published_or_final_version / abstract / Chemistry / Doctoral / Doctor of Philosophy

Synthesis and flash vacuum pyrolysis (FVP) of stabilised phosphorus ylides containing an o-amino functionalised benzene ring has been examined for the first time. Model studies using N-methyl-N-tosyl and N-mesyl-N-methyl ylides showed that the ylides could be prepared, although yields were variable, and had the expected spectroscopic properties. Upon FVP, however, the expected loss of Ph₃PO and the sulfonyl group was accompanied by unexpected transfer of the reactive site from nitrogen to carbon giving 3- substituted quinolines rather than the expected indole products. Moving to ylides with an α-cinnamoyl group (or heterocyclic analogue) did, however, result in the originally planned tandem cyclisation leading to ring-fused carbazole products. N-Benzyl was also found to be a suitable thermally labile group and a series of α-cinnamoyl N-benzyl-N-methyl ylides were prepared and characterised. For their synthesis, use of N-cinnamoylbenzotriazoles was found to be preferable to cinnamoyl chloride, requiring only half the amount of amino-functionalised phosphonium salt. While FVP of some of these ylides led to benzo-, furo- and thienocarbazoles in good yield, others again gave quinoline-type products pointing to a fine balance between the two alternative modes of cyclisation. It was noted that one of the furocarbazole products was very similar to a natural product, Eustifoline D, isolated from the medicinally active shrub Murraya euchrestifolia from Taiwan and its synthesis was planned. With a view to producing the required N-H carbazole, N,N-dibenzylamino amino ylides were prepared and were found to exhibit restricted rotation leading to broad NMR signals. Their FVP again led to both quinoline and carbazole products, with the former having usually, but not always, lost a phenyl group. Mechanistic pathways for the formation of the various products are proposed. Complete assignment of the complex ¹H NMR spectra of the various fused-ring heterocyclic products was achieved, assisted by simulations in many cases. The ylide precursor required for Eustifoline D was prepared in five steps and 10% overall yield from 5-methylanthranilic acid. The final FVP step gave a quinoline as the major product, but the minor product was Eustifoline D, spectroscopically identical to the natural product.

The research presented in this thesis is a project funded by the EU-network of the Marie Curie project NANO-HOST in collaboration with partner institutes. The aims of this network are to develop innovative methods for the preparation, recovery and reuse of single-site, nanostructured catalytic materials, and further on apply them in combination with specifically engineered reactors for a sustainable production process for making high value fine chemicals. One part of this project was to prepare chiral diphosphine ligands and their complexes for currently challenging reactions, such as asymmetric carbonylations (homogeneous catalysis). Catalytic studies of these chiral diphosphine ligands were carried out in asymmetric hydroxy-and alkoxy-carbonylations and hydrogenation reactions. The second part of this project was the heterogenisation of these chiral homogeneous complexes through collaborations with the network partners and furthermore their catalytic behavior was studied.

The synthesis of ligands has been identified as the limitation for wider application of catalytic asymmetric synthesis. Indeed, synthesis of phosphorus-based ligands, has been often shown to be challenging and not always efficient. It has also been observed that subtle changes in the ligand structure can lead to big differences in the catalytic activity of the ligand when coordinated to a metal. Therefore, it was considered useful to develop a methodology in order to obtain a library of phosphines. The first chapter of the thesis is a review of recent development in catalytic phosphine synthesis. In the second chapter of this thesis, the microwave mediated Suzuki cross coupling reaction has been investigated. In particular, attention has been focussed on the coupling of different arylboronic acids to chloroarylphosphine oxides, which are, in general, considered challenging coupling partners for this type of reaction. The reaction conditions have been optimised starting from the coupling of phenylboronic acid to tris(4-chlorophenyl)phosphine oxide. Different solvents, bases, and catalysts have been then tested and the better conditions have been developed for this substrate. Indeed, it was shown that the coupling occurs in only 30 minutes at 140°C, leading to reasonably high yields. These conditions were then applied to two other different chloroarylphosphine oxides with a range of boronic acids, with the aim to verify if the optimised reaction conditions could be applied to other substrates and it was noticed that good yields could be attained. This methodology led us to obtain a library of phosphine oxides. It was then decided to investigate the reduction of phosphine oxides under microwave irradiation. This reaction occurs under conventional heating but it can take several hours. It was observed that reaction times could be importantly reduced when reducing some phosphine oxides under microwave heating. It was found that some phosphine oxides are reduced rapidly under conventional conditions but for more difficult substrates to reduce there are significant advantage to microwave method. We decided to investigate the microwave mediated P-C bond forming reaction, with the aim to rapidly synthesise a library of phosphines cleanly. The conditions were optimised at first using o- trifluoromethylbromobenzene as the substrate. Once the appropriate reaction conditions and catalyst were identified, the reaction was run on other substrates to verify that this could be a general methodology for the synthesis of phosphines. It was found that it is indeed a general method for the synthesis of monophosphines. However, the synthesis of diphosphines with the microwave assisted P-C bond forming reaction on dibromo- and diiodo- aryl compounds proved to be very challenging. The fourth chapter presents different attempts for the synthesis of the new ligand Ph-DuPHOS. The synthesis of this ligand was considered interesting because of the previous results of other phospholane-based ligands, such as Ph-BPE and Me-DuPHOS. However, the synthesis of this ligand has proven to be elusive. A monodentate P-N phospholane-based ligand was prepared and its catalytic activity was tested in the rhodium catalysed hydrogenation of alkenes. Moreover, a bidentate (1,2-bisphospholano)xylene ligand was also prepared and its catalytic activity was also tested in the rhodium catalysed hydrogenation of alkenes. This latter ligand was also used in the hydroxycarbonylation of styrene, since for this reaction bulky diphosphines are required to give branched selectivity. In hydroxycarbonylation it is very rare to give good branched selectivity and there were no examples of substantial enantioselectivity prior to this work. The high regioselectivity and moderate e.e.’s observed suggest promise for future studies. Finally, mechanistic studies on the hydroxycarbonylation of styrene have been carried out in order to investigate the intermediates involved in the reaction as well as the role of the promoters. The possibility of (1-chloroethyl)benzene was proposed as the active intermediate of the reaction. Our results have disproved this possibility, suggesting that the reaction is likely to proceed through the hydride mechanism.

Digitized by Kansas Correctional Industries

Since late 1990s, organocatalysis has been widely explored in many aspects and achieved various difficult transformations. In this field, carbonyl catalysis, which could be traced back to 1860 has been developed with impressive progress including asymmetric variants. Over the years, major activation modes were developed for carbonyl catalysis including exploiting temporary intramolecularity to form catalytic tethers and transient intramolecular nucleophiles, dioxirane formation and imine formation. On the other hand, electrophilic activation is also an important area of organocatalysis where impressive progress has been achieved. However, limited examples were reported to achieve the electrophilic activation via carbonyl catalysis. Organophosphorus compounds are crucially important in many aspects in organic chemistry. Many approaches were developed for asymmetric organophosphorus compounds. In this work, different types of organophosphorus compounds were used as the substances for aldehyde-catalyzed hydrolysis reactions. The first part of this thesis illustrated the strategy to combine carbonyl catalysis and electrophilic activation. The hydrolysis of organophosphorus compounds containing P(=O)-N bond were investigated based on Jencks and Gilchrist’s preliminary results with formaldehyde as the catalyst to promote the hydrolysis of one inorganic substance, phosphoramidate. This Chapter describes a systematic research to identify a superior catalyst, o-phthalaldehyde, and develop catalytic hydrolyses of various organophosphorus compounds containing P(=O)-NHR subunits. Gratifyingly, the reaction proved efficient with phosphinic amides and phosphoramidates. Moreover, chemoselectivity was also studied and selective hydrolysis of the P(=O)-N bonds in the presence of P(=O)-OR bonds could be accomplished. The second part of this thesis demonstrated the further development of one of the major modes of carbonyl catalysis. Formaldehyde was identified as the efficient catalyst to react with α-amino phosphonates to form the transient intramolecular nucleophile, which facilitated the subsequent hydrolysis reactions. In this Chapter, different primary and secondary α-amino phosphonates with phenol as the leaving group, were tested in the reaction conditions. As a result, a vast of mono esters of α-amino phosphoric acids could be formed as the products. Finally, the last portion of this thesis applied the methodologies developed in Chapter 2 to prebiotic chemistry. A prebiotic-related aldehyde, glycolaldehyde was studied as the catalyst for the hydrolysis of organophosphorus compounds containing P(=O)-N bond, including phosphinic amides and phosphoramidates. Additionally, other prebiotic important substances, diamidophosphate (DAP) and monoamidophosphate (MAP) were also investigated for potential glycolaldehyde-catalyzed phosphorylation reaction under aqueous conditions. In the presence of catalytic amount of glycolaldehyde, 1) when water was used as the nucleophile, the hydrolysis of DAP and MAP were significantly improved; 2) when other phosphate nucleophiles were added to compete with water, DAP could act as a phosphorylating reagent to phosphorylate other phosphate nucleophiles. Overall, the results presented in this thesis investigated two different activation modes, electrophilic activation and transient intramolecular nucleophiles, for carbonyl catalysis to hydrolyze different organophosphorus compounds, phosphinic amides, phosphoramidates and α-amino phosphonates. The application of carbonyl catalysis to prebiotic chemistry was also achieved especially with the phosphorylation reaction with DAP.

Sample preparation is a key step in chemical analysis, and includes isolation of target analytes, removal of interferences, preconcentration, and/or modification of target analytes (if needed). Sample preparation is also the most time-consuming and error-prone step in the whole analytical process. Traditional sample preparation techniques involve hazardous solvents. Considering the environmental and health safety, it is desirable to reduce or eliminate organic solvents in sample preparation. Solid phase microextraction (SPME) was introduced as a solvent free sample preparation technique. Capillary microextraction (CME) is one of the formats of SPME that can be easily coupled to high performance liquid chromatography (HPLC). In SPME and CME a solvent free sample preparation is accomplished by using a sorbent coating instead of hazardous organic solvents commonly used in conventional extraction techniques. This research is focused on the development and systematic examination of novel niobia-, titania- and silica-based organic-inorganic hybrid sol-gel sorbents for CME. Conventionally silica and titania based precursors were used in organi-inorganic hybrid sol-gel sorbents for CME, here novel niobia based precursor was used in creating organic-inorganic hybrid sol-gel sorbents. Poly tetrahydrofuran (polyTHF) as well as electrically neutral and charged organic ligands were used to prepare the sorbents for CME coupled to HPLC. Characterization of created sol-gel sorbents, evaluation of extraction performance, and enrichment of environmentally and biomedically important analytes including organophosphorus compounds were performed. CME performances of the created sorbents were characterized by specific extraction (SE) (a measure of extraction efficiency) and desorption efficiency (DE) (a measure of completeness desorption of extracted analytes). Scientific findings of this research has shown that sol-gel niobia-polyTHF sorbent provides 60 to 70 % higher SE values for different environmentally important analytes compared to analogously prepared silica-polyTHF sorbent. This superior extraction performance can be attributed to the presence of surface Lewis acid sites undergoing Lewis acid-base interactions with analytes representing Lewis bases. The prepared sorbents also have the ability to undergo van der Waals interactions due to the presence of polyTHF. Absence of Lewis acid sites on silica surface resulted in inferior extraction efficiency compared to niobia-polyTHF sorbents. Extraction efficiency of the created sol-gel based niobia-polyTHF was also explored in the enrichment of organophosphorus pesticides and compared with that of the state-of-the-art titania-based sorbent. Sol-gel niobia-polyTHF sorbent has provided 40 to 50 % higher SE values in the enrichment of organophosphorus pesticides compared to sol-gel titania-polyTHF sorbent which can be attributed to the presence of bronsted acid sites on niobia surface (but lacking on titania) along with Lewis acid sites. To explore relative contributions of electrostatic, Lewis acid-base and van der Waals interactions between sol-gel sorbents and analytes, two sol-gel sorbents, one containing a positively charged octadecyl ligand and the other a neutral octadecyl ligand were created. Positive charge was imparted by using N-octadecyldimethyl [3-(trimethoxysilyl) propyl] ammonium chloride (C18 (+ve)) as ligand bearing co-precursor. Similarly N-octadecyl trimethoxysilane was used to impart a neutral C18 ligand in sol-gel coating. Experimental results have shown that sol-gel Nb2O5-C18 (+ve) sorbent has superior extraction efficiency compared to sol-gel based Nb2O5-C18 and purely inorganic Nb2O5 sorbents in enrichment of organophosphorus compounds (nucleotides and organophosphorus pesticides). Electrostatic interactions between the positive charge of organic ligand (C18 (+ve)) and negative charge of phosphate group has contributed to the higher extraction performance of sol-gel based Nb2O5-C18 (+ve) sorbent. TiO2-C18 (+ve) sorbent was also created to compare with the novel sol-gel niobia based sorbents, since titania-based sorbents are considered as the state-of-the-art extraction material in the enrichment of organophosphorus compounds. Established research results has shown that sol-gel based Nb2O5-C18 (+ve) sorbent has provided 40 to 50 % higher specific extraction values for organophosphorus compounds compared to sol-gel based TiO2-C18 (+ve) sorbent. Desorption efficiency of sol-gel Nb2O5-C18 (+ve) and TiO2-C18 (+ve) sorbents were 96% vs 90%. This superior DE of sol-gel Nb2O5-C18 (+ve) sorbent can be attributed the higher Lewis acid strength of titania than nioiba. The developed sol-gel niobia based sorbents have also shown high pH stability compared to traditional sol-gel silica based sorbents. The created sol-gel sorbents were characterized by less than 5% run to run RSD values and also less than 5% capillary to capillary RSD values which indicated the high reproducibility of developed method. The developed sol-gel niobia sorbents are applicable to sample preparation in different fields including biomedical, environmental, forensic, defense etc.

The alkoxytriphenylphosphonium ion intermediate of the Mitsunobu reaction for the esterification and inversion of configuration of an alcohol can be generated using the Hendrickson reagent, triphenylphosphonium anhydride trifluoromethanesulfonate, 27. While 27 was used in place of the Mitsunobu reagents (triphenylphosphine and a dialkyl azodicarboxylate) for the esterification of primary alcohols, the reaction failed with secondary alcohols such as (-)-menthol giving predominately elimination rather than the desired SN2 displacement. The difference between the two reactions was shown to be related to the more 'ionic' conditions generated when the Hendrickson reagent 27 was employed. An extreme sensitivity of the Mitsunobu reaction to the presence of salts was discussed and may indicate a mechanism involving ion pair clustering. Five-, six- and seven-membered cyclic analogues of the Hendrickson reagent 90-92 were prepared. A kinetic comparison of the cyclic analogues 90-92 revealed that a considerable increase in the rate of esterification could be achieved when the five-membered ring analogue 90 was used in a non-polar solvent such as toluene. Selected acyclic analogues of the Hendrickson reagent 27 possessing tributyl 118, tricyclohexyl 130 and diphenyl-2-pyridyl 137 functionalities were synthesised. However when 118, 130 and 137 were used for the attempted esterification of (-)-menthol, elimination was the major reaction pathway. Diphenyl-2-pyridylphosphonium anhydride triflate 137 was found to be a useful reagent for the synthesis of acyclic dialkyl ethers from primary alcohols. A polymeric version of the five-membered ring analogue 56, prepared by reaction of the polymer-supported 1,2-bis(diphenylphosphinyl)ethane 57 with triflic anhydride, was used for the preparation of simple esters and amides. A new dehydrating agent, polymer-supported triphenylphosphine ditriflate 157, was readily prepared from the oxidised form of commercially available polymer-supported triphenylphosphine and triflic anhydride. A wide range of dehydration-type reactions, such as ester, amide, anhydride, peptide, ether and nitrile formation, were performed in high yield using polymer-supported triphenylphosphine ditriflate 157. The reagent 157 was easily recovered and re-used several times without loss of efficiency. The use of 4-dimethylaminopyridine allowed the esterification of secondary alcohols with 157 to proceed without elimination and gave esters in high yield but with retention of configuration. Both reagents 56 and 157 provide an alternative to the Mitsunobu reaction, where the use of azodicarboxylates and chromatography to remove the phosphine oxide by-product can be avoided. However, the Mitsunobu reaction retains its supremacy for the inversion of configuration of a secondary alcohol. Preliminary investigations on the phosphityation of alcohols via the Hendrickson reagent 27, 1,3-benzodioxole formation using the Mitsunobu reaction and azodicarboxylate alternatives in the Mitsunobu reaction are described.

Thesis (Ph. D.)--University of Hong Kong, 2007. / Title proper from title frame. Also available in printed format.

Thesis (M. Phil.)--Hong Kong University of Science and Technology, 2003. / Includes bibliographical references (leaves 135-142). Also available in electronic version. Access restricted to campus users.