Finding eigenvalues of a Hamiltonian involving Sz, Sz^2 and Sx

  • Context: Graduate 
  • Thread starter Thread starter Physicslad78
  • Start date Start date
  • Tags Tags
    Eigenvalues Hamiltonian
Click For Summary
SUMMARY

The discussion centers on finding the eigenvalues of a Hamiltonian for an S=5/2 particle, represented as H = a.Sz + b.Sz^2 + c.Sx, where Sz and Sx are spin operators. Users suggest employing the eigenvalue equation |H - E I| = 0 and utilizing software tools like Mathematica and Maple for symbolic eigenvalue calculations. The matrix is identified as a tridiagonal symmetric 6x6 matrix, and while some users report difficulties with the Solve function in Mathematica, they emphasize the complexity of manual diagonalization.

PREREQUISITES
  • Understanding of Hamiltonians in quantum mechanics
  • Familiarity with spin operators Sz and Sx
  • Knowledge of eigenvalue problems and matrix diagonalization
  • Experience with computational tools such as Mathematica or Maple
NEXT STEPS
  • Learn how to use Mathematica for symbolic eigenvalue calculations
  • Explore the properties of tridiagonal matrices in quantum mechanics
  • Study the Solve function in Mathematica for solving polynomial equations
  • Investigate similarity transformations and their applications in diagonalization
USEFUL FOR

Quantum physicists, computational scientists, and students studying quantum mechanics who are working on eigenvalue problems involving Hamiltonians.

Physicslad78
Messages
46
Reaction score
0
I have the Hamiltonian for an S=5/2 particle given by:

H= a.Sz + b.Sz^2 +c.Sx where Sz and Sx are the spins in z and x directions respectively. The resulting matrix is tridiagonal symmetric but I can't find the eigenvalues..Any idea how to diagonalise it.

N.B: a is a variable and must be kept as a in matrix whereas b and c can be assigned values.


Thanks guys.
 
Physics news on Phys.org
I guess you'd just use the normal eigenvalue equation
\left|H - E I\right| = 0
where I represents the identity matrix... that, or Mathematica ;-) (seriously, if you have access to some software tool which can find symbolic eigenvalues, it's probably not worth the work to diagonalize a 6x6 matrix)

If you have trouble with that, post the details of the calculations you did. Sometimes it's just as simple as a little math error.
 
Mathematica doesn't solve it..Eigenvalues produces a root solution and solving a 7 equation system with 7 unknowns (if I want to find the eigenvectors) doesn't help either..Maybe I need a similarity transformation and I have tried some but in vain..dont know what to do..:(
 
You can use the Solve function to explicitly find the roots in the solution in terms of a, but it is rather ugly :/ One thing's for sure, I wouldn't want to have to come up with that by hand!
 
I would use some kind of program, it's not simple to find the eigenvalue of a 6x6 matrix unless there happens to be some real tricks...and I don't think there are for tridiagonal matrices (I don't know of any). There should be 6 eigenvalues (and eigen-vectors) though, not 7...o.O
 
Thanks guys..I have tried Solve as well but it just runs forever! MatterrWave, yeah I meant 7 unknowns (6 are the elements of spinor and the 7th is the eigenvalue)..I will see if I can meddle with it and get an answer
 
Uhm, Maple has a eigenvalue/eigenvector finder, though I'm not sure if it will be any good at a 6x6 matrix. I'm going to try and see. :p

EDIT: wow, even constructing these matrices takes a while...I may do it later <_<
 
Last edited:
Thanks a lot Matter wave...waiting for ur results if u do it or if any..lol..meanwhile will try sorting it out and have got a non separable differential equation waiting for me as well..:)
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 27 ·
Replies
27
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
732
  • · Replies 0 ·
Replies
0
Views
3K
  • · Replies 4 ·
Replies
4
Views
4K
  • · Replies 0 ·
Replies
0
Views
1K