Abstract

This project aims to look at the development of the Wadsworth-Emmons cyclopropanation reaction and compare it to alternative methods of cyclopropanation in order to understand why it may be used preferentially. Current applications of the WE cyclopropanation reaction are explored to see the efficacy and yield that result.

Key of Abbreviations

Seen below is a collection of all abbreviations used within this project and their subsequent meaning.

φ3: Tertiary

Bn: Benzyl group

CCR: Corey-Chaykovsky Reagent

d.e:

DME: Dimethoxyethane

e.e:

Et: Ethyl group

EWG: Electron Withdrawing Group

HWE: Horner-Wadsworth-Emmons

iPr: isopropyl

Me: Methyl

NOE:

Ph: Phenyl group

THF: Tetrahydrofuran

WE: Wadsworth-Emmons

1: Introduction

Cyclopropane [1] was used as an anaesthetic until it was discovered to be highly reactive and dangerous when combined with oxygen. The reactivity of cyclopropane is mainly due to the high amount of ring strain and the bond strength between the carbons being weaker than normal carbon bonds, allowing the ring to open easily. Cyclopropane structures are often found within compounds in nature (figure 1.) [2] that are observed to have medicinal and commercial applications, e.g. degradable insecticides such as the pyrethroid family that are less toxic to other animals in the environment [3] , [4] . Cyclopropanation can also produce a number of hallucinogens and opoid drugs that are most widely used recreationally for there effects on human mental and visual perception, exploiting their psychedelic properties. This type of use has historically been seen within Native American tribes and ancient civilisations. Recently more research shows that they may potentially be useful in therapeutic doses in the treatment of pain, depression, alcoholism and other behavioural problems. An example of this is Codorphone, an analgesic that can be both an agonist and antagonist at μ-opioid receptors in the body and shows a higher potency than codeine.

As a result, efficient synthetic routes that produce high yields have been developed in order to produce synthetic analogues of these natural compounds, with special attention being paid to the cyclopropanation step. Over the years there have been many methods of cyclopropanation, from using zinc carbenoids (Simmons-Smith reaction [5] ) to stabilised ylides (Corey-Chaykovsky Reaction [6] ), all producing varying ratios of isomers of the product. Often, one enantiomer is the more biologically active molecule, therefore stereoselective reactions are required to obtain high yields of the desired product. The Wadsworth-Emmons cyclopropanation reaction is an example of a reaction that is selective for the generation of the trans-isomer and has advantages over other stereoselective reactions.

1.1: Chemistry of the Cyclopropane Ring

Cyclopropane is the smallest of the cycloalkanes that can be formed and consists of three sp3 hybridised carbon atoms bonded to each other to form a triangular ring. Although it is the smallest cycloalkane it is also the most reactive due to the bond angle within the ring.

The ideal bond angle for sp3 hybridised carbons is 109.5° as it at this angle that orbitals can overlap correctly and form the highest strength carbon-carbon bonds possible. As the propane ring is planar and made up of only three carbon atoms, a bond angle of 109.5° is not possible and is reduced to a bond angle of ~60°. In order to achieve bond angles of 60° the sp3 orbitals need to form ‘bent’ bonds [8] where their p-characteristics are increased, causing the carbon-hydrogen bond length to shorten (figure.2). It is this significant difference between the ideal bond angle and the actual bond angle exhibited that causes a high amount of ring strain. Also, the bond angle and the modified overlap of orbitals results in carbon-carbon bonds which are weaker than normal.

In figure.3, as the ring size increases from cyclopropane to cyclohexane, the ring strain decreases because the ideal bond angle is reached (when in a planar conformation) and larger rings can assume a non-planar conformation. In comparison to cyclopropane, the most strained and planar ring, cyclohexane has ideal bond angles and can form both ‘chair’ and ‘boat’ folded conformations in order to be the least strained ring possible.

Further strain in the cyclopropane ring is due to the planar conformation of the molecule, where the two hydrogens present on each carbon atom are in an ‘eclipsed’ position (figure.4)2. The carbon-hydrogen bonds are locked into this high energy conformation, as the carbon-carbon bonds of the ring are unable to rotate to form a more staggered conformation and reduce torsion strain.

The total strain felt by the molecule leads to the ring structure being highly unstable and is ultimately responsible for the high reactivity of cyclopropane. Due to instability, the cyclopropane ring is able to break open very easily and releases a lot of energy in the process. This ring strain causes cyclopropane to release more energy on combustion than a standard strain-free propane chain.

2: Precursors of the Wadsworth Emmons Cyclopropanation Reaction

As with most reactions the Wadsworth-Emmons cyclopropanation reaction is simply a different application of an older reaction, the Horner-Wadsworth-Emmons reaction, which has the purpose of forming E- alkenes selectively. This in turn is a derivation of the original Wittig reaction first discovered in 1954 by Georg Wittig in which phosphonium ylides are used in order to form alkenes products from aldehyde or ketone reactants.

2.1: Wittig Reaction

The Wittig reaction is very useful in that it will form a carbon-carbon double bond in one site specifically on the desired molecule but the stereoselectivity of the reaction is controlled by the type of phosphonium ylide used. An ylide is a species that carries a positive and a negative charge on adjacent atoms of the molecule, and in this instance it is a phosphorus atom that carries the positive charge. As seen in scheme 12, the negatively charged carbon atom of the ylide acts as a nucleophile towards the carbonyl of the ketone (electrophile) and forms a betaine species. The betaine cyclises into an oxaphosphetane ring which quickly collapses to form a very strong phosphorus-oxygen double bond and results in the production of an alkene and a triphenyl phosphine oxide.

When an unstablised ylide is present in the reaction, the kinetic isomer (Z-alkene) is produced preferentially (figure. 5)2 as the intermediate oxaphosphetane ring forms irreversibly. As the stereochemistry of the substituents are ‘locked’ into a syn-conformation (figure. 7), when elimination of the triphenyl phosphine oxide occurs the alkene formed has it’s substituents on the same side of the plane.

When the negatively charged carbon is adjacent to an electron withdrawing group (EWG), in figure 6 this is represented by the ester substituent, the ylide group becomes more stable as the charge can be dispersed. This leads to the formation of the enolate resonance form and is referred to as a stabilised ylide. Unlike the unstabilised ylide, the oxaphosphetane ring that is formed is now a reversible reaction allowing interconversion between Z-orientation to E-orientation of groups in the ring before an elimination step occurs. Under the right conditions, the interconversion step can become faster than the elimination of phosphine oxide step (collapse of the ring) allowing the reaction to proceed via the thermodynamic product route. The E- isomer is the thermodynamic product as the anti- conformation (Figure.7) of the oxaphosphetane ring has the substituents on opposite sides of the molecule, reducing steric effects and producing a conformation lower in energy. The elimination of the Z-isomer is slower than that of the E-isomer, allowing the oxaphosphetane ring to open and rotation about the carbon-carbon bond to occur to form more of the E-isomer.

Although the Wittig reaction works efficiently with simple carbonyl reactants, the more sterically hindered a ketone reactant is, the slower the reaction can become. This does not necessarily have a negative effect on the yield of product but will affect the suitability of the reaction in time sensitive application, e.g. the commercial industry wants a high yield of product with a moderately fast synthetic route in order to keep costs low.

2.2: Horner-Wadsworth-Emmons reaction

The Horner-Wadsworth-Emmons reaction [9] (scheme 2.) is the preferred method to select for the E-alkenes product and instead of using phosphonium ylides (figure 6.) uses much more nucleophilic phosphonate-stabilized carbanions with an EWG attached, usually in the form of phosphonate esters. Firstly, the phosphonate ester is deprotonated using sodium hydride leading to the generation of an enolate species. This enolate/stabilised carbanion is then reacted with the chosen aldehyde or ketone to give product. As a result of its more nucleophilic nature, similar or better yields are produced and faster rates of reactions for aldehydes and ketones that are more sterically hindered are observed. The stereoselectivity of the reaction for E-alkene can be further increased by modifying the reaction conditions and the substituent groups of the phosphonate ester (figure 8.) [10] . The larger the alkyl groups attached to the phosphate and ester functional groups the greater the proportion of E-isomer obtained. Using the same idea, the larger the substituent group attached to the aldehyde/ketone reagent a more improved E-selectivity is seen, for example a phenyl ring. Increasing the temperature of the reaction to room temperature (23°C) and changing the solvent from THF to DME will also encourage E-selectivity.

Figure 9 shows that the pka of the HWE reagent is lower than that of the Wittig reagent and this is due to the ester EWG on the adjacent carbon to the acidic hydrogen. The EWG helps to stabilise the carbanion that will be formed by the loss of hydrogen making the phosphonate ester a stronger acid than the phosphonium salt, whose conjugate base will be less stabilised.

As seen in Scheme 2, along side the E-alkene, there is a water soluble phosphate molecule present in solution. Due to its solubility, the recovery of the pure product from the solution can be done via a simple work up and this is one of the advantages of the HWE reaction over the use of stabilised ylides where a φ3 phosphine oxide is formed. Through Wadsworth and Emmons’ investigations into the formation of alkenes such as stilbene in 19619, it was reported that the use of phosphonate carbanions was a more cost effective process that led to faster rates of reactions. It also produced very good yields in more mild conditions in comparison to stabilised phosphonium ylides. Phosphonate carbanions have a greater scope in number of different ketone and aldehyde reagents that they can successfully react with.

Comparing both methods, when using stabilised ylides the resulting solution will contain a mixture of the isomers and therefore a suitable method is needed in order to separate them.

Scheme 2.

2.3: Wadsworth-Emmons Cyclopropanation

Similarly to the HWE reaction and keeping in mind the steric effects of large substituents, the reaction uses phosphonate-stabilised carbanions like the phosphonoacetate anion with epoxide and lactone reagents in order to form trans-cyclopropane rings within molecules. As seen in scheme 3 [12] , the phosphonate carbanion acts as a nucleophile towards the electrophilic carbon of the epoxide resulting in the opening of the strained ring. Due to the negative charge present on the oxygen atom the phosphoryl group undergoes 1, 4 migration on to the oxygen, forming another carbanion. The carbanion can then cyclise leading to the γ-elimination of the phosphono- γ-oxyalkanoate and the closure of the cyclopropane ring. These stabilised phosphonates give a similar yield of trans-cyclopropane to reactions using phosphonium ylides but with faster reaction times and improved diastereoselectivity [13] .

3: Other Methods of Cyclopropanation

In order to understand how effective the WE cyclopropanation reaction is and its advantages, other methods with slightly different approaches to the same problem can be looked at.

3.1: Simmons-Smith Reaction

First developed in 1958, the Simmons-Smith reaction uses the chemistry of carbenes groups, producing cyclopropyl rings from the interaction between alkenes and a carbene derivative, zinc carbenoid. A standard carbene ‘is a neutral species containing a carbon atom with only six valence electrons’2 and can be inserted into σ-bonds and/or Ï€-bonds of other reagents. Examples of these carbenes are :CH2 and :CCl2, where the whole carbene reagent is incorporated into the final product structure. In cases using reagents like :CCl2, further steps are required to remove the halide atoms if a standard cycloproyl ring is desired. In comparison, the zinc carbenoid is a species capable of forming carbenes but does not react in exactly the same way as them. The zinc carbenoid is formed by the insertion of a zinc atom into a molecule diiodomethane using a copper catalyst, as seen below, and its mechanism of action is compared to that of a singlet carbene where the reaction is concerted. The carbon structure (-CH2) within the zinc carbenoid is incorporated into the cyclopropyl ring whilst the resulting metal halide is released into solution. This is done via an intermediate complex formed between the alkene, carbene and metal halide so that the carbene is not released on its own. One of the advantages of this method of cyclopropanation is the ease with which the stereochemistry of the product can be controlled. As the reaction is stereospecific, in order to obtain a product with a trans-cyclopropane ring, an alkene with E- stereochemistry can be used as the original stereochemistry will be retained. The rate of reaction of this can be dramatically increased by the presence of allylic alcohols with the same stereochemistry as the alkene, as the zinc atom can coordinate with the oxygen in a transition state to add the carbene to the same face of the molecule. The example below (figure. 10) [14] shows that this reaction is extremely effective at producing high yields of trans-cyclopropane product.

Scheme 4.

This reaction exhibits easy control over stereoselectivity and undergoes a relatively simple mechanism, making it easy to understand why this is one of the most popular methods of cyclopropanation. A disadvantage that the WE cyclopropanation reaction does not share is that there will have to be further steps taken in order to remove the product from the solution containing the zinc halide (insoluble) whilst preventing impurities being obtained. Although in some instances the Simmons-Smith reaction has greater stereoselectivity than the WE cyclopropanation reaction with comparable yields.

3.2: Corey-Chaykovsky Reaction [15] 

This reaction uses sulphonium ylides, as opposed to the phosphonium ylides of the Wittig reaction, reacting with enones in order to form cyclopropyl structures in the molecule. Firstly there is the generation in situ of the dimethyloxosulfonium methylide, often called the Corey-Chaykovsky Reagent (CCR), from dimethyl sulfoxide and methyl iodide reacting to give a trimethyl sulfoxonium iodide salt. This salt is then deprotonated using a strong base like sodium hydride resulting in the CCR.

In the mechanism of cyclopropanation, the CCR acts as a methylene transfer agent, with the carbanion acting as a nucleophile towards the alkene carbon-carbon double bond of the enone. This 1, 4 addition is followed by cyclisation within the molecule using the new carbon double bond reacting as a nucleophile toward the now electrophilic ylide carbon to form the cyclopropyl structure and a sulfonium cation (Scheme 5).

In an attempt to make the reaction stereoselective more substituted sulfonium ylides with specific chirality can be used to encourage the formation of a specific enantiomer as they transfer other substituents to the enone as well as methylene (figure.) [16] .

Scheme 5.

4: Uses of the Wadsworth-Emmons Cyclopropanation Reaction

Though there are only a few specific examples of WE cyclcopropanation in action, a good idea of its efficacy can be obtained.

4.1: Synthesis of Belactosin A [17] 

(+)-Belactosin A is a naturally occurring antitumor antibacterial compound that acts as an alkylating agent in chemotherapy treatment. As an alkylating agent [18] it adds alkyl groups to electronegative groups such as phosphates or the amines found on guanine nucleotide bases, which are present in all cells of the body, although it is used to target mutating cancer cells. Belactosin A specifically stops the cell cycle of cancer cells at the G2/M phase, where normal DNA will have been replicated and the cell undergoes mitosis. In mutated DNA, the areas on the nucleotide bases affected by alkylation form cross bridges with other atoms on the complementary base of the opposite strand of DNA. These ‘bridges’ prevent the DNA strands from separating at these specific points, stopping steps such as transcription. As a result this prevents the mutated DNA from being copied, cells from dividing into more cancer cells and halts proliferation of these cells through out the body. It affects mutated DNA cells more readily as they undergo cell cycle at a faster and uncontrolled rate and their repair mechanisms are less effective.

Armstrong and Scutt reported a good yield of 63% of the cyclopropane intermediate with greater than 95%e.e. By using H1 NMR and NOE they determined that the product obtained was the trans-isomer.

4.2: Synthesis of (R,R)- 2-Methylcyclopropanecarboxylic Acid [19] 

In agriculture and veterinary practices, insecticides like Cyromazine and Pyrethrum Extract4 are formed using the WE cyclopropanation reaction in the synthesis of (+) – and (-)-chrysanthemum dicarboxylic acids from anhydro sugars [20] . The pyrethroids are active molecules that prevent normal transmission and excitation along the nerve cells in insects by acting on sodium/potassium channels. This results in immediate death to agricultural pests such as locusts and parasitic insects such as ticks and fleas on household pets. Due to the number of nerve cells and the speed of transmission these insecticides are up to 100 times more effective on insects than humans. Consequently, Pyrethrum extract can be used pharmaceutically without detrimental effects in the treatment of worms and scabies.

Using the WE cyclopropanation method to obtain the biologically active enantiomer, Brione and company obtained excellent product yields of ~85-90% with exceptional trans-selectivity (>98%). These results were obtained under the conditions of 150°C with HexLi/MeTHF solvent.

5: Conclusion

Whilst the WE cyclopropanation reaction proves itself to be a useful step in the mechanism of formation of a few interesting biologically active compounds, the fact remains that it is an underused method. This is shown in the small volume of literature that can be obtained specifically for this reaction. As with most reactions the right balance of factors and reaction conditions are needed to get the most efficiency, and excellent yields have proved that the WE cyclopropanation reaction is capable of this in the cases of (R, R)- 2-Methylcyclopropanecarboxylic Acid and Belactosin A. Perhaps one of the reasons it is overlooked as a synthetic route is the presence of better known reactions like the Simmons-Smith reaction, as there are still some small areas that are not fully known e.g the degree of specificity of the reaction. The field of chemistry is one based on the evolution of ideas and continued search for improved yields and rates of reaction, especially in a growing area such as the synthesis of synthetic analogues of natural compounds. In the same way that the WE cyclopropanation reaction was derived from the Wittig reaction, it could provide as a good basis for future improved methods of cyclopropanation that arise from the modification of its reagents.