Kinetic (Hive Bee)
12-29-02 08:23
No 393849
      Proposed alternative to Friedel-Crafts acylation  Bookmark   

I've been thinking about something along these lines for a while now, but it was only this morning I realised that there is quite a good chance it would work.

The idea has been touched upon in TFSE, and the theory seems valid. Generally, to produce 'phenones' (eg propiophenone, acetophenone), Friedel-Crafts acylation is the best option, but with certain ring substituents (such as when trying to acylate benzodioxole with propionyl chloride and FeCl3 as the catalyst) the yields are terrible if anything at all. Because of this, I propose using a Grignard or Grignard-type reaction with the substituted phenylmagnesium halide [or diphenyl zinc?] and an acid chloride. Starting from benzodioxole and propionyl chloride, the reaction will hopefully look something like:

1. Brominate benzodioxole with N-Bromosuccinimide to form 4-Bromo-1,2-(methylenedioxy)benzene (this product is commercially available but ridiculously expensive).

2. Form the 4-Magnesiumbromo-1,2-(methylenedioxy)benzene Grignard in the usual way, or possibly make the zincorganic compound (adapting the procedure from Post 346231 (Antoncho: "Zealot: a new zincorganic synthesis of P2P's", Novel Discourse)) to give the less reactive Grignard-type organometallic.

3. React the Grignard with one equivalent of propionyl chloride to give the very handy 3,4-methylenedioxypropiophenone (Hopefuly to stir up some more interest, I'll mention that this can be easily made into safrole for bees that way inclined, see ../rhodium/chemistry /propiopropen.html)

The first step seems to have been carried out in good yield before, but the rest is all speculation; hopefully there is good reason for it to work the way I plan, but I still have a couple of questions:

For step three, would the Grignard be too reactive, and end up forming a mixture of desired product, tertiary alcohol, and unreacted starting material? Assuming one equivalent of acid chloride is used, and knowing that that acid chlorides are more electrophilic than ketones, and also the steric hindrance of adding a second methylenedioxyphenyl group to the formed ketone, hopefully the major product will be 3,4-methylenedioxypropiophenone. If the Grignard is too reactive, then the zinc compound could possibly be used.

Any other comments or suggestions will be most useful.

I haven’t done a literary search yet, but if anyone happens to have any useful references offhand on this type of reaction, I’d really appreciate it.
(Chief Bee)
12-29-02 09:06
No 393853
      Ketones from organozincs and organocadmiums  Bookmark   

Yes, you definitely want to form the organozinc (or possibly organocadmium) reagent instead of the plain grignard for sake of achieving good yields.

I have nothing handy right now on organozincs, but read this review of ketone synthesis with the aid of organocadmiums: ../rhodium/pdf /organocadmiums2ketones.pdf - The chemistry is relatively similar.
(Official Hive Translator)
12-29-02 21:40
No 394050
      Similarities and differences...  Bookmark   

The chemistry IS relatively similar, xcept the reactivity.

Cadmium cpds are the least reactive of all metallorganics (Cd<Zn<Mg<Li<K). You, e.g., CAN'T make P2P from Bz2Cd and Ac2O (as was recently found out by fallen_Angelsmile  (so i guess that the method on Rh's page should bee supplemented w/at least a warning).

But if using acyl chloride, that should bee an advantage rather than a drawback - ?

(Chief Bee)
12-29-02 22:07
No 394061
      P2P using alkylcadmiums (Bn2Cd + MeCOCl fails)  Bookmark   

Antoncho: Retrieving the actual references for that reaction myself, it seems like the original articles concerned the synthesis of 1,3-diphenyl-2-propanones from benzyl cadmiums and phenylacetyl chlorides... What kind of product was isolated by your comrade?

../rhodium/pdf /archive/bncocl-bn2cd-1.pdf
../rhodium/pdf /archive/bncocl-bn2cd-2.pdf

Would someone dare to react dimethylcadmium with phenylacetyl chloride to see if it works that way around? It most definitely works with phenylacetaldehyde, as in ../rhodium/chemistry /phenylacetone.html#phenylacetaldehyde - I have that from a trustworthy source.

(Hive Bee)
12-30-02 15:01
No 394289
      Back to Friedel-Crafts!  Bookmark   

Thanks for your replies; especially the organocadmium link which has saved me a lot of time and money before attempting the synthesis, which I now think may be doomed to failure. The document mentions that both organocadmium and organozinc reactions run much better in hydrocarbon solvents than they do in most ether solvents, due to ester formation between the acid chloride and the ether. Since benzodioxole is an ether, and possibly a more reactive one than most (it seems to be a lot easier to demethyleneate than other aromatic ethers are to demethyate; Shulgin acylates a number of aromatic ethers in high yield using AlCl3 as the catalyst, which would destroy benzodioxole), won’t ester formation between benzodioxole and propionyl chloride be a significant side reaction, enough to invalidate this route? So that may be another methylenedioxypropiophenone synthesis down the drain…

What I have just noticed though, and I can’t believe I missed the first time round, is the acylation of benzodioxole using tin tetrachloride as the catalyst - in Post 361935 (Rhodium: "Everything you need for methylone.", Novel Discourse), Rhodium mentions an article which gives 58% yield of methylenedioxypropiophenone, which is fantastic considering the terrible results I’ve had so far. Here is the abstract:

Reaction 16 of 34
Reaction ID 1651712
Reactant BRN 115506 benzo[1,3]dioxole
385632 propionyl chloride
Product BRN 165729 1-benzo[1,3]dioxol-5-yl-propan-1-one
No. of Reaction Details 1
Reaction Classification Preparation
Yield 58 percent (BRN=165729)
Reagent SnCl4
Solvent CH2Cl2
Time 1 hour(s)
Temperature 20 C
Ref. 1 5702207; Journal; Daukshas, V. K.; Gaidyalis, P. G.; Pyatrauskas, O. Yu.; Udrenaite, E. B.; Gasperavichene, G. A.; Raguotene, N. V.; PCJOAU; Pharm.Chem.J.(Engl.Transl.); EN; 21; 5; 1987; 341-345; KHFZAN; Khim.Farm.Zh.; RU; 21; 5; 1987; 569-573.

So would ‘Pharm.Chem.J’ be ‘Journal of Pharmaceutical Chemistry’, and contain an English translation of the entire article? I don’t know if I have access to this journal, but if someone could confirm the name I’ll have a look tomorrow if possible. If it's as simple as it looks, I suspect it will be the best route to methylenedioxypropiophenone from benzodioxole.
[Edit: ‘Pharm.Chem.J’ is an abbreviation of 'Pharmaceutical Chemistry Journal'. No wonder I couldn't find it... I'll have another look in the library soon, since it's only available online from 2000 onwards.]

Rhodium, I would love to attempt synthesising P2P from dimethyl cadmium and phenylacetylchloride, but due to certain factors beyond my control (3 in fact; the toxicity of cadmium and it’s salts, the legal status of the finished compound, and my incredible lack of money), I doubt I ever will. If I ever get the chance, I would try synthesing phenyl-2-butanone using diethyl zinc and phenylacetylchloride, just to see what would happen.
(Hive Bee)
12-31-02 11:59
No 394544
      Since I'm bored...  Bookmark   

I've been thinking of other routes to my chosen compound, since this was supposed to be an alternative to Friedel-Crafts, so for anyone who’s interested, here's another possibility. From an adaptation of ../rhodium/chemistry /phenylacetone.html#grignard:

1,2. Brominate benzodioxole as in my first post, then form the Grignard.

3. React this with an excess of propionitrile to form the intermediate imine.

4. Hydrolyse to give the desired ketone.

This method is used in ../rhodium/pdf /p2p.acetonitrile.pdf, with phenylmagnesium bromide and propionitrile, to give propiophenone in a reported 94% yield. The only data given for substituted phenylmagnesium bromide compares 3,5-dibromophenylmagnesium bromide and acetonitrile with phenylmagnesium bromide and acetonitrile; the respective yields of the corresponding acetophenones being 1% and 42% (with ether as solvent), or 52% and 68% (with benzene as solvent). The question is, what will happen if 3,4-methylenedioxyphenylmagnesium bromide is used with propionitrile? The ring is more activated than its unsubstituted counterpart, whereas the ring in 3,5-dibromophenylmagnesium bromide is deactivated.

Is the lower yield of 3,5-dibromoacetophenone caused by the ring deactivation (therefore increasing the likelihood of the activated methylenedioxyphenylmagnesium bromide giving good yields), or is it caused by some other side reaction, as the bromine atoms in the 3- and 5-positions could interfere in the reaction?