It’s the most famous corkscrew in history. Now an electron microscope has captured the famous Watson-Crick double helix in all its glory, by imaging threads of DNA resting on a silicon bed of nails. The technique will let researchers see how proteins, RNA and other biomolecules interact with DNA.

The structure of DNA was originally discovered using X-ray crystallography. This involves X-rays scattering off atoms in crystallised arrays of DNA to form a complex pattern of dots on photographic film. Interpreting the images requires complex mathematics to figure out what crystal structure could give rise to the observed patterns.

The new images are much more obvious, as they are a direct picture of the DNA strands, albeit seen with electrons rather than X-ray photons. The trick used by at the Italian Institute of Technology in Genoa, Italy, and his team was to snag DNA threads out of a dilute solution and lay them on a bed of nanoscopic silicon pillars.

The team developed a pattern of pillars that is extremely water-repellent, causing the moisture to evaporate quickly and leave behind strands of DNA stretched out and ready to view. The team also drilled tiny holes in the base of the nanopillar bed, through which they shone beams of electrons to make their high-resolution images. The results reveal the corkscrew thread of the DNA double helix, clearly visible. With this technique, researchers should be able to see how single molecules of DNA interact with other biomolecules.

DNA cords

But at present, the method only works with “cords” of DNA made up of six molecules wrapped around an seventh acting as a core. That’s because the electron energies are high enough to break up a single DNA molecule.

Using more sensitive detectors that can respond to lower-energy electrons should soon allow the team to see individual double helices, and even unwound single strands of DNA. “With improved sample preparation and better imaging resolution, we could directly observe DNA at the level of single bases,” says di Fabrizio.

Earlier this year a University College London team led by felt their way along individual strands of DNA using the Braille-like technique of atomic-force microscopy (). Like the Italian team, they were able to detect the twisting groove that separates the twin strands of the double helix.

Journal reference: , doi.org/jt3

Erwin Schrödinger, an Austrian physicist, conjured up his famous cat in
1935 to illustrate the difference between classical and quantum physics. In the
quantum world of atoms, it is possible for a system to be in two states at the
same time. Not so in familiar classical physics.

Schrödinger’s cat is hidden from observers in a box. In the box is a
radioactive atom that kills the cat if it decays, but spares it if it does not.
Since the atom is in a quantum “mixed” state, both decaying and not decaying,
the experimenter must assume that the cat’s state is also mixed—both alive
and dead.

The problem is that the mere act of observing atoms changes their
state. As soon as the observer looks in the box, the cat must be either dead or
alive. The act of switching from a mixed state to a determined state is known as
decoherence.

Now, however, Serge Haroche, Marc Brune and Jean-Michel Raimond at the Ecole
Normale Supérieure in Paris have overcome some of the problems. The team
constructed a microwave field enclosed by superconducting mirrors. They then
took single atoms in either of two electronic states, e or g, and fired them
through the field. The idea is that, as a single atom passes through the field,
it changes the quantum character of the field. Because the two states of the
atoms alter the microwaves in different ways, the field ends up in a mixed
state.

The researchers found that, by firing one atom in and then a second one close
behind, they could glimpse the microwave field in its elusive mixed state, and
measure the time, just tens of microseconds, before it decohered. (Physical
Review Letters, vol 77, p 4887). The first atom changes the field’s state;
the second interacts with the field’s state, and shifts its own state. When it
comes out, the second atom’s state is compared with the first, yielding clues
about the nature of the field.

According to quantum mechanics, all particles, including atoms, have
wave-like properties. This idea is already exploited in electron
microscopes, which use beams of electrons to produce images with a much higher
resolution than is possible with light microscopes. More recently, physicists
have turned their attention to atoms, bending and diffracting beams of atoms
just like beams of light (“Atoms through the looking glass”, New
ĐÓ°ÉÔ­´´, 20 April, p 30).

In conventional holograms, such as the security image on credit cards, light
waves pass through the transparent areas between patterns of dark lines. This
arrangement acts as a diffraction grating, causing the waves to interfere with
one another so that some cancel each other out and others add together to create
more intense light. The result is a recognisable image.

To create a diffraction grating for beams of atoms, Junichi Fujita and Shinji
Matsui of NEC, working with researchers at the University of Tokyo, punched a
computer-generated pattern of holes in a thin sheet of silicon nitride. The
pattern was calculated to diffract the neon atoms so that they would write the
letters “NEC” when they hit a surface. In principle, the holes could be punched
to create any pattern.

In all, the printed letters were made of just 52 000 atoms, accumulated over
two hours, the researchers report in the 29 July issue of Physical Review
Letters (vol 77, p 802). The edges of the letters were slightly blurred,
spread over about 65 micrometres. But with a finer pattern of holes cut
into the silicon nitride mask, it should be possible to produce atomic holograms
with a resolution of just 10 nanometres—six thousand times sharper.

David Pritchard of the Massachusetts Institute of Technology, a leading
expert in atom optics, says that atomic holography could one day be used to
create microelectronic circuits. The optical lithography techniques now used to
etch circuits are limited by the wavelength of light to a resolution of around
150 nanometres.

The goal of fusion researchers is to reproduce the reactions that power the
Sun, where nuclei of light elements fuse to release large quantities of
heat. Most fusion research is carried out in tokamaks—doughnut-shaped
vessels in which gases are heated to form plasmas contained by powerful magnetic
fields.

Until a few years ago, researchers avoided experimenting with the fuel
mixture that will eventually be used in commercial reactors: an equal mix of the
hydrogen isotopes deuterium and tritium. Fusing deuterium and tritium releases
neutrons that irradiate the apparatus, and because the researchers were
interested mainly in investigating the hot plasma, they wanted to avoid having
to deal with the additional problem of radiation.

But by 1993, having completed the initial experimental programme, the
Princeton researchers were ready to begin testing real fusion fuel in the TFTR.
They have tried a variety of plasma conditions, achieving fusion powers of up to
10.7 megawatts.

It is not enough, however, for a tokamak merely to release energy. Like a
candle flame, it must also heat up the plasma if it is to go on burning. In the
case of the deuterium-tritium reactions at the TFTR, energetic alpha particles
produced in the fusion reactions must remain within the plasma, heating the
mixture by giving up their energy to electrons.

Since fusion research began in the 1950s, it has been accepted that the alpha
particles would heat the plasma. But this had never been demonstrated, and there
were even suspicions that fusion might destabilise the plasma. Now, in an
extensive analysis of trials run in TFTR since 1993, Gary Taylor and his
colleagues have shown that the alpha particles generate heat without causing
instability (Physical Review Letters vol 76, p 2722).

They found that an equal mix of deuterium and tritium raises the temperature
of the electrons in the plasma by 5 to 7 million °C. This is only a marginal
increase compared with the base temperature of 90 million °C, and certainly
nowhere near high enough for “ignition”, when the fusion reactions become
self-sustaining.

In the past year, the Princeton researchers have been trying a
new configuration of magnetic fields that could double the power output of the
TFTR. Some theorists believe that the new configuration could generate
conditions in the plasma’s core that might even approach ignition.

But these experiments come at a time when cutbacks in US government support
for tokamak fusion are restricting the TFTR to working for only five or six
months in the year. “We are only just beginning to do real fusion research,”
says Taylor. “Before, they called it fusion research, but it was actually only
plasma physics.”

That has always been the theory, but now researchers have come up with a way to make the quantum equivalent of a crackly connection crystal clear.

The idea comes from Charles Bennett of IBM’s research division in New York and six other physicists in the US, Canada and Israel. Three years ago, Bennett caused a stir when he suggested a way to move quantum data from place to place without breaking the laws of physics. Until then, the idea had always run into trouble because of Heisenberg’s uncertainty principle, which says that the exact properties of subatomic particles cannot all be measured and therefore that copying quantum information from place to place is impossible.

Bennett and his colleagues suggested that quantum data could travel between pairs of “entangled” atoms or photons-pairs whose destinies are shared, even though they may be far apart (This Week, 3 April 1993). Measuring the properties of one half of the entangled pair provides information about the other. If Alice wants to send information to Bob, she prepares an entangled pair of atoms or photons in a pure quantum state, and sends one half to him. If she manipulates her atom, it changes the atom at Bob’s end, sending one quantum bit, or “qubit”, of data.

So far so good, but what if Bob’s half of the entangled pair was disturbed when Alice first sent it to him to set up the connection? Neither of them can find out about the quality of the connection by measuring its properties, because if they do they will be “cut off”.

But now even this problem can be overcome, say the researchers in Physical Review Letters (vol 76, p 722), if Alice sends Bob a large number of prepared pairs of atoms or photons in batches, and the two of them then sacrifice some connections by checking them in order to ensure a supply of good ones. This is possible, they say, because the rules of quantum mechanics will allow Alice and Bob to test pairs of connections while still losing only one.

Suppose Alice sends Bob two atoms and keeps two. She and Bob each measure the properties of the pair at each end. If their measurements agree, the connection is good. The act of measuring loses one atom at either end and the connection between them, but it tells Alice and Bob about the remaining link. By working their way through all of the connections, they keep the links that look good and eliminate all the ones they know are bad.

Critics say that this is only marginal progress. The approach will still be unreliable unless Alice’s initial supply of paired particles is remarkably good in itself, they say. But Bennett is optimistic that quantum information can be made reliable by building some redundancy into the system.

Dana Anderson, Carl Weiman, Eric Cornell and their colleagues at the Joint Institute of Laboratory Astrophysics in Boulder, Colorado, have used a laser beam to keep atoms they had fired into the pipes from touching the glass walls, where they would stick. The laser beam was tuned so it has almost the right wavelength to excite the atoms to a higher energy level. In trying unsuccessfully to do so, it attracts them into the beam. And because the laser light is at its most intense along the axis of the pipe, it acts like a rail along which the atoms travel.

The Colorado physicists fired several tens of thousands of rubidium atoms into one end of a tube 3 centimetres long with a hollow core 40 micrometres across. When the laser was turned on and the pipe was bent into a curve, a beam of atoms shot out of the other end. The atoms stopped emerging as soon as the laser was switched off (Physical Review Letters, vol 72, p 3253).

According to the laws of quantum mechanics, atom beams can be treated like waves, just like light beams. This means that the tricks of optics – reflection, refraction, diffraction and interference – can all be repeated with atoms. Now that physicists have a simple way to bend atom beams in any direction they like, it will be easier to perform these tricks.

Anderson is particularly enthusiastic about the possibility of using the pipes to build atom interferometers. When light beams are split in two and subsequently recombined, they interfere with one another, making characteristic patterns of light and dark bands when projected on a screen. Similarly, says Anderson, a beam of atoms could be split in two, then sent through two parallel loops of the hollow fibre before being recombined. Such a device could be used as part of a gyroscope, revealing minute changes in rotation through subtle changes in the interference pattern.

Although optical interferometers are already used for this purpose, says Anderson, “atom interferometers would be far more sensitive”. This is because atom “waves” have very short wavelengths. And unlike light, atoms are sensitive to electric and magnetic fields, so an atom interferometer could be used to detect tiny electromagnetic fields.

All laser light is generated by a cascade of photons passing in phase in a single direction. As the beam travels through the medium, more photons join the cascade and the beam becomes more intense. But in most lasers, for every photon that becomes part of the laser beam, tens of thousands radiate uselessly in other directions.

The trick has been to build on the new art of “photonics”, and use photonic wire, where radiation is suppressed in all directions but one. The photonic wire, developed by Seng-Tiong Ho and his colleagues at Northwestern University, Illinois, is far narrower than the wavelength of the light it carries. Inside, there is only one direction for photons of light to travel: down the axis.

In its present form, the laser is a loop with a diameter of 4 micrometres. The photonic wire is made of silica with a high refractive index, standing proud of a silicon wafer. Embedded in the wire are three “quantum wells” 0.01 micrometres deep made of indium gallium arsenide. These are the source of the photons. The laser beam would remain trapped inside the tiny ring laser were it not for a second photonic wire, wrapped in a U-shape around the edge of the laser ring. A quantum effect known as tunnelling allows a small fraction of the photons to bleed out of the source laser into the second wire.

This way, 70 per cent of available photons are forced into the laser beam. The high efficiency means that less energy gets wasted as heat. As microelectronic manufacturers try to get more components onto their chips, and look to optical as well as electronic processing of data, they need to minimise heat generation.

Obviously a glue must bind to other surfaces. To do so, it should “wet” any surface it comes into contact with. Whether it does so will depend on the material’s surface free energy, which can be thought of as the energy of the bonds that the molecules at the surface could form if they had any other molecules to make them with. Conventional adhesion theory says that a glue will stick most strongly to surfaces with a large surface free energy, as these will have the strongest incentive to form new bonds.

To test this idea, Chaudhury and his colleagues studied the adhesion of transparent sticky tape to surfaces made from an alkane, a fluoroalkane and a silicone. Of these three materials, the fluoroalkane has the lowest surface free energy and the silicone the highest. As expected, the energy freed as the glue bonded to the surfaces was greatest for the silicone and least for the fluoroalkane.

The researchers then peeled back the tape from eacb of the surfaces using weights slung over a pulley. Contrary to all expectations, the tape proved hardest to pull from the fluoroalkane surface and easiest to remove from the silicone.

Videotapes of the peeling front as seen through a microscope revealed why: as well as unbinding from the substrate, adhesive slips over it. The peeling front was a complicated series of ripples, with fingers of still stuck glue trailing behind the main fracture (Science, vol 269, p 1407).

Slippage requires less energy than straightforward unpeeling, so the ease with which a tape unpeels will depend mainly on the ease with which the glue can slide over the surface. Previous studies, the researchers point out, have shown that thin fluid films slide more easily over alkane than fluoroalkane surfaces.

Chaudhury says that as well as being of help to glue designers, the new view of adhesion should be useful to researchers trying to devise ways of countering glues found in the natural world. “One might extend this concept to coating ships’ hulls to prevent barnacles sticking to them,” he says.