Topics in Algebra solution

Sung Jong Lee, lovekrand.github.io

November 23, 2020

Supplementary Problems.

1. Let R be a commutative ring; an ideal P of R is said to be a prime ideal of R if ab ∈ P ,

a, b ∈ R implies that a ∈ P or b ∈ P . Prove that P is a prime ideal of R if and only if R/P
is an integral domain.

Proof. Note that

[ab ∈ P =⇒ a ∈ P or b ∈ P ]

is equivalent to

[(a + P )(b + P ) = ab + P = P =⇒ a + P = P or b + P = P ] .

Therefore, P is a prime ideal if and only if R/P is an integral domain.

2. Let R be a commutative ring with unit element; prove that every maximal ideal of R is
a prime ideal.

Proof. Let M be a maximal ideal of R. Then R/M is a field, and hence an integral domain.
Therefore, by Problem 1, M is a prime ideal.

3. Give an example of a ring in which some prime ideal is not a maximal ideal.

Solution. The trivial ideal (0) is a prime ideal, but not maximal.

4. If R is a finite commutative ring (i.e., has only a finite number of elements) with unit
element, prove that every prime ideal of R is a maximal ideal of R.

Proof. Let P be a prime ideal of R. Then R/P is an integral domain. Since R is finite,
R/P is also finite. Since every finite integral domain is a field, R/P is a field. Now it
follows that P is maximal.

5. If F is a field, prove that F [x] is isomorphic to F [t].

1
Proof. Let φ : F [x] → F [t] be a mapping defined as

φ(f (x)) = φ(a0 + a1 x + · · · + an xn ) = a0 + a1 t + · · · an tn = f (t).

Clearly it is an onto isomorphism from F [x] to F [t].

6. Find all the automorphisms σ of F [x] with the property that σ(f ) = f for every f ∈ F .

Proof. Suppose σ is an automorphism of F [x] such that σ(f ) = f for every f ∈ F . Then
σ is determined by the image of x. That is, the polynomial σ(x). Since F [σ(x)] ⊂ F [x].
For this mapping to be surjective, σ(x) cannot have degree of larger than 2. So, we are
left with the case σ(x) = ax + b where a 6= 0, b ∈ F . This is surjective, as g(x) = (x − b)/a
will do the inverse map and hence, F [σ(x)] = F [x]. Therefore, σ’s mapping x to ax + b,
a 6= 0, b ∈ F are the automorphisms of F [x].

7. If R is a commutative ring, let N = {x ∈ R : xn = 0 for some integer n}. Prove

a) N is an ideal of R.

Proof. This is exactly the lemma introduced in Problem 7, Section 3.11.

b) In R = R/N if xm = 0 for some m then x = 0.

Proof. Suppose xm = 0 for some m. Equivalently, xm ∈ N . Now by the definition of N ,

(xm )n = 0 for some n. Consequently, (xm )n = xmn = 0 which implies that x ∈ N ⇐⇒
x = 0.

8. Let R be a commutative ring and suppose that A is an ideal of R. Let N (A) = {x ∈

R : xn ∈ A for some integer n}. Prove
a) N (A) is an ideal of R which contains A.

Proof. N (A) clearly contains A. Let x, y ∈ N (A). Suppose m and n are the integers
satisfying xm , y n ∈ A. As A being an ideal of R,
m+n
X 
m+n m + n k m+n−k
(x + y) = x y = (y m+n + xy m+n−1 + · · · + xm−1 y n+1
k
k=0
+ x y + xm+1 y n−1 + · · · + xm+n−1 y + xm+n )
m n

= y m y n + (xy m−1 )y n + · · · + (xm−1 y)y m + xm y n +

+ xm (xy n−1 ) + · · · + xm (xn−1 y) + xm xn ∈ A

so that x + y ∈ N (A). Also, (−x)2m = x2m = xm xm ∈ A so that −x ∈ N (A). Further, for

any r ∈ R, (rx)m = rm xm ∈ A. Thus, N (A) is an ideal of R.

b) N (N (A)) = N (A).

2
Proof. Clearly N (A) ⊂ N (N (A)). Suppose x ∈ N (N (A)). Then xn ∈ N (A) for some n.
Further, (xn )m ∈ A for some m. Since xnm = (xn )m , x ∈ N (A). Therefore, N (N (A)) ⊂
N (A) so that N (N (A)) = N (A).

9. If n is an integer, let Jn be the ring of integers mod n. Describe N for Jn in terms of n.

Proof. Let the prime factorization of n be n = pa11 pa22 · · · pakk . We claim that N (A) =
(p1 p2 · · · pk ). Suppose x ∈ (p1 p2 · · · pk ). Then x = bp1 p2 · · · pk . Let a = max{a1 , a2 , · · · ak }.
Then xa = (bp1 p2 · · · pk )a and since n | (bp1 p2 · · · pk )a , n | xa ⇐⇒ xa = 0 in Jn . Now
conversely, assume that x ∈ N (A). That is, xm = 0 for some integer m. If x = 0, it is
done. If x 6= 0, assume that p1 p2 · · · pk - x for the sake of contradiction. Consequently,
there exists a prime pi such that pi - x. Hence, pi - xm for all positive integer m and hence
n - xm , xm 6= 0. But this is a contradiction. Hence it is must that p1 p2 · · · pk | x and
hence,x ∈ (p1 p2 · · · pk ).

10. If A and B are ideals in a ring R such that A ∩ B = (0), prove that for every a ∈ A,
b ∈ B, ab = 0.

Proof. Note that ab ∈ A ∩ B, as A and B are ideals of R. Therefore, ab = 0.

11. If R is a ring, let Z(R) = {x ∈ R : yx = xy all y ∈ R}. Prove that Z(R) is a subring
of R.

Proof. Choose a, b ∈ Z(R). Then (a + (−b))y = ay + (−b)y = ya + y(−b) = y(a + (−b))

for all y ∈ R. Hence a + b ∈ Z(R). Also, (ab)y = a(by) = a(yb) = (ay)b = (ya)b = y(ab)
so that xy ∈ Z(R). These shows that Z(R) is a subring of R.

12. If R is a division ring, prove that Z(R) is a field.

Proof. It is trivial that Z(R) is commutative. Hence Z(R) is a commutative division ring,
and hence a field.

13. Find a polynomial of degree 3 irreducible over the ring of integers, J3 , mod 3. Use it
to construct a field having 27 elements.

Solution. Let p(x) = x3 − x − 1. It is clearly an irreducible polynomial of degree 3 in J3 .

Consequently, J3 [x]/(p(x)) is a field, with 27 elements.

14. Construct a field having 625 elements.

Solution. Let p(x) = x5 − x − 1. If it had a quadratic factor f (x), then J5 [x]/(f (x)) ' J25
so that

w5 = w + 1, w = w25 = (w + 1)5 = w5 + 1 = w + 2,

3
which is a contradiction. Therefore, p(x) is irreducible in J5 . Now consider the field
J5 [x]/(p(x)). Then it is a field, with 54 = 625 elements.

15. If F is a field and p(x) ∈ F [x], prove that in the ring

F [x]
R= ,
(p(x))

N (Nilradical of R) is (0) if and only if p(x) is not divisible by the square of any polynomial.

Proof. Suppose N = (0). For the sake of contradiction, assume that p(x) is divisible by
some square of a non-costant polynomial t(x). Then t(x)2 d(x) = p(x) for some d(x) ∈ F [x].
Note that t(x)d(x) is not in (p(x)). But since (t(x)d(x))2 ∈ (p(x)), t(x)d(x) ∈ N which
contradicts the fact that N = (0).
Conversely, assume that p(x) is not divisible by the square of any polynomial. With
the fact that F [x] is an UFD, p(x) can be expressed as product of unique irreducible
polynomials(upto associates), which are all distinct. Consider t(x) which is not in (p(x)).
Then t(x) must be missing an irreducible factor of p(x). Consequently, t(x)n cannot contain
that missing factor for any n. Thus, t(x)n 6∈ (p(x)) for all n. Therefore, N = (0).

16. Prove that the polynomials f (x) = 1 + x + x3 + x4 is not irreducible over any field F .

Proof. It it easy to see that f (x) = 1 + x + x3 (x + 1) = (x3 + 1)(x + 1). Therefore, f (x) is
not irreducible over any field F .

17. Prove that the polynomial f (x) = x4 + 2x + 2 is irreducible over the field of rational
numbers.

Proof. Apply Eisenstein’s Criterion. Let ai denote the coefficients of xi . Then 2 - a4 , 2 |

ai , i ≤ 3 but 22 = 4 - a0 = 2. Thus, given f (x) is irreducible over Q.

18. Prove that if F is a finite field, its characteristic must be a prime number p and F
n
contains pn elements for some integer. Prove further that if a ∈ F then ap = a.

Proof. Let m denote the number of elements in F . Then viewing F as a additive group, m·
1 = 0. Hence F must be a field of finite characteristic, with p, a prime as its characteristic.
Suppose m has another prime factor q other than p. Then by Cauchy’s theorem, there is
an element x of order q. Note that (p, q) = 1. Hence, pr + qs = 1 for some integers r and
s. Consequently x(pr + qs) = x ⇐⇒ x = 0, which is a contradiction. Hence, m = pn for
n n
some n. Now viewing F × as a multiplicative group, ap −1 = 1 so that ap = a.

19. Prove that any nonzero ideal in the Gaussian integers J[i] must contain some positive
integers.

4
Proof. Let A be a nonzero ideal of J[i]. Say, a + bi ∈ A, where a and b are not both zero.
Then (a − bi)(a + bi) = a2 + b2 ∈ A so that A contains a positive integer. Hence proved.
20. Prove that if R is a ring in which a4 = a for every a ∈ R then R must be commutative.
Proof. Note that (−x)4 = x = −x so that 2x = 0 for all x ∈ R. So expanding (x + x2 )2 ,
we have
(x + x2 )2 = x4 + 2x3 + x2 = x + x2 .
This shows that elements of the form x + x2 is idempotent. We know that any idempotent
elements are central elements. That is, they lie in Z(R). Let x = a + b. Then
a(x + x2 ) = (x + x2 )a ⇐⇒ a2 b + a(b + b2 ) = ba2 + (b + b2 )a ⇐⇒ a2 b = ba2 . (1)
Since b was arbitrary, elements of the form x2 also lies in Z(R). Since Z(R) being the
subring of R, a = (a + a2 ) − a2 is also in Z(R). Now a was arbitrary, and hence, Z(R) = R.
Therefore, R is commutative.
21. Let R and R0 be rings and φ a mapping from R into R0 satisfying
a) φ(x + y) = φ(x) + φ(y) for every x, y ∈ R.
b) φ(xy) = φ(x)φ(y) or φ(y)φ(x).
Prove that for all a, b ∈ R, φ(ab) = φ(a)φ(b) or that, for all a, b ∈ R, φ(ab) = φ(b)φ(a).
Proof. Let a ∈ R. We define Wa and Ua as follows:
Wa = {x ∈ R : φ(ax) = φ(a)φ(x)}, Ua = {x ∈ R : φ(ax) = φ(x)φ(a)}.
It is easy to see that both Wa and Ua are additive subgroups of R and R = Wa ∪ Ua ,
by the definition of φ. Since no group can be written as union of two subgroup, either
R = Wa or R = Ua . This is equivalent to φ(ab) = φ(a)φ(b) either φ(ab) = φ(b)φ(a), for
every a, b ∈ R.
22. Let R be a ring with a unit element 1, in which (ab)2 = a2 b2 for all a, b ∈ R. Prove
that R must be commutative.
Proof. We compute ((1 + a)b)2 , (a(1 + b))2 and ((1 − a)(1 − b))2 in two ways each. Observe
that
((1 + a)b)2 = (1 + a)2 b2 ⇐⇒ bab = ab2 ,
((a(1 + b))2 = a2 (1 + b)2 ⇐⇒ aba = a2 b,
and
((1 − a)(1 − b))2 = (1 − a)2 (1 − b)2 ⇐⇒ ab − ab2 − a2 b = ba = bab − aba
⇐⇒ ab = ba.
Therefore, R is commutative.

5
23. Give an example of a noncommutative ring (of course, without 1) in which (ab)2 = a2 b2
for all elements a and b.

Proof. Consider the ring R defined as:

  
a b 
R= a, b ∈ Z2
0 0 
   
p q r s
Then for any a = ,b= ,
0 0 0 0
2
(pr)2 p2 rs
  
2 pr ps
(ab) = =
0 0 0 0

where
2  2  2  2
(pr)2 p2 rs
   
2 2 p q r s p pq r rs
a b = = =
0 0 0 0 0 0 0 0 0 0

so that (ab)2 = a2 b2 . But R is not commutative as

         
1 0 1 1 1 1 1 0 1 1 1 0
= 6= = .
0 0 0 0 0 0 0 0 0 0 0 0

24. a) Let R be a ring with unit element 1 such that (ab)2 = (ba)2 for all a, b ∈ R. If in R,
2x = 0 implies x = 0, prove that R must be commutative.

Proof. Similarly with Problem 22, we compute ((1 + a)b)2 , (a(1 + b))2 and ((1 − a)(1 − b))2
in two ways each. Observe that

((1 + a)b)2 = (b(1 + a))2 ⇐⇒ ab2 = b2 a,

((a(1 + b))2 = ((1 + b)a)2 ⇐⇒ a2 b = ba2 ,

and

((1 − a)(1 − b))2 = ((1 − b)(1 − a))2 ⇐⇒ 2ab − a2 b − ab2 = 2ba − b2 a − ba2
⇐⇒ 2(ab − ba) = 0 =⇒ ab = ba.

Therefore, R is commutative.

b) Show that the result of a) may be false if 2x = 0 for some x 6= 0.

6
Proof. Consider the ring R defined as:
  
 a b c  
R = 0 a d  a, b, c, d ∈ Z2
0 0 a 
 
   
p q r x y z
It consists of the unit element I3 . Further, suppose a = 0 p s and b =  0 x w.
0 0 p 0 0 x
Then
2 
(px)2
 
px py + qx pz + qw + rx 0 (py + qx)(pw + sx)
2
(ab) =  0 px pw + sx  =  0 (px)2 0 ,
0 0 px 0 0 (px) 2

2 
(px)2
 
px py + qx pz + sy + rx 0 (py + qx)(pw + sx)
(ba)2 =  0 px pw + sx  =  0 (px)2 0 ,
0 0 px 0 0 (px) 2

so that (ab)2 = (ba)2 . We also see that 2I3 = 0 but I3 6= 0. Moreover, R is not commutative
since
       
0 1 0 0 1 0 0 0 1 0 1 0 0 1 0
0 0 1 0 0 0 = 0 6= 0 0 0 = 0 0 0 0 0 1 .
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

c) Even if 2x = 0 implies x = 0 in R, show that the result of a) may be false if R does not
have a unit element.

Proof. Consider the ring R defined as:

  
 0 a b  
R = 0 0 c   a, b, c ∈ Z3
0 0 0 
 

This ring has no unit element, and 2x = 0 holds only for x = 0. Moreover, power of
every product a and b is zero. That is, (ab)2 = 0 = (ba)2 for all a, b ∈ R. But R is not
commutative since
       
0 1 0 0 0 1 0 0 2 0 0 1 0 1 0
0 0 2 0 0 2 = 0 0 0 6= 0 = 0 0 2 0 0 2 .
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

7
25. Let R be a ring in which xn = 0 implies x = 0. If (ab)2 = a2 b2 for all a, b ∈ R, prove
that R is commutative.

Proof. We shall compute (a(a + b))2 and ((a + b)b)2 in two different ways each. Observe
that

(a(a + b))2 = a2 (a + b)2 ⇐⇒ aba2 = a2 ba,

((a + b)b)2 = (a + b)2 b2 ⇐⇒ b2 ab = bab2 ,

so that

(ab − ba)3 = 0 =⇒ ab = ba.

Therefore, R is commutative.

26. Let R be a ring in which xn = 0 implies x = 0. If (ab)2 = (ba)2 for all a, b ∈ R, prove
that R must be commutative.

Proof. We can get (ab − ba)5 = 0 which leads to ab = ba. Refer ”Commutativity Theorems
Examples in Search of Algorithms”, John J Wavrik, Dept of Math Univ of Calif - San
Diego.

27. Let p1 , p2 , · · · , pk be distinct primes, and let n = p1 p2 · · · pk . If R is the ring of integers

modulo n, show that there are exactly 2k elements a in R such that a2 = a.

Proof. By the Chinese Remainder Theorem,

R = Zn ' Zp1 × Zp2 × · · · × Zpk .

Note that for each Zpi , there are exactly 2 elements in Zpi satisfying a2 = a. Therefore,
there are total of k times of 2, 2k elements in R satisfying a2 = a.

28. Construct a polynomial q(x) 6= 0 with integer coefficients which has no rational roots
but is such that for any prime p we can solve the congruence q(x) ≡ 0 mod p in the integers.

Proof. From the theory of Quadratic residues, x2 ≡ −1 mod p has solution iff p ≡ 1 mod
4. Also, x2 ≡ 2 mod p has solution iff p ≡ 1, 7 mod 8 and x2 ≡ −2 mod p has solution
iff p ≡ 1, 3 mod 8. Therefore, for every prime p, it must have either −1, 2 or,−2 as its
quadratic residue. Thus, q(x) = (x2 + 1)(x2 + 2)(x2 − 2) is a polynomial with integer
coefficients which has no rational roots, but has a root in Zp .