Publication of the week, number 50, 7th November 2014

This week another new use for older compounds from the Baran labs. The ASAP discusses the use of quinone diazides in steroid synthesis.

diaz01

Now these things have been around for quite some time and are simply made by diazotisation of the corresponding aminophenol. So here we have a carbene source, what is the best way to generate this particular one? Certainly rhodium catalysis is asked for and several were examined in a model reaction with cyclooctene. The best was [Rh2(esp)2] at a level of 5 mol% in dichloromethane producing the cyclopropane in reasonable yield, 66%. The catalyst structure is:

diaz03

and is readily available.

With a cyclopropane in hand the next step was to investigate the selective C-C bond cleavage process, as several products could result. The authors managed to find conditions which allowed this selective cyclopropane opening by bringing the “quinone part of the molecule into play”, for example:

diaz02

So far so good. In order to really make use of these carbene precursors a convenient precursor was required, one which allows the generation of the diazide under mild high yielding conditions, not at least when one is considering their use in the synthesis of steroid frameworks:

diaz04Steroids

From the authors” Since its resonance structure is an oxyaryldiazonium zwitterion (above), it was anticipated that methoxymethyl (MOM) and triazene groups would effectively shield the oxyanion and diazo species, respectively, and be easily unmasked under mildly acidic conditions. This masked quinone diazide was anticipated to arise from the union of triazenes with trans-bicyclic enones by conjugate addition followed by trapping with a one-carbon electrophile.” Through a convergent route the following triazene was synthesised:

diaz05

Treatment with TFA unmasked the diazide which was treated with the rhodium catalyst to give a high yield of the cyclopropane (drawn above). Reaction with lithium chloride produced the expected aromatic steroid system, compound 1 and hydrogenation gave the ring expanded system 2. Whose structures were confirmed by x-ray.

diaz06

From the authors”In conclusion, this work represents the first systematic study of quinone diazides, a ubiquitous entity in materials science, in the context of organic synthesis. The most significant findings resulting from this inquiry include: 1) a general rhodium-catalyzed intermolecular cyclopropanation of quinone diazides with a variety of different olefin classes, 2) divergent manipulations of the resulting cyclohexadienone adducts, 3) a strategy to enlist quinone diazides in complex intramolecular settings using a triazene-enabled “mask”, and 4) application to a fundamentally different approach to the steroid skeleton.”

So there we have it another wonderful bit of chemistry from the Baran labs. Now I’m not an expert in steroid synthesis, but I guess that this approach may have had Fieser jumping in his chair. The approach that Baran and his group have to their chemistry I find very refreshing and it is always a joy to read the ensuing publications.

4,883 total views, 1 views today

Prof. dangerdackel (199 Posts)


4 thoughts on “Publication of the week, number 50, 7th November 2014

  1. I have used [Rh2(esp)2] for cyclopropanations with N2CHCO2Et. The catalyst is pricey not only because of Rh(II) but also the ligand synthesis is multistep, and it is patented and exclusively licensed. What I like about [Rh2(esp)2] is its great robustness and good solubility which allow very low catalyst loadings (0.1%) and it works better also for all-in neat conditions (whereas in many cyclopropanations the diazo compound is typically syringe-pumped into the reaction. This makes the cyclopropanations very convenient – set up and leave on a melting ice bath over a weekend…

  2. Thanks for the post. I didn’t appreciate this article from just skimming – no flashy yields and little advertisement of the interesting product selectivity.

    1. Thanks for the comment. What have flashy yields got to do with it? Where is it written that yields should always be flashy (by that I take it you mean >99%?).

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.