5.

61 Fall 2007

Separable Systems

page

QUANTUM IN
1D Systems
x = d p i dx = i , p x

: SEPARABLE SYSTEMS
= (x y z +jy +k z ) = i x r = p x p y p z = i p +j +k i x i y i y

3D Systems

, p x x = i

y , p y = i

z , p z = i

2 2 d 2 = p T = 2m 2m dx 2 (x)
= O ( x ) dx ( x ) O
*

2 2 2 2 2 2 2 = p T = + + 2m 2m x 2 2m y2 2m y 2 ( x, y, z )
= O ( x, y, z ) dx dy dz ( x, y, z ) O
*

By fiat, operators corresponding to different axes commute with one another.

= yx xy

zy = yp z p

z p x = p x p z p

etc.

Further, operators in one variable have no effect on functions of another: ( y) = f ( y) x z f ( x) = f ( x) p z x = p x f * ( z ) etc. xf p f * (z) p The Time Independent Schrdinger Equation becomes:

2 2
2 2 + + + V x , y , z ( ) ( x, y , z ) = E ( x, y , z )
2 2 2 2 m x y z 2 the Laplacian

2 2 , y , z ) ( x, y , z ) = E ( x, y , z ) 2m + V ( x
2 2 , y , z ) Hamiltonian operator in 3D +V ( x H = 2m

( x, y , z ) = E ( x, y , z ) H
Separation of variables

3D Schrdinger equation (Time Independent)

5.61 Fall 2007

Separable Systems

page

IF

, y , z ) + Vy ( y ) + Vz ( z ) = Vx ( x ) V (x

2 2 2 2 2 2 + + + + + H ( x, y, z ) = V x V y V z ( ) ( ) ( ) x y z 2m y 2 2m z 2 2 then 2m x = H + H + H x y z
Schrdingers Eq. becomes:

H x + H y + H z ( x, y , z ) = E ( x, y, z )
Then try solution of form

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

(separation of variables) Where we assume that the 1D functions satisfy the appropriate 1D TISE:

( x) = E ( x) H x x x x ( y ) = E ( y )
H
y y y y

( z ) = E ( z )
H z z z z
First term:

( x ) ( y ) ( z ) = ( y ) ( z ) H ( x ) = ( y ) ( z ) E ( x ) H x x y z y z x x y z x x

= E x x ( x ) y ( y ) z ( z )
Same for

H y

and

H z

H = E ( x ) ( y ) ( z ) = ( E + E + E ) ( x ) ( y ) ( z ) H y z x y z x y z x x + H y + Hz
E = Ex + E y + Ez
Thus, if the Hamiltonian has this special form, the eigenfunctions of the 3D Hamiltonian are just products of the eigenfunctions of the 1D Hamiltonian and the situation is equivalent to doing three separate 1D problems.

is separable into Conclusion: Wavefunctions multiply and the energies add if H

5.61 Fall 2007

Separable Systems

page

=H +H +H . H x y z
The 3D product states are naturally normalized if the 1D wavefunctions are normalized:

( x ) ( y ) ( z ) ( x ) ( y ) ( z ) dx dy dz =
* * * x y z x y z

( x ) ( x ) dx ( y ) ( y ) dy ( z ) ( z ) dz = 1
* * * x x y y z z

Notice that for we are free to choose any eigenfunction, nx , of H x and any eigenfunction, n , of H . together with any eigenfunction, ny ,of H y z z
Thus, while in 1D we usually had one quantum number, in 3D we will have three (nx,ny,nz). Further, for two product states to be orthogonal, we do not have to have all three 1D functions be different. If any one of the three 1D wavefunctions (x,y or z) is orthogonal to its counterpart, then the two 3D wavefunctions are also orthogonal. For example, consider the two wavefunctions 111 ( x, y, z ) = x1 ( x ) y1 ( y ) z1 ( z ) and 311 ( x, y, z ) = x 3 ( x ) y1 ( y ) z1 ( z ) .

111 ( x, y, z ) 311 ( x, y, z ) dx dy dz 1* ( x ) 1* ( y ) 1* ( z ) 3 ( x ) 1 ( y ) 1 ( z ) dx dy dz 1* ( x ) 3 ( x ) dx 1* ( y ) 1 ( y ) dy 1* ( z ) 1 ( z ) dz = 0
*
x y z x y z x x y y z z

0 x 1 x 1 Because the product states are orthogonal and normalized, theyre orthonormal and we summarize this by writing: nxnynz * x, y, z mxmymz x, y, z dx dy dz = nx ,mx ny ,my nz ,mz

Example 1:3D Harmonic Oscillator Lets consider a particle in 3D subject to a Harmonic potential in x,y and z. Further, assume the force constants in each direction are different. This might be true, for example, for a particle trapped inside a protein: the resorting force for moving it in the x direction will be different from y or z because the protein has a different shape along x than y or z (see below right). Multidimensional

5.61 Fall 2007

Separable Systems

page 4

Harmonic potentials are also important for describing the vibrations of polyatomic molecules. For example, in HCO x might correspond to the CH stretch, y to the CO stretch and z to the HCO bend. In either case, the general potential is given by

V ( x, y , z ) =

1 1 1 k x x 2 + k y y 2 + k z z 2 = V x ( x ) + V y ( y ) + Vz ( z ) 2 2 2

Now, because the potential is a sum of an x potential, a y potential and a z potential, we can easily write the Hamiltonian down as a sum:

=H +H +H H x y z

x2 1 p 2 + kx x Hx = 2m 2
y2 1 p 2
+ ky y Hy = 2m 2 z2 1 2 = p + kz z H z 2m 2
where each 1D Hamiltonian describes a particle subject to a Harmonic potential with the appropriate spring constant (kx, ky or kz). Based on the discussion above, we can immediately write down all the eigenfunctions and eigenvalues:

n n n ( x , y , z ) = N x H n ( x x ) e
1/2
x y z x

x x2
2

N y H n y ( y y ) e
1/2

y y2
2

z 1/2 N H z e ( ) z nz z

1 4

z z2
2

1 1 1 Enxn y nz = n x + x + n y + y + n x + z 2 2 2

( mk x ) x =

12

x =

kx etc. m

Notice again that for the 3D problem we have 3 quantum numbers, and the energy and wavefunction depend on all three simultaneously. Degeneracies In 3D, there are a number of interesting things that can happen that we didnt see in 1D. One example of this is that, in 3D it is possible for two different

5.61 Fall 2007

Separable Systems

page

eigenfunctions of the Hamiltonian to have the same energy. When this happens, these two states are called degenerate. If there are three, four states with the same energy, they are said to be threefold, fourfold degenerate. This never happened for the Particle in a Box or the Harmonic Oscillator. In fact, one can show that for bound states in 1D, one never has degeneracy; every state has its own energy. However, in 3D we can find degeneracy very easily. For example, if our spring constants are all the same:

kx = k y = kz k

x = y = z

3 Enxn y nz = n x + n y + n z + 2
3 The ground state has an energy E000 = 2 . There is only one way I can get this

energy ( n x = 0, n y = 0, nz = 0 ), so it is not degenerate. However, there are three ways I can get the first excited state energy spring constants to be equal,
5 2

: n x = 1, ny = 0, nz = 0 ,

n x = 0, ny = 1, nz = 0 or n x = 0, ny = 0, nz = 1 . So we find that if we choose all the

E000 =

3 2 5 2

nondegenerate level
3-fold degenerate level

E100 = E010 = E001 =

etc. We will typically draw this with a picture like:

E E110 E011 E100 E101

E200

E020

E002

6fold degenerate

E010 E000

E001

3fold degenerate

Nondegenerate

Note the wavefunctions are distinct:

100 ( x, y, z ) 010 ( x, y , z ) 001 ( x, y , z )

This leads to an interesting effect. Suppose we make up a wavefunction that is a sum of two degenerate states, say

( x, y , z ) = a 010 ( x, y, z ) + b 001 ( x, y , z )

5.61 Fall 2007

Separable Systems

page

where a and b are constants. Then it turns out that is also an eigenstate of the Hamiltonian! To see this,

= H (a + b ) = a H + b H H 010 001 010 001

=a

5 2

010 + b

5 2

001 = 5 2 ( a 010 + b 001 )

= 5 2

This illustrates the important point that any sum of degenerate eigenstates is also an eigenstate of the Hamiltonian with the same eigenvalue. Setting the spring constants equal amounts to assuming the well is symmetric with respect to x, y and z. If the symmetry is broken, i.e. k x k y k z then the 100, 010 and 001 states will become nondegenerate. In this case, we will usually say that the degeneracy has been lifted or that the degenerate states have been split. The latter language comes from the pictorial view; if the spring constants are only slightly different, then the energy levels might look like:

E E110 E011 E100 E101 E200 E020 E002

E010 E000

E001

Here, you can see that the n=1 levels are almost degenerate (theyve been split) and the n=2 levels are almost degenerate, but not quite, because the force constants are slightly different. Notice that it is possible to break some degeneracies but keep others. For example , if we choose:

kx = k y k kz
Then the energies become:

x = y z

1 Enxn y nz = ( n x + n y + 1) + z nz + 2
and we find

5.61 Fall 2007

Separable Systems

page

1
E000 = + z 2
1
E100 = E010 = 2 + z 2
3
E001 = + z 2

nondegenerate level 2-fold degenerate level nondegenerate level

5.61 Fall 2007

Separable Systems

page

Example 2: Particle in 3D box

z c

V ( x, y , z ) = Vx ( x ) + V y ( y ) + Vz ( z )

Vx ( x ) = 0 Vy ( y ) = 0 Vz ( z ) = 0
a x
Inside the box:

0 xa 0 yb 0 zc
otherwise

Vx ( x ) ,V y ( y ) ,Vz ( z ) =

Outside the box:

2 2 2 2 + + ( x, y , z ) = E ( x, y , z ) 2 2 2 2m x y z

(x , y, z )= 0

=H +H +H where We can again apply separation of variables since H x y z

,H ,H are each 1D particle in a box Hamiltonians. So the solutions to the H x y z
3D equation are products of the 1D solutions

( x , y , z ) = n ( x ) n
x

( y ) n ( z )
z

where from the 1D problem we have the solutions

n x 2 nx ( x ) = sin x a a n y 2 2 n y ( y ) = sin y b b 2 n z nz ( z ) = sin z c c

1 2 1

1 2

n x = 1, 2,3,... n y = 1, 2,3,...

2 2 nx En x = 8m a 2

2 n y En y = 8m b2 2 nz2 En z = 8m c 2

nz = 1, 2,3,...

Where the energy is now a function of all three Quantum numbers:

5.61 Fall 2007

Separable Systems

page

En x n y n z
Degeneracies

2 n2 2 nx nz2 y = E n x + E n y + E nz = 2+ 2+ 2 8m b c a

Degeneracies occur in a similar fashion for the PiB. e.g. if a = b = c in our 3D box:

En x n y n z =

h2 2 n2 + n2 y + nz ) 2 ( x 8ma

3h 2 E111 = 8ma 2

is nondegenerate

8 2 x y z 111 ( x, y , z ) = 1 ( x ) 1 ( y ) 1 ( z ) = 3 sin sin sin a a a a

But

E211 = E121 = E112

h2 = 22 + 12 + 12 ) 2 ( 8ma

3fold degeneracy

Note the wavefunctions are again distinct:

211 (x, y , z ) 121 (x , y , z ) 112 (x , y , z )

E E222 E221 E211 E212 E121 E111 E122 E112
Nondegenerate Nondegenerate 3fold degenerate

5.61 Fall 2007

Separable Systems

page 10

If the symmetry is broken, i.e.

a bc

then the degeneracy is lifted.

E112

h2 1 1 4 h2 1 4 1 = 2 + 2 + 2 E121 = 2+ 2+ 2 8m a 8m a b c b c

Summary
3D box

8 2 n x x n y y n z z nx n y nz ( x , y , z ) = sin sin sin abc a b c En x n y n z

2 n2 h 2 nx nz2 y = 2+ 2+ 2 8m b c a

n x = 1, 2,3,... n y = 1, 2,3,... nz = 1, 2,3,...