[Dan]

I am trying to prepare 2,6-cycloheptadienone by dehydrogenation of hycloheptanone following the IBX dehydrogenation methodology reported by the Nicolaou group:

JACS, 2000, 7596: http://pubs.acs.org/doi/pdf/10.1021/ja001825b
JACS, 2002, 2245: http://pubs.acs.org/doi/pdf/10.1021/ja012127%2B

In these papers, they report this exact reaction with yields around 80% (but only a general prep is given).

In my hands this appears to be completely irreproducible, and the cause of the problem is unclear. The IBX was prepared in-house (Oxone oxidation of 2-iodobenzoic acid) and is clean by NMR. Inorganic impurities are unlikely - it was washed thoroughly with water, acetone and ether.

Over the first few hours at 80°C in DMSO or 2:1 toluene/DMSO (as reported), TLC looks ok. Starting material disappears and I see the formation of two new spots - one of these is the desired dienone (confirmed by NMR etc, isolated in 4% yield from one run), the other I assume is intermediate 2-cycloheptenone.

Following stirring overnight (literature reaction time is 15-24 h) the mixture turns dark brown, TLC shows mainly starting material and aqueous workup provides a crude mixture that is mostly (~80% by NMR) starting material with a crude yield around 25%. No product is observed by TLC in the aqueous washings.

To me this points towards product decomposition. I am going to run some experiments at lower temperature, but I was wondering if anyone has experience with this reaction and might have some tips. Is anyone aware of any water sensitivity? The DMSO I am using is not fresh and therefore we can assume it is wet, and so far I have run these reactions under air (I do not normally exclude oxygen for oxidation reactions).

Any help is appreciated, other synthetic routes (which I am exploring) are substantially longer.

[discodermolide]

I am trying to prepare 2,6-cycloheptadienone by dehydrogenation of hycloheptanone following the IBX dehydrogenation methodology reported by the Nicolau group:

JACS, 2000, 7596: http://pubs.acs.org/doi/pdf/10.1021/ja001825b
JACS, 2002, 2245: http://pubs.acs.org/doi/pdf/10.1021/ja012127%2B

In these papers, they report this exact reaction with yields around 80% (but only a general prep is given).

In my hands this appears to be completely irreproducible, and the cause of the problem is unclear. The IBX was prepared in-house (Oxone oxidation of 2-iodobenzoic acid) and is clean by NMR. Inorganic impurities are unlikely - it was washed thoroughly with water, acetone and ether.

Over the first few hours at 80°C in DMSO or 2:1 toluene/DMSO (as reported), TLC looks ok. Starting material disappears and I see the formation of two new spots - one of these is the desired dienone (confirmed by NMR etc, isolated in 4% yield from one run), the other I assume is intermediate 2-cycloheptenone.

Following stirring overnight (literature reaction time is 15-24 h) the mixture turns dark brown, TLC shows mainly starting material and aqueous workup provides a crude mixture that is mostly (~80% by NMR) starting material with a crude yield around 25%. No product is observed by TLC in the aqueous washings.

To me this points towards product decomposition. I am going to run some experiments at lower temperature, but I was wondering if anyone has experience with this reaction and might have some tips. Is anyone aware of any water sensitivity? The DMSO I am using is not fresh and therefore we can assume it is wet, and so far I have run these reactions under air (I do not normally exclude oxygen for oxidation reactions).

Any help is appreciated, other synthetic routes (which I am exploring) are substantially longer.

If you are getting mostly SM back, what is the other spot on TLC?, some sort of "complex" between IBX and SM which decomposes on work-up?
I would try at a lower temperature or don't leave it overnight. If you have to leave it ON stick it in the freezer and carry on the next day.

[Doc Oc]

Can't help you with your issue, only want to add that I also was unable to reproduce some of the chemistry reported in these papers.  I ended up having to find a different route because nothing worked.

[Dan]

If you are getting mostly SM back, what is the other spot on TLC?, some sort of "complex" between IBX and SM which decomposes on work-up?

The TLC analysis is pre-workup. To clarify:

I have 3 spots:

A: cycloheptanone (starting material)
B: 2-cycloheptenone (intermediate)
C: 2,6-cycloheptadienone (desired)

After ~6 h, the ratio by TLC looks something like A:B:C 1:2:3 (qualitative)
After 15 h, the ratio by TLC looks something like A:B:C 8:1:1 (confirmed by crude NMR)
Due to the low crude yield (~25% total), I am inclined to suspect that B and C are degrading under the reaction conditions (and not that A is regenerated). The remaining mass balance is in the aqueous washings, very polar and intractable.

I also was unable to reproduce some of the chemistry reported in these papers.

Yeah, I've been asking around and a few people have had issues with this too (or know someone who has). I'm ploughing through some classical chemistry for the 70s on the side. Seems to be working quite well.

I'm going to have a shot at lower T and exclude air/water and see if it helps.

[Dan]

Found a follow-up publication for a protocol at RT, which I'm going to try.

ACIE 2002, 993 Link

It is noted in this paper that anhydrous DMSO is detrimental, so the problem is not water. It may be light though - I have not been covering my flasks and IBX is light-sensitive (though it is not very sunny here and these reactions have been running overnight...).

I will update tomorrow in case the information is useful to anyone.

[discodermolide]

Found a follow-up publication for a protocol at RT, which I'm going to try.

ACIE 2002, 993 Link

It is noted in this paper that anhydrous DMSO is detrimental, so the problem is not water. It may be light though - I have not been covering my flasks and IBX is light-sensitive (though it is not very sunny here and these reactions have been running overnight...).

I will update tomorrow in case the information is useful to anyone.

On the wiki page http://en.wikipedia.org/wiki/2-Iodoxybenzoic_acid
in the glycol cleavage part, the mechanism actually suggests that DMSO interacts with the IBX!
Maybe better to avoid DMSO:

[Dan]

DMSO should be fine - it is an IBX-DMSO complex that is proposed to be the active oxidant. I am currently looking into alternative ligands that provide more reactive complexes that should do this reaction at RT (namely IBX-[methoxypyridine-N-oxide], see ACIE 2002, 993).

IBX its self is like brick dust and doesn't dissolve in much else.

[OC pro]

It seems God´s methodology is not working. Nothing new to me. I would not spend too much time with this one.  You should give this N-oxide a try but then I would give up on this.
Could you just do a bromination in positions 2 and 6 and then elimination to the corresponding dienone?

[Dan]

Could you just do a bromination in positions 2 and 6 and then elimination to the corresponding dienone?

I'm doing this as well. You need to proceed via a ketal though, otherwise you get side reactions through Favorskii pathways.
JOC 1979, 4285 Link
Sequence is (i) ethylene glycol protection; (ii) bromination; (iii) elimination; (iv) hydrolysis.

So you can see why I was keen to have a high yielding 1-step procedure...

[g-bones]

IBX and DMP tend to work best with partial hydrolysis (which I think someone mentioned).  I agree it may not be worth figuring this stuff out.  If you need the material you may have to bite the bullet and give a Reich oxidation or even Saeguesa-Ito a try.  Just enough so you can do whatever forward chemistry you want with it.  I agree that alpha bromination and elimination is also worth a shot and as long as you are careful about your base (k-carb or DBU or something) you should be alright. 

Also here is a pretty sweet one step Saegusa-Ito oxidation without the silyl enol ether intermediate:

http://pubs.acs.org/doi/abs/10.1021/ja206575j
   

[OC pro]

The α,α'-dibromination and subsequent elimination has been used extensively in steroid synthesis and it works well. Have also seen LiBr in DMF for elimination. I don´t think ketalization is necessary. Even with low yields, starting material is cheap so why wasting time on this chemistry. Bromination and then elimination without protection/deprotection.

[Dan]

g-bones: Thanks for the link I already saw that one and saw that:

"Cycloheptanone and cyclooctanone led to a mixture of dehydrogenation products, with 2,6-cycloheptadien-1-one and 2,7-cyclooctadien-1-one formed as the major products in 26 and 25% yields, respectively, based on GC
MS and 1H NMR spectroscopic analysis."

I didn't really fancy trying to re-optimise it. Then again, 25% is much better than what I see with IBX anyway.

Doc Oc: I'm not spending much time on the IBX chemistry, just a TLC every few hours. Incidentally, I observe no reaction of cycloheptanone with 4.5 eq IBX-MPO complex at RT in DMSO over 16 h. Warming up now.

I am trying the bromination route in parallel (following a lit prep via the ketal), and it looks good so far. I may try later without ketalisation if I need more material.

[GS]

I am trying to prepare 2,6-cycloheptadienone by dehydrogenation of hycloheptanone following the IBX dehydrogenation methodology reported by the Nicolaou group:

JACS, 2000, 7596: http://pubs.acs.org/doi/pdf/10.1021/ja001825b
JACS, 2002, 2245: http://pubs.acs.org/doi/pdf/10.1021/ja012127%2B

In these papers, they report this exact reaction with yields around 80% (but only a general prep is given).

In my hands this appears to be completely irreproducible, and the cause of the problem is unclear. The IBX was prepared in-house (Oxone oxidation of 2-iodobenzoic acid) and is clean by NMR. Inorganic impurities are unlikely - it was washed thoroughly with water, acetone and ether.

Over the first few hours at 80°C in DMSO or 2:1 toluene/DMSO (as reported), TLC looks ok. Starting material disappears and I see the formation of two new spots - one of these is the desired dienone (confirmed by NMR etc, isolated in 4% yield from one run), the other I assume is intermediate 2-cycloheptenone.

Following stirring overnight (literature reaction time is 15-24 h) the mixture turns dark brown, TLC shows mainly starting material and aqueous workup provides a crude mixture that is mostly (~80% by NMR) starting material with a crude yield around 25%. No product is observed by TLC in the aqueous washings.

To me this points towards product decomposition. I am going to run some experiments at lower temperature, but I was wondering if anyone has experience with this reaction and might have some tips. Is anyone aware of any water sensitivity? The DMSO I am using is not fresh and therefore we can assume it is wet, and so far I have run these reactions under air (I do not normally exclude oxygen for oxidation reactions).

Any help is appreciated, other synthetic routes (which I am exploring) are substantially longer.

Years ago at another employer I was dealing with precisely the same synthetic problem.  In my hands, IBX reaction was not reproducible, and was applicable only on the microscale anyway.  On another hand, the older reaction sequence [ethyleneketal, bromination, dehydrobromination, hydrolysis of ketal] proceed very smoothly, with decent yields, and was scale-limited only by the available glassware. Considering that IBX preparation also takes time, my vote goes for the good old 1965 JORC p. 2109 prep.

[Dan]

Thanks for the paper. I ended up making my own modifications and telescoping that classical sequence. It's pretty good (~50% overall from 10 g cycloheptanone).

I never got the IBX preps to work properly. I did get quite nice conversion with the IBX-MPO complex at 60°C for about 24 h, but the result was not very reproducible and did not perform well on more than about 50 mg.

[GS]

Thanks for the paper. I ended up making my own modifications and telescoping that classical sequence. It's pretty good (~50% overall from 10 g cycloheptanone).

I never got the IBX preps to work properly. I did get quite nice conversion with the IBX-MPO complex at 60°C for about 24 h, but the result was not very reproducible and did not perform well on more than about 50 mg.

It worked at about 45-50% overall even on the 250g scale, if memory serves.