Figure 1.
Their design, which is shown schematically in figure 1, consists of a
string of ions stored in a linear radio-frequency trap and cooled
sufficiently that their motion, which is coupled together due to the
Coulomb force between them, is quantum mechanical in nature. Each qubit
would be formed by two internal levels of each ion, a laser being used
to perform manipulations of the quantum mechanical probability
amplitudes of the states; conditional two-qubit logic gates being
realized with aid of the excitation or de-excitation of quanta of the
ions' collective motion. At Los Alamos we are building a prototype
quantum computer based on this scheme. For various reasons, singly
ionized Calcium seems to be the best choice of ion. A simplified energy
level diagram is shown in figure 2, with transition wavelengths and
decay times. The engineering problems associated with making such a
device work are formidable. Firstly an ion trap had to be designed and
built; in January of 1997, Calcium ions were confined in the ion trap
that has been built by the quantum information group
here at Los Alamos. Next the ions have to be cooled down to their
ground state; we are currently performing numerical simulations of the
laser cooling procedure to establish the optimum laser powers and
wavelengths to do this job. Once the ions are cooled, the computation
will be performed by a series of laser pulses directed at one or other
of the ions; each pulse must transfer population from one level of the
ion to a different level, without exciting any third level, or
affecting any of the adjacent ions. Sometimes the laser will be used to
excite a quantum of the ions' oscillations, requiring the that the
laser be in a standing wave. Chosing the best combination of lasers and
energy level of the ion, and then inventing ways in which these
operations can be performed reliably with the available technology is a
complicated problem of atomic and optical physics. Of course, it will
be impossible to perform all of the operations required to execute a
quantum algorithm completely reliably, and so there is also a lot work
being done here on the theory of fault
tolerant quantum computation.
Figure 2.