We look for zero products of the form (A±B)(C±D) where A, B, C and D are all products of basis elements of V. Letting a, b, c and d be the corresponding binary number and ignoring the sign or the term for now we see that the product will have 4 terms which must cancel by twos to produce 0. This means that a⊕b⊕c⊕d = 0. ‘⊕’ is exclusive-or. Our search logic has already produced the zero factors of the form x(A+B) which means that we will not find new values for a⊕b. Ditto c⊕d. This leaves 16 values (times two for sign ambiguity) for b that might lead to new 0 products.

Searching this small space for CL(4) Clifford numbers that my inversion code cannot invert finds 10 values:
1±γiγjγk for i<j<k, and 1±γ0γ1γ2γ3. Generalizations to higher order Clifford spaces are unclear from this.

Indeed for i<j<k (1+γiγjγk)(1−γiγjγk) = 0 = (1−γiγjγk)(1+γiγjγk). This detour has not revealed any more zero divisors not already evident from symmetry. It has ruled out a category of zero divisors. It reminds us to include some we have omitted, however. A superset of these, however, is the zero-divisors which are related to known divisors by orthogonal transformations on V.


((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . 0) -1 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . -1) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . -1) (0 . 0) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . -1) ((0 . 0) 0 . 0) (0 . 0) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . 0) 0 . 1)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . 0) 1 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . 1) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 1) (0 . 0) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 1) ((0 . 0) 0 . 0) (0 . 0) 0 . 0)
((((1 . 0) 0 . 0) (0 . 0) 0 . 0) ((0 . 0) 0 . 0) (0 . 0) 0 . -1)