Self-assembly is finding applications in everything from space to medical implants.
If humanity is going to fulfil its dreams about space — exploring it, mining it, colonising it — then we need to learn how to build in it. The kinds of large structures required for launching deep-space missions could be made much simpler if we could construct them in orbit rather than packaging them up for launch from Earth. And to build in one of the most extreme environments we know, we’ll have to take human involvement out of the picture. Automated construction equipment is one solution. But there may be another possibility: self-assembly.
‘Solar power is a huge problem for communication satellites,’ says Kim Ward, head of space engineering and technology at Harwell-based research centre RAL Space. ‘The traditional way of doing solar panels is they all have to be joined together and have a fancy deployment. But if you could have 30 or 40 free-flying solar panels that joined together then you could assemble huge panels. And if they’re really smart, they could get hit by a bit of space debris but then reconfigure and regenerate.’
RAL Space is working with Qinetiq Space and technology firm Magna Parva on a proposal for the European Space Agency (ESA) to develop what it calls a “hex swarm”, a fleet of seven dinner plate-sized hexagonal spacecraft that could be launched into space in a stack and then autonomously join together to form a larger structure. To convince ESA it will work, Ward’s team is building a test rig to that can demonstrate how the craft will use locate and then dock with each other.
Ward calls the hexagonal craft “worker bees”, referring to their fairly limited capabilities: they would align using the Sun as a reference point, find each other using infrared signals and rendezvous with one of two techniques RAL Space is testing (and not yet revealing). But they would be overseen by a “queen bee” that could communicate with them and with mission control.
As well as forming the basis of large structures like solar panels or communication dishes, the craft could share power or computing resources. ‘The whole idea of assembling structures on the Moon or Mars is equally relevant,’ says Ward. ‘In a sense it’s easier because you’re doing it in two dimensions, but you still have the same problem of things moving around that join together to make potentially very large structures.’
Several studies over the last few years have proposed similar self-assembling or reconfigurable spacecraft concepts, although Ward claims RAL Space is the first group to run a demonstration project. Another recent idea put forward by Skylar Tibbits, founder of the Self-Assembly Lab at MIT, is to build components made from specially designed materials that can change their shape and purpose in space, for example from a solar panel to a parabolic antenna.
‘If you can reconfigure from one state to another and make extremely functional systems in between then you have more robust scenarios,’ he says. ‘Right now, a lot of the technologies are one-off and so you ship up a lot of equipment, you need human space walks and construction up there.’
Tibbits points out that this idea is different from that of true self-assembly — there’s no coming together of constituent parts. And RAL Space’s vision of robotic components that link up is far from the principle of natural molecular self-assembly used to build nanostructures. But what they all share is the notion that adding an aspect of autonomy to a structure or material and removing human intervention can increase its functionality. And this concept is increasingly being employed across several other areas of manufacturing.
‘If you can reconfigure from one state to another and make extremely functional systems in between then you have more robust scenarios’
Skylar TIbbits, MIT Self-Assembly Lab
Scientists have long looked enviously at the efficiency with which nature manufactures things. The processes by which proteins in living cells are formed, folded and combined don’t require external energy or direction and are instead driven by the natural movement of molecules and the various forces between them. The other advantage of this molecular self-assembly is the complexity of the structures that can be formed at the nanolevel, something that traditional “top-down” manufacturing techniques cannot replicate.
For this reason, molecular self-assembly has become a popular technique for scientists experimenting with adding special functions to material surfaces or creating tiny structures with incredible properties. For example, a team at the University of Sheffield is using it to develop self-propelled nano-devices that can deliver drugs to a tumour within the body (see box).
Self-assembling nano devices seek out tumours
Drug delivery has already become pretty sophisticated. Using magnetic fields or other stimuli, doctors can guide tiny packets of cancer drugs through the body to the site of a tumour, meaning other organs aren’t exposed to their toxic effects. But scientists at Sheffield University want to go a step further and create self-assembling autonomous nano-devices that guide themselves to the tumour by following the trail of chemical signals it releases into the bloodstream.
The devices are based on spherical polymer molecules each with a catalyst patch to drive a reaction that creates thrust. These particles naturally self-assemble into groups that form the structure of the drug delivery device. ‘The neat thing about it is depending on how they join together you get different types of motion,’ says research leader Dr Stephen Ebbens. ‘So if they join with the catalysts perfectly aligned they’ll move in a straight line. If you have an offset you’ll get devices that spin around. And this all emerges from this natural self-assembly process.’
If Ebbens’ team can control the self-assembly process by changing the molecules’ design in a certain they will be able to tailor the devices to move in the desired way. ‘One of our ideas is to grow a long chain where the individual units are flexibly linked to make a snake-like structure,’ he says. Alternatively, making larger round devices that spin could be used to mix substances in microfluidic devices more precisely.
But this kind of self-assembly isn’t quite yet an option for commercial mass production. That’s why microchip companies are hoping to instead use the concept to make manufacturing tools that could produce the next generation of electronic components at much tinier scales than is currently possible — thereby helping to maintain the “Moore’s Law” development of ever-faster processors.
This method of “directed self-assembly” (DSA) relies on materials known as block copolymers, chains of at least two types of monomer molecules that are grouped into sections. The different monomer groups are linked but naturally want to separate, causing the molecules to arrange themselves into specific shapes — spheres or cylinders for example — depending on their structure. These copolymers can be used to create lithographic masks with features smaller than the wavelength of light, which in turn could mass-produce chips with much finer detail than those made by traditional photolithography.
Until last year, researchers had only been able to create two-dimensional masks, but a team from the Massachusetts Institute of Technology (MIT) has developed a way to build several layers of self-assembled structures to produce more complex 3D configurations. The technique involves dissolving the polymer and spin-coating it onto a silicon substrate covered with tiny posts that repel the polymer and so guide the self-assembly process.
‘We can put a set of cylinders going in one direction and above it a layer of cylinders going in the other direction and have junctions in between the two,’ says Prof Caroline Ross, who has been working on DSA since 1997. ‘So you can imagine making a three-dimensional structure in one go rather than building it up layer by layer.’ Ross’s team have also been able to use DSA to produce square shapes typical of existing electronics design, rather than the hexagonal shapes the self-assembled molecules naturally want to form.
For DSA to be used in commercial manufacturing, companies need to find ways to scale and speed up the technique and reduce the number of defects that occur. But over a decade of research like that at MIT has reportedly helped push the concept onto the R&D agenda at major firm such as Intel. And molecular self-assembly has also started to work its way into other areas of manufacturing research, not just in the creation of tiny devices but also in the production of materials themselves. A team at Manchester University, for example, is using the principles of natural self-assembly to mimicking the DNA building process in order to create a more efficient chemical manufacturing process (see box).
Molecular machines mimic DNA building process
Proteins and other molecules in living cells don’t form spontaneously but rather are constructed by tiny “molecular machines” that grab small molecules and join them together in specifically ordered chains. Researchers at Manchester University led by Prof David Leigh hope that by replicating this concept with their own molecular machines, they can produce a new, more efficient way of synthesising chemicals that doesn’t rely on multi-step batch processing that uses up lots of energy.
Their first machine is based on the ribosome that links amino acids into proteins according to the code laid out in DNA. The machine’s movements are driven by the random motion of molecules and its structure means that each stage of the process occurs in the right order. In this way, molecular manufacturing is another example of taking the human intervention out of the process and instead using a form of self-assembly. But the complexity offered by such a precise and programmable manufacturing method means the obvious application for it might be in creating new materials rather than replacing current techniques for making existing chemicals.
Self-assembly is even being used to create substances that mimic some of the properties of biological tissue but without the difficulties created by it actually being alive. Scientists have been able to coat water droplets with a membrane of fatty molecules known as lipids to approximate cells and, because the lipids have both water-loving and water-repelling parts, when the droplets are surrounded by oil they self-assemble into a specific structure. Inserting other biological molecules into the membranes gives these “cells” useful and controllable properties, for example conducting ionic currents or transducing light into electricity. Unlike living tissue, however, these droplet structures don’t need a supply of oxygen or food and they can’t die or grow uncontrollably.
A team from Oxford University recently developed a means to manufacture these structures on a larger scale by building their own version of a 3D printer that injected the water droplets into a bath of oil and lipids in a pre-detemined pattern. Gabriel Villar, one of the researchers and now an employee of technology transfer firm Cambridge Consultants, says the technique could be used to print functional structures for biocompatible medical implants, without the hassle of printing and maintaining fully living tissue.
‘By including the right kinds of biomolecules and chemicals you might be able to get these droplets to talk to a living tissue in a very well defined way,’ he says. ‘With living cells it’s very difficult to tell exactly what they’re going to do. One big problem is to stop cells from becoming carcinogenic, from creating tumours in a printed structure, or differentiating into the wrong kind of cell.’
Again, the advantage of these bio-inspired, self-assembled structures over simple electrical machines is their complexity, he adds. ‘The power in using biological processes is you can really manipulate matter much more generally. You can use them to control the flow of matter or to construct new molecules in a controlled way or to analyse or detect a presence of a specific molecule in a solution or in the air.’
The development of 3D printing could be important in scaling up production of self-assembling materials and structures but also in applying the principles to much larger objects. Skylar Tibbits of MIT’s Self-Assembly Lab has actually coined the term “4D printing” to describe the use of 3D printers to make objects that can change shape or function after they’ve been printed, although again he stresses that these items are transformational rather than truly self-assembled.
Using a Stratasys Connex printer that can make structures from two different polymers and experimental Autodesk design software known as Project Cyborg, Tibbits has been able to create a plastic string that very slowly folds up into a pre-programmed shape when placed in water. Its joints are made from a polymer that expands by 150 per cent when wet, and combining that with sections of a rigid, non-absorbing plastic creates an object that can change its shape without the need for motors and actuators — just the ambient energy input of submerging it.
You can imagine how this idea might be used to produce bigger strings or flat sheets that can fold themselves into large objects such as furniture without the need for human assembly, just as protein molecules fold themselves into shapes that give them functionality. But Tibbits has more unusual concepts in mind, not just the transforming space structures but also pipes that change their shape to reduce water flow or even that grow and constrict rapidly to pump water along — an idea he is working on with Boston-based engineering firm Geosyntec.
However, it’s not entirely clear how much of a role 3D printers and additive manufacturing will play in spreading self-assembly. For example, Tibbits questions whether 4D printing will actually be the technique he uses to produce the transforming pipes because of its size limitations.
‘Right now you can print up to a few feet in all dimensions on the biggest of the Connex machines and a lot of people are looking at how we can scale up printing,’ he says. ‘In the future we might be able to use a process like that for the pipes but it’s more about the mindset that you can program these materials in elegant ways to transform.’
‘The multi-functionality and the self-assembly comes out of the freedom of design that you can only get from an additive approach.’
Richard Hague, Centre for Additive Manufacturing
At the other end of the scale, 3D printing isn’t yet precise enough to manipulate molecules, for example to set up self-assembly for electronics manufacturing. ‘With 3D printers you’re looking at tens of microns and it’s a different magnitude,’ says Caroline Ross. On the other hand, the researchers at Oxford University overcame several issues to use 3D printing technology to build their water-droplet networks. They used an extruded glass tube the width of a human hair to place the droplets precisely and developed control software to compensate for the way this print head would drag the droplets after it had printed them.
Richard Hague, director of the EPSRC Centre for Additive Manufacturing at Nottingham University, argues that self-assembly essentially is a form of additive manufacturing. ‘Most additive manufacturing is at the micro scale really, so moving to the nanoscale throws up a whole bunch of different problems and you’re looking at a different bunch of physics.’
But, he adds, there is immense possibility for the technology as it progresses. ‘With the additive approach the most important thing is the design freedoms you get, the ability to make these complicated parts. The multi-functionality and the self-assembly comes out of the freedom of design that you can only get from an additive approach.’