Studies directed at the stereoselective synthesis of flavonoids through the hydrogenation of prochiral precursors

Loading...
Thumbnail Image
Date
2008-11
Authors
Van Tonder, Johannes Henning
Journal Title
Journal ISSN
Volume Title
Publisher
University of the Free State
Abstract
English: The in vitro studies of biologically active flavonoids are hampered by the inaccessibility of all monomeric units in enantiomerically pure form. Although a number of these scalemic flavonoids can be obtained synthetically, current methodologies are tedious and often result it low yields and e.e.s being obtained. Stereoselective conjugated hydrogenation of the more readily available prochiral a,b-unsaturated carbonyl flavonoid motifs, i.e. flavones, isoflavones and flavonols, provide a plausible solution to this problem and was therefore investigated during the first part of this dissertation. Since the stereoselective 1,4-reduction of an a,b-unsaturated system also contains an element of regioselectivity, a regioselective hydrogenation study was conducted first before the reaction would be extended to include stereoselective aspects. Because Wilkinson’s catalyst is commercially available, has the potential to be modified with chiral ligands and is a well known hydrogenation catalyst, it was chosen as catalyst for the initial evaluation of the idea. While Wilkinson’s catalyst is well known for hydrogenation of ordinary alkenes, no literature on it being used for the reduction of a,b-unsaturated carbonyl compounds could be found, so the effectiveness of this catalyst in the hydrogenation of differently substituted nonflavonoid substrates were first investigated. In this regard a solvent study on chromone, evaluating acetone, tetrahydrofurane and dichloromethane (DCM), showed DCM to be the solvent of choice for this reaction and led to the formation of 4-chromanone in 7.8 %, 14.3 % and 46.2 % respectively. By using 2-cyclohexenone as substrate, optimum hydrogenation conditions were determined to be ca. 30 bar and 100 oC, because reactions at higher temperatures (110 oC) and pressures (40 bar) were inhibited by rhodium fall-out. In order to evaluate the effect of different levels of substitution around the olefinic double bond on the reaction rate for cyclic- as well as acyclic olefins, 2-cyclohexenone, 3-methyl-2- cyclohexenone, 3-buten-2-one, 3-penten-2-one, 4-methyl-3-penten-2-one, 3-methyl-3-penten- 2-one, crotonaldehyde, and chalcone were subjected to the hydrogenation reaction. While no hydrogenation could be achieved for the trisubstituted acyclic ketones, i.e. 4- methyl-3-penten-2-one and 3-methyl-3-penten-2-one, as well as crotonaldehyde, the trisubstituted cyclic equivalent, 3-methyl-2-cyclohexenone, indeed gave the saturated ketone albeit with a very low reaction rate [kobs = 0.000174644 s-1 (80 oC and 20 bar)]. The hydrogenation rates of the remaining acyclic substrates, 3-buten-2-one and 3-penten-2-one (kobs = 0.0245703 and 0.000226852 s-1 respectively at 80 oC and 10 bar), followed the expected order of monosubstituted > disubstituted, while the effect of a cyclic structure proved to be rather insignificant (kobs = 0.000254025 s-1 for 2-cyclohexenone). Aromatic disubstitution, however, reduced the reaction rate by ca. 50 % (kobs = 0.000143572 s-1 for chalcone). While the reaction rate (kobs) for the volatile substrates could easily be determined by GC and GC/MS analysis, the reactor set-up and analytical methodology were changed for the flavonoid like solid substrates. For these solids, reactions were followed by NMR and t½ (time to achieve 50 % conversion) used as indication of the reaction rate. In order to be able to compare t½ with the previously measured reaction rate, the hydrogenation of chalcone was repeated and it was found that the concentration of the reactant has a major influence on the rate of the reaction. In a study where the concentration of chalcone was varied between 0.083 M and 0.50 M the optimum concentration was determined as 0.166 M (t½ = 27 min.) for reactions at 80 oC, 20 bar hydrogen pressure and a catalyst concentration of 0.72 mM. Extension of the reaction to chromone indicated the heterocyclic ring to have a profound influence on the reaction rate (t½ = 72 h vs. 27 min. for chalcone), while the flavonoid substrates, flavone and 4',7-dimethoxyisoflavone, having trisubstituted double bonds could not be hydrogenated at all using Wilkinson’s catalyst. In a completely different approach, the introduction of chirality into a planar flavonoid molecule (flavone or isoflavone type compound) by means of an arene metal complex was investigated. If a bulky tricarbonylmetal centre could be directed to one face of the A-ring of a flavonoid unit, that face of the adjacent unsaturated heterocyclic C-ring would become inaccessible to a hydrogenating reagent. In order to investigate the feasibility of such an approach, the tricarbonylchromium(0) complexes of several mononuclear and flavonoid type substrates were to be synthesised. Compounds like benzene, toluene, anisole, chlorobenzene, acetophenone, and chromanone were therefore subjected to thermolysis (72 h) with hexacarbonylchromium(0) in refluxing dibutyl ether-THF and it was found that while the activated substrates showed excellent reactions (85 - 98 % conversion), conversions for the compounds containing deactivating substituents [acetophenone (40 %) and chromanone (28 %)] were rather low. With the knowledge of the mononuclear substrates in hand, the study was extended to the flavonoids where selectivity between the two aromatic rings would be a major issue in the success of the envisaged methodology. Although successful from a reaction point of view (products obtained in 33 and 43 % yield respectively), reaction of the carbonyl containing substrates 4',7-dimethoxyisoflavone and flavone, with hexacarbonylchromium(0), however, yielded only the ‘unwanted products’, tricarbonyl(B-h6-4',7-dimethoxyisoflavone)- chromium(0) and tricarbonyl(B-h6-flavone)chromium(0). In an effort to move complex formation to the A-ring of the flavonoid moiety, substrates with increasing levels of reduced heterocyclic rings like chroman-4-ol, flavan-4-ol, flavan, and 7-methoxyflavan, were subjected to the reaction with hexacarbonylchromium(0). Although the first two compounds still contained a 4-subsitutent with a negative inductive effect, it is known that a benzylic OH group is capable of directing complexation towards the adjacent aromatic ring and it was hoped that this influence would facilitate reaction onto the A-ring. Formation of cis- and trans-tricarbonyl(h6-chroman-4-ol)chromium(0) in an ca. 3:1 ratio (14 % yield) from the reaction of chroman-4-ol with the chromium reagent, confirmed the benzylic–OH to be capable of directing the attacking chromium moiety to the anticipated face of the adjacent aromatic ring. Reaction of the flavan-4-ol substrate, however, indicated the A-ring still not to be the preferred binding site, since tricarbonyl(B-h6-flavan-4-ol)chromium(0) and tricarbonyl(A-h6-flavan-4-ol)chromium(0) were formed in 2.8 and 1.7 % yields, respectively. Finally, reaction of both flavan and 7-methoxyflavan, the latter with an activated dihydroxylated A-ring, yielded products originating from complexation onto both the A- and B-rings, i.e. tricarbonyl(A-h6-flavan)chromium(0), tricarbonyl(B-h6-flavan)chromium(0), tricarbonyl(A-h6-7-methoxyflavan)chromium(0), and tricarbonyl(B-h6-7-methoxyflavan)- chromium(0) as well as the bimetallic complex from the 7-methoxyflavan.
Afrikaans: Die in vitro bestudering van die biologiese aktiwiteit van flavonoïede word ernstig in die wiele gery deur die ontoeganklikheid tot monomeriese verbindings wat in enantiomeries suiwer vorm beskikbaar is. Hoewel verskeie metodes waarvolgens monomeriese flavonoïede in opties suiwer vorm gesintetiseer kan word, wel bekend is, is baie van hierdie metodes omslagtig en is die opbrengste en ee’s nie altyd aanvaarbaar hoog nie. Stereoselektiewe 1,4- hidrogenering van die meer algemeen beskikbare prochirale a,b-onversadigde karboniel bevattende flavonoïede soos flavone, isoflavone en flavonole bied ‘n voor die hand liggende oplossing vir hierdie probleem en is dus tydens die huidige studie ondersoek. Weens die feit dat regio- en stereoselektiwiteit tydens die 1,4-reduksie van a,b-onversadigde karbonyl verbindings hand aan hand gaan, is besluit dat laasgenoemde aandag sal kry sodra bevestig is dat die regioselektiewe aspek van die reaksie suksesvol bemeester is. Aangesien Wilkinson se katalisator kommersiëel beskikbaar is en maklik m.b.v. modifisering met chirale ligande in ʼn stereoselektiewe katalisator omgeskakel kan word, is op hierdie katalisator vir die aanvanklike evaluering van die idee besluit. Hoewel Wilkinson se katalisator bekend is vir die hidrogenering van eenvoudige alkene, kon geen benutting daarvan vir die reduksie van a,b-onversadigde karboniel verbindings in die literatuur gevind word nie, en is besluit om die effek van verskillende grade van substitusie om die dubbelbinding te ondersoek voordat die werklike flavonoïed substrate aangepak sou word. Ten einde die beste oplosmiddel vir die reaksie te bepaal is die hidrogenering van chromoon in asetoon, tetrahidrofuraan (THF) en dichlorometaan (DCM) uitgevoer en is die chromanoon produk in 7.8, 14.3 en 46.2 % opbrengs onderskeidelik verkry. Vervolgens is optimisering van die kondisies van druk en temperatuur m.b.v. die hidrogenering van 2-sikloheksenoon uitgevoer en is gevind dat die beste omskakeling by 30 bar en 100 oC bereik word, aangesien hoër temperature (110 oC) en drukke (40 bar) tot die vorming van rhodiumswart aanleiding gegee het. Ten einde die invloed van verskillende vlakke van substitusie om die dubbelbinding op die reaksie te bepaal, is die reaksiesnelheid van sikliese- sowel as asikliese substrate, soos 2-sikloheksenoon, 3-metiel-2-sikloheksenoon, 3-buten-2-oon, 3-penten-2-oon, 4-metiel-3-penten-2-oon, 3-metiel-3-penten-2-oon en chalkoon, bepaal. Hoewel tri-gesubstutieerde liniêre substrate soos 4-metiel-3-penten-2-oon, 3-metiel-3-penten- 2-oon, asook krotonaldehied geen reaksie getoon het nie, het die ooreenstemmende sikliese ekwivalent, 3-metiel-2-sikloheksenoon, wel die versadigde ketoon gelewer. Die reaksietempo [kobs = 000174644 s-1 (80 oC en 20 bar)] in hierdie geval was egter aansienlik laer as by die ongesubstitueerde 2-sikloheksenoon. Hidrogenering van die oorblywende asikliese substrate, 3-buten-2-oon en 3-penten-2-oon (kobs = 0.0245703 en 0.000226852 s-1 onderskeidelik by 80 oC en 10 bar), het die verwagte tempo van mono-gesubstitueerd > digesubstitueerd gevolg, terwyl die invloed van ‘n siklise struktuur weglaatbaar geblyk te gewees het (kobs = 0.000254025 s-1 vir 2-siklohexenoon). Die waargenome reaksietempo het in die geval van ‘n aromatiese digesubstitueerde verbinding (chalkoon) egter met ca. 50 % gedaal (kobs = 0.000143572 s-1). Waar GC en GC/MS metings met vrug gebruik kon word om die reaksietempo (kobs) vir die vlugtige substrate te bepaal, moes die reaksieopstelling sowel as die analitiese metodologie aangepas word vir die vastestof flavonoïed substrate. In laasgenoemde geval is KMR benut om t½ (tyd om 50 % omskakeling te bereik), wat as aanduiding van reaksietempo gebruik is, te bepaal. Ten einde die t½ waardes met die vorige reaksietempos te kan vergelyk, is die hidrogenering van chalkoon herhaal en is gevind dat die substraat konsentrasie ‘n beduidende invloed op die snelheid van die reaksie uit oefen. Tydens ‘n studie waar die konsentrasie van chalkoon tussen 0.083 M en 0.50 M gewissel is, is vasgestel dat 0.166 M (t½ = 27 min) die optimum konsentrasie verteenwoordig indien die reaksie by 80 oC, 20 bar waterstof druk en 0.72 mM katalis konsentrasie uitgevoer word. Uitbreiding van hierdie ondersoek na chromoon het aangedui dat die heterosiklise ring ‘n nadelige invloed op die reaksietempo het (t½ = 72 h vs. 27 min vir chalkoon), terwyl die flavonoïed substrate, flavoon en 4’,7- dimetoksie-isoflavoon, wat tri-gesubstitueerde dubbelbindings vertoon, geen reaksie met Wilkinson se katalisator getoon het nie. Aangesien dit geblyk het dat flavonoïed substrate nie m.b.v. die genoemde hidrogeneringskatalisatore in opties aktiewe vorm berei sou kon word nie, is die moontlikheid van die benutting van areen-metaal kompleks vir chirale induksie vervolgens ondersoek. Indien kompleksering van ‘n lywige trikarbonielmetaal eenheid aan een aansig van die A-ring van ‘n flavonoïedeenheid bewerkstellig kan word, sou hierdie vlak van die aangrensende onversadigde heterosikliese C-ring ontoegangklik wees vir ‘n hidrogeneringsreagens. Ten einde die uitvoerbaarheid van hierdie benadering te bepaal is ‘n reeks trikarbonielchroom(0) komplekse van monosikliese en flavonoïed substrate gesintetiseer. Verbindings soos benseen, tolueen, anisool, chlorobenseen, asetofenoon en chromanoon is met heksakarbonielchroom(0) in dibutieleter-THF onder terugvloei verhit (72 h) en dit is gevind dat die geaktiveerde substrate uitstekende opbrengste lewer (85 – 98 % omskakeling), terwyl met substrate met deaktiverende substituente (asetofenoon en chromanoon) slegs 40 en 28 % omskakeling respektiewelik bereik kon word. Die ondersoek is voorts uitgebrei na flavonoïedsubstrate waar selektiwiteit tussen die twee aromatiese ringe van kardinale belang vir die sukses van hierdie benadering sou wees. Alhoewel kompleksering tussen heksakarbonielchroom(0) en 4’,7-dimetoksie-isoflavoon en flavoon onderskeidelik wel waargeneem is, is slegs die ‘ongewensde B-ring produkte’, trikarboniel(B-h6-4’,7-dimetoksie-isoflavoon)chroom(0) en trikarboniel(B-h6-flavoon)- chroom(0) in 33 en 43 % opbrengs onderskeidelik, verkry. In ‘n poging om metaal kompleksering na die A-ring te verskuif, is substrate met toenemende vlakke van versadigdheid in die C-ring, soos chroman-4-ol, flavan-4-ol, flavaan en 7-metoksieflavaan, aan die reaksie met heksakarbonielchroom(0) blootgestel. Hoewel die eerste twee substrate steeds oor ‘n 4-substituent wat ‘n negatiewe induktiewe effek uitoeffen beskik het, is dit ook bekend dat ‘n bensiliese OH groep kompleksering na die aangrensende aromatiese ring kan rig en is gehoop dat hierdie eienskap daartoe sal bydra dat reaksie by voorkeur met die Aring sal plaasvind. Isolasie van die cis- en trans-trikarboniel(h6-chroman-4-ol)chroom(0) produkte in ‘n ca. 3:1 verhouding (14 % opbrengs) uit die reaksie tussen chroman-4-ol en die chroom reagens, het die rigtende effek van die bensiliese OH bevestig. Tydens reaksie van flavan-4-ol met die chroom reagens is egter vasgestel dat die A-ring in hierdie geval steeds nie die verkose ring vir kompleksering is nie en is trikarboniel(B-h6-flavan-4-ol)chroom(0) en trikarboniel(A-h6-flavan-4-ol)chroom(0) in onderskeidelik 2.8 en 1.7 % opbrengs verkry. Ten einde enige moontlike deaktiverende effekte op die A-ring te verwyder en dit selfs met een en twee aktiverende groepe te vervang, is flavaan en 7-metoksieflavaan laastens aan die reaksie onderwerp. Beide A- en B-ring gekomplekseerde produkte, nl. trikarboniel(A-h6- flavaan)chroom(0), trikarboniel(B-h6-flavaan)chroom(0), trikarboniel(A-h6-7- metoksieflavaan) chroom(0) en trikarboniel(B-h6-7-metoksieflavaan)chroom(0) asook die bimetaliese kompleks van 7-metoksieflavaan, is egter vanaf beide substrate verkry.
Description
Keywords
Dissertation (M.Sc. (Chemistry))--University of the Free State, 2008, Flavonoids -- Synthesis, Asymmetric synthesis, Plant pigments, Catalytic hydrogenation, Synthesis, Ab-unsaturated ketone, Flavonoids, Wilkinson’s catalyst, Arene-tricarbonylchromium(0), Kinetics, Anaerobic complexation
Citation