THE SECURITY of medical records or Internet bank accounts could one day be
maintained by a new breed of hypersensitive microchips, say physicists at
Toshiba and the University of Cambridge.
The scientists have worked out how to modify transistors so that they can
detect a single photon, the Conference on Lasers and Electro-optics (CLEO) in
San Francisco heard last week. Detecting single photons is crucial in the
emerging field of quantum cryptography, in which secret messages—such as
an encryption key which could unlock medical or e-commerce data—are
superimposed on a single photon. Any hacker who intercepted the photon in an
optical fibre would alter its quantum state, and alert the receiver and sender
to the interception. They could then change the key.
But until now, detecting light signals has been performed either with fragile
and bulky vacuum tube-based photomultipliers, or solid state photodiodes, which
can be noisy at the frequencies used in optical communication. So Andrew Shields
and his colleagues at Toshiba Research Europe in Cambridge, and at the nearby
Cavendish Lab, have tried a new approach. Their idea, says Shields, is to apply
the curious physics of the “quantum dot”—a tiny islet of semiconducting
material that is good at trapping one or two mobile electrons.
Advertisement
In an ordinary transistor, the through current is changed by altering the
voltage applied to its “gate” electrode. So to create their detector, the
Cambridge team grew indium arsenide quantum dots underneath a transparent gate
on a gallium arsenide transistor. The dots, replete with electrons, react to a
visible or infrared photon by freeing one of the electrons and allowing it to
enter the transistor, lowering the latter’s resistance. By measuring that
resistance change, the arrival of quanta can be detected. “Each response of the
device is to a single photon,” Shields says.
Markus Kuhn, a security researcher over at the university’s Computer
Laboratory, believes the detector could also be of use in quantum computers, but
he is sceptical about prospects for quantum cryptography. “No proposal I’ve seen
explains how a digital signature can be created, and making sure you only send
one photon is technically very difficult,” he says.