Higher-dimensional gamma matrices

In mathematical physics, higher-dimensional gamma matrices are the matrices which satisfy the Clifford algebra

 \{ \Gamma_a ~,~ \Gamma_b \} = 2 \eta_{a b} I_N

with the metric given by

 \eta = \parallel \eta_{a b} \parallel = \text{diag}(%2B1,-1, \dots, -1)

where  a,b = 0,1, \dots, d-1 and  I_N the identity matrix in  N= 2^{[d/2]} dimensions.

They have the following property under hermitian conjugation

 \Gamma_0^\dagger= %2B\Gamma_0 ~,~ \Gamma_i^\dagger= -\Gamma_i
~(i=1,\dots,d-1)

Contents

Charge conjugation

Since the groups generated by \ \Gamma_a ,  -\Gamma_a^T ,  \Gamma_a^T are the same we deduce from Schur's lemma that there must exist a similarity transformation which connect them. This transformation is generated by the charge conjugation matrix. Explicitly we can introduce the following matrices

 C_{(%2B)} \Gamma_a  C_{(%2B)}^{-1} = %2B \Gamma_a^T
 C_{(-)} \Gamma_a  C_{(-)}^{-1} = - \Gamma_a^T

They can be constructed as real matrices in various dimensions as the following table shows

D  C^*_{(%2B)}= C_{(%2B)}  C^*_{(-)}= C_{(-)}
 2  C^T_{(%2B)}=C_{(%2B)};~~~C^2_{(%2B)}=1  C^T_{(-)}=-C_{(-)};~~~C^2_{(-)}=-1
 3  C^T_{(-)}=-C_{(-)};~~~C^2_{(-)}=-1
 4  C^T_{(%2B)}=-C_{(%2B)};~~~C^2_{(%2B)}=-1  C^T_{(-)}=-C_{(-)};~~~C^2_{(-)}=-1
 5  C^T_{(%2B)}=-C_{(%2B)};~~~C^2_{(%2B)}=-1
 6  C^T_{(%2B)}=-C_{(%2B)};~~~C^2_{(%2B)}=-1  C^T_{(-)}=C_{(-)};~~~C^2_{(-)}=1
 7  C^T_{(-)}=C_{(-)};~~~C^2_{(-)}=1
 8  C^T_{(%2B)}=C_{(%2B)};~~~C^2_{(%2B)}=1  C^T_{(-)}=C_{(-)};~~~C^2_{(-)}=1
 9  C^T_{(%2B)}=C_{(%2B)};~~~C^2_{(%2B)}=1
 10  C^T_{(%2B)}=C_{(%2B)};~~~C^2_{(%2B)}=1  C^T_{(-)}=-C_{(-)};~~~C^2_{(-)}=-1
 11  C^T_{(-)}=C_{(-)};~~~C^2_{(-)}=-1

Symmetry properties

A  \Gamma matrix is called symmetric if

 ( C \Gamma_{a_1 \dots a_n} )^T = %2B  ( C \Gamma_{a_1 \dots a_n} )

otherwise it is called antisymmetric. In the previous expression  C can be either  C_{(%2B)} or  C_{(-)} . In odd dimension there is not ambiguity but in even dimension it is better to choose whichever one of  C_{(%2B)} or  C_{(-)} which allows for Majorana spinors. In  D=6 there is not such criterion and therefore we consider both.

D C Symmetric Antisymmetric
 3  C_{(-)}  \gamma_{a}  I_2
 4  C_{(-)}  \gamma_{a} ~,~ \gamma_{a_1 a_2}  I_4 ~,~  \gamma_\text{chir} ~,~ \gamma_\text{chir} \gamma_a
 5  C_{(%2B)}  \Gamma_{a_1 a_2}  I_4 ~,~  \Gamma_a
 6  C_{(-)}  I_8 ~,~ \Gamma_\text{chir} \Gamma_{a_1 a_2} ~,~ \Gamma_{a_1 a_2 a_3}  \Gamma_a ~,~ \Gamma_\text{chir}~,~ \Gamma_\text{chir} \Gamma_a ~,~ \Gamma_{a_1 a_2}
 7  C_{(-)}  I_8 ~,~ \Gamma_{a_1 a_2 a_3}  \Gamma_a ~,~ \Gamma_{a_1 a_2}
 8  C_{(%2B)}  I_{16} ~,~ \Gamma_{a} ~,~ \Gamma_\text{chir} ~,~ \Gamma_\text{chir}\Gamma_{a_1 a_2 a_3} ~,~ \Gamma_{a_1 \dots a_4}  \Gamma_\text{chir} \Gamma_a ~,~ \Gamma_{a_1 a_2} ~,~ \Gamma_\text{chir} \Gamma_{a_1 a_2} ~,~ \Gamma_{a_1 a_2 a_3}
 9  C_{(%2B)}  I_{16} ~,~ \Gamma_{a} ~,~ \Gamma_{a_1 \dots a_4} ~,~ \Gamma_{a_1 \dots a_5}  \Gamma_{a_1 a_2} ~,~ \Gamma_{a_1 a_2 a_3}
 10  C_{(-)}  \Gamma_{a} ~,~ \Gamma_\text{chir} ~,~ \Gamma_\text{chir} \Gamma_a ~,~ \Gamma_{a_1 a_2}
~,~ \Gamma_\text{chir} \Gamma_{a_1 \dots a_4}  ~,~ \Gamma_{a_1 \dots a_5}  I_{32} ~,~ \Gamma_\text{chir} \Gamma_{a_1 a_2} ~,~ \Gamma_{a_1 a_2 a_3}
~,~ \Gamma_{a_1 \dots a_4}  ~,~  \Gamma_\text{chir} \Gamma_{a_1 a_2 a_3}
 11  C_{(-)}  \Gamma_a ~,~ \Gamma_{a_1 a_2} ~,~ \Gamma_{a_1 \dots a_5}  I_{32} ~,~ \Gamma_{a_1 a_2 a_3} ~,~ \Gamma_{a_1 \dots a_4}

Example of an explicit construction in chiral base

We construct the  \Gamma matrices in a recursive way, first in all even dimensions and then in odd ones.

d = 2

We take

 \gamma_0= \sigma_1 ~,~ \gamma_1= -i \sigma_2

and we can easily check that the charge conjugation matrices are

 C_{(%2B)}= \sigma_1 = C_{(%2B)}^* = s_{(2,%2B)} C_{(%2B)}^T = s_{(2,%2B)} C_{(%2B)}^{-1} ~~~~ s_{(2,%2B)}=%2B1
 C_{(-)}= i \sigma_2 = C_{(-)}^* = s_{(2,-)} C_{(-)}^T = s_{(2,-)} C_{(-)}^{-1} ~~~~ s_{(2,-)}=-1

We can also define the hermitian chiral  \gamma_\text{chir} to be

 \gamma_\text{chir}= \gamma_0 \gamma_1 = \sigma_3 = \gamma_\text{chir}^\dagger

generic even d = 2k

We now construct the  \Gamma_a (  a=0,\dots d%2B1 ) matrices and the charge conjugations  C_{(\pm)} in  d%2B2 dimensions starting from the  \gamma_{a'} ( a'=0, \dots, d-1 ) and  c_{(\pm)} matrices in  d dimensions. Explicitly we have

 \Gamma_{a'} = \gamma_{a'} \otimes \sigma_3  ~(a'=0, \dots, d-1) ~~,~~ \Gamma_{d} = I \otimes (i \sigma_1),~~ \Gamma_{d%2B1}= I \otimes (i \sigma_2)

Then we can construct the charge conjugation matrices

 C_{(%2B)} = c_{(-)} \otimes \sigma_1 ~~~~,~~~~ C_{(-)} = c_{(%2B)} \otimes (i \sigma_2)

with the following properties

 C_{(%2B)}= C_{(%2B)}^* = s_{(d%2B2,%2B)} C_{(%2B)}^T = s_{(d%2B2,%2B)} C_{(%2B)}^{-1} ~~~~ s_{(d%2B2,%2B)}= s_{(d,-)}
 C_{(-)}= C_{(-)}^* = s_{(d%2B2,-)} C_{(-)}^T = s_{(d%2B2,-)} C_{(-)}^{-1} ~~~~ s_{(d%2B2,-)}=-s_{(d,%2B)}

Starting from the values for d=2,  s_{(2,%2B)}=%2B1,~~~ s_{(2,-)}=-1 we can compute all the signs s_{(d,\pm)} which have a periodicity of 8, explicitly we find

 d=8 k  d=8 k%2B2  d=8 k%2B4  d=8 k%2B6
 s_{(d,%2B)} +1 +1 −1 −1
 s_{(d,-)} +1 −1 −1 +1

Again we can define the hermitian chiral matrix in d%2B2 dimensions as

 \Gamma_\text{chir}=  \alpha_{d%2B2} \Gamma_0 \Gamma_1 \dots \Gamma_{d-1} =  \gamma_\text{chir} \otimes \sigma_3
~~~~ \alpha_d= i^{d/2-1}

which is diagonal by construction and transforms under charge conjugation as

 C_{(\pm)} \Gamma_\text{chir} C_{(\pm)}^{-1} = \beta_{d%2B2} \Gamma_\text{chir}^T
~~~~ \beta_d= (-)^{d(d-1)/2}

generic odd d = 2k + 1

We consider the previous construction for  d-1 (which is even) and then we simply take all  \Gamma_{a} ( a=0, \dots, d-2 ) matrices to which we add  \Gamma_{d-1}= i \Gamma_\text{chir} ( the  i is there in order to have an antihermitian matrix).

Finally we can compute the charge conjugation matrix: we have to choose between  C_{(%2B)} and  C_{(-)} in such a way that  \Gamma_{d-1} transforms as all the others  \Gamma matrices. Explicitly we require

 C_{(s)} \Gamma_\text{chir} C_{(s)}^{-1} = \beta_{d} \Gamma_\text{chir}^T = s \Gamma_\text{chir}^T