Scribd
Upload
3 views3 pages

A Lecture 06

The document discusses the group structure of Z_m, focusing on the concept of primitive roots and when Z_m is a cyclic group. It establishes definitions, lemmas, and theorems related to primitive roots, particularly for prime numbers and powers of 2, highlighting that Z_2^l is not cyclic for l ≥ 3. The document concludes with remarks on the distinct behavior of powers of 2 compared to other primes in number theory.

Uploaded by

hollis.chung93
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views3 pages

A Lecture 06

The document discusses the group structure of Z_m, focusing on the concept of primitive roots and when Z_m is a cyclic group. It establishes definitions, lemmas, and theorems related to primitive roots, particularly for prime numbers and powers of 2, highlighting that Z_2^l is not cyclic for l ≥ 3. The document concludes with remarks on the distinct behavior of powers of 2 compared to other primes in number theory.

Uploaded by

hollis.chung93
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

3.

Primitive Roots

We study the group structure of Zm for any integer m ¥ 2. In particular, we wish to
know when it is a cyclic group. This leads to the notion of the primitive root.

3.1. The cases of primes and powers of 2. We start with the definition of primitive
roots.
Definition 3.1. Let a, m P Z, m ¥ 2, hcf pa, mq  1. a is said to be a primitive root
modulo m if the group of units Zm is cyclic and the congruence class a is a generator.
Remark 3.2. We make some comments about this definition.

(1) Assume a and m and coprime. The order of a modulo m is defined to be the
order of a in the group of units Zm . For any integer n, an  1 pmod mq iff n is
a multiple of the order of a modulo m. In this terminology, a is a primitive root
modulo m iff a is coprime to m and the order of a modulo m is φpmq.
(2) Knowing that a is a primitive root modulo m allows us to write
Zm  t a k | k P Z, 0 ¤ k φpmq u.
In other words, every integer coprime to m is congruent to ak for some k P Z. This
will be extremely helpful in many different situations. See Exercises 3.2 and 3.3.
(3) If a is a primitive root modulo m, then Zm is cyclic of order φpmq hence has
φpφpmqq generators. More precisely, any primitive root modulo m lies in the
congruence class a k for some k with 0 ¤ k φpmq and hcf pk, φpmqq  1.

We have seen in Remark 2.20 that it is essential to understand Zm when m is a power of
a prime in order to understand the general case. We first consider the situation when m
is a prime. We need the following lemma:
Lemma 3.3. Let f pxq P krxs where k is a field. Suppose that deg f pxq  n. Then f has
at most n distinct roots in k.

Proof. The proof goes by induction on n. For n  0 the assertion is trivial. Assume that
the statement is true for polynomials of degree n  1. If f pxq has no roots in k, we are
done. If α is a root, since k rxs is a Euclidean domain, we can write f pxq  px  αqq pxq r,
where r is a constant. Setting x  α we see that r  0. Thus f pxq  px  αqq pxq and
deg q pxq  n  1. If β  α is another root of f pxq, then 0  f pβ q  pβ  αqq pβ q, which
implies that q pβ q  0. Since by induction q pxq has at most n  1 distinct roots, f pxq has
at most n distinct roots. 

The following theorem is useful in many situations.


30
Theorem 3.4. Let K be a field and K  the group of non-zero elements under multipli-
cation. Suppose G is a finite subgroup of K  , then G is cyclic.

Proof. We prove by strong induction on n  |G|. If n  1 there is nothing to prove. Now


we assume any subgroup of K  with order smaller than n is cyclic.

For any d with d  n and d n, we write Gd  tg P G | g d  1u. We claim Gd is a


subgroup of G. Indeed, 1 P Gd because 1d  1. If g1 , g2 P Gd , then pg1 g2 qd  g1d g2d 
1 because multiplication is commutative in the field K. Therefore Gd is closed under
multiplication. Moreover, if g P Gd , then pg 1 qd  pg d q1  1, hence Gd is closed under
taking inverse. These conclude that Gd is a group, thus a subgroup of G. Each element of
Gd is a solution to xd  1  0 in K, so |Gd | ¤ d by Lemma 3.3. By induction hypothesis
we know Gd is a cyclic group.

Let ψ pdq be the number of elements of order d in G. Each such element is contained in
Gd , so ψ pdq is also the number of elements of order d in Gd . If |Gd | d then ψ pdq  0.
Otherwise Gd is a cyclic group of order d and ψ pdq  φpdq. So we always have ψ pdq ¤ φpdq.
°
On one hand ψ pnq ψ pdq  n since the order of any element of G is a divisor of
dn,d n°
n. On the other hand φpnq dn,d n φpdq  n by Proposition 1.28. Since for each d n
we have ψ pdq ¤ φpdq, we must have ψ pnq ¥ φpnq ¡ 0. In other words, there are elements
of order n in G, hence G is cyclic. 

The following immediate consequence has fundamental importance. It was first proved
by Gauss.

Corollary 3.5. Let p be a prime, then Zp is a cyclic group; i.e. there exist primitive
roots modulo p.

Proof. By Proposition 2.9, Zp is a field. Then the result follows from Theorem 3.4. 

Next we study the case of prime powers. We will show that primitive roots exist for powers
of odd primes, but the situation is completely different for powers of 2. The necessity of
treating 2 differently from the other primes occurs repeatedly in number theory.

Proposition 3.6. Let l be a positive integer. Then Z2l is not cyclic unless l  1 or 2.

Proof. It is easy to see that 1 is a primitive root modulo 2, and 3 is a primitive root
modulo 4. From now on we assume that l ¥ 3. We claim that

a2
l 2
 1 pmod 2l q
for every odd integer a. It means that the order of every element in Z2l is strictly smaller
than φp2l q, hence Z2l cannot be cyclic.
31
We prove this claim by induction on l. When l  3, Z8  t1, 3, 5, 7u. We can check them
one by one and conclude a2  1 pmod 8q for any odd integer a. Now we assume the claim
l 2
holds for l, then we can write a2  1 b  2l , thus

a2
l 1
 p1 b  2l q2 1 b  2l 1
b2  22l .

The last two terms are divisible by 2l 1 , hence a2
l 1
 1 pmod 2l 1q, i.e. the claim holds
for l 1. 

Remark 3.7. For enthusiasts: for any l ¥ 3, we actually have Z2l  Z2  Z2l2 which is
the direct product of two cyclic groups. We do not prove this fact but it is not difficult.

32

You might also like