Page 190 - Principles and Applications of NanoMEMS Physics
P. 190

178                                                     Chapter 4


             molecules per state because the net polarization of spins is only about one
             part in one million. Herein lies one of the main limitations of NMR-based
             QC  [193],  [200]:  The fact that the pseudo-pure state signal decreases
             exponentially with the  number  of  qubits prepared,  while the noise level
             remains  constant,  precludes the methods for extracting  pseudo-pure states
             from working for more than about 10 nuclear spins.
               Thus, the use of pseudo-pure states enables one to obtain a result despite
             the highly random nature of the initial state. The question then becomes,
             how does one transform an initial random state into a pseudo-pure state with
             deviation  000 ... 000  000 ... 000 ? A technique, among various,  that  is
             employed applies magnetic field gradient to the sample in order to make the
             frequency  of  the precessing spins  position-dependent and, thus,  make  it
             possible to distinguish different parts of the sample.  In  particular,  the
             gradient field induces a position-dependent phase change along the sample.
             This is the basis of NMR imaging [193].
               Another issue that derives from the ensemble nature of the sample, is that
             care must be taken to reduce unintended coupling between qubits [193]. The
             established technique to accomplish this is called “refocusing” [193], [194].
             The  fundamental  idea  is to apply a pulse at the midpoint of the evolution
                                                             °
             period to a given spin,  of such a  phase (typically  180 ) as to undue the
             evolution it has experienced over the time period due to the influence of the
             undesired coupling [193].
               One common issue with QC is the effect of decoherence. In the case of
             NMR-based QC decoherence is  characterized in terms of two parameters,
             namely, the energy relaxation rate, T 1, and the phase randomization rate, T 2
             [194]. T 1 captures the energy lost by precessing spins to various mechanisms
             such as couplings to other spins, and to phonons and paramagnetic ions, and
             chemical reactions such as ions exchanges with the solvent. This source of
             decoherence may, by properly choosing the molecules and liquid samples,
             be extended to tens of seconds. T 2 captures energy losses due to short- and
             long-range spin-spin couplings, the effects of fluctuating magnetic fields due
             to the spatial anisotropy of the chemical shifts, local paramagnetic ions, or
             unstable laboratory fields. These factors, by properly choosing the quality of
             the samples and laboratory equipment allow  a  decoherence time  of  one
             second or more for molecules in solution [194].



             4.3.1.3  The Semiconductor Solid-State Qubit

               Given  the  predominance  of  solid state silicon electronics technology,
             there is a strong motivation to discover and develop paradigms for quantum
             computing that exploit qubits embedded in silicon wafers. An early example
             of this is  the scheme  for  a silicon-based nuclear spin quantum computer
   185   186   187   188   189   190   191   192   193   194   195