Problem set 3. Due Wednesday March 10. We will be addressing two different topics this time.

We will put electronic structure theory into perspective and define the general problem that has to be solved. We discuss antisymmetry of the wave function in some detail and introduce a basis of Slater determinants, which are the antisymmetrized equivalent of the pure Hartree products of one-electron orbitals. This allows us to formulate the linear variation problem in the space of all determinants that can be formed from a finite set of spin-orbitals. This defines the Full CI solution in quantum chemistry, and it serves as a benchmark for accurate approximation methods. The reading material is Szabo and Ostlund pages 39 to 64. Our second topic concerns the vibrational problem, primarily in one dimension (diatomics), and the use of operator algebra to recast the problem and to find solutions. This provides a convenient introduction to the language of second quantization, that we will later on use also for the electronic structure problem. You will receive a set of hand written lecture notes. The problem set below has some problems on electronic structure theory, but the majority of the problem set focuses on operator algebra for vibrational problems. The last two questions require you to use Maple. This is a symbolic algebra package that has been developed at the University of Waterloo. On Nexus you can find version 8 under Programs Math and Science Maple 8. You can find a brief tutorial on Maple with some examples on the Nooijen web site. The file maple_basics.mws can be opened in the Maple program. First go through a brief tutorial in Maple, then see what is done in the sample file, and off you go to do some calculations in Maple. The problem set is quite extended, not in the least because of the last two problems.

Problem Set 1. Szabo and Ostlund: 2.1, 2.2, 2.4, 2.5, 2.7 2. The N-particle antisymmetrization operator can be given in the following

= , where the sum runs over all N! permutations of N coordinates, and form A () pi P i
i =1

N!

pi is the parity of the permutation that indicates if it consists of an even or odd number of
= N !A . FYI the antisymmetrizer is transpositions or interchanges. Show that AA
N! '= 1 . In this case applying the sometimes normalized differently as A () pi P i N ! i=1

antisymmetrization operator twice yields the same result, which one might expect. It then acts as a projection operator on the antisymmetric subspace.
= TT ....T . It was 3. Every permutation can be written as a product of transpositions P i 1 2 k

shown in class that every transposition is Hermitian, and is its own inverse. Show that the
= P 1 , and that it has the same parity as P . Use these permutation operator is unitary: P i i i

results to show that the antisymmetrization operator is Hermitian. The following problems are on vibrational problems and using the bose operator algebra. , 1 4. Show the following. You can use the commutation relations b b = , the definitions
= 1 ( q = 1 (q = i , as well as the Hermiticty of p and q ip ), b + ip ), where p b q 2 2

in your proofs.
and b are Hermitian conjugates, i.e. b = b a. The operators b
*

n n1 n n1 b. b,(b ) = n(b ) , (b) , b = n(b)

5. The orthonormal eigenfunctions of the harmonic oscillator are defined as n = 1 n (b ) 0 , where 0 is normalized to unity. Using the relations discussed in the n!

previous question show: a. n is indeed normalized to 1. n = n + 1 n + 1 b. b n = n n 1 c. b d. n m = 0, n m . From a. and d. it follows that the functions n are orthonormal.

and b , remembering that these 6. Write the following operators in terms of the b

operators are defined in terms of the dimensionless coordinates q = conjugate momentum operator i a. Position, x b. Momentum i=
x
=2 2 2m x 2

m x and the =

. q

c. Kinetic energy operator d. Potential energy

e. The Hamiltonian

=2 2 1 + m 2 x 2 2 2m x 2

7. Consider the Hamiltonian of the following form H =

1 2 1 2 + q + q 4 . We will use 2 2 q 2

the linear variational principle using a finite set of harmonic oscillator basis states n = 1 n . We want to (b ) 0 , to find approximate energy eigenvalues of H n!

m . To this end we wish to evaluate the general form for the matrix elements n H
and b . You can make use of the express the Hamiltonian in terms of the operators b

relations expressed under question 4, in particular:

(b ) n = (b ) n b + n(b ) n1 , (b ) n b = b (b ) n + n(b ) n1 . b
operators in a product of We want to write the operator q 4 in a form that all of the b . So the operators b act first. This is called normal order. operators are to the left of b

Examples are given below. Let us do it in steps. = q 1 (b + b) 2

1 1 1 b + bb + 1) 2 = (b + b)(b + b) = (b b + b b + bb + bb) = (b b + 2b q 2 2 2 3 = q 1 2 1 b + bb + 1)(b + b ) (b + b) = (b b + 2b q 2 2 2 1 b b + 3b + 3b + 3b bb + bbb ) = ... = (b b b + 3b 2 2

3 for yourself, filling in the appropriate steps. Finally Derive the formula for q 4 = q 1 3 1 3 ) 2 b + 3b + 3b + 3b (b ) 2 + (b )3 )(b + b ) = ... (b + b) = ((b ) + 3(b q 4 2
1 2 1 2 + q + q 4 in 2 2 q 2

Now derive the complete formula for the Hamiltonian H =

second quantization. As a check on your results confirm that your Hamiltonian is Hermitian. m , where you can make repeated use of the Next evaluate the matrix elements n H n = n + 1 n + 1 , b n = n n 1 , n m = . For example: relations b nm

b b m = m nb b m 1 = m n b m = m m + 1 n m + 1 = nb n ,m +1m m + 1 m = mH n Again you can check your results by confirming that n H

Let us now consider the linear variational principle using the following expansion for the trial function = n cn . You can create the matrix elements for the Hamiltonian in
n=0 M

Maple, in analogy to what is done in the sample Maple page on the web site. Choose

= 0.1 . Diagonalize the Hamiltonian for M=1,, 10 and plot the eigenvalues as a
function of the size of the basis set. You can see an example in the Maple sample file, of how to set things up. Confirm that your results obey the interleafing theorem. Also calculate the eigenvectors of the Hamiltonian in Maple. Do you observe something remarkable? In this context, why do about half the eigenvalues stay the same with every increase of the basis set? Or phrased differently why do all of the eigenvalues alternately lower , then stay the same as you increase the basis set 1 function at a time? 1 You can generate a double well potential by using the potential V (q ) = + q 2 q 2 + q 4 2
1 2 1 2 b q2 + q4 . + q q2 + q4 = b with corresponding Hamiltonian H = 2 2 q 2

Taking = 1 and = 0.05 defines a nice double well potential. Plot the total potential, including the pure harmonic part of the potential as a function of q . Next evaluate the m as before and plot the new energy eigenvalues. Here you have matrix elements n H obtained the numerical solutions of the double well potential. What happens to the eigenvalues if you increase the values for and/or decrease the value of . Always plot the potential to see if you are on the right track. Can you understand the pairing of the eigenvalues by sketching the eigen functions in a deep double well potential? This could be a model of the umbrella tunnelling vibration in the ammonia molecule. This full problem set is lengthy, but it shows you clearly how nice it can be to have a convenient Mathematics program!

8. Let us assume a ground state Hamiltonian H =

1 2 1 2 + q and the associated 2 q 2 2

, b . An electronically excited state has the same harmonic frequency, but its operators b equilibrium geometry is displaced by a distance 2d , in the reduced dimensionless coordinates. The excited state Hamiltonian is given by
1 2 1 H = E = 2 + = ( q 2 d ) 2 2 q 2

a. Show that the excited state Hamiltonian in terms of the displaced operators
 = b d , b =b d takes the form H  b  + 1 = .  = E + =b b 2 , b , and b. Show that the displaced operators satisfy the same commutation relations as b
 , 0   of the electronic excited state in terms of b express the vibrational eigenstates n c. What are the excitation energies of the various vibrational levels in the electronically excited state w.r.t. the ground state vibrational level? What is the 0-0 energy? What is the vertical excitation energy? d. The intensities in the absorption spectrum are proportional to the squares of the
 =  . Show that 0 n Franck-Condon factors, which are defined as 0 n
dn  . 00 n!

e. Use the completeness relation

 n

 = e d 2 / 2 .  n  0 = 0 0 = 1 to derive 0 0 0n

 n . What are the transition f. In the emission spectrum the transitions are given by 0
energies, and the corresponding intensities? g. Using Maple, and the following parameters: = = 1, E = 10, d = 1.0 . Plot the absorption spectrum and the emission spectrum on the same energy scale. Also investigate what happens if the displacement is smaller (d=0.5, or larger d=3.0).