all blogs

Methane activation by iron

28 May 2014 - Chemical engineering

methane conversion.PNGNatural gas (methane) is an important chemical feedstock. It can be converted into syngas via CH4 + H2O -> CO + 3 H2 and then into methanol via CO + 2H2 -> MeOH. Translated, this route implies easy methane hydrolysis with one hydrogen equivalent as a bonus. Methanol then enters chemical industry as the starting material for a host of secondary chemicals. Do not fix something that is not broken? Guo et al. in a recent report try a different approach to methane activation eventually arriving at ethylene.

They note that the big hurdle for methane is overcome is the high carbon-hydrogen bond strength and low polarizability. Oxidative coupling of methane to ethylene (2CH4 + O2 -> CH2CH2 + 2H2O) suffers from over-oxidation leading to carbon dioxide formation and coke formation.

The Guo team has come up with a catalyst they call Fe@SiO2, made by mixing fayalite (the iron mineral Fe2SiO4) and silicon dioxide in a ball mill, fusing it at 1973 K for 6 hours and leaching it with nitric acid. At 1363 K this catalyst system was able to convert methane to ethylene and hydrogen gas and some benzene and naphtalene with 48% efficiency. Without the catalyst or with SiO2 not containing any iron a similar setup yielded only 95% coke. Several known catalysts based on iron/silicon did not work either. The main product from conventional methane pyrolysis is acetylene and again coke.

The rationale behind the apparent success of just this catalyst system according to Guo is the isolation of the iron sites in the silicon matrix (HAADF data included). This prevents C-C coupling reactions taking place at the surface and hence coke formation. During the catalyst lifetime the iron atoms also get better embedded in the matrix. External carbon also dissolves in the matrix forming Fe-C bonds that further help stabilize the iron center. The role of the iron center is to eject methyl radicals from the surface. These radicals have been detected by (take a deep breath) "vacuum ultraviolet soft photoionization molecular-beam mass spectrometry". The radicals then combine to form ethane and then split of hydrogen to form ethylene. The source of the external carbon is not fully explained. Does it limit the final 48% yield? By adding ethane to the gas input the yield also increases again no explanation is given.

Fullerene with wings

15 May 2014 - Orgo

C70 anthracene bis adduct.PNGNeti et al. explain here what happens if you react C70 fullerene with anthracene. The compounds were mixed and heated above the anthracene melting point at 240° for some time, resulting in a neat bis-Diels-Alder adduct (68% yield after column chromatography) with the anthracene units at opposite ends of the cage. The work was a repeat of an identical reaction but then with C60 that was reported 15 years ago by the same group (DOI). As with C60 the reaction is reversible: one unit is lost at 180°C and the other at 220°. As with C60 the anthracene units can be exploited in templated activation of fullerenes and as protective groups.

The B18 dianion

13 May 2014 - Chemical Zoo

B18 dianion.PNGNew in the chemical zoo, well at least computational: the B18 dianion as reported by Moreno et al. here. This anion is predicted to have a pentagonal six-boron core surrounded by a bowl-shaped B12 ring. The pentagon can move freely inside the ring as in a Wankel engine and the molecule is therefore fluxional although the imaginary temperature in the simulation had to be cranked up to 900 Kelvin. The inner core will not actually fall out because of a continuous process of bond making and bond breaking. Based on the total number of pi electrons the molecule should be antiaromatic but in this particular configuration the core is again aromatic.

A solvent for selenium

11 May 2014 - Nanotech

A solvent for selenium.PNGAs reported earlier Webber et al. have recently demonstrated that a mixture of a thiol and an amine can dissolve an impressive amount of tellurium. That was not all that there was to it. The same mixture also works for selenium. For example a 1:4 mixture by volume of ethanethiol and ethylenediamine can be 38% in Se. With regards to a mechanism Webber noted that conductivity increased during dissolution (ions!), that the thiol was deprotonated (proton NMR) and that selenium ring structures were formed (Raman).

But this is not all there is to it. Another research group appears to make the same thiol-amine-selenium claim. In an accepted manuscript Walker and Agrawal ( DOI) demonstrate that a very similar 2:1 mixture of ethanethiol and oleylamine can dissolve up to 5% selenium. Both teams report that solubility in either solvent is negligible. In the Walker mechanistic scheme the thiol is oxidized to a disulfide (spotted by GCMS). Selenium is reduced to the Se82- anion. The solution has to be basic and that explains the amine. The ammonium ion and the Se82- anion combine to form the dark-red solution. Webber does not mention disulfides.

So who can claim priority? Both teams cite Liu et al. (DOI) who in 2012 used hexadecanethiol, oleylamine and selenium and already described it as a alkylthiol reduction. In 2014 the Webber article is older by three weeks but Walker & Agrawal have not been in a hurry: they confidently write their solution is stable for up to three years.

The belt wars continue

10 May 2014 - Orgo

5CPP.PNGTwo research groups have almost simultaneously reported the synthesis of 5-cycloparaphenylene (5CPP), the chemical compound with 5 phenylene groups joined together in a loop via 1,4-connnections and of some relevance because it also makes up the equatorial belt of fullerene. Team Evans, Darzi & Jasti (DOI) beat team Kayahara, Patel & Yamago (DOI) by just a month. Both teams are veterans in the field with a track record of research on similar but larger nanohoops (earlier report here).

Team Jasti considered itself lucky to find traces of a 5CPP precursor amidst a 10CPP synthesis based on a dimerisation. This precursor (already strained by 32 kcal/mole) was then reduced by sodium naphthalenide to a dianion and reaction with methanol produced a neutral methanol-subtstituted 5CPP. Removing those groups with lithium diisopropylamide gave 5CPP itself as a "dark-red solid with a deep ruby colour" solution in toluene and stable at RT. Based on the crystal structure, Team Jasti reports a benzenoid structure with elongated phenyl-phenyl bonds and identical bond lengths for the phenyl hexagons. The displacement of each tertiary carbon atom out of the hexagon plane is 15.6°. The strain energy has increased to 119 kcal/mole.

In the meanwhile the Yamago team in their publication presented their strategy towards 5CPP as a hybrid of earlier Jasti and Yamago schemes and coincidently arrived at an almost identical 5CPP precursor with 4 methanol groups instead of just two. After reduction with tin chloride 5CPP was isolated as a "dark-purple solid".

Is 5CPP relevant? Here is a quote from Graham Bodwell in his Nature Chemistry News & Views item on the topic: "Just as the most skilled golfer makes the most difficult of shots look easy, the Jasti and Yamago groups have made the synthesis of 5CPP look straightforward. Ironically, this can make it very difficult to convey the significance of the accomplishment to the scientific community."