MAT 3343, APPLIED ALGEBRA, FALL 2003

Handout 2: Ideals of Integers

(Supplement to Chapter 1.2)
Peter Selinger

1 Ideals of Integers
Recall that Z = {0, −1, 1, −2, 2, −3, 3, . . .} is the set of integers. If n ∈ Z is any
integer, we write nZ for the set
nZ = {nx | x ∈ Z}.
So for example, 2Z is the set of even numbers, 3Z is the set of multiples of 3, and
0Z is the one-element set {0}. Notice that a ∈ nZ if and only if n divides a. In
particular, we have n ∈ nZ and 0 ∈ nZ, for all n.
Remark 1.1. If nZ = mZ, then n = m or n = −m. To prove this, first notice
that in this situation, n ∈ mZ and m ∈ nZ. Thus m|n and n|m. This implies that
n = m or n = −m.
Notice that, as shown in examples in class, the intersection of two sets nZ and
mZ is again of the form kZ, for some k. For example:
4Z ∩ 6Z = 12Z
4Z ∩ 5Z = 20Z
0Z ∩ 5Z = 0Z.
Also, if A and B are sets of integers, let us write A + B for the set
A + B = {a + b | a ∈ A and b ∈ B}.
So, for example, 4Z + 6Z is the set of all integers of the form 4x + 6y, where
x, y ∈ Z. This happens to cover precisely the even integers, so we find that
4Z + 6Z = 2Z. Again, trying different examples, we find that the sum of two sets
nZ and mZ always seems to be of the form dZ, for some integer d. For example:
4Z + 6Z = 2Z
4Z + 5Z = 1Z
0Z + 5Z = 5Z.
Our goal is to show that this always works. It will be useful to characterize the
sets of the form nZ a little more abstractly.

1
Definition. A subset I ⊆ Z is called an ideal if it satisfies the following three
conditions:

(1) If a, b ∈ I, then a + b ∈ I.
(2) If a ∈ I and k ∈ Z, then ak ∈ I.
(3) 0 ∈ I.

The point is that, as we will show now, the ideals in Z are exactly the subsets of
the form nZ. In other words, the sets of the form nZ are characterized by the
three properties (1)–(3) in the definition. This is proved by the combination of
Lemmas 1.2 and 1.3 below.
Lemma 1.2. Any set of the form nZ is an ideal.

Proof. Let I = nZ. We have to show I satisfies the three properties in the defini-
tion of an ideal.

(1) Take arbitrary elements a, b ∈ I. We have to show that a + b ∈ I. Because

a ∈ nZ, we know that a = nx for some x ∈ Z. Because b ∈ nZ, we know
that b = ny for some y ∈ Z. Then a + b = nx + ny = n(x + y) ∈ nZ = I.
(2) Take arbitrary elements a ∈ I and k ∈ Z. We have to show that ak ∈ I.
But a ∈ nZ, hence we know that a = nx for some x ∈ Z. It follows that
ak = (nx)k = n(xk) ∈ nZ = I.
(3) We have to show that 0 ∈ I. But clearly, 0 = 0n, so it follows that 0 ∈
nZ = I. 
Lemma 1.3. Any ideal I of integers is of the form nZ, for some n ∈ Z.

The idea of the proof is simple: let n be the smallest positive element in I, and
then prove (using the three properties of I) that I = nZ.
The reason the real proof is a bit longer is that we have to worry about a number of
details: for instance, we have to worry about what happens if there are no positive
elements in I (in this case we can’t let n be such an element!).
The proof makes use of the following principle, which says that if there is a posi-
tive integer with a certain property, then there is a smallest such integer.
Principle 1.4 (Well-foundedness principle). If A is a non-empty set of positive
integers, then A has a smallest element.

2
 Proof of Lemma 1.3: We know, from property (3), that 0 ∈ I. In case I = {0},
we are done, because I = 0Z is of the desired form. Otherwise, there must be
some non-zero element k ∈ I. Let A = {x ∈ I | x > 0}. Note that A is non-
empty, because from property (2), −k ∈ I, and hence either k or −k is in A. By
the well-foundedness principle, A has a smallest element. Let n be the smallest
element in A.
Next, we want to show that I = nZ. We do this by first showing than nZ ⊆ I,
then that I ⊆ nZ.
To show that nZ ⊆ I, take an arbitrary element a ∈ nZ. By definition of nZ,
we know that a = nx, for some x ∈ Z. Then from n ∈ I and x ∈ Z, it follows
by property (2) that nx ∈ I, thus a ∈ I. Since a was arbitrary, this shows that
nZ ⊆ I.
To show that I ⊆ nZ, we use the method of contradiction. Thus, assume that
I 6⊆ nZ. Then there exists some k ∈ I such that k 6∈ nZ. Note that k 6∈ nZ
implies k 6= 0. Let B = {x ∈ I | x > 0 and x 6∈ nZ}. From property (2), we
know that −k ∈ I and −k 6∈ nZ, so either k ∈ B or −k ∈ B. Thus, B is
non-empty, and by the well-foundedness principle, it has a least element, say a.
We distinguish three cases:

Case 1: Suppose a > n. Then a − n is positive. Also, from a ∈ I and n ∈ I, it

follows by property (2) that −n ∈ I and by property (1) that a − n ∈ I.
Also, since a 6∈ nZ, it follows that a−n 6∈ nZ. So a−n ∈ B, contradicting
the fact that a is the smallest element of B.
Case 2: Suppose a = n. This contradicts the fact that a 6∈ nZ.
Case 3: Suppose a < n. But a ∈ I, therefore a ∈ A, contradicting the fact that n
was the smallest element in A.

All three cases lead to a contradiction, which implies that our assumption that
I 6⊆ nZ was false. Therefore I ⊆ nZ. Together with nZ ⊆ I, this implies that
I = nZ, which finishes the proof of the lemma. 

Now that we know that an ideal is exactly the same thing as a set of the form nZ,
we want to show that the intersection of two ideals is again an ideal, and similarly
for sums.
Lemma 1.5. (a) If I and J are ideals, then so is I ∩ J.

3
 (b) If I and J are ideals, then so is I + J.

In the following proof, only one of the six cases is given. The remaining cases are
left as homework.

Proof. (a) Suppose I and J are ideals. To show that I ∩ J is an ideal, we must
show that it satisfies the three properties in the definition of an ideal. We
prove each property in turn.

(1) Suppose a, b ∈ I ∩ J. We want to show that a + b ∈ I ∩ J. By

assumption, we know that a ∈ I and a ∈ J and b ∈ I and b ∈ J. By
property (1) of I, we have a + b ∈ I. By property (1) of J, we have
a + b ∈ J. It follows that a + b ∈ I + J.
(2) . . .
(3) . . .
(b) Suppose I and J are ideals. We need to show that I + J is an ideal.
(1) . . .
(2) . . .
(3) . . .

We have now proved what we stated in the beginning: any intersection or sum
of two sets of the form nZ and mZ is again of the form kZ. We summarize this
result in the following theorem:
Theorem 1.6. If n, m ∈ Z, then there exist integers k and d such that

nZ ∩ mZ = kZ,
nZ + mZ = dZ.

Proof. By Lemma 1.2, we know that nZ and mZ are ideals. By Lemma 1.5, we
know that nZ ∩ mZ and nZ + mZ are also ideals. By Lemma 1.3, we know that
they are of the form kZ and dZ, respectively. 

Also note that, by Remark 1.1, the numbers k and d in Theorem 1.6 are essentially
unique: they are determined up to a sign.

4
2 Least common multiple, greatest common divisor
Let us compute some instances of Theorem 1.6. We compute k and d for various
different values of n and m.
4Z ∩ 6Z = 12Z 4Z + 6Z = 2Z
6Z ∩ 6Z = 6Z 6Z + 6Z = 6Z
8Z ∩ 5Z = 40Z 8Z + 5Z = 1Z
9Z ∩ 6Z = 18Z 9Z + 6Z = 3Z
3Z ∩ 5Z = 15Z 3Z + 5Z = 1Z

We observe that the numbers in the first column appear to be greatest common
divisors, and the number in the right column appear to be least common multiples.
Definition. A common divisor of two integers n and m is an integer d such that
d|n and d|m. Further, d is called a greatest common divisor if, whenever e is
another common divisor of n and m, then e|d.
Definition. A common multiple of two integers n and m is an integer k such that
n|k and m|k. Further, k is called a least common multiple if, whenever e is another
common multiple of n and m, then k|e.

The following lemma shows that the numbers d and k in Theorem 1.6 are indeed
a greatest common divisor and a least common multiple.
Lemma 2.1. (a) aZ ⊆ bZ if and only if b|a.
(b) If nZ + mZ = dZ, then d is a greatest common divisor of n and m.
(c) If nZ ∩ mZ = kZ, then k is a least common multiple of n and m.

Proof. (a) Exercise.

(b) Suppose that nZ + mZ = dZ. First, we need to show that d is a common
divisor of n and m. But we have nZ ⊆ dZ and mZ ⊆ dZ. It follows
from part (a) that d|n and d|m, hence d is a common divisor. Next, we
need to show that it is a least common divisor. So suppose that e is another
common divisor, i.e., that e|n and e|m. By part (a), we have nZ ⊆ eZ and
mZ ⊆ eZ. Since eZ is closed under addition, it follows that nZ+mZ ⊆ eZ,
and therefore dZ ⊆ eZ. Finally, by part (a) again, it follows that e|d. Thus,
d is a greatest common divisor.

5
 (c) Suppose that nZ ∩ mZ = kZ. First, we show that k is a common multiple
of n and m. But we have kZ ⊆ nZ and kZ ⊆ mZ. It follows from
part (a) that n|k and m|k, so k is a common multiple. Now, suppose that
e is another common multiple of n and m. then n|e and m|e. By part
(a), we have eZ ⊆ nZ and eZ ⊆ mZ. It follows from set theory that
eZ ⊆ nZ ∩ mZ = kZ, hence, again by part (a), we have k|e. Thus, k is a
least common multiple. 
Corollary 2.2. Any pair of integers n, m have a greatest common divisor and a
least common multiple.

Proof. Theorem 1.6 and Lemma 2.1. 

Greatest common divisors and least common multiples are unique up to a sign.
For instance, if d and d 0 are both greatest common divisors of n and m, then we
must have d|d0 and d0 |d, which implies d = d0 or d = −d0 . When we speak of
the greatest common divisor, we always mean the one that is not negative. It is
also denoted as gcd(n, m). The situation is similar for least common multiples,
and the unique non-negative least common multiple of n and m is often written as
lcm(n, m).
Remark. In our definition of the greatest common divisor d of n and m, we have
not actually required that d is greater than any other common divisor, but only
that it is a multiple of any other common divisor. In this way, we do not have to
make special arrangements in the case where n and/or m are 0. For instance, any
integer is a common divisor of 0 and 0, but gcd(0, 0) = 0. The name “greatest”
common divisor is actually bad terminology, but it is nevertheless standard. A
similar remark applies to least common multiples.
Theorem 2.3. If d = gcd(n, m), then there exists integers a and b such that
d = an + bm.

Proof. This follows directly from the fact that nZ + mZ = dZ. Namely, we have
d ∈ nZ + mZ, and thus, d = an + bm, for some a, b ∈ Z. 

In general, finding a and b such that d = an + bm is not as easy as it looks. For

instance, we have gcd(89, 144) = 1. Therefore, there exist integers a and b such
that 89a + 144b. Try to find them yourself! This can be quite a lot of work. We
will later learn a method for finding a and b efficiently.

6
3 Exercises
Problem 1 Finish the proof of Lemma 1.5. Try to imitate the style used in the
first part.

Problem 2 Prove that n|m and m|p implies n|p.

Problem 3 Prove Lemma 2.1(a). There are two directions to prove: aZ ⊆ bZ ⇒

b|a, and b|a ⇒ aZ ⊆ bZ.

Problem 4 Consider Principle 1.4, the well-foundedness principle. It states that

any non-empty subset of positive integers has a least element. Answer the follow-
ing questions. In each case, if the answer is “no”, give a counterexample (i.e., a
set which does not have a least element).

(a) Is the principle still true if we drop the word “non-empty”?

(b) Is the principle still true if we drop the word “positive”?

(c) Is the principle still true if we replace the word “integer” by “rational num-
ber”?
(d) Is the principle still true if we replace the word “positive” by “non-negative”?
Recall that a number x is positive if x > 0, and non-negative if x > 0.