Skip to content

Bond breaking dance

Picture from Wikipedia - Carbon disulphide orbital

Chemical reactions alter the arrangement of atoms and their constituent electrons in reactant molecules to form new product molecules. Almost every change occuring in our world is at some level the result of such a reaction. Visualizing a chemical reaction at the single-molecule level and ‘seeing’ how this rearrangement takes place in real time would give us deeper insight and greater understanding of the fundamental dynamics of the process.

Atoms consist of negatively charged electrons orbiting around a positively charged nucleus. Molecules consist of two or more atoms, each with its own nucleus and set of electrons, and in addition ‘valence’ electrons which have an orbit encompassing all the constituent atomic nuclei. These ‘outermost’  electrons of the molecule are responsible for the chemical ‘bond’ between the constituent atoms, which keeps the molecule tied together. It is these valence electrons which participate in a chemical reaction, breaking old bonds and forming new ones.

Quantum mechanics forbids the assignment of exact positions to electrons in an atom or molecule. Instead, there are fuzzy clouds called ‘orbitals’ surrounding the nuclei which represent probability distributions in space for finding the electrons. The valence electron orbital determines the size and shape of a molecule, since it is the ‘outermost’ cloud. Researchers at the National Research Council in Ontario, Canada have now been able to image the dynamical evolution of the valence electron orbital in carbon disulphide (CS2) molecules, during a chemical reaction where CS2 dissociates into CS and S.

CS2 is a linear molecule where the sulphur (S) atoms are at the ends, bonded to the carbon (C) atom in the center. Gaseous CS2 molecules in a chamber are first all aligned in a particular direction using an intense laser pulse i.e. a short burst of monochromatic light. An ultraviolet laser pulse then excites the aligned CS2 molecules to a state where the CS bonds are stretched and bent, from where they can dissociate into CS and S. Then, another ultraviolet ‘probe’ laser pulse is incident on the excited CS2 molecules after a controllable delay, which ionizes the molecules to break the valence electrons free and release them as energetic ‘photoelectrons’ that are detected.

The velocity distribution of these emitted photoelectrons i.e. the directions in which they travel outward from the aligned CS2 molecules, is completely determined by the ‘shape’ of the valence electron orbital at the exact moment of ionization. One can thus reconstruct the valence electron orbital at different times by measuring the photoelectron spatial distribution for different delays between the excitation and ionization laser pulses.

The CS2 molecule remains in the excited state for about a picosecond (a millionth of a millionth of a second) on average, before dissociating into the products CS and S. Even light, the fastest thing in the universe, only travels a distance roughly equal to the width of a human hair in that time. The alignment, excitation and ionization laser pulses used in the experiment all had a duration of about a tenth of a picosecond and the delay between the excitation and ionization pulses was varied from zero to one picosecond.

The reconstructed valence orbital had an hourglass-like shape in the initial linear form of the CS2 molecule, right after excitation. Thereafter, the orbital evolved over half a picosecond into a distorted dumbbell-like shape due to the stretching and bending of the CS bonds. The orbital shape oscillated back and forth between these two configurations for 1-2 picoseconds before the CS2 molecule decayed into CS and S. The experimental results matched very well with extensive computer simulations and theoretical calculations based on quantum mechanics.

The authors have essentially developed a camera that allows one to zoom into a molecule and watch the dance of the electrons choreographed by nature during a chemical reaction.


Hockett, P., Bisgaard, C., Clarkin, O., & Stolow, A. (2011). Time-resolved imaging of purely valence-electron dynamics during a chemical reaction Nature Physics, 7 (8), 612-615 DOI: 10.1038/nphys1980

Workplace stress: Opposites attract

Picture by Michael Schmid (Wikipedia)

Individual atoms are like tiny bar-magnets pointing in a certain direction, and the arrangement of a large number of these atomic magnets on a periodic lattice determines the macroscopic magnetic properties of a bulk material. In most materials, atoms at different lattice sites are oriented randomly and experience thermal fluctuations, resulting in no net magnetization. Under certain conditions however, in some materials like iron, nickel and cobalt, there can be an order in the orientation of atoms at different lattice points.

For example, if all the atomic magnets point in the same direction (“ferromagnetism”), the whole material acquires a macroscopic magnetization and behaves as the familiar magnet that we stick onto our refrigerators. This long range order arises out of a strong interaction (“exchange coupling”) between neighbouring atomic magnets in the lattice. The magnitude and sign of this exchange energy is a result of the quantum mechanical behaviour of the atoms and details of the lattice. If this exchange energy has the opposite sign, it can result in “anti-ferromagnetism”, where all the neighbouring atoms are anti-aligned to each other i.e. an atom pointing ‘up’ is surrounded by atoms pointing ‘down’ and vice versa. Imagine moving along a line in such a lattice – one would see a periodic arrangement of ‘up, down, up, down’ repeat forever.

Researchers at the Hitachi corporation in Japan have now seen such anti-ferromagnetic behaviour in the stress levels of humans in the workplace. They measured the stress levels of 630 people working in 11 different divisions of a software development company using standard questionnaires. They also monitored face-to-face interactions using “sociometric badges”  that the employees were told to wear for a 4-week period. The badges were equipped with accelerometers, infrared sensors and microphones to measure the movement and speech of every employee. A face-to-face communication ‘lattice’ was then constructed by making the people as nodes and establishing links between people who had on average more than five minutes per day of communication time.

The scientists calculated the correlation between the stress levels of employees as a function of the distance (number of links) between them on the ‘communication lattice’. Remarkably, they found a negative correlation between the stress levels of nearest neighbours (people separated by a single link). For next-nearest neighbour employees (separated by two links), the correlation was positive. For employees separated by three links, the correlation was found to be negative again, and the alternating positive-negative pattern continued till the maximum separation of five links. The result was replicated in all the different divisions of the company.

This is clear evidence of anti-ferromagnetic order, and it seems in the workplace, relatively low stress is experienced by people around other highly stressed people and vice versa. This is counter-intuitive and even more surprising considering the results of similar experiments conducted in groups of friends or family i.e. people who know each other well, where a positive correlation and ‘diffusion’ of stress between individuals was found. There, being around stressful people meant increased stress for the person.

The authors attribute this difference in the ‘macroscopic’ behaviour of the group to the difference in the ‘microscopic’ details of the human-human interactions. In analogy with statistical physics, they model human-human interaction as an “exchange coupling” and simulate the minimum energy configuration for a particular communication lattice. The situation of the ‘friends-group’ and the ‘workplace-group’ was reproduced just by taking opposite signs of the exchange coupling, suggesting a major difference in the nature of human-human interaction in the two cases. People tend to be competitive and less cooperative in their interactions in the workplace, while they tend to be kind, helpful and empathetic towards friends and family.


Watanabe, J., Akitomi, T., Ara, K., & Yano, K. (2011). Antiferromagnetic character of workplace stress Physical Review E, 84 (1) DOI: 10.1103/PhysRevE.84.017101

Single-atom speed-up

Picture from Wikipedia - Catalytic converter on a Dodge van (platinum is used as catalyst inside)

Catalysts are substances that speed up the rate of chemical reactions, without being consumed in the reaction themselves. Their influence extends over a wide range of areas, from reducing toxic exhaust in our cars (cataytic converter) to controlling metabolic processes inside our bodies (enzymes). Chemical reactions have an inertia or “activation energy”, a barrier that the reactants must climb over in order to form the products. Catalysts lower the height of this barrier by forming reaction intermediates, which enable faster conversion of reactants into products.

Noble metals like gold and platinum (Pt) are important heterogenous catalysts i.e. they are used in their solid form to speed up reactions in the gaseous state. The reacting gases “adsorb” (stick) onto active sites on the metal particles, where the activation energy is lowered. An important example is the oxidation of carbon monoxide (CO) into carbon dioxide (CO2). CO is a colorless, odourless, tasteless but extremely toxic gas that reacts with haemoglobin in the blood directly, compromising its capacity to carry oxygen. Concentrations as low as 500 parts per million in air can be fatal. 

CO in the presence of oxygen converts to CO2, which is non-toxic; however, the reaction rate is extremely small. For metal catalysts to work efficiently, it is crucial to have a large surface area (for a given mass) for the reacting gases to adsorb onto. It is only the atoms on the surface of the metal particles which catalyse the reaction; atoms in the bulk remain inactive. Imagine a spherical particle – the smaller the size, the higher the fraction of atoms on the surface. Nanometer (one billionth of a meter) sized particles, consisting of tens to hundreds of atoms, have proven to be good catalysts. The ultimate limit, however, would be having individual atoms of the metal separated from each other, dispersed on a given surface, to maximize catalytic efficiency.

Researchers at the Chinese Academy of Sciences in Dalian and the Tsinghua University in Beijing have now achieved that holy grail of catalysis. They developed a catalyst consisting of single Pt atoms on a surface of iron oxide, where one out of every 1430 atoms of iron, on average, was replaced by Pt. The sample precipitated out as a solid from a reaction between liquid solutions containing compounds of iron and Pt. The novel single-atom based catalyst was shown to have a record “turnover frequency” (TOF) of 1360 for CO oxidation, which means that on average about 1360 molecules of CO were converted to CO2 per second, for every Pt atom present. This speed-up is about 100 times better than those achieved by commercial Pt-based catalysts and about 3 times better than the best known gold-based catalysts. The catalytic performance was also found to be robust after hours of reaction time.

Extensive computer simulations were also performed to determine the mechanism of the catalysed oxidation reaction. First, an oxygen molecule (O2, containing two atoms) adsorbs onto a Pt atom, which is followed by the adsorption of a CO molecule onto the same Pt atom. The CO molecule reacts with one of the oxygen atoms at this low activation energy site to release a molecule of CO2. Then, another CO molecule sticks to the Pt atom and reacts with the remaining oxygen atom there to produce another CO2 molecule, after which the entire cycle repeats.

The authors believe that this single-atom speed-up can be extended to other precious-metal based catalytic systems, and greatly reduce the cost of commercial noble-metal catalysts.


Qiao, B., Wang, A., Yang, X., Allard, L., Jiang, Z., Cui, Y., Liu, J., Li, J., & Zhang, T. (2011). Single-atom catalysis of CO oxidation using Pt1/FeOx Nature Chemistry, 3 (8), 634-641 DOI: 10.1038/nchem.1095

Evidence for Rodinia

Picture from Wikipedia - Break up of pangaea into modern continents

Most of us are aware of Pangaea, earth’s super-continent that existed about 250 million years ago before breaking up into the present day continents. Geologists have known that there have been previous super-continents before Pangaea which broke up and coalesced back again.

The first super land mass was Columbia, formed about 2 billion years ago. Columbia broke up into fragments which accreted back together again about 1 billion years ago, forming the super-continent Rodinia. It was conjectured that about 750 million years ago, Rodinia broke up into several land masses including Laurentia, the core of the present North American continent, and Gondwana, consisting of most of the present Southern Hemisphere continents. Antarctica was thought to be entirely a part of the ancient Gondwana land mass. Now, a team of geologists in the USA, UK and Australia have found evidence that a part of present day East Antarctica, the Coats Land crustal block, was once part of the ancient Laurentia, bordering present day Texas.

The scientists compared the concentrations of different lead isotopes (Pb-204, Pb-206 and Pb-207) in the igneous (lava) rocks from Coats Land to that of the Red Bluff granites in the Franklin Mountains in Texas, and found remarkable agreement between the two. Rocks from the Keweenawan province of North America, near the great lakes, also had similar isotopic concentrations. The composition of lead isotopes in igneous rocks is determined by the initial lead concentration of its magma source, and from the radioactive decay of uranium and thorium. Interestingly, rocks from the Umkondo province in the Kalahari region of Southern Africa, which was earlier a part of Gondwana, had a discernably different lead isotope composition. This is clear evidence that the Coats Land crustal block had the same origins as Laurentia, but separate from Gondwana.

Incorporating previously obtained paleomagnetic data, the exact location of the Coats Land cluster block about 1 billion years ago was pinned down to the southwestern end of Laurentia, next to present day Franklin Mountains in Texas. It seems likely that the Kalahari region of the ancient Gondwana collided with Laurentia around 1 billion years ago, forming the super-continent Rodinia. The authors suspect that Gondwana broke away from Laurentia again about 500 million years ago, resulting in the Coats Land cluster block also breaking away.

It seems modern day Antarctica is made up partly from Laurentia and partly from Gondwana, the two pedestals of ancient Rodinia.


Loewy, S., Dalziel, I., Pisarevsky, S., Connelly, J., Tait, J., Hanson, R., & Bullen, D. (2011). Coats Land crustal block, East Antarctica: A tectonic tracer for Laurentia? Geology, 39 (9), 859-862 DOI: 10.1130/G32029.1

Rise of the planet of the aged

Picture from Wikipedia - Cerebral cortex (brain)

In the new movie ‘Rise of the planet of the apes’, actor James Franco plays a scientist trying to cure debilitating conditions developed in old-age such as Alzheimer’s. He ends up developing a virus which can cause mutations in apes enabling them to regenerate brain cells (neurons). Instead of finding a cure for Alzheimer’s though, it ends up catapulting apes above humans as the dominant species on the planet. Researchers at Yale University working with monkeys have now identified the neuro-physiological basis for decline of working memory with old age.

The pre-frontal cortex (PFC) is a part of the brain in primates that guides behaviour and thought using working memory, enabling us to remember things like ‘where are the car keys’ or ‘what was that phone number’. A network of neurons in the PFC start ‘firing’ (generating voltage spikes) when a ‘cue’ (some external stimulus or information) is received, and the firing persists for some time (many seconds) even after the cue is removed. This persistent firing is responsible for our working memory.

The Yale scientists trained monkeys of three age groups (young, middle-aged and old) to perform a spatial working memory task where they were shown a ‘cue’ at one out of eight locations for half a second, and then asked to remember the exact location after another 2.5 seconds, in order to obtain a reward (juice). They observed that the persistent firing rate of PFC neurons, during the delay period after the cue, decreased with increasing age.

PFC neurons have a high concentration of cAMP (cyclic adenosine monophosphate), a “messenger” molecule derived from ATP (adenosine triphosphate, the energy unit of the cell) used for intracellular signal transduction. An important function of cAMP is to activate proteins that act as channels for the flow of ions like potassium into and out of the cell. The authors suspected that this cAMP signaling and opening of ion channels might be interfering with the firing of PFC neurons, since it was known that cAMP activity is amplified with advancing age. To test this hypothesis, they administered drugs to the old monkeys in the vicinity of the PFC neurons which either inhibited cAMP activity or blocked the ion channels. Remarkably, in both cases they observed that the persistent firing rate of neurons, during the delay period after the cue, was almost restored to the same level as that of the young monkeys.

One of the drugs, guanfacine, is now being tested in elderly humans with cognitive defects related to the PFC. The potential to restore cognitive abilities of our rapidly ageing population could result in a paradigm shift in our view of human society.


Wang, M., Gamo, N., Yang, Y., Jin, L., Wang, X., Laubach, M., Mazer, J., Lee, D., & Arnsten, A. (2011). Neuronal basis of age-related working memory decline Nature, 476 (7359), 210-213 DOI: 10.1038/nature10243

One way traffic

Picture by Netweb01 (Wikipedia)

The diode is a basic two-terminal device which allows electrons to flow in one direction but not in the opposite direction. One would be hard pressed to find a modern electronic gadget that does not employ this elemantary component in its circuitry. It is envisioned that photonics, or computation using light zooming around in waveguides on an integrated chip, would soon replace electronics offering improved speed and lower power consumption. An obvious goal then, is to develop the optical version of the diode. Now, researchers at the California Institute of Technology and the University of California, San Diego have demonstrated such “non-reciprocal” light propagation in a waveguide on a silicon chip.

Traditionally, it has been very difficult to get photons to obey the rules of a one-way street. Exotic magneto-optic materials and large magnetic fields are required to exploit the “Faraday effect” of light and make it behave differently while travelling in different directions in a bulk medium. However, doing the same on nanoscale dimensions on a tiny chip has proved to be far from trivial. The recent work achieves non-reciprocal behaviour using conventional materials and standard nanofabrication technology used to manufacture computer chips today.

The key insight is to engineer a “complex” (here referring to the mathematical term) optical potential along the waveguide which couples to light traveling only in one direction. This is achieved by having periodic “bumps” of silicon and germanium/chromium along the waveguide, which offer a different array of scattering surfaces to light travelling in the forward direction as compared to the backward direction. For a loose analogy, imagine riding an escalator – it is easy to walk in the same direction as the escalator is moving, but much harder to move against its motion. Light travelling in the forward direction encounter the silicon bumps just before the germanium/chromium ones, and the same pattern repeats after another small propagation distance. On the contrary, for backward propagating light the germanium/chromium bumps appear just before the silicon ones, and again the same pattern repeats after another small propagation distance. 

The net effect is that the transverse spatial distribution of the light or the “optical mode” is preserved for one propagation direction and is modified for the opposite direction, and one can filter out a particular mode at the end of the waveguide.

Feng, L., Ayache, M., Huang, J., Xu, Y., Lu, M., Chen, Y., Fainman, Y., & Scherer, A. (2011). Nonreciprocal Light Propagation in a Silicon Photonic Circuit Science, 333 (6043), 729-733 DOI: 10.1126/science.1206038

Kim’s the name

Picture from Wikipedia - Distribution of Korean family names. Blue is Kim, Green is Lee, Orange is Park, Red is Choi, Purple is Jung, Gray is all others.

A fascinating statistical study of Korean family names has revealed that “Kim” has been on top of the popularity charts for the past one and a half milennia. Researchers at the Umea University in Sweden and the Sungkyunkwan University in Korea took data from special Korean family books which traditionally record the genealogical tree. They determined that the distribution of Korean family names is well described by the Random Group Formation (RGF) model, which also predicts the word-frequency distribution of novels written by an author to a very good approximation.

The RGF model captures the features of the group-size distribution when a large collection of objects is divided into a number of groups. In the present case, the persons are the objects and groups are formed by people sharing the same family name. The current population of Korea is about 48 million and there are about 250 distinct family names currently in use. The model assumes optimal mixing or the maximum entropy condition, which essentially implies that marriage between individuals of any two family names is equally likely and there is no segregation or isolation of any group.

The key prediction of the RGF model is that the group-size distribution is solely dependent on the total numer of objects, or population size. For example, the word-frequency distribution of a novel will only depend on the total number of words it contains. The RGF model is shown to predict the number of new Korean family names added and also fit the general family name distributions of the past 500 years very well.

Another feature of the RGF model is that the size of the largest group is always proportional to the total size of the entire data set. In the word-frequency case, this means that the occurence of the most common word in an English text – “the” – is proportional to the total number of words in the text. In the case of Korean family names, it implies that the frequency of the most common name – “Kim” – in a randomly selected group of Koreans should always be proportional to the size of the group chosen, irrespective of the historical time or the group/population size. This is verified to remarkable accuracy for group sizes varying over six orders of magnitude and over the time period 1500-2000 AD. The study suggests that about 20% of the population has shared the name “Kim” for the past 1500 years, and in 500 AD about 10,000 people out of the 50,000 population had the “Kim” family name.

The authors speculate that these findings points to some core stability in Korean culture that has remained intact for more than a thousand years. It is interesting to note that the fraction of people sharing other less common family names fluctuates quite a bit and is not always constant.

Baek, S., Minnhagen, P., & Kim, B. (2011). The ten thousand Kims New Journal of Physics, 13 (7) DOI: 10.1088/1367-2630/13/7/073036

This post was chosen as an Editor's Selection for