You are on page 1of 6

Evan Nash

MATH 818
Homework 1 Bin Chen
7.4.30 Let I be an ideal of the commutative ring R and dene
rad I = {r R | r
n
I for some n Z
+
}
called the radical of I. Prove that rad I is an ideal containing I.
Proof. We rst note that if i I is arbitrary, then i
1
I, so indeed I rad I.
Next, we shall show that rad I is a subring of R. Because I rad I, rad I is
non-empty, so we must demonstrate that it is closed under addition, additive
inverses and multiplication.
Let a, b rad I be arbitrary such that a
m
I and b
n
I. We rst seek to
show that (a + b)
k
I for some k Z
+
. I claim that k = m + n satises this.
To see this, we consider the expansion:
(a + b)
m+n
=
m+n

i=0
_
m + n
i
_
a
i
b
m+ni
=
_
n

i=0
_
m + n
m + i
_
a
i
b
ni
_
a
m
+
_
m

j=0
_
m + n
n + j
_
a
mj
b
j
_
b
n
Because a
m
, b
n
I and I is an ideal, the terms
_
n
i=0
_
m+n
m+i
_
a
i
b
ni
_
a
m
and
_

m
j=0
_
m+n
n+j
_
a
mj
b
j
_
b
n
are both in I. As I is closed under addition, we there-
fore have that (a + b)
m+n
I and so rad I is closed under addition.
We must next show that if a rad I then a rad I. If a
m
I, then
(a)
2m
= (1)
2m
a
2m
= a
2m
. Then a
2m
= (a
m
)
2
I as I has closure. Thus,
rad I is closed under additive inverses.
To nish the proof, we shall show that r (rad I) rad I for arbitrary r R.
This will jointly demonstrate that rad I is closed under multiplication and that
rad I is an ideal because R is commutative. We let a rad I be arbitrary with
a
m
I and consider ra. Observe the following algebra:
(ra)
m
= r
m
a
m
r
m
I I
This demonstrates that ra rad I and so rad I is an ideal of R.
1
Evan Nash
MATH 818
Homework 1 Bin Chen
7.4.32 Let I be an ideal of the commutative ring R and dene
Jac I to be the intersection of all maximal ideals of R that contain I
where the convention is that Jac R = R.
a) Prove that Jac I is an ideal of R containing I.
Proof. We let {M
i
} be the set of maximal ideals of R that contain I so that
Jac I =

i
M
i
. We must demonstrate that Jac I is nonempty, closed un-
der subtraction and multiplication and that r(Jac I) Jac I for all r R.
Certainly Jac I is nonempty as it must contain I, so we let a, b Jac I be
arbitrary. We let M be an arbitrary maximal ideal of R that contains I. Then
Jac I M, so a, b M. Because M is an ideal, it is closed under subtraction,
so a b M. As M was arbitrary, this means that a b M
i
for all i. Thus,
a b

i
M
i
= Jac I and so Jac I is closed under subtraction.
We let r R be arbitrary and seek to show that r(Jac I) Jac I. This will
demonstrate that Jac I is closed under multiplication and that Jac R is an
ideal as R is commutative. We let a Jac I and r R be arbitrary. As
a Jac I, we know that a M, so ra M as M is an ideal. Because M was
arbitrary, it must be the case that ra M
i
for all i, so ra Jac I and thus
r(Jac I) Jac I and the claim is proven.
b) Prove that rad I Jac I.
Proof. Let Jac I =

i
M
i
for maximal ideals M
i
R. Then suppose r rad I
is arbitrary so that r
m
I for some m Z
+
. Consider some arbitrary M in
the set of maximal ideals that contain I. If we can show that r M, then
rad I M
i
for all M
i
as M was arbitrary, so rad I Jac I and we will be
done. We shall assume that r / M and show that this delivers a contradiction.
Because r
m
I and I Jac I M, we know r
m
M. By Corollary 14, we
also have that M is a prime ideal of R as R is commutative. Then because
r
m
M, we know that either r M or r
m1
M. But because r / M,
we must have r
m1
M. But then we observe once again that either r M
or r
m2
M and we may follow a similar line of reasoning. After m 2
iterations of this argument, we are left with the case that r
2
M, which
means it is unavoidable that r is in M. This contraction shows that r M, so
rad I Jac I and the claim is proven.
2
Evan Nash
MATH 818
Homework 1 Bin Chen
c) Let n > 1 be an integer. Describe Jac nZ in terms of the prime factorization
of n.
We note that the maximal ideals of Z are precisely those subrings of the form
pZ for p a prime. This is because any subring of Z is of the form nZ and the
only subrings that will contain pZ must be of the form nZ where n divides p.
The only possibilities are pZ and Z itself, so pZ is indeed a maximal ideal.
Thus, if n = p

1
1
p

2
2
p

k
k
for distinct primes p
i
, we set m = p
1
p
2
p
k
. I claim
that mZ is Jac nZ. We note that every maximal ideal of Z that contains nZ
will be of the form p
i
Z. Thus, the intersection of all of these maximal ideals
will be p
1
p
2
p
k
Z = mZ as claimed.
7.4.37 A commutative ring R is called a local ring if it has a unique maximal ideal.
Prove that if R is a local ring with maximal ideal M then every element of
R M is a unit. Prove conversely that if R is a commutative ring with 1 in
which the set of nonunits forms an ideal M, then R is a local ring with unique
maximal ideal M.
Proof. Let R be a local ring and suppose M is the unique maximal ideal of
R. Suppose that there exists some r / M such that r is a nonunit. Then (r),
the ideal generated by r, is not contained in M. But because r is not a unit,
there is no element a R with ar = 1, so 1 / (r). Thus (r) = R, so (r) is a
proper ideal of R. By Proposition 11, (r) must therefore be contained in some
maximal ideal of R as R has 1. But (r) M, so this implies that there is
some maximal ideal of R other than M. This dees our assumption that M is
unique, so every element of R not in M must be a unit, proving the claim.
Conversely, suppose that R is a commutative ring with 1 in which the set
of nonunits forms an ideal M. We note by Proposition 9 that every ideal
containing a unit equals R, so every proper ideal of R must have only nonunits.
Thus, every proper ideal of R is contained in M, so M must be the unique
maximal ideal of R.
3
Evan Nash
MATH 818
Homework 1 Bin Chen
7.5.3 Let F be a eld. Prove that F contains a unique smallest subeld F
0
and that
F
0
is isomorphic to either Q or Z/pZ for some prime p.
Proof. We rst show that F contains a unique smallest subeld. Let {S
i
} be
the set of subelds of F and let F
0
=

i
S
i
. I claim that F
0
is a subeld of
F. As every subeld must contain 1, 1 F
0
so we let a, b F
0
be arbitrary
and seek to show that a b F
0
and ab
1
F
0
. This will demonstrate that
F
0
is a subring with 1 that is closed under multiplicative inverses, making it a
subeld. If S {S
i
} is arbitrary, then a, b S, so a b S and ab
1
S as S
is a subeld. Because S was arbitrary, we have that a b S
i
and ab
1
S
i
for all i and so a b F
0
and ab
1
F
0
, conrming that F
0
is a subeld of F.
As F
0
is contained in every subeld of F, it must also be the unique smallest
subeld of F.
To show the second part of the claim, we consider the characteristic of F. I
claim that this value is either 0 or p for some prime p. It is certainly possible
that the characteristic of F is 0 as the eld Q fullls this property. We therefore
consider the possibility that the characteristic of F is n for some composite
number n so that n1 = 0. Then n = ab and so ab1 = 0, which implies that
a1 b1 = 0. But because a, b < n, we know that a1 = 0 and b1 = 0, which
implies that F contains zerodivisors, which is impossible as F is a eld. Thus,
if the characteristic of F is not zero, it must be prime.
We then observe that 1 is contained in every subeld of F and so because a
subeld must be closed under addition and additive inverses, every subeld
of F must contain this set of all possible sums of 1 and its additive inverse
1. We will notate this set of elements of the form n1 with n Z as (1). In
the case that the characteristic of F is p, the map : (1) Z/pZ dened
by (a1) = a (mod p) can be easily shown to be an isomorphism. If the
characteristic is zero, the map : (1) Z dened by (a1) = a is also an
isomorphism. Thus, if the characteristic of F is p, then (1) is a subeld as Z/pZ
is a eld for all prime p. This eld will be contained in every other subeld
and so it is the smallest subeld of F. If the characteristic is zero, then (1) is
isomorphic to Z, and Corollary 16 says that the smallest eld that contains Z is
its eld of fractions, Q, which will similarly be contained in all other subelds
of F. Thus, the smallest subeld of F is isomorphic to either Q or Z/pZ for
some prime p.
4
Evan Nash
MATH 818
Homework 1 Bin Chen
7.5.6 Prove that the real numbers, R, contain a subring A with 1 A and A maximal
(under inclusion) with respect to the property that
1
2
/ A.
Proof. We shall prove this through use of Zorns Lemma. Let S be the set of
all subrings of R with 1 that do not contain
1
2
. We know that S is non-empty
as Z is a subring of R with 1 that does not contain
1
2
. Then let C = {R
i
} be
an arbitrary chain in S ordered by inclusion. I claim that

i
R
i
is a subring of
R with 1 not containing
1
2
.
To show this, we let a, b

i
R
i
arbitrary so that a R
a
and b R
b
for some
R
a
, R
b
C. Because C is ordered by inclusion, we may assume without loss of
generality that R
b
R
a
so that a, b R
a
. Then because R
a
is a subring of R,
we know that a b R
a
and ab R
a
. This further implies that a b

i
R
i
and ab

i
R
i
, so

i
R
i
is a subring of R. Because 1 R
i
and
1
2
/ R
i
for all i,
we also know that 1

i
R
i
and
1
2
/

i
R
i
. Then

i
R
i
is a maximal subring
of the chain C with 1 and not containing
1
2
, and so Zorns Lemma tells us that
R must contain a maximal subring with 1 that does not contain
1
2
.
8.1.3 Let R be a Euclidean Domain. Let m be the minimum integer in the set of
norms of nonzero elements of R. Prove that every nonzero element of R of
norm m is a unit. Deduce that a nonzero element of norm zero (if such an
element exists) is a unit.
Proof. Let a R be an element of norm m. Because R is a Euclidean Domain,
there exist some q, r R such that 1 = qa +r and r = 0 or N(r) < N(a) = m.
If r is nonzero, then N(r) < m is impossible as m is the minimum integer in
the set of norms of nonzero elements of R. Thus, r must equal zero and so
1 = qa for some q R, which means that a is a unit.
If a R is a nonzero element with norm zero, then its norm must be the
minimum integer in the set of norms of nonzero elements as norms are by
denition non-negative. The proof above therefore shows that a must be a
unit, as claimed.
5
Evan Nash
MATH 818
Homework 1 Bin Chen
8.1.10 Prove that the quotient ring Z[i]/I is nite for any nonzero ideal I of Z[i].
Proof. We rst note that Z[i] is a Euclidean Domain through use of the norm
N(a + bi) = a
2
+ b
2
. Then Proposition 1 tells us that every ideal I of Z[i] is
principal so that I = (d) for some nonzero d I of minimum norm. Then
consider some g Z[i] with N(g) > N(d). If g = dq for some q Z[i], then
g I, so we shall assume that g = dq + r for some nonzero r Z[i] such that
N(r) < N(d). Then g r = dq, so g r I. This implies that g r + I = I,
so g + I = r + I. Thus, every element of Z[i] is either in I or is in the same
coset of I as some element of Z[i] with norm less than that of d. But there
are only a nite number of elements in Z[i] with norm less than d, so there are
only a nite number of cosets of I in Z[i]. Thus, Z[i]/I is nite and the claim
is proven.
6

You might also like