Org Prep Daily

May 18, 2007

Mitsunobu aryl ether formation, 4-piperidinol with 2-hydroxypyridine

Filed under: procedures — milkshake @ 5:32 am


5-fluoro-2-hydroxypyridine 1.844g (16.3 mmol) mixture with N-Boc-4-piperidinol 5.032g (25 mmol) and PPh3 6.82g in anhydrous THF 60mL was stirred under Ar until complete dissolution (20 min). The solution was cooled to 0C on ice bath and DEAD 40% solution in toluene 10mL (Aldrich, 26 mmol) was added dropwise over a 20 min period. The ice cooling bath was allowed to expire over 90 min and the mixture was stirred at RT for 36 hours (over weekend). The unreacted PPh3 was destroyed by addition of a small amount of concentrated H2O2 (50% solution, 0.5mL) followed by few spoons of silica. The mixture was stirred for 10 min, then filtered through a pad of silica (2x2in). The silica was washed with ethyl acetate 400mL. The combined filtrates were evaporated and the residue was dried on highvac. The residue was diluted with toluene 20mL and cyclohexane 20mL was added. The mixture was allowed to crystallize for 1 hour, the precipitate (PPh3=O) was removed by filtration (and washed with some 2:1 cyclohexane-toluene mixture), the filtrates were evaporated. The residue was dissolved in a small volume of cyclohexane-toluene 1:1 mixture, applied onto a column of silica (80g, in hexane) and eluted with ethyl acetate linear gradient, 0 to 20% EtOAc in hexane. (Combiflash LC station with UV detection).  Y=4.526g (93.5%) of a white crystalline solid

1H(d6-DMSO, 400MHz): 8.126(d, 3.0Hz, 1H), 7.673(td, t:8.4Hz, 3.1Hz, 1H), 6.848(dd, 9.0Hz, 3.6Hz, 1H), 5.083(m, 1H), 3.688(dt, d:13.8Hz, t:4.6Hz, 2H), 3.145(br m, 2H), 1.928(m, 2H), 1.522(m, 2H), 1.401(s, 9H)

In this case the used phenol was much more expensive then the alcohol, so the piperidinol (and DEAD/PPh3) was used in excess.


  1. 93.5%?

    Comment by wcm — May 18, 2007 @ 8:41 am

  2. Too low for you?
    (Please note that there is an excess of the alcohol which is the component that tends to crap up during Mitsunobu; if you forget to add the phenol before adding DEAD or if you don’t cool enough or if you add DEAD too fast you end up with lots of the elimination product).

    Comment by milkshake — May 18, 2007 @ 9:21 am

  3. It would be better if it was 93.56%

    🙂 smile. Your just a chemist.

    Comment by wcm — May 18, 2007 @ 10:24 am

  4. Great Job

    I like your way to remove the excess of PPh3, it’s from you or do you have a reference for this??
    Just a small notice for the younglings, generally the hydroxypyridine in position 2 and 4 are in equilibrium between the pyridone form (NHC=O, major) and the hydroxypyridine (N=C-OH, minor), but the chemistry remains the same.

    Comment by hetchem — May 19, 2007 @ 2:33 pm

  5. no reference but I needed PPh3O couple times for a TLC comparison and I noticed that if you add 30% or 50% H2O2 to PPh3 (in MeCN or Me2CO or tHF) the oxidation is so fast and exothermic that it is likely to jump back at you from the test tube.
    I think Mitsunobu complex is a rather hard electrophile that prefers O-alkylation – but I have done some alkylation of 4-hydroxypyridine with a bromoacetate ester and it went to N-alkylated product almost exclusively. Also in DMSO 1H spectra of 4-hydroxypyridine you see the signals of 4-pyridone, not the hydroxypyridine

    Comment by milkshake — May 19, 2007 @ 3:00 pm

  6. 2- or 4- hydroxy pyridines often are seen as the pyridones, same with bi-pyridyl and terpyridyl analogues. Pyridine chemistry is fun and frustrating, all at once.

    Comment by Milo — May 20, 2007 @ 8:43 pm

  7. Any good ways to remove PPh3O if it is too close to the desired product?

    Comment by diketene — May 22, 2007 @ 12:14 am

  8. I don’t know. If your product is that polar, the PPh3=O crystallization would probably not work either – because of a poor solubility of your product in benzene-cyhex mixture – but you should try as well, and see what happens. Even a partial removal of phosphine oxide is a good thing. Also, if you switch from hex-EtOAc eluent to CH2Cl2-EA eluent (with less EtOAc) you often get a separation of things that are close in EtOAc-hex. This is because the hexane-based systems separate compounds not only on the basis of polarity but also based on their solubility (in hexane, it is typically quite poor) whereas with DCM mixture it is polarity-based only.
    When everything else fails, you can substitute PPh3 with PBu3 (or even PMe3 or polymer-supported PPh3 if you have the money to spend).Also, people have been using Dppe instead of PPh3 in Mitsunobu

    Comment by milkshake — May 22, 2007 @ 12:27 am

  9. I don’t know your molecule, if you don’t have too nucleophilic functionnal groups you can use Merrifield resin with sodium iodide in acetone to scavenge your PPh3=O, this methodology has been published for transition metal chemistry by Lipshutz in 2001 it works also for PPh3

    Comment by hetchem — May 22, 2007 @ 12:57 pm

  10. Thanks, milkshake and hetchem. I am a lazy guy. As flash column did not work well, I went with preparative HPLC. I will keep your tricks in mind for future use.

    Milkshake, I really think you should offer a “ask milkshake” service here.

    Comment by diketene — May 23, 2007 @ 12:53 am

  11. Diketene: Fluorous-tagged phosphines and azo compounds are an ideal way to solve the often vexing purification issue of Mitsunobu chemistry. A simple fluorous solid phase extraction after the reaction removes BOTH the TPPO and the reduced azo reagent in one step. Here is a link to a detailed application note for more detail:

    Click to access FTI_AppNote_FMitsunobu.pdf

    In the spirit of full disclosure…I work for the company that sells the technology, but I can honestly say that Mitsunobu chemistry is one of the most popular applications of our technology.

    Good luck ye fellow chemists!

    Comment by FTI Guy — May 23, 2007 @ 11:16 am

  12. So do you think it is a Mitsunobu-type SN2, or an SNAr?

    Comment by atompusher — May 29, 2007 @ 3:45 pm

  13. SN1 + SN2. But no SNAr with normal phenols.

    Comment by milkshake — May 29, 2007 @ 4:03 pm

  14. Mitsunobu reaction between two alcohol is possible?Please send the ref.

    Comment by Paresh — July 18, 2007 @ 3:36 am

  15. In the above example, the Mitsunobu was performed between alcohol and a phenol.

    The only case of Mitsunobu between two alcohols that I know of is the intramolecular cyclization of 1,2-diols to epoxides. (It has been used in sugar chemistry.) With 1,2-diols, the less hindered OH of the two gets activated and undergoes the inversion – but stereo/regioselectivity might not be perfect so depending on the substrate you can easily get a mixture of stereoisomers from a chiral diol. Find the relevant references yourself; I am not eager to do the lit search for you.

    Comment by milkshake — July 18, 2007 @ 4:07 am

  16. I have two questions, for a similar synthesis.

    What is an estimated heat of reaction for this synthesis?

    I end up with mixed solids. PPh3=O and reduced DIAD. What is a good way to separate these solids?

    Feel free to email me

    ~ Alex

    Comment by Alex — December 16, 2007 @ 10:45 am

  17. I don’t know whats the kJ per mol but the formation of the azadicarboxylate-phosphine adduct is instant and very exothermic so you need to start the reaction on ice bath because overheating will cause sidereactions.

    For workup I would try to remove the bulk of phosphine oxide – if your material is soluble in toluene-cyclohexane mix I would try to precipitate the phosphine oxide that away, then directly apply the filtrates on a column. But if your product is not as soluble and co-precipitates with PPh3O, you can try Mitsunobu with tributyl phosphine because Bu3PO is less crystalline and more polar and can be washed away with water to some extent.

    Also DEAD is faster than DIAD and now they sell DEAD solution again, so I suggest you give it a try, the reduced DEAD product will be slighly slower on the column.

    I would have to know what you are making to answer your second question in more detail/

    Comment by milkshake — December 16, 2007 @ 12:06 pm

  18. I think that this reaction works well because of the loan pair nitrogen in the ring. If you don’t that you can add some triethylamine which can deprotonate the phenol.

    Check this out.

    Comment by Thomas Dorn — June 26, 2008 @ 5:40 pm

  19. This looks like good modification.

    I think the actual reason why I got a good yield in this case was that I used an excess (1.5 eq.) of the alcohol because the hydroxypyridine was the most expensive component here – wheresas the protected piperidinol is reasonably cheap.

    My impression from another similar Mitsunobu was that the alcohol elimination on piperidinol was the main side-reaction so adding little more of it made sense.

    Comment by milkshake — June 26, 2008 @ 11:09 pm

  20. would you suggest any way to remove reduced DIAD without column chromatography & TLC system for reduce DIAD.
    thanks & regards.

    Comment by mahesh — October 3, 2008 @ 2:10 am

  21. Unfortunately I have no good idea how would one go about it – I always ended up running a column afterwards.

    I suppose in those cases when the Mitsunobu product is very greasy and the reaction is clean, one could dilute the crude react mixture in toluene and add limited quantities silica into the mix to bind up the more polar sideproducts and then filter off and wash the silica, an imperfect alternative to column chromatography which they sometimes do in the process… The problem with DIAD is that its reaction is slower and the reduced DIAD product is less polar so sometimes it gets fairly close to the product on silica.

    Comment by milkshake — October 3, 2008 @ 2:20 pm

  22. Its good, I also did similar type of reaction.
    Now I want to do similar reaction, one -OH is vinylic i.e attached to 5 membered ring through a carbon atom,one more -OH (oMs form also we have)is a long ether chain, I am facing problem with other conditions, can I do mistunoubu rx here.

    Comment by singh — November 28, 2008 @ 3:27 am

  23. is it good to use Phthalimide as nucleophile than convert phthalimide to amine by hrdrazine hydrate OR we can directly use amine as nucleophile.? what do you suggest aboutcolum chromat. is ti on alumina orr silica gel

    Comment by nilesh jain — June 17, 2010 @ 5:08 am

    • generally silica gel (fine grade) separates compounds much better than alumina, unless you want to separate basic compounds like amines which bind to silica. But I have done basic compounds on silica in systems with ammonia and I got a good separation also.

      You cannot use amine directly in Mitsunobu though one can generate mesylate (or better yet, triflate) and quench it with the amine in situ.

      Both Phtalimide and azide are suitable NH3 equivalents for Mitsunobu but I personally never used either of them in this reaction

      Comment by milkshake — June 17, 2010 @ 2:40 pm

  24. Is it possible to perform mitsnoubu reaction between phenol and primary amine (of amino acid) to get an aryl amine. please send the ref if possible.

    Thanking you in anticipation.

    Comment by pradeep — September 4, 2010 @ 6:29 am

    • no – it is not possible. Please look up the reaction mechanism: Mitsunobu turns aliphatic alcohol into a leaving group, not phenol. It also deprotonates the phenol (or carboxylic acid) – so you have a system that mimics alkylation of phenolate anion with alkylating agent which is made in situ from the alcohol.

      What you want to do in your case is to prepare triflate or nonaflate ester from your phenol. Then you use it in Pd(0) (or Cu(I))- catalyzed arylation of your aminoacid.

      But I will not do the literature search for you.

      Comment by milkshake — September 4, 2010 @ 12:04 pm

  25. I understand that this is a three-year-old procedure, but having done several reactions on 2-hydroxy pyridines via Mitsunobu, I know that N-alkylation products are also possible. How are you sure this is O-alkylation and not N? The NMR’s are approximately the same, I think I remember the N-alkylated product having some peaks more downfield, but i am not sure.

    Comment by gdawg — September 25, 2010 @ 2:22 pm

    • This was one of several 4-(2′-pyridyloxy)-piperidine building blocks that I made for a medchem project, I prepared the others by SNAr from fluoropyridines (NaH, Boc-4-piperidinol, xylene). I am pretty sure that I got O-alkylation product in this case. (I think N-subst 2-pyridone would be far more polar and sticking to silica), nothing unusual showed up on NMR, HPLC, TLC. And the final compound made from this building block was pretty potent, which would not have been the case if the product was N-subst 2-pyridone.

      I think Mitsunobu reagent-alcohol complex is a fairly hard electrophile, so it should preferentially alkylate the oxygen. And with 4-piperidone it is also rather hindered – so that one should favor O-substitution as well

      Comment by milkshake — September 25, 2010 @ 3:30 pm

  26. Hi- I came across this page via google.

    Was interested in using Mitsunobo to couple a dihydroxyquinoxaline -which also contains pyridyl groups- to a benzyl alcohol. The complication is that the this special dihydroxyquinoxaline exists as its double salt with HBr. Considering the mechanism to the M. reaction, would it -in your opinion- be reasonable to just add 2 equiv. of a tertiary amine (e.g. Hünig’s base) to make this one run? Hope this question isn’t too convoluted!

    Comment by fentonh — February 28, 2011 @ 12:25 am

    • HBr salt of NR3 would protonate your Mitsunobu reagent complex (you would not obtain the necessary phenolate anion). Also there would be solubility problems with the salt.

      Freebase your substrate (dissolve the salt in water, add aqueous ammonia, collect the precipitate, dry it on highvac). Benzyl alcohols are somewhat tricky so you may need to start your Mitsunobu at -20C and gradually warm it up to RT.

      Comment by milkshake — February 28, 2011 @ 3:07 am

  27. Thanks for your comments. Yes what you are saying about protonation of the M. complex makes sense. But the NR3 _is_ the phenol (or catechol) Can’t I just add Et3N to the reaction mixture to mop up the HBr from the pyridyl dihydroxyquinoxaline? The complication with free basing as you describe is that the free base is zwitterionic, and very soluble in water. I should really show the structures of the molecules involved 🙂

    Comment by fentonh — February 28, 2011 @ 8:25 am

    • You don’t need to believe me, just try it – it is your material… You may also consult a pK table, to see what is more acidic (OH of your phenol, versus protonated triethylamine).
      By the way what kind of solvent do you plan to use for your Mitsunobu? If your substrate is so polar chances are that it won’t dissolve in THF, MeCN, DCM or any other solvent commonly used for the reaction.

      Comment by milkshake — February 28, 2011 @ 1:19 pm

  28. Hi Milkshake,
    do you think that 4-hydroxyquinazoline can behave like your 2-hydroxypyridine in a similar mitsunobu reaction?

    Comment by madforit — May 25, 2011 @ 12:55 pm

    • I think it might. But make sure that your staring material is soluble in THF.

      Comment by milkshake — May 25, 2011 @ 7:48 pm

  29. is mitsunobu reaction possible in case of 2 phenols?

    Comment by Santosh Taduru — April 14, 2012 @ 8:28 am

    • No it is not. Look up the reaction mechanism for your benefit.

      Comment by milkshake — April 14, 2012 @ 6:09 pm

  30. I made a very similar Mitsunobu with Hydroxyproline instead of Hydroxypiperidinol I always get a side product that has the Mass “product plus DIAD” did you ever observe that? And do you have an idea what it is?

    Comment by Sunny — July 14, 2014 @ 11:49 am

    • I do not have experience with hydroxyproline but I suppose it could dehydrate (in the absence of a good nucleophile) to 4,5-dehydro proline. This would be a fairly reactive enamine, which could possibly add things like the dihydro-DIAD byproduct, in the 5 position next to nitrogen. This is just a speculation. But I have definitely seen dehydratation happening with N-Boc-4-piperidinol, and since this piece is far less expensive than the phenol counterpart, I used it in excess, to get a good yield.

      Comment by milkshake — July 14, 2014 @ 5:34 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: