Model for a Qubit system using the Hamiltonion Operator

Lambda96
Messages
233
Reaction score
77
Homework Statement
Calculate the eigenvalues and eigenvectors of the Hamilton operator
Relevant Equations
none
Hi,

unfortunately, I am not sure if I have calculated the task a correctly.

Bildschirmfoto 2023-06-06 um 14.56.03.png

I calculated the eigenvalues with the usual formula ##\vec{0}=(H-\lambda I) \psi## and got the following results

$$\lambda_1=E_1=-\sqrt{B^2+\nabla^2}$$
$$\lambda_2=E_2=\sqrt{B^2+\nabla^2}$$

I'm just not sure about the normalized eigenvectors, I got the following as eigenvectors

$$\vec{\psi}_1= \left(\begin{array}{c} \frac{B- \sqrt{B^2+\triangle^2}}{\triangle} \\ 1 \end{array}\right)$$
$$\vec{\psi}_2= \left(\begin{array}{c} \frac{B+ \sqrt{B^2+\triangle^2}}{\triangle} \\ 1 \end{array}\right)$$

For the normalization, I must divide the vectors by their norm. Since writing this vector via Latex is very time-consuming, I have calculated this with Mathematica, instead of the triangle, I have simply written an A. For the eigenvector ##\vec{\psi_1}## I got the following:

Bildschirmfoto 2023-06-06 um 15.33.35.png

Since this eigenvector looks very messy, I am not sure if I have calculated the task correctly now.
 
Physics news on Phys.org
First you need to be less confusing with your symbols. Use \Delta for ##\Delta## and not ##\nabla##, ##\triangle## or ##A##.
According to our rules, to receive help, you need to show your work. The derivation of the normalized eigenvectors is straightforward but I cannot guide you to it unless I see what what you did. Try to use LaTeX, but if you must use Mathematica, did you know that you can export its equations in LaTeX?

Addendum on edit:
Your expression for ##|\psi_1\rangle## is not correct because it is not a constant times the unnormalized ##|\psi_1\rangle##. If you factor out the denominator, you are left with $$
\begin{pmatrix}
\frac{-B-\sqrt{B^2+\Delta ^2}}{\Delta} & 1
\end{pmatrix}$$while the unnormalized vector is
$$
\begin{pmatrix}
\frac{B-\sqrt{B^2+\Delta ^2}}{\Delta} & 1
\end{pmatrix}.$$I suggest that you replace the radical with a single symbol, e.g. ##R\equiv \sqrt{B^2+\Delta^2}.## It will make the algebra more transparent.
 
Last edited:
Thanks kuruman for your help 👍

Sorry, because of all the different symbols, unfortunately I still had in mind that it is the Nabla operator ##\nabla##. I had then changed all symbols by ##\triangle## and forgotten the change in the eigenvalues. But as you already said correctly, it should be ##\Delta##
kuruman said:
Try to use LaTeX, but if you must use Mathematica, did you know that you can export its equations in LaTeX?
I did not know that, thanks for the tip 👍
I have now been able to confirm by the following calculation that these are the eigenvectors of A

$$H \vec{\psi}_1=E_1 \vec{\psi}_1$$
$$H \vec{\psi}_2=E_2 \vec{\psi}_2$$

The normalized eigenvectors are then as follows.

$$\hat{\psi}_1=\frac{1}{\sqrt{\biggl( \frac{(B-\sqrt{B^2+\Delta^2})}{\Delta} \biggr)^2+1}} \cdot \left(\begin{array}{c} \frac{B-\sqrt{B^2+\Delta^2}}{\Delta} \\ 1 \end{array}\right)$$

$$\hat{\psi}_2=\frac{1}{\sqrt{\biggl( \frac{(B+\sqrt{B^2+\Delta^2})}{\Delta} \biggr)^2+1}} \cdot \left(\begin{array}{c} \frac{B+\sqrt{B^2+\Delta^2}}{\Delta} \\ 1 \end{array}\right)$$
 
This is how I would do it.

The unnormalized eigenvector corresponding to ##E_1## is $$|\psi_1\rangle=
\begin{pmatrix}
\frac{B-R}{\Delta} \\
1
\end{pmatrix}~~~R\equiv \sqrt{B^2+\Delta^2}.
$$Let ##N## be the normalization constant by which you must multiply the eigenvector. Then $$\begin{align} 1= & N^2\langle \psi_1 | \psi_1\rangle = \left(\frac{B-R}{\Delta}\right)^2+1= \frac{B^2-2BR+R^2+\Delta^2}{\Delta^2}=\frac{2R^2-2BR}{\Delta^2 }=\frac{2R(R-B)}{\Delta^2} \nonumber \\
& \implies N=\frac{\Delta}{\sqrt{2R(R-B)}}

\nonumber \end{align}$$and the normalized eigenvector is
$$|\psi_1\rangle=\frac{1}{\sqrt{2R(R-B)}}
\begin{pmatrix}
{B-R} \\
\Delta
\end{pmatrix}.
$$The normalized eigenvector corresponding to ##E_2## is obtained by swapping the column vector elements and introducing a relative minus sign. Obviously, the overall phase doesn't matter.
$$|\psi_2\rangle=\frac{1}{\sqrt{2R(R-B)}}
\begin{pmatrix}
-\Delta \\
B-R
\end{pmatrix}.$$It is immediately clear that the above vectors are orthonormal. In your expression that is not obvious. That is what I meant when I said that the substitution ##R=\sqrt{B^2+\Delta^2}## in such problems makes the algebra transparent.
 
Thanks again kuruman for your help and for the tip with the calculation of the normalization constant ##N## 👍👍.

I would have only one question, concerning the normalized eigenvector ##\hat{\psi}_2##.

I would have now calculated the normalization constant exactly as for ##\hat{\psi}_1## and got the following

$$N=\frac{\Delta}{\sqrt{2R(R+B)}}$$

I would then get the following as the normalized eigenvector.

$$\hat{\psi}_2=\frac{\Delta}{\sqrt{2R(R+B)}} \cdot \left(\begin{array}{c} B+R \\ 1 \end{array}\right)$$

Unfortunately, I don't quite understand why I also get the eigenvector when I swap the columns for the normalized eigenvector ##\hat{\psi}_1## and multiply by minus. I wanted to verify with the calculation ##H\hat{\psi}_2=E_2 \hat{\psi}_2## that it is an eigenvector of H and unfortunately I got the following.

$$H \cdot \left(\begin{array}{c} -\Delta \\ B-R \end{array}\right)=\frac{\Delta}{\sqrt{2R(R+B)}} \cdot \left(\begin{array}{c} -\Delta R\\ -\Delta^2-B(B-R) \end{array}\right)$$

The first line is correct, only with the second line I have problems.

Is it possible to determine the normalized eigenvectors of a 2x2 matrix with this trick, or only in certain cases?
 
First of all this $$\hat{\psi}_2=\frac{\Delta}{\sqrt{2R(R+B)}} \cdot \left(\begin{array}{c} B+R \\ 1 \end{array}\right)$$is dimensionally incorrect. The first element in the column vector has dimensions of energy while the second element is dimensionless. That cannot be. Also, I think you should have a term ##(R-B)## in the denominator, not ##(R+B).## Recheck your work.

Lambda96 said:
Is it possible to determine the normalized eigenvectors of a 2x2 matrix with this trick, or only in certain cases?
It is possible to use this trick with any 2×2 matrix. In the more general case where the eigenvectors have complex coefficients, let the normalized eigenvectors be
##|\psi_1\rangle=a_1|1\rangle +b_1|2\rangle~~~(|a_1|^2+|b_1|^2=1).##
##|\psi_2\rangle=a_2|1\rangle +b_2|2\rangle~~~(|a_2|^2+|b_2|^2=1).##
To ensure that they are orthogonal, you must have
$$0=\langle\psi_2|\psi_1\rangle=
\begin{pmatrix}
{a^*_2} & b^*_2
\end{pmatrix}
\begin{pmatrix}
{a_1} \\ b_1
\end{pmatrix}=a^*_2a_1+b^*_2b_1.
$$Once you obtain ##a_1## and ##b_1## for the first eigenvector, you can immediately write ##a_2\rightarrow -b^*_1## and ##b_2\rightarrow a^*_1## because then $$a^*_2a_1+b^*_2b_1\rightarrow (-b^*_1)^*a_1+ (a^*_1)^*b_1=-b_1a_1+a_1b_1=0.$$Of course, if the coefficients are real as in this case, you don't have to worry about the complex conjugates, just swap elements and introduce a relative negative sign.
 
Thanks kuruman for your help and your explanation for finding the eigenvectors of a 2x2 matrix👍👍, for future tasks, this will simplify the calculation :smile:
 
Back
Top