Help with quinoline synthesis

We have been puzzled by a synthesis problem for a while now where our target molecule is a 4-hydroxyquinoline derivative with a formyl group at the third position.
So far the attempts at the most obvious pathway, a simple Reimer-Tiemann formylation, have been unsuccessful with no detectable product. A paper in JOC reported a 16% yeild for a similar compound but we have been unable to replicate the results for our compound.

Now I'm not looking for the work to be done for me but I'm just interested to know if anyone has some other synthesis ideas on this molecule that we could explore.


I have attached a picture which should give a better idea on what we are aiming at with some proposed pathways.

Refs for the pic:

1. Chem. Rev., 1942, 30 (1), pp 113–144
2. J. Am. Chem. Soc., 1946, 68 (7), pp 1232–1238
3. Tet. Lett., 2002, Vol. 43, Is. 8, p. 1395-1396
4. Bull. Korean Chem. Soc. 2006, Vol. 27, No. 1, p. 121 - 122

Thank you ScienceSquirrel for the reference! The Duff reaction was tried out, but although it did give the desired aldehyde it did so in only 15% yield. Now I don't know if it could be improved, but so far the message I got from the lab supervisor was that I should look for other routes as TFA is notoriously expensive.

Other things I tried out so far were the synthesis of the 3-methyl substituted 8-bromo-4-hydroxyquinoline through the Conrad-Limpach reaction. This was found problematic as the imine was a bit hard to make. First the ester and aniline was combined with few drops of conc. H2SO4, put in vacuo over P2O5 overnight, resulting mixture added to Ph2O, refluxed for 4h, yield 12%. Other attempt at the pure imine formation was by aniline + ester in toluene, reflux with dean-stark overnight gave 32% pure imine after flash. This in refluxing Ph2O gave 45% yield of the 3-methyl compound based on the used imine. A synth.comm. article mentioned alternative solvents for the C-L reaction, but at least the 2,6-di-tert-butylphenol they dubbed as "the best solvent" wasn't working for our substrate as even a 12h reflux didn't yield any precipitate after cooling down. Addition of equal volume of diethylether did precipitate something crystalline but it did so with the formation of a _lot_ of tar and workup was found impossible (or I just lacked patience.. the tar clogged frits immediately). I'll try out a way to make the imine similar to the Gould-Jacobs reaction (aniline and ester heater together neat or maybe with a hint if acid catalyst) and then heat it in Ph2O as at least I know it's relatively insoluble in that.

The 3-methyl substituted product that I managed to isolate was then subjected to SeO2 oxidation (if anyone has anything better in mind for selective methyl group oxidation to the aldehyde I'd love to know as SeO2 is another expensive and rather toxic reagent) in refluxing xylene. This gave a yield of 25% which although better than anything previously tried, still sucks.

The third pathway is looking like the best option so far. This is the synthesis of the 3-substituted nitrile by the Gould-Jacobs reaction from the aniline and a ethyl (ethoxymethylene)cyanoacetate (both of which are reasonably priced too! ), and the subsequent reduction of this nitrile to the aldehyde. Preparation of the nitrile was easy (first try at 5 mmol scale yielded 34% of the nitrile, pure by NMR straight after drying), although it's amazingly insoluble in everything. I can only hope it doesn't matter in the DIBALH reduction. Other options for the reduction are available but the reagents are more exotic and expensive (LDBBA, PtO2)..

That's it for now, hopefully someone will find this interesting. =D


Links for the mentioned reactions if someone isn't familiar with them:

http://en.wikipedia.org/wiki/Gould-Jacobs_reaction
http://en.wikipedia.org/wiki/Conrad-Limpach_synthesis

If you feel inovative, you can try a Mannich reaction with morpholine followed by the oxidation of the Mannich base hydrochloride with DMSO.
The Mannich on your 8-bromo-4-hydroxyquinoline should be high yielding. On plain 4-hydroxyquinoline the yield using morpholine is reported to be 95% [1]. Similarly on 4-hydroxy-2-methylquinoline with diethylamine [2]. There must be plenty other examples in the literature, but I have not bothered much with the search.
Make sure you use morpholine or some other amine having poorly electrophilic methylene groups in the Mannich base (don't use dimethylamine!).

The Kornblum on benzylammonium salts has been reported [3]. Since amines are very poor leaving groups, the reaction requires harsh conditions (up to 180 °C). The Mannich base has a hydroxy group ortho to the aminomethylene group, so that it is prone to elimination forming the corresponding ortho-quinomethane. These are highly electrophilic and should easily O-alkylate DMSO and do the Kornblum thing. This might allow the reaction to proceed at lower temperatures.
I would also not be surprised at all if a one-pot reaction starting from the 8-bromo-4-hydroxyquinoline, trioxane and morpholine hydrochloride in DMSO turns out feasible.
On the other hand, heating the Mannich base in DMSO might only give crap. But you can't be sure until you try. If for some strange coincidence it turns out a success, make sure you do not forget citing Sciencemadness in your report/article. Don't make me writing to the editors!

[1] Bulletin of the Korean Chemical Society 1999, 20, 973-976. (freely availble online)

[2] Canadian Journal of Chemistry 1966, 44, 1863-1865. (old issues are freely available, I think)

[3] J. Org. Chem. 1970, 35, 2207–2212.

Thank you fledarmus! The article isn't available to me online, but I'll be sure to check it on monday when I visit the library.

Very interesting stuff on the Mannich Nicodem, thanks a lot! I was going to try the Kornblum on the 3-methyl substituted one (after bromination with NBS), but that is for when all else fails. This would be much better though, skipping few steps too. Then again it might be too good to be true, but as you said, I need to try it to know (and I'll remember scimad if it does )! And all that Me2S aroma will mix nicely with the mercaptan stuff going on in the fumehood next to me.

So the DMSO oxidation was tried out but unfortunately no aldehyde was found in the extracts after workup. I used some previously prepared mannich bases, made with piperidine and morpholine. These were around 2 months old, but were still pure by NMR so I doubt the fault was in the starting materials. Hydrochlorides were prepared by gassing with HCl in DCM suspension/solution.
On 1mmol scale the DMSO amount used in the JOC article wasn't enough to result in a stirrable mixture so I don't know if this contributed to the black, tarry reaction mixtures after the 20h heating period (under argon). Smell of dimethyl sulfide was very strong (and I had to clean the whole schlenk line because I used the argon line and not a balloon, silly me) but this was to be expected. Workup was done with 1M HCl / brine / EtOAc. It might be that an inert, high-boiling solvent could help in preventing over oxidation/tar formation but unfortunately I don't have enough time to explore this pathway as much as I'd like.

Nitrile still refuses to get reduced even with fresh DIBAl (I even tried refluxing in one attempt for ~30min, no reduction ). The 4-OH-quinoline apparently does a good job at complexing with the aluminium hydrides rendering the reagent inert.
Oh well, time to start working through a whole different approach..

Anyways, thanks for all the suggestions!

Quote:

Nitrile still refuses to get reduced even with fresh DIBAl (I even tried refluxing in one attempt for ~30min, no reduction ). The 4-OH-quinoline apparently does a good job at complexing with the aluminium hydrides rendering the reagent inert.

Yeah I used +2eq DIBAL for the nitrile in both THF and DCM, neither showed any reduction products. The Mg-formylation was also attempted but this only resulted in some funky polymer (dunno if it was the THF, had some trouble dissolving everything in acetonitrile).

SnCl2/HCl method was looked into as well and it actually seemed like it would have worked as a yellowish precipitate formed shortly after introducing the nitrile into the Et2O/SnCl4 mix but only starting material was obtained after workup.

I could give the nickel one a go as I'm going to try PtO2 again under more forcing conditions so I can run them in parallel, but most likely I will find myself with the amine considering how easily the imine would be reduced..

Well, to my amazement the Raney nickel did the job in the microwave reactor. A 50% yield was obtained from the nitrile with a very simple workup.
The platinum oxide in return gave nothing with the same conditions.

Thanks again Nicodem.

Now to do the same with some napthyridines...