Shrinking in the heat – lanthanoid hexacyanidocobaltates, a.k.a. LnCo(CN)6

While most materials expand when heated, a few show the opposite behaviour, known as negative thermal expansion (NTE). The record holder, which has the greatest rate of volume contraction upon warming over a wide temperature range, is single-network cadmium cyanide (we previously blogged about the double-network version). This unusual phenomenon is useful – if you combine an NTE material with a normal material in just the right ratio, you can make a mixture with zero thermal expansion. Such a composite is immune to the undesirable effects of thermal expansion, from the buckling of railway tracks in the heat, to quartz-crystal clocks gaining or losing time when it’s too hot or cold.

A related group of frameworks which also show NTE are the lanthanoid hexacyanidocobaltates, with chemical formula LnCo(CN)6, where Ln can be any element from the lanthanoid row of the periodic table from lanthanum (La) to lutetium (Lu), Co is cobalt, and CN is cyanide. The cyanide is tightly bound to the cobalt, so these materials aren’t particularly toxic, unlike potassium cyanide.  They can be easily crystallised from a solution as shown in this time-lapse video, recorded over the course of a few hours.

LaCo(CN)6·5H2O is easy to grow, and forms nice hexagonal crystals. By the end of the video, they are about 2 mm in size and have started to merge together. Once the water is removed from the crystal structure by heating, the NTE properties are activated.

The key to the NTE is the networked structure of the materials. The lanthanoid and cobalt atoms are ‘bridged’ together by the cyanide (CN) ions. As the temperature rises, these linkers start to vibrate more and more in skipping-rope-like transverse (sideways) vibrations. Just like the ends of a skipping rope get closer together when you vibrate it, these transverse motions cause the average metal-metal distances to decrease, thereby causing contraction of the crystal.

The LnCo(CN)6 frameworks’ lanthanoid and cobalt metal atoms are linked together by cyanide bridges. ‘Skipping rope’ transverse thermal vibrations of these bridges bring the metal atoms closer together and cause the contraction of the whole material as it heats up.

The LnCo(CN)6 frameworks’ lanthanoid and cobalt metal atoms are linked together by cyanide bridges. ‘Skipping rope’ transverse thermal vibrations of these bridges bring the metal atoms closer together and cause the contraction of the whole material as it heats up.

Crystallography (powder diffraction in this case) allows the dimensions of the unit cell to be precisely measured, which lets us monitor the thermal expansion of materials by taking diffraction measurements at a number of different temperatures. The results for a series of LnCo(CN)6 compounds are shown in the plot below, and demonstrate that we can tune the NTE properties just by making the material with different lanthanoid metals. Swapping Lu for La doubles the steepness of the plot. This is because La bonds more loosely to the cyanide linkers than Lu does, making a more flexible framework with bigger transverse vibrations, and therefore bigger NTE.


The crystal structure of LaCo(CN)6 is made up of trigonal prismatic lanthanum (green) and octahedral cobalt (blue) metal atoms linked together by cyanide bridges (grey/blue). The plot shows the negative thermal expansion (NTE) behaviour of several different LnCo(CN)6 frameworks (with Ln = La, Sm, Y, Ho and Lu) measured using X-ray powder diffraction at the Advanced Photon Source. Larger lanthanoid metals, such as La, give a more flexible framework and a stronger NTE effect. The neutron diffraction measurements (black) allow us to go to lower temperatures, revealing how the NTE behaviour dies out, deviating from the nice straight line as the transverse vibrational modes “freeze out”. These results were published in S.G. Duyker et al. (2013), Angew. Chem. Int. Ed., 52: 5266–5270.

DNA Polymerase in Celebration of Evolution Day

Today’s Anton tells us about a structure behind one aspect of evolution.

What does it look like?

The Klenow fragment of DNA polymerase I from E. coli (in blue) with template and new DNA strands (green and magenta respectively).

The Klenow fragment of DNA polymerase I from E. coli (in blue) with template and new DNA strands (green and magenta respectively).

What is it?

Today is evolution day, the day where the publication of Charles Darwin’s seminal work On the Origin of Species in 1859 is celebrated. DNA and its specific sequence of bases: A, T, C, and G, have a lot to do with evolution but since we’ve already blogged about the structure of DNA for Rosalind Franklin’s birthday, I thought I’d write about the protein that replicates the DNA when a cell divides: DNA polymerase.

DNA replication is semi-conservative: that is, each strand of DNA acts as the template for the new copy and one of copy each goes to the two new daughter cells. There are billions of A, T, C, and G’s that make up a genome thus it is of the utmost importance that DNA polymerase is extremely accurate in its job. Typically DNA polymerase will make one mistake for every billon bases copied. There are a number of reasons why DNA polymerase is so accurate. 1. DNA polymerase can select the correct bases for the correct pairing (i.e. A pairs with T and G pairs with C); 2. There is a proof reading mechanism and 3. Repair systems if a mistake is spotted.

Many those one in a billion mistakes will never show as a unique characteristic but some could result in a random mutation that is seen as observable characteristics (the phenotype) in an organism. If that mutation is in the reproductive cells then the mutation could be passed onto the next generation and with the favour of natural selection propagate for many generations.

As a mistake in copying DNA can prove to be disastrous as well a beneficial there is a lot of redundancy in DNA replication with many organisms having more than one DNA polymerase. E. coli for instance has five DNA polymerases. The structure shown today is the Klenow fragment of E. coli DNA polymerase I. The Klenow fragment is the part of the polymerase where replication and proof reading occur but other domains have been removed. The structure resembles the shape of a hand with the DNA being held between a finger and thumb.

Where did it come from?

DNA polymerase I from E. coli was discovered by Arthur Kornberg in 1956 (and later won the Nobel Prize for Physiology and Medicine in 1959 for the discovery). The structure shown here was solved in 1993 [1] by Lorena Beese, Victoria Derbyshire, and Thomas Steitz (who won the Noble prize for Chemistry in 2009 with Venkatraman Ramakrishnan and Ada Yonath for their work on the ribosome). The structure was taken from the Protein Data Bank (PDB number: 1KLN).

[1] L.S. Beese, V. Derbyshire, T.A. Steitz (1993) Structure of DNA polymerase I Klenow fragment bound to duplex DNA. Science 260: 352-355.

JD Bernal and the structure of water

Can we understand the structure of a liquid?  Dr Alison Edwards’ celebrates one of the pioneers of X-ray crystallography and how he thought about the structure of the most fundamental of liquids, water.

What does it look like?

JD Burnal and his model of water.  Image re-produced from 'Burnal and the structure of water' by J.L. Finney

JD Burnal and his model of water. Image re-produced from ‘Burnal and the structure of water’ by J.L. Finney

J.D. Bernal is shown building a water (liquid) model in which randomness was introduced by using a number of different sized connecting rods to separate the spheres which depict the water molecules. This later work stemmed from Bernal’s insight that further work was required was due to his conviction that the proposed 1933 model was “too crystalline”.

What is it and where did the structure come from?

The structure of water was determined by the application of Bernal’s crystallographic understandings to a liquid system. This occurred in two bouts: the seminal work in 1933 with R.H. Fowler: A Theory of Water and Ionic Solution with Particular Reference to Hydrogen and Hydroxyl Ions, appears in the First volume of The Journal of Chemical Physics (p. 515-548). It has citations numbering in the thousands and was highly influential in the formation of the thinking about the role played by water in biological systems. Later work in the sixties refined and improved the fit of his more randomised proposed model to available data. An excellent account of the original and later work can be accessed here.

On the 113th anniversary of the birth of J.D. Bernal, we should celebrate the application of careful thought to the solution of interesting problems!

Bernal was a thinker and scholar of such high renown that he was known to his friends as “Sage” – there are well-documented stories of the breadth and depth of his knowledge both within science and more broadly in the human condition. His scholarly output was prodigious and eminently readable and the biographies and articles about him are entertaining. The Biographical Memoir of the Fellows of the Royal Society (DOI: 10.1098/rsbm.1980.0002 is freely accessible at ) and was written by one of his early students: Dorothy M.C. Hodgkin O.M. F.R.S. it opens:

“John Desmond Bernal lived his life to the full. He liked to say that his biography ought to be written in four colours, on interleaved pages, to show his different activities, fitted together. Oral tradition differs as to the colours of the pages; black and white, certainly for science, red for politics, blue for arts, purple or yellow for his personal life”

In a science where “big data” is now a dominant feature it is instructive to remember that very significant insights into matter can be achieved by thinking clearly about the limited data which were available.

Magnetic monopoles in the pyrochlore lattice

What does it look like?

Image generated by the VESTA (Visualisation for Electronic and STructual analysis) software

The red atoms here are oxygen, and blue are the transition metal with the purple atoms representing the position of the rare earth atoms. Image generated by the VESTA (Visualisation for Electronic and STructual analysis) software

What is it?

The pyrochlore crystal structure is a cubic that can be described by the spacegroup Fd-3m. The general formula for these materials are A2B2O and A2B2O7 where the A and B species are typically either rare earth ions, or transition metal species. The A and B sites form an array of interlinked tetrahedral, which are 3 sided shapes consisting of triangles and it is this that often leads to exotic magnetic properties.

Dy2Ti2O and Ho2Ti2O are two such pyrochlore systems which have been called “spin-ice”. This stems from the “2in-2out” magnetic spin structure which can be likened to the water-ice structure where for each oxygen ion, two protons (or hydrogens) must be in the near position and two in the far position. Dy2Ti2O and Ho2Ti2O are also part of a larger family of geometrically frustrated pyrochlores in which the magnetic “spins” cannot simultaneously satisfy their antiferromagnetic (antiparallel) spin arrangements with all of their neighbouring spins due to the geometric constraints, similar to Jarosite (Fig. 1).

(left) The rare earth tetrahedra structure of the pyrochlore lattice can be used to understand the magnetism in these materials. (right) The “3in-1out” (red) and “1in-3out” (blue) “monopoles” can be seen joined by an infinitely thin Dirac string [Morris et al., Science 326, 411 (2009)].

(left) The rare earth tetrahedra structure of the pyrochlore lattice can be used to understand the magnetism in these materials. (right) The “3in-1out” (red) and “1in-3out” (blue) “monopoles” can be seen joined by an infinitely thin Dirac string [Morris et al., Science 326, 411 (2009)].

Both Dy2Ti2O and Ho2Ti2O received a lot of attention in the media in 2009 due to the observation of the elusive magnetic monopoles as emergent quasi-particles. We all know that every magnet has a north and south pole and that if we cut these dipole magnets in half, we fail to separate the ends into monopoles, but instead create two new magnets each with a north and south pole. However due to the topology of the spin structure in Dy2Ti2O and Ho2Ti2O, under certain applied field and low temperature sample conditions, magnetic monopoles could be formed within the lattice.  The application of a small magnetic field could flip one of the spins such that one tetrahedron was a “3in-1out” (South) pole and its neighbouring tetrahedron was a “1in-3out” (North) pole (See Fig. 1). It was shown that with no additional energy input, these two “monopoles” could move away from each other such that locally there appeared to be an isolated monopole. In actual fact these two monopoles were connected by an extremely long, tensionless Dirac string. However the sensational news of observing a “magnetic monopole” for the first time became a hot topic on news websites around the world – even being mentioned in “The Big Bang Theory”.

Where did the structure come from?

This crystal structure was taken from van de Velde et al., Powder Diffraction 5 (4), pp229-31 (1990) and displays the pyrochlore structure for the related Tb2Ti2O7 material. Each of the A2Ti2O7 materials (A = rare earth) forms a cubic structure with a lattice parameter close to 10 Å.

Don’t Panic – Valium / Diazepam

Our latest post from Dr Dave Turner.

What does it look like?

Image generated with Crystal Maker software

Image generated with Crystal Maker software

What is it?

Inspiration for compounds to include in the Crystallography365 blog can be tricky and, in my experience, tends to reflect my mood when coming up with the next topic. The structure that leapt to mind today was valium – read into that what you will.

Diazepam, originally sold under the name valium since 1963 (and still commonly known as this), is a drug used to treat a variety of conditions such as anxiety, panic attacks, muscle spasms and opiate withdrawal amongst others. Diazepam is a member of the benzodiazepine family of psychoactive drugs which interact with the GABA neurotransmitter. The core structure of these drug molecules is the three rings that can be seen in the left-hand picture.

Crystals of diazepam are held together primarily by weak interactions between adjacent molecules in the absence of any charged groups or strongly polar, complementary functionalities.

Where did the structure come from?

The structure was determined and published in the Journal of the American Chemical Society in 1972 (A. Camerman and N. Cameran, J. Am. Chem. Soc., 1972, 94, 268-272), when the full range of properties of the drug was coming to light. Understanding the geometrical structure was important to determine if and how the stereochemistry of the molecule influenced its biological behaviour. Structural determination allowed comparison with other drugs with similar properties and characterisation by crystallography remains a key part in drug design.

Botox: Toxic medicine

What is it?

Botulinum toxin A –- more commonly known today as Botox – is a neurotoxin produced by the bacterium Clostridium botulinum. It acts on nerves at the point at which they meet the muscles that they control, where it disrupts transmission of chemical messenger molecules causing muscular paralysis. Botulinum toxin is one of the most potent human toxins and C. botulinum infection causes botulism, a serious paralytic illness. Thankfully botulism is rare. More common today is the deliberate localised injection of the toxin for cosmetic purposes, where its paralytic properties can be used to banish wrinkles, laughter lines, and the ability of professional actors to convincingly convey normal human facial expressions. Botox also has a number of important medical applications and is now routinely used to treat eye misalignment (strabismus), excessive sweating (hyperhidrosis), migraine and urinary incontinence, with no side effects upon acting ability yet reported.

What does it look like?

botoxBotulinum toxin A is made up of three functional domains which each play a distinct role in delivery and action of the toxin. The receptor-binding domain (yellow and red) is important for entry into the nerve cells where the toxin is initially encapsulated within a small membrane bordered compartment called an endosome. Next the toxin’s translocation domain (green) induces a pore in the compartment to allow transport of the enzymatic domain (blue) out of the endosome and into the cytoplasm of the cell where it cuts target proteins in the nerve to block chemical messenger release and induce paralysis. This image was generated using PDB ID 3BTA1 and the molecular graphics software PyMOL.

Where did the structure come from?

The crystal structure of botulinum toxin A was first published 1 in the scientific journal Nature in 1998, using toxin isolated from liquid cultures of Clostridium botulinum.

1. Lacy et al., Nature (1998) 5: 898-902


Collagen: Triple strength

collagenWhat is it?

The best designs – in architecture, engineering, software development – are those that effectively solve a functional problem. The same is true in biology, and scientists have long since appreciated that a protein’s design is intrinsically related to the particular function that it carries out. This idea is exemplified by collagen, a highly abundant protein in the body found in the connective tissue of tendons, skin and ligaments. Collagen has a triple helical structure in which three individual chains are wound together to form a tightly organized protein rope. This organization lends collagen a high tensile strength enabling it to function as a robust structural protein that provides a scaffold within tissue and connects it to bone.

What does it look like?

Collagen is made up of three polypeptide chains (red, green, blue) wound tightly in a triple helical structure. The individual polypeptide chains have a very distinctive and unusual sequence in which every third amino acid is a glycine (the smallest amino acid) that faces into the interior of the protein rope. This pattern allows the three chains to pack snugly against each one another without clashing. Collagen also has an unusually high proline content. Proline is a rather geometrically constrained amino acid and its abundance helps the individual chains to spontaneously adopt a stable left handed helical conformation.

Where did the structure come from?

This structure is of a model peptide of Type III collagen that assembles as a homotrimer. It is a small part of the much larger protein that naturally exists in the body but it reveals the triple helical rod like structure of the protein. It was reported in Nature Structural Biology in 1999 and is PDB ID 1BVK1.

1. Kramer et al., Nature Struc. Biol 1999. 6:454-457