Left Divisors of Zero in the Ring of Upper Triangular Matrices

Left Divisors of Zero in the Ring of Upper Triangular Matrices

Understanding the left divisors of zero in the ring of upper triangular matrices is a fascinating exploration in ring theory. Given the notation begin{bmatrix} Z Z_2 0 Z_2 end{bmatrix}, we are dealing with matrices of the form:

[ M begin{pmatrix} m a 0 b end{pmatrix} ]

where m in Z and ab in Z/2Z end{pmatrix}, and the right column is considered modulo 2 to stay in Z/2Z.

Definition of Left Zero Divisors

A matrix M is a left zero divisor if there exists a non-zero matrix N such that MN 0. Let's explore this concept with a general element N of the form:

[ N begin{pmatrix} n c 0 d end{pmatrix} ]

Product of Matrices

Computing the product MN yields:

[ MN begin{pmatrix} mn overline{mcad} 0 overline{bd} end{pmatrix} ]

where the bars on the right column indicate elements are interpreted modulo 2.

Conditions for Zero Product

The requirement for MN 0 is:

[ begin{pmatrix} 0 bar{0} 0 bar{0} end{pmatrix} ]

This implies the need for certain elements to be zero:

Case 1: m 0

In this case, we have:

[ MN begin{pmatrix} 0 overline{ad} 0 overline{bd} end{pmatrix} ]

Here, we can achieve MN 0 by setting d 0 regardless of a and b. This makes the matrices of the form:

[ M begin{pmatrix} 0 a 0 b end{pmatrix} ]

left zero divisors, including when a b 0.

Case 2: m eq 0 and m text{ is even}

For this case, we need n 0. The product simplifies to:

[ MN begin{pmatrix} 0 overline{0ad} 0 overline{bd} end{pmatrix} ]

By setting d 0, we get zero in both the top-right and bottom-right entries. We can also set c 1. Therefore, we have:

[ begin{pmatrix} 2k a 0 b end{pmatrix}begin{pmatrix} 0 1 0 0end{pmatrix} begin{pmatrix} 0 bar{0} 0 bar{0} end{pmatrix} ]

which shows that matrices of the form:

[ M begin{pmatrix} 2k a 0 b end{pmatrix} ]

are left zero divisors, subsuming the previous case when 2k is arbitrary.

Case 3: m text{ is odd}

Here, we need n 0. The product is:

[ MN begin{pmatrix} 0 overline{cad} 0 overline{bd} end{pmatrix} ]

We need bd 0, meaning either b 0 or d 0 or both. Let's explore the subcases:

Subcase 3.0: b 0

In this subcase, MN begin{pmatrix} 0 overline{cad} 0 0 end{pmatrix} ]. We need overline{cad} 0. Choosing a 0 and c 1, d 1 works, and choosing a 1 and c d 1 also works. Therefore, we have:

[ M begin{pmatrix} 2k 1 a 0 0 end{pmatrix} text{ where } a in Z/2Z text{ and } k in mathbb{Z} ]

Subcase 3.1: b 1

In this subcase, c 1 and MN begin{pmatrix} 0 overline{1ad} 0 0 end{pmatrix} ]. Since d 0, it is impossible for overline{1ad} 0. Therefore, this case does not generate any new left zero divisors.

In conclusion, we have two main types of left zero divisors:

[ M begin{pmatrix} 2k a 0 b end{pmatrix} text{ and } M begin{pmatrix} 2k 1 a 0 0 end{pmatrix} ]

where (a, b in Z/2Z), and (k in mathbb{Z}). This completes our exploration of left zero divisors in the ring of upper triangular matrices.