What is meant here, when it is said that "diagonal generalized Gell-Mann matrices are not unique"?

307 Views Asked by At

According to the answer given here, the diagonal generalized Gell-Mann matrices are not unique. But what exactly is meant by this?

  • Are they just saying that we can multiply the diagonal matrices by a constant and still have a valid generator of the group since the matrices will still be orthogonal? If so, then why are the other generalized Gell-Mann matrices not also being called non-unique (you can multiply them by a constant and still have an orthogonal set of matrices).

  • Or are they saying that we could have completely different diagonal matrices as long as they are traceless and satisfy orthogonality under the Hilbert-Schmidt norm: $(A,B) =Tr(AB)$ ? For example we could have $\textrm{diag}(1 ,0 , -1)$ and $\textrm{diag}( 1/\sqrt{3} , -2/\sqrt{3} , 1/\sqrt{3})$, or even something more radically different such as $\textrm{diag}(a,b,c)$ and $\textrm{diag}(d,e,f)$ where none of $a,b,c$ are $0$?

1

There are 1 best solutions below

0
On

If deriving the n×n matrices in the same way Gell-Mann did, there is only one choice for the diagonal matrices.

Gell-Mann generated the two diagonal 3×3 matrices by starting with the Pauli $z$ matrix and padding it with zeros to make it 3×3, and letting the other diagonal matrix be arbitrary:

$$\tag{1} \lambda_3 = \begin{pmatrix} 1 & 0 & 0\\ 0 & -1 & 0\\ 0 & 0 & 0 \\ \end{pmatrix}, \lambda_8=\begin{pmatrix} a & 0 & 0\\ 0 & b & 0\\ 0 & 0 & c \\ \end{pmatrix}. $$

Then he used the following constraints:

\begin{eqnarray} \textrm{Tr}(\lambda_3\lambda_8) = 0 \tag{2}\\ \textrm{Tr}(\lambda_8) = 0 \tag{3}\\ \textrm{Tr}(\lambda_8^2) = 2 .\tag{4} \end{eqnarray}

The first constraint forces us to choose $b = a$,
the second constraint forces us to choose $c = -2a$,
and the third constraint forces us to choose $a = \frac{1}{\sqrt{3}}$.

Let's now derive the three diagonal 4×4 matrices in the same, way, by starting with the 3×3 matrices padded with zeros, and adding an arbitrary diagonal matrix:

$$\tag{5} \Lambda_3 = \begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & -1 & 0 & 0 \\ 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 \\ \end{pmatrix}, \Lambda_8=\frac{1}{\sqrt{3}}\begin{pmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & -2 & 0 \\ 0 & 0 & 0 & 0 \\ \end{pmatrix}, \Lambda_{15} = \begin{pmatrix} a & 0 & 0 & 0 \\ 0 & b & 0 & 0 \\ 0 & 0 & c & 0 \\ 0 & 0 & 0 & d \\ \end{pmatrix}. $$

Then we have the following constraints:

\begin{eqnarray} \textrm{Tr}(\Lambda_3\Lambda_{15}) = 0 \tag{6}\\ \textrm{Tr}(\Lambda_8\Lambda_{15}) = 0 \tag{7}\\ \textrm{Tr}(\Lambda_{15}) = 0 \tag{8}\\ \textrm{Tr}(\Lambda_{15}^2) = 2 .\tag{9} \end{eqnarray}

The first constraint forces us to choose $b = a$
The second constraint forces us to choose $c = a$
The third constraint forces us to choose $d = -3a$
The fourth constraint forces us to choose $a = \frac{1}{\sqrt{6}}$.

This matches when you see on Pg 368 of this book (PDF), Eq. 11 of this paper (PDF), Pg 7 of this talk (PDF), Eq 3 of this paper (PDF), and the definition of "generalized Gell-Mann matrices" given here.

In general we carry over padded versions of the $n-2$ diagonal matrices from the $(n-1)$×$(n-1)$ case, and the orthogonality conditions force us to make all entries of the last diagonal matrix to be equal except for the last entry. The last entry $x$ has to be the negative of the sum of the rest of the entries (which are all the same, call them $a$ so $x = -(n-1)a$), due to the condition that the matrices are all traceless. The condition that the trace of each matrix multiplied by itself is 2, gives us only one choice for $a$.