Post a Synthesis

[mississippichem]

I figured I would start a thread for us chemistry folks to post some reaction schemes and discuss synthesis design, as well as relevant laboratory techniques.

 

I'll start things off with the total synthesis of Sildenafil [trade name "Viagra"]. The drug is administered orally as the ammonium-citrate salt. This is the lab scale synthesis and is covered in detail in Bioinorganic and Medicinal Chemistry Letters Vol. 6.

 

This scheme is brought to you courtesy of: www.ch.ic.ac.uk

 

 

Notice the sulfonamide that shows up in the last step. I chose this synthesis because of the simple functional group transformations that lead to a highly marketable drug. I also like the beta-dicarbonyl+hydrazine step to give the pyrazole; A clever variation on the typical ketone -> hydrazone reaction.

Edited January 30, 2011 by mississippichem

[Horza2002]

Hey, I think this idea is a very good one. Just want to clean something up though, are you proposing we develop a new synthetic scheme for this or just to disuss the synthesis. Is this the route that it is produced commerially? As some of these steps don't seem very efficient.

[mississippichem]

Hey, I think this idea is a very good one. Just want to clean something up though, are you proposing we develop a new synthetic scheme for this or just to disuss the synthesis. Is this the route that it is produced commerially? As some of these steps don't seem very efficient.

 

If you have an idea for improving a synthesis, that's fine. I just wanted to start a thread for general synthetic discussion; whatever that may encompass. Feel free to post your own syntheses, not just drugs either; anything really. I just posted this one to start things off.

 

No this isn't the industrial synthesis. It's the second patent filed for the synthesis of the drug at the lab scale. I'm willing to bet that the industrial synthesis starts at the pyrazole though.

[insane_alien]

remember, on an industrial synthesis you can boost efficiency of a reaction step by the use of recycle loops. theoreticaly to 100% although to do so would require infinite money and resources. but getting close can be done much more cheaply.

[Horza2002]

Yes, while that is true, that is not always possible to use recycling reactions. My guess is that they would start there as well.

 

I've just had a look more closely at these steps...and actually, they are all very simple (ester hydrolysis etc). This is quiet a good synthesis as it stands.

Edited January 30, 2011 by Horza2002

[hypervalent_iodine]

Great post mississippi! I've always loved hydrazine I've never actually seen the synthesis of Viagra before, it's quite simple and eloquent when you look at it.

 

I thought I'd include a synthesis that the group I work in helped complete before I started with them (found at http://www.organic-c...07December.shtm). The group I work in (Williams group) worked mostly on the Wittig reaction at the end of the synthesis, which is nice in itself since the ylide can have a wider application outside of this synthesis. Anyway, I like this because it's wonderfully entioselective and it scares first years to see what organic chemist's have to deal with.

 

Overall reaction:

 

 

 

First synthesis leading to (2)

 

 

 

Second synthesis leading to (15)

 

 

 

 

 

Final synthesis leading to (3)

 

 

remember, on an industrial synthesis you can boost efficiency of a reaction step by the use of recycle loops. theoreticaly to 100% although to do so would require infinite money and resources. but getting close can be done much more cheaply.

 

 

 

As Horza said, this is true but using cheap catalysts in your reactions are not always possible. Also, you are incorrect in saying 'theoretically to 100%'. You almost always have byproducts in your reaction, some that form reversibly some that don't. Quantitative yields are nice, but hard to get.

Edited January 30, 2011 by hypervalent_iodine

[mississippichem]

Great post mississippi! I've always loved hydrazine I've never actually seen the synthesis of Viagra before, it's quite simple and eloquent when you look at it.

 

I thought I'd include a synthesis that the group I work in helped complete before I started with them (found at http://www.organic-c...07December.shtm). The group I work in (Williams group) worked mostly on the Wittig reaction at the end of the synthesis, which is nice in itself since the ylide can have a wider application outside of this synthesis. Anyway, I like this because it's wonderfully entioselective and it scares first years to see what organic chemist's have to deal with.

 

 

In step 5, is your rhodium catalyst a chiral [ce]Rh_{2}(DOSP)_{4}[/ce] type species?

 

 

I've never seen epoxidations with those catalysts, but I have seen enatioselective C-C addition over a double bond also in the presence of a terminal azo-group with that same rhodium species. New enatioselctive epoxidations are always nice to see..."Sharpless" assymetric epoxidations are nice but the tetrakisisopropyltitanium(IV) reagent can be a pain to filter out.

 

By the way, nice microwave assisted allyl migration heading into 15. I wonder if that proceeds as a [2+2] "pseudo-concerted" followed by a C-O cleavage? It would seem typical of microwave rearrangements but I'm just speculating.

Edited January 31, 2011 by mississippichem

[hypervalent_iodine]

In step 5, is your rhodium catalyst a chiral [ce]Rh_{2}(DOSP)_{4}[/ce] type species?

 

 

I've never seen epoxidations with those catalysts, but I have seen enatioselective C-C addition over a double bond also in the presence of a terminal azo-group with that same rhodium species. New enatioselctive epoxidations are always nice to see..."Sharpless" assymetric epoxidations are nice but the tetrakisisopropyltitanium(IV) reagent can be a pain to filter out.

 

By the way, nice microwave assisted allyl migration heading into 15. I wonder if that proceeds as a [2+2] "pseudo-concerted" followed by a C-O cleavage? It would seem typical of microwave rearrangements but I'm just speculating.

 

You're talking about the cyclopropanation step (there are no epoxides in there). It's not actually a new discovery. The group that Williams was collaborating with (Davies group) used Rh2(S-DOSP)4 and Rh2(S-PTAD)4 to achieve the assymetric [4+3] addition (which then underwent a Cope rearrangement to for the 7 membered ring). This was based on previous work that found it worked as a chiral catalyst for the synthesis of tropanes in a similar fashion. Looking at the paper (I copied an excerpt below), the best yields and ee's were obtained using Rh2(S-PTAD)4 in conjunction with a combination of a methyl ester at R1 and OTBS at R2 on the vinyldiazoacetate, with the remaining to R groups unfunctionalised. This just changes the electrostatic and steric environment of the substrate, which in turn affects how the Rh catalyst approaches it and how the reactants come together in space to undergo the pericyclic reaction.

 

 

The model shown in the paper (below) suggests that the Rh catalyst aligns with the vinylcarbenoid with the OTBS facing away from the phthalimido groups (for steric reasons). The diene then attacks from the front face in the [4+3] addition and the compound then undergoes the Cope rearrangement. Also, the reaction actually has quite a nice scope, in that it could be performed using a multitude of different dienes with good yields and ee's.

 

 

The microwave step, from what I can tell and have read, induces a Claisen type rearrangement.

Edited January 31, 2011 by hypervalent_iodine

[mississippichem]

You're talking about the cyclopropanation step (there are no epoxides in there). It's not actually a new discovery. The group that Williams was collaborating with (Davies group) used Rh2(S-DOSP)4 and Rh2(S-PTAD)4 to achieve the assymetric [4+3] addition (which then underwent a Cope rearrangement to for the 7 membered ring). This was based on previous work that found it worked as a chiral catalyst for the synthesis of tropanes in a similar fashion. Looking at the paper (I copied an excerpt below), the best yields and ee's were obtained using Rh2(S-PTAD)4 in conjunction with a combination of a methyl ester at R1 and OTBS at R2 on the vinyldiazoacetate, with the remaining to R groups unfunctionalised. This just changes the electrostatic and steric environment of the substrate, which in turn affects how the Rh catalyst approaches it and how the reactants come together in space to undergo the pericyclic reaction.

 

Yeah, sorry! cyclopropanation! Duh ! I made that post 6 beers deep. I should stop that.

 

I guess that makes a lot more sense that this Rhodium catalyst does that. Sci-finder gives a load of references. I kept typing in "enatioselective rhodium catalyzed expoxidation" and to my dismay it couldn't find any relevant references. I know why now.

 

[4+3] eh? I'm not familiar with such, but I am with Cope rearrangements. I'll look into it. It makes sense for a Rh(II) bimetallic though since cyclic voltammetry shows that similar complexes tend to undergo a 1 electron reversible oxidation couple. (I did some electrochem work with Ru, Rh, and Os bimetallic aromatic nitrogen heterocycle complexes; which electronically, should behave similarly to your catalyst).

Edited January 31, 2011 by mississippichem

[Dan_]

Hi all,

the idea of a synthesis post seems very good to me. I will hopefully contribute with another example later and just wanted to make a few suggestions:

 

Also, the reaction actually has quite a nice scope, in that it could be performed using a multitude of different dienes with good yields and ee's.

 

The catalysis of that [4+3]Cycloaddition with a Rh(II) precatalyst (not pseudo-[4+3]? Is the Cyclopropane-Intermediate 4 really an intermediate? Could it be isolated?) seems a powerful synthetic tool for the stereoselective preparation of seven-membered rings, but how can I know for sure whether it is, when no yield or ee/ratio of diastereomers is given? (An ee of 90 % is good, thoug. What was the conversion?) I would be very grateful if that could be implemented in the next posts, as well as reagents and reaction conditions. Otherwise it makes no sense to discuss about alternatives or optimization of certain steps of a synthesis.

 

Another nice thing would be to post a retrosynthesis first and discuss the various reasonable and then selected disconnetions and respective transforms. Especially with more complex targets, this is perhaps the best approach to a fruitful discussion...

 

I will be happy to contribute with a synthesis later.

[hypervalent_iodine]

Okay, here is the retrosynthesis from the paper:

 

 

 

And here are the yields and ee's for a few model dienes:

 

 

 

As you can see, there are some pretty good ee's in there on a the dienes they looked at.

 

Looking at previous papers, I can say only speculate whether the cyclopropane 'intermediate' can be isolated (I might actually email my supervisor to ask him if he knows). There were previous studies performed by the other group working with Williams on the project (Davies group) in the lead up to the completed synthesis of vibsanin that managed to isolate similar cyclopropanes. The Cope rearrangement seen in the formation of the cycloheptadiene systems occurs when the reaction vessel is allowed to warm to room temperature. The cyclopropane then adopts a boat conformation, which is what allows for the good stereoselectivity, and undergoes the rearrangement. So I suppose it is possible, but the spontaneity of the Cope rearrangement even at RT suggests to me that it would be difficult. There are two papers in particular that deal with the topic of the cyclopropanation:

 

http://pubs.acs.org/doi/abs/10.1021/ja974201n

 

http://www-scopus-com/record/display.url?eid=2-s2.0-0027488634&origin=inward&txGid=Hzq14-vMO3yYPrAMd-H-ylv%3a2

[hypervalent_iodine]

Regarding the cyclopropane and whether or not it may be isolated, the short answer (as per my supervisor) is not often, as I had suspected.

[mississippichem]

 

Some of my undergraduate work from last year. I worked organic synthesis in a coordination chemistry lab. I like this reaction because it is high yielding and gives nice yellow crystals. The [ce]SeO_2[/ce] oxidation was my idea, as I had seen similar Ar-methyl oxidations in the literature. Interestingly enough, we actually got a small amount of over-oxidation to the carboxylic acid which technically shouldn't happen. 13C NMR in CDCl_3 showed a peak at 193.579 which can only be accounted for by a small amount of the carboxylic acid. The thiazole and imidazole were also prepared with 2-aminothiophenol and ortho-diaminobenzene respectively instead of 2-aminophenol.

 

I have read in literature that the final ring closing can also be accomplished with [ce] O_2 [/ce] gas bubled through xylene. We decided to to use the silver(I) oxide instead though.

 

1H NMR (CDCl_3): 1.20(s), 1.65(s), 7.19(s), 7.61(m), 7.81(m), 8.31(d), 8.71(d)

 

IR: 401, 628, 703, 738, 1038, 1076, 1438, 1451

 

The thiazole analogue was later chosen to be reacted with the ruthenium center because the oxazole was too labile at the desired pH.

 

I should add that the whole scheme scales up very well. The first time I ran it I made about half a gram. The next time I made 40 grams and got a very similar yield.

Edited February 5, 2011 by mississippichem

[Dan_Ny]

There has been some tumult about Maoecrystal V (having three vicinal quaternary streocenters). And the diels-alder reaction in the final step really is beautiful...

Gonh, J.; Lin, G.; Sun, W.; Li, C.-C.; Yang, Z., J. Am. Chem. Soc., 2010, 132 (47), pp 16745–16746

 

Actually, there have been other, interesting approaches to this crowded molecule:

 

Baitinger, I. ; Mayer, P.; Trauner, D. Org. Lett. 2010, 24, 5656-5659.

 

(As well as many other papers...)

 

But nothing beats a good Diels-Alder. Up to 4 stereocenters in one step, perfect atom economy etc. pp.. Is there a better reaction?

Edited March 9, 2011 by Dan_iel