You are on page 1of 57

ARBAMINCH UNIVERSITY

SCHOOL OF POST GRADUATE STUDIES


COLLAGE OF NATURAL SCIENCE
DEPARTMENT OF MATHEMATICS

MSc. THESIS
ON
COMPLETION OF SEMI- LOCAL RINGS

BY
ABERA GEBREKIDAN ANTENYISTEGN

A THESIS SUBMITTED TO COLLEGE OF NATURAL


SCIENCE, DEPARTMENT OF MATHEMATICS FOR THE
AWARD OF THE DEGREE OF
MASTERS OF SCIENCE
IN
MATHEMATICS(ALGEBRA)

August, 2021
ARBA MINCH UNIVERSITY
SCHOOL OF GRADUATE STUDIES
DEPARTMENT OF MATHEMATICS

Certificate
This is to certify that the thesis entitled “ Completion of semi-local rings” in

partial fulfillment of requirements for the award of the masters of science (MSc.) in

Mathematics(Algebra) submitted to the department of Mathematics, Arbaminch

university carried out by ABERA GEBREKIDAN (I.D PSNS/1525/09) under my

supervision. Therefore I recommend that the student’s Thesis can be presented

for review and open oral presentation.

ARBA MINCH UNIVERSITY DAWIT CHERINET(PHD)


Advisor,
ARBA MINCH
Department of Mathematics
March, 2021 Arba Minch University.

i
Declaration
I here by declare that this MSc. Specialty or equivalent thesis dissertation is my
original work and has not been presented for a degree in any other university,
and all sources of material used for this thesis / dissertation have been duly
acknowledged.

Name:ABERA GEBREKIDAN ANTENYISTEGN

Signature: ———————————————-

Date: —————————————————-

ii
ARBA MINCH UNIVERSITY
SCHOOL OF GRADUATE STUDIES
DEPARTMENT OF MATHEMATICS
APPROVAL

As member of Board of Examiners of the final MSc. Graduate thesis open


defense Examination, we certify that we have read and evaluate this graduate
thesis prepared by ABERA GEBREKIDAN entitled by “ COMPLETION OF
SEMI-LOCAL RINGS” and recommended that it can be accepted as fulfilling the
thesis requirement for the degree of MSc. In Mathematics (Algebra).

APPROVED BY

Advisor
Name :Dr.Dawit Cherinnet Signature ————————– Date—————-

Examiner
Name ————————————– Signature ——————– Date—————-

Chair person
Name ————————————– Signature ——————– Date—————-

Head of department
Name ————————————– Signature ——————– Date—————-

iii
ARBA MINCH UNIVERSITY
SCHOOL OF GRADUATE STUDIES
DEPARTMENT OF MATHEMATICS
APPROVAL OF REVISED THESIS

Name of the candidate —————————————————————————-


College/institute ———————————————————————————–
Department/School——————————————————————————–
Thesis Title:——————————————————————————————
Date of Review ———————————————————————————-
1. Comment on thesis and open presentation
———————————————————————————————————–
———————————————————————————————————–
2. Suggestions made by the reviewer/s
———————————————————————————————————–
———————————————————————————————————–
3. Modification(s) to be made
———————————————————————————————————–
———————————————————————————————————–
4. Final decision by the Reviewer/s
———————————————————————————————————–
———————————————————————————————————–
Name of the Reviewer/s —————————————————————————
Signature ————————– Date———————————————————
Department Head’s Name ———————————————————————–
Signature and Stamp ————————– Date——————————————–

iv
Acknowledgements
Next to GOD, I would like to thank my adviser Dr.Dawit Cherinet for all his help
and support during the work of my thesis. Next I would also like to thank every
one who read and give constractive and helpful comments on doing this thesis.

Abera Gebrekidan

v
Abstract
In this thesis we aim to build an understanding of the theory behind Completion of
semi-local rings. We shall discuss local rings that are general commutative rings
and conclude by study stating ( and proving a special case thereof ) which concerns
formal power series rings, an ideal-adic topology , completion, Exact tensor
products and the theorem of transition. A good understanding of elementary rings
and Module theory is assumed. Henceforth, all rings are assumed commutative
ring with unity.

vi
Contents

Certificate i

Declaration ii

Acknowledgements v

Abstract vi

Contents vii

1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Statement of the Problem . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Objective of the Study . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.1 General Objective of the Study . . . . . . . . . . . . . . . . 5
1.3.2 Specific Objectives of the Study . . . . . . . . . . . . . . . . 5
1.4 Significance of the Study . . . . . . . . . . . . . . . . . . . . . . . . 5
1.5 Delimitation of the Study . . . . . . . . . . . . . . . . . . . . . . . 6
1.6 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 Preliminaries 7
2.1 Nested Ideals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Prime Divisor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Primary ideals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Noetherian Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.5 Jacobson radicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.6 Ring of Quotients . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.7 Local Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.8 primary Decomposition of Ideals . . . . . . . . . . . . . . . . . . . . 12
2.9 The Notions of Height and Altitude . . . . . . . . . . . . . . . . . . 13
2.10 Integral Dependence . . . . . . . . . . . . . . . . . . . . . . . . . . 14

vii
2.11 Noetherian Normal Rings . . . . . . . . . . . . . . . . . . . . . . . 14
2.12 Exact Tensor Products . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.12.1 Exactness Properties of Tensor Producct . . . . . . . . . . . 18

3 COMPLETIONS 23
3.1 Formal Power Series Rings . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.1 Ring structure . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1.2 Topological structure . . . . . . . . . . . . . . . . . . . . . . 25
3.2 An Ideal −adic Topology . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Completions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.4 The theorem of Transition . . . . . . . . . . . . . . . . . . . . . . . 44

References 47

viii
CHAPTER 1

Introduction

1.1 Background

The history of local rings begins with Krull’s paper [9]. Here he
defines a ”Stellenring”as a Notherian ring with only one maximal
ideal (a Noetherian ring is a commutative ring with units which
satisfies the maximum condition for ideals). The name Stellenring
was chosen because such rings are often associated with points on
algebraic and analytic varieties. Chevalley [1] renamed them ”local
rings” since a ring associated with a point on a variety gives local
properties of the variety.
To illustrate the geometric aspect of local rings let us consider an
affine n - dimensional space An over the complex field C. Let x1 , . .
.,xn be the set of coordinates for An and p a point in An . If Rp is the
set of rational functions in x1 , . . .,xn which are regular at p, then Rp
is Neotherian in which mp = {f : f Rp , f (p) = 0} is the only maximal
ideal. Thus Rp is a local ring associated with point p of Ap . An
irreducible variety v going though p defines a prime ideal of Rp , B =
{f : f Rp , f (v) = 0} and vice versa a prime ideal of Rp defines variety
though p. Further the ring Rp /B is again a local ring which we call
1
the local ring of p on v. Thus a ring-theoretic and sheaf –theoretic
study of the set of local rings of points on V might be expected to
yield properties of V. This can be adapted to algebraic varieties over
other defining fields, abstract varieties, and also to analytic varieties
if Rp is replaced by the analytic (holomorphic) functions of P. The
following two sections of this study are will be devoted to basic results
for general commutative rings (section 1) and to the development of
an important tool which is used throughout. This is the completion
of rings with respect to a simple topology (section 2). The fact that
the completed rings have a nice relationship to the original ring is
itself a major reason for the usefulness of the theory of local rings.
In section two we will also studies semi-local rings, because they are
a generalization of local rings but from theoretical necessity. For
instance, a ring which is a finite integral extension of a local integral
domain or more generally a ring which is a finite module over a local
ring is not in general a local ring but rather a semi-local ring.
0
Geometrically, this is accounted for the fact that if V is a finite
covering variety of a variety V, then to each point P of V there
0
corresponds a finite number of points of V , but in general more
than one.
We let R be a Neotherian ring , and I be an ideal of R. Denote by
R̂ the I - adic completion of R.
In this thesis all rings are commutative with unity. Local rings are
define to be Neotherian ring, while quasi local rings are not necessary
Neotherian. Which we write (R, M ) is a quasi local ring, we mean
that R is a quasi local ring with maximal ideal M. In this case R̂
denotes the I- adic completion of R.
In 1986, Lech solved the problem of characterizing completions
of local domains proving that a complete local ring (T, m) is the
completing of a local domain if and only if (1) m = (0) or m ∈AssT
/
2
and (2) no non zero integer T is a zero divisor [6].
Heit mann, in 1993,continued this work by finding all completions of
local unique factorization domain [4].
to be precise, a complete local ring T is the completion of local
unique factorization domain if and only if it is a field,a discrete
valuation ring or a ring of depth at least two with no non zero integer
being a zero divisor.Following this trend, we asked the corresponding
questions for reduced rings :given a complete local ring, when is it
is the completion of a reduced local ring? In this thesis we prove a
theorem that answer this question.
In mathematics ring of formal power series are a generalization of
ring of polynomial as formal objects, where the number of terms is
allow to be infinite; this implies giving up the possibility to substitute
arbitrary values for indeterminates. A formal power seriescan be
loosely thought of as an object that like a polynomial, but with
infinitely many terms.The set of all power series in X with coefficients
in a commutative ring R form an other ring that is written R[[x]],
and called the ring of formal power series in the variable X over R.
One can characterize R[[x]] abstractly as the completion of the
polynomial ring R[x] equipped with a particular metric. This
automatically gives R[[x]] the structure of of a topological ring (and
even of a complete metric space). But the general construction of a
completion of a metric space is more involved than what is needed
here, and would make formal poer series seem more complicated than
they are.

3
1.2 Statement of the Problem

There are many theorems in the completion of local rings which


are not proved. Besides, there are no enough examples, definitions
and illustrations on this area. This study, designed to address the
following basic concepts on Completion of semi-local ring.

X To give briefly explanation on Formal power series of rings.

X To give briefly explanation on an ideal-adic topology

X To state and proof some theorems and corollaries that related


to Completion of local rings.

X To proof existence of the theorem of transition.

4
1.3 Objective of the Study

1.3.1 General Objective of the Study

The general objective of this study is to study more about semi-local


rings and its completions.

1.3.2 Specific Objectives of the Study

The specific objectives of this study are:-

X To study an ideal-adic topology.

X To define formal power series on rings.

X To study completion property of local rings.

X To proof existence of the theorem of transition.

1.4 Significance of the Study

The following are the importance of doing this thesis/ research/

X It can be used as a starting idea for any other researcher with


the same area of interest.

X It also helps any interested body to assure Completion of semi-


local rings.

X It also helps to analyze formal power series rings, an ideal-adic


topology completion property of local rings,a and the theorem
of transition .
5
1.5 Delimitation of the Study

For a research manageable, the researcher has restricted the main


idea on Completion of semi-local rings.

1.6 Methodology

The thesis will be involve collecting information from different


reference books.All information obtain will be recorded and the
following activities will be done to collect materials and procedures
used by different authors will be examined in details.

• A detail study of theory behind completion of semi-local rings.

• Concepts and facts have carried out.

• Examples, theorems, corollaries and definitions of basic


concepts will be given to be illustrate the main idea of the thesis.

6
CHAPTER 2

preliminaries

2.1 Nested Ideals

Definition 1. Let R be a ring, a be an ideal,and K: R −→ R/a the


quotient map. Given an ideal b ⊇ a, form the corresponding set of
co sets of a.

0 0
b/a = {b + a|b ∈ b} = K(b)

2.2 Prime Divisor

Definition 2. A prime ideal q of R is called a prime divisor of I if


there is a multiplicatively closed subset S of R which does not meet
I such that qRs is maximal prime divisor of aRs .

Definition 3. A prime ideal P is a minimal prime ideal of I if and


only if it is minimal among prime divisor of I.
7
Definition 4. A prime divisor of I which is not minimal is called an
imbedded prime divisor of I.

2.3 Primary ideals

Definition 5. An ideal q of a ring R (q 6= R) is called primary if


every zero divisor of R/q is nilpotent, or equivalently,a, b ∈ q, a ∈ q,
b∈
/ q (a, b∈ R) imply that nil-potency of modulo q. This definition
can be expressedas follows:
Let P be the radical of q,then q is primary if and only if a, b ∈ q, b
/ P (a, b ∈ R) imply an ∈ q for natural number n.

Example 1. In the ring of Z, (pn ) is a primary ideal if p is a prime


number.

Theorem 2.3.1. Let φ be a homomorphism from a ring R in to a


0 0 0 0
ring R . If q is a primary ideal with prime divisor p in R , then q =
0 0
φ−1 (q ) is a primary ideal with prime divisor p = φ−1 (p ).

2.4 Noetherian Rings

Definition 6. The commutative ring R is said to be Noetherian or


to satisfy the ascending chain condition on an ideals or (a. c. c) on
ideals if there no infinite increasing chain of ideals in R when I1 ⊆

8
I2 ⊆ I3 ⊆ . . . is an increasing chain of R, then there is a positive
integer m such that Ik = Im for all k ≥ m.

Theorem 2.4.1. Let M be a module over a ring R. Then the following


three conditions are equivalent to each other.

1. M is Noetherian

2. N1 , . . ., Nn , . . . are sub modules of M such that Ni ⊆ Ni+1


for any i = 1, 2, . . ., then thereis an n such that Ni = Nj for
j ≥ n.

3. Every sub module of M is a finite module.

Example 2. The ring of integers and the polynomial ring over a


field are Noetherian.

Theorem 2.4.2. Let M be a finite module over a Noetherian ring R


0
and N, N be (R-) submodules of M.Let a be an ideal of R.Then
there is a natural number r such that

0 0
an N = an−r (ar N
T T
N)

for natural numbers n which are greater than r.(Lemma of Artin Ree
)

Corollary 1. If M is a Noetherian module over a ring R, then R/(0 :


M ) is a Notherian ring and M is a finite R/(0 : M )-module.

9
2.5 Jacobson radicals

Definition 7. The intersection of m of all maximal ideals of a ring


R is called the Jacobson radicals of R.

Example 3. The formal power series ring [[R]], where R is any


X
Noetherian ring is such an example.We can show that f = an xn
n≥0
is in the Jacobson radical of R if and only if a0 is the Jacobson radical
of R,this x is in the Jacobson radical.

Theorem 2.5.1. If m is the Jacobson radical of a Noetherian ring


\
R, and if M is a finite R- module, then mn M = 0; in particular
\ n
n
m = 0.
n
Lemma 1 (Nonunit Criterion). Let R be a ring, n the set of non
units.Ten R is a local if and only if n is an ideal, so, then n is the
maximal ideal

2.6 Ring of Quotients

Definition 8. Let R be a ring and S be a multiplicatively closed


set. The ring S −1 R (or Rs ) is called the ring of quotient (or fraction)
of R by S.
S −1 R has the following properties:

1. 0 ∈ S ⇒ Rs = {0}
10
r
2. The map ϕ :R7−→ Rs , r7−→ 1 (1 ∈ R) is ring homomorphism.

3. The canonical homomorphism 7−→ is monomorphism if and only


if S contains no zero divisors.

Theorem 2.6.1. 1. If a is an ideal of R,then R(x)/aR(x) =


(R/a)(x)

2. If q is a primary ideal with prime divisor p, then pR(x) is prime,


T
qR(x) is primary to pR)x, qR(x) = q, pR(x) R = P.
T T
3. If a1 , ..., an are ideals in R, then (a1 , ..., an )R(x) =
T T
a1 R(x) , ..., an R(x)
0
4. an ideal m of R(x) is a maximal ideal of R(x) if and only if
0
there exists a maximal ideal m of R such that m = mR(x)

5. R < R(x)

Theorem 2.6.2. Let S be a multiplicatively closed subset of a ring R


sch that 0 ∈
/ S. Then ⊗R Rs is exact.

2.7 Local Rings

Definition 9. A ring having just one maximal ideal is called a local


ring. We often express the fact that R is a local ring with maximal
ideal m by saying that (R, m) is a local ring; if this happens, then the
field K = R/m is called the reduce field of R. We will say that (R, m,
K) is a local ring to mean that is a local ring m = rad (R) and K =
11
R/m. If (R, m) is a local ring , then the elements of R not contained
in m are units; conversely a (non-zero) ring whose non-units form an
ideal is a local ring.

Definition 10. A (non-zero) ring having only finitely many maximal


ideals is a semi local ring.

Definition 11. A Noetherian quasi-semi-local ring is called a


semi-local ring; a Noetherian quasi-local ring is called local ring

Definition 12. A ring R is called a quasi-semi-local ring if it has


only a finite number of maximal ideals; it is called a quasi-local ring
if it has only one maximal ideal.

Note. When we say that (R1 , R2 , . . ., Rr ) is quasi-semi-local ring


(or semi- local ring or quasi-local ring or local ring), we mean that
R is quasi-semi-local ring (or semi- local ring or semi-local ring or
quasi-local ring or local ring) and that the maximal ideals of R are
the Pi .

2.8 primary Decomposition of Ideals

Definition 13. We say that a ring R is a graded ring if R is the


direct sum of additive subgroups R1 , R2 , . . .,. . . such that Ri
Rj ⊆ Ri+j (then it follow that 1 ∈ Ro ). An R-module M is called
a graded module if M is the direct sum of Ro -sub moduleMo , . .
12
.M , . . .such that Ri Rj ⊆ Ri+j . Elements of Rn or Mn are called
homogeneous elements of degree n.

Corollary 2. A prime ideal p of a Noetherian graded ring R is a prime


divisor of a graded ideal a of R if and only if there is a homogeneous
element a of R such that a : aR = p.

Corollary 3. Let M be a module over a ring R. If M ⊗ Rm = 0 for


every maximal ideal m of R, then M = 0.

2.9 The Notions of Height and Altitude

Definition 14. We say that ring R is of altitude r if there is a chain


of prime ideals Pi such that PO ⊃ P1 ⊃ . . . . ⊃ Pr but there is no
such chain with more terms. If there is no such r, we say that R is
of infinite altitude.

Definition 15. A system of parameters x1 , . . ., xr of a local ring


R is called regular system of parameters if it generates the maximal
ideal of R. A local ring which has regular system of parameters is
called a regular local ring.

Theorem 2.9.1. Let (R, m) be a local ring. For a given element x of


m, there is a system of parameter of R containing x if and only if
altitude R/xR = altitude R − 1,each of the following conditions is
sufficient:

13
1. altitude R/xR < altitude R

2. There is no prime ideal of altitude 0 which contain x.

3. There is no prime ideal of depth equal to altitude R, which


contains x.

2.10 Integral Dependence


0
Definition 16. Let R be a sub ring of a ring R . An element a of
0
R is said to be integral over R if there are elements c1 , . . ., cn. of
R such that an + c1 an−1 + . . . + cn = 0, that is a is a root of a
0
monic polynomial over R. We say that R is integral over R if every
0
element of R is integral over R.
0
Lemma 2. An element a ∈ R is integral over R if and only if there
00 0 00 00
is a sub ring R of R such that R is a finite R-module and a ∈ R .

2.11 Noetherian Normal Rings

Definition 17. A ring R is called normal if it is a Noetherian domain


and integrally closed.

Example 4. Any principal ideal domain is normal

Example 5. If K is afield, then the ring of formal power series [[ x1 ,


. . ., xn ]] are normal rings
14
Example 6. If K is afield, then the ring of polynomial [ x1 , . . .,
xn ] are normal rings.

Remark 1. A local ring (r, m) which is not a field is a valuation


ring if and only if height m ≥ 1 and m is principal. In this case,
height m = 1.

Theorem 2.11.1. Let P be a prime ideal of a Noetherian ring and


assume that P contains an element a which is not a zero divisor.
Then Pis a prime divisor of aR if and only either height P = 1 and
Rp is a valuation ring or there exists a, b ∈ R such that b/a is integral
over R and such that the conductor of R in Rb/a coincides with P.

Corollary 4. Let R be a Noetherian ring and let a, b elements of


R such that a is not a zero divisor, b/ a R and such that b/ a is
integral over R. Then either there is a minimal prime divisor p of
a R such that Rp is not a normal ring or there exists an imbedded
prime divisor of a R

2.12 Exact Tensor Products

Let R be a ring (R with 1) and let M,N,p be R-module.

Definition 18. A mapping f : M × P is said to be bilinear if

1. For each x ∈ M , the map y 7−→ f (x, y) of N in to P is a linear


(i.e,)f (x, r1 y1 + r2 y2 ) = r1 f (x, y1 ) +r2 f (x, y2 ) (x ∈ M, y1 y2
∈ N)
15
2. for each y ∈, the map x 7−→ f (x, y) of Min to Pis linear
(i.e,)f ( r1 x1 + r2 x2 , y) = r1 f ( x1 , y) +r2 f ( x2 , y) (y ∈ N, x1
x2 ∈ M)

Proposition 1. Let M, N be R-module. Then there exist a pair (T,


g) consisting of an R- module T and a bilinear mapping
g : M × N 7−→ T with the following properties
Given any R-module P and any bi linear map f : M × N 7−→ P ,
0 0
there exists a unique R-linear map f : T 7−→ P such that f = f og.
0 0
Moreover, if (T, g) and (T ,g ) are two pairs with this property, then
0 0
there exist a unique isomorphism j : T 7−→ T such that jog = g .
Let C = R⊕ (M × N ) be free module on M × N . From a linear
X
combination over R of elements of M × N . { ai (xi , yi ): ai ∈ R,
f inite
xi ∈ M, y ∈ N}.
Let D is a submodule of C generated by all elements of C of the
following type
0 0
(x + x , y) - (x, y) - (x , y)
0 0
(x, y + y ) - (x, y) - (x, y )
(ax, y) - a(x,y)
(x, ay)- a(x,y) are deffer from zero.
Let T = C/D, for each pair (x, y) ∈ x⊕y denotes its image in T.Then
T is generated by the form

1. x ⊗ y
0 0
2. x + x ⊗ y = x ⊗ y + x ⊗ y
16
0 0
3. x ⊗ (y + y ) = x ⊗ y + x ⊗ y

4. ax ⊗ y = a (x ⊗ y)

5. x ⊗ ay = a x ⊗ y

Thus, the mapping g : M ⊗ N −→ T define by


g(x, y): = x ⊗ y is bilinear.
Definition The module T constructed above is called tensor product
of M and N.
Example Let R = Z, M = Z, N = Z / 2Z = Z2
0 0
let M = 2Z, N = N
Now, for any x ∈ N
In M ⊗ N ,
2 ⊗x = 2.1⊗x = 1 ⊗2x = 1 ⊗0 = 0
0 0
but 2 ⊗x 6= 0 in M ⊗ N .
0 0
Definition 19. Let f : M −→ M ; g : N −→ N be homomorphism.
Define
h : M × N −→ M ⊗ N by h(x, y) = f (x) ⊗ f (y).Then h is bilinear.
=⇒ h induces an R-module homomorphism
0 0
f ⊗ g : M ⊗ N =⇒ M ⊗ N such that (f ⊗ g)(x ⊗ y) = f (x) ⊗ g(y),
0 0 00 0 0 00
x ∈ M, y ∈ N. Moreover, if f : M −→ M ,if g : N −→ N be
R-module homomorphism.
0 0 0 0 0 0
(f of ) ⊗ (g og) (x⊗y) = f (f (x)) ⊗ g (g (y)) = (f ⊗g ) (f (x)⊗g(y))
0 0 0 0 0 0
= (f ⊗ g ) o (f ⊗ g) (x, y) = (f of ) ⊗ (g og) = (f ⊗ g )o (f ⊗ g).

17
2.12.1 Exactness Properties of Tensor Producct

Definition 20. Let f : M × N −→ P be bilinear map for each


x ∈ M , then
g 7→ f (x, y) is an R-linear. Thus f gives rise to a map
M −→ Hom(N, P )
fx : N −→ P , fx (y) = f (x, y)

0 0 0 0
1. fx (y + y ) = f (x, y + y ) = f (x, y) + f (x, y ) = fx (y) +fx (y )

2. fx (ry) = rfx (y)

Definition 21. On the other hand, given any R-module


homomorphism.
φ :M HomR (N, P ). Then the map (x, y) 7→ φ(x)(y),x ∈ M ,
y ∈ N .This map is bilinear. Also it is possible
M ⊗ N −→ P .
Thus,Hom ( M ⊗ N, P ) ∼
= Hom(M, Hom)(N, P )).

Proposition 2. Let
0 00
M fMgM →0 (2.12.1)

− →

be an exact sequence of an R-module and homomorphism, let N be
R-module. Then the sequence

0 00
M ⊗ N f ⊗ 1M ⊗ N g ⊗ 1M ⊗ N → (2.12.2)
−−−→ −−−→

is an exact(where 3.4.1 is the identity map N on to N)


18
Remark 2. It is not in general true that if:
0 00
M f M g M is exact, then

− →

0 00
M ⊗ N f ⊗ 1 M ⊗ N g ⊗ 1 M ⊗ N is exact.
−−−→ −−−→
Example 7. Let R = Z, then
0 −→ Z −→ Z where f (x) =2x is exact.
Put N = Z/2Z = Z2
Tensoring the sequence with N. We get the sequence.
0 −→ Z ⊗ N f ⊗ I Z ⊗ N
−−−→
let x ⊗ y ∈ Z ⊗ N. Then
(f ⊗ I) (x ⊗ y) = f (x) ⊗ I(y) = 2x ⊗ y = x ⊗ 2y = x ⊗ 0 = 0
=⇒ x ⊗ y ∈ ker(f ⊗ I)
=⇒ ker(f ⊗ I) 6= {0}
f ⊗ I is not injective.
Therefore the sequence is not exact.

Theorem 2.12.1. Let R be a ring and let R∗ be a ring which is an


R-module, and such that ⊗R R∗ is exact.
Let M be an R-module. Then we have the following, where Ni are
submodules of M and a is an ideal R:

Nr ) ⊗ R∗ = (N1 ⊗ R∗ ) (Nr ⊗ R∗ ).
T T T T
1. (N1 . . . . . .

2. (N1 : N2 ) ⊗R∗ = ((N1 ⊗ R∗ ) : N2 ⊗ R∗ )),provided that N2 has


a finite basis.

19
3. (N1 : a) ⊗ R∗ = (N1 ⊗ R∗ ) : aR∗ , provided that a has a finite
basis.

4. If an element a of R is not a zero divisor with respect to N1 ,


then a is not a zero divisor with respect to N1 ⊗ R∗ .

5. If the elements n1 , ..., nr of M are linearly independent over R,


then n1 ⊗ 1, . . ., nr ⊗ 1 in M ⊗ R∗ are linearly independent
over R∗ .

Proof. In order to prove (1), we have only to prove the case where 1.2.
From (N1 + N2 ) /N1 = N2 / (N1 N2 ),we have (N2 ⊗ R∗ )/(N1 N2 )
T T

⊗R∗ = ((N1 + N2 ) ⊗ R∗ )/(N1 ⊗ R∗ ) = (N2 ⊗ R∗ )/((N1 ⊗ R∗ ) (N2 ⊗


T

R∗ )),which proves (1).Since N1 has a finite basis, in order to prove


(2), we may assume that N2 = bR with an element b (N1 , by virtue
of (1) above and (1.2). The map φ defined by φ (x modulo N1 :
T
bR) = (bxmoduloN 1 bR) is isomorphism from R/(N1 : bR) on to
bR/(N1 bR) by (1.5). There fore R∗ /(N1 : bR)R∗φ⊗R ∼

= (b ⊗ R∗ )/
T

((N1 bR)⊗R∗ ) = (b⊗R∗ ) /((N1 ⊗R∗ ) (b⊗R∗ ))φ∗−1 ∼ = R∗ /((N1 ⊗


T T

R∗ ) : (b ⊗ R∗ )),where φ∗ is defined similar as φ. Hence we have (2).


Since N1 has a finite basis, in order to prove (3), we may assume
that N2 = bM with an element b (N1 , by virtue of (1) above and
(1.2). The

Theorem 2.12.2. Let R be a ring and let R∗ be a ring which is an


R-module, and such that ⊗R R∗ is exact.

20
Let M be an R-module. Then we have the following, where Ni are
submodules of M and a is an ideal R:

Nr ) ⊗ R∗ = (N1 ⊗ R∗ ) (Nr ⊗ R∗ ).
T T T T
1. (N1 . . . . . .

2. (N1 : N2 ) ⊗R∗ = ((N1 ⊗ R∗ ) : N2 ⊗ R∗ )),provided that N2 has


a finite basis.

3. (N1 : a) ⊗ R∗ = (N1 ⊗ R∗ ) : aR∗ , provided that a has a finite


basis.

4. If an element a of R is not a zero divisor with respect to N1 ,


then a is not a zero divisor with respect to N1 ⊗ R∗ .

5. If the elements n1 , ..., nr of M are linearly independent over R,


then n1 ⊗ 1, . . ., nr ⊗ 1 in M ⊗ R∗ are linearly independent
over R∗ .

Theorem 2.12.3. Let R be a ring and let R∗ be a ring which an


R-module, assume furthermore that there is ideal a of R such that a
6= R and aR∗ = R∗ . Then R ⊆ R∗ and, for any ideal a of R, it holds
that aR∗
T
R = a.

Theorem 2.12.4. Assume that R and R∗ are Noetherian rings such


that R∗ is an R- module. Let φ(a) = a.1 (in R∗ ) Let R∗ be the set
of maximal ideals of R∗ and let R be the set of prime ideals m of R
such that m = φ−1 (m∗ ) with R∗ ∈ R. Then ⊗R R∗ is exact if and
only if the following is true:

21
if q is a primary ideal with prime divisor m ∈ R and if b is an
element of R such that q : bR = m, then qR∗ : bR∗ = mR∗ .

22
CHAPTER 3

COMPLETIONS

3.1 Formal Power Series Rings

Definition 22. the set of all power series in X with coefficients in


a commutative ring R form another ring that is written R[[x]], and
called the of formal power seriesin the variabble X over R.

3.1.1 Ring structure

As a set ,R[[x]] can be constructed as the set RN of all infinite


sequence of elements of R, indexed by the natural numbers (taken
to be include 0). Desgnating a sequence whose term at index n is an
by (an )n∈N , one defines addition of two such sequences by

(an )n∈N + (bn )n∈N = (an + bn )n∈N

and multiplication by
23
n
X
(an )n∈N × (bn )n∈N = ( ak bn−k )n∈N .
k=0

This types of products is called the cauchy product of the two


sequences of coefficients, and is sort of discrete convolution.With
these operations, RN be comes a commutative ring with zero element
(0, 0 0, . . .) and multiplicative identity (1, 0 0, . . . ).
The product is in fact the same used to define the product
of polynomials in one indeterminate,which suggests using similar
notation. One embeds R in to R[[X]] by sending any (constant) a ∈
Rto the sequence (a, 0, 0, ...)and designates the sequence (0, 1, 0, 0, ...)
by X;then using the above definitions every sequence with only
finitely many nonzero terms can be expressed in terms of these
special elements as

n
X
2 n
(a0 , a1 , a2 , ...an , 0, 0, ...) = (a0 + a1 X + a2 X + ... + an X = ai X i ;
i=0

these are precisely the polynomials in X. Given this, it is quite


natural and convenient to designate a general sequence an by (an )n∈N
X
by the formal expression ai X i , even though the latter is not an
i∈N
expression formed by the operations of addition and multiplication
defined above (from which only finite sums can be constructed). This
notational convention allows reformulation the above as

24
3.1.2 Topological structure

Having stipulated conventionally that



X
(ao ,a1 ,a2 , a3 , . . ) = ai X i
i=0
One would like to interpret the right hand side as well-defined
infinite summation. To that end, a notion of convergence in RN is
constructed. There are several equivalent ways to define the desired
topology.
Informally, two sequence (an ) and (bn ) become closer and closer if
and only if more and more of their terms agree exactly. Formally,
the sequence of partial sums of some infinite summation converges if
for every fixed power of X stabilizes: there is a point beyond which
all further partial sums have the same coefficient.
This topological structure, together with the ring operations
described above, form a topological ring.this is called the ring of
formal power series over R and is denoted by [[R]]. The topology
has the use full property that an infinite summation converges if and
only if the sequence of its terms converges to 0, which just means
that any fixed power of X occurs in only finitely many terms.
Topological structure allows much more flexible use of infinitely
summations. For instance the rule for multiplication can be restated
simply as
X X X
i i
( ai X ) × ( bi X ) = ai bj X i+j ,
n∈N n∈N n∈N
since only finitely many terms on the right affect any fixed X n .
25
Infinite product also defined by the topological structure; it can be
seen that an infinite product converges if and only if the sequence of
its factors converges to 1.
Consider the ring of formal power series
Z[[X]][[Y ]], then the topology of above construction only relates to
the indeterminate Y, since the topology that was put on Z [[X]]
has been replacedby the discrete topology when we defining the
X
topology of the whole ring.So XY i converges to the power series
n∈N
X
suggested,which can be written as 1−Y ; however the summation
X
XY i Would be considered to be divergent, since every term
n∈N
affects the coefficient of Y (which coefficient is itself a power series in
X).This asymmetry disappears if the power series ring in Y is given
the product topology where each copy ofZ [[X]] is given its topology
as a ring of formal power series rather than the discrete topology. As
a consequence, for convergence of a sequence of elements Z[[X]][[Y ]]
it then suffices that the coefficient of each power of Y converges to a
formal power series in X, a weaker condition that stabilizing entirely
; for instance in the second example given here the coefficient of Y
1 Y
converges to 1−X , so the whole summation converges to 1−X .

This way of defining the topology is in fact the standard one for
repeated constructions of rings of formal power series, and gives the
same topology as one would get by taking formal power series in
all indeterminates at once.In the above examples that would mean
constructing Z[[X, Y ]],and here a sequence converges if and only if
26
the coefficient of every monomial X i Y j stabilizes. This topology,
which is also the I - adic topology,where I= (X, Y) is the ideal
generated by X and Y,still enjoys the property that a summation
converges if and only if it terms tends to 0.

Definition 23. Let R be a ring. We define the ring of formal power


serie in X over R to be

X
[[R]] = { ri X i |ri ∈ R}
i=0

X ∞
X
i
Given any a = ai X |ri ∈ R and b = bi X i |ri ∈ R, we define
i=0 i=0
their sum as
X∞ ∞
X ∞
X
i i
ai X |ri ∈ R and b + bi X |ri ∈ R = (ai + bi )X i
i=0 i=0 i=0
and their product as
X∞ ∞
X ∞
X ∞
X
i i i
( ai X ) ( bi X ) = ci X , ck = ai b j X i
i=0 i=0 i=0 i+j=k

Theorem 3.1.1. There is a one - one correspondence between all


maximal ideals m of R and all ideals m∗ of R[[x]] in such away
that m∗ corresponds to m if and only if m∗ is generated by m and
the xi .

Proof. We are obviously that, when m is a maximal ideal of R, the


ideal mR[[x]] + xi R[[x]] is the maximal ideal of R[[x]].Let m∗ be a
P

maximal ideal of R[[x]] and let m be set of constant terms of elements


of m∗ . Then m is obviously an ideal of R. If we know that m 6= R,
then m∗ is is contained in mR[[x]] +
P
xi R[[x]], and we see that
m∗ = mR[[x]] +
P
xi R[[x]] and m is the maximal. Therefore it is
27
sufficient to show that m 6= R. Assume the contrary, Then there is
X

P
an element ai of m with ao = 1, set b = −( ai ). Then we can
1
n n
0
X X
consider f = b which is equal to ci where each ci is defined
0 0
m
0
X
to be the nth degree part of b for an m.n hence for any m ≥ n.
0
n n
0
X X X X
m+n
Since ( ai ) ( b) = 1 - b , we see that ( ai ) ( ci ) = 1
0 0
+ ( (termsof degree>n)), which means that th e nth degree part
P
X X
of ( ai ) ( ci ) is 1 or zero according as n is zero or not. Thus
X
we see that the element ai is a unit, which contradicts to the
X
assumption that ai ∈ m∗ , and the proof is completed.

Corollary 5. An element x ∈ R[[x]] is a unit if and only if the constant


term is a unit in R.

X
Proof. (=⇒) f : R[[x]] −→ R defined by x = ai xi 7−→ a0 is a
i=0
ring map.
Hence x is a unit.
=⇒ f (x) = a0 is a unit.

X ∞
X
i
(⇐=) 1 − x is a unit for all x ∈ ai x with inverse bi xi .
i=0 i=0

X
=⇒ Let x = ai xi ∈ R[[x]], where a0 is a unit. We want to
i=0

X
construct y = bi xi such that xy = 1 or after expanding
i=0
xy = a0 b0 +(a1 b0 + a0 b1 )x + . . . + 0x + 0x2 + . . ..
We there fore need a0 b0 = 1 (re call that a0 is a unit).
28
We want to have
a1 b0 + a0 b1 = 0, so our only choice for b1 is

−a1 b0
b1 = a0 = a1 a20

we want to must a2 b0 + a1 b1 + a0 b2 = 0 so we have

−a2 b0 −a1 b1
b2 = a0 = a2 a−2 −3
0 + a1 a0

By applying in this fashion we have


n
X
bn = ai bn−i , for n ≥ 0
i=0

an important special case is that the geometric series formula is valid


in R[[x]];

X
(1 − x)−1 = x−n
n≥0

Since b0 is a unit in R this some argument gives us that for x =


X∞
ai xi , a0 is a unit in R.
i=0

Theorem 3.1.2 (Formal Hilbert Basis Theorem ). . Let R be a


Noetherian ring. then R[[X1 ,. . . Xn ]] is Noetherian.

Proof. We prove the theorem for the case of indeterminate X. The


general case then follows easily by induction.If f = an+1 X n+1 + . .
. ∈ R[[x]] is a power series, we shall say that the degree of f is n and
that is anti-leading coefficient an . If f = 0, then we shall consider
29
the degree of f to be infinite and its anti-leading cofficient to be 0
let I M R[[x]] be an ideal. We need to show that I is finitely
generated. Fix an f1 ∈ I of minimal degree (we can clearly do
this since the degree must be positive).We first define a sequence
of power series {fn /n ∈N} ⊆ I inductively. Suppose that we have
already chosen elements f1 , . . ., fi−1 . Denote their degrees by di
and their anti-leading coefficients by ai .Now if (f1 , . . ., fi−1 ) 6= I,
then choose an fi ∈ I/(f1 , . . ., fi−1 ) of minimal degree. If this
process terminates then clearly, Ib is finitely generated. If not, then
consider the ideal (a1 , . . ., ai ) M R generated by the anti-leading
coefficients of the first i functions in the sequence.We have that
(a1 ) ⊆ (a1 , a2 ) ⊆ . . .⊆ (a1 , . . ., ai )⊆ . . . is an ascending chain
of ideals in R. But R is Noetherian so this chain must stabilize at
say i = n. We claim that I is generated by f1 , . . ., fn .
Fix some g ∈ I. Let ag be its anti-leading coefficient and dg its
degree.It is clear thatag ∈ (a1 , . . ., ai ). Hence we may write ag
n
(0) (0)
X
= ri X dg −di ai for some ri ∈ R. First suppose that ag ≥ d.g .
i=1
Define the power series
n
(0)
X
go = ri X dg −di fi
i=1
We observe that go also has degree dg and anti-leading coefficient ag .
It thus follows that the degree ofg -go is greater than dg . Now, g
and go are both in I so we must have that g -go ∈ I. We see that
the anti leading coefficient of g -go is in the ideal generated by a1 ,
. . ., an . We can then repeat this process to produce a power
30
series g1 with anti-leading coefficient and degree equal to those of g -
go . Continuing in this fashion , we inductively define a sequence of
power series go , . . ., gm such that
n m
(0) dg +m−di
X X
m
gm = ri X fi for some ri ∈ R. Now,g - gi has degree
i=1 i=1
greater than dg + m and we clearly see that
∞ ∞ X n
(i)
X X
g= gi = rj X dg +i−dj fi
i=1 i=1 i=1
Since the inner sum is finite, we may swap the summations to see
thatg is finitely generated by fi , . . ., fn .
Now suppose that ag < dn . We now that ag ∈ (a1 , . . .,an ) and there
must exist some 1≤ k ≤ n such that ag ∈ (a1 , ...ak ). Hence dg ≥ dk .
k
(0) (o)
X
We may write gg = ri ai for some ri ∈R. Now define
i=1
k
(0)
X
h= ri X dg −di fi Which has the same anti-leading coefficient and
i=1
degree as g. We see that g - h has degree greater than dg . Again, we
inductively define a sequence in this fashion until we reach a power
series of degreedn . We can then apply the result from the previous
case to see that g ∈ (f1 , . . .fn ).

Corollary 6. If R is a semi-local ring, then so is R[[x]], too; If R is


local ring, then so is R[[x]].

Proof. Let R be a ring. A formal power series in the n variables x1 ,


X
..., xn is a formal infinite sum of the form a(i) x1 i1 ... xinn where
a(i) ∈ R and where (i) = (i1 , ..., in ) with each ij ≥ 0. Addition
and multiplication are performed as as (definition 23); with these
31
operations, these series form a ring R[[x1 , ..., xn ]].
Set P = R[[x1 , ..., xn ]] and a = hx1 , ..., xn i. Then

X
a(i) x1 i1 ... xinn 7−→ a(0)

is a canonical surjective ring map P −→ R with kernel a; hence, P/a


= R.
Given an ideal m ⊂ R, set n = a + mP . Then (Nested ideal) yields
P/n = R/m.
Suppose R is a local ring with maximal ideal m. Then any power
series f 6∈ n is the form f = a(1 − g) with a∈ R× and g ∈ a. h = a−1
(1 + g + g 2 + ...); this sum makes sense as the component of degree
d involves only the first d + 1 summands. clearly f.h = 1. Hence
the non-units constitute n. Thus P is local with maximal ideal n by
(lemma 2).

3.2 An Ideal −adic Topology

Definition 24. Let a be an ideal of a ring R and let M be an


R -module, let F the family of an M We introduce a topology
on m,which may not To − topology, taking F to be a base of
neighborhoods of zero of R, which means that open sets of M are
unions of arbitrary number of sets of the form b + an M ( ∈ M ).
This topology is called the a - adic topology of M . The addition is
continuous (f (x, y) = x + y is a continuous function from M × M
32
in to M ),and the multiplication of elements of R if also continuous
(i.e,f (x, y) = xy is continuous function from R × M in to M , where
the topology of R is assumed to be the a - adic topology of R).

an M
T
Theorem 3.2.1. The a-adic topology of M is To if and only if
= 0. If the adic topology of M is To , then the following distance
function makes M a metric space : r(x, x) = 0, r(x, y) = 2−n if and
only if x − y ∈ an M and x − y 6∈ an−1 M .

Theorem 3.2.2. A sub module N of the R-module M is an open set


of M in the a -adic topology, if and only if N : M contains some
power of a. In that case, N is also a closed set.

Proof. If N is an open set, the 0 ∈ N implies an M ⊆ N for some


n.Conversely, if an M ⊆ N, then since N is a sub module ban M ⊆ N
for any b ∈ N, and N is the union sets b + an M , which proves that
N is open. If N is open, then the complement M − N of N is the
union of x + with x ∈ M − N, and each x + N is open, hence M −
N is open. Thus N is closed.

Theorem 3.2.3. The closure N ∗ of a sub module N of M in M with


\
the a- adic topology coincides with (N + an M ).
n

Proof. Since N + an M is an open set, it is closed set by (3.2.1).Hence


\
N ⊆ (N + a M ) for any n, and n ⊆ (N + an M ).
∗ n
\
Conversely, let x be an arbitrary element of (N + an M ). Then x
= bn + an with bn ∈ N , an ∈ an M for each n, which show that for
33
any n, x + an M meets N. Since the x + an M form a basis for the
neighborhoods of x. it follows that x is in the closure N ∗ of N.Thus
the assertion is proved

Corollary 7. For a submodule N of M, the a- adic topology of M/N


is To if and only if N is a closed subset in the a- adic topology of M.

Proof. The a- adic topology of M/N is To if and only if (an M +


T

N )/N = 0 or equivalent (an M + N )/N = N, which proves the


T

assertion, we note that when N is a submodule of M, the a-adic


topology of N may be different from the topology of N as a subspace
of M.Hence the following theorem is really noteworthy.

Theorem 3.2.4. If M is a Noetherian module, then for any sub module


N of M, the a-adic topology of N coincides with the topology of N
as a subspace of M with the a-adic topology.

Proof. By virtue of corollary (1), we may assume that R is


Noetherian.It is obvious that an N ⊆ an M
T
N . The lemma of
Artin-Rees (2.2.2) implies that an M N = an−r (ar M N ) ⊆
T T

an−r N ,thus we prove the assertion.

Remark 3. Let R∗ be the ring R⊕M in the principle of idealization.


Then the a -adic topology of M coincides with the topology of M as
a sub-space of R∗ with (a ⊕ M )-adic topology (which is equivalent
to aR∗ -adic topology.)

34
Definition 25. A semi- local ring which may not be Noetherian,
we define the Jacobson -radical-adic topology to be the natural
topology.When M is a finite module over a semi-local ring R with
Jacobson radical m, them m-adic topology of M is defined to be the
natural topology of M. This definition is justified by (3.2.3).

Theorem 3.2.5. Assume that M is a finite module over a semi-local


ring R with Jacobson radical m. Then an arbitrary submodule N of
\
M is a closed subspace of M and N = (N + mn M ).
n

Proof. Since M/N is Noetherian, it is a To -space by (2.3.1), and we


prove the assertion by (3.2.2), (corollary 5) add (3.2.3).
0
Theorem 3.2.6. Assume that a semi -local rings R is a finite module
0
over a semi-local ring R. Then the topology of R as a semi-local ring
0 0
coincides with that of R as a finite R-module. If furthermore, R
0 0
contains R(hence R dominates R),then R is a closed subspace of R .

0 0
Proof. Let m and m be the Jacobson radicals of R and R
0 0
respectively. R /mR is a finite R-module, hence is a finite
R/m-module.Therefore, there exists a natural number n such that
0 0 0 0 0
m n ⊆ mR , which shows that m ns ⊆ ms R ⊆ m s , and the first
assertion is proved.

35
3.3 Completions

This section is concerned with completions of semi-local rings and


with completions of finite modules over semi-local rings. But we
begin with a case a bit more general so that the readers can see
some general facts in the case of semi local rings which may not be
Noetherian.

Definition 26. Let R be a ring and let M is an R-module. Assume


that M is a metric space with a distance function r(x, y) such that:

1. For any r(x, y) = r (x − y, 0) for some any x, y ∈ M.


S
2. For any positive real number , the set  of x such that r(x, 0)
<  forms an R-module of M.

Definition 27. If r(a, b) > r(c, d), then r(a, b) = r(a+c, b+d). With
this metric, we can discuss completions as usual.Namely :A sequence
{Cn } of elements Cn (n = 1, 2, . . .) of M is called a cauchy sequence
if for any given positive number , there is a natural number N such
that, for any m and n which are greater than N, r(Cm , Cn ) < .

Definition 28. An element a is said to be a limit of Cauchy sequence


if

lim r(cn , a) = 0;
n→∞

36
if such an a exists for a given {cn }, then a is a unique and is denoted
by lim cn .

Definition 29. M is said to be complete if every Cauchy sequence


in M has a limit in M.

Definition 30. An R-module M ∗ , having the following properties,


is called the a completion of M:

1. M ∗ is a metric space with a distance function r∗ such that


r∗ (x∗ ,y ∗ ) = r∗ (x∗ + ,y ∗ , 0) for any x∗ , y ∗ ∈ M ∗

2. (M ∗ is complete.

3. M is a dense subspace of M ∗

4. If x∗ and y ∗ are limits of Cauchy sequences {Xn } and {yn }


respectively, then r∗ =(x∗ ,y ∗ ) = lim r(cn , dn ) = 0.
n→∞

Definition 31. A Cauchy sequence {cn } is called regular sequence


if
1n
r (cj , cn ) < 2 for any n and for any j > n. Then as a general fact
in metric spaces, we have.
For any Cauchy sequence {cn } in M, there is a regular sequence {dn }
such that lim r(cn , dn ) = 0.

Now we prove the existence of completion

Theorem 3.3.1. M has a completion M ∗ which is unique up to


isomorphisms (with regards to topologies) . Then set ∗ elements
S

37
x∗ of M ∗ such that r∗ (x∗ , 0)<  forms an R-submodule of M ∗ and
is the closure of  in M ∗ . M ∗ /  ∗ is naturally isomorphic to M/
S S


S S
 . If M is a ring and if  are ideals, then M is naturally a ring

whose multiplication is such that (lim c∗n ) (lim d∗n ) =lim(c∗n d∗n ) for any
Cauchy sequences {d∗n } and {c∗n } in M ∗ ,and the ∗ are ideals of M ∗ .
S

Proof. Let C and C ∗ be the set of Cauchy sequences and regular


sequences in M respectively. In C,we introduce the following
0 0
operations: {cn } +{ cn } = {cn + cn }, a {cn } ( a ∈ R), and when
0 0
M is a ring (a above), {cn } { cn } = {cn cn }. Thus C be comes an
R-module;when M is a ring, C be comes a ring. Let n be the set of
Cauchy sequences which have zero as limits. Then n is a submodule
of C; when M is a ring, n is an ideal. We shall show that C/n is
a completion of M,identifying an element a of M with the class of
Cauchy sequences which have a as the limit.We define a function
r∗ ( x∗ , y ∗ ) ( x∗ , y ∗ ∈ C/n) to be lim r(xn , yn ) with representatives
n→∞
∗ ∗ ∗ ∗ ∗
{xn },{yn } of x , y .We see that r ( x , y ) = r (xn , yn ) for sufficiently
large n except for the case where x∗ = x∗ . With this r∗ , C/n be
comes a metric space as is easily seen. Furthermore,if x∗ is a class
in C/n containing {cn }, then x∗ =lim cn in this topology. Thus
M is dense in C/n. Let {x∗n } be a Cauchy sequence in C/n. Let
{xni } be regular sequence in M such that x∗n = lim xni . Then by
the regularity of the {xni }, the sequence {xnn } be comes a Cauchy
sequence, hence has a lim x∗ in C/n. Then we see that x∗ is a limit of

38
the sequence {x∗n }. Thus C/n is a completion of M. The uniqueness
of M ∗ is rather obvious, because element of M ∗ must be one- one
correspondence with elements of C/n.

Theorem 3.3.2. Assume that a is an ideal of R and that the metric


of M is given by the a -adic topology.If a has a finite basis, then the
topology of the completion M ∗ is the a -adic topology

Proof. It is sufficient, by virtue of (3.3.1), to show that the closure


00 00
Nn∗ of a M in M ∗ is a M ∗ . Let x∗ be arbitrary element of Nn∗ .
Then x∗ = lim xi with {xi } such that x1 ∈ an M and such that
xi − xi+1 ∈ an+i = an (ai M ).Let a1 , ..., at be a basis for an . Then
X X
xi − xi+1 = aj bji with bji ∈ ai M . For each j, the series bji
j i
t
X
is convergent (i.e.,{ bji } is a Cauchy sequence) and express an
i=1 X
element bj of M . then we see that x∗ =
∗ ∗
aj b∗j , which is in an M ∗ .
Thus Nn∗ ⊆ an M ∗ .

Theorem 3.3.3. Assume that the a -adic topology of R is To and that


a has a finite basis a1 , . . ., ar . Let x1 , . . ., xr be indeterminates and
consider the formal power series ring R[[x]] . Then the completion
R∗ of R is (isomorphic to ) the R[[x]]/n∗ , where n∗ is closure of the
P
ideal n = (xi − ai ) R[[x]] in R[[x]] with the (n + aR[[x]]) adic
P
topology ( = the (a[[R]]) + xi R[[x]] )- adic topology.

39
Proof. We consider the map φ from R[[x]] into R∗ such that if f
P
fi (x) ∈ R[[x]](fi (x) being homogeneous form of degree i), then
P
φ(f ) = fi (a).Then we see that φ is a homomorphism from R[[x]]
0
on to R∗ and that the krrnel n of φ contains x−a, hence f = fi (x)
P

0 P
is in n if and only if fi (a) = 0. Let fu (x) be the leading form of
fu (a) ∈ au+1 , and there is a
P P
f . Then fi (a) = 0 implies that
homogeneous form hu+1 (x) degree u + 1 such that
fu (a) = hu+1 (a) .f - fu (x) + hu+1 (x) has leading degree greater than
u and f ≡ fu(x) + hu+1 (x) modulo n. Thus, repeating the same, we
0
see that if f ∈ n , then f ∈ n + ( xi R[[x]])n for any n.
P

Conversely, if f ∈ (n + ( xi R[[x]])n ), then we see that


T P
\ 0
\ 0 X
n
xi R[[x]])n ) =
P
fi (a) ∈ a = 0. Thus we have n = (n + (
\ n n
n ∗
(n + a R[[x]]) = n .
n

Theorem 3.3.4. Let (R, p1 , . . ., pr ) be a semi- local rings. Then the


completion R∗ of Ris the direct sum of completions Ri∗ of local rings
Ri = Rpi .

\
Proof. Let m be the Jacobson radical of R, i.e., m = pi . Then for
i
each natural number n, R/mn is the direct sum of pni and there are
X
i,n ∈ R such that 1 - i,n ∈ mn ,i,n ∈ pnj i 6= j, hence 1 - i,n ∈
i
pj .Then for each i,i,n - i,n+1 ∈ pn1
n
. . . pnr = mn and therefore
T T

the sequence {i,n } has a limit i in R∗ . Since


X
lim i,n = 1,
i

40
i = 1. Since i,n j,n ∈ mn (i 6= j), we have i j = 0 if
P
we have
i 6= j.. Hence we have 2i = i . Thus R∗ is the direct sum of ideals
R∗ i , which are rings with identities i . We shall prove now that
R∗ i is (isomorphic to) the completion of Ri . Let {cn } be a Cauchy
0 0 0
sequence in R and lim cn i,n = lim cn 2i,n = i (lim cn i,n ). Hence the
0
limit of {cn i,n } is in R∗ i .It is easy to see that the map lim cn →
0
lim cn i,n gives a one-one correspondence between the completion
Ri∗ and R∗ i . Then we see easily that Ri∗ is isomorphic to R∗ i .

Now we go back to the completions of modules:

Theorem 3.3.5. Assume that R is a semi-local ring and that M is a


finite R-module.Let R∗ be the completion of R. Then M ⊗R R is
the completion of M,whose topology as a finite R∗ - module coincides
with its topology as the completion of M.

Proof. Let m be the Jacobson radical of of R. Then mR∗ is the


Jacobson of R∗ . Furthermore R∗ /mn R∗ = R/mn R. Now,
(M ⊗R∗ ) /mn (M ⊗R∗ ) = (M /mn M ) ⊗ (R∗ /mn R∗ ) =M /mn M .
m M = 0, we see that M ⊗ R∗ contains M (by the
T n
Since
identification m = m⊗ 1 for any m ∈ M), and furthermore that
mn (M ⊗R∗ ) M = mn M . This shows that M is a space of M
T

⊗ R∗ . An arbitrary element c∗ of M ⊗ R∗ can be expressed as


X
mi ⊗ ri∗ with mi ∈ M and ri∗ = lim ri , n with ri , n ∈ R.Then c∗
n
P
is the limit of { mi ri , n} as is easily seen. Thus M is dense in M

41
⊗ R∗ . Let {c∗n } be a regular sequence in M ⊗ R∗ . Let ui , . . . ut
be a basis for M. Since (M ⊗ R∗ ) /mn (M ⊗ R∗ ) = M/mn M , there
is a sequence {mn } in M such that c∗n - cn ∈ mn (M ⊗ R∗ ). cn+1 -
cn is in mn M, hence is expressed as ui min (min ∈ mn ). For each i,
P

min has limit ri∗ in R∗ . Furthermore,we see that ui ⊗ ri∗ is the


P P

limit of the sequence {c∗i }. Thus M ⊗ R∗ . is complete.

Corollary 8. Let a be an ideal of a semi local ring R. Let R∗ be


the completion of R. Then the completion of a is aR∗ and aR∗ is
isomorphic to a ⊗ R∗ .Furthermore aR∗ R = a and R∗ /aR∗ is the
T

completion of R/a.

Proof. The first assertion follows from (3.2.3), hence the second
assertion follows from (3.3.5). Since a is a closed set by (3.3.4)
and since aR∗ is its closure in R∗ , we see that aR∗ R = a. The
T

completion of R/a is R/a ⊗ R∗ , which is obviously R∗ /aR∗ .

Similarly, we havefor modules:

Corollary 9. Let N be a submodule of a finite module M over a


semi local ring R.Then the closure N ∗ of N in the completion M ∗ of
M is the completion of N and we have N ∗ M = N . Furthermore,
T

M ∗ /N ∗ is the completion of M/N .

Proof. The first assertion follows from (3.2.3), hence the second
assertion follows from (3.3.5). Since N is a closed set by (3.3.4)
42
and since N ∗ is its closure in N , we see that N ∗
T
M = N . The
completion of M/N is M/N ⊗ is M ∗ , which is obviously M ∗ /N ∗ .

If we identify M ∗ with M ⊗ R∗ , then N ∗ is surely identified with


N ⊗ R∗ in M ⊗ R∗ as easily seen. that N ∗ is the completion of N
implies that the usual tensor product N ⊗ R∗ is isomorphic to the
sub module N ⊗ R∗ in M ⊗ R∗ . Hence we have

Corollary 10. If R is a semi-local ring,then ⊗R R∗ is exact.

Proof.

Corollary 11. Let (R, p1 , ..., pr ) be a semi- local ring and let R∗ be
the completions of R. Then we have altitude R∗ = altitude R

Proof. altitude R = max{altitudeRPi }. On the other hand, R∗ is


the direct sum of completions of RPi . Therefore it is sufficient to
prove the assertion in the case where R is a local ring with maximal
ideal m = pi . We prove the assertion by induction on altitude R∗ .
If altitude R∗ = 0, then R∗ has discrete topology, hence R = R∗
and altitude R∗ = altitude R in this case. Assuming that altitude
R∗ > 0, let x1 , ..., xs be a system of parameters of R. Then the xi
generate an ideal which is primary to mR∗ in R∗ . Therefore we see
that altitude R∗ ≤ altitude R and that there is an element xi , say
x1 , such that R∗ / x1 R∗ has altitude less than altitude R∗ , hence
altitude R∗ / x1 R∗ = altitude R∗ − 1 by (2.8.1). Since R∗ /x1 R∗ is

43
the completions of R/x1 R, we see that altitude R∗ /x1 R∗ = altitude
R/x1 R by our induction assumption.Thus we have altitude R∗ =
altitude R∗ /x1 R∗ + 1 = altitude R/x1 R + 1 ≥ altitude R,hence we
have altitude R∗ = altitude R.

3.4 The theorem of Transition

Definition 32. We say that the theorem of a transition holds for


0
rings R and R if:

0
1. R is dominated by R

2. if q is a primary ideal in R such that the prime divisor of q is a


maximal ideal,say m, then

0 0
lengthR0 R / qR

is finite and it holds that


0 0 0 0
lengthR0 R / qR = (lengthR0 R / mR ) (lengthR R/ q ).
0 0
Theorem 3.4.1. Let R and R be Noetherian rings such that R ≤ R .
0 0
Assume that for any maximal ideal m of R, the length of R / mR is
0
finite (i.e., the prime divisors of mR are all maximal). Then each of
the following is a necessary and sufficient condition for the validity
0
of the theorem of transition for the rings R and R :

44
1. If q is a prime a primary ideal belonging to a maximal ideal m
0 0
in R and if q : bR = m for an element b of R, then qR : bR =
0
mR .
0
2. ⊗R R is exact.
0 0
3. For any maximal ideal m of R , the theorem of transition holds
0
for Rm0 T R and Rm
0 0 0
4. If a and b are ideals of R, then (a : b)R = aR : bR

Proof. We show first that the validity of theorem of transition implies


0
(1). Set q = q + b R. Then lengthR/q = lengthR/q 1 + 1,
0 0 0 0 0 0 0
hence lengthR /qR = lengthR /q R + lengthR /mR , which implies
0 0 0 0 0 0 0 T 0
that lengthR /qR = lengthq R /qR = lengthbR /(bR qR ) =
0 0 0 0 0 0
lengthR /(qR : bR ), which implies that mR = qR : bR . We
see the equivalence with (1).
0
(1) implies (2) by (3.4.4) we have Assume that R and R are
0
Noetherian rings such that R is an R- module. Let φ(a) = a.1
0 0 0
(in R ) Let R be the set of maximal ideals of R and let R be the set
0 0
of prime ideals m of R such that m = φ−1 (m ) with R ∈ R. Then
0
⊗R R is exact if and only if the following is true:
if q is a primary ideal with prime divisor m ∈ R and if b is an
0 0 0
element of R such that q : bR = m, then qR : bR = mR .
(2) implies (4) by (3.4.1), while (4) implies (1) obviously. Thus (1),
(2) and (3) are equivalent. It is obvious that (3) implies the validity

45
of the theorem of transition;the converse is also obvious except for
the fact that R∗ = (Rm T R ) ⊆ R∗∗ = Rm0 ,which is proved as follows
⊗R∗ R∗∗ is exact by (2) and (cor 1), hence (3.4.2) implies the inclusion
relation.

Corollary 12. Assume that the theorem of transition holds for


0
Noetherian rings R and R .

1. For any ideal a of R, the theorem of transition holds for the


0 0
rings R/a and R /aR .
0
2. Assume that S and S are multiplicatively closed subsets of
0 0
R and R respectively such that every maximal ideal p with
0 0 0
respect to S in R is a minimal prime divisor of pR for some
maximal ideal p with respect to S in R and conversely that if p
0
is maximal ideal with respect to S in R, then pR does not meet
0 0
S the the theorem of transition holds for the rings Rs and RS 0 .
0
3. If Rs and RS 0 . in (2) are semi-local rings,then Rs is a subspace
0
of Rs0 .

0 0 T
Proof. We note first that ⊗R R is exact by (3.5.1). Since aR R=
0
a by (3.4.2), the validity of (1) is obvious. ⊗Rs Rs0 is exact by (cor
1) . Therefore, by our assumption and by (3.4.2), Rs is a sub ring of
0 0
Rs0 , hence Rs is dominated by Rs0 . There fore (2) is proved (3.5.1)
00 0 T 00
(3). (3) is easy because a Rs0 Rs = a for any ideal a of Rs , hence
in particular for the Jacobson radical a.

46
References

[1] AUSLANDER, M. And BUCHSBAUM, D. A. Homological


dimension in Local rings, Trans. Am. Math. Soc. 85 (1957), pp.
390 - 405.

[2] AUSLANDER, M. And BUCHSBAUM, D. A. Unique


factorization in regular Local ring, Proc. Nat. Acad. Sci.
U.S. 45 (1959), pp.733-734.

[3] CHEVALLEY, C. On the theory of local rings Ann. Math. 44


(1943), pp. 690-708.

[4] CHEVALLEY, C. On the notion of local ring of quotients of a


prime ideal, Bull. Am. Math. Soc. 50 (1944), pp. 93-97. .

[5] COTTEN, I. H. Commutative rings with restricted minimum


condition, Duke. Math, J. 17 (1960), PP. 27-52.. Commutative
rings with restricted minimum condition, Duke. Math, J. 17
(1960), PP. 27-52.

[6] NAGATA, M. the theory of semi-local rings, Proc. Japan Acad.


26 (1950) pp. 131-140.
47
[7] NAGATA, M. Local rings (in Japanese), Subaru 5 (1953-54) pp.
104-114 and pp. 229-238.

[8] NAGATA, M. Some remarks on local rings, Nagoya Math.J.6


(1953), pp. 53-58.

[9] NAGATA, M. Basic theorems on general commutative rings,


Memo. Coll. Sci., Univ. Kyoto 29(1955), pp. 59-77.

[10] NAGATA, M. On the derived normal ringsof Noetherian integral


domains. Memo. Coll. Sci., Univ. Kyoto 29(1955), pp. 293-303.

[11] NAGATA, M. An example of normal ring which is analytically


ramified, Nagoya Math. J. 9 (1955), pp. 111-11

[12] NAGATA, M. Note on a chain condition for a prime ideals Mem.


Coll. Sci., Univ. Kyoto 32 (1959-60), pp. 85-90.

[13] NARITA, M. On the unique factorization theorem in regular


local rings, Proc. Japan Acad. 35 (1959), pp. 329-331.

[14] WAERDEN B. L. VANDER Modern algebra, II, Grundy. Math.


Wises. 34. Julius Springer. Berlin. 1931 (1st edition); 1940 (2nd
edition); etc.

[15] ZARISKI, O. and SAMUEL, P. Commutative algebra, 1, Van


No strand, New York, 1958.

48

You might also like