all blogs

Novel Fischer-Tropsch catalyst

25 February 2012 - Catalysis

It is not every week that an inorganic chemist is making headlines in the Dutch newspapers or even the Dutch television news bulletins. The inorganic chemist is Krijn de Jong from Utrecht University and the headline-making news is a new catalyst for the Fischer-Tropsch (FT) process. The Science article was a joy to read (DOI) but it does have problems.
Problem number 1 concerns the title Supported Iron Nanoparticles as Catalysts for Sustainable Production of Lower Olefins. The intended feedstock for olefin production is biomass and hence the sustainability claim. But the new iron oxide catalyst works on syngas, the carbon monoxide hydrogen mixture that can be produced from any carbon source (biomass included) at an industrial scale. No biomass was anywhere near the Utrecht laboratory.
What the new catalyst does particularly well is converting CO/H2 to directly to the lower olefins (ethylene, propylene) instead to the higher ones. And these lower olefins are what plastics are made from. With the present state of technology these lower olefins are only accessible in a very inefficient way by cracking higher olefins.
Regular iron-based FT catalysts tend to degrade mechanically in typical FT process conditions (high temperature, high pressure) and therefore the new catalyst has iron oxide particles deposited on carbon nanofiber (CNF) as a solid support. Surface oxidized nanofiber (surface area 150 square meter per gram) was impregnated with a water solution of ammonium ferric citrate and then heated for 2 hours at 255°C. Silicon carbide particles were added and this ensamble was reduced using hydrogen although is it unclear what is being reduced and why. Note that although the article consistently mentions iron particles the true catalyst is an iron oxide.

Sure enough, compared to bulk iron or compared to several other supported iron oxides, at 20 bar and 340 °C, Fe2O3/CNF displays selectivity towards formation of the lower olefins at the expense of methane formation combined with a high carbon monoxide turnover. On the other hand the production of carbon dioxide (clearly a useless by-product) appears to be up. problem number two with this article is that this piece of information is presented in the supporting information and not in the article itself.
Problem number three concerns the requirement for sulfur and sodium as trace elements. An unsuspecting reader may get the idea that catalyst design involves testing a huge amount of element combinations before one of them actually works. As it happens an interview in the Dutch newspaper NRC Handelsblad with Krijn de Jong (title translates roughly as plastics from garbage incineration) reveals more. Apparently the discovery of the catalyst by one of his Ph.D students is a nice case of serendipity because reactions were found to work well with one particular bottle of ammonium ferric citrate where the sodium and sulfur were hiding but not at all with a second bottle. Should this piece of information not be in the article as well? Does the reaction require both sulfur and sodium or just one of them?

Moerdijk requiem

10 February 2012 - Freak accident

Moerdijk remnant pump.gifThe Chemiepack company is bankrupt, its management is still facing jailtime (earlier report here) and the official report into the Moerdijk disaster (chemical packaging/mixing company went up in flames, no one was hurt though), the publication of which Chemiepack tried to block, is finally out (PDF, 5MB in Dutch). So what did the investigators find and what was Chemiepack trying to hide? A lot it seems.
The report explains in detail how the fire started: someone was transferring resin from a container to an industrial dispenser with the aid of a pump, because of the cold the pump got clogged and the same someone then defrosted the pump with a gas burner (against regulations of course). This ignited a overflow container filled with xylene, used to clean the pump. Operators on the scene then attempted put the fire out but failed because a) a powder fire extinguisher did not work and b) water was used which only helped dispersing the burning organic material.
To make matters worse no one thought of switching off the pump and the pipe carrying the resin to the dispenser burst due to a faulty weld, causing a spray of resin (up to 150 liter per minute) that was ignited by the burning xylene. The burning resin then made sure that the plastic containers in the vicinity of the fire collapsed adding their inflammable content to the fire as well. Nice detail: laboratory testing showed it is not possible to ignite the resin directly with a gas burner, burning xylene is a necessary go-between.
The fire department was then called in and their response was also assessed in the report. Contrary to the early news reports the fire department at an early stage decided to let the site burn out in a controllable way. They did not use water to try to put the fire out but used water merely to guard specific objects, for example a sea container filled with isopropyl alcohol. In that sense the fire fighters are vindicated. Chemiepack management on the other hand is blamed for ignoring all sorts of safety regulations. The pump in question did not just malfunction in the winter of 2011, it did malfunction every winter.
One of the recommendations relevant to us chemists concerns the use of plastic containers (up to 1000 liters) for packaging hazardous chemicals. They are often made from polyethylene which is a material resistant to chemicals but which melts already at 105°C. A burning wad of newspaper is sufficient for ignition.
With respect to the various chemicals stored on the site the report specifically mentions one chemical called 2-ethylhexyl nitrate present in a quantity of 160 tonnes. This chemical is used as a fuel additive and decomposes and ignites at 130°C. Apparently the owner knew of the hazards but none of the Chemiepack employees did.
And why did the pump clog up in the first place? The pump ran on compressed air and the exhaust apparently made a hell of a noise. To keep noise levels down an exhaust muffler was added. Expanding air efficiently cools this muffler and in combination with an already cold January day and condensation water the muffler clogged up with ice and reduced in turn pumping action to zero. Enter the gas burner.

Ranking the carbenes

11 February 2012 - Orgo

Zhang iconic carbenes  Some time ago Robert Moss found a new and convenient way to synthesise dichlorocarbene from a diazaridine. More carbenes followed prepared in a similar way and in a recent report Moss together with Zhang, Thompson & Krogh-Jespersen have pitched 6 different carbenes against 6 different alkenes in an ultimate reactivity showdown both experimentally and theoretically (DOI).
In calculations the order of reactivity in the carbene line-up is CCl2 > CClF > CF2 (difluorocarbene) > CClOMe > CFOMe > C(OMe)2 (dimethoxcarbene) In this order the substituents increasingly donate electrons in the empty carbon p-orbital, making it less electrophilic and more stable. The LUMO energy increases.
The 6 alkenes can likewise be ranked with decreasing nucleophilicity and HOMO energy (pi electrons more tightly bound) in the order tetramethylethylene, cyclohexene, 1-hexene,methyl acrylate , acrylonitrile and chloroacrylonitrile because the substituents are increasingly better electron-withdrawing groups.
In the good tradition of frontier molecular orbital theory a reaction is fast (low energy transition state) when in either combination the HOMO-LUMO energy gap is small. With nucleophilic carbenes it is all about carbene's HOMO and alkene's LUMO and with electrophilic carbenes it is the other way round. This explains why difluorocarbene will only react with the electron-rich alkenes and dimethoxycarbene only with the electron-poor one's.
The experimental activation energies for reactions of C(Cl)2 with all 6 alkenes give a clear picture: the activation energy increases with increasing carbene LUMO - alkene HOMO energy difference just like you would expect. With the most electron-poor alkene the activation energy drops again suggesting that now C(Cl)2 reacts as a nucleoplice and not an electrophile. Again predictably with nucleophilic CClOMe the situation is reversed although the report does not explain with this carbene fails to react with cyclohexene or 1-hexene.
The confusion sets in when the entropy of activation is considered which experimentally becomes less negative (making the reaction more favorable) on going from CCl2 to C(OMe)2. In line with Hammond's postulate the transition state should become later, tighter, and more sterically demanding as the carbene becomes more stable. According to the authors 'compelling explanations for this counterintuitive pattern and the noted discrepancies between computed and measured activation parameters are currently lacking.

The mechanics of ice

06 February 2012 - Skating fever

klik hier voor de afbeelding op ware grootteIt is freezing cold in the Netherlands making the locals nervous for the Big One. The country has plenty of surface water that all frozen up, makes a very large temporary ice skating super highway. If only all the ditches, rivers, canals and lakes the country has would freeze up nice and predictably with a 15 cm pitch black ice sheet. Reality is stubborn. Right now the ice is riddled with wind holes thanks to of course prominent wind action. These wind holes take a lot longer to get to any decent thickness. To make matters worse all ice is currently coated in a layer of snow conveniently hiding the wind holes not to mention all the cracks.
This blog braved one of the frozen up canals this week that luckily had a stretch de-snowed. Most of it black ice but with patches of grey ice, just like in today's pic. But what exactly are the little mounds of snow doing there?
A very useful website is lakeice.squarespace.com, completely dedicated to skating and lake ice! The closest thing on this website that might describe this phenomenon are ice pores where pores through the ice allow water to surface and then sublime.

The zinc-ion battery

04 February 2012 - Concepts

In the meanwhile (see earlier blog) Xu, Li, Du & Kang have a new battery design of their own they would like to call the zinc-ion battery (DOI). They combine discharge power with high capacitance. The cathode is manganese dioxide, with zinc as anode and aqueous zinc sulfate as the electrolyte. Zinc is oxidized to zinc(II) ions which then migrate to the MnO2 cathode. MnO2 has a peculiar crystal structure with long channels that can incorporate zinc ions by a process of intercalation (aka insertion). Zn2+ and MnO2 are then reduced to ZnMn2O4. In recharging mode the reverse reaction takes place and zinc migrates back where it came from (extraction).
Compare this to the Lithium-ion battery where lithium is doing all the shuttling between anode (graphite) and cathode (Lithium cobalt oxide). Lithium intercalates into the carbon electrode (see: Graphite intercalation compound). Another similar battery type is the Zinc-carbon battery where zinc is again oxidized to Zn2+ and manganese(IV) oxide is reduced to Manganese(III) oxide. According to the Wikipedia article the carbon present in this battery type just sits around and collects current but surely intercalation must be playing a role here as well.
The zinc-ion battery people find their new battery attractive because the electrolyte is non-corrosive (commercial batteries are either alkaline or acidic) and the other materials are nontoxic and low-cost. In one of those do-not-try-this-at-home experiments they demonstrate that the battery is also safe to use in a nailing experiment (no flash!) which must be what we think it is.

The liquid metal battery

03 February 2012 - Concepts

A magnesium-antimony Liquid Metal Battery? Sure, why not. Here is how Bradwell, Kim, Aislinn, Sirk and Sadoway made one (DOI) starting from a 90 mm hollow carbon rod. At the bottom of the cylinder is a tungsten pin as the positive current collector. The cylinder (diameter 40 mm) is lined with a layer of boron nitride for insulation, filled with a layer of pure antimony, then a layer of electrolyte salt (MgCl2-KCl-NaCl), then filled with a layer of magnesium metal and finally filled with a capping epoxy layer. A stainless steel pin on the top is the negative current collector.
This battery has an operating temperature of 700°C. In discharging mode magnesium is oxidized to Mg2+ ions which dissolve in the electrolyte salt. On the other side these magnesium ions are again reduced to magnesium which dissolves in the antimony phase. Both electrolyte salt and the resulting alloy at 700°C are liquids but do not mix. The driving force for the battery to run is a difference in chemical potential between solid magnesium and a magnesium / antimony alloy.
Compare this mode of operation to the way a common nickel-cadmium battery works: cadmium is oxidized to cadmium hydroxide on one end and nickel(III) oxide is reduced to nickel(II) hydroxide on the other side. Common potassium hydroxide is the electrolyte in between.
The main attraction of the liquid metal battery according to the researchers is that it is resistant to cracking, in this way it is a self-healing material. This type of battery already existed but this particular one uses less expensive metals.