Synthetic Endgame

Inspired by this paper from Melanie Sanford's rocking organometallic group at Michigan.

--

*Pun contest! I could also have called this post...

Gotta Make 'Em All

Poke-Ball and Stick Models

Chantix Charmander

Putting the Fun in C-H FUNctionalization

'All Thumbs' Synthesis

Encore, Encore!

Two hot areas of research served up second helpings online this week:

C-H Azidation: Remember John Hartwig's iron-meets hypervalent iodide combination from last March? It possessed the power to insert a late-stage amine equivalent into complex natural products. John Groves has raised the stakes, disclosing a "practical and complementary" Mn-porphyrin promoted version that takes solid sodium azide as the precursor. 

Source: Groves, JACS ASAP

The group finds it can enable late-stage azidation of a variety of complex bioactive substances (sclareolide, artemisinin, estrone, papaverine). Even more surprisingly, although likely a radical-induced transformation, using a chiral salen led to a single example of 70% ee material. Groves admits they have work to do, but the fact that this reaction operates with 1% loading in wet ethyl acetate at room temperature sure sounds promising!

Synthesis Machines: Over at Nature, Kobayashi published a flow reactor approach to syntheses of either enantiomer of rolipram, an anti-inflammatory. No MIDA-boronate 'handles' here; this is classic chemistry - olefination, 1,4 addition, reduction, hydrolysis, decarboxylation, cyclization - performed over heterogeneous catalyst beds encased in stainless steel tubes. The group spices up the synthesis by including their in-house chiral PyBOX-calcium catalyst to control the 1,4 addition, and developing a Pd / polysilane-catalyzed reduction for a troublesome nitro group. 

Kobayashi claims his synthetic engine can produce a gram of 96% ee material every 24 hours, and that the system remains stable and operable for about a week's time. In a complementary Commentary, Joel Hawkins of Pfizer presents a tantalizing future, where hood-sized continuous synthesis units chug through kilo quantities of drug precursors, using commercial reagents, sans column chromatography.

Oxidase Toolkit: C-H Azidation

Do you ever stare at your late-stage molecules, thinking "They're almost perfect, but I really wish I could add an amine right over there." Thanks to a new reaction, you might soon be able to.

Reporting in Nature, John Hartwig and coworkers have cracked the case: a mixture of iron (II), a tridentate nitrogen ligand, and a modified Togni reagent Zhdankin reagent reliably functionalize tertiary C-H bonds with an azide(N₃ group). The selectivity, yield, and mild conditions match pretty well with White's C-H oxidation, which utilized a similar catalytic manifold.

Hartwig's initial targets for this new reaction include two modified steroids and a gibberellic acid derivative. Sadly, precious few heteroatoms exist in these molecules to gum up the ironworks, but I'm certain they'll address that in the full paper. I'd especially like to point readers to Figure 3, in which the group shows subsequent transformations: heterocycle formation, amine reduction, chemical ligation, and capping with fluorescent tags.

These two reactions together, along with a variety of C-H halogenations and sulfidations, seem to support the growing "oxidase phase" approach to total synthesis. One could imagine that, in a few years, a naked carbon scaffold could be suitably decorated with O, N, S, or X at positions of the scientists' choosing. Wow.

Greener Nylon Synth? Just Add UV and Ozone!

"Any sufficiently-developed technology is indistinguishable from magic" - Arthur C. Clarke

Looks like we'll soon have a more straightforward way to make stockings, zip-ties, and tire belts. 

Adipic acid, a six-carbon diacid representing one of the "sixes" in Nylon 6-6, apparently takes quite a bit of industrial "elbow grease" to make. The current process, starting from cyclohexane, requires cobalt, manganese, copper, and vanadate salts, high pressures of oxygen gas, and hot nitric acid. Out the other side, its responsible for 5-8% of the nitrous oxide we humans spew into the atmosphere each year.

From Science 2014, Hwang and Sagadevan

Now, researchers Hwang and Sagadevan (National Tsing Hua University, Taiwan) believe they have a better method. Reporting in this week's Science, the two disclose a method that sounds so much simpler: flush a sample of cyclohexane with ozone and UV light, and, presto! Solid adipic acid at the bottom of your reactor. No metal salts, no nitrous oxide, no high pressures or temperatures.

Wow, that looks a lot simpler.

The researchers note that zapping ozone produces both singlet oxygen, ¹O₂, and a single singlet oxygen atom O(¹D). The highly reactive single singlet (say that three times fast!) can easily insert into C-H bonds, and, since it seems to prefer insertion next to an already-oxidized carbon, the diol, diketone, and finally diacid products are formed preferentially.

Applause, please: Look at this beautiful pictorial SI!
Twice, in two days.

Just for fun, Hwang and Sagadevan crack open some larger hydrocarbons, and check the selectivity of alkyl-functionalized rings and aromatics. There are tantalizing possibilities here that I'm sure, given the ease of this reaction setup, most organic chemists will already be trying: how do complex natural products* react under these conditions? If anyone tries it this weekend, please drop me a line.

--

*For that matter, I wonder if this pathway is operative in human tissues under physiological conditions? Sunlight does have some 300 nm band, and we certainly come into contact with ozone out in the wide world. Hmm.

Scalable Ingenol? Phil Strikes Again!

Update: Want the inside scoop? Check out Open Flask!

I'm officially declaring it: Every 6-7 months, we should expect another huge molecule to fall to Phil & Co:

May 2011: Cortistatin
November 2011: Taxane Cores
May 2012: Ten Meroterpenoids.
December 2012: Ouabagenin

July 2013: Seen the latest* over at ScienceExpress? I think this scheme sums the whole thing up quite nicely:

And that's why it's in Science, kids...
Source: Baran Group | ScienceExpress

Ingenol falls! LEO Pharma, in collaboration with Scripps, may soon make gram-scale batches of ingenol analogs - something that used to take entire groups years to make. This paper cheers from so many different bleachers, I can't even count 'em all:

Total synthesis accesses trace plant metabolite!

Investment in basic research reaps huge Pharma dividends!

Imitating nature makes stitching together complex terpenes look easy!

Enzymes, Schmenzymes...

This paper really does have something for everyone. A volatile intermediate gumming up the works. A surprise crystallization. X-Ray structures. Some allenic Pauson-Khand reactions. A low-temp vinylogous pinacol rearrangement. Even some C-H activation / oxidation tossed in at the end.

If you want some more ingenol goodies, head on over to Chemistry World's fantastic write-up.
And, of course, join me on PhilWatch somewhere around January 2014...

*Thanks again to Brandon for a copy.

New Tricks for Old Reagents: Oxygen Everywhere!

Passed a time, not too long ago, when if you wanted to oxygenate a selected C-C or C-H bond, you had to jump through several hoops: Over-oxidize (read: DESTROY!) then reduce again. Convert it to another functional group first, then use an expensive catalyst. Use toxic heavy metals (Cr, Hg, Pb, anyone?) in their highest oxidation states...and, oh yeah, heat the heck out of it.

The past decade has seen kinder, gentler oxidations emerge in rapid succession. Cobalt. P-450s. Iron. Now, two recent papers bring new wrinkles to the oxygenation of organic molecules in unexpected ways.

The first, from the Concellon / del Amo group in Org. Lett., relates a neat trick performed by Oxone, usually a reagent reserved to make other oxidants.

The researchers deal with their serendipitous discovery with humility and class:

"This work was not originally intended..[but]...was worth studying. [We] remark that Oxone is a crystalline solid oxidant, easy to handle, non-toxic...and, above all, stable and cheap."

All great reasons to run these reactions, which are formally derivatives of the classic Baeyer-Villiger reaction. They blast through a brief substrate table (26 entries, 33-95% yields), and seem pretty excited about investigating the mechanism.

The second reaction, hot off the Nature presses, involves another legacy reagent: phthaloyl peroxide. I suspect the Siegel group was looking for sp3 C-H activation conditions, but instead discovered a serendipitous site-selective arene activation, reliably producing phenols.

The reaction works across a broad functional group palette - azides, silyl groups, boronate esters, primary halogens - that other oxidants would tear apart. They ultimately do about 50 substrates, including 3 natural product-like scaffolds, with yields ranging from 45-95%.

Deciphering the mechanism requires Ken Houk's computational super-powers. The researchers discover a "reverse-rebound" mechanism operates, meaning an oxygen radical from phthaloyl peroxide adds into the ring, the electron bounces around in the pi cloud, and then ejects the ipso hydrogen in a two-step process. Interestingly, other radical oxygen oxidants (di-benzoyl peroxide) led to primarily sp3 oxidation, showing that the structure of the radical precursor plays a big role here.

Let the Sun Shine In

Have you ever wondered why so many chemicals are shipped in amber-tinted bottles? Turns out light can do some pretty sneaky unexpected chemistry.

Reporting in from scenic Switzerland, Gademann & Co recently disclosed a sweet protecting-group free synthesis of Taiwaniaquinol F, a 6-5-6 diterpenoid with "potent cytotoxicity" against certain cancer cell lines. They finish the target in 17 steps, highlighting a Wolff ring-contraction and a neato halogen-oxygen exchange. 

The authors note that the final isolation had to strictly exclude light, or else their product slowly decomposed. Luckily, they asked themselves "Decomposed...to what?"

Source: Org. Lett. ASAP | Gademann group

Not to keep you in the dark; their product underwent a spontaneous 1,5 C-H remote functionalization. When the group tried to do it "on purpose" - placing a flask in the sun for 15 minutes - they were rewarded with a 30% yield of cyclized Taiwaniaquinol A.

The authors aren't quite sure what's so special about this system, but note a tantalizing biosynthetic possibility: since Taiwaniaquinol F comes from the bark of a certain tree, and "A" from its leaves, there's a real possibility that this process happens all the time in the plant! No enzymes, no metals, no bases. Just sunshine. 

Nature's amazing.

Pd Bites Back! Isopropyl Surprise

Although we don't always admit it, we chemists love a good surprise.

That's the feeling I got reading through a recent JACS ASAP, from the Chen group at PSU. They've been playing with an easy-to-remove C-H activation auxiliary (PA) first developed by Daugulis back in 2005. They've taught it some pretty neat tricks thus far: C-H amination, alkoxylation, and even alkylation reactions using simple primary iodides.

Their latest extends the chemistry to methyl iodide, which, when combined with some silver salts and a phosphate additive, usually plunks a new CH3 on an unhindered, kinetically-accessible gamma-methyl (i.e. the chain grows by one). However, in the case of norbornene, something wild happens: the Pd catalyst activates a secondary C-H bond, sticks a methyl on, but doesn't stop there; it "bites" into the newly formed methyl and adds on two more, to produce an isopropyl group. Sweet!

     
       Source: JACS ASAP 2013 | Chen group, PSU

Chen's group shows that the position of the new i-Pr depends upon exo ("up") or endo ("down") relationships of the initial amine, but doesn't speculate much on mechanism.

O Norbornene Tree, O Norbornene Tree;
I do not grok your sterics.

Well, allow me! I wonder what happens if you keep feeding iodomethane into the system. I'm sure there's a limit (likely steric) that prohibits additional branching past a certain point, but currently nothing completely rules out a dendrimeric "norbornene tree" compound, right?

Readers, can someone disabuse me of this crazy notion?

Cyclophanes Fall Apart for Rhodium

When chemists design syntheses, they usually think in terms of building up, not tearing down. We alkylate, we esterify, we cyclize, adding always more molecular weight, joining synthons together. Occasionally, bonds must be broken. Considering most protecting groups, severing C-N, C-O, and C-X bonds keeps chemists of all stripes pretty busy. What about C-C scission? Not so much.

Sure, there's your decarboxylations, your ozonolyses, your samarium iodide reductions. Long ago, natural enzymes figured out how to push electron density around to slice up substrates: think about gramine fragmentation, Poitier oxidation, or the (now-defunct) Pictet-Spengler spiro mechanism. But you don't often see fully saturated sp3 carbon-carbon bonds falling apart; after all, we'd be dealing biology a mortal blow with such precarious engineering. 

Wouldn't it be cool, though, if we could just add a specific catalyst, a dash of water, and selectively crack a hydrocarbon?

I don't recall seeing this graphic in the SI...but I can hope.

Turns out you can...if the system's just right. Chinese University of Hong Kong chemistry professor Kin Shing Chan, with coworkers Ching Tat To and Kwong Shing Choi, reported just such a reaction in JACS ASAP yesterday. The scientists mix some Rh(III) porphyrin, some base, and a 100-fold excess of water, heating everything in the dark for 2-3 days. Out pops the "bibenzyl" compound (83%), which initially causes a bit of head-scratching: why don't they see C-H activation products? And where's the hydrogen coming from? 

(**SPOILERS BELOW**)

There's quite a bit of C-H activation, actually - it just doesn't go anywhere under the conditions. Swapping in deuterium oxide leads, unsurprisingly, to "D" incorporation on all the methylene groups and at the two newly-formed methyl groups. So the water indeed provides hydrogen, but not the way we'd usually think about it. No water splitting, no hydride formation, no "M-H." Instead, it's simply a radical quench: each metal-carbon bond, formed from C-C bond homolysis, grabs an "H-dot" from a neighboring water, and the remaining OH radicals shuffle away to produce hydrogen peroxide. And the rate looks pretty screwy, with second-order kinetics in metal, which the authors think means that two separate Rh(II) radical species - from Rh(III) reduction in situ - cooperate to cleave each side of the C-C bond simultaneously.

Source: Chan et. al., JACS ASAP 2012

Well, enough hype: only a few substrates (cyclooctane, cyclophane, strained substrates) have been shown to reliably react with these rhodium porphyrins, and the conditions (200 degrees Celsius, 3-4 days, in the dark?!?) aren't winning any immediate med chem converts. Hey, these things take a little time to become practical, so maybe one day you'll reach for your C-C "knife" of choice, and dial-in your molecular dissection.