Section 7: Ring theory

Matthew Macauley

Department of Mathematical Sciences

Clemson University
http://www.math.clemson.edu/~macaule/

Math 4120, Modern Algebra

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 1 / 46

What is a ring?
Definition
A ring is an additive (abelian) group R with an additional binary operation
(multiplication), satisfying the distributive law:

x(y + z) = xy + xz and (y + z)x = yx + zx ∀x, y , z ∈ R .

Remarks
There need not be multiplicative inverses.
Multiplication need not be commutative (it may happen that xy 6= yx).

A few more terms

If xy = yx for all x, y ∈ R, then R is commutative.

If R has a multiplicative identity 1 = 1R 6= 0, we say that “R has identity” or

“unity”, or “R is a ring with 1.”

A subring of R is a subset S ⊆ R that is also a ring.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 2 / 46

What is a ring?
Examples
1. Z ⊂ Q ⊂ R ⊂ C are all commutative rings with 1.
2. Zn is a commutative ring with 1.
3. For any ring R with 1, the set Mn (R) of n × n matrices over R is a ring. It has
identity 1Mn (R) = In iff R has 1.
4. For any ring R, the set of functions F = {f : R → R} is a ring by defining

(f + g )(r ) = f (r ) + g (r ), (fg )(r ) = f (r )g (r ) .

5. The set S = 2Z is a subring of Z but it does not have 1.

  
a 0
6. S = : a ∈ R is a subring of R = M2 (R). However, note that
0 0
   
1 0 1 0
1R = , but 1S = .
0 1 0 0

7. If R is a ring and x a variable, then the set

R[x] = {an x n + · · · + a1 x + a0 | ai ∈ R}

is called the polynomial ring over R.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 3 / 46

Another example: the quaternions k
−i j

Recall the (unit) quaternion group: −1 1

2 2 2
Q8 = hi, j, k | i = j = k = −1, ij = ki.
−j i
−k

Allowing addition makes them into a ring H, called the quaternions, or Hamiltonians:

H = {a + bi + cj + dk | a, b, c, d ∈ R} .

The set H is isomorphic to a subring of M4 (R), the real-valued 4 × 4 matrices:

  
 a −b −c −d
 
−d

b a c 
H = c d a −b  : a, b, c, d ∈ R ⊆ M4 (R) .

 d −c 
b a


Formally, we have an embedding φ : H ,→ M4 (R) where

" # " # " #
0 −1 0 0 0 0 −1 0 0 0 0 −1
1 0 0 0 0 0 0 1 0 0 −1 0
φ(i) = 0 0 0 −1 , φ(j) = 1 0 0 0 , φ(k) = 0 1 0 0 .
0 0 1 0 0 −1 0 0 1 0 0 0

We say that H is represented by a set of matrices.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 4 / 46
Units and zero divisors
Definition
Let R be a ring with 1. A unit is any x ∈ R that has a multiplicative inverse. Let
U(R) be the set (a multiplicative group) of units of R.

An element x ∈ R is a left zero divisor if xy = 0 for some y 6= 0. (Right zero divisors

are defined analogously.)

Examples
1. Let R = Z. The units are U(R) = {−1, 1}. There are no (nonzero) zero divisors.
2. Let R = Z10 . Then 7 is a unit (and 7−1 = 3) because 7 · 3 = 1. However, 2 is
not a unit.
3. Let R = Zn . A nonzero k ∈ Zn is a unit if gcd(n, k) = 1, and a zero divisor if
gcd(n, k) ≥ 2.
4. The ring R = M2 (R) has zero divisors, such as:
    
1 −2 6 2 0 0
=
−2 4 3 1 0 0

The groups of units of M2 (R) are the invertible matrices.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 5 / 46

Group rings
Let R be a commutative ring (usually, Z, R, or C) and G a finite (multiplicative)
group. We can define the group ring RG as
RG := {a1 g1 + · · · + an gn | ai ∈ R, gi ∈ G } ,
where multiplication is defined in the “obvious” way.
For example, let R = Z and G = D4 = hr , f | r 4 = f 2 = rfrf = 1i, and consider the
elements x = r + r 2 − 3f and y = −5r 2 + rf in ZD4 . Their sum is
x + y = r − 4r 2 − 3f + rf ,
and their product is
xy = (r + r 2 − 3f )(−5r 2 + rf ) = r (−5r 2 + rf ) + r 2 (−5r 2 + rf ) − 3f (−5r 2 + rf )
= −5r 3 + r 2 f − 5r 4 + r 3 f + 15fr 2 − 3frf = −5 − 8r 3 + 16r 2 f + r 3 f .

Remarks
The (real) Hamiltonians H is not the same ring as RQ8 .
If g ∈ G has finite order |g | = k > 1, then RG always has zero divisors:

(1 − g )(1 + g + · · · + g k−1 ) = 1 − g k = 1 − 1 = 0.

RG contains a subring isomorphic to R, and the group of units U(RG ) contains

a subgroup isomorphic to G .
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 6 / 46
Types of rings
Definition
If all nonzero elements of R have a multiplicative inverse, then R is a division ring.
(Think: “field without commutativity”.)

An integral domain is a commutative ring with 1 and with no (nonzero) zero divisors.
(Think: “field without inverses”.)

A field is just a commutative division ring. Moreover:

fields ( division rings

fields ( integral domains ( all rings

Examples
Rings that are not integral domains: Zn (composite n), 2Z, Mn (R), Z × Z, H.
Integral domains that are not fields (or even division rings): Z, Z[x], R[x], R[[x]]
(formal power series).
Division ring but not a field: H.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 7 / 46

Cancellation
When doing basic algebra, we often take for granted basic properties such as
cancellation: ax = ay =⇒ x = y . However, this need not hold in all rings!

Examples where cancellation fails

In Z6 , note that 2 = 2 · 1 = 2 · 4, but 1 6= 4.

       
1 0 0 1 4 1 0 1 1 2
In M2 (R), note that = = .
0 0 0 0 1 0 0 0 1 0

However, everything works fine as long as there aren’t any (nonzero) zero divisors.

Proposition
Let R be an integral domain and a 6= 0. If ax = ay for some x, y ∈ R, then x = y .

Proof
If ax = ay , then ax − ay = a(x − y ) = 0.
Since a 6= 0 and R has no (nonzero) zero divisors, then x − y = 0. 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 8 / 46

Finite integral domains

Lemma (HW)
If R is an integral domain and 0 6= a ∈ R and k ∈ N, then ak 6= 0. 

Theorem
Every finite integral domain is a field.

Proof
Suppose R is a finite integral domain and 0 6= a ∈ R. It suffices to show that a has a
multiplicative inverse.
Consider the infinite sequence a, a2 , a3 , a4 , . . . , which must repeat.
Find i > j with ai = aj , which means that

0 = ai − aj = aj (ai−j − 1).

Since R is an integral domain and aj 6= 0, then ai−j = 1.

Thus, a · ai−j−1 = 1. 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 9 / 46

Ideals
In the theory of groups, we can quotient out by a subgroup if and only if it is a
normal subgroup. The analogue of this for rings are (two-sided) ideals.

Definition
A subring I ⊆ R is a left ideal if

rx ∈ I for all r ∈ R and x ∈ I .

Right ideals, and two-sided ideals are defined similarly.

If R is commutative, then all left (or right) ideals are two-sided.

We use the term ideal and two-sided ideal synonymously, and write I E R.

Examples
nZ E Z.
  
a 0
If R = M2 (R), then I = : a, c ∈ R is a left, but not a right ideal of R.
c 0
The set Symn (R) of symmetric n × n matrices is a subring of Mn (R), but not an
ideal.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 10 / 46

Ideals
Remark
If an ideal I of R contains 1, then I = R.

Proof
Suppose 1 ∈ I , and take an arbitrary r ∈ R.

Then r 1 ∈ I , and so r 1 = r ∈ I . Therefore, I = R. 

It is not hard to modify the above result to show that if I contains any unit, then
I = R. (HW)

Let’s compare the concept of a normal subgroup to that of an ideal:

normal subgroups are characterized by being invariant under conjugation:

H ≤ G is normal iff ghg −1 ∈ H for all g ∈ G , h ∈ H.

(left) ideals of rings are characterized by being invariant under (left)

multiplication:
I ⊆ R is a (left) ideal iff ri ∈ I for all r ∈ R, i ∈ I .
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 11 / 46
Ideals generated by sets
Definition
The left ideal generated by a set X ⊂ R is defined as:
\ 
(X ) := I : I is a left ideal s.t. X ⊆ I ⊆ R .

This is the smallest left ideal containing X .

There are analogous definitions by replacing “left” with “right” or “two-sided”.

Recall the two ways to define the subgroup hX i generated by a subset X ⊆ G :

“Bottom up”: As the set of all finite products of elements in X ;

“Top down”: As the intersection of all subgroups containing X .

Proposition (HW)
Let R be a ring with unity. The (left, right, two-sided) ideal generated by X ⊆ R is:
Left: {r1 x1 + · · · + rn xn : n ∈ N, ri ∈ R, xi ∈ X },
Right: {x1 r1 + · · · + xn rn : n ∈ N, ri ∈ R, xi ∈ X },
Two-sided: {r1 x1 s1 + · · · + rn xn sn : n ∈ N, ri , si ∈ R, xi ∈ X }.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 12 / 46

Ideals and quotients
Since an ideal I of R is an additive subgroup (and hence normal), then:
R/I = {x + I | x ∈ R} is the set of cosets of I in R;
R/I is a quotient group; with the binary operation (addition) defined as

(x + I ) + (y + I ) := x + y + I .

It turns out that if I is also a two-sided ideal, then we can make R/I into a ring.

Proposition
If I ⊆ R is a (two-sided) ideal, then R/I is a ring (called a quotient ring), where
multiplication is defined by
(x + I )(y + I ) := xy + I .

Proof
We need to show this is well-defined. Suppose x + I = r + I and y + I = s + I . This
means that x − r ∈ I and y − s ∈ I .
It suffices to show that xy + I = rs + I , or equivalently, xy − rs ∈ I :

xy − rs = xy − ry + ry − rs = (x − r )y + r (y − s) ∈ I .

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 13 / 46

Finite fields
We’ve already seen that Zp is a field if p is prime, and that finite integral domains
are fields. But what do these “other” finite fields look like?

Let R = Z2 [x] be the polynomial ring over the field Z2 . (Note: we can ignore all
negative signs.)

The polynomial f (x) = x 2 + x + 1 is irreducible over Z2 because it does not have a

root. (Note that f (0) = f (1) = 1 6= 0.)

Consider the ideal I = (x 2 + x + 1), the set of multiples of x 2 + x + 1.

In the quotient ring R/I , we have the relation x 2 + x + 1 = 0, or equivalently,

x 2 = −x − 1 = x + 1.

The quotient has only 4 elements:

0+I , 1+I , x +I , (x + 1) + I .

As with the quotient group (or ring) Z/nZ, we usually drop the “I ”, and just write

R/I = Z2 [x]/(x 2 + x + 1) ∼
= {0, 1, x, x + 1} .

It is easy to check that this is a field!

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 14 / 46
Finite fields
Here is a Cayley diagram, and the operation tables for R/I = Z2 [x]/(x 2 + x + 1):

+ 0 1 x x +1
0 1
0 0 1 x x +1 × 1 x x +1

1 1 0 x +1 x 1 1 x x +1

x x x +1 0 1 x x x +1 1
x x +1
x +1 x +1 x 1 0 x +1 x +1 1 x

Theorem
There exists a finite field Fq of order q, which is unique up to isomorphism, iff q = p n
for some prime p. If n > 1, then this field is isomorphic to the quotient ring

Zp [x]/(f ) ,

where f is any irreducible polynomial of degree n.

Much of the error correcting techniques in coding theory are built using mathematics
over F28 = F256 . This is what allows your CD to play despite scratches.
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 15 / 46
Motivation (spoilers!)
Many of the big ideas from group homomorphisms carry over to ring homomorphisms.

Group theory
The quotient group G /N exists iff N is a normal subgroup.
A homomorphism is a structure-preserving map: f (x ∗ y ) = f (x) ∗ f (y ).
The kernel of a homomorphism is a normal subgroup: Ker φ E G .
For every normal subgroup N E G , there is a natural quotient homomorphism
φ : G → G /N, φ(g ) = gN.
There are four standard isomorphism theorems for groups.

Ring theory
The quotient ring R/I exists iff I is a two-sided ideal.
A homomorphism is a structure-preserving map: f (x + y ) = f (x) + f (y ) and
f (xy ) = f (x)f (y ).
The kernel of a homomorphism is a two-sided ideal: Ker φ E R.
For every two-sided ideal I E R, there is a natural quotient homomorphism
φ : R → R/I , φ(r ) = r + I .
There are four standard isomorphism theorems for rings.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 16 / 46

Ring homomorphisms
Definition
A ring homomorphism is a function f : R → S satisfying

f (x + y ) = f (x) + f (y ) and f (xy ) = f (x)f (y ) for all x, y ∈ R.

A ring isomorphism is a homomorphism that is bijective.

The kernel f : R → S is the set Ker f := {x ∈ R : f (x) = 0}.

Examples
1. The function φ : Z → Zn that sends k 7→ k (mod n) is a ring homomorphism
with Ker(φ) = nZ.
2. For a fixed real number α ∈ R, the “evaluation function”

φ : R[x] −→ R , φ : p(x) 7−→ p(α)

is a homomorphism. The kernel consists of all polynomials that have α as a root.

3. The following is a homomorphism, for the ideal I = (x 2 + x + 1) in Z2 [x]:

φ : Z2 [x] −→ Z2 [x]/I , f (x) 7−→ f (x) + I .

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 17 / 46

The isomorphism theorems for rings
Fundamental homomorphism theorem
If φ : R → S is a ring homomorphism, then Ker φ is an ideal and Im(φ) ∼
= R/ Ker(φ).

R φ
Im φ ≤ S
(I = Ker φ) any homomorphism

q g
quotient remaining isomorphism
process (“relabeling”)

R Ker φ
quotient
ring

Proof (HW)
The statement holds for the underlying additive group R. Thus, it remains to show
that Ker φ is a (two-sided) ideal, and the following map is a ring homomorphism:

g : R/I −→ Im φ , g (x + I ) = φ(x) .

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 18 / 46

The second isomorphism theorem for rings
Suppose S is a subring and I an ideal of R. Then R
(i) The sum S + I = {s + i | s ∈ S, i ∈ I } is a subring of R
and the intersection S ∩ I is an ideal of S. S + IA
| AA
(ii) The following quotient rings are isomorphic: ||| A
S B I
BBB }}
(S + I )/I ∼
= S/(S ∩ I ) . }}
S ∩I

Proof (sketch)
S + I is an additive subgroup, and it’s closed under multiplication because

s1 , s2 ∈ S, i1 , i2 ∈ I =⇒ (s1 + i1 )(s2 + i2 ) = s1 s2 + s1 i2 + i1 s2 + i1 i2 ∈ S + I .
|{z} | {z }
∈S ∈I

Showing S ∩ I is an ideal of S is straightforward (homework exercise).

We already know that (S + I )/I ∼
= S/(S ∩ I ) as additive groups.
One explicit isomorphism is φ : s + (S ∩ I ) 7→ s + I . It is easy to check that φ : 1 7→ 1
and φ preserves products. 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 19 / 46

The third isomorphism theorem for rings

Freshman theorem
Suppose R is a ring with ideals J ⊆ I . Then I /J is an ideal of R/J and

(R/J)/(I /J) ∼
= R/I .

(Thanks to Zach Teitler of Boise State for the concept and graphic!)

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 20 / 46

The fourth isomorphism theorem for rings

Correspondence theorem
Let I be an ideal of R. There is a bijective correspondence between subrings (&
ideals) of R/I and subrings (& ideals) of R that contain I . In particular, every ideal
of R/I has the form J/I , for some ideal J satisfying I ⊆ J ⊆ R.

R R/I

I1 S1 I3 I1 /I S1/I I3 /I

I2 S2 S3 I4 I2 /I S2/I S3/I I4 /I

I 0

subrings & ideals that contain I subrings & ideals of R/I

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 21 / 46

Maximal ideals

Definition
An ideal I of R is maximal if I 6= R and if I ⊆ J ⊆ R holds for some ideal J, then
J = I or J = R.

A ring R is simple if its only (two-sided) ideals are 0 and R.

Examples
1. If n 6= 0, then the ideal M = (n) of R = Z is maximal if and only if n is prime.

2. Let R = Q[x] be the set of all polynomials over Q. The ideal M = (x)
consisting of all polynomials with constant term zero is a maximal ideal.

Elements in the quotient ring Q[x]/(x) have the form f (x) + M = a0 + M.

3. Let R = Z2 [x], the polynomials over Z2 . The ideal M = (x 2 + x + 1) is

maximal, and R/M ∼ = F4 , the (unique) finite field of order 4.

In all three examples above, the quotient R/M is a field.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 22 / 46

Maximal ideals
Theorem
Let R be a commutative ring with 1. The following are equivalent for an ideal I ⊆ R.
(i) I is a maximal ideal;
(ii) R/I is simple;
(iii) R/I is a field.

Proof
The equivalence (i)⇔(ii) is immediate from the Correspondence Theorem.

For (ii)⇔(iii), we’ll show that an arbitrary ring R is simple iff R is a field.

“⇒”: Assume R is simple. Then (a) = R for any nonzero a ∈ R.

Thus, 1 ∈ (a), so 1 = ba for some b ∈ R, so a ∈ U(R) and R is a field. X

“⇐”: Let I ⊆ R be a nonzero ideal of a field R. Take any nonzero a ∈ I .

Then a−1 a ∈ I , and so 1 ∈ I , which means I = R. X 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 23 / 46

Prime ideals
Definition
Let R be a commutative ring. An ideal P ⊂ R is prime if ab ∈ P implies either a ∈ P
or b ∈ P.

Note that p ∈ N is a prime number iff p = ab implies either a = p or b = p.

Examples
1. The ideal (n) of Z is a prime ideal iff n is a prime number (possibly n = 0).
2. In the polynomial ring Z[x], the ideal I = (2, x) is a prime ideal. It consists of all
polynomials whose constant coefficient is even.

Theorem
An ideal P ⊆ R is prime iff R/P is an integral domain.

The proof is straightforward (HW). Since fields are integral domains, the following is
immediate:
Corollary
In a commutative ring, every maximal ideal is prime.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 24 / 46

Divisibility and factorization
A ring is in some sense, a generalization of the familiar number systems like Z, R,
and C, where we are allowed to add, subtract, and multiply.

Two key properties about these structures are:

multiplication is commutative,
there are no (nonzero) zero divisors.

Blanket assumption
Throughout this lecture, unless explicitly mentioned otherwise, R is assumed to be an
integral domain, and we will define R ∗ := R \ {0}.

The integers have several basic properties that we usually take for granted:
every nonzero number can be factored uniquely into primes;
any two numbers have a unique greatest common divisor and least common
multiple;
there is a Euclidean algorithm, which can find the gcd of two numbers.

Surprisingly, these need not always hold in integrals domains! We would like to
understand this better.
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 25 / 46
Divisibility

Definition
If a, b ∈ R, say that a divides b, or b is a multiple of a if b = ac for some c ∈ R. We
write a | b.

If a | b and b | a, then a and b are associates, written a ∼ b.

Examples
In Z: n and −n are associates.
In R[x]: f (x) and c · f (x) are associates for any c 6= 0.
The only associate of 0 is itself.
The associates of 1 are the units of R.

Proposition (HW)
Two elements a, b ∈ R are associates if and only if a = bu for some unit u ∈ U(R).

This defines an equivalence relation on R, and partitions R into equivalence classes.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 26 / 46

Irreducibles and primes
Note that units divide everything: if b ∈ R and u ∈ U(R), then u | b.

Definition
If b ∈ R is not a unit, and the only divisors of b are units and associates of b, then b
is irreducible.

An element p ∈ R is prime if p is not a unit, and p | ab implies p | a or p | b.

Proposition
If 0 6= p ∈ R is prime, then p is irreducible.

Proof
Suppose p is prime but not irreducible. Then p = ab with a, b 6∈ U(R).

Then (wlog) p | a, so a = pc for some c ∈ R. Now,

p = ab = (pc)b = p(cb) .

This means that cb = 1, and thus b ∈ U(R), a contradiction. 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 27 / 46

Irreducibles and primes

Caveat: Irreducible 6⇒ prime

√
Consider the ring R−5 := {a + b −5 : a, b ∈ Z}.
√ √
3 | (2 + −5)(2 − −5) = 9 = 3 · 3 ,
√ √
but 3 - 2 + −5 and 3 - 2 − −5.
Thus, 3 is irreducible in R−5 but not prime.

When irreducibles fail to be prime, we can lose nice properties like unique
factorization.

Things can get really bad: not even the lengths of factorizations into irreducibles
need be the same!

For example, consider the ring R = Z[x 2 , x 3 ]. Then

x 6 = x 2 · x 2 · x 2 = x 3 · x 3.

The element x 2 ∈ R is not prime because x 2 | x 3 · x 3 yet x 2 - x 3 in R (note: x 6∈ R).

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 28 / 46

Principal ideal domains
Fortunately, there is a type of ring where such “bad things” don’t happen.

Definition
An ideal I generated by a single element a ∈ R is called a principal ideal. We denote
this by I = (a).

If every ideal of R is principal, then R is a principal ideal domain (PID).

Examples
The following are all PIDs (stated without proof):
The ring of integers, Z.
Any field F .
The polynomial ring F [x] over a field.

As we will see shortly, PIDs are “nice” rings. Here are some properties they enjoy:
pairs of elements have a “greatest common divisor” & “least common multiple”;
irreducible ⇒ prime;
Every element factors uniquely into primes.
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 29 / 46
Greatest common divisors & least common multiples
Proposition
If I ⊆ Z is an ideal, and a ∈ I is its smallest positive element, then I = (a).

Proof
Pick any positive b ∈ I . Write b = aq + r , for q, r ∈ Z and 0 ≤ r < a.

Then r = b − aq ∈ I , so r = 0. Therefore, b = qa ∈ (a). 

Definition
A common divisor of a, b ∈ R is an element d ∈ R such that d | a and d | b.

Moreover, d is a greatest common divisor (GCD) if c | d for all other common

divisors c of a and b.

A common multiple of a, b ∈ R is an element m ∈ R such that a | m and b | m.

Moreover, m is a least common multiple (LCM) if m | n for all other common

multiples n of a and b.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 30 / 46

Nice properties of PIDs

Proposition
If R is a PID, then any a, b ∈ R ∗ have a GCD, d = gcd(a, b).
It is unique up to associates, and can be written as d = xa + yb for some x, y ∈ R.

Proof
Existence. The ideal generated by a and b is

I = (a, b) = {ua + vb : u, v ∈ R} .

Since R is a PID, we can write I = (d) for some d ∈ I , and so d = xa + yb.

Since a, b ∈ (d), both d | a and d | b hold.

If c is a divisor of a & b, then c | xa + yb = d, so d is a GCD for a and b. X

Uniqueness. If d 0 is another GCD, then d | d 0 and d 0 | d, so d ∼ d 0 . X 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 31 / 46

Nice properties of PIDs
Corollary
If R is a PID, then every irreducible element is prime.

Proof
Let p ∈ R be irreducible and suppose p | ab for some a, b ∈ R.

If p - a, then gcd(p, a) = 1, so we may write 1 = xa + yp for some x, y ∈ R. Thus

b = (xa + yp)b = x(ab) + (yb)p .

Since p | x(ab) and p | (yb)p, then p | x(ab) + (yb)p = b. 

Not surprisingly, least common multiples also have a nice characterization in PIDs.

Proposition (HW)
If R is a PID, then any a, b ∈ R ∗ have an LCM, m = lcm(a, b).
It is unique up to associates, and can be characterized as a generator of the ideal
I := (a) ∩ (b).

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 32 / 46

Unique factorization domains
Definition
An integral domain is a unique factorization domain (UFD) if:
(i) Every nonzero element is a product of irreducible elements;
(ii) Every irreducible element is prime.

Examples
1. Z is a UFD: Every integer n ∈ Z can be uniquely factored as a product of
irreducibles (primes):
d
n = p1d1 p2d2 · · · pk k .
This is the fundamental theorem of arithmetic.
2. The ring Z[x] is a UFD, because every polynomial can be factored into
irreducibles. But it is not a PID because the following ideal is not principal:

(2, x) = {f (x) : the constant term is even}.

√ √
3. The ring R−5 is not a UFD because 9 = 3 · 3 = (2 + −5)(2 − −5).
4. We’ve shown that (ii) holds for PIDs. Next, we will see that (i) holds as well.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 33 / 46

Unique factorization domains
Theorem
If R is a PID, then R is a UFD.

Proof
We need to show Condition (i) holds: every element is a product of irreducibles. A
ring is Noetherian if every ascending chain of ideals

I1 ⊆ I2 ⊆ I3 ⊆ · · ·

stabilizes, meaning that Ik = Ik+1 = Ik+2 = · · · holds for some k.

Suppose R is a PID. It is not hard to show that R is Noetherian (HW). Define

X = {a ∈ R ∗ \ U(R) : a can’t be written as a product of irreducibles}.

If X 6= ∅, then pick a1 ∈ X . Factor this as a1 = a2 b, where a2 ∈ X and b 6∈ U(R).

Then (a1 ) ( (a2 ) ( R, and repeat this process. We get an ascending chain

(a1 ) ( (a2 ) ( (a3 ) ( · · ·

that does not stabilize. This is impossible in a PID, so X = ∅. 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 34 / 46

Summary of ring types
all rings
RG Mn (R)
commutative rings H
Z×Z Z6
integral domains
Z[x 2, x 3 ] R−5 2Z
UFDs
F [x, y ] Z[x]

PIDs
F [x] Z

Zp fields C
Q
Z2 [x]/(x 2 +x +1)
A F256
R
√ √
R( −π) Q( m)
√
3
Q( 2, ζ)

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 35 / 46

The Euclidean algorithm
Around 300 B.C., Euclid wrote his famous book, the
Elements, in which he described what is now known as
the Euclidean algorithm:

Proposition VII.2 (Euclid’s Elements)

Given two numbers not prime to one another, to find their greatest common measure.

The algorithm works due to two key observations:

If a | b, then gcd(a, b) = a;
If a = bq + r , then gcd(a, b) = gcd(b, r ).

This is best seen by an example: Let a = 654 and b = 360.

654 = 360 · 1 + 294 gcd(654, 360) = gcd(360, 294)

360 = 294 · 1 + 66 gcd(360, 294) = gcd(294, 66)
294 = 66 · 4 + 30 gcd(294, 66) = gcd(66, 30)
66 = 30 · 2 + 6 gcd(66, 30) = gcd(30, 6)
30 = 6 · 5 gcd(30, 6) = 6.

We conclude that gcd(654, 360) = 6.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 36 / 46

Euclidean domains
Loosely speaking, a Euclidean domain is any ring for which the Euclidean algorithm
still works.

Definition
An integral domain R is Euclidean if it has a degree function d : R ∗ → Z satisfying:
(i) non-negativity: d(r ) ≥ 0 ∀r ∈ R ∗ .
(ii) monotonicity: d(a) ≤ d(ab) for all a, b ∈ R ∗ .
(iii) division-with-remainder property: For all a, b ∈ R, b 6= 0, there are q, r ∈ R
such that
a = bq + r with r = 0 or d(r ) < d(b) .

Note that Property (ii) could be restated to say: If a | b, then d(a) ≤ d(b);

Examples
R = Z is Euclidean. Define d(r ) = |r |.
R = F [x] is Euclidean if F is a field. Define d(f (x)) = deg f (x).
√
The Gaussian integers R−1 = Z[ −1] = {a + bi : a, b ∈ Z} is Euclidean with
degree function d(a + bi) = a2 + b 2 .

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 37 / 46

Euclidean domains
Proposition
If R is Euclidean, then U(R) = {x ∈ R ∗ : d(x) = d(1)}.

Proof
⊆”: First, we’ll show that associates have the same degree. Take a ∼ b in R ∗ :

a|b =⇒ d(a) ≤ d(b)

=⇒ d(a) = d(b).
b|a =⇒ d(b) ≤ d(a)

If u ∈ U(R), then u ∼ 1, and so d(u) = d(1). X

“⊇”: Suppose x ∈ R ∗ and d(x) = d(1).

Then 1 = qx + r for some q ∈ R with either r = 0 or d(r ) < d(x) = d(1).

If r 6= 0, then d(1) ≤ d(r ) since 1 | r .

Thus, r = 0, and so qx = 1, hence x ∈ U(R). X 

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 38 / 46

Euclidean domains
Proposition
If R is Euclidean, then R is a PID.

Proof
Let I 6= 0 be an ideal and pick some b ∈ I with d(b) minimal.

Pick a ∈ I , and write a = bq + r with either r = 0, or d(r ) < d(b).

This latter case is impossible: r = a − bq ∈ I , and by minimality, d(b) ≤ d(r ).

Therefore, r = 0, which means a = bq ∈ (b). Since a was arbitrary, I = (b). 

Exercises.
√
(i) The ideal I = (3, 2 + −5) is not principal in R−5 .
(ii) If R is an integral domain, then I = (x, y ) is not principal in R[x, y ].

Corollary
The rings R−5 (not a PID or UFD) and R[x, y ] (not a PID) are not Euclidean.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 39 / 46

Algebraic integers
The algebraic integers are the roots of monic polynomials in Z[x]. This is a subring
of the algebraic numbers (roots of all polynomials in Z[x]).

Assume m ∈ Z is square-free with m 6= 0, 1. Recall the quadratic field

√  √
Q( m) = p + q m | p, q ∈ Q .

Definition
√
The ring Rm is the set of algebraic integers in Q( m), i.e., the subring consisting of
those numbers that are roots of monic quadratic polynomials x 2 + cx + d ∈ Z[x].

Facts
Rm is an integral domain with 1.
Since m is square-free, m 6≡ 0 (mod 4). For the other three cases:
 √  √
 Z[ m] = a + b m : a, b ∈ Z m ≡ 2 or 3 (mod 4)
Rm = √ √
  1+ m  
= a + b 1+2 m ) : a, b ∈ Z

Z 2 m ≡ 1 (mod 4)
R−1 is the Gaussian integers, which is a PID. (easy)
R−19 is a PID. (hard)

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 40 / 46

Algebraic integers

Definition
√ √
For x = r + s m ∈ Q( m), define the norm of x to be
√ √
N(x) = (r + s m)(r − s m) = r 2 − ms 2 .

Rm is norm-Euclidean if it is a Euclidean domain with d(x) = |N(x)|.

Note that the norm is multiplicative: N(xy ) = N(x)N(y ).

Exercises
Assume m ∈ Z is square-free, with m 6= 0, 1.
u ∈ U(Rm ) iff |N(u)| = 1.
If m ≥ 2, then U(Rm ) is infinite.
√
± 1, ± 1± 2 −3 .

U(R−1 ) = {±1, ±i} and U(R−3 ) =
If m = −2 or m < −3, then U(Rm ) = {±1}.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 41 / 46

Euclidean domains and algebraic integers
Theorem
Rm is norm-Euclidean iff
m ∈ {−11, −7, −3, −2, −1, 2, 3, 5, 6, 7, 11, 13, 17, 19, 21, 29, 33, 37, 41, 57, 73} .

Theorem (D.A. Clark, 1994)

The ring R69 is a Euclidean domain that is not norm-Euclidean.

√
Let α = (1 + 69)/2 and c > 25 be an integer. Then the following degree function
works for R69 , defined on the prime elements:
(
|N(p)| if p 6= 10 + 3α
d(p) =
c if p = 10 + 3α

Theorem
If m < 0 and m 6∈ {−11, −7, −3, −2, −1}, then Rm is not Euclidean.

Open problem
Classify which Rm ’s are PIDs, and which are Euclidean.
M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 42 / 46
PIDs that are not Euclidean

Theorem
If m < 0, then Rm is a PID iff
m ∈ {−1, −2, −3, −7, −11, −19, −43, −67, −163} .
| {z }
Euclidean

Recall that Rm is norm-Euclidean iff

m ∈ {−11, −7, −3, −2, −1, 2, 3, 5, 6, 7, 11, 13, 17, 19, 21, 29, 33, 37, 41, 57, 73} .

Corollary
If m < 0, then Rm is a PID that is not Euclidean iff m ∈ {−19, −43, −67, −163}.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 43 / 46

Algebraic integers

Figure: Algebraic numbers in the complex plane. Colors indicate the coefficient of the
leading term: red = 1 (algebraic integer), green = 2, blue = 3, yellow = 4. Large dots mean
fewer terms and smaller coefficients. Image from Wikipedia (made by Stephen J. Brooks).

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 44 / 46

Algebraic integers

Figure: Algebraic integers in the complex plane. Each red dot is the root of a monic
polynomial of degree ≤ 7 with coefficients from {0, ±1, ±2, ±3, ±4, ±5}. From Wikipedia.

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 45 / 46

Summary of ring types
all rings
RG Mn (R)
commutative rings H
Z×Z Z6
integral domains
Z[x 2, x 3 ] R−5 2Z
UFDs
F [x, y ] Z[x]
PIDs
R−43 R−67
R−19 R−163
Euclidean domains
Z F [x]
fields C
Zp
R−1 R69
A Q Fpn
√
R( −π, i) R
√
Q( m)

M. Macauley (Clemson) Section 7: Ring theory Math 4120, Modern algebra 46 / 46