‘V— V—‘rm. -'— —W_ A SPECTRQFHOTOMETRIC AND SPECTROFLUORGMETRIC STUDY OF FLAVQNOL-ALUMINUM {I-IE'LATES. IN ABSZGLUTE ETHYL ALCQHOL Thai: for the Degree of DI}. D. MICHIGAN STATE UNIVERSITY Frederick Lewis Urbach 1964 LIBRARY LI Michigan State University ABSTRACT A SPECTROPHOTOMBTRIC AND SPECTROFLUOROMETRIC STUDY OF FLAVONOL-ALUMINUM CHELATES IN ABSOLUTE ETHYL ALCOHOL by Frederick Lewis Urbach The electronic absorption and emission spectra of flavonol in several solvents have been interpreted in terms of an internally hydrogen-bonded flavonol species. The i changes in the absorption and emission spectra of flavonol which occur with the formation of flavonol—aluminum chelates are also interpreted. The chelation reactions between aluminum ions and flavonol have been studied in absolute ethyl alcohol. A series of fluorescent polynuclear chelates is formed. Spec— trOphotometric and spectrofluorometric evidence is presented for the existence of species with the stoichiometries: 6 aluminum ions to l flavonol, 2 aluminum ions to l flavonol and 1 aluminum ion to l flavonol. Without the addition of base to the solutions flavonol reacts with aluminum ion first to form the 6:1 species and with excess flavonol then forms the 2:1 aluminum to flavonol species. The addition of base to give a hydroxide to metal ion ratio of 2.5 is necessary to form the 1:1 flavonol to aluminum chelate. The chelates are stable in dilute acid solutions. The presence of acid or water enhances the fluorescence greatly. In eXcess base the chelate which forms is nonfluorescent. Potentiometric Frederick Lewis Urbach titrations of chelate solutions with base indicate that the addition of flavonol to aluminum ions to give the 2:1 alum- inum to flavonol species proceeds by the displacement of two solvated ethoxide ions from the coordination sphere of the metal ion. Further addition of flavonol occurs by the elimination of one solvated ethoxide ion and one solvated ethanol molecule. The preparation of a solid flavonol- aluminum chelate was.attempted and a substance which ap- peared by analysis to be a 3:1 flavonol to aluminum chelate was isolated. A SPECTROPHOTOMETRIC AND SPECTROFLUOROMETRIC STUDY OF FLAVONOL-ALUMINUM CHELATES IN ABSOLUTE ETHYL ALCOHOL BY Frederick Lewis Urbach 'A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1964 VITA Name: Frederick Lewis Urbach Born: November 21, 1938 in New Castle, Pennsylvania Academic Career: Beaver Falls High School, BeaVer Falls, Pennsylvania (1953-1956) The Pennsylvania State University, University Park, Pennsylvania (1956— 1960) Michigan State University, East Lansing, Michigan (1960—1964) Degree Held: 8.8. The Pennsylvania State university (1960) ii ACKNOWLEDGEMENTS The author wishes to expresshis gratitude to Dr. Andrew Timnick for his encouragement and advice throughout this investigation and course of study. The author also acknowledges the financial support of the National Science Foundation and the Socony—Mobil Oil Company. Special thanks go to the author's wife, Carrie, for her encouragement and devotion. iii VITA . TABLE OF CONTENTS ACMOWIJEDGEMBNTS O O O O O O O O O O O O O O O 0 LIST OF FIGURES O O O O O O O O O O O O O O O 0 INTRODUCTION 0 O O O O O O O O O O O O O O O O O HISTORICAL O O O O O O O O O O O O O O O O O O 0 EXPERIMENTAL O O O O O O O O O O O O O O O O O O INSTRUMENTATION . . . . . o o . . o o o . o CHEMICAIIS O O O O O O O O O O O O O O O O O PREPARATION OF FLAVONOL . . o . o o . . o . PREPARATION OF STOCK SOLUTIONS . . . . o . EXPERIMENTAL PROCEDURES . . o . o . . . o 0 RESULTS Calibration of the Spectrofluorometer Spectrophotometric Titrations Apparent pH Measurements Temperature Control AND DISCUSSION 0 O O O O O O O O O O O O The Electronic Absorption Spectrum of Flavonol The Electronic Absorption Spectrum of the Flavonol Ion The Electronic Absorption Spectra of Flavonol-Aluminum Chelates The Fluorescence Spectrum of Flavonol Determination of Molar Absorptivities of Flavonol and the Flavonol Anion Determination of the Stoichiometries of the Flavonol-Aluminum Chelates The Determination of Flavonol-Aluminum Chelate Molar Absorptivities Potentiometric Titrations of the Chelate with Base iv Page ii iii vi mefiH 10 11 12 12 13 13 13 14 15 24 25 29 36 37 62 62 The Effect of Acid on the Chelates The Effect of Water on the Flavonol- Aluminum Chelates Preparation of a Solid Flavonol-Aluminum Chelate SUMMARY AND CONCLUSION . . . o . . . . . . . . . LITERATURE CITED 0 O O O O O O O O O O O O O O O Page 67 7O 72 77 83 Figure 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. LIST OF FIGURES Absorption Spectra of Flavonol in Carbon Tetrachloride and in Chloroform . . . . . . . Absorption Spectra of Flavonol in Acetone andinISOOCtane00000000000000 Absorption Spectra of Flavonol in Ethanol andinmethan01ooooooooooooooo Absorption Spectrum of the Flavonol Anion inEthanO].0.000.000.0000... Changes in the Absorption Spectrum of Flavonol in Ethanol Caused by the Addition of Aluminum (III) Solution . . . . . . . . . .Fluorescence Spectra of Flavonol in Carbon Tetrachloride and in Chloroform . . . . . . . Fluorescence Spectra of Flavonol in Isooctane andinACetoneooooooooooooooo Fluorescence Spectra of Flavonol in Ethanol andinMethanOl............... Fluorescence Spectrum of Flavonol-Aluminum ChelateSinBthan01ooo0.000.000. Continuous Variations Curves for Flavonol- Aluminum Solutions in Ethanol . . . . . . . . Spectrophotometric Titration Curves for Aluminum (III) Titrated with Flavonol in the Absence of Hydroxide Ions . . . . . . . . Spectrophotometric Titration Curve for Flavonol Titrated with Aluminum (III) in the Absence 0f HYdrOXide 10113 0 o o o o o o o o o Spectrophotometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 1:1 . . . . Spectrophotometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 2.5:1 . . . vi Page 16 17 18 26 27 30 31 32 35 38 40 41 42 43 Figure 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Spectrophotometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 3:1 . . . . . Spectrophotometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio of 4:1 . . . Spectrofluorometric Titration Curves for Aluminum (III) Titrated with Flavonol in the Absence of Hydroxide Ions . . . . . . . . Spectrofluorometric Titration Curve for Flavonol Titrated with Aluminum (III) in the Absence of Hydroxide Ions . . . . . . . . Spectrofluorometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 1:1 . . . . . Spectrofluorometric Titration Curves for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 2.5:1 . . . . Spectrofluorometric Titration Curve for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 3:1 . . . . . Spectrofluorometric Titration Curve for Aluminum (III) Titrated with Flavonol with the Hydroxide to Metal Ion Ratio 4:1 . . . . . Spectrofluorometric Titration Curves for Aluminum (III) Titrated with Flavonol to Indicate the Existence of a 6:1 Aluminum to Flavonol Species . . . . . . . . . . . . . Spectrofluorometric Titration Curve for Flavonol Titrated with Aluminum (III) to Indicate the Existence of a 6:1 Aluminum to FlavonolSpeCieS............... Variation of Chelate Fluorescence with Time after Mixing for a Series of Aluminum to FlavonolRatiOS............... Potentiometric Titration Curves for Aluminum (III) Titrated with Hydroxide Ions in Absolute EthanOl0.0000000000000000. vii Page 44 45 48 49 50 51 52 53 58 59 61 63 Figure Page 27. Potentiometric Titration Curves for the Titrations of Various Aluminum to Flavonol Ratios with Hydroxide Ions in Absolute Ethan01...o................ 64 28. Potentiometric Titration Curves for the Titrations of Various Aluminum to Flavonol Ratios with Hydroxide Ions in Absolute Ethan01 O O O O O O O O O O O O O O O O O O O O 65 29. Dependence of the Absorbance and Fluorescence of the 2:1 Aluminum to Flavonol Chelate on ApparentpH.................. 68 30. Variation of the Absorbance and Fluorescence of the 1:1 Aluminum to Flavonol Chelate with the Addition of Hydrochloric Acid . . . . . . . 69 31. Dependence of the Absorbance and Fluorescence of the 2:1 Aluminum to Flavonol Chelate on theAdditionofWater............. 71 32A. Infrared Absorption Spectrum of the Solid Flavonol Aluminum Chelate . . . . . . . . . . . 75 328. Infrared Absorption Spectrum of the Solid FlavonOl Aluminum Chelate o o o o o o o o o o o 76 viii INTRODUCTION Flavones are naturally occurring substances which are generally found in plants as the polyhydroxy or polymethoxy substituted compounds. The structure and numbering conven- tion of the parent compound, flavone, is given here. No monohydroxyflavones are known to exist in nature, however many have been synthesized. The simplest naturally occurring hydroxy substituted flavones are dihydroxy com- pounds. The natural occurrence, syntheses and reactions of flavonoid compounds has been reviewed (16,44,47,59,62). The ultraviolet absorption spectra (2,11,27,53) and the infrared absorption spectra (8,21,25,35,51) of these compounds have also been studied. With vicinal dihydroxy groups, or hydroxy groups in positions 3 or 5, the possibility of forming chelates with metal ions arises. The formation of chelates between poly- hydroxyflavones and metal ions has been studied for a vari- ety of flavonoids and metal ions in several solvents. Early studies in this laboratory involved the complex- ation of various lanthanide ions with morin (2',3,4',5,7 pentahydroxyflavone) using 1:1 dioxane water as a solvent (14,56,58). The present study is concerned with the chela- tion reactions of 3-hydroxyflavone (flavonol) and aluminum ions in absolute ethanol. Flavonol was chosen as the chelating agent to avoid the problems inherent with having several sites of chelation on the same molecule. The chelation of aluminum ions was chosen when preliminary experiments indicated that a series of stable fluorescent complexes was formed. Absolute ethyl alcohol was selected as a solvent on the basis of pre- liminary experiments which indicated that the presence of water had a serious effect on the complexation reactions of flavonol with metal ions. HISTORICAL The earliest report of a chelation reaction with poly- hydroxy flavones was the discovery by Goppelsroeder (18) that in the reaction between aluminum ion and morin (2',3,4',5,7, pentahydroxyflavone) a species was formed which possessed an intense greenish fluorescence. Since then research with metal salt chelates of flavones has been concerned mainly with the following three areas of study: 1. The use of metal salt chelates to elucidate struc- tural factors of flavone compounds. 2. The use of chelates as spectrophotometric reagents for the qualitative identification and quantitative determination of flavones. 3. The study of flavones as spectrophotometric or Spectrofluorometric reagents for metal ions. An early report by Neelakantam and Row (38) attempted to correlate the structural factors of naturally occurring polyhydroxyflavones with the appearance of fluorescence with aluminum or beryllium salts. They employed visual observa- tion and proposed that the 3-hydroxy group was not absolutely necessary for the emission of fluorescence. Horhammer and Hansel (23) studied photometrically the stoichiometry of the rutin and quercetin complexes formed with zirconium, aluminum and copper in aqueous alcohol. They differentiated between quercetin and its 3-glycosides by the stability of the zirconium complex of the aglycon in the pre- sence of citric acid. These authors (24) later extended their examination of the stoichiometries of the zironconium com- plexes with flavones to include flavonol and myricetin. Swain (54) tabulated the shift in wavelength of the main ultraviolet band of phenolic compounds including hydroxyflavones caused by the addition of aluminum chloride to alcoholic solution of these compounds. The structure of flavones was investigated by Brune (10) by the use of the formation of metal salt complexes. He em— ployed the ultraviolet spectrum of the beryllium salt complexes to differentiate between flavonols and flavanones. Neu (40) differentiated between flavones and flavone glycosides by the formation of colored copper or cadmium complexes with the former which were stable to acetic acid. Jurd and Geissman (28) studied the effect of aluminum chloride on the absorption Spectra of many flavonoids. Kanno (29) studied the reaction between germanium and flavonoids spectrophotometrically. The suitability of several metal salt complexes for flavone determinations was examined by Hagedorn and Neu (19, 20,39). The reaction between hypericin or hyperoside and aluminum chloride was proposed as a photometric determination of these flavones in aqueous ethanol extracts. The amperometric titration of some flavonoid compounds with copper sulfate solution was studied by Detty, Heston and Wender (13). By this method the authors determined the stoichiometries of complexes with several flavones and fla- vanones in essentially aqueous solutions. Morin has by far received the most attention of all the flavones as a photometric reagent for metal ions. Among the metal ions which have been determined with morin are aluminum (55), thorium (36,37). molybdenum (VI) (1), uranium (VI) (49), indium and gallium (52). Other photometric determinations which have been re- ported are: titanium (IV) with quercetin (l7), gallium with quercetin (45), indium with quercetin (3), boron with quercetin (22), tin with flavonol sulfonic acid (42), uran- ium (VI) and thorium (IV) with norwogonin (31), zirconium with galangin (32), vanadium (V) (6), and uranium (VI) (7) with rutin, and uranyl ion with flavonol (30). Reactions between metal ions and hydroxyflavones yield- ing fluorescent species have also received attention. Coyle and White (12) reported a fluorimetric determination of tin with flavonol. White, Hoffmann and Magee (61) examined the emission, excitation and absorption spectra of several fla- vone—metal chelates. Morin has also been used as a fluori- metric reagent for aluminum (60), beryllium (15,48), and certain lanthanide ions (14,58). Korenman st El' (34) studied the colorimetric and fluorescence reactions of sev- eral metal ions with quercetin. Flavonol has also been sug- gested as a fluorimetric reagent for tungsten (5). Ballantine and Whalley (4) reported the preparation of stable solid complexes of substituted flavonols with aluminum chloride. These were reported to be the triligand Chelates. EXPERIMENTAL Instrumentation The following instruments were employed to make the appropriate measurements: Beckman Zeromatic pH meter with combination glass— saturated calomel'electrodetto measurenthetapparent pH of the ethanol solutions. Beckman Model DB Spectrophotometer equipped with a Sargent SR recorder to obtain visible and ultra— violet absorption spectra. Cary Model 14 Recording Spectrophotometer to obtain visible and ultraviolet absorption spectra. Beckman lRSA Infrared Spectrophotometer to obtain infrared absorption spectra. Spectrofluorometer (52), employing a Hanovia mer- cury arc source, a Bausch and Lomb excitation monochromator, a Beckman DU with an AC power supply as an emission monochromator and a IPUAS photo- multiplier tube as a detector. Chemicals The following chemicals were used without further purification in the syntheses and preparations of reagent solutions described in this study: Acetone Reagent Grade Commercial Solvents Corp. Aluminum chloride Anhydrous, Analytical Reagent Grade,Mallinckrodt Chemical Works Aluminum nitrate, Analytical Reagent Grade nonahydrate Allied Chemical Ammonium Hydroxide Reagent Grade Fisher Scientific Company Benzaldehyde The Matheson Co., Inc. Carbon tetrachloride Baker's Analyzed Reagent J. T. Baker Chemical Co. Chloroform Ethylenediaminetetracetic acid, disodium salt, monohydrate Ethanol, absolute Ethanol, 95% Hydrochloric Acid Hydrogen peroxide 30% Isooctane O-Hydroxyacetophenone Methanol Nitric Acid Perchloric Acid 70-72% Quinine Sulfate Sodium Hydroxide Zinc Sulfate Preparation of Flavonol Baker's Analyzed Reagent J. T. Baker Chemical Co. Analytical Reagent Grade J. T. Baker Chemical Co. Commercial Solvents Corp. Commercial Solvents Corp. Baker's Analyzed Reagent J. T. Baker Chemical Co. Baker's Analyzed Reagent J. T. Baker Chemical Co. Eastman Kodak Distillation Products “ K and K Chemical Company Baker's Analyzed Reagent J. T. Baker Chemical Co. Baker's Analyzed Reagent J. T. Baker Chemical Co. Baker's Analyzed Reagent J. T. Baker Chemical Co. Mallinckrodt Chemical Works Baker's Analyzed Reagent J. T. Baker Chemical Co. Reagent Grade Allied Chemical Company The flavonol was synthesized by the method of Oyamada (43). Benzaldehyde and o—hydroxyacetophenone were condensed in the presence of sodium hydroxide according to Kostanecki (33) to give 2'—hydroxychalcone. lized from aqueous ethanol. The chalcone was recrystal- This product was treated with hydrogen peroxide and sodium hydroxide to yield flavonol. 11 The flavonol was recrystallized several times from aqueous ethanol. The flavonol preparation was analyzed by the Spang Microanalytical Laboratory. Found: 75.79% C, 4.13% H; cal- culated: 75.61% C, 4.23% H. Stock solutions of flavonol in absolute ethanol were prepared by direct weighing of this product. Preparation of Stock Solutions Standard ethylenediaminetetracetic acid (EDTA) solution was prepared by direct weighing oftthe‘disodium as t hydrate. A zinc sulfate solution was standardized by titration with the EDTA solution using Eriochrome Black T indicator. Solutions of aluminum salts were prepared by dissolving the appropriate amount of either anhydrous aluminum chloride or hydrated aluminum nitrate in absolute ethanol to give ap- proximately 0.01 M solutions. Aliquots of these solutions were evaporated to dryness on a steam bath and the residue was dissolved in concentrated hydrochloric acid with heating. After dilution with distilled water, a measured excess of standard EDTA solution was added. The pH of the solutions were then adjusted to between 7 and 8, a few drops of Erio+ chrome Black T indicator were added, and the excess EDTA was titrated immediately with the standard zinc sulfate solution. The concentrations of the aluminum solutions were -3 calculated and the stock solutions of 1.00 x 10 M aluminum salts were prepared by dilution. 12 Stock solutions of sodium hydroxide in ethanol were pre- pared by dissolving sodium hydroxide pellets in ethanol. These solutions were standardized against potassium acid phthalate using phenolphthalein as the indicator. Stock solutions of hydrochloric acid in ethanol were prepared by dilution of aliquots of concentrated aqueous hydrochloric acid. These solutions were standardized with the standard sodium hydroxide solution using phenolphthalein as the indicator. Experimental Procedures Calibration of the Spectrofluorometer. The spectrofluorometer was calibrated with a solution containing 24.1 x 10"3 mg/ml of quinine sulfate in 0.1 N sulfuric acid. The following instrument settings were em- ployed in the calibration procedure: entrance slit, excita- tion monchromator, 1.0 mm.; exit slit, excitation mono- chromator, 0.5 mm.; excitation wavelength, 365 mu; emission wavelength, 540 mu; slit, emission monochromator (DU), 0.3 mm.; selector switch, DU, 0.1; percent transmittance, 60%; AC power supply, power switch, on, sensitivity switch, full. With the phototube shutter closed, the dark current was ad- justed to zero. The calibration solution was now placed in the sample compartment and with the shutter open, the sensi- tivity control on the DU was adjusted to give a zero reading. 13 Spectrophotometric Titrations An aliquot of the reagent to be titrated was placed in each of a series of volumetric flasks. To these aliquots were added successively larger increments of the titrant. The contents of the flasks were then diluted to volume with sol- vent and the appropriate measurements were made on these solutions. No correction for dilution was necessary with this procedure. Apparent pH Measurements In the apparent pH measurements, the meter was calle- brated with an aqueous pH 7 buffer solution. The electrode was then allowed to equilibrate for a minumum of twenty min- utes in absolute ethanol before any readings were taken. A circuit ground and a solution ground were employed in these measurements. Temperature Control Unless otherwise stated, no temperature control was employed during any of the measurements. All standard solu- tions were prepared at 25 i 0.1.. RESULTS AND DISCUSSION 15 The Electronic Absopption Spectrum of Flavonol The ultraviolet absorption spectrum of flavonol as ob- tained in several solvents is presented in Figures 1-3. The characteristic shape of the absorption spectrum is essentially the same in methanol, ethanol, chloroform, carbon tetrachlor- ide, acetone and isooctane. In the flavonol molecule the carbonyl group is con- jugated directly with the A ring and is conjugated with the B ring (2—phenyl ring) through a double bond. Except for some high energy ethylenic TT-fTT‘ transitions, occurring below 200 mp, the electronic absorption spectrum of flavonol can be attributed to a variety of TT—eT1‘ transitions origi- nating in the carbonyl group and varying in energy depending on the conjugation with the rest of the molecule. The molar absorptivities of all the observed transitions are greater than 10,000 absorbance units per mole indicating that these are indeed TT-a11* transitions. In earlier studies (2,53) it was assumed that flavonol undergoes a keto-enol tautomerization in solution. From this assumption the 344 mp absorption band was assigned to the keto form and the 306 my band was attributed to the enol absorption. Observations in this study indicate that the keto-enol tautomerization does not occur in solution and the 16 .suomoonmU cm can woauoHnumwuma conwmu cw doco>cam mo swuummm coaumuomnd oo.H .H mmpeHm .m: CH £u0d0H0>m3 oom owe omv ovv owe 00¢ 0mm own OVm own com 0mm omw ovm .ONN cow — — b L p p/ _ L p _ _ p p p O z ’ \I a s a ’ x s I OH 0 O . a . i a . . L \ . \ uo~.o . \ . I. a . .. i , . i x _ uom.o . i \ s N I , a. \\ x a 4‘ a o . I. / \ ice o w" — s . 9 a \ / O o s l. J .. \ TOm.o m. /, x w , a / lom.o I s I Io>.o mofiuoanumuvma conwmu lom.o CH Hoco>mam .2 mica x o.v III Ewomoonnu Iom.o ca Hoco>mHm .2 mloa x o.v Ill. 17 com .msmuuoouH cm can occumud ca Homebmam mo cwuummm coauawomnd - .L2 as nuucoam>m3 L p . . . _ p b P .l .N MMQOHm _ omv omv ovv omw oov on own own own oom 0mm 0mm ov~k.o- CON I l-‘ ""|'-"-"’ ocmuuoomH IOH N ocfi Ill scoumud OH X o.¢.lll ca Hoco>mHm .2 m CH Hoco>mHm .2 ml 0 10H.o ro~.o Jom.o Toe.o lom.o [00.0 lou.o lom.o om.o oo.H SDU'QQJOSQV l8 .HOCmnuo2 CH pCm HOszuu Ca H0C0>mah mo swuuwmm Coauauomnd L .m mesons .12 CH CuaCmHa>m3 i _ _ e L! Li — oom omv owe owv omv oov 0mm 00m own own oom 0mm owN _ oem o- sow P _ CH H0C0>mHm CH H0C0>mam .2 .2 Hocmnuwz mIOH x m.m III socmrpm mica x o.e_ill o TOH.0 lom.o rom.o iocoo _. :1 tom 0 eoueqxosqv / rom.o Iow.o Iom.o .[om.o oo.H l9 assignment of the absorption bands to these tautomers is incorrect. Below are tabulated the ratios of the absorbance values of flavonol at 344 my and 306 mp in five of the solvents em- ployed in this study. Solvent A344/A306 Ethanol 1.42 Chloroform 1.40 Carbon tetrachloride 1.45 Isooctane 1.38 Methanol 1.37 It is found that this ratio is essentially constant in a variety of solvents. If these absorbances could be attrib- uted to the enol and keto forms of flavonol, then the ratios of the absorbances should vary in different solvents since the keto to enol ratio will vary in different solvents. The fact that this ratio is essentially constant in several sol- vents indicates that the 344 and 306 mp bands are not charac— teristic of the keto and enol tautomers but are simply due to two electronic transitions in the same molecule. Evidence for the flavonol species which exists in solu- tion is obtained from the infrared absorption spectrum of flavonol in carbon tetrachloride solution. In this solvent flavonol exhibits a weak hydroxyl stretching frequency at «3333 cm.-l. The fact that the hydroxyl stretching frequency is very weak and is shifted from the normal "free" hydroxyl stretching frequency of 3650 cm."1 indicates that the 20 3-hydroxyl group is weakly hydrogen-bonded to the carbonyl group in the 4 position. The position of the stretching fre- quency of the carbonyl group is also shifted to lower ener- gies indicating that it does engage in hydrogen-bonding with the adjacent hydroxyl group. The carbonyl stretching fre- quency for the unsubstituted flavone occurs at 1660 cm.-1. Briggs and Colebrook (8) give further evidence for in- ternal hydrogen bonding by the increase in the carbonyl fre- quencies when 3-hydroxyf1avones are acylated. Acylation of the 3-hydroxy group results in a shift in the carbonyl fre— quency of approximately 40 cm.'1 to higher energies. This shift presumably occurs because the acyl substituent is in- capable of hydrogen bonding with the carbonyl. Shaw and Simpson (50) used partition chromatography in the study of hydrogen bonding in flavones and their data indi- cate hydrogen bonding between the 3—hydroxyl and the 4- carbonyl groups. From these data and observations it is concluded that flavonol exists in solution entirely in the enol form which is stablized by conjugation and hydrogen-bonding. Any inter— pretation regarding the ultraviolet absorption spectrum of --H flavonol must be based on this type of species. The electronic absorption spectrum of unsubstituted flavone consists of bands at 200, 250 and 295 mp. The 21 introduction of a hydroxyl group at the 3 position produces radical changes in this spectrum. A new band appears at 344 mp with other bands located at 306 mp, 238-242 mp and 201 mp. Formerly the appearance of the new long wavelength band in 3- hydroxyflavones was attributed to the possibility of keto- enol tautomerization with these compounds. It is apparent that the long wavelength absorption band (in the region 340— 370 mm) which occurs when a 3-hydroxy group is present in the flavone nucleus appears because certain TT-eTT' transitions of the carbonyl group are made easier by the presence of the 3-hydroxy group. It is proposed that the 3-hydroxy group affects the transitions mainly in two ways. In.‘n1—>1T* transitions the excited state receives pre- dominant contributions from polar structures since an electron is excited into an antibonding orbital. In the case of fla- vonol some probable polar excited states may be described by the following resonance structures. From these representaef tions it is obvious that the presence of hydrogen-bonding in 3-hydroxyflavones stabilizes these polar excited states of the molecule. Since the energy of the excited state is low— ered, the transition to this excited state occurs at a higher wavelength. The second method by which the 3-hydroxyl group affects the long wavelength transition is the auxochromic effect of a 22 strong electron donating group adjacent to a conjugated car- bonyl group (26). It has been shown that the presence of electron donating groups such as -OR, -Cl, -Br or -OH in the ocposition of dienones produces a bathochromic shift in the main‘Ff—é Tf‘ transition. The effect of the hydroxyl group is the greatest of these substituents. It is not known whether this is due to the greater electron donating ability of the hydroxyl group or because of the possibility of hydro- gen-bonding. These two effects of the 3-hydroxyl group, the electron- donating ability and the stabilization of certain resonance structures by hydrogen-bonding, cannot be separated. It ap- pears likely, however, that the combination of these two effects is responsible for the large shift in the long wave- length band of 3-hydroxyflavones. Further evidence that this argument is correct is ob- tained from an examination of the absorption spectra of other substituted flavones. With hydroxy groups in the 2' org4' positions, resonance stabilization, as shown below, similar to that proposed for the 3-hydroxy compounds is possible. These substituents produce a bathochromic shift of the long wavelength band to 328 and 334 mp respectively. 9 FKIB illl”. 03’ With the compound 4' methoxy -3— hydroxyflavone the H 6%) effects of the two substituents appear to be additive as the 23 long wavelength absorption occurs at 356 mp. The same addi- tive effects occur with morin and quercetin whose long wave- length absorption occurs at 380 mp and 375 mp respectively. The ultraviolet absorption spectrum of 3-aminoflavone, in which hydrogen-bonding can occur similar to that in fla- vonol, is very similar to the spectrum of flavonol. The positions of the major absorption bands for 3-aminoflavone are 242, 305 and 362 mp. These are analogous to the 240, 306 and 344 mp bands of flavonol. The shift of the long wavelength band to 362 mp upon the substitution of an amine group at the 3 position indicates that the stabilization of the excited state is greater in this molecule than in fla- vonol. This may be attributed to the greater electron dona- ting ability of the amine group or to a more favorable struc- ture for the formation of a five-membered chelate ring as shown below. The similarity between the absorption spectrum of fla- vonol and that of 3-aminoflavone is presented as a further argument against the explanation of the spectrum of flavonol in terms of a keto-enol tautomeric equilibrium. If the keto- enol concept was correct then the 362 mp band in the spectrum of 3-aminof1avone must be assigned to an imine form, analo- gous to the keto form of flavonol. It is unlikely that this 24 type of tautomeric equilibrium involving an imine structure can occur. With 3-acetoxyflavone and 3-methoxyflavone the spectra revert to that of flavone. This indicates that the excited state is not stabilized to any extent by these substituents even though the methoxy group could be expected to have some electron donating ability. The 3-bromoflavone spectrum indicates a slight batho- chromic shift of the long wavelength band to 308 mp. This may be attributed to the electron-donating ability of the Br substituent. The Electronic Absorption Spectrum of the Flavonol Ion In the presence of an excess of hydroxide ions the: 3- hydroxy group in flavonol is ionized and the ultraviolet absorption spectrum is transformed into the characteristic spectrum of the flavonol ion (Figure 4). The major bands in this spectrum occur at 412, 315, 278, 238 and 200 mp. The typical bathochromic shift observed in the ionization of a phenolic hydroxy group is attributed to the production of a very strong electron donating substituent, -0-, at the 3 position. 25 The Electronic Absorption Spectra of Flavonol- Aluminum Chelates The changes in the flavonol absorption spectrum which occur upon the addition of successive increments of aluminum ion solution may be observed in Figure 5. The fOrmation of flavonol-aluminum chelates results in the appearance of a new absorption band at 405 mp which is characteristic for the chelates. The appearance of the chelate band is accom- panied by the corresponding decrease in the bands attributed to the molecular form of flavonol. In the intermediate solu- tions all three transitions of flavonol in the 330 - 360 mp region may be observed. By observing the spectra of a series of flavonol-aluminum solutions it is apparent that the 326 mp band of the flavonol chelate arises from the same transi- tion which produces the inflection at 330 mp in the flavonol spectrum. When all of the flavonol has been converted to the chelate the characteristic spectrum has major absorption bands at 405, 326, 245, 232, and 201 mp. The band at 326 mp is not shifted significantly upon chelation and for this reason it is assumed that this transition is not affected by the 3-hydroxy group. This band is attributed to a trans- ition in the benzoyl grouping, the carbonyl group conjugated with the A ring. During the successive additions of aluminum ions to a flavonol solution, two isosbestic points appear in the spec- tra. These occur at 364 mp and 287 mp. 26 00m owe omv ovv owv 00¢ com o P .HOCMCpm CH COHCC HOCo>mah on» mo asuvommm COHuawonnd P — P .L2 CH CvoCmHm>m3 em own _ P mm o p _ L com r 2v .58on 0mm 0mm ovm CNN CON L b _ _ ECHUOm CH H OCO>M H .m CH .2 .2 N m mUonwphm Ioa x o.H IOH x N.m o roa.o uo~.o Iom.o ..lo¢.o Iom.o T8.0 .iuoe.o .rom.o rom.o oo.H eouquosqv 27 .coHusaom AHHHV sacH52H< mo aoHuHee< as» an pmmsmu HOCMCum CH HOCo>mHm mo Emuuummm COHquomn< may CH nmmCan .m MCDme .L2 CH CumCmHmbmz oom owe owe oev owe oov on 0mm oem 0mm com 0mm omN oem omm com _ _ _ . P PI _ _ C _ _ o r os.o A I o~.o .N 1 om.o .m . l oe.o l om.o . .I om.o in oa.o HCOH.m Huw .N .l om.o HNH .H 0Humm Hoco>mHm ow AHHHV ECCHESHC I.om.o Hoco>mHs 2H .z mica x o.~ l.oo.H eoueqrosqv 28 The shift in the position of the long wavelength ab- sorption band upon chelation cannot be attributed to the presence of a strong electron donor group at the 3-position since the chelation of an electropositive metal ion would result in an electron withdrawing effect. This shift in wavelength must then arise from the effect of the metal ion on the carbonyl oxygen. A multi-valent metal ion can accept an electron pair from the non-bonded electrons on the car— bonyl oxygen in contrast to a proton which can only interact weakly with these electrons in the formation of a hydrogen _bond. This effect, together with the increased stabiliza- tion of polar excited states in the chelate by the contri- bution of this resonance form are probably responsible for the wavelength shift upon chelation. ... ,8? a. AI++ In the presence of excess base the wavelength of max- imum absorption of the chelate was shifted from 405 to 400 mp. This indicates a change in the nature of the chelates upon the addition of hydroxide ions to the remaining co- ordination sites of the aluminum ions. The absorption spectra of the chelates were obtained using an aluminum stock solution prepared with anhydrous aluminum chloride. ‘This stock solution exhibited no ab— sorbance except for a sharp peak at 203 mp. The aluminum nitrate stock solutions exhibited absorbance bands at 274, 223, and 203 mp. These bands were presumably due to the absorption of the nitrate ion. 29 The Fluorescence Spectrum of Flavonol The fluorescence spectrum of flavonol as obtained in several solvents is presented in Figures 6-8. The location of the emission bands and the relative intensities of the maximum emission is summarized below: Solvent Emission Peaks Relative Intensities (mp.) 4x10-5M Equal Absorbance Solution at 365 mu. Carbon Tetra- chloride 526 l.00 1.00 Chloroform 520 0.64 0.51 Isooctane 525 0.53 0.96 Acetone 410,535 0.083 0.14 Methanol 410,535 0.078 0.072 Ethanol 410,535 0.073 0.073 Since many more factors influence the emission of fluorescence than do the absorption of radiation, the fluorescence spectrum of flavonol is not subject to simple interpretation. The effect of the solvent on the fluorescence spec- trum of flavonol depends on the polar nature of the solvent. In polar solvents, two weak bands are emitted, one at 410 mp and the other at 535 mp. In nonpolar solvents, the blue emission band is absent and the green emission band is in- tensified and shifted toward higher energies. The shift in the green emission toward lower wavelengths in nonpolar solvents is readily explained by the solvent interactions. 30 .EhomoonCU CH pCm mproHCu Imwpma Conwmu CH HOCo>MHm mo.muuumam wUCmumwwosHm . s .Ls cH rueaeau>mz omt 0mm 000 0¢0 0N0 000 00m 00m 0¢m owm 00m_00¢ 00¢_0¢¢ 0N¢ 00¢ 00m 00m 2 n p _ , _ _ P p — b - .0 HmeHm r .CE mmm I CmemHm>m3 COHumuHUxm EwomoonCU CH H0C0>mHm .2 mIOH x 0.¢ .N mUHonCUMHDme Conumu CH HOCO>mHm .2 mIOH x 0.¢ .H 0 .0H .om .0m .04 -om roe -oa -Om {om -oos IOHH Aatsuequl eoueosexontg .0C0umu< CH pCm owuuoomH CH H0C0>mHm mo mnuummm wUCmummuosHm .h mmeHm .L2 CH £00COH0>03 of. 0am 000 0¢m 0N0 000 00m 00m 0¢m 0mm 00m 00¢ 00¢ 0¢¢ 0~¢ 00¢ 00m 00m b - p r . — L p _ r — p — - 31 il.|\ 0 low 10m 10¢ . low low low .15 mom low I CumCmHm>m3 COHHmuHUxm .EE ¢.0 lom I a I HHHm Do mCmuuoomH CH H0C0>mHh .2 I0H x 0.¢ .N m l00H .ssm.H u pHHm so .mcoumu4 C 0Co>m . x . . H H H2 2 mI0H 0 ¢ H I0HH Kitsuequx eoueoseronta 32 .HOCCCumE CH UCC HOCmnum CH H0C0>0Hm mo muuummm qumummuoaHm .Lz CH CumaHm>m3 .m mmanm 005 000 000 0¢m 0N0 com 00m 00m 0¢m 0mm 00m 00¢ 00¢ 0¢¢ 0~¢ 00¢ 00m 00m r .xll» _ p _ _ _ _ xi? _ _ \k _ _ . P 0 foa 10m 10m 10¢ .~ Tom 0H low 105 .LE mmm I00 I CpoCMHm>m3 COHuwuHuxm om .ae o.H u uHHm so .Hocmnumz C 0Co>m . . . H H Hm EmloH x 0 m N 100H .ee m.H u uHHm be .Hocmnum CH HOC0>mHm .EmIOH K 0.¢ .H IOHH thsuaqul aouaasaxonta 33 The emission spectrum, as the absorption spectrum, must have its origin in 1T-4TT' transitions in the molecule. In a polar solvent the polar excited state is stablized and the energy required for the transition is lowered. The emission in polar solvents therefore occurs at longer wavelengths compared to the emission in nonpolar solvents where no sta- bilization of the excited state occurs. Solvent interactions also account for the decrease in the intensity of the green emission upon going to a more polar solvent. The polar solvent molecules interact with the polar excited state of flavonol and some of the absorbed energy is lost through external conversion. The appearance of a second emission band at 410 mp for flavonol solutions in polar solvents has not been explained. Several possible interpretations have been considered. Per- haps in polar solvents the internal hydrogen bonding in the flavonol molecule is partially destroyed by the competitive hydrogen bonding of the solvent molecules. This effect would produce two types of flavonol species in solution, and the two emission bands could be assigned to these species. It is unlikely that the appearance of the blue emission band can be correlated with the appearance of a significant absorbance due to the flavonol anion in the absorption spec- trum of flavonol obtained in polar solvents. The fluores- cence spectrum of the flavonol anion in ethanol, as obtained with a hundred-fold excess of sodium hydroxide, consists of a symmetrical band centering at 520 mp. The fluorescence 34 intensity of this band is so weak that it has been neglected in this study. Under the conditions when a 4.0xlO-5M solution of flavonol in ethanol gave a fluorescence intensity of 100 units, the emission of a comparable solution of the anion could not be detected. The emission spectrum of flavone in ethanol has been reported (7) to be a blue emission in the 430 mp region. When the 3-hydroxy group of some substituted flavonolw was methylated, it was reported (7) that the fluorescgnce spec— trum now consisted of primarily a blue emission at 430 mp, with only a small peak occurring in the 530 mp region. From these reports it is apparent that the 3-hydroxy group is re- sponsible for the shift in the emission wavelength from 430 mp to the 530 mp region. This shift from 430 mp to 530 mp corresponds to a shift of 4.6 kilokaysers. The shift in the long wavelength absorption band upon the substitution of a 3-hydroxy group in the flavone molecule, 295 mp to 344 mp, corresponds to a shift of 4.8 kilokaysers. It is assumed therefore that the same process is responsible for the batho- chromic shifts in the absorption and fluorescence spectra in going from flavone to 3-hydroxyflavone. The 410 mp emission band is postulated to be the tran- sition which is shifted to 454 mp and intensified to give the blue fluorescence characteristic of the flavonol-aluminum chelate. A typical emission spectrum for the flavonol- aluminum chelate is shown in Figure 7. This emission band is not symmetrical and tails off slowly toward higher 35 .Hocunpu CH mmumHmCU ESCHECHmHh mo eswuumam wUCwUmwuosHm .m mmwam .Lz CH numCmHm>m3 00> 000 000 0¢¢ 0mm 000 00m 00m 0¢m 0mm 00m 00¢ 00¢ 0¢¢ 0~¢ 00¢ 0mm 00m _ b, . C _ C #‘ C _ _ _ _ . . b _ _ .LE mom I CumCmHm>m3 COHumpHuxm 0 [tea 1.0N 1.0m i.0¢ r 00H 1.0HH Ratsuaaux eoueosexontg 36 wavelengths. The chelate emission spectrum can be excited by either 405 or 365 mp radiation. Determination of Molar Absorptivities of Flavonol and the Flavonol Anion In ethanol solutions flavonol ionizes to a slight ex- tent and therefore some of the flavonol exists in these solutions as the anionic form. For an accurate determination of the molar absorptivity of flavonol at 344 mp, the concen- tration of flavonol must be corrected for that amount exist- ing as the ion. In order to make this correction the molar absorptivity of the flavonol anion was determined by measur- ing the absorbance at 412 mp of a series of standard solu- tions of flavonol containing a hundred-fold excess of sodium hydroxide. A plot of these absorbance values versus con- centration gave a straight line which passed through the origin. The slope of this line was the molar absorptivity of the flavonol anion. The absorption spectra of a series of standard solu— tions of flavonol were now recorded. Using the molar ab- sorptivity of the anion obtained previously the amount of flavonol existing as the anion was determined and the con- centration of the molecular form was calculated by subtract- ing the concentration of the anion from the total concentra- tion of flavonol. A plot of the absorbance values at 344 mu versus the corrected concentrations yielded a straight line which passed through the origin. The slope of this line was 37 the molar absorptivity of flavonol at 344 mp. These deter— minations of molar absorptivities were carried out at 25.0 i 0.1% The values obtained for the molar absorptivities are tabulated below: Molar Absorp tivity 10356 Flavonol 344 = 1.69 x 104 4.228 Flavonol anion 412 = 1.45 x 104 4.161 The value for the molar absorptivity of flavonol is slightly higher than a previously reported value (2) for flavonol in ethanol solution. This is presumably due to the correction of the concentration for the flavonol existing as the anion. Determination of the Stoichiometries of the Flavonol-Aluminum Chelates Without the addition of base to the chelate solution, Job's method of continuous variations (Figure 10) indicates that a species with the stoichiometry two aluminum ions to one flavonol is formed. Upon the addition of base to give a hydroxide to metal ion ratio of 2.5 : 1, the Job's method curves seem to suggest that a species containing two fla- vonols per aluminum ion is formed but the results are in- conclusive. Spectrophotometric titrations of aluminum ion with flavonol and flavonol with aluminum ion were employed to 38 1. 405 mp. 2. 344 mp. 0.60-— 0050 — 0040 ‘— 0.30- 0.20 - 0.10 " -0010 _T -0. 20 '— -o.3o 4 -0040 "‘ I I I I I I I I 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mole Fraction of Aluminum (III). FIGURE 10. Continuous Variations Curves for Flavonol- Aluminum Solutions in Ethanol. 39 determine the stoichiometry of the chelates. The titrations of the metal ion with flavonol followed the increase in ab- sorbance values of the chelate peak at 405 mp and the in— crease in the flavonol absorbance at 344 mp. The titrations of flavonol with aluminum ions followed the increase in ab- sorbance values at 405 mp and the decrease in absorbance values at 344 mp. In the absence of base, the spectrophotometric titra- tion curves (Figures 11, 12) indicated that the initial stoichiometry was the same as that obtained by Job's method. On standing, however, further chelation of the metal ions took place and over a period of two weeks the stoichiometry of the chelate approached 1:1, flavonol to aluminum ions. The presence of base promoted the further chelation of the flavonol and spectrophotometric titrations of alum- inum ions with flavonol were carried out at several hydroxide to metal ion ratios to determine the effect of base on the stoichiometry of the chelates (Figures 13-16). In all of these titrations the order of mixing of the reagents was the same. The aliquot of the aluminum solution was placed in a volumetric flank and the flavonol aliquot was added to it. The hydroxide solution was then added and the solution was diluted to volume. If the base was added to the metal ion solution prior to the addition of the fla- vonol, the amount of chelate formation was diminished greatly. These solutions did not attain the same absorbance values as identical solutions which were mixed in the above 4O .uCOH mponwpxm mo mUCman< map CH H0C0>mHm Cqu noumuuHe AHHHV ECCH59H< Com uo>uso coHumuuHe UHuumsouoCdouuumdm .HH mmoon .AHHHV ECCHBCHC ou H0C0>0Hm mo OHumm 0H0: o.¢ m.m o.m m.~ o.~ m.H o.H m.o . P p w- b h b P .Ls moe .N .m& avm .H mCOH ECCHECH< CH .2 OH x 0.N ml. 10H.o I0N.0 Iom.0 I0¢.o 10H.H 41 mConunsm mo CUCmmCC «C» CH AHHHV .mCOH sCCHsCHC Cqu emu ImuyHB H0C0>CHm uom m>usu COHumuuHB UkuoEouonmouuqum .NH MMDme .HoCo>mHm on AHHHV CCCHCCHC mo oHuom mHoz m.~ o.~ m.H o.H m.o o _ C FI H _ _ - - n . . Io~.o Ioa.o . x . :om.o w S O I 1 a. c 0.. w Iom.0 e noo.H OH I .13 Cam .m , uo~.H 0 O o .IIYIII. o \I, as moe H I HOCo>mHm CH .2 mIOH x o.m 42 .HHH OHumm COH Hmuwz o» 00onwvwm mnu Cqu HOCo>mHm CuHs umpmuuHe AHHHV CCCHCCHC Com mo>usu CoHpmuuHe UHuumCouoCCouuumCm .mH mesoHn .AHHHV CCCHCCHC on HoCo>mHm mo oHumm mHoz 0.4 m.m o.m m.~ o.~ m.H o.H m.o o h . _ . _ I _ _ _ . I I o -OH.o I u -o~.o ll) C|||D|||0 Ooi|ID‘N 0 C a a 4‘ 11 \ -0m.o -ov.o nom.o I .o .H cm [05.0 Iom.o Iom.o .Ls moa .m woo.H .xs Cam .H mCOH ECCHECHC CH .2 mI0H x 0.~ I0H.H eouquosqv 43 0.¢ oHnmoN OH¥M- COH HMfiU—L On... OUHXOHUWI 05 Suvfl3 HOGO>MHh Sufi“? pmumuuHe AHHHV ECCHESH< wow mm>unu COHumwuHB UHHumEouonmouuumam .¢H HMDme .AHHHV ECCHECH< op HOCo>mHm mo OHumm 0H0: m.m 0.m m.N 0.N m.H 0.H m.0 0 _ p p P p I P _ i O . . -OH.o I0~.o o Iom.0 0 ..~ uoa.o O . :l o 0 OH 0 . . rom.o H o I0>.0 . nom.o I00.0 OLE I Vflm N .l OOOH .15 mos .H mCOH ESCHECH< CH .2 mIOH x 0.~ I QDUEQJOSQV 44 0.¢ .Hum oHume CoH Hmuoz 0» mConunmm «Cu CHHz HoCo>mHm Cqu pwumuuHB AHHHV ECCHBCHC Mom mm>uao COHumwuHB UHHumEouonmouuumdm m.m . caHHHv §CH53H< B HOGO>MHM M0 OHDMN 0H0: 0.m _ m.~ L .N .H 0.~ _ m.H 0.H m.0 _ . P .12 mov .N .12 Cam .H mCOH ECCHECH< CH .2 mIOH x 0.~ .mH mmDUHm I0H.o 10m.0 I0m.0 l0¢.o I00.o I0b.0 lom.0 100:0 I00.H eoueqxosqv 45 0.¢ .H"¢ mo OHHCC COH Hmum2 o» mponuvwm,mnv.CuH3 H0C0>mHm Cqu pmwmuvHa AHHHV ECCHECH< mom nobwau COHumHuHB.UHupwEoponmouuummm .AHHHV ECCHESHC ow HOC0>mHm mo 0Humm wHo2 m.m 0.m m.~ 0.~ m.H 0.H _ . P _ _ _ .mH mmeHh m.0 0 _ ‘ I .H .15 mos .m .13 Cam .H OH x o.~ mCOH ECCHECHC CH .2 ml I0H.o I0N.0 Iom.o I0¢.0 l0m.0 lom.o I230 Iom.0 lom.0 l00.H IOH.H BDUEQJOSQV 46 order even after standing for ten days. The addition of hydroxide ions to the metal ion aliquot prior to the fla— vonol addition presumably allows the hydroxide ions to add to the metal ion and form a metal ion species which is un— favorable to chelation. When the titration of aluminum with flavonol was car— ried out with the hydroxide to metal ion ratio below 2.5 : 1, the stoichiometry of the chelate which formed was found to be two aluminum ions to one flavonol. On standing, further chelation occurred more rapidly in the presence of base than did in the solutions to which no base was added. When the hydroxide to metal ion ratio was 2.5:1, a 1:1 flavonol to aluminum chelate was formed. In these solutions only a slight amount of further chelation occurred on stand- ing and the indicated stoichiometry did not change. In one titration at this hydroxide to metal ion ratio, a stoichi- ometry of 2:1 flavonol to aluminum ion was indicated but this could not be duplicated. At a hydroxide to metal ion ratio of 3:1 or greater the chelation of aluminum ions by flavonol appeared to be severely inhibited. The absence of any breaks in these spectrophotometric titration curves made the determination of the stoichiometry impossible. This inhibiting effect is probably due to the competition of hydroxide ions for the coordination sites on the metal ion. The fluorescence intensities of the solutions employed in the spectrophotometric titrations were also measured. 47 These data were plotted as spectrofluorometric titrations following both the chelate fluorescence at 450 mp and the flavonol fluorescence at 530 mp (Figures 17—22). The chelate emission was excited by both 365 and 405 mp radiation. The fluorescence of the flavonol was obtained by excitation with 365 mp radiation. The spectrofluorometric curves in general do not indi- cate the stoichiometries of the chelates as well as the spectrOphotometric titration data. This can be attributed to the complex nature of the origin of fluorescence. The chelate absorbance values are, in general, directly pro- portional to the amount of chelate which is present. On the other hand the emission of fluorescence is not only de- pendent on the concentration of the species, but is influenced by self—absorption of the emission radiation and by quenching effects of the solvent and other species in the solution. In spite of these difficulties the fluorescence data have proved to be valuable supplements to the spectrophotometric data in the interpretation of the spectral properties of the chelates. In one case the fluorescence data showed the existence of a chelate species which was not indicated by spectrophoto- metric data. The spectrofluorometric titrations curves may be in- terpreted in terms of the fluorescence characteristics of the various chelate species. A comparison of the titration curves which were obtained spectrophotometrically and spec- trofluorometrically on the same set of solutions indicates '48 .mCOH mUonuphm mo OUCOmnd on» CH.H0Co>mHm CHH3 owumuuHe AHHHV CCCHCCHC Com mm>uso CoHumuuHe oHuumsouoCHmouuuow .aH meome oaHHHv aafifigflafl OD. HOCOKKMHM HO Oflflmm 0H0: 0.¢ m.m com m.N oom mofl oofl m.o o . _ _ b b L F . . . 0 IOH - - - - - . rom . . . Iom . o . . Io¢ Iom ON 0 rlom o 4 I05 .U/0 9 low 0 .18 mom I CumCMHm>mz COHumyHuxm .H I00 .15 0mm .N I00H .1C oma .H o mCOH ESCHECHC CH .2 mloH x 0.N I0HH Karsuequl eoueosexontg 49 .nCOH mponuphm mo CUCCmCC 0CH CH AHHHV asCHECHC Cqu pmymuuHa H0C0>mHm mom m>wsu COHumwuHB UHuumEouoaHmouuummm .mH mmDUHm .HoCo>CHm on AHHHV sCCHsCHC mo OHCCC mHos 0.m 0.~ m.H 0.H m.0 0 b - P b . F .LE mom I CmemHm>m3 COHpmpHUxm HOCO>mHm CH .2 mIOH X 0.0 I OOH I.OHH °nm 05p 19 fiqtsuequl aoueosaxontg SO .HHH OHHCC COH Hmuw2 on mponuphm ecu Cqu H0C0>mHm Csz pmumuuHB AHHHV CBCHECH< Mom mw>uzu COHUCHHHB UHuumEouosHmouuuwam .mH mmDGHm .AHHHV ESCHSCH< on HOC0>mHh mo OHumm mHo2 0.¢ m.m 0.m m.~ 0.N m.H 0.H m.0 0 p . . _ 1 _ r _ 0 IS - o . o Iom H D l s H p. 0 . o Iom . . . -3 x O I . JIom . 18 ON ‘ O u I05 I00 .H r .LE mmm I CmemHm>m3 COHumuHUxm 00 .lE . - L 0mm N rOOH . E 0m¢ .H u o mCoH ESCHECHC CH .2 mI0H x 0.~ I0HH Kirsuaqux eoueosexonta 51 .Hum.~ oHumm CoH Hope: 0» mConucsm mCu Cqu HoCo>mHm Cqu CmumuuHe AHHHV sCCHCCHC Com mm>usu COHHmuuHe UHunmsouoCHmouuumCm .om mmoon _ . p k _ _ _ 0 I0H 'q 0 . I u I I o a I . o O I o o .ION . . . . . Iom . - . . IOC 0 ON 0.H o 0 O Iom I00 Ion . I00 .15 mom I CvoCmHm>m3 COHumuHuxm .Iom .LE . L 0mm N ,I00H . E 0m¢ .H mCOH ESCHECH< CH .2 mI0H x 0.N I Kitsuaqul eoueaseaontg 52 .Hum oHumm CoH Hmums 0» muonuusm mCu Cqu HoCo>mHm Cqu nmumuuHe HHHHV CCCHCCHC Com m>usu CoHumuuHe UHunmsouoCHmouuumCm .Hm mmoon .AHHHV BCCHECH< on H0C0>mHm mo 0Huwm mHo2 0.¢ m.m 0.m m.~ 0.~ 0.H 0.H m.0 0 wI r H . L 1 _ 0 I0H ION I00 I0¢ Tom I00 I00 I00 I00H .LE m0m I numCmHm>mz CoHpmuHuxm mCOH ESCHECH< CH .2 mI0H x 0.0 °nm 035 as thsuequx eoueosesontg 53 .Hue oHumm COH Hmuas 0» mConqum «CH Csz HoCo>mHm CHH: omumuuHe AHHHV sCCHsCHC Com o>uso CoHumuuHe oHuumsouoCHmoupumCm .NN mCoon .AHHHV CCCHCCHC on HoCo>mHm mo oHumm mHos 0.¢ m.m o.m m.~ o.~ m.H c.H m.o o _ l- b F p p , - L I00 I00 I00 I c: c: H °nm ogg as fiqrsuequx eoueoseJontd .LE m0m I CumCmHm>mz COHumuHuxm mCOH ECCHECH< CH .2 mI0H x 0.0 54 that the absorption and emission properties of the chelates are not directly related to one another. In the titrations of aluminum ions with flavonol, the absorbance data indicate that with successive additions of flavonol the concentration of the chelate species increases. The fluorescence data, conversely, indicates that the highest chelate fluorescence occurs when only a small amount of flavonol has been added, giving a large ratio of aluminum ions to flavonol. The chelate emission decreases rapidly with further addition of flavonol, attains a poorly defined break, and then decreases only slightly with further addition of reagent after the break. The emission of the flavonol during these titrations paralleled closely the absorbance data for the flavonol band at 344 mp. Prior to the break the flavonol emission exhi- bited a slight increase because of the contribution of the chelate emission to the emission of the flavonol at 530 mp. After the break in the titration curve the flavonol emission increased as expected since excess reagent was present. The negative deviation from linearity exhibited in these cUrves was presumably due to attenuation of the emission by self- quenching.of flavonol. The reagent emission, therefore, was as expected and requires no further interpretation. The anomalous results obtained by following the chelate emission may be interpreted in terms of the fluorescence characteristics of the various chelates. The abnormally high chelate fluorescence intensity in those solutions which 55 contained an excess of aluminum ions indicated the existence of a species with a high aluminum ion to flavonol ratio. With further addition of flavonol, the aluminum ion to fla- vonol ratio in the chelate was lowered, presumably to the 2:1 aluminum ion to flavonol species which exhibited a much smaller fluorescence intensity. Further addition of flavonol past the break decreased the fluorescence intensity only slightly. This effect may be attributed to further formation of the less fluorescent chelate or to a slight quenching effect of excess flavonol. With an increasing hydroxide to metal ion ratiO‘ up to 2.5:1, the following effects on the spectrofluorometric titra- tion curves may be observed: the intense fluorescence prior to the break is diminished significantly, the position of the break shifts toward higher flavonol to aluminum ratios, and the fluorescence intensity after the break is decreased only slightly. The effect of the base appears to be much more signif- icant on the fluorescence of the chelates withmahigher ratio of aluminum to flavonol than on those chelates which have a 1:1 or greater ratio of flavonol to aluminum. An increase in the hydroxide to metal ion ratio up to 2.5:1 appears to aid further chelation of aluminum ion by flavonol. Spectrophotometric data indicate that the chelation of aluminum with flavonol is severely inhibited by the pre- sence of excess base presumably because of the competition of the hydroxide ions with flavonol for the coordination 56 sites on the metal ion. When the hydroxide to metal ion ration is 3:1 or greater the flavonol-aluminum chelates no longer emit any blue fluorescence. The only fluorescence emission which occurs in these solutions when excited by 365 mp radiation is that emission at 530 mp due to the un- chelated flavonol. Examples of the fluorescence titration curves which followed the flavonol emission in the presence of excess base are shown in Figures 21-22. These curves show the increase in the unchelated flavonol, exhibiting the negative deviation from linearity which is typical of concentrated solutions of a fluorescent species. There was no emission when these solutions were subjected to 405 mp radiation. The effect of base on the fluorescence of the flavonol— aluminum chelates is not unusual. It is quite reasonable that the emission of fluorescence by a chelate is affected signif- icantly by changes in the remaining sites in the coordina- tion sphere of the metal ion. The addition of negative ligands such as hydroxide ions to the coordination sphere of the aluminum ion will make the aluminum ion appear less electrOpositive toward the flavonol ion with which it is Chelated. This effect would weaken the flavonol-aluminum ion bond and would allow for more dissipation of absorbed energy through vibrational losses. That the presence of excess base influences the elec- tronic transitions in the chelate has been demonstrated by the shift in the wavelength of the absorption peak attributed 57 to the chelate under these conditions. This same effect may be responsible for the decrease in fluorescence intensity by decreasing the probability of the absorption transition which gives rise to the emission. Spectrofluorometric titrations were employed to de- termine the stoichiometry of the highly fluorescent species indicated in the previous titration curves. Since the un- usually high fluorescence intensity occurred when the ratio of aluminum ions to flavonol was high, spectrofluorometric titrations of aluminum ions with flavonol (Figure 23) were carried out in the region where the metal ion to flavonol ration varied from 2:1 to 20:1. These titration curves indicate the existence of a species with a stoichiometry of six aluminum ions to one flavonol. The spectrofluorometric titration of flavonol with aluminum ions in the region where the metal ion to flavonol ratio varies from 1:1 to 10:1 gives two breaks corresponding to the 2:1 and 6:1 aluminum to flavonol species (Figure 24). In both of the above titrations the flavonol aliquot was added to the aluminum solution to maintain at all times an excess of aluminum to flavonol. This was done to prevent the formation of higher chelates which would retard the for- mation of the 6:1 species. Further evidence for the existence of the 6:1 aluminum ion to flavonol species is given by observing the chelate fluorescence of various ratios of aluminum to flavonol with time. It is postulated that the addition of flavonol to 58 100 8.0 x 10"5 in Aluminum long 5_ l. Excitation Wavelength E _ ’ 365 mp. 53 90 ' 2. Excitation Wavelength g 80_. . 405 mp. 44 m >.' 70—1 " .p H ‘8 60- - B 2. .5 50C 0 8 40- m 52’ 3o . m 1 u 8 F* 20- m 10‘ , O I I I I I I I ”I O 20 10 6 5 4 3 2.5 2 Mole Ratio of Aluminum (III) to Flavonol. FIGURE 23. Spectrofluorometric Titration Curves for Aluminum (III) Titrated with Flavonol to Indicate the Existence of a 6:1 Aluminum to Flavonol Species. 59 2.0 x 10"5 M. in Flavonol 100— DU slit = 0.3 mm. Excitation Wavelength - 365 mp. KO O l (I) O l 70- 60- 50- 40- 301 Fluorescence Intensity at 450 mp. 20- 10 O / 1 I I I I I I I I I 0 l 2 3 4 S 6 7 8 9 10 Mole Ratio of Aluminum (III) to Flavonol. FIGURE 24. Spectrofluorometric Titration Curve for Flavonol Titrated with Aluminum (III) to Indicate the Existence of a 6:1 Aluminum to Flavonol Species. 60 aluminum ion solutions initially forms the 6:1 aluminum ion to flavonol species. Therefore in solutions which contain aluminum ions in excess of the 6:1 ratio required for this species, the chelate fluorescence should increase initially as the 6:1 species is formed and then remain constant since there is no tendency toward further chelation or toward the formation of chelates with lower aluminum ion to flavonol ratios. On the other hand, when flavonol is present in ex- cess of the amount necessary for the formation of the 6:1 species, the chelate fluorescence should initially be at a high value since the 6:1 species will form first. The chelate fluorescence should then decrease with time as the excess flavonol chelates with the aluminum ions and forms the less fluorescent 2:1 aluminum to flavonol species. An examination of Figure 25 indicates that this postu- late is correct. The chelate fluorescence of solutions con- taining an aluminum ion to flavonol ratio of 2:1, 3:1 and 4:1 decreases with time from an initial value. When the ratio of aluminum to flavonol was 7:1 or 7.5:1 the chelate fluorescence intensity rapidly attained a high value and re- mained constant with time. In all of these solutions the fluorescence emission was excited with 405 mp radiation and was measured at 450 mp. During the time the fluorescence intensity changes were being measured, the chelate absorbance at 405 mp was _ also recorded. These absorbance values indicated that fur- ther chelation took place during the time the fluorescence 61 10 Aluminum to Flavonol Ratio a 7.5:1 9—++—e—4+—er i,______‘ 0.. O 6 lO-H Aluminum to Flavonol Ratio 2 7.0:1 > ?,14>——¢P——“—°—"3 H—4r———4r———4¥———4h———Q————O :3 I :3 0‘} .8 z a? 1'3 30~p 8 . C 20 _ m 8 Aluminum to Flavonol Ratio = 4.0:1 3101 o s H 2' o- .,..| I? & gZOH I: m .C U10 ‘ Aluminum to Flavonol Ratio = 3.0:1 O __ 412 —: 2 10 ‘ 0 Aluminum to Flavonol Ratio = 2.0:1 O I I I I I I I I - I" '7 2 4 6 8 10 12 14 16 18 20 Time (Minutes) FIGURE 25. Variation of Chelate Fluorescence with Time after Mixing for a Series of Aluminum to Flavonol Ratios. 62 intensities decreased. This eliminates the possibility that the decrease in the fluorescence values was due to a decrease in the amount of flavonol which was chelated. The Determination of Flavonol-Aluminum Chelate Molar Absorptivities In the presence of a fivefold or greater excess of aluminum ions, flavonol exists entirely in the chelated form, as indicated by the constant absorbance values at 405 mu. The chelate species was assumed to contain one flavonol ion and the concentration of the chelate was taken as equal to the concentration of the flavonol added. The molar absorp- tivities of the chelate,per flavonol unit, were calculated from the limiting absorbances at 405, 344, and 326 mp. The following values were obtained by this method: £405 a 2.07 3 E 3 4 . x 10 , e = 3.4 x 10 , 326 = 6.5 x 10 . 344 Potentiometric Titrations of the Chelate with Ease Aluminum ions may be titrated with base potentiomet- rically in absolute ethanol. The end point is well defined and occurs at a hydroxide to metal ion ratio of 2.5:l.0. (Figure 26) There was no indication of any precipitate for- mation during the titration of either the aluminum solutions or the chelate solutions with base. With the addition of flavonol to the aluminum solu- tions, an additional break in the curve (Figures 27-28) 63 11.. 2.0 x 10'4 M. in Aluminum Ions lO_I Apparent pH I T’ I I I I I I 0 0.5 1.0 1.5, 2.0 2.5 3.0 3.5 4.0 Mole Ratio, Hydroxide Ions to Aluminum Ions. FIGURE 26. Potentiometric Titration Curves for Alum- inum (III) Titrated with Hydroxide Ions in Absolute Ethanol. 64 .HOCmem quHomQC CH mCOH mpdwouphm CuHB mOHumm HOCo>mHm O». ESCHEHHH< m30flhfl> MO MGOHNMh—UHB 003. .HOH WOESU SOH#MH#HB UflufimeOHucmfiom .FN HMDWHHW .AHHHV ECCHECHC o» mponuUMm mo 0Humm qu2 N H 0 m N H 0 m N H 0 IL _ _ FKVFI Cf _ .IxLI C. h ‘ I TH I 032 Hum I 03mm H3. I 03mm Hoco>mHm op ESCHECH< HOCo>mHm ow ECCHECH< HOCo>mHm o» ECCHESH< mCOH EWCHECHC CH .2 ¢I0H x 0.N Hd queJeddv 65 .HoCmCum muCHomn< CH mCOH prxowphm Cqu mOHumm HOCo>MHm on ECCHECH< mCoHum> mo mCoHumuuHe CC» Com mm>usu COHumuuHe UHuumsoHqupoC .am mmDon .HHHHV CCCHCCHC on mConunsm mo oHumm mHos m m H o m N H o H _ h 1 _ b _ o I.H I_~ I m I a I.m I m I.a muH I OHumm NHH I OHCCC HOCo>mHm Op ESCHESH< HOCO>0Hm OH ESCHEDH< I.0 I m . . IIOH meow EzcmezH< CH 2 HIOH x o m Hd quexeddv 66 appears prior to a hydroxide to metal ion ratio of 2.5:1.0. Increasing the amount of flavonol present to give a ligand to metal ion ratio of 1:1 shifts this first break to a hy- droxide to metal ion ratio of 2.0:l.0. Further addition of flavonol does not shift this break but does cause it to be better defined. In all of these titrations a second break always occurs at a hydroxide to metal ion ratio of 2.5:l.0. The addition of flavonol to aluminum solutions in ab— solute ethanol always causes a rise in the apparent pH. This is contrary to what is expected since in order for flavonol to chelate with a metal ion a proton must be released from the flavonol molecule. This observation together with the fact that less base is required to titrate flavonol-aluminum mixtures than is required to titrate aluminum ion alone can be explained by assuming that the incoming flavonol molecule displaces two coordinated ethoxide ions from the aluminum ion species. One of the released ethoxide ions consumes the proton which is released from the flavonol and the second ethoxide consumes one of the protons released by the solvolysis of the metal ion. This must be the process by which the fla- vonol adds to give the 2:1 aluminum to flavonol species since this is the species which predominates at hydroxide to metal ion ratios below 2.5 to 1.0. Further addition of flavonol must proceed by the replacement of one coordinated ethanol molecule and one coordinated ethoxide ion since the amount of base required to reach the second end point is not depend- ent on the amount of flavonol present. 67 The Effect of Acid on the Chelates The addition of ethanolic hydrochloric acid to solu- tions containing the 6:1 aluminum to flavonol species up to an acid to metal ion ratio of 10:1 does not affect the amount of flavonol which is chelated as determined by the absorbance at 405 mp. The fluorescence intensity of the chelate is en- hanced about 10 per cent with this concentration of acid. The dependence of the absorbance and fluorescence of the 2:1 aluminum to flavonol chelate on apparent pH is pre- sented in Figure 29. Even though the more acidic solutions decrease the amount of flavonol which is chelated according to the absorbance values at 405 mp, the fluorescence inten- sity of the chelate is greatly enhanced by the presence of acid. The dependence of the absorbance and fluorescence of the 1:1 flavonol to aluminum chelate on the addition of small amounts of hydrochloric acid is shown in Figure 30. The acid affects the amount of flavonol which is chelated insignificantly but the fluorescence intensity is again greatly enhanced. Solutions of the 1:1 chelate which have been prepared by the addition of base followed by an equiva- lent amount of acid are more fluorescent than the 2:1 alum- inum to flavonol solutions. The enhancement of the fluorescence of the chelates with the addition of acid must be attributed to the effect of the acid on the species which complete the coordination sphere of the metal ion. 68 1.00 100 All Solutions are 4.0 x 10"5 M. in Alum- 0°90 _ inum (III), 2.0 x10“6 M. in Flavonol. — 90 1. Fluorescence Intensity of 450 mp. 0080.4 2. Absorbance at 405 mp. p.80 Absorbance Fluorescence Intensity O , L O I ' I I I l l r l T O 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Apparent pH. FIGURE 29. Dependence of the Absorbance and Fluorescence of the 2:1 Aluminum to Flavonol Chelate on Apparent pH. 69 1.0 0.90-- I. 0080— .—. a} 0.70— I. U) 3 0.60"J 2. I ‘fi L— 4., rd Q) 0050— —- U C 8 ‘4 0.40- i— O V} g 0030"“ I—0 0.20—1 ,1 0.10 _ O IIIIIfi I II 0 1 2 3 4 5 6 7 89 Milliliters of 1.0 x 10"3 M. HCl Added. 100 90 80 70 60 50 40 30 20 10 FIGURE 30. Variation of the Absorbance and Fluorescence of the 1:1 Aluminum to Flavonol Chelate with the Addition of Hydrochloric Acid. (1.) Fluorescence Intensity at 450 mu. 70 The acid would tend to neutralize any coordinated ethoxide ions and thereby remove any negative ligands from the aluminum ion except for the flavonol. The net result would be to make the aluminum ion appear more electroposi- tive toward the flavonol. This would result in the aluminum ion exerting a stronger attraction toward the electrons in the flavonol molecule and it is probably this effect which enhances the fluorescence. The Effect of Water on the Flavonol-Aluminum Chelates The effect of the addition of water to chelate solu- tions was measured. In Figure 31 is presented the fluores- cence intensity at 450 and the absorbance values at 405 due to the chelate. The addition of water has an inhibiting effect on the chelate formation but enhances the fluores- cence of the chelate greatly. This observation is contrary to the expected result. The solvent would become more polar with the addition of water and more polar solvents allow for more external conversion of the absorbed energy. The en- hancement of the chelate fluorescence can only be explained by assuming that the water molecules replace some solvated ethanol molecules from the aluminum ions and make the chelate species more favonable for fluorescence. The precipitate which forms in the solutions containing above 50 per cent water is probably due to the insolubility of the aluminum ion species. When water was added to ethanol’ solutions containing only the aluminum salt, white precipitates 110 ;i_1oo s 5: 9o 21‘ .p m 80 5* H 70 In 3 60 4.) s 50 a) 2 m 40 U 8 u 30 O 3 m 20 10 0 FIGURE 71 of the Addition of Water. 1 __l.10 _ -—l.OO l. _ ._0.90 _ -0.80 4.0 x 10"5 M. in 7 Aluminum (III), “’0°7o 2.0 x 10-5 M. in, - Flavonol. I.0.60 U ‘ 1. Fluorescence at — 450 mp. “‘0'50 . 2. Absorbance at — 405 mp. "'0-40 — . —0.30 2. y _ ‘ __0.20 1 .-0.10 i I I‘ I I I O O 10 20 3O 40 50 Per Cent Water Added, by Volume. 31.- Dependence of the Absorbance and Fluorescence Absorbance at 405 mp. 2:1 Aluminum to Flavonol Chelate on the 72 were formed. Flavonol, however, remained in solution well past the 50 per cent water content. Preparation of a Solid Flavonol-Aluminum Chelate Several attempts were made to prepare a solid flavonol- aluminum chelate throughout the course of this study. Some of the procedures which were unsuccessful are listed here: 1. Precipitation of an aluminum-flavonol chelate from an acetic acid- sodium acetate buffered solution analogous to the preparation of tris (8-hydroxyquinoline) aluminum. The low solubility of the flavonol in this medium caused all preparations to be contaminated with excess reagent. 2. Air evaporation of concentrated mixtures of flavonol and aluminum ion in ethanol to try to exceed the solubility of the chelate. No precipitates were obtained in these attempts without taking the solution to dryness. The extent of chelate formation was probably limited to the 2:1 aluminum to flavonol species which is probably a charged species. 3. Refluxing of mixtures of flavonol and aluminum ion in ethanol to promote the chelation reaction followed by evaporation to a smaller volume to precipitate the chelate. No precipitates were obtained by this method. Some darkening of the solutions occurred indicating the decomposition of fla- vonol. 4. The addition of stoichiometric amounts of base to various ratios of flavonol and aluminum ion in ethanol to try to force the formation of higher chelates which might 73 precipitate. A solid phase was repeatedly obtained by this method but analysis for aluminum and flavonol content in— dicated that the solid was contaminated with the sodium salt of flavonol. No method of separating the two substances could be devised. The most successful preparation of a solid chelate was carried out in the following manner: To fifty milliliters of ethanol containing 0.200 moles of flavonol were added twenty milliliters of 0.01 M sodium hydroxide. The flavonol salt which formed was kept in solution by stirring and warm- ing the solution. To this mixture was now added dropwise with stirring ten milliliters of ethanol containing 0.066 moles of aluminum chloride. A fine yellow precipitate was obtained. The amount of precipitate was increased by the evaporation of the solution to one-half the volume by a stream of air. The precipitate was collected on filter paper and washed with cold ethanol and was dried at lOSo’C. After the initial drying period there was no further weight loss. There was no apparent discoloration of the precipitate at this temperature. All attempts to recrystallize this pre- cipitate from a variety of solvents were unsuccessful. The solid chelate was prepared for the analysis of aluminum by decomposing the chelate in a 1:1 nitric acid- perchloric acid mixture. This treatment destroyed the chelating agent. In samples where the flavonol was not destroyed by this method it interfered in the EDTA titration. The acid solution was neutralized, an excess of standard 74 EDTA solution was added, the pH was adjusted to between 7 and 8, and the excess EDTA was titrated with a standard zinc sul- fate solution using Eriochrome Black'ras the indicator. The flavonol content was determined by destroying the chelate with concentrated hydrochloric acid in ethanol and comparing the absorbance of flavonol at 344 mu to a standard curve prepared under the same conditions. The results of the analyses are: Found: %Al, 3.76, 3.40; % flavonol, 93.7, 93.9; calculated for 3:1 flavonol to aluminum chelate: %Al, 3.64; % flavonol, 96.7. The infrared absorption spectrum of this solid prepar- ation was obtained by potassium bromide pellet technique and is presented in Figure 32. This spectrum was compared to the infrared spectra of flavonol and the sodium salt of fla- vonol obtained by the same technique. The chelate spectrum was significantly different from the other spectra to assume that it was a different species. The melting point of the solid was determined in a combustion tube in a flow of nitrogen by the use of a thermo- couple for the temperature measurement. The solid melted sharply to a deep yellow liquid without apparent decomposi- tion in the region 330-3400. 75 .OHMHOSU ESCHECH< H0C0>MHm UHHom mzu mo Enuuummm COHuauomnd UmHCHMCH mCOHUH2 CH CvanHm>m3 0 C 0 0 ¢ .mHm UHHom mCu mo Eswuuomm COHuauonnd pmHCHMCH MCouUH2 CH CumaHm>m3 mH ¢H mH NH HH .mNm mmDOHm 0H ..._ _ _ _ _ _ _ 4. 0H 0 <3 :3 0 KO uI ¢ In eoueqitmsuell dues 18d 0 t\ O 0 om 00H SUMMARY AND CONCLUSION 78 From spectrophotometric evidence it has been concluded that flavonol does not undergo keto-enol tautomerization in solution. Observations have been presented which indicate that flavonol exists in solution entirely in the enol form, which is stabilized by conjugation and hydrogen bonding. The differences in the ultraviolet absorption spectrum of fla- vonol and the unsubstituted flavone molecule have been inter— preted in terms of the flavonol species which is thought to exist in solution. The appearance of a long wavelength ab— sorption band in the spectra of flavonols which is absent in flavones has been attributed to the auxochromic effect of a strong electron donating group adjacent to a conjugated carbonyl group. This auxochromic effect is thought to be supplemented by the stablization of certain excited states by internal hydrogen bonding between the 3-hydroxy group and the carbonyl group. The fluorescence spectrum of flavonol has been observed in a variety of solvents. Differences in intensities and wavelengths of maximum emission in the different solvents were explained in terms of solvent interactions with the molecule. In polar solvents the polar excited states of fla- vonol were stabilized and the maximum emission was shifted to higher wavelengths. The interaction of the polar solvent molecules with the flavonol molecules, however, allowed for dissipation of the absorbed energy through external converé sion and the fluorescence intensity was greatly reduced in the polar solvents. A blue emission band also appeared in 79 the fluorescence spectrum of flavonol in polar solvents which was absent in nonpolar solvents. It is thought that it is this emission band which is shifted and intensified upon chelation of the flavonol to give the brilliant bluecbhelate fluorescence. The stoichiometries of the flavonol-aluminum chelates which formed in absolute ethanol were investigated. Upon the addition of flavonol to a solution of aluminum ions the first chelate species which formed was a 6:1 aluminum to flavonol species. Since it is unlikely that a flavonol mol- ecule could bring about the formation of a polymeric aluminum species in any manner, it is assumed that aluminum ions exist in ethanol solutions as six—membered species. The formation of the 6:1 aluminum to flavonol chelate then occurs by the addition of one flavonol to the six membered metal ion species in solution. Other authors have postulated the existence of six-membered aluminum species in water (9, 46). Ohnesorge (41) stated from potentiometric titration data that the simplest aluminum species in ethanol is at least a binuclear species and is probably some polymer of this. The fact that the fluorescence data indicate a 6:1 species while the absorbance data do not, can be explained in terms of the more sensitive nature of fluorescence. The absorbance peak which is attributed to the chelate arises from the perturbation of an electronic transition in the fla- vonol molecule. Therefore an increase in the chelate ab— sorbance peak only indicates that more flavonol is chelated 80 with the metal ion. A variation in stoichiometry will only be indicated by a spectrophotometric titration when there is a large enough difference in the stability of successive complexes and if the absorption characteristics of the spe- cies vary. On the other hand, the emission of a chelate is much more sensitive to the environment of the metal ion. If successive chelates have enough difference in their fluores- cence characteristics then the spectrofluorometric titration will yield a break even though no break is indicated from the complementary absorbance data. With the further addition of flavonol to aluminum ion solutions in excess of the 6:1 aluminum ion to flavonol ratio, a species is formed which has the empirical stoichiometry of two aluminum ions to one flavonol ion. -This 2:1 aluminum ion to flavonol species forms readily without the addition of sodium hydroxide to the solutions and is also the initial species formed in solutions where the hydroxide to metal ion ratio is less than 2.5:1. On standing there is a tendency for further chelation in these solutions. The presence of base promotes the formation of higher chelates and when the hydroxide to metal ion ratio is 2.5:1 the species which forms immediately has a stoichiometry of 1:1 flavonol to aluminum ion. Above a hydroxide to metal ion ratio of 3:1, the base inhibits the formation of the chelates and destroys the fluorescence of the chelates. The 6:1 chelate species exhibits the most intense fluorescence. The 2:1 aluminum to flavonol species is more 81 fluorescent than the 1:1 species. The presence of acid or water enhances the fluorescence of the chelates greatly. This effect has been attributed to the effect of these sub- stances on the coordination sphere of the metal ion. The fact that in absolute ethanol there is not much of a tendency to form chelate species containing more than one flavonol per aluminum ion indicates that the aluminum ions exist in some polymeric form in this solvent. When the aluminum salt solution is prepared in water and the etha— nolic flavonol solution is added to it, a 3:1 flavonol to aluminum species is formed immediately even in the absence of base (57). If some of the coordination sites on the aluminum ions were occupied by the formation of metal- oxygen bridges, then the number of bidentate flavonol ions that could add would be reduced. Several attempts were made to determine the formation constants of the flavonol-aluminum chelates but because of the complex nature of the species no evaluation of these constants could be done. The most severe limitation of any estimation of these formation constants was the fact that no equilibrium appeared to exist between the chelates and the unchelated reactants. The chelates, once formed, showed no tendency to dissociate. On standing, further chelation occurred slowly until one of the reactants was essentially consumed. Another limitation on the attempts to measure forma- tion constants was the inability to obtain meaningful 82 measurements of the hydrogen ion concentration in absolute ethanol. Any equilibrium constant involving the chelation of metal ions by flavonol must involve the hydrogen ion concentration. An attempt to maintain a constant hydrogen ion concentration to evaluate apparent formation constants failed when no suitable buffer in ethanol could be found for the apparent pH region where the chelates formed read- ily. The evaluation by the fact that the were not known. The mined from titration the 6:1 species, the of formation constants was also negated actual stoichiometries of the chelates empirical stoichiometries were deter- data. With the possible exception of actual chelate stoichiometries are probably some multiple of these empirical ratios. LITERATURE CITED 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 84 Almassy, Gy., and Vigvari, M., Acta Chim. Acad. Sci. Hung., 22, 243 (1959); C.A., £4, 7429C. Aronoff, S., J. Org. Chem., 5, 561 (1940). Babko, A.K., and Kish, P.P., Dopovidi Akad. Nauk. Ukr. R.S.R., 1961, 1323; C.A., 26, 10900e. Ballantine, J.A., and Whalley, W.B., J. Chem. Soc., 1956, 3224. Bottei, R.S., and Trusk, B.A., Anal. Chem.,‘3§, 1910 (1963). _ Brahm, D., and Jain, B.D., Proc. Indian Acad. Sci., Sect. A‘gl (3), 142 (1963). Brahm, D., and Jain, B.D., J. Indian Chem. Soc., 42, 269 (1963). Briggs, L.H., and Colebrook, L.D., Spectrochim. Acta, 18, 939 (1962). Brosset, C., Biedermann, G., and Sillen, L.G., Acta Chem. Scand., 8, 1917 (1954).' Brune, w., Arch. Pharm., 288, 205-16 (1955). Chen, Fa Ching, and Lin, Chih-Hsiung, Formosan Sci.,‘§, 81 (1952); Cvo , fl, 3972C. ‘. X '. Coyle, C.F., and White, C.E., Anal.Chem., 32, 1486 (1957). I Detty, W.E., Heston, 8.0., and Wender, J.H., J. Am. Chem. Soc., 11, 162 (1955). Fleck, L.L., Ph.D. Thesis, Michigan State University, East Lansing, Michigan (1961). Francois, H., and Grandclement, A.M., Compt. Rend. Journee Etude Protect. Contre Dangers Beryllium, Gif- Sur-Yvette, France, 1961, 159, 161; C.A., 62, 7445f. Geissman, T., "The Chemistry of Flavonoid Compounds," Macmillan, New York, (1962). Golovina, A.P., Alimarin, I.P., and Stepanov, N.F., Nauch. Doklady Vysshei Shkoly, Khim. i Khim. Tekhnol., 1958, no. 2, 285; C.A., £3, l698lg. Goppelsroeder, F., Z. anal. Chem., 1, 195 (1868). Hagedorn, F., and Neu, R., Arch. Pharm., 286,486 (1953). 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 85 Hagedorn, P., and Neu, R., Arch. Pharm., 287,70 (1954). Hergert, H.L., and Kurth, E.F., J. Am. Chem. Soc., 15, 1622 (1953). Hiiro, K., Bull. Chem. Soc. Japan, 34, 1748 (1961). Harhammer, L., and Hansel, R., Arch. Pharm., 285, 438 (1952). ‘ Hérhammer, L., and Hansel, R., and Hieber, W.,.Z. anal. Chem., 148, 251 (1955). Inglett, G.E., J. Org. Chem., 23, 93 (1958). Jaffe, H.H., and Orchin, M., "Theory and Applications of Ultraviolet Spectroscopy," Wiley, New York, (1964). p. 213. Jatkar, S.K., and Mattoo, B.N., J. Indian Chem. Soc., 33, 623, 641 (1956). Jurd, L., and Geissman, T., J. Org. Chem., 31, 1395 (1956). Kanno, T., Sci. Repts. Research Inst., Tohoku Univ., Ser. A 11, 141 (1959); C.A., 23, 19666i. Kanno, T., Sci. Repts. Research Inst., Tohoku Univ., Ser. A.$39 no. 2, 91 (1961); C.A., 52, 20757f. Katyal, M., and Singh, R.P., J. Indian Chem. Soc., 49, 191 (1963). Katyal, M., and Singh, R.P., Proc. Indian Acad. Sci., Sect. A, 21 (4), 259 (1963). Konstanecki, St. v., and Feurstein, W., Ber., 1898, 31 710. Korenman, I.M., Sheyanova, F.K., Shcherbakova,:Z.F., Trudy po Khim. i Khim. Tekhnol., 3, 303 (1960); C.A., 56, 14231. Looker, J.H., and Hanneman, W.W., J. Org. Chem., 21, 381 (1962). Milkey, R.G., and Fletcher, M.H., Anal. Chem., 38, 1403 (1956). Milkey, R.G., and Fletcher, M.H., J. Am. Chem. Soc., ‘12, 5425 (1957). 86 38. Neelakantam, K., and Row, L.R., Proc. Indian Acad. Sci., 14A, 307 (1941). 39. Neu, R., and Hagedorn, P., Naturwissenschaften, 40, 411 (1953). 40. Neu,R., Z. anal. Chem., 153, 95 (1956). 41. Ohnesorge, W.E., and Capotosto, A., J. Inorg. Nucl. Chem., 24, 829 (1962). 42. Oka, Y., Tanaka, R., and Umekara, M., Nippon Kagaku Zasshi, 83, 699-703 (1962); C.A., 22, 10761. 43. Oyamada, T., Bull. Chem. Soc. Japan, 12, 182 (1935). 44. Paech, K., and Tracey, M. V., "Modern Methods of Plant Analysis," Springer-Verlag, Berlin, (1955) vol. 3, pp. 450- 496. 45. POpa, Gr., Negoiu, D., Luca, C., and Baiulescu, Gh., Rev. Chim. Acad. Rep. Populaire Roumaine, 6, 87 (1961); C. A., 56, 66490. 46. Pungor, E., and Zapp, E.E., Z. anal. Chem., 197, 404 (1963). 47. Rao, P. S., and Seshadri, T. R., Proc. Indian Acad. Sci., 17A,119 (1943). 48. Rutkowski, W., Chem. Anal. (Warsaw),‘§ (3), 389 (1963); C.A., £32: 34749. . 49. Samsoni, Z., At. Kozlemenyck, 2, 53 (1960); C. A., 59, 13488f. 50. Shaw, B.L., and Simpson, T.H., J. Chem. Soc., 1952, 4638, 5027. 51. Shaw, B.L., and Simpson, T.H., J. Chem. Soc., 1955, 655. 52. Shkrobof, E.P., Zh. Analit. Khim., _1, No. 2, 186 (1962); C.A., 26, l4917g. 53. Skarzynski, 8., Biochem. Z., 301, 150-69 (1939). 54. Swain, T., Chemistry & Industry, 1954 1480. 55. Szabo, 2.6., and Beck, M.T., Acta Chim. Acad. Sci. Hung., 4, 211 (1954). 56. Van Eenenaam,‘R.H., M.S. Thesis, Michigan State Uni- versity, East Lansing, Michigan (1964). 57. 58. 59. 60. 61. 62. 87 Wambach, D.A., Private communication. Weiler, J.E., M.S. Thesis, Michigan State University, East Lansing, Michigan (1961). Wheeler, T.S., Record Chem. Progr. (Kresge-Hooker Sci. Lib.), _1__8_, 133 (1957). White, C.E., and Lowe, C.S., Ind. Eng. Chem., Anal. Ed., 12, 229 (1940). White, C.E., Hoffmann, D.E., and Magee, J.S., Spectro- chim. Acta, 2, 105 (1957). Zechmeister, L., "Fortschritte der Chemie Organischer Naturstoffe," Springer-Verlag, Vienna, (1959) vol. 17, pp. 1-590 $115511? ‘I ...4 LA—_. \ IIIHHIH IHHHII 93 03 75 5279 l WWIIIIIIWHIIIIIU IIIIIIIII