## Scientists Report Interactions Between Individual Photons: First Step Toward a Quantum Computer

This result—interactions between single photons—could be used to make information processing devices that employ quantum-mechanical effects to improve their performance. Further, these devices could form the building blocks needed to construct a "quantum computer," a theoretical machine that, researchers believe, could outperform any computer based on conventional technology.

Scientists and security specialists have given the subject of quantum computing much attention since the 1994 discovery of an algorithm—a mathematical technique—for factoring large numbers on a quantum computer. This algorithm, devised by Peter Shor at AT&T Bell Labs in Murray Hill, New Jersey, means that in theory, a quantum computer could outperform any conceivable classical computer. Such a machine would have wide ranging implications for everything from national security to automated teller machines, because the encryption codes that protect electronic data rely on huge numbers that even the most powerful conventional computer cannot factor. A quantum computer would make such codes obsolete.

The area of quantum information and computation had caught the attention of the quantum optics group at Caltech, headed by Professor of Physics H. Jeff Kimble. Together with Seth Lloyd at MIT, the quantum optics group recognized that their experiments were closely related to something called quantum logic gates. Quantum logic gates are the building blocks needed to construct a quantum computer.

Conventional computers work by sending classical, prescribed "bits" of information—the distinction between two alternate states, such as zero or one, no or yes—as pulses of electrical current through wires, transistors, and other components. The basic building blocks of an ordinary computer are logic gates, which process the bits of information. The processing either passes the bits through unchanged or "flips" them, changing zeroes to ones and vice versa. Theorists have proposed that a quantum computer might work by recreating the components and logic gates of conventional computers in a quantum mechanical way, using the quantum states of atomic particles to carry and manipulate information. The basic components in a quantum computer would be quantum logic gates.

Motivated by the recent excitement over Shor's algorithm, Kimble and his group demonstrated that the strong interaction between photons and an atom in an optical cavity can provide the basis for building optical quantum logic gates. Their optical quantum logic gate operates by processing the polarization states of a pair of photons, with the polarization state of each photon encoding one bit of information.

Any legitimate logic gate must display an essential feature called conditional dynamics, which means that the output of each gate must depend upon both inputs to the gate. Or, in an optical quantum logic gate, the output state of each photon must depend on the input state of both photons.

In their experiment, reported in the December 18, 1995 issue of Physical Review Letters, Kimble's group showed strong conditional dynamics for an atom in an optical cavity formed by two highly reflective mirrors, one of which allowed partial transmission of light. The scientists sent pairs of photons through the cavity, and investigated the states of the photons when they reemerged, showing that the output state of each photon depended on the polarization of both input photons. This is just what is required to implement quantum logic.

In effect, the cavity functioned as a rudimentary logic gate at the single photon level. The photons served as the current needed to carry bits of information; and changing the photons' polarization was analogous to flipping the bits in conventional computers. The Caltech result, by Kimble and graduate students Christina Hood, Hideo Mabuchi, Quentin Turchette, and research fellow Wolfgang Lange, was accompanied by a paper in the same issue of Physical Review Letters by a group at the National Institute of Standards and Technology in Boulder, Colorado. Together these papers represent the first demonstrations of conditional dynamics at the single-quantum level—the level suitable for implementing discrete quantum logic.

While this result is a significant first step, many complex problems remain to be solved before even primitive networks of quantum logic gates could be built, much less an entire computer. Indeed, researchers have not yet determined whether large and complex quantum computers could ever be built using current technological strategies.

In this regard, the work of Kimble's group is especially important because they have focused on an optical implementation of quantum logic. Because single photons can be transmitted long distances through optical fibers, optical quantum logic gates could be used for specialized applications in optical communication, even if they prove not to be useful for quantum computers.

This work was supported by the National Science Foundation and the Office of Naval Research.

Contact: Jay Aller (818) 395-3631 aller@caltech.edu