DAVID GRIER is a choreographer with a difference. He doesn鈥檛 bark orders at a group of dancers on a stage. As a physicist at the University of Chicago, Grier鈥檚 ballet is on a far smaller scale, using holograms and lasers to twirl objects just a few micrometres in diameter. Peer through Grier鈥檚 microscope and you would see a Lilliputian dance troupe 鈥� hundreds of glass beads tumbling in formation, doing an eightsome reel or shimmying through a bath of solution.
Grier鈥檚 precision choreography is on the point of revolutionising how scientists handle objects on the microscale. He has invented computer-generated holograms that can split laser light into any number of different beams and use them to move cells around at will. Hundreds of shafts of light could soon craft microscopic cogs and levers and then assemble them into micromachines, or sort chromosomes in record time. One day they might power micromachines themselves, spinning their gears on a light beam. 鈥淚t鈥檚 animating matter with light,鈥� says Grier.
The basic techniques used are not new. Arthur Ashkin and Steven Chu, physicists at Bell Laboratories in New Jersey, invented 鈥渙ptical tweezers鈥� in 1986. These harness the forces that light exerts on partially transparent objects to grip and move all manner of tiny things. Since then, biologists and physicists have used optical tweezers to measure the stretchiness of DNA, probe the molecular motors that power bacteria, and gauge the forces between tiny particles in solution.
Advertisement
The forces that trap a particle in optical tweezers arise from the variation of light intensity within a laser beam. The intensity of a standard laser beam is greatest along the beam axis and falls off towards the edges. This ensures that any transparent object caught in the beam experiences a small force pulling it inwards towards the brightest point at the centre of the beam.
This force can be explained by refraction. When a laser beam hits a particle slightly off-centre, refraction bends more light outwards, away from the axis of the beam, than inwards (see Diagram, p 35). So on average the refracted photons gain momentum in that direction, and because momentum is always conserved, the particle gains an equal amount of momentum in the opposite direction and moves towards the centre of the laser beam. If it overshoots, the bulk of the laser light hits off-centre on the other side of the particle, and is refracted in the opposite direction, pushing the particle back towards the centre.
That鈥檚 fine for trapping in two dimensions. But optical tweezers can also defy gravity and hold particles aloft. To do this, the laser light must be tightly focused by a lens down to a point just above the particle. The photons diverging from this focal point then hit the particle and are refracted downwards, gaining momentum in that direction. As a result, the particle gains momentum upwards and hovers at the focal point. So the particle is trapped in three dimensions.
Optical tweezers have given scientists exquisite control over microscopic objects. To move a particle sideways, they simply shift the laser beam. And to set the particle spinning, they use an 鈥渙ptical spanner鈥�, a laser beam in which the energy spirals along the beam rather than travelling along the beam鈥檚 axis (快猫短视频, 14 February 1998, p 34). But despite the flood of experimental results, scientists are still frustrated.
The trouble is that they can only generate one or two tweezers at a time in a given microenvironment. While that is fine for moving one or two cells or grabbing both ends of a strand of DNA, it is painfully slow when you have lots of cells. 鈥淚f you want to do more sophisticated manipulations, roll things around and turn them over, you start wishing that you had more traps,鈥� says Grier.
This limitation began to annoy Grier in 1992 when his research hit a brick wall. He and his colleagues had been looking at the behaviour of colloids 鈥� suspensions of microscopic particles 鈥� and they were attempting to measure the forces involved when you bring together charged polystyrene spheres in water. The physics seemed straightforward: two negative charges should repel each other, while a negative and a positive should attract.
To study the effect in detail, Grier鈥檚 team managed to split a laser beam to make two optical tweezers. Grabbing a polystyrene sphere with each tweezer, Grier nudged them together and then released them. The charged spheres behaved exactly as he expected 鈥� those with like charges swam away.
But something strange happened when Grier added more polystyrene beads with the same charge. To his amazement, instead of repelling they attracted each other. The result baffled theorists. Even today, no one can explain this bizarre behaviour, says Grier, but he believes he is close.
To help tease out an explanation, Grier needed to assemble clusters containing large numbers of particles. He knew that other groups had managed to create multiple tweezers by rapidly scanning a single optical tweezer from one point to another, by means of a rotating mirror. But these scanning tweezers do not work well with sensitive materials or when forces on individual particles are important, as they were in Grier鈥檚 colloids. 鈥淓ach time the beam arrives it gives the sample a smack,鈥� Grier explains. 鈥淵ou also need a very strong beam since you only visit each site for a short time.鈥�
Grier and his student Eric Dufresne, now at Harvard University, mulled over the problem of making multiple optical tweezers for a long time. They believed a multi-tweezer would make all sorts of experiments possible, but they were stumped. Then one day in 1997 Dufresne was browsing through a catalogue from a company called American Science & Surplus whose motto is 鈥渋ncredible stuff at incredible prices鈥�. It鈥檚 the kind of place you go to buy a bag of surplus glass lenses, Petri dishes, tractor batteries or even a radio-controlled rat.
鈥淓ric found these holograms you shine a laser pointer through and get 鈥榶our logo here鈥�,鈥� recalls Grier. One hologram stood out because it produced a four-by-four array of spots from a single laser beam. Dufresne realised this was the breakthrough he needed. By aiming the 16 laser beams that emerged from the hologram through the lens of a microscope, he might be able to trap 16 particles at once.
But Grier didn鈥檛 believe it would work. 鈥淭here was absolutely no way this was going to work,鈥� he remembers. 鈥淥ptical tweezers are really, really sensitive to distortions in the beam. They smear out the focus, they reduce the intensity variations and therefore reduce your ability to trap the particle.鈥� Even worse, the toys were designed to work with red light, and Grier鈥檚 optical tweezers use green lasers because they produce less heat, which can damage delicate cells. Despite his scepticism, he ordered a hologram. 鈥淭hey only cost five bucks, so we got one and put it in the right place, and it worked. First time. It was really surprising,鈥� he says.
It did not take Grier and Dufresne long to realise that they were onto something big. What sets a hologram apart from a normal photograph is that it records both the intensity and the phase of light, rather than just the intensity. To make a hologram, you split a laser beam into two and project one beam directly onto a photographic plate, while bouncing the other off an object nearby. The plate then records the interference pattern that results when the two light beams recombine. When light is shone onto the plate, it produces an exact image of the original object in 3D. But Grier and Dufresne knew that they did not need real photographic plates or even real objects to produce 3D images. They could generate them on computers instead.
Central to their idea is a device called a spatial light modulator (SLM), which works in much the same way as a computer鈥檚 liquid crystal display. Instead of changing colour when an electric field is applied, the pixels in an SLM become thicker. Laser light passing through a thicker pixel will take slightly longer to reach the other side, and this creates a tiny shift in the phase of the light. By altering the phase of light pixel by pixel, Grier and Dufresne realised they could create exactly the interference they needed to create multiple optical tweezers.
And because the holograms are generated on a computer, it is easy to manipulate them. Do it fast enough and you can move the light beams around at will in much the same way you can make an animated cartoon by flicking quickly through a series of drawings. What鈥檚 more, holograms produce a three-dimensional image, so you can focus the light beams and direct the tweezers anywhere you like.
It might sound straightforward, but the hard part is working out the interference pattern that will split the light into many beams and focus them at exactly the right spots. Indeed, it took Grier and his colleagues a few years to develop the sophisticated computer algorithms needed to generate the holograms, but the finished instrument is actually pretty simple. It requires little more than a computer, an SLM, a microscope and some optics (see Diagram), and can produce up to 200 tweezers at once. Grier has even set up a company called Arryx, which has managed to cram all the optics, which cover an entire bench in his laboratory, into a device that fits inside a microscope.
This ability to manipulate so many micrometre-sized particles at once could be invaluable in many areas of research. One of the first people to buy holographic tweezers from Arryx was Wolfgang Losert, a physicist at the University of Maryland in College Park. Although Losert and his colleagues are still getting to grips with the new technique, they are already using the tweezers to try to control the growth of microscopic crystals.
At present it is impossible to grow two identical crystals, due to random microscopic variations. But Losert is hoping to change that. He plans to monitor crystal growth through a microscope and intervene as soon as he spots a blemish on the embryonic crystal. He hopes holographic tweezers will allow him to direct a beam of light to heat the microscopic pattern and nudge it to grow the desired way. The ultimate goal is to create identical crystals for use in micromachines.
And this is just the first of many potential applications for holographic tweezers. Losert recently managed to pick up a vesicle 鈥� a large membrane-bound sac 鈥� by clamping 20 independent tweezers along its periphery. It is a task that would have been impossible with a single tiny tweezer. Now his group is planning to investigate the physical properties of such structures. 鈥淚t鈥檚 a huge improvement in the sense that you can ask new questions,鈥� says Losert. 鈥淧eople have used single laser tweezers to measure forces between two objects and stretch cells in one plane. Now you can actually hold a small object with one tweezer, deform it with another and bring something up to the object with a third tweezer.鈥�
Losert has already starting pushing optical tweezers further, exploring an effect discovered earlier this year by Grier and his colleague Jennifer Curtis. By manipulating the phase of the light in each beam, Grier and Curtis found they could give the beam a sort of twist. These twisty beams can transfer some of their momentum to the objects in their grip and make them race around in circles.
This is a big improvement on traditional optical spanners, which spin objects on the spot. And scientists are still dreaming up uses for this new ability. So far Grier has been using arrays of spinning glass beads to pump fluids around tiny lab-on-a-chip devices that are being developed for high-speed chemical analysis (快猫短视频, 26 April, p 21). And he envisions twisty beams one day spinning the tiny gears in micromachines.
Beyond simply controlling micromachines, holographic optical tweezers could even be used to sculpt them. Several groups have recently shown that the intense light at the focus of an optical tweezer can be used to harden a liquid polymer. By moving the tweezer around from point to point, a group of Japanese physicists led by Satoshi Kawata at Osaka University has created springs, gears and even a sculpture of a bull the size of a red blood cell. Building these structures pixel by pixel is painstaking work, but Grier believes that with 200 or more optical tweezers operating simultaneously, the technique will become dramatically faster and more practical.
It seems like every time Grier鈥檚 group tries something new, it opens up a new possibility. Recently he asked students in his group to study the seemingly unremarkable task of seeing how particles would scatter off a lattice of optical tweezers. They discovered that the traps deflected the particles at different angles, depending on how strongly they interacted with light, which in turn depends on the size of the particle. Bigger particles experience greater forces, so an array of light traps can spit out a 鈥渟pectrum鈥� of particle sizes in much the same way as a prism can disperse the components of a beam of white light. 鈥淭his size sensitivity sets optical sorting apart from all other sorting techniques鈥�, says Grier.
Already Grier鈥檚 group has used the technique to separate young yeast cells, which are small, from larger old cells. They have also been experimenting with using holographic tweezers to sort chromosomes and other cell components. 鈥淚t鈥檚 the basis of a whole new industry,鈥� says Grier.