all blogs

Tackling oil sands

23 July 2010 - Petrochemistry

Oil sand extraction.jpgThe black goo on the left in the picture is crude Alberta oil sand, an ugly mixture of sand, clay, bitumen and water. Present-day isolation of the bitumen (crude oil!) requires an equally ugly process: hot-water extraction to get rid of the sand, coking at plus 500 °C followed by catalytic hydrocracking to get lighter fractions and catalytic hydroprocessing (300-500 °C, 18MPa H2) to get rid of heteroatoms and metals. The overall process requires a lot of energy, water and produces a lot of waste.
Brough at al. (DOI) in a new publication report an alternative method using supercritical carbon dioxide. Oil sand, toluene and a catalyst (Rh/C, Ru/C) are mixed in a high pressure vessel at 100°C and hydrogen (6 MPa) and carbon dioxide (12MPa) are added. In the resulting petroleum product the API gravity (density of a petroleum liquid compared to water) has increased from 8 to 16° (the Wikipedia article will explain why the unit is a degree....), sulfur and metal (Ni, V) content has decreased and H/C ratio has increased. The researchers are quite pleased with themselves: the yellow substance in the central picture looks like a regular oil and the with the sand in image three you can almost start a beach.

Novel silicon based fragrances

22 July 2010 - Organosilicon chemistry

Patchouli odorants Sunderkotter 2010c  If you are into the Patchouli fragrance you might be interested in the the new silicon-based compound tricyclopropyl(1-hydroxy-1-methylethyl)silane 1 which you can smell at an even lower threshhold value of 0.14 nanogram per liter air than the main natural patchouli ingredient patchoulol (2, 0.93 ng/L). The novel compound was created by a group of chemists from the University of Würzburg and Philip Kraft of the flavor and fragrance company Givaudan (Sunderkötter et al. DOI) with main ingredients bromocyclopropane, tetrachlorosilane, ethyl vinyl ether and a lot of lithium (elemental, tert-butyllithium, methyllithium). The basic idea behind the molecular design is that two compounds smell alike when key molecular parts are roughly of similar shape and with similar polar and apolar spots on its van der Waals surface. Substituting carbon for silicon has two advantages: the carbon-silicon bond is longer allowing the molecule to be constructed from less atoms and at the same time silicon decreases vapor pressure.

The inevitable Word Cloud

18 July 2010 - Datamining

No website nowadays is complete without a word cloud. Numerous sites offer to create for you a word cloud pic but then it is based on a single page not an entire website. NNNS computer labs stepped in, scrambled up some code, did the calculations and the results are just in. It will come as no surprise that in the chemical elements competition carbon and hydrogen are leading the way. They are followed by sodium, palladium, chloride, copper and gold.
reactions copper catalyst gold palladium reaction complex product addition compound acid compounds hydrogen bond research molecular carbon molecule groups sodium mechanism chloride organic chemistry wikipedia electron method formation water chiral system total metal chemical ring synthesis

Gold-catalysed Sonogashira?

15 July 2010 - Catalysis

Palladium free Sonogashira coupling  A year ago Buchwald & Bolm investigated some alleged iron-catalysed coupling reactions and found that copper present in minute quantities as a contaminant was the true catalyst (blog here). In a similar case, Lauterbach et al. in a new publication (DOI) argue that several gold-catalyzed coupling reactions reported in the recent literature are in fact co-catalyzed by hidden palladium. Gold just cannot do it on its own.

The coupling reaction in question is the Sonogashira reaction between an aryl halide (ArX) and an alkyne (R-CC-H) that is catalysed by a combination of palladium and copper. In it Pd(I) supplies the oxidized L2XPdAr complex and Cu(I) supplies the acetylide and after a transmetallation step and a reductive elimination step the product R-CC-Ar forms. Now gold(I) is isoelectronic with Pd(I) and could actually replace both Pd and Cu. This idea was tested out by Arellano et al. in 2007 (DOI) and by Li et al. in 2008 (DOI) with respectable results.

But Lauterbach is not convinced as in his hands a simple gold-only Sonogashira ends up with a meager 2% yield. Something is going on and some more testing is done. If the reaction sequence in the catalytic cycle is oxidative addition followed by transmetallation then a reaction between PhI and AuCl(PPh3) should result in oxidative addition but it does not. Likewise if the cycle is initiated by transmetallation the compound (PPh3)AuCC-Ar should be oxidized by ArI but again this reaction does not work. What does work on the other hand is the addition of small amounts of palladium. This makes the conclusion inescapable that in the so-called gold-catalysed Sonogashira hidden palladium (as contaminant in gold or base or solvent) is the true catalyst with gold just replacing copper.

In this way the palladium-free Sonogashira reaction takes on a new level of complexity just like the so-called copper-free Sonogashira reported on in an earlier blog.
Lauterbach, T., Livendahl, M., Rosello?n, A., Espinet, P., & Echavarren, A. (2010). Unlikeliness of Pd-Free Gold(I)-Catalyzed Sonogashira Coupling Reactions Organic Letters, 12 (13), 3006-3009 DOI: 10.1021/ol101012n

FLP twice over.

12 July 2010 - FLP chemistry Part 6

FLP alcarazo 2010  Alcarazo et al. describe a new Frustrated Lewis Pair (see FLP part V) with new Lewis acid hexaphenylcarbodiphosphorane (1) and the familiar Tris(pentafluorophenyl)boron (DOI). At room temperature the compounds react in nucleophilic substitution, at -78°C they do nothing but add hydrogen at that temperature and salt 3 forms as a result of heterolytic hydrogen cleavage. The new pair is also good at C-O bond cleavage (THF), C-F cleavage as in the fluoropentane example and in Si-H cleavage. Remarkably the pair is still active when 1 is protonated or alkylated such as in compound 6. It is not reactive enough for hydrogen but it can still cleave methanol. A frustrated Lewis pair twice over.

Exit organon

09 July 2010 - R&D in the Netherlands

intermediair vac chemicus.jpgThe future of the MSD branch in the Dutch city of Oss ( ex. Schering-Plough ex. Organon International, ex. AkzoNobel ) looks gloomy. Half of the 4000 jobs will go between now and 2012, partially transferred to the USA. Usually de press does not bother to check exactly what sort of jobs disappear when a company downsizes but this time we are assured all high-end research and development is involved.

Oss, like Chicago used to be a city of slaughterhouses (the locals all own a knife and know how to use one on cattle or on fellow humans) and Organon was founded in 1923 selling insuline produced from refuse. From then on there followed a whole range of hormones. In 1962 the company introduced the birth control pill lyndiol based on the hormone Lynestrenol and later on in 1982 Marvelon (desogestrel). One Internet source recounts the story that de production of the pill in the 1950's was not welcomed at all in Catholic Oss and that packaging was outsourced to another company. After consulting with the local Bishop the pill was branded as a menstrual cycle regulator with temporary infertility as a side-effect. The Internet sources are scarce but a certain Max de Winter is credited here as the father of the lynestrenol pill and one Van den Broek here for the invention of desogestrel. This de Winter incidentally as a Jew before his career at Organon was forced to leave the Shell company in Amsterdam in 1942 and survived Auschwitz only on account of his technical skills by forced labor at Siemens, very much like Primo Levi.

Fokke & Sukke (a duck and a canary) also know their history and remark in today's paper that if only the mother of the present MSD CEO would have used the pill nobody at Organon today would have been fired. And there is of course the blame game: Both the Nafarma (the Dutch pharma lobby) and the company's management and works council blame the Dutch government for not supporting innovative companies (Volkskrant July 8). Management also laments the slow drug approval procedure that exists in the Netherlands.

And in what way has the Dutch government been supporting R&D thus far? By mass-subsidizing commercial companies or simply put: paying the bills (25 million euro in the last 10 years to organon alone, not counting EU subsidies). And what self-respecting R&D CEO is willing to spend his own hard-earned money on R&D when the competition gets it for free from the government next year?. Already the Dutch Ministry of Economic Affairs and even the Oss municipal council are negotiating with MSD the rescue of some business units and the there is talk of setting up a science park at the Organon location. With respect to the science park the MSD CEO has already done the math: the government will again heavily subsidise the new start-ups and MSD can decide at will whether to participate. So instead of just handing over money to technology companies what should the government do? Just make sure there is a large reservoir of skilled people (invest the money in schools & universities) and make sure people want to actually live and work in the Netherlands. That is combat housing shortages, combat traffic jams, combat taxes and combat crime.

Update 21 July 2010: newspaper Volkskrant reports employees of pharmaceutical company Solvay in Weesp also fear R&D shutdown. The paper also reports Oss will get its science park as a result of parliamentary debate.

Update 24 September 2010: Duphar R&D will close down. Organon shut-down postphoned to year-end. New owner for Crucell?

Some like it flat

07 July 2010 - Flat carbon flat nitrogen

flat nitrogen flat carbon  We are quite used to a trigonal pyramidal molecular geometry for trivalent nitrogen and a tetrahedral one for tetravalant carbon. It does not have to be that way. For example in the compound class of the fenestranes the central carbon atom is flattened and so is the central nitrogen atom in triisopropylamine. Two recent publications describe more flattening: of nitrogen and of carbon.

Livant et al. synthesised tris(1,3-dihydroxy-2-propyl)amine (DOI) and determined its X-ray structure. The degree of planarity for this molecule can be measured as the height of the nitrogen atom above the carbon plane (0.082 angstrom) and by the sum of CNC angles (359°). In an ordinary amine these values are 0.4 A and 321°. The molecule has two internal hydrogen atoms and every oxygen atom is both a hydrogen-bond donor and an acceptor. The nitrogen center is flattened for two reasons: a steric contribution is crowding of three hydroxymethyl groups when the nitrogen atom would move away from planarity. The hydrogen-bond network may also play a role. The electronic contribution is a favorable interaction between the nitrogen lone pair and the antibonding C-O sigma orbitals. The shortening of the C-N bonds - 1.454 angstrom vs 1.47 angstrom - is taken as evidence. The basicity is very low - pKa 3.08, compare triethylamine pKa = 10 - and the protonated molecule is flattened as well more resembling trigonal planar than tetrahedral.

Liddle et al. present (DOI) a distorted trans-planar 4 coordinate carbon atom in a complex molecule constructed from a diphosphazene CH2(P(Ph)2=N(dipp))2, buLi and TMEDA. It is basically a methyl group with two phosphazene groups and twice deprotonated and replaced by lithium. The Angewandte helpfully added the header planar 4-coordinate carbon to the article but what exactly is the geometry? The phrase The root-mean square deviation from the mean plane of C1, P1, P2, Li1, and Li2 is only 0.34 angstrom, and C1 deviates from this plane by 0.007(2) angstrom is not at all help full. The atoms making up the trans-planar geometry are not identified. The P-C-P angle of 132° and the Li-C-Li angle of 161° certainly do not imply overall planarity. The longest Li-C bond length is 2.53 angstrom but as that same lithium atom also connects to one of the dipp isopropyl carbon atoms by 2.66 angstrom one must wonder if that Li-C bond is a true bond.
Jie, Y., Livant, P., Li, H., Yang, M., Zhu, W., Cammarata, V., Almond, P., Sullens, T., Qin, Y., & Bakker, E. (2010). An Acyclic Trialkylamine Virtually Planar at Nitrogen. Some Chemical Consequences of Nitrogen Planarity The Journal of Organic Chemistry, 75 (13), 4472-4479 DOI: 10.1021/jo100628v
Cooper, O., Wooles, A., McMaster, J., Lewis, W., Blake, A., & Liddle, S. (2010). A Monomeric Dilithio Methandiide with a Distorted trans-Planar Four-Coordinate Carbon Angewandte Chemie International Edition DOI: 10.1002/anie.201002483

The NNaNMgNNaNMg cycle

03 July 2010 - Chemical misadventures

THF Mg kennedy 2010  In a previous episode Kennedy & Mulvey successfully isolated the sedated anion of THF as a zincate that was stabilized by both TMP and TMEDA. The potential of this unual compound containing a OCZnNNa cycle was then demonstrated in reactions with common electrophiles. Replacing zinc by magnesium in this procedure should be an easy walk in the park but it did not work out that way (DOI). Whereas sedated THF took 2 weeks to form in the zinc reaction pot, the magnesium version only took 5 minutes to form crystals. Two sets of crystals in fact. Turns out that THF is completely shredded to pieces. The first set of crystals belongs to dimagnesiated butadiene stabilized by TMP, sodium and TMEDA. In the second set of crystals magnesium, sodium and nitrogen form an inverted crown ether for the oxygen dianion originally belonging to THF as well. In terms of a mechanism Mulvey cautiously offers THF deprotonation twice over forming the oxide and a dicarbene.