in Table 3.1.The data reveal that the spectral parameters for the ligand complexes of interest arecomparable to those previously obtained for similar vanadyl species.88’9Table 3.1. Visible and IR spectral parameters of VO(ma)2and VO(ka)2.Complex ) (cm-’) £ (M-’c&) uvo (cur’)VO(ma)2 625, 86O 15, 30 995VO(ka)2 610, 8401 17,32 980pH 5.0I pH5.5In all these visible spectra the tail end of a high intensity charge-transfer band is seenbelow 500 nm in the UV region of the spectrum. Vanadium(V) species however, lacking delectrons, show no transitions in the visible region. A pH titration of N}Lt[V02(ma)2] inthe UV region is shown in Figure 3.4. A peak at = 273 nm has a maximum absorption25corresponding to lowest pH. Between pH 1.5 and 4.0 the gradual decrease in intensity ofthis absorption is observed, with a single isosbestic point noted at 292 run. In the pH range4.5 - 7.5 the spectra are almost identical, indicating that only one major species exists inthis region. Above pH 8.0 the increase in intensity of an absorbance at ) = 318 nm isobserved, with a concomitant reduction in the intensity of the band at 273 nm. A singlepseudo-isosbestic point is noted at 295 nm. The bands correlate to those of protonated ( =273 nm) and deprotonated Q. = 318 nm) maltol ligand, as determined in this study, and it islikely that the bands are due to it -÷ it transitions within the ligand. In acidic solution thebands possibly denote species involving protonated maltol ligand and vanadyl andvanadate-maltol complexes. In the neutral pH region it appears likely that anionic and/orprotonated [V02(ma)2] is prevalent, from 5’V NMR data reported in Section 3.4., and thespectra are assigned thus. In basic solution vanadates predominate, hence the bandobserved at 318 nm presumably corresponds to uncomplexed free ligand, L-.26Figure 3.4.0CCC0‘P0C0000Variable p1-I UV specta ofNH4[V02(nia)](2mM, 0.15 M NaC1, 25 C,under Ar). Arrows indicate the direction of peak intensity with increasingpH..0g.‘PCC25C 260 300 320 350273.2. ElectrochemistryElecuochemical data are useful in obtaining information about the redox propertiesof metal complexes. Of the common oxidation states of vanadium, only Vm, V”, andVVare of any significance when considering physiological systems, V being far too reducingto exist under normal conditions. In acid, the predominant species for these oxidationstates are V, VO2 and cis -VO2. As the p14 is raised, hydrolysis occurs and a numberof monomeric and oligomeric vanadium species are formed. The variable pH reductionpotentials for various vanadium species are shown in Figure 3.5.wvszwFigure 3.5. Reduction potentials (versus S.H.E.) for various vanadium species as afunction of p11. Boundary lines coirespond to values where species inadjacent regions are in equal concenuations. Short dashed lines indicateuncertainty in the location of the boundary.28In the systems under observation V’’ - V” is most likely to be the only importantredox couple. Previous electrochemical studies, performed in H20, have shown that thepotential for this couple in five coordinate vanadyl species generally lies between +0.3 V(vs S.H.E.), for example in ((3,5-Dimethylpyrazolyl)borato)dichlorooxovanadium(IV),8991and +0.9 V. as observed for (N,N-Disalicylidenecthylenediamine)oxovanadium(IV), andis dependent upon the ligands bound to the vanadyl centre.In cyclic voltammetry, the potential of a small, stationary, working electrodeê ischanged linearly with time starting from a potential where no electrode reaction occurs andmoving to potentials where reduction or oxidation occurs. After traversing the potentialregion of interest, the direction of the linear sweep is reversed. The reversibilty of a cyclicvoltammetric wave is usually compared to that of a known system. One such standard isthe ferrocenium/ferrocene redox couple, a one electron reversible process with tE =72 mV(versus S.H.E.), Ic/ia =1, at 100 mV/s.lTo investigate the redox behaviour of these systems the variable pH electrochemistryof VO(ma)2was studied using cyclic voltammetry (Figure 3.6.). The measurements werereferenced to the ferrocenium/ferrocene reduction potential. For 5 <pH < 8, an irreversiblewave is observed; E >>72 mV, i /i1 >1 (Table 3.2.). As the acidity of the solution isincreased, the wave becomes increasingly quasi-reversible (jc/ial). The wave obtained atpH 2.5 closest resembles reversibility as defined by the ferrocenium/ferrocene redox couple(zE —72 mV, IC!i1=l), and from ESR measurements reported in Section 3.3., is likely tocorrespond to a solution containing predominantly VO(ma)2. In the extremes of acid andbase the waves obtained appear similar to those of oxovanadium(IV) and vanadium(V)respectively, as determined in this study. The E1f2 value of 0.77 V versus the standardhydrogen electrode at pH 2.5 lies within the range of the expected values for the V” -+ VWi.e.Ptasusedinthisstudy.I Where: E Ecathodic -Edic, ic = cathodic current, 1a = anodic current.29redox couple.t The process occurring can be summarized as in equation 3.2. The markeddependence of (quasi-) reversibility with pH can be explained by the increasing extent ofhydrolysis of the complex with increasing basicity, and by the steady conversion ofVO(ma)2to various vanadium(V) species, as implicated from ESR and 51V NMR studiespresented herein. Hence, depending on the pH of the solution, a mixture of species isobtained. The redox couple corresponding to vanadium TV ++ III frequently has potentialsranging from 0.2 V to - 1.0 V; in this study no waves were observed in the region - 1.4 Vto 0.0 V or above 0.75 V, indicating that there is only one redox couple of importanceinvolved in these systems.[V1’O(ma)2] — [VVO(ma)z]4 + e (3.2)In the variable pH electrochemistry of a corresponding solution of VO(ka)2 theobserved waves are irreversible in all pH regions; àE >>72 mV, /i >1 in each case(Figure 3.7.). In addition to the reasons stated for solutions of VO(ma)2, these observationsmay be attributed to the effect of the hydroxyl group of the kojate ligand. This may reactwith neighbouring vanadium centres to form kojato chain complexes, hence causing theirreversible redox behaviour observed with VO(ka)2. Further studies are necessary toestablish this.t Potentials reported versus the standard hydrogen electrode are derived from those experimentally obtainedusing AgfAgCl electrodes as follows: E (S.H.E.) = E (AgIAgCI) + O.2223.30253D3402PjsAt4.57w.075556065775O•075 oFigure 3.6. Variable pH electrochemistry of VO(ma)2(0.15 M NaCI ,25 C., under Ar,scan rate 100 mV/s. electrode: Ag/AgCI)31Table 3.2. Redox potentials (Volts) obtained from a cyclic voltammetric investigationof VO(ma)2at variable pH (25 C, 0.15 M NaC1, under Ar).pH Ei (V) liE (mV)b ic/ia2.5 0.77 71 1.003.0 0.73 68 1.033.5 0.70 65 1.034.0 0.69 64 1.015.0 0.67 55 >>1a Potentials reported vs standard hydrogen electrode (E Ag!AgC1 + 0.222); ± 0.01 V.b =72mV,ic/ia=1.00forFc/Fc.The variable pH electrochemistry ofNH4[V02(ma)]gave irreversible waves for pH>4 (Figure 3.8.). As the pH is lowered, the waves become increasingly quasi-reversible.At pH 2.5: E1 =0.70 V, E =67 mV and i/i =—1. This corresponds to the V”/V” redoxcouple, and quite reasonably, appears similar to that for VO(ma)2. ESR observationspresented in Section 3.3. indicate that as acidity is lowered in a solution ofNH4[V02(ma)]increasing amounts of VO(ma)2 are observed. In addition, the first ESR signal detectedoccurs at pH <4, at which point quasi-reversibility is also manifested. Hence, the findingsappear to indicate that VO(ma)2exhibits reversible electrochemical behaviour in aqueoussolution, whilst its hydrolysis and oxidation products do not. Differences between E1values of VO(ma)2 and reduced NH4[V02(ma)]determined here presumably are due tothe differing amounts of VO(ma)2 present in either solution, the “concentration” ofVO(ma)2being both pH and concentration dependent.32===30 707.540’ 804.59050 i 6.075 0.075 0Figure 3.7. Variable pH electrochemistry of VO(ka)2 (0.15 M NaCI ,25 C., under Ar,scan rate = 100 mV/s, electrode: AgIAgC1)332OA{Figure 3.8. Variable pH electrochemistry of NH4VO2(ma)](0.15 M NaCI ,25 C.,under Ar, scan rate = 100 mV/s, electrode: Ag/AgCI)pH605.550’4.54030.25.0175 0•343.3. ESR SpectroscopyVanaclyl complexes have a low lying singly degenerate ground state due to their nearidealized C4v geometries, axial distortion due to the oxo group, and 3d1 electronicconfiguration. This makes vanadyl complexes valuable as spin probes in ESR. The d0vanadium(V) species do not exhibit ESR spectra. In general, the VW oxidation staterequires a noncubic field for ESR observability, fulfilled in vanadyl complexes. Anothersimplifying feature of these spectra is the absence of interelectronic effects, since there isonly one d electron.94 ESR is a valuable speciroscopic tool, being several orders ofmagnitude more sensitive than most other spectroscopic techniques. The ESR spectra ofvanadyl complexes at room temperature exhibit characteristic widely-spaced eight linepatterns due to the coupling of the unpaired electron with the large moment of the —100%abundant 51V nucleus (I = f2).In addition to the intensity and line shape of the absorption, two fundamentalparameters can be derived from the ESR spectral measurement. These are the Landéeffective electron g-factor and the electron-nuclear spin coupling constant A. The isotropicESR parameters A0 and g0 can be determined from the position and spacing of the eightresonant lines for the room temperature spectrum. Frozen solutions of axial d’ systems(e.g. vanadyl complexes) show two sets of overlapping eight line resonance components,one set due to the parallel (N) features and the other due to the perpendicular (J..) features.Ideally, ESR spectra can yield much useful information to the coordination chemist:(1) identification of the metal, oxidation state, and spin state; (2) identification of thebinding site (ligands) and local symmetry; (3) determination of the concentration of theparamagnetic ion.95 The hyperfine coupling constants A, given in lO cm-1, are highlysensitive to the ligand field around the vanadyl centre; linewidths to the size of the complexand to the viscosity of the solution.96’7 This has been demonstrated in the fine structure35observed in the ESR spectra of vanadyl albumin, which clearly shows two distinct bindingenvironments.98 These large protein complexes exhibit near rigid limiting spectra due toslow molecular tumbling.72 It is also possible to obtain spectra of samples containingminute concentrations of vanadium, thus allowing for the study of tissue and sub-cellularfractions. Studies of this kind have been performed on organs from vanadium treateddiabetic rats,”00 and as a probe of pH in intact vanadocytes from Ascidia cerawdes.101The use of ESR is probably one of the most powerful tools in the study of vanadylcontaining species.102The hydrolysis of the vanadyl ion [VO(H2O)5]in aqueous solution is ofconsiderable importance when considering the ESR spectra of vanadyl complexes. TheVO2 spectrum loses intensity as the pH is increased from 2 to 4 due to the formation of anESR-silent hydrolysis species, [VO(OH)]2.A sharp decline in the intensity of the spectraafter pH —4 marks the onset of the formation of a grey precipitate of [VO(OH)]. By pH—6 precipitation is complete, and the solution is ESR-silent. No further signals areobservable until pH >11, when significant amounts of ESR active VO(OH)3-forms.These observations have an important ramification; between pH —5 - 10, ESR signalsobserved in solution must be solely due to vanadyl-ligand complexes, and have noattributable contribution from vanadyl hydrolysis products.The variable pH ESR spectra of VO(ma)2are depicted in Figure 3.9. Below pH 1.5the spectra (not shown) appear similar to those observed in this study for VO(H2O)5,observations that concur with results obtained in the visible spectra of VO(ma)2reported inSection 3.1 Analysis and comparison of the spectral parameters confirm that in highlyacidic solution VO(ma)2 is converted to VO(J42O)5. In the region 2 <pH <9, aspectrum distinct from that of VO(H2O)5is observed, having greater intensity in moreacidic solution. Analysis of the spectral parameters, given in Table 3.3., reveals nodifference in g0 or A0 throughout the pH range. Hence, it may be concluded that there is36only one major vanadium(IV) species throughout this region, and based on a 1:2metal:Iigand dominant species determined in the visible spectroscopic study, it is VO(ma)2.Spectral integration reveals that between p144-7 the spectra arc all of equal intensity, andhence “concentration”. Assuming that in the neutral pH range ESR signals are solely dueto vanadyl-ligand complexes, then in neutral pH 65% of the total vanadium in solution ispresent as VO(ma)2.This is detennined from spectral integrations relative to VO(H20)5,which is assumed to correspond to 300% of the total vanadium in solution under highlyacidic conditions.103 As basicity is raised spectral intensity wanes, and by pH 10 thespectra has all but disappeared into the baseline. This is consistent with conversion ofvanadium(IV) to vanadium(V), as indicated by the 5V NMR spectra reported in Section3.4. The corresponding variable pH ESR spectra of VO(ka)2 (not shown) revealed similartrends; spectral parameters are reported in Table 3.3.pE9. 3.9. Variable pH ESR spectra of VO(ma)2 (0.15 M NaC1, 25 C, under Ar,spectra depicted on an equal intensity scale).3000 3200 3400 3600 380037In the variable pH ESR spectra of ESR-silent d° NH4[V02(ma)](not shown) nospectra are observed above pH 4. As the acidity of the solution is increased a typicalvanadyl ESR spectra is observed in increasing intensity. Analysis of the isotropic spectralparameters indicates that this species is VO(ma)2, as the parameters are identical (Table3.3.). Below pH 1.5 VO(H2O)forms, as evidenced by the appearance of the spectra, andthe spectral parameters. It is apparent therefore that the vanadium(V) species has beenreduced and that there exists a facile route for the interconversion between thesevanadium(IV) and coffesponding vanadium(V) ligand complexes in aqueous solution.A closer examination of the room temperature spectrum of VO(ma)2 is given inFigure 3.10., and shows the presence of a minor species in the spectrum, indicated in thefigure by an arrow. The nature of this minor species may be attributed to an isomer ofVO(ma)2with trans ketonic oxygens. The presence of suchcis and trans isomers ofvanadyl complexes has previously been resolved by ESR spectroscopy.’°5 An alternativeexplanation is that the distortion is due to VO(H2O)5.Further studies are needed toconfirm its identity.38H3 o H30%1k%%II 0th.bis(makolato)oxovanadjumay) srans.bis(maholato)oxovanadjumav)X.Band ESR298 K1G3Figure 3.10. 1soopic specua of VO(ma)2,VO(ka)2,NR4[V02(ma)]and VO(H2O)5.eb•VO(ma)2VO(ka)2NH4!V02(ma)]m= “‘2I I I I3000 3200 3400 3800 380000039[o..v.o=1o3.7’ IFigure 3.11. Chem 3.DTM representation of the structure ofK[V02(ma)].45In the corresponding ESR spectrum of VO(ka)2,Figure 3.10., it was not possible tounequivocally determine if such a minor species exists. However, comparison of thespectra of these pyrone complexes with that of VO(H2O)5reveals a spectral distortion,particularly evident in the ni = - /2, /2 and /2 lines.40Table 3.3. Isotropic ESR spectral parameters of VO(ma)2, VO(ka)NH4[V02(ma)1.Complex g0(&) A0(b) (10 cm-1)VO(ma)2 1.963 103.7VO(ka)2 1.963 104.5NH4[V0(ma)] 1.963 103.6VO(H2O)5 1.968 106.7a±OO(j1b±01 (10 cm-’).The anisotropic spectral parameters for the complexes of interest were determined byspectral simulation of the frozen glass spectra obtained at —110 K. The anisotropicspectrum, and corresponding simulated spectrum, of VO(ma)2are shown in Figure 3.12.Often X-band spectra can be described by axially symmetric tensors, in which case g =gzz, g.L = g, = gyy, A11 = and A.L = = Ayy. Simulation of the anisotropic spectumof VO(ma)2 using these parameters failed to adequately describe the experimentalspectrum. In a study of the analogous anisotropic ESR spectral parameters of VO(acac)2,the authors determined that the complex exhibited slight rhombicity.t’°6 Rhombicity ismanifested in cases where xx - g> 0.006 and A1 = 4 x10- cm-1 (for axes, seeFigure 3.12. inset).28 Upon closer examination of the frozen glass spectrum of VO(ma)2,the manifestation of rhombicity is evident from the position and spacings of spectral lines,particularly in the central region of major overlap (the minor VO(ma)2 isomer furthercomplicates the spectrum). The spectrum is satisfactorily simulated with a slight rhombicdistortion with parameters g = 1.979, g = 1.973, g = 1.938 (g - g = 0.006), A =t VO(acac)2refers to bis(acetylacetonato)oxovanadium(IV).4155, A,3, =60 and A21 = 169 (A - =5 (10-4 cm’)). By this method the frozen glassspectrum of VO(ka)2was similarly simulated with slight rhombic distortion, whilst thefrozen glass spectrum ofNH4[V02(ma)],quite reasonably, is coincident with that ofVO(ma)2,and has identical parameters (Table 3.4.).Table 3.4. Anisotropic (simulated) spectral parameters for VO(ma)2, VO(ka)NHJVO2(ma)]and VO(H20)5’.Complex g,(a) g(a) g(1) Acj) AyyO’) Azz0)(10-4cm) (1O-4cm’) (1O-tcml)VO(ma)2 1.979 1.973 1.938 55 60 169VO(ka)2 1.979 1.973 1.938 58 62 170NH4 1.979 1.973 1.938 55 60 169[V02(ma)]VO(H2O)5 1.986 1.986 1.934 70 70 180a± 0.001.b± 1 (l0- cm1).42Figure 3.12. Experimental and simulated anisotopic ESR specua of VO(ma)2.X.Band ESR110KVO(ma)2ExperimentalSimulated-4250G43In general, the ESR spectra of V02+ complexes are strongly dependent on the donoratoms bound to the metal; e.g. they are different for nitrogen and oxygen donors. There arealso observable, but smaller, differences with different types of ligands that contain thesame donor atom. An analysis of the spectral parameters for these vanadium-pyronecomplexes, Tables 3.3. and 3.4., reveals a distinct similarity between parameters. It is firstnoted that the g values remain invariant between the maltolato and kojato complexes, eventhough the latter has an electron-withdrawing hydroxymethyl group. In general, this hasbeen found for other ligand types in oxovanadium complexes. For example, fluorine orphenyl substitution does not affect the g values of -diketone, -ketimine and -diketonatecomplexes.95’°7”8The hyperfine splittings , A and are more sensitive tosubstituent effects, as evidenced in the 1 - 3 x104 cnr1 change upon hydroxymethylsubstitution for VO(ka)2relative to VO(ma)2. It is noted that, except for the most powerfulelectron-withdrawing or donating groups, the changes in Ac, in oxovanadium(IV)complexes upon substitution are usually small.95 Chasteen has correlated ESR spectraldata with ligand type for a variety of comp]exes. The determined spectral parameters aresimilar to those previously determined for 04 ligand coordinated oxovanadium(W).In a preliminary experiment, the blood plasma of a V0(ma)2treated S1’Z-diabetic ratwas investigated at -110 K by ESR spectroscopy. As blood plasma contains manyparamagnetic impurities, particularly Mn2 whose ESR signal appears in the region ofinterest, it was necessary to improve the spectrum to determine if any vanadyl signal couldbe observed. This was achieved by subtraction of the ESR spectrum of the blood plasmafrom an STZ-control (i.e. diabetic, but V0(ma)2 untreated) rat from the experimentalspectrum. The final spectrum (experimental minus control) is given in Figure 3.13., andclearly shows prominent V0 lines, in addition to those caused by Mn2. The signals arelikely to correspond to protein-bound vanadyl; the spectrum is annotated according tospectral lines expected for a vanadyl-rat transferrin complex.72-1”044Figure 3.13. Possible ESR spectral lines due to vanadyl-rat transferrin.(Subtraction of non-treated STZ-diabetic rat plasma spectrum from VO(ma)2treated STZ-diabetic rat plasma spectrum)2600 3000 3400 3800 4200 4600(6)453.4. S1V NMR SpectroscopyVanadium has a particularly attractive nucleus for NMR spectroscopy. However, itis only the d° (VV) oxidation state that is observed. Paramagnetic dt vanadium(IV) signalsdo not appear in the experiment; the large electron-nucleus hyperfine splitting constant forparamagnetic species causes large chemical shifts, and extreme line broadening.111Nuclear properties of 51y are given in Table 3.5.Table 3.5. NMR properties of the 51V nucleus.112Natural abundance 99.76 %Nuclear spinRelative receptivity (1H = 1.00) 0.38Chemical shift range relative to VOCI3 +200 to -1000 ppmVanadium has a relatively large shielding range that manifests itself in the sensitivityof the 51V nucleus to minor changes in its chemical environment. The chemical shift (6) isprobably the most important parameter discussed in 5V NMR spectroscopy, and isstrongly dependent on the electronegativity of the ligands surrounding the nucleus. Ingeneral, the electronegativity criterion can be used to distinguish between a nucleussurrounded by thio-ligands (6 -300 to +200 ppm) or oxo- or nitrogen ligands (6 -700 to-300 ppm).113 There is also a marked increase in shielding, and hence 6, observed whenintroducing steric strain via 5- or 6- membered chelate rings, with bulky ligands, and to alesser extent with increase in coordination number.114 Line widths for the vanadiumcomplexes encountered in these spectra are relatively broad, No correlation of line width toligand environment is noted in the literature, except the general observation that themagnitude of line width parallels ligand symmetry about the V’ centre, narrower in moresymmetric environments.57 Line widths are not noted in this study, apart from suchgeneralizations.46‘v NMR spectroscopy has been employed in many investigations of biologicalinterest. For example, the ‘V nucleus has been utilized as a metalloenzyme probe of thetransferrins,115”6to probe vanadate-protein interactions,112,1 13.1‘7”20422 and to investigatevanadate as a phosphate analogue. 51V NMR is the speciroscopic probe of choicefor investigating vanadate systems.As a reference for further studies, the hydrolysis behaviour of vanadate was studiedby pH titration. In acidic solution the dominant vanadium(V) species is “free” VO, andappears as a broad line at 8= -544 ppm (Figure 3.14.), indicating the low symmetry of thisspecies.l At pH 4 the dominant peaks appear at 8= 422, -497 and -514 ppm in a 1:2:2ratio and are all attributable to decavanadate. The structure of decavanadate consists of ten(VO)6 octahedra sharing edges, with two oxygen atoms in the interior of the moleculesurrounded by six vanadium atoms.126 The result is three different vanadiumenvironments, and hence three chemical shifts.114 Minor resonances are also seen at 6 =-559, -569 and -580 ppm, and are due to the following vanadate species: monomericvanadate (V1), tetrameric vanadate (Vt) and pentameric vanadate (V5) respectively.”4These are observed with greater intensity at pH 7, but are shifted slightly. The dominantspecies at neutral pH is V4, 8= -576 ppm; while a new peak corresponding to dimericvanadate, V2, appears at 8 = -572 ppm. All these peaks are relatively narrow and sharp,indicating high symmetry around vanadium in these complexes. At pH 9, the major speciesare V1 and V4, whilst V, dominates at pH 11. The deprotonation of these species accountsfor the variation in chemical shift for V, (6= -559 to -536 ppm) and V2 (8= -572 to -560ppm). Chemical shifts are summarized in Table 3.6., and are in good agreement (±1 ppm)with those obtained by Pettersson er a!. in an extensive study using a combination ofpotentiometric titrations and 51V NMR spectroscopy.127I ESR spectroscopic utrations in the acidic regime for this complex cearIy show that reduction to avanadium(TV) species also occurs, and indeed V(IV) is the dominant oxidation state.47vipE____________2.1.0JI: 9.0__7.0_10_AaJL 4.0________________2.0••—400 -440 -480 -520 —560 -600 6 (ppm)Figure 3.14. ‘V NMR specta of a pH titradon of vanadate (10 mM, 0.15 M NaC1, 25•C under Ar).48Table 3.6. Chemical shifts (6) of vanadates present in aqueous solution (Data obtainedin this study).Vanadate Proposed formula 8 (ppm) pH rangeV1 HV042- -559to -536 3-12V2 HV2073 -572 to -560 5 - 10V4 V4O12- -576 5-9V5 V5015- -584 5-9V10 V1oO6- -422 -497 -514 3-6VO2 -544 1-4The reaction of vanadate with increasing concentrations of maltol was studied inorder to examine the nature and stoichiometry of the species formed in solution. To a 10mM solution of vanadate at pH 7,* were added increasing equivalents of maltol, and the‘v NMR spectra observed (Figure 3.15.). With no maltol added, the only observablespecies are the vanadate oligomers, Vi (6= -559 ppm), V2 (8= -572 ppm), V4 (8= -576ppm) and V5 (6= -584 ppm). For M:L ratios of 1:’,2 to 1:10, the gradual appearance ofpeaks due to increasing complexation, nominally assigned as peak a (6= -496 ppm) andpeak b (8= -509 ppm), are observed with higher ligand concentrations. There is no changein the chemical shifts of the vanadate oligomers, and no other resonances are detected. Asthe ligand concentration is increased, and the metal-ligand complexes are formed inincreasing amounts, one can observe the concomitant reduction in the intensity of the peakassigned to V4 (and to a lesser extent V5). These observations indicate that peaks a and bare due to vanadate complexation with maltol.§ From ESR observations, reduction to vanadium(IV) does not occur above p114.49aJi1____ _______1:21:1——1:0.5_________jV4_____________3:0flIl1j14lII4IiI II1ipiiII.I.pflIpIIIpI(IpIIIJIIIIjIIIIeIIoIIjI IlIIIipiI1.iIipI..1 ,,WJ1i .j-400 440 -480 -520 -560 -600 ‘6 (ppm)Figure 3.15. ‘V NMR spectra of the reaction of vanadate (10 mM) with increasingconcentrations of maltol (0 mM, 5 mM, 10 mM, 20 mM, 50 mM, 100 mM.Conditions: 0.15 M NaC1, pH 7,25 ‘C, under Ar).50The concentrations of the various vanadate species observed in the 51V NMR spectrawere determined by integration. The NMR sample contained a known total amount ofvanadate; hence, having first calculated the relative ratio of each integrated resonance, theconcentration of each species could be determined, and is equal to the product of totalvanadate concentration and the relative ratio of that integrated peak. Following the methodof Tracey2t’124 the signal corresponding to a, 8= -496 ppm, was analyzed for maltolstoichiometry; the integrations of peak b were too small to determine stoichiometry withany degree of accuracy. Using this method, a 1:2 vanadate:ligand stoichiometry isdetermined for resonance a.The reaction of vanadate and maltol at increasing vanadate concentrations wasstudied by ‘v NMR, the spectra obtained shown in Figure 3.16. The total maltolconcentration was 10 mM and the pH was 5.5, chosen to study the interaction of thecomplexes with decavanadate. With a large excess of ligand (metal:ligand ratios of 1:20 to1:5) the only observable resonance is peak a (6 = -496 ppm), which further supportsevidence that a 1:2 complex is the major vanadate-maltol species. For a metal:ligand ratioof 1:2, peaks a, b (6= -508 ppm) and V1 (8= -560 ppm) are observed. For a 1:1 ratio,appreciable vanadate formation is noted; V1,V4,V5 and V10 signals are seen in their usualpositions. In excess vanadate (5:1 ratio) peak b is not observed, and peak a appears as ashoulder on the 8 = -497 ppm decavanadate peak. A concentrated aqueous solution ofNH4[V02(ma)]yields a lone resonance at 8= -495 ppm. Studies in non-aqueous solventsfurther support evidence that the [V02(ma)] anion persists as the major species of asolution ofNH4[V02(ma)2]; Peak a is therefore assigned to anionic [V02(ma)2].51aV0 lviAb 1 N:Z.JL5:11:1-—--1:21:51._____________________________________________________________________—1:20InIj I IjIII IIIII4IIIIIIIIIjII ,.ie.J..-400 -440 -480 -520 -360 -600 5 (pp)Figure 3.16. ‘V NMR specua of the reaction of maltol (10 mM) with increasingvanadate concentations (0.5 mM, 1 mM, 2 mM, 5 mM, 10 mM, 50 mM.Conditions: 0.15 M NaCI, pH 5.5,25 C, under Ar).52The metal-centered oxygen atom transfer reactions of vanadium have been likened tothose of organic functional groups: VOL2-OR (R =aikyl) to an organic esteT, L2OV-O-VOL2 to an acid anhydride.1”Thus, an “anhydricle” can be esterified, acylated or basehydrolyzed. A number ofNH4[V02(ma)]alkylated derivatives have been synthesized andcharacterized (i.e. VO(OR)(ma)2,R= Et, Me, iPr).45 In addition, the formation of adinuclear species, [(ma)0V-O-VO(ma)],and a protonated complex, VO(OH)(ma)2havebeen proposed in organic solvent solutions of these derivatives.45 51V NMR chemicalshifts of these complexes are summarized in Table 3.7.The aqueous chemistry of NH4VO2(ma)]was studied by variable pH 51V NMR;the spectra are shown in Figure 3.17. In pH 2 solution two very broad resonances areobserved at 8 = -460 ppm and 6= -510 ppm. As the pH is raised to 3.5, the two peaks haveincreased in intensity, and are located at 8= -483 and -509 ppm. The integrated peak areasindicate that the two species are in approximately equal proportion at this point. Theposition and nature of these two signals indicate that they are due to vanadate-maltolcomplexes, as they lie outside the ranges of vanadate hydrolysis products. The chemicalshifts of the two peaks continue to coalesce until the pH is raised to 4.5, upon which thesignals are seen at 8= -496 and -509 ppm, corresponding to peaks a, anionic [V02(ma)]-,and b in the previous investigations. A minor peak at 8= -519 ppm is also observed at thispoint, which is assigned to decavanadate (8 = —-520 ppm). From Figure 3.16. it ispresumed that the 6 = -497 ppm decavanadate peak is located under, and contributes to,resonance a. In the neutral pH region, vanadate hydrolysis products appear in the spectra,and are assigned to Vi (8= -560 to -537 ppm), V2 (6= -570 to -560 ppm) and V4 (6= -576ppm). For pH >3.5, resonance a grows in intensity, whilst peak b decreases. Above pH 8,V1 begins to dominate to the detriment of the other signals. In basic solution, V1 is themajor species.53The partial protonation of the maltol ligands in the coordination sphere of vanadiumand their displacement by H20 as the solution is acidified result in the formation of[VO2(ma)(HO ](x = 1 or 2) which, in the acidic regime, can be assigned as a contributorto resonance b (6= -509 to -529 ppm). This is implicated by the relative intensities of thepeaks as the pH is reduced. These observations are in agreement with those previouslyobserved in the variable pH 27A1 NMR spectra of AP’ complexes with pyrone andpyridinone Iigands.’30”’In addition, from chemical shifts noted in Table 3.7, it is alsopossible that in acidic solution the resonance has some contribution from the protonatedspecies [VO(OH)(ma)2]. In acidic solution the resonance is very broad, and maycorrespond to overlap from several peaks. It is also noted that in neutral to basic solutionresonances a and b do not change in intensity, under conditions in which neither a monoligand or protonated complex would be expected to exist. Therefore, it is postulated thatpeak b is due to another species in this pH region; a possibility being another isomer of cisbis(maltolato)diooxovanadate(V), with one maltol ligand inverted by 180 so that ahydroxy oxygen is trans to a V=O (see Scheme I). In effect, several distinct vanadatespecies have a similar chemical shift, 8 = -509 ppm, that manifest themselves underconditions in which formation of that species is favoured. Further studies are necessary tounequivocally assign this resonance with increasing pH. Chemical shifts of these vanadatemaltol complexes are given in Table 3.7.It has been shown using ESR that an appreciable amount of NH4[V02(ma)]isreduced in acidic solution to VO(ma)2 (Section 3.3.). Therefore there will be an increase inparamagnetic vanadium(IV) complexes at low pH. The variation in chemical shifts ofpeaks a and b observed with increasing acidity in the pH study is attributed toparamagnetic shifting. The lower the pH, the greater the degree of reduction to V(JV)complexes, hence the greater concentration of paramagnetic species in solution and themore exaggerated the variation in the chemical shifts.54pE______ _______12.0__ __ __ _ __ __2.0—400 —440 —480 —520 —560 —600 (ppm)Figure 3.17. Variable pH V NMR spectra ofNH4[V02(rna)] (5 mM, 0.15 M NaC1,25 ‘C, under Ar).55Table 3.7. Chemical shifts and assignments for vanadate-maltol complexes formed inaqueous solution. (0.15 M NaC1, 25’ C).Chemical Shift (6) Assignment45-509 ppm anothercis-[V02(ma)]-fand/or[V02(H0)13ma))and/or [VO(OH)(ma)]-496 ppm cis-[V02(ma)]-420 ppm [VO(OMe)(ma)]*-470 ppm [(ma)0V-O-VO(ma))t i.e. a cis-[V02(ma)2]- isomer, with a hydroxy oxygen trans to a V=O.§ i.e. in MeOHIn a similar experiment the variable pH 51V NMR spectra of VO(ma)2 were studied.In basic solution a peak is observed at 8= -537 ppm corresponding to V1 (Figure 3.18.). Asthe pH is lowered, peaks at 8= -496 and -510 ppm appear. These clearly correspond topeaks a and b in the previous study. As the acidity of the solution is increased, V1 isshifted to -560 ppm, and disappears below pH 5.5. In acidic solution, peaks a and bseparate and are paramagnetically shifted. No spectra are observed below pH 2. Theseresults prove that the same vanadium(V) complexes form in a solution of V0(ma)2, evenunder relatively acidic conditions. The most likely mechanism for the formation of thesevanadate complexes is by the oxidation of VO(ma)2 with 02 to form [V02(ma)2]-, andsubsequently vanadate oligomers, as implicated in Section 3.1. Comparison of the 51VNMR spectra obtained in the variable pH studies of VO(ma)2 andNH.4[V02(ma)](Figures3.17. and 3.18.) reveals that the spectra are nearly coincident at identical pH values. Theformation of the V1 vanadate oligomer is somewhat more favoured in the case of V0(ma)2however.56Iv’ab I:*t 10.0S .0-7 • 05.5453 • 5.Li1Il 2 • 0I’I.II..lIiII1.nhpII4IIIiIIl uIj!lIrj.I 41141141 I’14t’’I’J’’4’.,.II’’’’l..n.j-400 -440 -480 -520 -560 -600 8 (ppm)Figure 3.18. Variable pH V NMR specta of VO(ma)2(10 mM, 0.15 M NaCI, 25 C).57The 5V NMR spectra obtained in the corresponding variable pH study of VO(ka)2are very similar to those observed for VO(ma)2 (Figure 3.19.). In this case however signalintensity is appreciably less, as the complex is less water soluble. A broad peak centredaround 8= -495 ppm (&) and a shoulder at 6 = -510 ppm (W) were detected, in addition toV1 peaks. The peaks corresponding to vanadate-kojate complexes occur in a similarposition to those observed in the mahol systems. This is not entirely unexpected, as theimmediate 51-vanadium nuclear environment is very similar in both complexes (i.e. an 05,6donor set). Corresponding ESR spectral parameters for the two ligands are also verysimilar. Tracey et al has observed similar chemical shifts in similar positions in the 51VNMR spectra of a number of carboxylate ligands.’ In addition, the signals observed inthe experiment are broad, and have a low signal/noise ratio, and hence are more difficult toassign. Comparison of the variable pH 51V NMR spectra of VO(ka)2and NH4[V02(ma)](Figures 3.17. and 3.19.) reveals that the spectra are nearly identical at corresponding pHvalues.In a preliminary investigation a blood plasma sample obtained from a V0(ma)2treated STZ-rat was studied by 51V NMR spectroscopy. The spectrum (not shown) shows alone resonance at 8= -555 ppm, corresponding to V1. Interpretation of this result must betreated with caution, as the number of scans necessary to obtain a signal (100,000 scans)necessitates a lengthy time period in which the sample was exposed to air. Hence, thesignal could be due to vanadyl to vanadate conversion in vivo, or may be due to airoxidation of vanadyl in the sample. Further studies are necessary to help elucidate themechanism and cause of oxidation. It is significant, however, that signals due tovanadium(V)-transferrin, which typically has resonances at 8 = -529 ppm and -531ppm,115’6 were not observed.The various equilibria descibed in this section are summarized in Scheme I.580pB10 . 08.0____________ ______— ______2.0I .I4.-400 ‘440-480 -520 -560 -600 6 (ppm)Figure 3.19. Variable pH V NMR spectra of VO(ka)2 (10 mM, 0.15 M NaC1, 25 C).59CH3oJOHl+ 0H 0ZoJ.OHH’?.5O9ppjCH3oJo 7- O_LO%114 I -CH3°,,f_________\= -$O9ppm 7/” [6= 496ppm) wCH3 / .H20OZ0\/0o’L0s%IICH %..dLy0’OH2OH2rCH3 ROH.5O9ppmOJ,ORoil 4CH3\0CH3 o H3o0,i0b[=42Opj[ESR SignScheme 1. Equilibria, V NMR chemical shifts and possible routes of fomadon forvanadium(V) and vanadium(IV) complexes with the maltolate anion. (0.15M NaCI, 25 C).603.5. Stability Constant DeterminationsStability constants are essentially equilibrium constants for metal complexformation, and have long been employed as a quantitative indication of the affinity of aligand for a metal ion in solution.63 With the relatively recent introduction of powerfulcomputational methods for the analysis of titration data,75’132 the accuracy of calculatedconstants have considerably improved. Although there is still some concern regarding thevalidity of all published work,133 provided a careful and thorough experimental techniqueis employed, such as that described herein, reliable and reproducible data can be obtained.Stability constants are frequently quoted in terms of stepwise, or successive, protonationconstants for the ligand and either overall or successive formation constants for the metalligand complex. The relationship between successive (K) and overall (f) formationconstants, is that the latter are a cumulative product of the former. Due to the computerprograms employed for data reduction (BEST, PKAS (Chapter 2.4.4.)) stability constantsquoted in this study are generally given in terms of the overall stability constants, J3.Fundamentally, a stability constant describes an equilibrium quotient involving theactivities of reacting species in solution at equilibrium. This is often defined as the ratio ofproducts of the activities of the reaction products, each raised to the appropriate power, tothe product of the activities of the reactants. The determination of the activities of thesecomplex ionic species is both complicated and time consuming. However, it has beenobserved that concentration parallels activities when the ionic strength is regulated by anon-reacting electrolyte in great excess.56 Although some physical chemists haveoccasionally questioned the validity of the “concentration constant” approximation, it hasbecome common practice to measure equilibrium constants as a quotient of concentrationsat constant ionic strength maintained by a supporting electrolyte.63 The two most commontechniques employed in stability constant determination are potentiometry and61spectrophotometry. Spectrophotometry is applicable to studies performed at lowconcentration or pH, making it useful in cases where ligands are sparingly soluble, and canbe used with almost all systems, even proteins and macromolecules. However, the relativeease of use and applicability of data obtained for computation make potentiometry the mostattractive, and popular, technique.132The ligands maltol and kojic acid contain the 3-hydroxy-4-pyronate moiety as thecoordinating group to the metal centre, and have been investigated previously as ligandprecursors for Lewis acids. When deprotonated both of these ligand precursors behave asbidentate Lewis bases, which form relatively stable, neutral, tris-ligand complexes withgroup 13 metals such as aluminium(lll).1’Therefore they are expected to form relativelystrong complexes with the hard Lewis acid VO. The positioning of a hydroxy group inan a-position relative to a keto group is found widespread in many naturally occurringcompounds, and makes these ligands attractive as potential physiological ligands.46Comparison with the analogous 3-hydroxy-2-methyl-pyridinones shows that the pyronesare stronger acids (pyridinone pKa’s —9- 10; pyrone piCa’s ‘7 8).135The remarkable ease of hydrolysis of vanadyl chelates, and the potential foroxidation and/or hydrolysis of uncoordinated vanadium(IV), has limited the number ofstudies on vanadyl chelate formation constants. Indeed, a survey of the literature reveals averitable dearth of information, and the accuracy of the results obtained in earlier studiesmust be questioned (due to outmoded computational methodology, poor experimentaltechnique, and lack of consideration for the complexity of the system). Vanadyl chelatecomplexes hydrolyze as, or more, easily than any of the other metal chelates of the firsttransition series, and often prefer to combine with hydroxyl ions rather than a second moleof ligand. However, in comparison, vanadyl chelates are often found to be more stable thanother transition metal chelates. This may be a consequence of the influence of the highlyelectronegative oxygen bound to vanadium; the resulting higher electropositive character of62the vanadium would make it a stronger Lewis acid, and lead to more stable chelates.Thermodynamic studies have been scarce, but indicate a high +AS’ value as a majorcontributor to chelate stability in these comp1exes.’Vanadyl complexes with catechol as the ligand have been thoroughly investigated,and values of log = 16.8 and log 1,O,2 = 31.5 determined.81’137 These representparticularly high vanadyl complex stability constants, and bis(catecholate) species arefound to be stable to hydrolysis even at pH >12. The formation of vanadyl chelates withphenolic and/or carboxylate containing ligands yield typical values of log i,oj = —5-12 andlog = _1220.70,138.140 These figures are representative of the range of determinedstability constants. In a series of papers, Gilard et al 78,1411U have determined stabilityconstants of various vanadyl-amino acid complexes, that fall into the range given. Thesestudies are noteworthy owing to the experimental scrutiny and thoroughness of the authors.Similarly, the work of the Umeâ group on various vanadate systems represents a highstandard of data collection and computation.’45’146 However, when comparing reportedresults it is prudent to remember that the type of supporting ionic media, the ionic strengthand the solution temperature all contribute to the calculated equilibrium constants;comparison between values determined under different conditions is only indicative, notprecise. Smith and Martell have collected a library of stability constant data, for a varietyof metal-ligand complexes.147 As previously mentioned, it is the formation ofoxovanadium(IV) hydrolysis products in aqueous solution that makes assignments of thecomplexes formed difficult. The nature of species present in neutral or basic solutions isstill somewhat controversial, but it is apparent that hydrolysis products predominate withincreasing basicity, and should be included in any equilibrium model. In acidic solution“free”, unhydrolyzed, VO(H2O)5 predominates. With increasing basicity the speciesVO(OH), logf3i,.i.o = - 5.791, and (VO)2OH,log = - 6.70, form in neutral pH.¶ log ‘s of the vanadyl hydrolysis species are quoted re-calculated at experimental conditions (0.15 MNaCI, 25 ‘C) by the author according to the method of Bacs and Mesmer.3363Two VO2:OH stoichiometries have been proposed to exist in the pH range 5 - 11; (i) 2:5and (ii) 2:6. Assuming [(VO)2(OH5)-predominates,27its formation constant isj3s,o = -22.06. In the basic regime, the monomeric species [VO(OH)3]predominates, log 1,.3,O =—18 (not included in BEST calculations; this species is certainly important for pH>1 1).26.273(i) Vanadyl-maltol systemThe protonation constant of maltol was determined by PKAS analysis: log oj,i =8.42 ± 0.01 (0.15 M NaC1, 25 C). Literature values determined for the pICa of maltol are8.62 (0.6 M NaC1, 25 ‘C)’34 and 8.38 (0.1M KNO3,25 ‘C).’47 A representative titrationcurve for the vanadyl-maltol system is given in Figure 3.20. The curve corresponding tothe vanadyl-maltol titration is shifted from the maltol ligand curve by about 2 units,indicating two moles of ligand consumed in the vanadyl titration. The sharp pH increase ata = 2 in the vanadyl-maltol curve also suggests formation of a 1:2 complex. A bufferregion appears between pH 8 - 10, which is indicative of the formation of hydrolysisproducts.”32 The thermodynamic reversibility of the system was investigated by titrationwith standard HC1 solution, as depicted in Figure 3.20.; the results of the forward andreverse titrations are in good agreement, with the two curves almost superimposable.6412-10-8-+CIII;’4.20-2-101 2345a—0-— Maltol-G V02+ - MaltolReverseFigure 3.20. Titration curves obtained from the potentiometric study of the system: V02+- maltol. Conditions: 0.15 M NaC1, 25 C, metal:ligand =1:2. (a =(molesbase added-moles acid present)/moles ligand present).The vanadyl-pyrone system is quite complicated, and calculations involvingrefinement of values of all possible species that may form (i.e. x, y and z quitegeneral) gave results with no physical meaning. The following strategy was employed.The species log 3i,o,i and log D1,o.2 were refined initially, with the appropriate vanadylhydrolysis constants, ligand protonation constant and pK.t33 Other plausible species weregradually introduced into the refinement, and were discarded if the fit worsened.Combinations of plausible species were refined, until finally no improvement in fit couldt pKw refers to the water dissociation constant. pKw =13.765 ,as calculated by Baes for conditions ofO.15M NaC1, 25C.33fr65be realized. The results of the refinement using the model generating the lowest a, andhence the best fit, are given in Table 3.8.Table 3.8. Composition and log f3 values for species in the vanadyl-maltol system, asrefined in the model of best fit as determined by BEST data reduction.&Composition x, y, z log ,[VO(ma)] 1 0 1 8.49(2)[VO(ma)2] 1 0 2 15.10(4)[VO(OH)(ma)] 1 - 1 1 1.36 (5)[VO(OH)(ma)2]- 1 - 1 2 5.51 (7)[VO(OH)(ma)]- 1 -2 1 -4.57 (4)[(VO)2(OH)2(ma)2] 2-2 2 5.29 (1)a The number in parentheses refers to the standard deviation between successive runs.Log 1 quoted as the mean value.(ii) Vanadyl-kojic acid systemAnalysis of the kojic acid pKa titration data gave log 3, =7.77 ± 0.01 (0.15 M NaC1,25 C). Literature values determined for the pKa of kojic acid are 7.75 (0.1 M KNO3,25‘C)’47 and 7.61 (0.6 M NaC1, 25 .C).1MA representative titration curve of the vanadyl-kojate system is given in Figure 3.21.The figure shows a sharp inflection in pH at a =2 for the vanadyl-kojic acid titration curve,and a two unit difference between that and the curve for kojic acid alone. This indicates theformation of a 1:2 complex. The thermodynamic reversibility of the vanadyl-kojatetitration was checked by reverse titration, with good agreement between forward andreverse curves. Data analysis was performed using the same procedure as for the maltolsystem, and the results of the refinement are given in Table 3.9.661:.Qo I II 22 4 6—0-— Kojic Acid--.0-- V02+ - Kojic AcidReverseFigure 3.21. Titration curves obtained from the potentiometric study of the system:VO2-kojic acid. Conditions: 0.15 M NaC1, 25 C, metai:ligand 1:2. (a=(moles base added - moles acid present)/moles ligand present).Table 3.9. Composition and log f3 values for species refined using BEST for thevanadyl-kojic acid system.Composition x, y, z log f3 ,,[VO(ka)] 1 0 1 9.88 (1)[VO(ka)2] 1 0 2 16.37 (5)[VO(OH)(ka)] 1 - 1 1 1.36 (5)[VO(OH)2(ka)]- 1 -2 1 4.57 (4)674-3- —-3(iii) NH4[V02(ma)JGiacomelli and co-workers have determined that it is possible to protonate the oxogroup in cisbis(ligand)dioxovanadates.129 To this end, NH4[V02(ma)]was investigatedby potentiometric titration, a representative curve shown in Figure 3.22. The protonationconstant for the complex was determined by the Henderson-Hassleblach equation: log ni,..1.2 = 6.83 (6). This value is comparable to that obtained by Giacomelli et al, log6.3, for bis(8-quinolinato)dioxovanadate(V). A second protonation constant for their anionwas also determined (by acid solubility measurements) as log , = 3.4. It was notpossible to determine a value for a second protonation constant in this study, either throughpotentiometric or spectrophotometric means. However, this does not entirely discount thepossible formation of such a species at pH <3.9-8-7.—+I.;’ 5—o-—- Forward —a-— ReverseFigure 3.22. Titration curve obtained from the potentiometric determination of the (first)protonation constant forNH4[V02(ma)](0.15 M NaC1, 25 C, under Ar).I I-2 .1 0 1 2 3a68The refined stability constants in Tables 3.8. and 3.9. offer a plausible hypothesis forseveral of the species present in each solution. The 1:1 and 1:2 complexes can beconfidently defined, and show low standard deviations. Species important above pH >6,where appreciable hydrolysis occurs, show higher standard deviations, and are harder toascertain. Of considerable importance to the accuracy of the model is the nature of thevanadyl hydrolysis species in neutral to basic conditions. The uncertainty in the valueattributed to Pz-5,o is certainly a concern for investigators;27 however, refinementsattempted without this constant gave a very poor fit. It is apparent that hydrolysis, of bothfree and bound vanadium, occurs in these regions, as proposed by the visible spectroscopy,‘v NMR and ESR studies presented herein. The inclusion of the f2.-s.o hydrolysisconstant in vanadyl-ligand stability constant determinations is a practice observed in otherlaboratories.’48The metal-ligand equilibria for the non-hydrolyzed species in these systems mayadequately be described by equations 3.3 and 3.4.V02+ + L- VO(L) f3 (3.3)VO(L) + L- -. VO(L)2 Pi (3.4)— 1(VO)L,(2)i— [VO][LjThese pyrone ligands are found to have a reasonably high affinity for vanadyl,evidenced by the determined log 3 values. In comparison, the results show that thesepyrones have similar stabilities to those found for other phenolic and/or carboxylatecontaining ligands, i.e. log = 5.59 for (malonato)oxovanadium(IV), log Di,o, = 9.48for bis(malonato)oxovanadium( V).’47 A comparison between the two ligands reveals that69the acidity of maltol (pKa = 8.42) is somewhat lower than that of kojic acid (pICa = 7.77).This difference may be attributed to the inductive effect of the methyl group in maltol.Vanadyl binds most effectively to hard electronegative atoms, i.e. F, Cl, 0 and N.Fluoro and oxygen donor containing compounds are especially stable. In general,complexes with oxygen as the donor atoms are found to follow the expanded Irving-Williams series.’49 The presence of one or two nitrogen donors places VO lower in theseries.27 When the stability constants of these vanadyl-y-pyrones are compared to thestabilities of other divalent metal-y-pyrone complexes, the resulting series is indeed foundto follow the order: V02+ >C’j2>2+ >Ni2>Co >Mn2. in addition, it is also notedthat the magnitude of the stability constants previously determined for divalent metalkojato complexes are in general larger than those for the analogous maltolato complexes,observations that are also noted in this study.t 136,147,150453 Presumably, it is the effect thatthe hydroxyl group in kojic acid has upon the stability of the metal-ligand complexes, incomparison to the methyl group of maltol, that is responsible for the slight difference in thestability constants for the two ligands.Speciation diagrams are an effective indicator of the nature of the complexes formedin solution, although they are by no means a substitute for spectroscopic investigation.Information garnered from such diagrams must be treated purely as a gauge for solutionspeciation. The quality of information is directly proportional to the quality of the stabilityconstant data from which they are constructed. In this example, the poorly understoodvanadyl hydrolysis, and the nature of minor hydrolysis species, questions the completevalidity of the proposed diagrams in the basic regime. Speciation diagrams for complexformation in the vanadyl-maltol system are shown calculated at mM vanadyl concentrationfor a 1:2 metal-ligand system from stability constants obtained in this study (Figure 3.231).t However, it must be noted that several of these studies were performed before the introduction of newcomputational methods, or contain results obtained under less than rigorous experimental conditions.Hence, these trends are indicative at best.70In highly acidic pH, free V02+ is observed. The log i,o,i complex dominates in the regionpH 2- 6; the log Di,o,z complex begins to form by pH 4, and has a maximum“concentration” at pH —7. This has obvious ramifications for the chemist interested inphysiological modeling. In basic solution hydrolyzed complexes, and vanadyl hydrolysisproducts, dominate. The diagrams succinctly demonstrate the effects of concentration oncomplex formation. At low concentrations in neutral solution, hydrolysis products are ableto effectively compete with the ligand for vanadyl. The result is that very little of the 1:2product is obtained at this concentration level. In the basic regime, hydrolysis productsdominate, as proposed from results presented herein.In the corresponding vanadyl-kojic acid systems, Figure 3.23.ii, similar observationsare made. However, in this case the higher overall stability of the kojato complexes isdemonstrated by the dominance of the log (31,o,i and log (31,0,2 curves. The difference instability of the maltol and kojate complexes is manifested in a less dramatic concentrationdependence for complex formation in the latter case.An obvious flaw in the computer software currently available for determiningstability constants is the lack of a provision to include an electron as a fourth variable in thedata computation and curve fitting procedures. Until this is possible, and a completeunderstanding of vanadyl hydrolysis is reached, the absolute dependibilty of data obtainedfor species in the neutral to basic regime will remain questionable.71100- --a-Iog[H+](i)100-St.. C5 0 Se ,25- / ‘.S. ‘ —— _,S0- -. —1 2 3 4 5 6 7 8.tog[H+](ii)Figure 3.23. Speciation diagrams forVO2-Ligand (1:2, 10-2 mM VO). (I) Maltol(ii) Kojic Acid. Conditions: 1=0.15 M NaCI, 25 C, under Ar.a. VO, b. [VO(L)], c. [VO(L)2],d. [VO(L)2(OH)], e. [(VO)2(OH5].72Chapter 4Suggestions for Future WorkThe results presented herein clearly show that facile processes occur for theinterconversion between VO(ma)2and anionic [V02(ma)1 in aqueous solution. This isevidenced by ESR signals observed in solutions of ESR silent NH.jV02(ma)]and VNMR signals in solutions of NMR silent VO(ma)2. Dissolved oxygen is the presumedoxidant in the conversion of VO(ma)2 to [V02(ma)],as shown by a retardation of theprocess in solutions maintained under an inert atmosphere. An investigation of themechanism and kinetics of this oxidation are appropriate subjects for future study Thisoxidation process has obvious implications in vivo as dissolved oxygen is prevalentphysiologically, particularly in plasma. Therefore it is assumed that a similar process willoccur in the mammalian system.The similarities in the chemistry of the (bismaltolato)oxovanadium(IV) and(biskojato)oxovanadium(IV) complexes is demonstrated in their similar behaviour andvisible, ESR and ‘V NMR spectral properties.Recent studies in collaboration with Dr. 3. H. McNeill of the Faculty ofPharmaceutical Sciences at UBC have concluded that while VO(ma)2and VO(ka)2displayfavourable insulin mimicking properties when administered to STZ-treated diabetic rats,the vanadate analogue NH4[V02(ma)]does not, and can indeed be severely toxic. Thisappears surprising in view of the similar, pH dependent, chemistry reported herein. As thepH of the mammalian system changes dramatically from stomach to intestines, it isapparent that in vivo investigations are necessary to determine the physiological site ofabsorption, and the oxidation state during absorption. ESR, and less likely 51v NMR, asdemonstrated in this study, are suitable probes for studying the redox behaviour, ligandenvironment, and oxidation state of the compounds in the blood and organs of treated73animals, and could be effectively employed in this case. As both ESR and NMR can beused to monitor the diffusion of molecules acmss membranes and red blood cells,154 theycould be the basis for model studies into absorption of vanadyl-complex molecules fromthe gastrointestinal tract into the blood.74References(1) Jandhyala, B. S.; Hom,G. 3. Life Sciences 1983,33, 13254340.(2) Chasteen, N. D.; Lord, E. M.; Thompson, H. 3. In Frontiers in BioinorganicChemistry; A. Xavier, Ed.; VCH Publishers: 1986;133-141.(3) Smith, M. 3. Experienda 1989,45,452.(4) Byrne, A. R.; Kosta, L. Sci. Total Environ. 1978,10, 17-30.(5) Bayer, W. F.; Kneifel, H. Z. Natwforsch. 1972,27,207.(6) Vilter, H. Phytochemisrry 1984,23, 1387.(7) Cornman, C. R.; Kampf, 3.; Soo Lah, M.; Pecoraro, V. L. bzorg. Chem. 1992,31,2035-2043(8) Cantley, L. C.; Josephson, L.; Warner, R.; Yanaqisawa, M.; Lechene, C.; Guidotti,G. JBiol. Chem. 1977,252,7421.(9) Shechter, Y. In Vanadium in Biological Systems; N. D. Chasteen, Ed.; KiuwerAcademic: Dordrecht, 1990;100-128.(10) Biggs, W. R.; Swinehart, 3. H. Metal Ions Biol. Syst. 1976,6, 142-196.(11) Kustin, K.; Macara, I. G. Comments Inorg. Chem. 1982,2, 1-22.(12) Chasteen, N. D. Struc:. Bond. 1983,53, 105-138.(13) Boyd, D. W.; Kustin, K. Adv. Inorg. Biochem. 1984,6,311-365.(14) Nechay, B. R.; Nanninga, L. B.; Nechay, P. S. E.; Post, R. L; Grantham, J. 3.;Macara, L 0.; Kubena, L. F.; Phillips, T. D.; Nielsen, F. H. Fed. Proc. 1986,45,123-132.(15) Wever, R.; Kustin, K.Adv. lnorg. Chem. 1990,35, 81-115.(16) Rehder, D. Angew. Chem. mt. Ed. Engl. 1991,30, 148-167.(17) Butler, A.; Canano, C. 3. Coord. Chem. Rev. 1991,109,61-105.(18) Tracey, A. S.; Gresser, M. 3. Proc. Nat!. Acad. Sci. USA. 1986,83, 609-613.75(19) Gresser, M. 3.; Tracey, A. S.; Stankiewicz, P.3. Adv. Protein Phosphatases 1987,4,35.(20) Tracey, A. S.; Gresser, M. 3.; Liu, S. I. Am. Chem. Soc. 1988,110,5869-5874.(21) Tracey, A. S.; Galeffi, B. Inorg. Chem. 1989,28, 1726-1734.(22) Gresser, M. 3.; Tracey, A. S. In Vanadiumin Biological Systems; N. D. Chasteen,Ed.; Kiuwer Academic: Dordrecht, 1990;63-80.(23) Liochev, S.; Ivancheva, E.; Fridovich, L Arch. Biochem. Biophys. 1989,269, 188-193.(24) Secco, F. Inorg. Chem. 1980,19,2722-2725.(25) A1-Bayati, M.; Raabe, 0. G.; Gin, S. N.; Knaak, 3. B. .1. Am. Coil. Tox. 1991,10,233-241.(26) Henry, R. P.; Mitchell, P. C. H.; Prue, 3. E. I. Chem. Soc. Dalton Trans. 1973, 1156.(27) Vilas Boas, L. F.; Costa Pessoa, 3. In Comprehensive Coordination Chemistry; S.0. Wilkinson, R. D. Gillard and 3. A. McCleverty, Ed.; Pergamon Press: Oxford,1987; Vol. 3;453-584.(28) Chasteen, N. D. Biol. Magn. Res. 1981,3,53-119.(29) Hon, P. K.; Belford, R. L.; Pfluger, C. E. J. Chem. Phys. 1965,58, 1195.(30) Harris, W. R.; Carrano, C. J. I. Inorg. Biochem. 1984,22,201-218.(31) Harris, W. R.; Friedman, S. B.; Silberman, D. J. Jnorg. Biochem. 1984,20, 157-169.(32) Sabbioni, E.; Marafante, E. Bioinorg. Chem. 1978,9,389-407.(33) Baes, C. F.; Mesmer, R. E. The Hydrolysis ofCwions; John Wiley & Sons: NewYork, 1986.(34) Atkinson, M. A.; Maclaren, N. K. Sci. Am. 1990,263,62-71.(35) Shechter, Y. Diabetes 1990,39, 1-5.(36) Shechter, Y.; Karlish, S. 3. D. Nature 1980,284,556-558.76(37) Degani, A.; Shechter, Y.; Gochin, M.; Karlish, S. 3. D. Biochemistry 1981,10,2433-2438.(38) Heyliger, C. E.; Tahiliani, A. 0.; McNeil, 3. H. Science 1985,227, 1474-1477.(39) Ramanadham, S.; Mongold, 3.3.; Brownsey, R. W.; eros, 0.; McNeill, 3.11.American Journal ofPhysiology 1989,257, H904-H91 1.(40) McNeil, 3. H.; Yeun, V.0.; Hoveyda, H. R.; Orvig, C. J. Med. Chem. 1992,35,1489-1491.(41) Dai, S.; Yeun, V. 0.; Orvig, C.; McNeill, 3. H. Pharmacol. Comm. 1993,3,311-321.(42) Yuen, V. G.; Orvig, C.; McNeil, 3. H. Can. I. Physiol. Pharmacol. 1993,71,263-269.(43) Yuen, V. 0.; Orvig, C.; McNeill, 3. H. Can. I. Physiol. Pharmacol. 1993,71,270-276.(44) Paulson, D. J.; Kopp, S. 3.; Tow, 3. P.; Peace, D. G. I. Pharmacol. Exp. Ther. 1987,240,529-534.(45) Gelmini, L.; Glover, N.; HerringF.G.H.;Rettig,S.J.;Orvig, C. Manuscript inPreparation 1993.(46) Beelik, A.; Purves, C. B. Can. I. Chem. 1955,33, 1361-1374.(47) Zhou, Y. M.Sc. Thesis, University of British Columbia, 1993.(48) Kustin, K.; Pizer, P. Inorg. Chem. 1970,9, 1536.(49) Pettit, L. D.; Swash, 3. L. M. I. Chem. Soc. Dalton Trans. 1978,286.(50) Che, T. M.; Kustin, K. Inorg. Chem. 1980,19,2275.(51) Holland, R. M.; Tapscott, R. E. I. Coord. Chem. 1981, 17.(52) Fisher, D. C.; Barclay-Peet, S. 3.; Balfe, C. A.; Raymond, K. N. lnorg. Chem. 1989,28,4399.(53) Branca, M.; Micera, G.; Dessi, A.; Sanna, D. I. Inorg. Biochem. 1992,45, 169-177.77(54) Buglyó, P.; Kiss, T. I. Coord. Chem. 1991,22,259.(55) Rossotti, F. 3. C.; Rossotti, H. I. Chem. Ed. 1965,42,375-378.(56) Rossotti, H. The Study ofIonic Equilibria; Longman: New York, 1978.(57) Rehder, D. Magn. Reson. Rev. 1984,9, 125-233.(58) White, L. K.; Belford, R. L I. Am. Chem. Soc. 1976,98,4428(59) Gagne, R. R.; Koval, C. A.; Lisensky, G. C. Inorg. Chem. 1980,19,2854-2855.(60) Cacheris, W. P.;Franczyk, T. S. In Raymond Research Group: University ofCalifornia, Berkeley, 1990.(61) May, P. M.; Williams, D. R.Talanza 1982,29,249.(62) Bates, R. 0. Determination ofpH; John Wiley: New York, 1973.(63) Martell, A. E.; Motekaitis, R. Determination and Use ofStability Constajus; VCHPublishers: New York, 1988.(64) Motekaitis, R. 3.; Martell, A. E. Can. I. Chem. 1982,60,2403-2409.(65) Motekaitis, R. 3.; Martell, A. E. Can. I. Chem. 1982,60, 168-173.(66) Ortolano, T. R.; Selbin, 3.; McGlynn, S. P. J. Chem. Phys. 1964,41,262-269.(67) Selbin, 3. Chem. Rev. 1965,65, 153-175.(68) Selbin, 3. Coord. Chem. Rev. 1966, 1, 293-314.(69) Bafihausen, C. 3.; Gray, H. B. Inorg. Chem. 1962, 1, 111-122.(70) Tomiyasu, H.; Gordon, G. I. Coord. Chem. 1973,3,47-56.(71) Lever, A. B. P. Inorganic Electronic Spectroscopy Elsevier: Amsterdam,1986, p.260(72) Chasteen, N. D.; Grday, 3. K.; Holloway, C. E. Inorg. Chem. 1986,25,2754-2760.(73) Branca, M.; Micera, 0.; Dessi, A.; Sanna, D. I. lnorg. Biochem. 1992,45, 169-177.(74) Ferrer, E. G.; Williams, P. A. M.; Baran, E. J. I. Inorg. Biochem. 1993,50,253-262.78(75) Meloun. M.; Havel, 3.; Hogfeldt, E. Computation ofSolution Equilibria: A Guide toMethods in Potentiometiy, Extraction, and Spectrophotometry; Ellis Horwood Ltd:Chichester, 1988.(76) Geraldes, C. F. 0.; Casim, M. M. C. A. I. Inorg. Biochem. 1989,35,79-93.(77) Williams, P. A. M.; Baran, E. 3. Biol. Trace Eleme,u Res. 1993,36, 143-150.(78) Costa Pessoa, 3.; Vilas Boas, L F.; Gillard, R. D.; Lancashire, R. 3. Polyhedron1988,7, 1245-1262.(79) Wuthrich, K. Helv. Chim. Acca. 1965,48, 1012-1017.(80) Cooper, S. R.; Koh, Y. B., Raymond, KJ. Am. Chem. Soc. 1982,104,5092-5102.(81) Jezowska-Bojczuk, M.; Kozlowski, H.; Zubor, A.; Kiss, T.; Branca, M.; Micera, 0.;Dessi, A. I. Chem. Soc. Dalton Trans. 1990,2903-2907.(82) Cornman, C. R.; Colpas, G. 3.; Hoeschele, 3. D.; Kampf, 3., Pecararo, V.LJ. Am.Chem. Soc. 1993, 114, 9925-9933.(83) Saito, K. In Coordination Chemistry; Volume 20; D. Baherjea, Ed. 1980.(84) Porte, A. L.; Stewart, C. P.1. Chem. Soc. Dalton Trans. 1972, 1661-1666.(85) Ortolano, T.R., Selbin, 3., McGlynn, S.P. 1. Chem. Phys. 1964,41(1), 262-268.(86) Baithausen, C. 3.; Gray, H. B. Inorg. Chem. 1962,1, 111-122.(87) Selbin, 3. Can. 1. Chem. 1966,293-315.(88) Borovik, A. S., Dewey, T. M., Raymond, K. N.Inorg. Chem. 1993,32,413-421.(89) Bonadies, 3. A., Carrano, C. JJ. Am. Chem. Soc. 1986,108,4088-4095.(90) Israel, Y.; Meites, L. In Standard Potentials in Aqueous Solution; A. 3. Bard, R.Parsons and 3. Jordan, Ed.; Dekker: New York, 1985.(91) Bonadies, 3. A.; Butler, W. M.; Pecoraro, V. L.; Carrano, C. 3. Inorg. Chem. 1987,26, 1218-1222.(92) Collison, D.; Powell, A. K.; Eardley, D. R.; Mabbs, F. E.; Turner, S. S. Inorg.Chem. 1993,32,664-671.79(93) Bard, A. 3.; Faulkener, L. R. Electrochemical Methods - Fundamentals andApplications; John Wiley & Sons: New York, 1980.(94) Wertz, 3. E.; Bolton, 3. R. Electron Spin Resonance - Elementary Theory andPractical Applications; Chapman and Hall: New York, 1986.(95) Boucher, L. J.; Tynan, E. C.; Yen, T. F. In Electron Spin Resonance ofMetalComplexes; T. F. Yen, Ed.; Plenum: New York, 1969; 111-130.(96) Kivelson, D.; Lee, S.-K. 1. Chem. Phys. 1964,41, 1896-1903.(97) Wilson, R.; Kivelson, D. .1. Chem. Phys. 1965,44, 154-166.(98) Casey, 3. D.; Chasteen, N. D. I. Inorg. Biochem. 1980,13, 111-126.(99) Sakurai, H.; Hiram, 3.; Michibata, H. Biochem. Biophys. Res. Commun. 1987,149,411-416.(100) Sakurai, H.; Tsuchiya, K.; Nukatsuka, M.; Kawada, 3.; Ishikawa, S.; Yoshida, H.;Komatsu, M. J. Clin. Biochem. Nutr. 1990,8, 193-200.(101) Frank, P.; Carlson, R. M. K.; Hodgson, K. 0. Inorg. Chem. 1986,25,470-478.(102) Eaton, S. S.; Eaton, G. R. In Vanadium in Biological Systems; N. D. Chasteen,Ed.; Kiuwer Academic: Dordrecht, 1990;199-222.(103) Mustafi, D.; Telser, 3.; Makinen, M. W. I. Am. Chem. Soc. 1992, 114,6219-6226.(104) Ehde, P. M.; Andersson, I.; Petersson, L. Ada. Chem. Scand. A 1986,40,489-499.(105) Sakurai, H.; Shimomura, S.; Ishizu, K. Inog. Chim. Acta 1981,55, L65-L69.(106) Yordanov, N. D.; Zdravkova, M. Polyhedron 1993,12,635-639.(107) Kuska, H. A.; Rogers, M. T. Inorg. Chem. 1966,5,3113.(108) Beddoes, R. L.; Collison, D.; Mabbs, F. E.; Passand, M. A. Polyhedron 1990,9,2483-2489.(109) Chasteen, N. D.; Lord, E. M.; Thompson, H. 3.; Grady, 3. K. Biochim. Biophys. Acta1986,884,84-92.80(110) Vanadiwn in Biological Systems - Physiology and Biochemistry; Chasteen, N. D.,Ed.; Kiuwer Academic: Dordecht, 1990.(111) Harris, R. K. Nuclear Magnetic Resonance : A Physicochemical View Pitman:London, 1983.(112) Rehder, D. In Vanadiunz in Biological Systems; N. D. Chasteen, Ed.; KiuwerAcademic: Dordrecht, 1990;173-198.(113) Rebder, D.; Weidemann, C.; Duch, A.; Priebsch, W. Inorg. Chem. 1988,27,584-587.(114) O’Donnell, S.; Pope, M. T. I. Chem. Soc. Dalton Trans. 1976,2290-2297.(115) Butler, A.; Danzitz, M. 3.; Eckert, H. J. Am. Chem. Soc. 1987,109, 1864-1865.(116) Butler, A.; Eckert, H. I. Am. Chem. Soc. 1989, 111, 2802-2809.(117) Crans, D. C.; Shin, P. K. Inorg. Chem. 1988,27, 1791-1806.(118) Crans, D. C.; Bunch, R. L.; Theisen, L. A. I. Am. Chem. Soc. 1989, 111, 7597-7607.(119) Crans, D. C.; Wong, C.-H.; Daniels, L.; Pederson, R. L.; Durrwachter, 3. R.;Drueckhammer, D. 0. J. Org. Chem. 1989,54,70-77.(120) Crans, D. C.; Rithner, C. D.; Theisen, L. A. I. Am. Chem. Soc. 1990,112,2901-2908.(121) Rehder, D. Inorg. Chem. 1988,27,4312-4316.(122) Rehder, D.; Hoist, H.; Quaas, R.; Hinrichs, W.; Hahn, U.; Saeñger, W. J. Inorg.Biochem. 1989, 37, 141-150.(123) Tracey, A. S.; Jaswai, 3. S.; Gresser, M.; Rehder, D. Inorg. Chem. 1990,29,4283-4288.(124) Tracey, A. S.; Gresser, M. 3. Inorg. Chem. 1990,29,2267-2271.(125) Tracey, A. S.; Gresser, M. 3.; Liu, S. Biochemistry 1992,31,2677-2685.(126) Habayeb, M. A.; Hileman, 0. E. Can. I. Chem. 1980,58,2255-2261.81(127) Pettersson, L.; Hedman, B.; Andersson, 1.; Ingri, N. ChemicaScripta. 1983,22,254-264.(128) Blair, A. 3.; Pantony, D. A.; Minkoff, G. I. I. bzorg. Nuc. Chem. 1958,5,316-331.(129) Giacomelli, A.; Floriani, C.; Duane, A.; Chiesi-Villa, A. Inorg. Chem. 1982,21,3310-3316.(130) Finnegan, M. M.; Lutz, T. G.; Nelson, W. 0.; Smith, A.; Orvig, C. Jnorg. Chem.1987,26,2171-2176.(131) Nelson, W. 0.; Karpishin, T. B.; Rettig, S. 3.; Orvig, C. Inorg. Chem. 1988,27,1045-1051.(132) Zékány, L.; Nagyp1, I. Computational Methodsfor the Determination ofStabilityConstants; Plenum Press: New York, 1985.(133) Rossotti, H. S. Talanw 1974,21,809-829.(134) Hedlund, T.; Ohman, L-O. Ac:. Chem. A 1988,42,702-709.(135) Clevette, D. J.; Orvig, C. Polyhedron 1990, 9, 151-161.(136) Gerard, C.; Hugel, R. P.1. Chem. Research 1978,404-405,4875-4889.(137) Lal, K.; Agarwal, R. P. Bull. Chem. Soc. Japan 1967,40, 1148-1152.(138) Nechay, B. R.; Nanninga, L. B.; Nechay, P. S. E. Arch. Biochem. Biophys. 1986,251, 128-138.(139) Lorenzotti, A.; Leonesi, D.; Cingolani, A.; Bernardo, P. D. 1. Inorg. Nuc. Chem.1981,43,737-738.(140) Mont, G. E.; Martell, A. E. 1. Am. Chem. Soc. 1966,88, 1387-1393.(141) Costa Pessoa, 3.; Vilas Boas, L F.; Gillard, R. D. Polyhedron 1989,8, 1173-1199.(142) Costa Pessoa, 3.; Marques, R. L.; Vilas Boas, L. F.; Gillard, R. D. Polyhedron 1990,9, 81-98.(143) Costa Pessoa, 3.; Vilas Boas, L. F.; Gillard, R. D. Polyhedron 1990,9,2101-2125.82(144) Costa Pessoa, 3.; Antunes, 3. L.; Vilas Boas, L. F.; Gilard, R. D. Polyhedron 1992,11, 1149-1461.(145) Pettersson, L; Hedman, B.; Nenner, A. M.; Andersson, I. Acm. Chem. Scand. A1985,39,499-506.(146) Ehde, P. M.; Peuersson, L.; Glaser, 3. Acm. Chem. Scand. A1991, 45,998-1005.(147) Smith, R. M.; Martell, A. E. Critical Stability Constants; Plenum Press: New York,1974-1989.(148) Costa Pessoa, 3., Personal Communication.(149) Huheey, 3. E. Inorganic Chemistry; Third edition; Harper & Row: New York,1983, 276-277.(150) Bryant, B. E.; Fernelius, W. C. I. Am. Chem. Soc. 1954,76,5351-5352.(151) Okac, A.; Kolarik, Z. Collect. Czechoslov. Chem. Commun. 1959,266-272.(152) Gerard, C. Bull. Chem. Soc. France 1979,1,451-456.(153) Gerard, C.; Hugel, R. P. C. R. Acad. Sc. Paris 1982,295, 175-177.(154) Herring, G. Unpublished Results83


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items