Ben Sauerwine Practice for Qualifying Exams Thanks to Yossef and Shiang Yong for their input in this

problem. Problem Source: CMU August 2001 Qualifying Exam Electron in a spherical well An electron is in a spherical well of radius a and depth V0 , i.e., the non-relativistic Hamiltonian is

H=

v p2 + V (1) 2m

with m = the mass of an electron and

V = V0

ra

V =0 r>a

(2)

h = 1 . The solutions of the Schrodinger equation are of 2 the form R(r )Ylm ( , ) s , where R(r ) is the radial wave equation, Ylm ( , ) is a spherical harmonic, and s is a non-relativistic spinor. R(r ) is a solution to the equation
In this problem, we take
l (l + 1) 1 d 2 dR 2 r dr + 2m(E V ) r 2 R = 0 (3) r dr

where E is the energy of the state. Note that in spherical coordinates, the 2 operator has the form 1 d 2 dR 1 2 = 2 r + ( , ) (4) r dr dr r 2 where is an operator in the angular variables. The differential equation satisfied by spherical Bessel or Hankel functions is

1 d 2 dFl ( x ) l (l + 1) x + 1 Fl ( x ) = 0 (5) dx x 2 dx x2

(a) Explain what properties of the spherical harmonics are used to obtain equation (3) starting with the Hamiltonian given in equation (1). Note the form of the 2 operator given above.
h2 2 +V 2m H = E H =
2

[
[

+ 2 m ( E V ) = 0

Next, separation of variables is assumed so that:

2

+ 2m(E V ) R (r )Ylm ( , ) = 0

Now when one expands the 2 operator, one gets

1 d 2 d 1 r 2 dr r dr + r 2 ( , ) + 2m(E V ) R(r )Ylm ( , ) = 0

This is particularly convenient, since the spherical harmonics are orthogonal eigenfunctions of the spherical harmonic differential equation, here written ( , ) , with eigenvalue l (l + 1) . At this point, it is clear that by allowing the spherical harmonic differential equation to commute with the radial portion and act on the spherical harmonic, then dividing out the spherical harmonic portion will give (3) exactly. (b) For our spherical well problem, what conditions must R(r ) satisfy at: (i)
r=0

Since this wave function must be continuous in space, I see that it must without a doubt have a derivative of zero at the origin or otherwise the spherical harmonics approaching from either side could force a discontinuity as one approaches from either side: R ' (0 ) = 0 if l = 0 no constra int otherwise Second, I see that a discontinuity could be imposed by the spherical harmonics if the spherical harmonic is not spherically symmetric: R (0 ) = 0 if l 0 R (0 ) finite otherwise (ii)

r=a

At this boundary, the wave function must be continuous:

Rr > a (a ) = Rr < a (a ) and

(iii)

R' r >a (a ) = R' r < a (a )

The wave function must be normalizable

R( ) = 0

(c) The solutions for R(r ) for bound electrons are spherical Bessel functions jl (r ) inside the well and spherical Hankel functions hl ( r ) outside the well. From the form of the differential equation for the spherical Bessel or Hankel functions given above, what are and in terms of the quantities given? The energies depend on the principal quantum number n as well as l. How is this reflected in the radial equations? Obeying the Bessel functions inside the well, I have
1 d 2 dR l (l + 1) + 1 R=0 x x2 x 2 dx dx has solutions hl ( x ), jl ( x ) Define x = Cr 1 d d dr d = = dx dx dr C dr Substitute, 1 1 d 2 2 1 dR l (l + 1) + 1 2 2 R = 0 C r C dr C r 2 C dr C r
2 2

1 d 2 dR l (l + 1) + 1 2 2 R = 0 r C r C r 2 dr dr 1 d 2 dR 2 l (l + 1) R=0 + C r r2 r 2 dr dr

Identify = 2mE = 2m(E + V0 )

The principal quantum number as well as the orbital angular momentum will be tied up in the energy E n ,l of which both and are functions, as shown above. (d) Write down the conditions from which you can find the energy eigenstates for bound states of the spherical well. Valid solutions will be fixed by the boundary condition at the border of the sphere:

Ahl
l

( 2mE a ) = Bj ( 2m(E + V )a ) Ah ' ( 2mE a ) = Bj ' ( 2m(E + V )a )

n ,l l n ,l 0 n ,l l n ,l 0

v 2 (r ) = 1
all space

Many energies E may be possible for a given l, corresponding to the radial portion.
(e) Two electrons are placed in the well. Neglecting interactions and assuming that the electrons are distinguishable, write down the wave functions of the two-particle system. Your solutions should correspond to total orbital and total spin quantum numbers.

These will be given by Clebsch-Gordan coefficients. I have:

S {0,1}

v = valid v v

wavefunction count 1 Rn (r1 )Rm (r2 ) v n=0 m=0

L , S (x1 , x 2 ) =

i+L i L i k ; j M L k L M L i ,k (1 , 1 ) j , M L k ( 2 , 2 ) i =0 j = max (i L , 0 ) k = i M L = L S 1 1 v v f ; M S f S M S f ( x1 ) M S f ( x 2 ) 2 1 1 M S = S 2 f = 2 , 2 j M L k m M S f 1 i k n f

Note that I also have removed the case where the particles had the same wave function.
(f) There is a spin-spin interaction between the two electrons, which has the v r approximate form g S S1 S 2 . Show how this interaction arises from the

classical interaction between two magnetic dipoles and express g S in terms of the properties of the electron. Do not attempt a detailed derivation, but give a qualitative explanation of the origin of the spin-spin force.

Suppose two electrons were overlap one another but interact only via their magnetic moments (think of their spins as current loops). There would be a torque on one from the field on the other, and vice-versa.

 B : Bohr Magneton =

eh 2m
spin

v v I then have: E = B Bother v e v = Se me

and similarly,

For a magnetic dipole, v 0 v v v v B = [3 ( r )r ] 2 0 (r ) 3 3 4 r Then, v v v 2 v v E = B 0 3 [3( r )r ] 0 (r ) 3 4r 2 v v v v v v v = 0 3 [3( B1 r )( B 2 r ) B1 B 2 ] 0 ( B1 B 2 ) (r ) 3 4r v v v 2 e 2 v v v e2 v = 0 3 3 2 (S1 r )(S 2 r ) S1 S 2 0 2 S1 S 2 (r ) 3 me 4r me v r Taking only the portions depending only on S1 S 2 ,

0 e 2 2 0 e 2 v v v gS = (r ) S1 S 2 3 2 3 me2 4r me
(g) Treating the interaction of part (f) as a small perturbation, estimate the change in the energies of the lowest-energy states you listed in part (e).

all

space

g S *dV will give an estimate. Recall that

(S

v 2 v v v v + S 2 ) = S12 + S 22 + 2S1 S 2 v v v v v v (S1 + S 2 )2 S12 S 22 S1 S 2 = 2

1

so that only the total spin and individual spins take part. There is then shifting but no splitting in the states above. v v v v v v (S1 + S 2 )2 S12 S 22 = 3 h 2 (if S = 0) 1 h 2 (if S = 1) S1 S 2 = 2 4 4
and the integration is something like

 v v e 2 2 e 2 2 E = S1 S 2 4 Rn (r ) 0 3 2 0 2 (r ) r 2 dr 3 me 4r me 0 which is on the order of v v e 2 2 1 2 E = S1 S 2 given above 4 0 2 Rn (r ) (r ) r 2 dr 1 3 3 me 0 4r (h) The electrons are identical fermions. How are the wave functions of part (e) modified? Prove that if two electrons are placed in the same l-orbit, solutions with total spin S = 1 (triplet states) vanish. Give a qualitative explanation in terms of the Pauli Exclusion Principle.

The states in part (e) are anti-symmetrized if the electrons are considered identical. This means that similar states will be combined so that the exchange of particle numbers gives an overall -1 factor. In the case of identical l-orbit, then, I see from part (e) that all portions of the wave functions match under exchange except the spin part:
S 1 2 f = 1 , 1 M S = S 2 2

1 ; 2

MS f S

MS

v v f (x1 ) M S f (x 2 )

In an antisymmetric combination, when I switch the two particles to combine I need an 1 1 1 1 overall negative factor. In the possibilities ; {0,1} {0,1} , the 2 2 2 2 only qualifying coefficient pair that is nonzero and brings a negative factor in swapping is 1 1 1 1 1 1 1 1 is 0 0 = ; ; 0 0 . 2 2 2 2 2 2 2 2 In terms of the Pauli exclusion principle, one can think of a electrons in the same radial and angular momentum orbital. This orbital must be occupied by one spin-up and one spin-down electron.
(i) The system is put in a heat bath at temperature T. If an electron is put in the well and comes to thermal equilibrium, what is the ratio of probabilities that the electron is in the first excited versus the ground state of the system? Assume that the quantum energy levels are not changed by the heat bath.

Using the Canonical Ensemble in a two-state system, I can assume that this ratio is roughly

e e

1 1 kT 1 0 kT

I will assume that for the one-electron system, the lowest energy states are both for the lowest radial state and for different total angular momentum states, 0 and 1. The form of the momentum operator then gives
1 r2 1 r2 d 2 d 1 r dr + r 2 ( , ) + 2m(E V ) R1 (r )Y00 ( , ) = 0 dr d 2 d r dr + 2m(E0 V ) R1 (r ) = 0 dr

and
1 d 2 d 1 r 2 dr r dr + r 2 ( , ) + 2m(E V ) R1 (r )Y10 ( , ) = 0 1d 2 d 2 + 2m(E1 V ) R1 (r ) = 0 2 r dr r dr 1d 2 d 1 d 2 d 2 r dr 2 + 2m(E1 V ) R1 (r ) = r 2 dr r dr + 2m(E0 V ) R1 (r ) r dr 2 r 2 + 2m(E1 V ) R1 (r ) = [2m(E0 V )]R1 (r ) This integration to collapse the radius, however, is nontrivialat least I get an order of magnitude from the additional term: a a 2 2 + 2m(E1 V ) R1 (r )dr = [2m(E0 V )]R1 (r )dr r 0 0 2 ln (a ) + 2m(E1 V ) = 2m(E0 V )
h2 m where time-averaging over states assumed stationary will remove the seconds from the hbar units. E1 = E0 + ln a

e e

1 1 kT 1 0 kT

=e

h2 ln a kTme

(j) You were not given the value of T in part (i). Describe an experiment that would enable you to measure the temperature of the system. Briefly discuss

the possible errors and the design of your experiment to achieve the measurement. Assume that a is atomic size and V0 has the scale of electronvolts.

First, notice that the transition from the zero angular-momentum to one angularmomentum state here could plausibly come from excitation or emission of a circularlypolarized photon. Assume that the heat-transfer efficiency is perfect between the object and the heat bath. Using this probability, I see that no transition would occur from a photon incident on the well if the particle were already in an excited state and that absorption and re-emission would occur if the particle were in the ground state. In this case, then, I fire a fine beam of photons of the desired energy into the object, and then observe the proportion that is allowed to pass through. The temperature can then be deduced from part (i) since the proportion of photons that passes through is equal to the proportion of measurements that found the electron in the first excited-state.