Doctoral Degrees (Chemistry)

Permanent URI for this collection

http://hdl.handle.net/11660/406

Browse

Browsing Doctoral Degrees (Chemistry) by Author "Bezuidenhoudt, B. C. B."

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Evaluation of ligand modified palladium catalysts in the wacker oxidation of alkenes
    (University of the Free State, 2012-08) Saku, Duduetsang; Bezuidenhoudt, B. C. B.; Marais, C. M.
    The industrial application of Wacker oxidation of terminal olefins in aqueous aerobic mixtures with PdCl2 and CuCl/CuCl2 has largely been limited to shorter chain alkenes, that is, ethylene. As the alkene chain length increases, so do the challenges that render the reaction inapplicable for large scale production. Longer chain alkenes tend to isomerize due to the limited solubility in organic-aqueous mixtures. More so, the use of co-oxidants such as CuCl or CuCl2 in stoichiometric amounts results in the formation of toxic chlorinated by-products which make the system corrosive. Pd0 aggregation from the PdII active state, is also pertinent in these reactions hence the use of large amounts of a co-oxidant. Small TONs and TOFs have subsequently been reported. As one of the approaches to curb these challenges, ligand support and modification has recently been viewed with interest because it promises efficient stabilization of Pd0, wherein the efficiency of O2 to re-oxidize the Pd0 species is relied upon thereby avoiding Pd0 aggregation. Ligand support can also be used to alter the electronic environment of the PdII centre thereby affecting its activity and selectivity. The application of phosphorus-palladium complexes in this study is not only a new approach in Wacker oxidation but the utilization of the π-accepting and or σ-donating abilities of phosphorus compounds was also advantageous in altering the PdII electronic environment. No co-oxidants were used in this study w.r.t. the oxidation of 1-octene and the complexes evaluated were comparable to those reported in literature with PdCl2/DMA systems under similar conditions. Since oxygen is the preferred oxidant in all oxidation reactions because of its natural abundance, its reported enhanced selectivity and ease of separation from products, it was decided to evaluate the utilization of this reagent as first choice in the current investigation of ligand supported palladium catalysts in the Wacker oxidation. Due to the fact that the phosphite based palladium catalyst, PdCl2[P(OPh)3]2, is readily soluble in DMA, it was determined that no pre-stirring as for PdCl2 was required for this catalyst. In order to obtain the optimum reaction conditions for oxygen as oxidant with this catalyst, conditions like solvent, reaction temperature, O2 pressure and water, catalyst, and substrate concentration were varied. The optimized conditions were determined to be 0.5 mol% of catalyst in DMA:H2O (6:1) under 9 atm of O2 at 80°C, while the optimum substrate concentration was found to be 0.2M. PdCl2[P(OPh)3]2 showed the highest activity of the catalysts evaluated and gave a TOF of >1370 (mol/mol/hr), which compared favourably with other known catalysts like PdCl2, PdCl2(CH3CN)2 Pd(OAc)2, and Pd(CF3SO3)2 where TOF’s of 1429, 1420, 817 and 524 respectively, were obtained under the conditions optimized for PdCl2[P(OPh)3]2. While the palladium metallocycle [Pd(u-Cl)(C6H4O)(OC6H6)2]2 gave TOF’s (1380 mol/mol/hr) virtually the same as PdCl2[P(OPh)3]2, total conversion for the latter catalyst was only 93%, so it can be regarded as the second best of all the catalysts evaluated. The monomers thereof, PdCl[(C6H4O)(C6H6O)2P(OPh3)] and PdCl[(C6H4O)(C6H6O)2(PPh3)], revealed the least basic P(OPh3) to be more reactive (TOF >900 mol/mol/hr) than the TPP containing analogue, where the latter showed no activity within the first hour of reaction. While all the active catalysts showed good selectivities of >80%, the metallocycle [Pd(u-Cl)(C6H4O)(OC6H6)2]2 proved to be the best with a selectivity of 89%. Catalyst recyclability was also observed to at least 3 cycles, with selectivities maintained above 80%. No Pd0 ‘fall-out’ or aggregation was observed with any of the catalysts evaluated. For the palladium phosphinite catalysts 1,2-Ph(OPPh2)2PdCl2 and 1,3-Ph(OPPh2)2PdCl it was found that both were active in the Wacker oxidation of 1-octene albeit with very low rates for the latter complex (1,3-Ph(OPPh2)2PdCl). The low reactivity of 1,3-Ph(OPPh2)2PdCl was similar to that of the phosphines (PPh3)2PdCl2 and (3,5-CF3-PPh2Cl)2PdCl2 where (PPh3)2PdCl2 showed some conversion only after 3 hours and (3,5-CF3-PPh2Cl)2PdCl2 gave only 53% conversion after an hour. Through a comparison of the reactivity of 1,2-Ph(OPPh2)2PdCl2 with that of the hydrolyzed equivalent [μ-ClPd(PPh2OH)(PPh2O)]2, it seemed as if the phosphinite catalysts are prone to hydrolysis under the prevailing conditions as the final conversion of both these catalysts were almost the same (85 and 79% respectively). Hydrogen peroxide and tert-butylhydroperoxide (TBHP) were also evaluated as alternative oxidants with PdCl2[P(OPh)3]2 as catalyst and H2O2 was found to be the better of the two oxidants with conversion (99%), selectivity (86%), and TOF (1220) almost as good as those found for oxygen (100, 82% and 1370 respectively). In addition, the catalyst could also be recycled three times although degradation of the H2O2 was observed and additional peroxide (12 eq.) had to be added with each cycle of substrate. TBHP, however, suffered from moderate selectivities of only 60-65%, while the catalysts was deactivated during the first oxidation cycle and could therefore not be recycled at all. Although all phosphite catalysts promoted isomerization to internal 1-octene isomers to some extent, the cyclopalladated [Pd(u-Cl)(C6H4O)(OC6H6)2]2 catalysts proved to be the best in this aspect of the reaction w.r.t. oxygen as oxidant and led to very low quantities of isomerised products being observed (3 - 4%). It was also evident that the type and amount (for H2O2 and TBHP) of oxidant played a crucial role in enhancing or suppressing isomerization and hydrogen peroxide (at only 2% isomerization) was found to be the best oxidant in this regard followed by oxygen (13%).
  • Thumbnail Image
    ItemOpen Access
    New ring closing metathesis based methodology for the synthesis of monomeric flavonoids
    (University of the Free State, 2017-02) Pieterse, Tanya; Bezuidenhoudt, B. C. B.; Marais, C.
    Although the physiological activity of flavonoids stimulated investigations into more efficient synthetic methods for the preparation of these compounds, many of these routes entail multiple steps and require the utilization of stoichiometric and often poisonous reagents. Known methodologies are also hampered by difficulties around the isolation of the desired product and often lead to inseparable mixtures, low yields, and tedious synthetic processes. To circumvent these problems and to bring the synthesis of flavonoids in line with modern synthetic methodologies, it was decided to embark on a process of preparing the different classes of flavonoids through the application of a catalytic process, like ring closing metathesis (RCM), as key step in the methodology. Developing this methodology would have the added advantage that all the different classes of flavonoids would be reachable from readily available starting materials and the application of basically a single catalytic reaction in the final process step. For entry into the first class of flavonoids, i.e. compounds with a 2-phenylchromane skeleton, the preparation of flav-2-enes were investigated as key intermediate. In this regard, allyl phenyl ethers, prepared via Williamson etherification [K2CO3 (2.0 e.q), CH3CN, reflux], were subjected to Claisen rearrangement in a neat microwave assisted process to obtain the substituted allyl benzenes, 1-allyl-2-hydroxy-4-methoxybenzene and 1-allyl-2-hydroxy-4,6- dimethoxybenzene, in 44% and 88% yields, respectively. Subsequent esterification of the allyl phenols with substituted benzoyl chlorides [aq. NaOH (2.0 M, 40.0 mL) or 4- dimethylaminopyridine (0.2 eq.), dry pyridine (1.0 eq.), dichloromethane, reflux] afforded a series of the benzoates, i.e. 2-allylphenyl benzoate, 2-allylphenyl 4-methoxybenzoate, 2- allylphenyl 3,4-dimethoxybenzoate, 2-allylphenyl 3,4,5-trimethoxybenzoate, 2-allyl-5- methoxyphenyl 3,4-dimethoxybenzoate, 2-allyl-5-methoxyphenyl 3,4,5-trimethoxybenzoate, 2-allyl-3,5-dimethoxyphenyl 3,4-dimethoxybenzoate and 2-allyl-3,5-dimethoxyphenyl 3,4,5- trimethoxybenzoate in 68 – 98% yield. During methylenation of these esters through utilisation of the Tebbe reagent, it was found that the reaction is largely dependent on the concentration of the substrate, as well as reaction time and temperature. High yields (71 – 94%) were obtained with an increase in concentration of the ester and a brief period at elevated temperature (80 – 90°C). While a series of substituted diaryl vinyl ethers could be prepared, methylenation of substrates containing a phloroglucinol-type substitution pattern on what was to become the A-ring of the flavonoid failed. Ring closing metathesis of all the vinyl ethers in hand under standard metathesis conditions [Grubbs II, dichloromethane, reflux] led to the formation of flav-2-ene, 4'-methoxyflav-2-ene, 3',4'-dimethoxyflav-2-ene, 3',4',5'-trimethoxyflav-2-ene, 3',4',7-trimethoxyflav-2-ene and 3',4',5',7-tetramethoxyflav-2- ene in 41 – 96% yields. Attempts at the epoxidation of flav-2-ene with m-CPBA with and without a base (NaHCO3), did not yield any of the desired product. Construction of the isoflavonoid nucleus was first attempted through preparation of the isoflav-2-ene analogue via the deoxybenzoin intermediate, which could be prepared by phenylmagnesium bromide addition to the corresponding phenyl acetate. Although the phenyl acetates (methyl 4-methoxyphenyl acetate, methyl 4-trifluoromethylphenyl acetate, methyl 3-methoxy-4-trifluoromethanesulfonyloxyphenyl acetate and methyl 3,5- dimethoxyphenyl-4-trifluoromethanesulfonyloxyacetate) could be prepared in excellent yields (80 – 99%) via ozonolysis of the substituted allyl benzenes, the transformation of these compounds into the required deoxybenzoins was hampered by the inability (even at temperatures as low as -78 °C) to stop the reaction of the Grignard reagent with the substrate at the ketone stage. The methodology for the preparation of isoflavenes was therefore adapted to the synthesis of the isoflav-3-ene analogues, which could be constructed through a one-pot reaction of the substituted benzaldehyde with substituted α-bromoacetophenones followed by Wittig reaction with methyltriphenylphosphonium bromide to afford vinyl benzene intermediates, 4-methoxy-2-[(2-phenylallyl)oxy]-1-vinylbenzene, 1,5-dimethoxy-3-[(2- phenylallyl)oxy]-2-vinylbenzene, 1-{[2-(4-methoxyphenyl)allyl]oxy}-2-vinylbenzene, 4- methoxy-2-{[2-(4-methoxyphenyl)allyl]oxy}-1-vinylbenzene and 1,5-dimethoxy-3-{[2-(4- methoxyphenyl)allyl]oxy}-2-vinylbenzene, in 61 – 89% yield. Subsequent ring closing metathesis of the 7- and/or 4' substituted vinyl benzenes proceeded smoothly over Grubbs II catalyst in refluxing DCM and gave the isoflav-3-enes, (7-methoxyisoflav-3-ene, 4'- methoxyisoflav-3-ene and 4',7-dimethoxyisoflav-3-ene) in 57% to quantitative yields. RCM of the vinyl benzenes with a phloroglucinol-type substitution pattern, however, required elevated temperatures (refluxing toluene) and/or the addition of 1,4-benzoquinone in order to form the isoflav-3-enes, 5,7-dimethoxyisoflav-3-ene and 4',5,7-trimethoxyisoflav-3-ene, in decent yields (67 and 65%, respectively). Subsequent epoxidation of 7-methoxyisoflav-3-ene with m-CPBA and NaHCO3 in dichloromethane again failed to give any of the desired isoflavene epoxide. Although the neoflavonoid nucleus could be reached through Claisen rearrangement of 1- cinnamyloxybenzenes followed by vinylation of the phenolic hydroxy entity or Wittig mediated methylenation of 2-allyloxybenzophenones, followed by ring closing metathesis, this methodology was not viewed as being appropriate for application to oxygenated substrates as a number of process steps would be required to obtain oxygenated starting materials. It was therefore decided to follow a process where the appropriate acetophenones would be converted into the substituted styrenes by a Grignard reaction-dehydration process. Since electron-rich acetophenones are notorious for being lousy substrates in Grignard reactions the addition of aluminium triflate to the reaction mixture to enhance the reactivity of the reactant was investigated and it was found that the addition of Al(OTf)3 to the reaction mixture had a significant effect on the reaction of 4-methoxyphenylmagnesium bromide and 2-allyloxy-4-methoxyacetophenone. Not only did the presence of the Lewis acid increase the reaction rate, but it also led to the direct formation of the substituted styrene in 66% yield. Extending this reaction to the addition of phenylmagnesium bromide to 2-allyloxy-4,6- dimethoxyacetophenone and the addition of 4-methoxyphenylmagnesium bromide to 2- allyloxy-4,6-dimethoxyacetophenone and 2-allyloxy-4,5-dimethoxyacetophenone gave the substituted styrene products in moderate to high yields (52 – 94%). When 3,4- dimethoxyphenylmagnesium bromide was utilised in reactions with 2-allyloxy-4- methoxyacetophenone and 2-allyloxy-4,5-dimethoxyacetophenone, however, the analogous alcohols were obtained in 50% and 4% yields, respectively. When employing standard Grignard conditions (THF, -60 °C) i.e. without Al(OTf)3 activation, the tertiary alcohol products could be obtained 60% and 80% yields, respectively. Subsequent CuSO4-mediated dehydration of the alcohols yielded the desired styrenes (75% and 65%, respectively). Ring closing metathesis of all the styrene intermediates in hand proceeded smoothly and yielded the series of neoflav-3-enes, (4',7-dimethoxyneoflav-3-ene, 5,7-dimethoxyneoflav-3-ene, 4',5,7-trimethoxyneoflav-3-ene, 4',6,7-trimethoxyneoflav-3-ene, 3',4',7-trimethoxyneoflav-3- ene and 3',4',6,7-tetramethoxyneoflav-3-ene) in excellent yields (67% – quant.). Since it was shown that aluminium triflate had an enhancing effect on the addition of Grignard reagents to the acetophenones required for the synthesis of neoflavenes and that the styrenes could be obtained in a one-step process, it was decided to explore the scope of this novel process towards other ketones. During this investigation it was determined that the addition of Grignard reagents like phenylmagnesium bromide, benzylmagnesium bromide and ethylmagnesium bromide, to electron-rich ketones, i.e. 4-methoxyacetophenone, 2,4- dimethoxyacetophenone, 2,4-dimethoxypropiophenone and 4-chromanone, led to the formation of respective alkenes in 46 – 97% yields, while no product formation was observed for the less activated substrates like 4-chloroacetophenone and α-tetralone. It was furthermore observed that for the reaction of 4-methoxyacetophenone with ethyl - and benzylmagnesium bromide only the E-isomers of the product was formed, while only the Z-isomer was obtained during the addition of ethylmagnesium bromide to 2,4-dimethoxyacetophenone. The reaction of 2,4-dimethoxypropiophenone with phenylmagnesium bromide, on the other hand, yielded both geometric isomers in a 1:1 ratio. The stereoselectivity found during the reactions of 4- methoxy- and 2,4-dimethoxyacetophenone with ethyl - and benzylmagnesium bromide is probably explicable in terms of an E2 elimination process involving the preferred sterically less hindered gauche conformation of the transition state. Extending the reaction to the addition of phenylmagnesium bromide to α,β -unsaturated systems, like chalcone, indicated Al(OTf)3 to also have an activating effect in this regard, albeit to a marginal extent, since only an 8% increase in the yield of the 1,4-addition product was observed. Finally, indications were also found that Al(OTf)3 may also be utilized in catalytic quantities for this reaction when p-methoxy-1-phenylstyrene could be prepared in 82% yield by utilising Al(OTf)3 in 10 mol%; thus rendering the new methodology the first Grignard based Lewis acid catalysed process for the direct synthesis of alkenes.