tensor product states.pdf

Upload: shan-de-silva

Post on 06-Jul-2018

217 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/18/2019 Tensor Product States.pdf

    1/18

    Lecture 24:Tensor Product States

    Phy851 Fall 2009

  • 8/18/2019 Tensor Product States.pdf

    2/18

    Basis sets for a particle in 3D

    • Clearly the Hilbert space of a particle in threedimensions is not the same as the Hilbertspace for a particle in one-dimension

    • In one dimension, X and P are incompatible

    – If you specify the wave function in coordinate-

    space, !x|" #, its momentum-space state iscompletely specified as well:! p|" # = $ dx ! p| x #! x |" #

    – You thus specify a state by assigning anamplitude to every possible position OR  byassigning and amplitude to every possiblemomentum

    • In three dimensions, X, Y, and Z, arecompatible.

    – Thus, to specify a state, you must assign anamplitude to each possible position in threedimensions.

    – This requires three quantum numbers

    – So apparently, one basis set is:

    ! !    x x   =)(

    ! !    z  y x z  y x   ,,),,(   =

    or

     x, y, z{ }"   x, y, z   # ( x, y, z)$  R

    3

     

    " ( p) =   p " 

  • 8/18/2019 Tensor Product States.pdf

    3/18

    Definition of Tensor product

    • Suppose you have a system with 10 possible

    states

    • Now you want to enlarge your system byadding ten more states to its Hilbert space.

    – The dimensionality of the Hilbert space increasesfrom 10 to 20

    – The system can now be found in one of 20possible states

    – This is a sum of two Hilbert sub-spaces

    – One quantum number is required to specifywhich state

    • Instead, suppose you want to combine your

    system with a second system, which has tenstates of its own

    – The first system can be in 1 of its 10 states

    – The second system can be in 1 of  its 10 states

    • The state of the second system is independent of the state of the first system

    – So two independent quantum numbers are

    required to specify the combined state

    • The dimensionality of the combined Hilbertspace thus goes from 10 to 10x10=100

    – This combined Hilbert space is a

    (Tensor  ) Product of the two Hilbert sub-spaces

  • 8/18/2019 Tensor Product States.pdf

    4/18

    Formalism

    • Let H 1 and H 2 be two Hilbert spaces

    – We will temporarily ‘tag’ states with a label tospecify which space the state belongs to

    • Let the Hilbert space H 12  be the tensor-

    product of spaces H 1 and H 2 .

    • The Tensor product state |" 12#=|" 1#(1) %|" 2#

    (2)

    belongs toH 

    12.

    • The KEY POINT TO ‘GET’ IS:

    – Bras and kets in the same Hilbert space‘attach’.

    – BUT, Bras and kets in different Hilbertspaces do not  ‘attach’

    21

    )1(

    2

    )1(

    1   ! ! ! !    "#$

    1

    )1(H  !"  2

    )2(H  !" 

    2112  H  H  H     != H 12 is the ‘tensor product’

    of H 1 and H 2

    )1(

    1

    )2(

    1

    )2(

    1

    )1(

    1   ! " " !    #$%They just slide past

    each other

  • 8/18/2019 Tensor Product States.pdf

    5/18

    Schmidt Basis

    • The easiest way to find a good basis for a

    tensor product space is to use tensor productsof basis states from each sub-space

    If:  {|n1#(1) } ; n1=1,2,…, N 1 is a basis in H 1

    {|n2#(2) } ; n2=1,2,…, N 2 is a basis in H 2

    It follows that:

    {|n1, n2#(12) }; |n1, n2#

    (12)=|n1#(1)%|n2#

    (2) is a basis in H 12.

    If System 1 is in state:

    and System 2 is in state:

    Then the combined system is in state:

    • Schmidt Decomposition Theorem:All states in a tensor-product space can beexpressed as a linear combination of tensorproduct states

    !=

    =

    1

    1

    1

    1

    )1(

    1

    )1(

    1

     N 

    n

    n  na" 

    !=

    =

    2

    2

    2

    1

    )2(

    2

    )2(

    2

     N 

    n

    n   nb" 

     

    " 1," 2(12)

    =  " 1(1)#  " 2

    (2)=   a

    n1bn2

    n1,n2(12)

    n2=1

     N 2

    $n1=1

     N 1

    $

    !!= =

    =

    1

    1

    2

    2

    21

    1 1

    )12(

    2,1

    )12(

    12

     N 

    n

     N 

    n

    nn  nnc" 

  • 8/18/2019 Tensor Product States.pdf

    6/18

    Entangled States

    • The essence of quantum `weirdness’ lies in

    the fact that there exist states in the tensor-product space of physically distinct systemsthat are not tensor product states

    • A tensor-product state is of the form

    – Tensor-product states are called ‘factorizable’ 

    • The most general state is

    – This may or may-not be ‘factorizable’ 

    • Non-factorizable states are called ‘entangled’ 

    – For an `entangled state’, each sub-system has no independent objectivereality 

    !!= =

    =

    1

    1

    2

    2

    21

    1 1

    )12(

    2,1

    )12(

    12

     N 

    n

     N 

    n

    nn   nnc" 

    )2(

    2

    )1(

    1

    )12(

    ! ! !    "=

  • 8/18/2019 Tensor Product States.pdf

    7/18

    Configuration Space

    • The state of a quantum system of Nparticles in 3 dimensions lives in

     ‘configuration space’ 

    – There are three quantum numbersassociated with each particle

    • It takes 3N quantum numbers to specify astate of the full system

    – Coordinate Basis:

    – Wavefunction:

    – To specify a state of N  particles in d dimensions requires d·N quantum

    numbers

     

     x1, y

    1, z

    1, x

    2, y

    2, z

    2,..., x

     N  , y

     N , z

     N { }

    (We don’t know about spin yet)

    or we could just write

     

    r

    r1,r

    r2,...,

    r

    r N { }

    ( )   ! !  N  N 

      r r r r r r   rrrrrr

    ,...,,,...,,2121

      =

    So counting quantum numbers might be a goodway to check if you are using a valid basis

    This form is thecoordinate system

    independent

    representation

  • 8/18/2019 Tensor Product States.pdf

    8/18

    Tensor Products of Operators

    • THEOREM:

    Let  A (1) act in H 1 , and  B

    (2) act in H 2 ,

    Then the tensor product operator C (12) = A(1)%  B(2)

    acts in H 12 .

    • PROOF:

    • The action of C (12) on a tensor-product state:

    )1()1()1(

    mm

    m

    m  aaa A !=

    )2()2()2(mm

    m

    m  bbb B !=

     

     A(1)"  B

    (2)=   a

    mbn

    m,n

    #   am(1)

    am

    (1)"   b

    n

    (2)bn

    (2)

     

    =   ambn

    m,n"  a

    m

    (1)#   b

    n

    (2)

    ( )  a

    m

    (1)#   b

    n

    (2)

    ( ))12()12(

    ,

    )12( ,,nmnm

    nm

    nm  bababaC  !=

    )2(

    2

    )1(

    1

    )12(

    ,

    )12(

    21

    )12( ,,   ! ! ! ! nmnm

    nm

    nm  bababaC  "=

    )2(

    2

    )1(

    1

    )12(

    21   :,   ! ! ! !    "=

  • 8/18/2019 Tensor Product States.pdf

    9/18

    General form of Operators inTensor-product spaces

    • The most general form of an operator in H 12is:

    – Here |m,n# may or may not be a tensorproduct state. The important thing is thatit takes two quantum numbers to specifya basis state in H 12

    • A basis that is not formed from tensor-product states is an ‘entangled-state’ basis

    • In the beginning, you should always startwith a tensor-product basis as your ‘physicalbasis’ 

    – Then all operators are well-defined

    – Just expand states and operators onto tensor-product states

    – Then match up the bras and kets with theirproper partners when taking inner products

    )12()12(

    ,

    ',';,

    )12( ',',   nmnmcC nm

    nmnm!=

    )12()12()12(

    ',';,   ,,:   nmC nmc nmnm   !!=

  • 8/18/2019 Tensor Product States.pdf

    10/18

    Upgrading Subspace Operators

    • Any operator in H1

      can be upgraded to an

    operator in H12 by taking the tensor product

    with the identity operator in H2 :

    – If  A1 is an observable in H1, then it is also an

    observable in H12 (since it remains Hermitianwhen upgraded).

    – The spectrum of  A1 remains the same afterupgrading

    Proof:

    Let

    Then:

    )2()1()12( I  A A   !=

     

     A(1)"  I 

    (2)am

    (1)"  # 

    (2)( ) =   A(1) am (1)( ) "   I (2) #  (2)( )=  a

    m  a

    m

    (1)" # 

    (2)

    =

     am   am

    (1)

    "  # 

    (2)

    ( )Note that |" 2# is completely

    arbitrary, but |a 1 #%|" 

    2 #  is an

    eigenstate of A(12)

    )1()1()1(

    mmm  aaa A   =

  • 8/18/2019 Tensor Product States.pdf

    11/18

    Product of two Upgraded Operators

    • Let  A(1) and  B(2) be observables in their

    respective Hilbert spaces• Let  A(12)= A(1)% I (2) and  B(12)= I (1)% B(2) .

    • The product  A(12) B(12) is given by  A(1) %  B(2)

    – Proof:

     

     A1"  I 

    2( )   I 1 "  B2( ) =  A1 I 1 "  I 2 B2=  A

    1"  B

    2

  • 8/18/2019 Tensor Product States.pdf

    12/18

    Compatible observables

    • Let  A(1) and  B(2) be observables in their

    respective Hilbert spaces• Let  A= A(1)% I (2) and  B= I (1)% B(2) .

    • Theorem: [ A,B]=0

    Proof:

    • Conclusion: any operator in H 1 , is ‘compatible’ 

    with any operator in H 2 ,.

    • I.e. simultaneous eigenstates exist.

    – Let  A1 |a1#= a1 |a1# and  B2 |b2#= b2 |b2#.

    – Let a= a1 and  b= b2

    – Let |ab#= |a1# % |b2#.

    – Then  AB|ab#=ab|ab#.

    [ ]( )( ) ( )( )

    ( ) ( ) ( ) ( )0

    ,

    2121

    22112211

    21212121

    =!"!=

    !"!=

    !!"!!=

    "=

     B A B A

     I  B A I  B I  I  A

     I  A B I  B I  I  A

     BA AB B A

  • 8/18/2019 Tensor Product States.pdf

    13/18

     ‘And’ versus ‘Or’ 

    • The tensor product correlates with a system

    having property A and  property B– Dimension of combined Hilbert space is product

    of dimensions of subspaces associated with Aand B

    – Example: start with a system having 4 energylevels. Let it interact with a 2 level system. TheHilbert space of the combined system has 8

    possible states.

    • Hilbert spaces are added when a system canhave either property A or  property B

    – Dimension of combined Hilbert space is sum of dimensions of subspaces associated with A and

    B– Example: start with a system having 4 energy

    levels. Add 2 more energy levels to your model,and the dimension goes from 4 to 6

  • 8/18/2019 Tensor Product States.pdf

    14/18

    Example #1 : Particle in Three Dimensions

    • Let H 

    1 be the Hilbert space of functionsin one dimension– The projector is:

    – So a basis is:

    • Then H 3 =  (H 

    1)3 would then be the

    Hilbert space of square integrablefunctions in three dimensions.

    – Proof:

    • Note: H 3   is also the Hilbert space of 

    three particles in one-dimensional space

     

     I 1=   dx "    x x

     

     x { }

     

     I 3= I 

    1" I 

    1"  I 

    1

    =   dx #    x x   "   dy #    y y   "   dz #    z z

    =   dxdydz #    x x   "   y y   "   z z

    =   dxdydz #    x   "   y   "   z( )   x  "   y   "   z( )

    =   dxdydz #    x, y, z x, y, z

    ),,(,,   z  y x z  y xdxdydz    ! !   " =

     z  y x z  y x   !!",,

      ! !    z  y x z  y x   ,,),,(   =

  • 8/18/2019 Tensor Product States.pdf

    15/18

    Three-dimensional Operators

    • We can define the vector operators:

    • Note that:  X = X (1)% I (2) % I (3) and  P  y = I 

    (1)% P  (2)

     % I (3)

    so that [ X , P  y]=0.

    • With

    • We can use:

     z  P  y P  x P  P 

     z  Z  yY  x X  R

     z  y x  ˆˆˆ

    ˆˆˆ

    ++=

    ++=

    r

    r

    ( )   z  y x z  z  y y x x z  y x R   ,,ˆˆˆ,,   ++=r

     Z  R

    Y  R

     X  R

    =

    =

    =

    3

    2

    1

     z 

     y

     x

     P  P  P  P 

     P  P 

    =

    =

    =

    3

    2

    1

     

     R j , Rk [ ] =   P j ,Pk [ ] = 0

     R j 

    ,Pk 

    [ ]= ih" 

     jk 

     

    r

     R

    r

    r=

    r

    r

    r

    ror

  • 8/18/2019 Tensor Product States.pdf

    16/18

    Example #2: Two particles in OneDimension

    • For two particles in one-dimensional space,

    the Hilbert space is (H 1 )2.

    )2(

    2

    )1(

    121,   x x x x   !=

    ! !  2121   ,),(   x x x x   =

     

     I =   dx1dx

    2 "    x1, x2   x1, x2 =1

     

     X  j , X k [ ] =   P j ,Pk [ ] = 0

     X  j 

    ,Pk 

    [ ]= ih" 

     jk 

    ...

    212

    211

    etc

     P  I  P 

     I  X  X 

    !=

    !=

  • 8/18/2019 Tensor Product States.pdf

    17/18

    Hamiltonians

    • One particle in three dimensions:

    – Each component of momentum contributesadditively to the Kinetic Energy

    • Two particles in one dimension:

    ( )

    )(2

    1

    ),,(2

    1   222

     RV  P  P m

     Z Y  X V  P  P  P m

     H   z  y x

    vrr

    +!

    =

    +++=

    ),(22

      21

    2

    2

    2

    1

    2

    1 X  X V 

    m

     P 

    m

     P  H    ++=

  • 8/18/2019 Tensor Product States.pdf

    18/18

    Conclusions

    • The ‘take home’ messages are:

    – The combined Hilbert space of twosystems, dimensions d 1 and d 2, hasdimension d 1%2=d 1·d 2

    – A physical basis set for the combinedHilbert space, H 1%2 can be formed by

    taking all possible products of one basisstate from space H 1 with one basis statefrom H 2.

    – In a tensor product space, a bra from onesubspace can only attach to a ket fromthe same subspace:

    – For N  particles (spin 0) in d  dimensions,d·N  quantum numbers are required tospecify a unit-vector in any basis

     

    n1

    (1){ }" H 1,   n2 (2){ }" H 2n

    1,n

    2  :=   n

    1

    (1)#   n

    2

    (2)

    $   n1,n2{ }" H 12 :=H 1  #H 2

    21

    )1(

    2

    )1(

    1   ! ! ! !    "#$

    )1(

    1

    )2(

    1

    )2(

    1

    )1(

    1   ! " " !    #$%

     Nz  Ny Nx z  y x z  y x

     N 

    nnnnnnnnn

    r r r 

    ,,,,,,,,,

    ,,

    222111

    21

    K

    r

    K

    rr