An exponential map from CL(4,R) to Spin(3,1) which is surjective?

31 Views Asked by At

I am trying to find an algebra whose exponential maps surjectively to Spin(3,1).

I understand that its lie algebra is $\mathfrak{so}(3,1)$, but I think it leaves elements out.

I am working with the bivectors of CL(4,R).

$$ \mathbf{f} = E_{01}\gamma_0\gamma_1+E_{02}\gamma_0\gamma_2 + E_{03}\gamma_0\gamma_3+B_{12}\gamma_1\gamma_2+B_{13}\gamma_1\gamma_3+B_{23}\gamma_2\gamma_3 $$

Its exponential is

$$ e^{\mathbf{f}}=\cos \theta+ \left( \sin \theta \right) \left( E_{01}\gamma_0\gamma_1+E_{02}\gamma_0\gamma_2 + E_{03}\gamma_0\gamma_3+B_{12}\gamma_1\gamma_2+B_{13}\gamma_1\gamma_3+B_{23}\gamma_2\gamma_3 \right) $$

where $\theta = || E_{01}\gamma_0\gamma_1+E_{02}\gamma_0\gamma_2 + E_{03}\gamma_0\gamma_3+B_{12}\gamma_1\gamma_2+B_{13}\gamma_1\gamma_3+B_{23}\gamma_2\gamma_3||$


However, the official definition of spin relates to multiplying an even number of vectors together whose norm is +1 (or -1).

Such multiplication can produce linear combinasions of scalars, pseudo-scalars and bivectors.

Yet, the exponential of a bivector can only produce scalars and bivectors. Thus, it must be incomplete.


I note that if I add a pseudo-scalar $\mathbf{z}$ to my bivector $\mathbf{f}$, I get

$$ \mathbf{f} + \mathbf{z} = E_{01}\gamma_0\gamma_1+E_{02}\gamma_0\gamma_2 + E_{03}\gamma_0\gamma_3+B_{12}\gamma_1\gamma_2+B_{13}\gamma_1\gamma_3+B_{23}\gamma_2\gamma_3 + Z\gamma_0\gamma_1\gamma_2\gamma_3 $$

Its exponential is:

$$ e^{\mathbf{f} + \mathbf{z}}=\cos \theta+ \left( \sin \theta \right) \left( E_{01}\gamma_0\gamma_1+E_{02}\gamma_0\gamma_2 + E_{03}\gamma_0\gamma_3+B_{12}\gamma_1\gamma_2+B_{13}\gamma_1\gamma_3+B_{23}\gamma_2\gamma_3 + Z \gamma_0\gamma_1\gamma_2\gamma_3\right) $$


From this, it appears that the exponential map of $\mathbf{f}+\mathbf{z}$ is surjective to ${\rm Spin}(3,1)$.

Is this correct?

Is it bijective?

Finally, is $\mathbf{f}+\mathbf{z}$=$\mathfrak{so}(3,1) \oplus \mathfrak{u}(1)$?