On The Structure of Principal Ideal Rings: Thomas W. Hungerford

PACIFIC JOURNAL OF MATHEMATICS
Vol. 25, No. 3, 1968
ON THE STRUCTURE OF PRINCIPAL IDEAL RINGS

THOMAS W. HUNGERFORD
In this paper all rings are supposed commutative with

identity (and all ring direct sums are finite). We distinguish
between a principal ideal domain (PID) and a principal ideal
ring (PIR) which may not be an integral domain. Our chief
purpose is to prove:
THEOREM 1. Every principal ideal ring R is a direct sum
of rings, each of which is the homomorphic image of a princi-
pal ideal domain.
A number of special cases of this theorem are well known. In

order to prove the general case one needs several facts about complete
local rings and formal power series rings. For the reader's convenience,
this background material is summarized in § 1; the theorem is proved
in § 2. In the final section is shown a counterexample indicating that
in a sense Theorem 1 is best possible.
1* Preliminaries •
DEFINITION 2. A Noetherian ring R which has a unique maximal

ideal M is a local ring.
The following proposition, a proof of which is in Lambek [5,
p. 34], is extremely useful for characterizing local rings.
PROPOSITION 3. The following conditions on a ring R are equivalent:

( 1 ) R has a unique maximal ideal M;
(2) all nonunits of R are contained in a proper ideal M;
(3) the nonunits of R form an ideal M.
If R is a local ring, then Π Mί = 0; this fact is due to Krull [3,
Th. 2] and is also proved in Nagata [6, p. 12]. Consequently we
obtain
PROPOSITION 4. Let R be a local ring whose maximal ideal M is

principal, M — (α), then R is a principal ideal ring and every nonzero
k
element of R can be written in the form a n, with u a unit.
Proof. Let reR. Since f\Mi = Π (α*) = 0, there is a highest

power of a dividing r, say r = aku. Then u must be a unit, since
k+ι
all nonunits are in M and u e M implies α ( uy hence a \r. If J is
any ideal in i?, all elements of J can be written in the form α%, u a
unit. Choose such an element with i minimal. It is then easy to
54E
544 THOMAS W. HUNGERFORD
{
verify that J = (a ).
A standard procedure with a local ring R (maximal ideal M) is
to topologize it by taking the family {M* | i ^ 1} as a base of neighbor-
hoods of 0 in jR. It can be verified that this topology agrees with
the topology induced by the following metric on R: d(r, r) = 0;
d(r, s) = 2rn if and only if r — s e Mn and r — s & Mn+1. A complete
local ring is a local ring which is complete with respect to this
topology. The structure of complete local rings is well known and
will be described in Theorem 8 below.
If R is any ring, R[[x]] denotes the ring of formal power series
in one indeterminate with coefficients in R. The ring of polynomials
R[x) is considered as a subring of i?[[a;]] in the obvious way. We
shall need several facts about such power series.
PROPOSITION 5. Let R be a ring. Then:

(1) If R is an integral domain, so is !?[[#]];
(2) If R is Noetherian, so is i?f[x]];
(3) An element feR[[x]] is a unit if and only if the constant
term of / if a unit in R.
The proofs of (l)-(3) can be found in Zariski and Samuel [8,

Chapter 7] and in Nagata [6, Chapter 2],
COROLLARY 6. If R is a field, then R[[x]] is a principal ideal

domain.
Proof. All nonzero elements of a field are units; hence (3) im-
plies that all the nonunits of R[[x]] are contained in the ideal (x).
Hence by (1), (2) and Proposition 3 R[[x]] is a local integral domain
whose maximal ideal is principal. Therefore by Proposition 4 iϋ[[a;]]
is a principal ideal domain.
It should also be noted that (3) implies that every element of
R[[x]] whose constant term is irreducible in R is irreducible in
DEFINITION 7. A v-ring is a ring D such that D is a complete

local integral domain of characteristic zero, whose maximal ideal N
is generated by p(ihe p-ΐold of the identity element-i.e. 1 + 1 + +1
(p times)), where p Φ 0 is the characteristic of the residue field D/N.
By Proposition 4, a ^-ring D is a PID. Observe that (up to unit)
p is the only irreducible element of D.
THEOREM 8. Let R be a complete local ring whose maximal ideal

M is principal, M = (a). Then R is the homomorphic image of
D[[x]] under a homomorphism which sends x e D[[x]] onto aeR, where
ON THE STRUCTURE OF PRINCIPAL IDEAL RINGS 545
D is either the field R/M (if char R = char R/M), or a v-ring whose
residue field Dj(p) is naturally isomorphic to R/M (if char R Φ char
R/M).
This is a very specialized version of the structure theorem for

complete local rings due to Cohen [1, Theorem 9 and 12].
2* Proof of Theorem 1*
DEFINITION 9. A principal ideal ring R is special if R has a

unique prime ideal and this ideal is nilpotent.
Since every PIR is Noetherian and every maximal ideal is prime,
a special PIR is a local ring. Furthermore the nilpotency of the
maximal ideal implies that every Cauchy sequence must eventually be
constant and therefore convergent. Hence a special PIR is a complete
local ring.
LEMMA 10. Every principal ideal ring R is the direct sum of

principal ideal domains and special principal ideal rings.
This is just Theorem 33, page 245, of Zariski and Samuel [7].
In view of the lemma it suffices in Theorem 1 to assume that R is a
special PIR and thus a complete local ring with (nipotent) maximal
ideal M = (a). By Theorem 8, R is the homomorphic image of />[[#]],
where D is a field or v-ring. If D is a field #[[#]] is a PID by Corol-
lary 6. If D is a v-ring, D[[#]] is not a PID. Consequently, to com-
plete the proof of Theorem 1 we must show that when D is a v-ring,
R is the homomorphic image of a PID.
Let φ: D[[x\] —* R be a ring epimorphism such that φ(x) — a. Our
method is to factor out of JD[[#]] a principal ideal generated by an
irreducible element in such a way that the resulting quotient is a
PID. We shall choose this irreducible element to be in kernel φ.
Thus φ will induce a homomorphism from the quotient onto R and
the theorem will be proved.
As above we shall write p for the ordinary prime integer as well
as for the p-fold of the identity in R and in D (and therefore also
in D[[#]]). In the present case D is a v-ring and by Theorem 8 we
have: char R Φ char R/M = char D/(p) = p Φ 0. It follows that p e R is
a nonzero nonunit. Therefore by Proposition 4, p can be written in the
form aku, with u a unit in R; k > 0 since p is in the ideal of nonunits
M — (a). Choose c e D[[x]] such that <p(c) = u. Then the element
/ = — cxh + p of £>{[#]] is in the kernel of φ (since φ(x) — a and
φ(Ί>) = P) Furthermore since the constant term of / is p, which is
irreducible in D, / i s irreducible in
546 THOMAS W. HUNGERFORD
It is proved in Krull [4, p. 13] that if D if a v-ring, then

is a unique factorization domain. Therefore the ideal (/) generated
by the irreducible element / is a prime ideal (see, for example, Zariski
and Samuel [7, p. 21]). Thus the quotient S = D[[x]]/(f) is an integral
domain and φ: S —»R is an epimorphism.
To complete the proof we need only show that S is a PID. By
Proposition 4 this will be the case if we can show that S is a local
ring with principal maximal ideal. D[[#]] (and hence S) is Noetherian
by Proposition 5. If h e D[[x]] let h denote its coset in S. Since
φ(x) — a, which is a nonunit in R, x must be a nonunit in S. There-
fore (x) is a proper ideal in S. Also since / = 0 in S and / has
constant term p, pe(x). Now let h be a nonunit in S. Then h is a
nonunit in £>[[#]] and therefore has constant term which is a nonunit
in D. Since (p) is the ideal of all nonunits in D, the constant term
of h is in (p). But then h is a sum of an element in (x) and an
element in (p). Therefore h is the sum of an element in (x) and an
element in (p) g (x). Hence h e (x) and (x) contains all nonunits of S.
Thus by Proposition 3 S is a local integral domain with principal
maximal ideal. By Proposition 4, S is a PID and the proof is complete.
COROLLARY 11. Every special principal ideal ring is the homo-

morphic image of a principal ideal domain.
3* Further comments* In a certain sense, Theorem 1 is best

possible. Namely the following conjecture is false: every principal ideal
ring is the homomorphic image of a principal ideal domain. A counter-
example to this conjecture is given by the direct sum S = RφZ of
the reals and the integers. S is clearly a PIR which is not an integral
domain. Also every ^-module is an S-module by pullback along the
projection R 0 Z —> Z. In particular the additive group of rationale Q
is an S-module. As is well known Q is not a direct sum of cyclic
groups, and hence cannot be a direct sum of cyclic S-modules.
Suppose S were a proper quotient of an integral domain D, say
S — D/(c). Then every S-module is also a D-module which is annihi-
lated by the ideal (c). Thus every S-module is a torsion D-module of
bounded order and therefore by Kaplansky [2, Th. 6 and p. 36] is a
direct sum of cyclic D-modules. Therefore every S-module, (in par-
ticular Q) is also a direct sum of cyclic S-modules, a contradiction.
In closing we note the following easy consequence of Lemma 10.
COROLLARY 12. A principal ideal ring R is a direct sum of

principal ideal domains if and only if it has no nonzero nilpotent
elements.
ON T H E STRUCTURE O F P R I N C I P A L I D E A L RINGS 547
BIBLIOGRAPHY
1. I. S. Cohen, On the structure and ideal theory of complete local rings, Trans. Amer.
Math. Soc. 59 (1946), 54-106.
2. I. Kaplansky, Infinite Abelian Groups, University of Michigan Press, Ann Arbor,
1954.
3. W. Krull, Dimensionstheorie in Stellenringen, J. Reine Angew. Math. 179 (1938),
204-226.
4. , Uber die Zerlegung der Hauptideale in allgemeinen Ringen, Math. Ann.
105 (1931), 1-14.
5. J. Lambek, Lectures on Rings and Modules, Blaisdell, Waltham, Toronto, and
London, 1966.
6. M. Nagata, Local Rings, Interscience, New York and London, 1962.
7. 0. Zariski and P. Samuel, Commutative Algebra, volume I, Van Nostrand, Princeton,
1960.
8. , Commutative Algebra, volume II, Van Nostrand Princeton, 1960.
Received July 27, 1967.
UNIVERSITY OF WASHINGTON

On The Structure of Principal Ideal Rings: Thomas W. Hungerford

Uploaded by

Copyright:

Available Formats

On The Structure of Principal Ideal Rings: Thomas W. Hungerford

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

On The Structure of Principal Ideal Rings: Thomas W. Hungerford

Uploaded by

Copyright:

Available Formats

PACIFIC JOURNAL OF MATHEMATICS

Vol. 25, No. 3, 1968

ON THE STRUCTURE OF PRINCIPAL IDEAL RINGS

In this paper all rings are supposed commutative with

A number of special cases of this theorem are well known. In

DEFINITION 2. A Noetherian ring R which has a unique maximal

PROPOSITION 3. The following conditions on a ring R are equivalent:

PROPOSITION 4. Let R be a local ring whose maximal ideal M is

Proof. Let reR. Since f\Mi = Π (α*) = 0, there is a highest

PROPOSITION 5. Let R be a ring. Then:

The proofs of (l)-(3) can be found in Zariski and Samuel [8,

COROLLARY 6. If R is a field, then R[[x]] is a principal ideal

DEFINITION 7. A v-ring is a ring D such that D is a complete

THEOREM 8. Let R be a complete local ring whose maximal ideal

This is a very specialized version of the structure theorem for

DEFINITION 9. A principal ideal ring R is special if R has a

LEMMA 10. Every principal ideal ring R is the direct sum of

It is proved in Krull [4, p. 13] that if D if a v-ring, then

COROLLARY 11. Every special principal ideal ring is the homo-

3* Further comments* In a certain sense, Theorem 1 is best

COROLLARY 12. A principal ideal ring R is a direct sum of

Received July 27, 1967.

You might also like