[vu_nam]

Hi everyone, I tried to make 2-chloro 5-fluoro quinoline from 5-flouro quinoline. I used POCl₃ and refluxed in DCM at the second step but the percent yield was very low ~10 %. Later, I tried neat POCl₃ and increase temp. to 100°C and could obtain ~ 20% percent yield but at least 2 side products also appeared and it was difficult to eliminate. Could anyone suggest me better way to improve selectivity and percent yield of this reaction. Thank you!

[kriggy]

Maybe try Chichibabin reaction and then Sandmeyer? Im however not sure if chichibabin would work.

[TheUnassuming]

Do you know what the side products are? 

You can probably also try N-Chlorosuccinimide and AcOH with water as the solvent.  I don't have full access to the journal but these guys made similar substrates to what you are trying to make.

Monatshefte für Chemie / Chemical Monthly
November 1991, Volume 122, Issue 11, pp 935-941

[vu_nam]

    I think the side products are 4-chloro quinoline derivatives and maybe other chlorine derivative. What made me really confused that the senior who taught me to use the first procedure insisted that she only obtained 1 product, but some literatures I read reported at least 2 products and the ratio of them depend on substituents on quinoline.
   In the paper you mentioned, they used N-chlorosuccinimide to form 3-chloro quinoline from non halogen quinoline derivatie, while they made 2-chloro derivatives from 2-quinolinol. Is this any other possible way to proceed directly from N-oxide compound to the 2-chloro derivative? I read from a paper in which the selectivity of chrorination of pyridine is improved by adding of base such as Et₃N. I attach here the paper you mentioned and the one I found.

Edit by Borek: please don't post copyrighted material on the forum. Attached papers deleted.

[TheUnassuming]

You sir have great journal access.  Apologies for the mistake.
Looking at the correct chlorination position it looks like you are using the more precedented method.  Use of a base like that might serve to mop up HCl, which could be the pathway that gives you the other products. 
With the method you are using currently it looks like your yields are a little low, but not unreasonable for the reaction. 
In this paper: Bioorganic & Medicinal Chemistry Letters Volume 23, Issue 5, 1 March 2013, Pages 1472–1476.  they report 30-50% yields for the sequence. 
It was also used in: ChemMedChem Volume 4, Issue 2, pages 249–260, February 13, 2009, but I don't have access to this paper to see the yields. 
Another route is through lithiation and reaction with a Cl electrophile like they did in this paper: Letters in Organic Chemistry (2010), 7, (1), 90-93.  The catch is I don't know if lithiation will occur faster than the lithium/halogen exchange on your substrate.  Ive done Li/Br and I exchanges before and they happen relatively quickly(<30min), but I have no idea how fast it would occur with your F. 

[TheUnassuming]

Very cool Syn. Comm. paper by the way. 

[orgopete]

The question is being asked as though the problem exists with step two. But, no indication was given about the N-oxide. Was it isolated?

[vu_nam]

@TheUnassuming: Thank you very much! You could find a lot of valuable sources. I'm still lack of experience.
   In ChemMedChem Volume 4, Issue 2, pages 249–260, they proceeded from N-oxide to quinolinone and in last step formed 2-chloro quinoline with 35% overall yield.
   Your mentioned report from Letters in Organic Chemistry (2010), 7, (1), 90-93 really impress me. Following this, I found Tetrahedron 62 (2006) 6166–6171 suggesting another possibility, although I'm not sure the flouro substituent in the adjacent ring could be affected or not. I may tried use base as promoting agent before testing other ways.
@orgopete: I used N-oxide as crude intermediate, but a single spot which had very low Rf could be noticed from TLC. My priority is to reduce the amount of side product in step 2.

[orgopete]

Purifying the N-oxide will reduce the by-products and improve the yield.

[TheUnassuming]

Even if you want to one-pot the reaction when you scale up, isolating the oxide from the first step at least once would be a good idea.  It tells you definitively what % of your 20% yield comes from the first step and what comes from the second step.  Also you can run the second reaction on the pure substrate you isolate to know for sure whether your side products are from your reaction conditions or from components carried on from your first step.