Combine Hamiltonians of two different bases

  • Context: Graduate 
  • Thread starter Thread starter TheDestroyer
  • Start date Start date
  • Tags Tags
    Bases
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
13 replies · 2K views
TheDestroyer
Messages
401
Reaction score
1
I'm trying to solve a dynamical quantum mechanics problem related to the Cs atom, but I'm having trouble in the following, and I'm afraid I'm doing it wrong.

Say I have the matrix form of the Hamiltonian on a basis for a system [itex]| \psi \rangle[/itex] to be [itex]H_\psi[/itex], and another system with bases [itex]| \phi \rangle[/itex] with Hamiltonian [itex]H_\phi[/itex].

Now I would like to introduce interactions between the first and seconed systems, which will become (I suppose) [itex]| \psi \phi\rangle[/itex].

1) How do I combine those hamiltonians in matrix formalism before adding the new hamiltonian that introduces the interactions?

2) Say the first and second system are complete angular momenta basis in Zeeman eigen-states (so [itex]|F_1 m_{F_1} \rangle[/itex] and [itex]|F_2 m_{F_2}\rangle[/itex]). If the Wigner-D rotation matrix for the first Hamiltonian/basis is [itex]\mathcal{D}^{F_1}_{m_{F_1} m_{F_1}^\prime}[/itex], and for the second system is [itex]\mathcal{D}^{F_2}_{m_{F_2} m_{F_2}^\prime}[/itex]. How can I rotate each system and then combine them to introduce interactions between them?

Thank you for any efforts.
 
Physics news on Phys.org
TheDestroyer said:
Say I have the matrix form of the Hamiltonian on a basis for a system [itex]| \psi \rangle[/itex] to be [itex]H_\psi[/itex], and another system with bases [itex]| \phi \rangle[/itex] with Hamiltonian [itex]H_\phi[/itex].

Now I would like to introduce interactions between the first and seconed systems, which will become (I suppose) [itex]| \psi \phi\rangle[/itex].

You suppose correct. For this situation we create a Hilbert Space by the tensor product of basis of the two spaces containing Hamiltonians [itex]\hat{H}_\psi[/itex] and [itex]\hat{H}_\phi[/itex](which are spanned by [itex]| \psi \rangle[/itex] and [itex]| \phi \rangle[/itex] respectively).

TheDestroyer said:
1) How do I combine those hamiltonians in matrix formalism before adding the new hamiltonian that introduces the interactions?

So the new Hilbert Space is now spanned by basis [itex]| \psi_i \phi_j\rangle[/itex] or [itex]| \psi_i\rangle \otimes |\phi_j\rangle[/itex] (sum over i and j). The Hamiltonian of the system now becomes [itex]\hat{H}_\psi \otimes I + I \otimes \hat{H}_\phi + \hat{H}_{int}[/itex]. If you are not aware of this algebra, I would recommend Quantum Computation and Quantum Information by Neilsen and Chaung, Chapter 2.
 
Last edited:
Thank you so much for your answer. Let me though put an example to make this clear, because I got some result that I don't believe. Say I have the Hamiltonian matrices [itex]H_1[/itex] and [itex]H_2[/itex]

[tex]H_1=\left(<br /> \begin{array}{cccc}<br /> \text{s1} & 0 & 0 & 0 \\<br /> 0 & \text{s2} & 0 & 0 \\<br /> 0 & 0 & \text{s3} & 0 \\<br /> 0 & 0 & 0 & \text{s4} \\<br /> \end{array}<br /> \right)[/tex]

[tex]H_2=\left(<br /> \begin{array}{ccc}<br /> \text{t1} & 0 & 0 \\<br /> 0 & \text{t2} & 0 \\<br /> 0 & 0 & \text{t3} \\<br /> \end{array}<br /> \right)[/tex]

So now we have:

[tex]H_1 \otimes I_3 = \left(<br /> \begin{array}{cccccccccccc}<br /> \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} \\<br /> \end{array}<br /> \right)[/tex]

and

[tex]H_2 \otimes I_3 = \left(<br /> \begin{array}{cccccccccccc}<br /> \text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & \text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & \text{t3} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & \text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & \text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & \text{t3} & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & \text{t1} & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t2} & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t3} & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t1} & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t2} & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t3} \\<br /> \end{array}<br /> \right)[/tex]

So now the new Hamiltonian is

[tex]H_1 \otimes I_3 + I_4 \otimes H_2 = \left(<br /> \begin{array}{cccccccccccc}<br /> \text{s1}+\text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & \text{s1}+\text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & \text{s1}+\text{t3} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & \text{s2}+\text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & \text{s2}+\text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & \text{s2}+\text{t3} & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & \text{s3}+\text{t1} & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3}+\text{t2} & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3}+\text{t3} & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4}+\text{t1} & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4}+\text{t2} & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4}+\text{t3} \\<br /> \end{array}<br /> \right)[/tex]

Is that how it's done?
 
You have done correctly with some typos.

[tex]H_1=\left(<br /> \begin{array}{cccc}<br /> \text{s1} & 0 & 0 & 0 \\<br /> 0 & \text{s2} & 0 & 0 \\<br /> 0 & 0 & \text{s3} & 0 \\<br /> 0 & 0 & 0 & \text{s4} \\<br /> \end{array}<br /> \right),[/tex]

and define identity operator on first space

[tex]I_1=\left(<br /> \begin{array}{cccc}<br /> 1 & 0 & 0 & 0 \\<br /> 0 & 1 & 0 & 0 \\<br /> 0 & 0 & 1 & 0 \\<br /> 0 & 0 & 0 & 1 \\<br /> \end{array}<br /> \right).[/tex]
Similarly
[tex]H_2=\left(<br /> \begin{array}{ccc}<br /> \text{t1} & 0 & 0 \\<br /> 0 & \text{t2} & 0 \\<br /> 0 & 0 & \text{t3} \\<br /> \end{array}<br /> \right),[/tex]
and identity operator on second space
[tex]I_2=\left(<br /> \begin{array}{ccc}<br /> 1 & 0 & 0 \\<br /> 0 & 1 & 0 \\<br /> 0 & 0 & 1 \\<br /> \end{array}<br /> \right).[/tex]

The total Hamiltonian now should be [itex]\hat{H}_1 \otimes I_2 + I_1 \otimes \hat{H}_2 + \hat{H}_{int}[/itex]. We can evaluate terms as follows

[tex]H_1 \otimes I_2 = \left(<br /> \begin{array}{cccccccccccc}<br /> \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & \text{s1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & \text{s2} & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s3} & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{s4} \\<br /> \end{array}<br /> \right),[/tex]
as you have done correctly.
Second term is
[tex]I_1 \otimes H_2 = \left(<br /> \begin{array}{cccccccccccc}<br /> \text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & \text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & \text{t3} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & \text{t1} & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & \text{t2} & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & \text{t3} & 0 & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & \text{t1} & 0 & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t2} & 0 & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t3} & 0 & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t1} & 0 & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t2} & 0 \\<br /> 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & \text{t3} \\<br /> \end{array}<br /> \right).[/tex]
(you did correctly but it seems lot of typesetting made you overlook the order [itex]I_1 \otimes H_2[/itex] :)).
And you can add these matrices to get total Hamiltonian matrix (without interaction).
 
Hmmmmmmm... Thank you so much. Though I thought this approach would solve my problem, but apparently it doesn't.

I have Hyperfine states with [itex]F=3,4[/itex] (where [itex]F=I+J[/itex] is the sum of the total angular momentum and the nuclear spin), and I want to separate them, do a rotation with Wigner-D matrices, and combine them back. Do you know how this could be done?
 
TheDestroyer said:
I have Hyperfine states with [itex]F=3,4[/itex] (where [itex]F=I+J[/itex] is the sum of the total angular momentum and the nuclear spin), and I want to separate them, do a rotation with Wigner-D matrices, and combine them back. Do you know how this could be done?

I am sorry, I don't know what Wigner-D matrices are.
 
Thank you, I appreciate your help so far. You could though (if you're interested) look at this further. Wigner-D matrices are simply the matrices that provide rotations in the angular momentum basis with Euler angles.

So you have a basis of [itex]F=3[/itex] and [itex]m_F=-3,...,3[/itex], the Wigner-D matrix will be [itex]D_{m_F m_F^\prime}^{F=3}[/itex], where [itex]m_F, m_F^\prime[/itex] will be the matrix elements, and we'll have [itex]7\times 7=49[/itex] matrix elements for this basis.

That's the idea simply, so now, with that given, I know how to rotate the basis F=3 or F=4, but I don't know how to rotate them both when they're combined. You see where my problem is?
 
TheDestroyer said:
That's the idea simply, so now, with that given, I know how to rotate the basis F=3 or F=4, but I don't know how to rotate them both when they're combined. You see where my problem is?
Please give one demonstration. Say I need to rotate basis F=3. How will I proceed?
 
Thank you for your time. Wigner D-Matrix for that case will be the following, starting from the eigen-vector [itex]|F=3;m_F=-3 \rangle[/itex] to [itex]|F=3;m_F=3 \rangle[/itex]

[tex] D^{F=3}=\left(<br /> \begin{array}{ccccccc}<br /> e^{-3 i \alpha -3 i \gamma } \cos ^6\left(\frac{\beta }{2}\right) & -\sqrt{6} e^{-3 i \alpha -2 i \gamma } \cos ^5\left(\frac{\beta }{2}\right) \sin \left(\frac{\beta }{2}\right) & \sqrt{15} e^{-3 i \alpha -i \gamma } \cos ^4\left(\frac{\beta }{2}\right) \sin ^2\left(\frac{\beta }{2}\right) & -2 \sqrt{5} e^{-3 i \alpha } \cos ^3\left(\frac{\beta }{2}\right) \sin ^3\left(\frac{\beta }{2}\right) & \sqrt{15} e^{i \gamma -3 i \alpha } \cos ^2\left(\frac{\beta }{2}\right) \sin ^4\left(\frac{\beta }{2}\right) & -\sqrt{6} e^{2 i \gamma -3 i \alpha } \cos \left(\frac{\beta }{2}\right) \sin ^5\left(\frac{\beta }{2}\right) & e^{3 i \gamma -3 i \alpha } \sin ^6\left(\frac{\beta }{2}\right) \\<br /> \sqrt{6} e^{-2 i \alpha -3 i \gamma } \cos ^5\left(\frac{\beta }{2}\right) \sin \left(\frac{\beta }{2}\right) & \frac{1}{2} e^{-2 i \alpha -2 i \gamma } \cos ^4\left(\frac{\beta }{2}\right) (6 \cos (\beta )-4) & -\frac{1}{2} \sqrt{\frac{5}{2}} e^{-2 i \alpha -i \gamma } \cos ^3\left(\frac{\beta }{2}\right) (6 \cos (\beta )-2) \sin \left(\frac{\beta }{2}\right) & \sqrt{30} e^{-2 i \alpha } \cos ^2\left(\frac{\beta }{2}\right) \cos (\beta ) \sin ^2\left(\frac{\beta }{2}\right) & -\frac{1}{2} \sqrt{\frac{5}{2}} e^{i \gamma -2 i \alpha } \cos \left(\frac{\beta }{2}\right) (6 \cos (\beta )+2) \sin ^3\left(\frac{\beta }{2}\right) & \frac{1}{2} e^{2 i \gamma -2 i \alpha } (6 \cos (\beta )+4) \sin ^4\left(\frac{\beta }{2}\right) & -\sqrt{6} e^{3 i \gamma -2 i \alpha } \cos \left(\frac{\beta }{2}\right) \sin ^5\left(\frac{\beta }{2}\right) \\<br /> \sqrt{15} e^{-i \alpha -3 i \gamma } \cos ^4\left(\frac{\beta }{2}\right) \sin ^2\left(\frac{\beta }{2}\right) & \frac{1}{2} \sqrt{\frac{5}{2}} e^{-i \alpha -2 i \gamma } \cos ^3\left(\frac{\beta }{2}\right) (6 \cos (\beta )-2) \sin \left(\frac{\beta }{2}\right) & e^{-i \alpha -i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+5 (\cos (\beta )-1)+1\right) & -\frac{2 e^{-i \alpha } \cos \left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+\frac{15}{2} (\cos (\beta )-1)+3\right) \sin \left(\frac{\beta }{2}\right)}{\sqrt{3}} & e^{i \gamma -i \alpha } \left(\frac{15}{4} (\cos (\beta )-1)^2+10 (\cos (\beta )-1)+6\right) \sin ^2\left(\frac{\beta }{2}\right) & -\frac{1}{2} \sqrt{\frac{5}{2}} e^{2 i \gamma -i \alpha } \cos \left(\frac{\beta }{2}\right) (6 \cos (\beta )+2) \sin ^3\left(\frac{\beta }{2}\right) & \sqrt{15} e^{3 i \gamma -i \alpha } \cos ^2\left(\frac{\beta }{2}\right) \sin ^4\left(\frac{\beta }{2}\right) \\<br /> 2 \sqrt{5} e^{-3 i \gamma } \cos ^3\left(\frac{\beta }{2}\right) \sin ^3\left(\frac{\beta }{2}\right) & \sqrt{30} e^{-2 i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \cos (\beta ) \sin ^2\left(\frac{\beta }{2}\right) & \frac{2 e^{-i \gamma } \cos \left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+\frac{15}{2} (\cos (\beta )-1)+3\right) \sin \left(\frac{\beta }{2}\right)}{\sqrt{3}} & \frac{1}{2} \left(5 \cos ^3(\beta )-3 \cos (\beta )\right) & -\frac{2 e^{i \gamma } \cos \left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+\frac{15}{2} (\cos (\beta )-1)+3\right) \sin \left(\frac{\beta }{2}\right)}{\sqrt{3}} & \sqrt{30} e^{2 i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \cos (\beta ) \sin ^2\left(\frac{\beta }{2}\right) & -2 \sqrt{5} e^{3 i \gamma } \cos ^3\left(\frac{\beta }{2}\right) \sin ^3\left(\frac{\beta }{2}\right) \\<br /> \sqrt{15} e^{i \alpha -3 i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \sin ^4\left(\frac{\beta }{2}\right) & \frac{1}{2} \sqrt{\frac{5}{2}} e^{i \alpha -2 i \gamma } \cos \left(\frac{\beta }{2}\right) (6 \cos (\beta )+2) \sin ^3\left(\frac{\beta }{2}\right) & e^{i \alpha -i \gamma } \left(\frac{15}{4} (\cos (\beta )-1)^2+10 (\cos (\beta )-1)+6\right) \sin ^2\left(\frac{\beta }{2}\right) & \frac{2 e^{i \alpha } \cos \left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+\frac{15}{2} (\cos (\beta )-1)+3\right) \sin \left(\frac{\beta }{2}\right)}{\sqrt{3}} & e^{i \alpha +i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \left(\frac{15}{4} (\cos (\beta )-1)^2+5 (\cos (\beta )-1)+1\right) & -\frac{1}{2} \sqrt{\frac{5}{2}} e^{i \alpha +2 i \gamma } \cos ^3\left(\frac{\beta }{2}\right) (6 \cos (\beta )-2) \sin \left(\frac{\beta }{2}\right) & \sqrt{15} e^{i \alpha +3 i \gamma } \cos ^4\left(\frac{\beta }{2}\right) \sin ^2\left(\frac{\beta }{2}\right) \\<br /> \sqrt{6} e^{2 i \alpha -3 i \gamma } \cos \left(\frac{\beta }{2}\right) \sin ^5\left(\frac{\beta }{2}\right) & \frac{1}{2} e^{2 i \alpha -2 i \gamma } (6 \cos (\beta )+4) \sin ^4\left(\frac{\beta }{2}\right) & \frac{1}{2} \sqrt{\frac{5}{2}} e^{2 i \alpha -i \gamma } \cos \left(\frac{\beta }{2}\right) (6 \cos (\beta )+2) \sin ^3\left(\frac{\beta }{2}\right) & \sqrt{30} e^{2 i \alpha } \cos ^2\left(\frac{\beta }{2}\right) \cos (\beta ) \sin ^2\left(\frac{\beta }{2}\right) & \frac{1}{2} \sqrt{\frac{5}{2}} e^{2 i \alpha +i \gamma } \cos ^3\left(\frac{\beta }{2}\right) (6 \cos (\beta )-2) \sin \left(\frac{\beta }{2}\right) & \frac{1}{2} e^{2 i \alpha +2 i \gamma } \cos ^4\left(\frac{\beta }{2}\right) (6 \cos (\beta )-4) & -\sqrt{6} e^{2 i \alpha +3 i \gamma } \cos ^5\left(\frac{\beta }{2}\right) \sin \left(\frac{\beta }{2}\right) \\<br /> e^{3 i \alpha -3 i \gamma } \sin ^6\left(\frac{\beta }{2}\right) & \sqrt{6} e^{3 i \alpha -2 i \gamma } \cos \left(\frac{\beta }{2}\right) \sin ^5\left(\frac{\beta }{2}\right) & \sqrt{15} e^{3 i \alpha -i \gamma } \cos ^2\left(\frac{\beta }{2}\right) \sin ^4\left(\frac{\beta }{2}\right) & 2 \sqrt{5} e^{3 i \alpha } \cos ^3\left(\frac{\beta }{2}\right) \sin ^3\left(\frac{\beta }{2}\right) & \sqrt{15} e^{3 i \alpha +i \gamma } \cos ^4\left(\frac{\beta }{2}\right) \sin ^2\left(\frac{\beta }{2}\right) & \sqrt{6} e^{3 i \alpha +2 i \gamma } \cos ^5\left(\frac{\beta }{2}\right) \sin \left(\frac{\beta }{2}\right) & e^{3 i \alpha +3 i \gamma } \cos ^6\left(\frac{\beta }{2}\right) \\<br /> \end{array}<br /> \right)[/tex]

where the angles are the Euler angles, and the matrix is a unitary matrix, so I can use it to transform any Hamiltonian in the F=3 basis as follows:

[tex] H^\prime=D^\dagger H D[/tex]

The calculation of that matrix is another story, I got that from Mathematica. So my question is: assume I know that matrix for F=3 and F=4, and I know the Hamiltonians for both F=3 and F=4. How can I calculate the combined Hamiltonian after applying an arbitrary (but equal for each) transformations?

In other words: Say I use the matrix to rotate the F=3 Hamiltonian and F=4 Hamiltonian by some angles (a,b,c). Now I have 2 new Hamiltonians for each case. How can I combine them both?
 
Last edited:
The hamiltonian matrix is
[itex]\hat{H} = \hat{H}_1 \otimes I_2 + I_1 \otimes \hat{H}_2 + \hat{H}_{int}[/itex]. The rotation matrix in the new space should be defined as
[itex]D = D^{F=3} \otimes I_2 + I_1 \otimes D^{F=4}[/itex]. And now you can use your formula
[tex] \hat{H}^\prime=D^\dagger \hat{H} D.[/tex]
 
Last edited:
If I do this product the result's side length will become 7*9=63 elements... while my combined system from F=3 and F=4 has only 7+9=16 elements, since F=3 has side-length 2*3+1=7 and F=4 has 2*4+1=9.

So here's the problem right now.
 
TheDestroyer said:
If I do this product the result's side length will become 7*9=63 elements... while my combined system from F=3 and F=4 has only 7+9=16 elements.
This seems to be about the difference between the direct product and the direct sum. The dimension of the product is n*m while the dimension of the sum is n+m. You get the direct sum if you write the initial matrices on the diagonal of the resulting matrix and fill the off-diagonals with zeros.

If you combine two systems, you need the direct product. But what you want to do is to combine different possible states of a single system, don't you? There, the direct sum seems to be the right thing to me.

I've never used Wigner's matrix but I remember that I found Sakurai really insightful on it at the time. Maybe it helps you.

/edit: typo
 
Last edited:
TheDestroyer said:
while my combined system from F=3 and F=4 has only 7+9=16 elements, since F=3 has side-length 2*3+1=7 and F=4 has 2*4+1=9.
Your combined system also has 69 elements. Consider the first term of Hamiltonian. H (7*7 matrix) of system 1 tensored with Identity of system 2 (9*9 matrix) giving you a total of 63*63 matrix. Similarly the second term is 63*63 matrix.
 
Thank you for your answers, guys.

Like Kith said, I need to add components and not have the product. A 63*63 matrix is redundant, where it gives me, for example, a state [itex]| F=3,F=4,m_F=3,m_F=-4 \rangle[/itex], which is definitely wrong. What I'm doing is adding components... please consider re-assessing the situation.