skip to content

Department of Chemical Engineering and Biotechnology


Researchers from our Computational Modelling group have published a possible solution to why disordered carbon structures are reluctant to turn into graphite, a puzzle that perplexed Rosalind Franklin before her discovery of the structure of DNA.

Carbon exists in many different structural forms, the most commonly known being diamond, graphite and its two-dimensional analogue, graphene, but there are many other forms with less regular structures that give rise to different properties.

Disordered three-dimensional graphenes may sound exotic, but are just another form of carbon, found all around us. They are the carbon materials present in batteries, water filters, gas masks, high-temperature ceramics, electrochemical sensors and insulation. They also have more specialised uses, such as protecting the Parker solar probe spacecraft from burning up on its approach to the sun.

Rosalind Franklin, the scientist who would later deduce the helical geometry of DNA, first discovered this class of materials in 1951. While she is best known for her crystallography work that led her to discover the DNA double-helix, Franklin was initially a carbon scientist, investigating the porosity of carbon structures for their use in gas masks during the Second World War.

Most carbon-containing materials develop small, layered regions of graphene – flat sheets of carbon atoms arranged in hexagonal rings – when heated. Upon further heating, to thousands of degrees, Franklin found, to her surprise, a complete reluctance of the carbons to convert to the most stable form of carbon, graphite.

Explanations for this reluctance to graphitise have centred around either crosslinks within the structure, knotted ribbon-like structures or warping of the sheets into bowl- or saddle-shaped geometries. However, experiments were unable to resolve and combine these suggestions into a coherent model of the nanostructure.

Researchers from our Computational Modelling group in collaboration with Curtin University have now published a possible solution to Franklin's problem in Physical Review Letters. They turned to large scale simulations using Australia’s Pawsey supercomputer to self-assemble the largest and most accurate networks of disordered 3D graphenes to date.

They developed a new measure for the global curvature of the networks and found that for all densities, excess saddle-shaped graphene sheets are present. These saddle shapes are caused by the integration of seven- or eight-membered rings within the hexagonal graphene network. This warping allows it to connect in 3D and the researchers suggest it is the cause of the material's resistance to convert into graphite.

Image caption: New nanostructure proposed for disordered 3D graphenes with bowl, saddle and ribbon-like graphene sheets. Increasing density screw dislocations allow for winding up and layering of the network.

“What we discovered is that you've got these two extremes, on one side you've got this very porous structure, but as you increase the density you get to something that's completely the opposite,” says Jacob Martin, a PhD student in our Computational Modelling group who led the work. “We wanted to figure out how you go from something that's incredibly porous to something that's very non-porous and has these stacked layers.

“That’s what puzzled Rosalind Franklin, she talks in her paper about how on earth do you get these graphitic regions that somehow grow in size from a disordered network. You would think that if it's disordered, you'd just have everything jumbled around, you wouldn't get these areas of stacking.

“What we discovered is that as you go up in density, the sheets have to condense somehow and the way that happens is they crack open and you get a screwing up of the layers in a spiral staircase. You have a spiralling upwards, and that actually fills in all of those pores and allows the sheets to stack while also being connected in this beautiful geometry.”

This screw or helix defect is well known in graphite but has not been suggested in these disordered materials. A variety of other defects were discovered, which resolve many issues of the graphene network being both curved and layered.

Image caption: Three of the five defects observed in the disordered 3D graphene.

These results open up possibilities for understanding and engineering carbon materials for applications in supercapacitors, carbon fibres and high-temperature ceramics applications. However, more work is needed to experimentally confirm some aspects of the model.

The researchers suggest that carbon materials could be topologically tuned and optimised for a given product. Of particular industrial importance to making batteries and electrodes, as well as the growing area of graphene science, is the question of how you could guide a disordered carbon compound towards becoming graphite.

“Most materials will turn into non-graphitising carbons, it's incredibly rare for a material to turn into graphite, so that's why we mine most of it from the ground,” says Martin. “One of the issues with graphene science for example is that you have to get the best graphite, often now from a mine.

“So there's a lot of science on how to make graphite synthetically and this gives us the first indication of what's going on. It appears that it's connected in three dimensions and as long as it's being held in 3D it won't turn into graphite, so the question is how do you make it only connect in two of those dimensions. We've got some experimental ideas on how to explore and exploit that.

“There is a pleasing connection with Franklin's later work on DNA in that the solution to her earlier problem of non-graphitising carbon materials could also lie in topology and the famed helix structure.”

Read the full paper: Topology of Disordered 3D Graphene Networks, JW Martin, C de Tomas, I Suarez-Martinez, M Kraft and N Marks, Physical Review Letters, 2019

Leading Image caption: Disordered three-dimensional graphene network (1.5 g/cc, a similar density to charcoal). Shown as a surface mesh constructed from the graphene rings with the curvature coloured red for saddle shapes, blue for bowl shapes.


Louise Renwick

Communications and External Affairs Executive, Cambridge Centre for Advanced Research and Education in Singapore

+65 6601 5447

Authors available for comment

Jacob Martin, PhD student

University of Cambridge

Email |

Tel | +65 8618 6990

A/Prof Nigel Marks, Academic, Physics & Astronomy

Curtin University

Tel | +61 8 9266 1386

Mob | 0402 379 209

Email |

Web | 

Group |

Latest news

Cell ‘membrane on a chip’ could speed up screening of drug candidates for COVID-19

6 July 2020

Researchers have developed a human cell ‘membrane on a chip’ that allows continuous monitoring of how drugs and infectious agents interact with our cells, and may soon be used to test potential drug candidates for COVID-19.

New heart valve could transform open heart surgery for millions of patients globally

29 June 2020

The new polymeric heart valve, developed by scientists at the universities of Bristol and Cambridge, prevents the need for life-long blood thinning medication.

Shaping the world

23 June 2020

CEB researchers Beth Tennyson and Jana Weber named in WES Top 50 engineers awards