all blogs

Sloppy science and Stapel

28 November 2012 - Fraud

Yes, the official Diederik Stapel report is out (NOS News). Find it at Scribd or (here in English).
I have always wondered how the guy could get away with publishing over 30 fraudulent psychology papers. His fabricated data even ended up in his students Ph.D theses! It was very easy. He claimed he knew schools and teaching staff in these schools with students willing to fill out the questionnaires required for the research. After collecting the questionnaires he threw them in a bin and then made up the results. He was also more than happy to do the grunt work in the psychology laboratory, taking the tedious data collection and analysis away from his students. In chemistry students do all the hard labour but apparently in some psychology communities it is the other way round.
But do you know any chemistry professors that place themselves between you and the analytical balance and always report 100% yields back to you? Should you not become suspicious when denied access to analytical data other than the data fed back to you by your professor that also look way too clean? The really interesting part about the Stapel case is that none of the 30 or so co-authors of his papers or even the 10 people supervised by Stapel for their Ph.D thesis did get into trouble and can continue their careers.

As an aside, from the report it becomes evident that test subjects in all studies were students and possibly even exclusively psychology students whose presence at "testing days" was mandatory. Even if no fraud takes place how valuable are data taken exclusively from students?
Another interesting aside: the report recommends that in the future no publicity should be given to scientific fraud investigations because that would hamper fact finding. And you were thinking more publicity would prevent future fraud cases?

Although Stapel made a full confession and turned in his Ph.D title, do not expect any humility from him. In a media statement today he claimed he was self-delusional and that he was convinced 'he was only helping people'. The self-delusion part is true: this Friday his new book is in the bookshops.

Regenerating spent ammonia-borane II

27 November 2012 - The hydrogen economy

In a previous episode Sutton et al. introduced a new method for the regeneration of ammonia borane (AB) using hydrazine. Bad idea! say Reller and Mertens because hydrazine is (1) a hazardous material and (2) therefore predominantly sold as the less hazardous hydrate. And the hydrate cannot be used in AB regeneration. In their recent publication (DOI) Reller and Mertens suggest an alternative scheme for AB regeneration featuring an entirely different intermediate as summarized below:

1 - Test reaction: conversion of ammonia borane to polymerized BHN at 95°C
2 - hydrochlorination with aluminum chloride / hydrogen chloride in carbon disulfide at 40 bar forming ammonium chloride and boron trichloride
3- Hydrodechlorination with hydrogen at 130°C / 60 bar and Ni3B as catalyst. Trietylamine is added as thermodynamic driver forming the Et3NBH3 adduct. Byproduct HCl can be recycled.
4 - Exchange reaction back to H3NBH3 starting with liquid ammonia then heated at 80°C for 12 hours.

An efficiency of 60% was reported.

Bottlenecks reported: small window of opportunity in step 4 as dehydrogenation sets in already at 90°C but eventually ammonia can replace triethylamine.
Lucky breaks reported: despite its abundance chlorine did not poison the nickel boride catalyst.

Dendralene CRD

23 November 2012 - The CRD

The latest entry in the Chemical Reaction Database (CRD) features novel dendralene chemistry. (DOI). Wang et al. report the simple reaction of 3-dendralene with nitrosobenzene in chloroform to a single Diels-Alder reaction adduct. The other possible regioisomer does not form. The transition state (TS) according to Wang is actually a biradical and the favored TS has a pentadienyl radical and the unfavored one a less stable allyl radical group.

And how helpful is the Organic Letters publication in making information available? It certainly does not publish systematic names so OPSIN is of no use here. It is not an ACS policy (three randomly selected articles do have systematic names ) so it is just the authors being lazy or having a competence issue.Is there a SMILES to systematic name converter out there?

No, the image below does not look great. The components (reactants, products etc.) are assembled from the database records but getting them to work together in an appealing visual is tricky. A solution is in the making.

Dendralene reactivity 2012.PNG

Organic synthesis by electrospray

17 November 2012 - Orgoish

prep electrospray muller 2012  Graham Cooks has a fondness for electrospraying. Basic principle: eject a solution through a nozzle to which is applied a large voltage, liquid exiting the nozzle adopts a cone-shape and from its tip highly charged microdroplets are emitted from which solvent is rapidly evaporated. Convenient in electrospray ionization for generating charged macromolecules in MS. Stuff that Cooks like throw at a mass spectrometer through an electrospray includes milk and urine (DOI), blood (DOI) and living plants (DOI). In a novel exploit Cooks is electrospraying organic reactions (DOI).

More specifically the water solution to be electrosprayed contains indanone, 4-chlorobenzaldehyde and potassium hydroxide and the reaction taking place is a Claisen-Schmidt condensation. Compared to a standard bulk reaction the electrospray reaction is much faster due to rapid solvent evaporation. Taking into account droplet speed (200 m/s) and distance travelled to collecting unit (5 cm) the reaction is near completion (> 92%) in a matter of milliseconds.

Off course the reaction scale is minute, at the milligram scale, and the number of side-products has increased. Not easily understood from the article is the effect on reaction time of the negatively charged reaction intermediates in this reaction. The sign of the voltage does not seem to matter much. Another issue to tackle is reaction scale, if 4 nozzles can produce a milligram under a minute, how expensive will 4000 nozzles be when producing a gram per minute?

Fun with OPSIN

10 November 2012 - CRD update

In the meanwhile we are making progress on our Chemical Reaction Database (CRD). Entering chemical reaction data into a database is time consuming and every tool we can get our hands on to speed things up is very welcome. In a previous episode we have been looking at Open Babel for easy generation of images from a SMILES string.
Another tool is called OPSIN, introduced in 2011 by the Unilever Centre for Molecular Science Informatics in Cambridge (10.1021/ci100384d DOI). It is an open-source web service for converting IUPAC systematic names into SMILES, inChi and CML strings. The chemical journals do not publish them so getting the SMILES codes with every molecule involved in a reaction would require paying a visit to the PubChem sketcher and draw the molecule first.
The OPSIN tool should then make live easier. The Computer Lab surrendered the necessary code (under 20 lines!) to the CRD project and the first OPSIN results are really good. Collecting SMILES,InChi and CML from a single chemical name input takes a fraction of a second. So far it has managed 2'-aminoacetophenone, 2-aminoacetophenone, sodium hydroxide, 3-Methyl-1H-indazole without complaints. Di-µ-bromobis(tri-tert-butylphosphine)dipalladium returns a 400. The SMILES string is directly routed to the Open Babel tool so actually drawing molecules is a thing of the past.

Moerdijk chemical fire trial update

09 November 2012 - (un)justice

Remember the big Moerdijk disaster?. Not the low-level employee who started the fire that set chemical packaging company Chemie-Pack ablaze was prosecuted (clumsy attempt to unblock a pump with a gas burner) but his managers. Today the trial proceedings came to a close and the prosecutor is demanding jail time (2 to 4 years) for the CEO, the production manager and the safety guy (NRC, BNR). Charge: overall neglect of safety regulations and deliberate arson. The most detailed information can be found at the website of the public prosecution office itself (link), no thanks, journalists! Interesting details: personnel was not paid by the hour but by units processed, mobile phones where used in area's with explosion risk, and last but not least the prosecutors blamed management for allowing personnel or at least not restraining personnel from engaging in reckless forklift truck driving (!). Out of 4 years how much time is allotted for the forklift driving? A month? Two months? The prosecutor should understand forklifts are A fun to drive and B driving a forklift requires a special licence. Nowhere is alleged licences were missing so perhaps it would be a better idea to prosecute the organization issuing these licences.
Unfortunately, in this country judges are not elected (sorry, no jurors either!) and prosecutors can do as they please. To be continued.

Impossible catalyst for impossible reaction

02 November 2012 - Ammonia synthesis

klik hier voor de afbeelding op ware grootteOne percent (!) of the world's energy output is spent on the Haber-Bosch process for the production of ammonia Kitano et al. (DOI) remind us in a recent contribution to Nature Chemistry. Good thing they invented a new experimental catalytic system just a little more efficient than the previous one.

Converting nitrogen and hydrogen to ammonia is hugely energy demanding because of the N:::N triple bond with regular catalysts (Fe or Ru) requiring temperatures up to 600°C and high pressures. Their secret weapon is combining ruthenium metal with the very unusual electride (Ca24Al28O64)4+(e-)4, a zeolite-like compound that has trapped electrons where it should have counterions!

This electride can be produced by annealing monocalcium aluminate (12CaO.7Al2O3), a constituent found in common cement (also see Mayenite). Heating chases away two equivalents of oxygen leaving the electrons behind in the crystal's interconnected cavities. With the electrons freely and easily available, the electride is a great electron donor for the reduction of nitrogen. Ruthenium was added mixing the electride with Ru3(CO)12 and heating at 450°C.

Compared to commercially used Ru-Ba/activated carbon the new catalyst has a much higher ammonia synthesis rate (8-fold) even at a high operating temperature of 400°C with cutting the activation energy in half.

In regular catalyst systems hydrogen absorbs on the metal surface first and the resulting adatoms hinder absorbtion of nitrogen. Catalyst poisoning in the new system is limited because hydrogen can absorb as hydride ions into the crystals. When an hydride ion reacts with an activated nitrogen species on the metal surface, electrons are again left behind.