Page 189 - Principles and Applications of NanoMEMS Physics
P. 189

4. NANOMEMS APPLICATIONS: CIRCUITS AND SYSTEMS                177


             quantum gate to an initial state  000 ... 000  [193]. The result is the same,
             namely,  p =  (1− ( aa ′ )) 2 , except that now the deviation in question takes
                      1
             the form =δ  ε  000 ... 000  000 ... 000 .
               In general, if a state has a deviation proportional to a pure state  ψ  ψ ,
             in particular,  =δ  ε  ψ  ψ , it is called a pseudo-pure state. Physically, Cory

             et al. [200] stated that the justification for constructing a pseudo-pure state
             derives from the fact that the spins in the different molecules of a liquid are
             virtually independent of  one  another and that,  as a result, they may be
             construed as a large number of  copies  of a single type of molecule,  thus
             permitting the liquid to be approximated by a Gibbs ensemble. Because of
                                                             N
             this, instead of dealing with a density matrix of size  2 , which is the total
                                                                             n
             number of molecules, one can deal with a reduced density matrix of size  2 ,
             where n is the number of spins in a single molecule. Analytically, instead of
             the density matrix [200],
               Ψ  =  ³  p () ψψ  ψ d ψ ,                                                                         (3 )
                                                                            5
                    {}
                    ψ
                     ψ
                   p
             where  () is  the probability  density of the pure  state described by the
                            ψ
             spinor  ψ  and   {} denotes the set of all unit norm spinors, one uses the
             approximation [200],
                    (1 −  α  ) +I  2α  ψ  ψ
               Ψ  =                    ( 1 ≤−  α  ≤  ) 1 ,                                            (3 )
                                                                            6
                      (1 −  α )2 +  2α
                              n
             where  ψ  is a unit spinor. Thus, since the ensemble  average  of  an
             observable  O is obtained by taking the trace of its product by the density
             matrix,  ( ΨOtr  ),  a simplification is obtained from using the pseudo-state,
             since    the    ensemble    average    is    now     given     by,
                                                        O
                                                      tr
             tr ( ΨO  ) ( −∝ 1  α ) ( )+ 2Otr  α  ψ O  ψ , where  ()  is known. While  the
             pseudo-pure state continues to be made up of a statistical mixture of
             molecules,  since  by  Eq.  (3 ), each spinor determines a unique psudo-pure
                                    6
             density matrix, and each pseudo-pure density matrix determines a spinor that
             is unique to within an overall phase factor (assuming the polarization is  α
             known), each addition of the magnetizations of all the molecules reveals the
             predominance of  one  particular  state present, in effect  capturing each
             molecule’s state for the final spectrum without the necessity of wavefunction
             collapse [200].  The price paid as a result of using pseudo-pure states is the
             loss  of  a factor  of the order of one million in the  effective  number  of
   184   185   186   187   188   189   190   191   192   193   194