Green synthetic rout to Sildenafil

Sildenafil

= Viagra

Intended use

In the search for new cardiovascular medication, new inhbitors for PDE (phosphodiesterase) were screaend

• drugs like zaprinast weren’t succesful enough

Sildenafil was teste

• showed not potent along with ‘side effects’

• it was rapidly trialed and gave good results (physical and psychological)

• commericalized fast

Becasue of the rapid growth in sales of sildenafil, the environmental efficiency of its synthessis has been immediately considered

goal of sildenafil

aim = inhibition of PDE5

• there are 7 different types of PDE5 in the body

• 3 of them are able to selectively hydrolyse cGMP to GMP or cAMP to AMP

  • therefore selectivity is often more important than potency

  • the structure of the sildenafil should be similar to that of the enzyme substrates (cGMP and cAMP)

• PDE5 can be found in several parts of the human body

  • vascular and visceral smooth muscle are the ones responsible for erection

How Sildenafil works

1. during sexual stimulation, NO is released in the erectile tissue of the penis

2. NO activates GC (guanylate cyclase) which converts GTP in cGMP

3. cGMP causes relaxation of the smooth muscle in th blood vessels of the erectile tissue allowing more blood flow causing erection

4. normally, cGMP is broken down into GMP by the PDE5

5. In men with erectile dysfunciton, the NO/cGMP signaling patway may be weak or not sustained for long engough

6. the little cGMP produced will be broken down to quickly = insufficient smoot muscle relaction and bloodflow

7. sildenafil inhibits PDE5 preventing cGMP breakdown allowing cGMP to accumulate longer to maintain smooth muscle relaxation

—> sildenafil does not directly cause eraction, sexual stimulation is still needed

other facts

Sildenafil is much more selective for PDE5 over the other PDE enzymes

• except only 10 fold selectivity better than PDE6 which is located in the retina

• this can cause abnormalities in color vision

• also inhibition of PDE5 in other parts of the body (than the vascular and visceral smooth muscle) can cause side effects

Substituent at

• position 3 substituent (propyl) is supposed to mimic the ribose part of cGMP

• position 5’ (sulfonamide) is supposed to mimic the phosphate bond

  • this is on the other side of the molecule compared to cGMP

  • however the structure of the molecule allows it to fit well anyway

Intermediate A: initial route and problems:

Problems

• hydrazine

  • explosive

• Me₂SO₄

  • carcinogenic

• SOCl₂

  • corrosive and toxic

  • reacts violently with water producing toxic/ corrosive gasses: HCl and SO2

  • used in large excess as reactant and solvent

• SnCl₂ =

  • toxic heavy metal

  • poor atom economy: reducing agent that only delivers electrons, Sn and Cl are wasted

  • expensive: need stoichiometric amounts

  • but justified

    • other reductions were not efffective, especially hydrogenation

    • large excess of thionyl chloride led to inclusion of traces of sulfur that cause catalyst poisoning needed for hydrogenation

Quick fixes

• Only use stoichiometric amounts of SOCl₂ and use toluene as a reaction medium

  • this improved yield of amide formation

• This also allowed the use of catalytic hydrogenation in the second step

  • improved yield of reduction

  • only water as by product (much better AE)

  • both catalyst and solvent can be recovered

    • EtOAc is a sustainable solvent

acyl chloride formation mechanism

acyl chloride foramtion with DMF

release of gas —> drives equilibrium to the right

however: CO is toxic, and cannot see or smell it so dangerous

nitration of benzene

Intermediate B: initial route and problems and quick fixes

uses oxalyl chloride in DCM

• DCM

  • chlorinated solvent and considered carcinogenic

• oxalyl chloride

  • releases CO which is a toxic gas

Quick fix

• replace with SOCl2 with DMF cat

  • increases yield to quantitative

  • avoids elimination of CO

  • EtOAc much better solvent

Combining A and B: initial route

• amide formation

  • low yield: likely due to low nucleophilicity of the amine by the é-poor pyrazole ring

  • DCM as solvent

  • requires chromatography —> cannot be scaled up

• ring closure

  • challenging step but reasonable yield: would expect low reactivity due to low nuc of amide N and low electrophilicity of amide carbonyl, but intramolecular so feasible

  • H2O2 usually acts as oxidant but here as H-bond donor (stronger than h2o because 2 electron withdrwing oxygens)

    • increases electrophilicity of the amide carbonyl

  • requires workup (a lot of organic and aqueous waste) and chromatography —> no scale up

  • DCM as solvent

• chlorosulfonation

  • low yield

    • large volumes of acidic aqeous waste that needs to be neutralized

    • chance of hydrolysis during quench

  • DCM is needed because of FG sensitivity

• piperazine introduction

  • ok

• salt formation

  • amine (from piperazine) is turned into a salt using citric acid

    • the ammonium salts are easier purify by recrystalization in aquous acetone and easier to handle, , scale up, formulate and suitable as a pharmaceutical solid

    • also protonation of the amine reduces electron density and nucleophility of the N making it less prone to oxidation (eg by air) or side reaction

Overal 10% yield starting from intermediate A

—> not great for scale-up

role of DMAP

DMAP attacks on the acid chloride and the amine attacks on the formed DMAP intermediate

Combining A and B: Quick Fixes

• Amide formation

  • replaced DMAP and DCM by pyridine and EtOAc

  • much better solvents

  • higher yield

  • no need for column chromatography

• ring closure

  • stronger bases under water free conditions

  • quantitative yield

  • no column required

• chlorosulfonation

  • still low yield

    • large volumes of acidic aqeous waste that needs to be neutralized

    • chance of hydrolysis during quench

  • DCM for FG sensitivity

    • The reaction is quenched with water.

    • The product is extracted with dichloromethane CH2​Cl2​.

    • The DCM solution is then subjected to solvent exchange, meaning DCM is removed/replaced by toluene. because not isolated and needed for next step

• piperazine formation

  • toluene instead of ethanol

• salt formation

  • same

36% yield starting from A = much better

chlorosulfonation and chlorosulfon(yl)ation

problems with current synthesis

• current synthesis is completely linear

• cleanest reaction is in the middle of the synthesis (cyclization)

• chlorosulfonylation

  • uses SO3HCl which is highly toxic and corrosive

  • produces SO3 which is also toxic and corrosive

  • and this is used at then end of the sequence = higher chance of toxic impurities in final product

toward an environmetnal (economical synthesis)

solutions:

• move clean cyclization to the end

• move toxic chlorosulfonylation to the beginning and follow up with piperazine reaction befor connecting with A

  • the environmental issues associated with scale up of the chloro sulfonylation have been reduced by moving it to the beginning of the synthesis

    • a lot of acidic aqueous waste that needs to be neutralised

    • increased level of hydroylisis during quench

  • this allows isolation of a very clean final product

    • further toxic impurites can be removed ruing subsequent purificiaton steps

• the is allows a convergent synthesis

  • not all functional groups need to survive each step

Commercial route of Sildenafil

Intermediate A

• only need stoichiometric amount of SOCl2 by using toluene as solvent

• toluene can also be recovered and reused, minimizing solvent waste

• stoichimetric amount of SOCl2 allows hydrogenation in Ethylacetate

  • high AE

Sulfonamide

cholorosulfonylation has been conducted in the presence of SOCl2 to ensure the presence of sulfonylchloride for the reaction with methyl piperazine

• 2-ethoxy benzoic aic also has a low melting point

  • it can therefore be used melted, so we can reduce the amount of chlorosulfonic aicd and thionyl chloride and no organic solvent because no problmes of solubilization

• quench of the reaction is combined withe methylpiperazine coupling in heterogenous conditions

• pH of the final solution can be adjusted to precipitate the product, isolate it via filtration avoiding the need of any organic solvent (that would be needed for extraction)

• Thus no organic solvent is need for the preparation of the sulfonamide

Coupling reaction

several conditons were screened including acylchloride formation with SOCl2 and oxalyyl chloride but went with CDI

Although CDI cost is very high, it’s chosen for several reasons:

• allows clean and trustable chemical step

• allows combination of several chemical steps

  • nitro reduction by hydrogentation

    • also conducted in EtOAc

  • COOH activation

    • which previously required oxalyl chloride in DCM

    • after quick fix: SOCl2 with DMF (cat) in EtOAc

  • amine acivation = amide formation

    • which previously required DMAP in DCM

    • after quick fix pyrdine in EtOAc

• allows EtOAc to be the only solvent used

  • simpler recovery of solvent and minimize energy use

• minimize the emmision of volatile organic compounds (VOC)

  • because no Ethylchloride can be formed anymore because no SOCl2 (or oxalyl chloride) is used anymore that could react with EtOAc

    • would be the case if coupling would still occur through SOCl2 (for COOH activation)

  • these are difficult to prevent spreading in the environment and bad because halogenated

• allows for a good yield of 90% which has been optimized to 96%

disadvantages

• High cost

• poor atom economy: because all atoms of the reagent are lost

—> but this is overruled by massive number of other advantages

cyclization

• ‘clean’ cyclization can be performed under ‘highly concentrated’ conditions allowing waste to be minimized

• add the end of the process, water is added and the pH is adjusted to induce precipitation so sildenafil can be isolated by filtration

  • sildenafil free base becomes poorly soluble and precipitates$

Salt formation

• using citric acid with 2-butanone

• high yield and purity

  • no recrystalization anymore because separated by filtration after cyclization

Overal yield = 75%

CDI mechanism

N,N-carbonyldiimidazole = staab’s reagent

Solvent optimization

• Much less solvents

  • no halogenated solvents or highly volatiles solvents anymere = less emissions

• solvents can be recovered:

  • toluene and EtOAc are recovered since the commerical route has been established

• Further optimization

  • no more tBuOH

    • soluble in water and therefore hard to recover

  • 2-butanone has also been optimized for recovery

1300 to 7 L/kg (future target = 4L/kg)

E-factor and impact atom economy