12-21-03 18:22
No 478307
Synthesis of a (mostly) OTC ketamine/PCP analog
Happy holidays, fellow bees! Here is my gift to you--another crackpot synthesis idea! This time our target is a tasty-looking hybrid of ketamine and PCP that happens to be made out of OTC ingredients. How fun!
First, the molecule:
Molecule: our analog ("Clc1ccccc1C2(NC)CCCCC2=O.c1ccccc1C2(NC
On the top left is ketamine, on the bottom left is PCE, a very active analog of PCP. Our analog is on the right. Notice any resemblance?
Design Rationale:
1. Like ketamine, the cyclohexane ring has a 2-carbonyl group.
By taking advantage of this group, the whole synthesis becomes possible. We need it. SAR studies show 2-substituents on the cyclohexane ring don't cause much trouble. Hopefully, some of ketamine's quality (as opposed to pure potency) is attributable to this substituent, perhaps lending our analog similar tastiness in bioassay.
2. Unlike ketamine, benzene ring has no 2-chloro group.
It seems unnecessary, based on the potency of PCP and its many analogs. However, if you wish to buy precursors rather than making them, it may bee wise to include that 2-chloro group, for reasons soon made clear.
2. Isopropyl amino group.
According to the PCP SAR document, N-isopropyl group makes for a more potent and tasty (less sedating) compound than N-methyl or N-ethyl. It is also much easier to make without overalkylation by performing a reductive alkylation on the primary amine precursor with acetone.
Synthesis:
Step I
Starting from benzyl cyanide, alpha-acetyltolunitrile is prepared with NaOEt and EtOAc according to the instructions of Org Syn CV2, 487 (a-phenylacetoactetonitrile) (http://www.orgsyn.org/orgsyn/prep.asp?p
Molecule: Step I ("c1ccccc1CC#N.CC(=O)OCC>>c1ccccc1
Step II
Allyl bromide (http://www.orgsyn.org/orgsyn/prep.asp?p
Molecule: Step II ("c1ccccc1C(C#N)C(=O)C.BrCC=C>>c1c
Step III
The tricky one. A radical-chain ring closure is performed using MnO2/HOAc or Mn(OAc)2/KMnO4 systems. This is why a 2-carbonyl group is required on the cyclohexane ring--it provides the electron-withdrawing group necessary for this addition to work.
Molecule: Step III ("c1ccccc1C(C#N)(C(=O)C)CC=C>>c1cc
We now have a 2-substituted phenylcyclohexylcarbonitrile, and the next two steps can be stolen directly out of Rhodium's PCP document (../rhodium /pcp/pc
Step IV
The nitrile is partially hydrolyzed to a urethan in ethanolic HCl, then fully hydrolyzed to an amide in H2O/HCl.
Molecule: Step IV ("c1ccccc1C2(C#N)C(=O)CCCC2>>c1ccc
Step V
The amide is converted to an amine via Hoffmann rearrangement, using NaOCl (bleach).
Molecule: Step V ("c1ccccc1C2(C(=O)N)C(=O)CCCC2>>c1
Step VI
Now, we have the problem of N-alkylation to overcome. The most OTC and pharmacologically promising thing I could come up with was reductive alkylation with acetone to an N-isopropyl compound. Like in amphetamines, red. amination avoids the overalkylation and corresponding reduction of yield that occurs with use of alkyl halides. Use your favorite method (Raney nickel or Zn/Cu couple should serve nicely).
Molecule: Step VI ("c1ccccc1C2(N)C(=O)CCCC2.CC(=O)C>>
Discussion:
First, this method (should it work) would provide the most OTC pathway to ketamine analogs I know of. All the precursors can be made from scratch if necessary, and since some of the route is shared with an established method of P2P synthesis, while another section is common to the PCP via nitrile intermediate route, there are few steps not well covered already. The major advantages over PCP synth are 1) that with the combined activation of the nitrile and carbonyl groups super-strong bases are not required to alkylate the phenyl alpha-carbon, and 2) allyl bromide is much easier to make than 1,5 dibromopentane, which seems to bee the sticking point in OTC pcp. Benzyl cyanide can be made with NaCN and benzyl chloride, which in turn is made by chlorinating toluene. (It is because benzyl cyanide is such a well known amphetamine precursor that I recommended that only the 2-chloro version be bought) Similarly, allyl bromide can be made in fairly high yield from allyl alcohol, made in turn with oxalic acid and glycerin. There is a most excellent thread on (safe?) home cyanide production here also.
Secondly, this synth does not require sensitive organometallics, making it easier to run with more minimal apparatus. Steps I and II might be performed in the same pot, as they both require NaOEt. Perhaps their order might even be reversed to raise yield in the Claisen condensation phase.
Thirdly, this compound has been described in the literature, along with a HUGE set of others, in a real whopper of a paper JOC 1966, 2593, where our compound is (IIIe). It has a sister paper on page 2601, and combined they synthesize a huge number of ketamine analogs. Oh, what this researcher wouldn't do to find their counterparts in J. Med. Chem. or J. Pharm. Pharmacol., where they would describe the results of testing/tasting. Sadly, I do not know if such a paper exists.
Of the two steps not already part of hive lore, <II> and <III>, <III> is the trickiest one. It would be a simple conjugate addition, except the allyl's double bond is not conjugated. Turns out there are some tricks out there involving radical chains that can be used to add a non-conjugated double bond to the alpha-carbon of a ketone. Since this is the linchpin step, refs are provided to support it:
First, a classic paper.
J. Org. Chem. 1996, 7832. Mn(III)-Based Oxidative Free-Radical Cyclizations of Unsaturated Ketones, by Cole, Han, and Snider.
They use Mn(OAc)3 with Cu(II) to do many cyclizations in this paper, with much more congested targets than ours. Yields range from "That Sucks, dude" to "Cool!". A seminal paper, and one that attempts much more than we need to do.
Next, we have in-situ Mn(III) generation with KMnO4.
Tetrahedron 1995, 9917. Potassium Permanganate-Mediated Radical Reactions: Chemoselective Addition of Acetone to Olefins, by Linker, Kersten, and Linker. (husband-wife team?)
Yields by their protocol range from crappy to great also. One of the major side reactions is the adding of an acetate to the final molecule. No doubt a hydrolysis followed by dehydration would give an unsaturated compound that would be saturated in the final reductive amination step.
I quote:
A mixture of potassium acetate (1.75 g, 18 mmol), manganese(II) acetate tetrahydrate (12.5 mg, 0.05 mmol), acetone (22 ml, 300 mmol) and glacial acetic acid (18 ml) was heated to 70 C under argon atmosphere. After the addition of the olefin (5 mmol), solid potassium permanganate (0.3-0.4 equiv) was added in very small portions at 70 C over a period of 4-10 h until TLC indicated complete conversion of the olefin. The mixture was cooled to room temperature, diluted with water (100 ml) and extracted with DCM (3 x 100 ml). The combined organic extracts were washed with saturated bicarbonate solution (2 x 100 ml), water (100 ml), dried (Na2SO4), and concentrated. [...]
Another paper, in which they perform a two-step procedure very similar to steps II and III, offers some insight.
Tetrahedron Letters 2003, 659. Synthesis of the core bicylic system of hyperforin and nemorosone, by Kraus, Nguyen, and Jeon.
This has to been seen, so here:
Molecule: hello, there... ("O=C1C(C(=O)OCC)CCCC1.BrCC=C>>O=C
1. NaH, Allyl Bromide
2. 2 Mn(OAc)3, Cu(OAc)2
Together yield was 56%.
A similar run, with a phenyl instead of ethyl ester gave 60% in the first step, and 76% in the second, suggesting that the yield-limiting factor was in fact the alkylation step. In our compound, the nitrile group should be a much stronger activator than the dual carboxyls used in this synth, so it may well yield better without such a strong base as NaH.
In the paper
Tetrahedron Letters 1998, 3721. Electronic and Steric Effects of Various Silyl Groups in Radical Addition Reactions, by Hwu, King, Wu, and Hakimelahi
it is shown that radical addition/cyclization can be performed with MnO2 and a titch of acetic acid, using the MnO2 as a radical initiator/perpetuator. Though they investigate the conjugational effects of silyl groups, it appears that they are not necessary for the reaction to work. Addition of trimethylallylsilane (the least sterically hindered) to acetone proceeded in 81% yield. They reference a previous paper in JOC
1995, 2448 for the details of the MnO2 process.
Lastly, alkylation of acetals by alkenes is done in moderate yields "the old fashioned way"--i.e., with UV excitation, sensitized by acetone.
Journal of Organic Chemistry 1967, 805.
Hope this made someone's christmas just a bit merrier. I will bee posting some info on step <II> later, as well as anything else I can find relevant to step <III> after chasing down some of the refs in those papers. Ning out.
Catching a buzz @ the Hive
(Newbee)
12-22-03 05:00
No 478432
Ning, don't you think you forgot to draw a double bond on the product of the scheme "Step III" like there is in the "hello, there..." scheme? I can't imagine a radical cyclisation like Step III without forming either a double bond or a substituent (from the solvent: AcO, MeO...) on the position 4 of the cyclohexanone ring. That double bond quite complicates everything.
The conditions needed for the Hoffmann rearrangement are very non-friendly to ketones. A ketal type of protection would probably be needed and still it would not necessarilly lead to your amine. Some amides just don't work with Hoffmann, but I don't know if this is the case.
Have you checked the synth of ketamine? It is some wird rearrangement and actually it looks easier on the paper than your proposed synth does, though you would need more hard-to-get chemicals.
You did a nice research, but it is far from being a full proof synthesis and it won't be such until somebody checks it.
“The real drug-problem is that we need more and better drugs.” – J. Ott
(acetaminophanatic)
12-22-03 22:50
No 478617
The ketone point you raised worries me a little bit, I will check it out.
As for the double bond, no, it is not necesary. Or I should say, it isn't necessarily necessary. They used an oxidative cyclization, because they actually wanted a double bond there. On the other hand, a normal radical cyclization would not have it. I say it doesn't matter because of the reductive amination at the end, which should reduce all the non-aromatic double bonds. But don't take my word for it, look at the ref I posted where they perform that non-oxidative radical cyclization. If that doesn't make you happy, then when I get the chance, I'll post more refs.
Also, while I don't know too much about the hoffmann rearrangement, I do know that normal sodium hypochlorite oxidation of alcohols stops at the ketone/aldehyde stage, and I have several papers on bleach chemistry (gathered for the quinone project) that show that to be the case under both acidic and alkaline conditions. If you have any papers that show this to be a problem, I would like you to post or msg me the refs.
EDIT:
By the way, if you don't like radical additions in the main path, you can always take your crude allyl bromide and do an anti-markovnikov HBr addition with a little benzoyl peroxide or something to get 1,3 dibromopropane, then perform the cyclization in one step.
Catching a buzz @ the Hive
(Newbee)
12-23-03 01:49
No 478635
I can't see the mechanism going on in that radical cyclisation. The only one I could understand would leave a double bond or 4-substituent but if that paper says otherwise I will take your word for it.
From what I remember (I might be wrong) the conditions of alcohol oxidation to aldehyde/ketone goes on in just slightly acidic conditions (acetic acid) while the Hofmann requires strongly basic conditions (excess of NaOH) – that is the same as for the ‘haloform’ reaction (chlorine does not limit itself to methylketones like iodine!). But you may be right, if there is no excess of the hypochlorite maybe just Hoffman would happen. It would be necessary to find some refs for such alpha-amino-ketones made with a Hoffman rearrangement. They might be also useful for cathinones synth.
“The real drug-problem is that we need more and better drugs.” – J. Ott