0

The Stern Gerlach experiment in 3-Dimensions provides us with conditions on what properties Spin-Half vectors must satisfy, from which we can build our basis states in $x, y, \text{ and } z$. The conditions can be expressed as follows (where $i$ and $j$ are variables representing spatial dimensions):

$$\ \langle \pm_i | \pm_i \rangle=1, \langle \pm_i | \mp_i \rangle=0$$

$$\text{For } i \neq j : \big|\langle \pm_i | \pm_j \rangle \big|^2= \big| \langle \pm_i | \mp_j \rangle \big|^2=\frac{1}{2}$$

Of course there are infinitely many solutions to the 6 basis vectors, but as a convention we define:

$$|+_z\rangle = \begin{pmatrix} 1\\ 0 \\ \end{pmatrix}, |-_z\rangle = \begin{pmatrix} 0\\ 1 \\ \end{pmatrix}, |\pm_x\rangle = \frac{|+_z\rangle \pm |-_z\rangle }{\sqrt2}, |\pm_y\rangle = \frac{|+_z\rangle \pm i|-_z\rangle }{\sqrt2}$$

All other solutions can be seen as rotations of this convention about the Bloch Sphere.

From here it is clear that were this experiment only being done in 2-Dimensions, then complex numbers would not be required to define the vectors, since the $z$ and $x$ basis span 2-Dimensions and they can be constructed with only real numbers. Which leads me to wonder; what algebra would we have to use if we were in 4 spatial dimensions performing the Stern-Gerlach Experiment. Our conditions would be expressed identically as:

$$\ \langle \pm_i | \pm_i \rangle=1, \langle \pm_i | \mp_i \rangle=0$$

$$\text{For } i \neq j : \big|\langle \pm_i | \pm_j \rangle \big|^2= \big| \langle \pm_i | \mp_j \rangle \big|^2=\frac{1}{2}$$

However now $i$ and $j$ are indices over 4 spatial dimensions, let's say $w, x, y, z$. We need to define mathematical objects to represent our 8 basis vectors that satisfy these conditions; 6 of which is trivial, in that its identical to the 3-Dimensional case, but for the new dimension, can it be done with just complex numbers? Or do we have to resort to something else, potentially quaternions?

Qmechanic
  • 201,751
  • 3
    Your question doesn't really make sense because spin generalizes to other dimensions very differently from what you assume here (in particular, "spin half" is not a meaningful statement in higher dimensions, and in 2d you get anyons rather than the familiar distinction between half-integer and integer spin). See e.g. https://physics.stackexchange.com/a/422088/50583. – ACuriousMind Aug 26 '21 at 20:11
  • I agree that the proposal above is not how it's defined. But what's the problem with referring to spinorial representations as "spin half" in higher dimensions? – Connor Behan Aug 26 '21 at 21:25
  • I fear you are trying to replicate standard group theory in a language from hell. In 3D, you are considering the doublet representation of SO(3)~SU(2) which has 3 generators (Pauli matrices), so you have 2×3=6 states. If you wished to appreciate rotations in 4D, and note that SO(4)~SU(2)×SU(2), with 6 (two disjoint sets of Paulis) generators, you'd be looking at the doublet × doublet representation, so 12 states, a construction you might have covered in the Lorentz group. I'm not aware of a workable two Bloch spheres' direct product gimmick to get you where I suspect you dream of going. – Cosmas Zachos Aug 27 '21 at 15:37

0 Answers0