Modules along endomorphism rings

1 2 3
Md . Ashraful Alam , Arjuman Ara , Md . Rezaul Karim
1
Department of Mathematics, Bangladesh University of Engineering & Technology, Dhaka, Bangladesh
2
General Education Department, City University, Dhaka, Bangladesh
3
School of Science & Engineering, University of Creative Technology Chittagong, Chattogram, Bangladesh

Corresponding Author: ashraful.mat5@gmail.com

Abstract:

In this paper, all rings are associative with identity and all modules are nonzero unitary right R-modules. We introduce and investigate properties of

modules over endomorphism rings based on established properties that every homomorphism of R from right ideal of R to itself is a left multiplication and

endomorphism of right ideal of R is isomorphic to R. We use the fact that the class of all fully invariant submodules of a right R-module M is non-empty

and closed under intersections and sums when it is known that a right ideal of R is a fully invariant submodule of R to the right if it is an ideal of R.

Keywords: Associative ring; endomorphism ring; left multiplication; fully invariant submodule.

1 Introduction:

Let M R and N R be two right R -modules over a ring R and set of all homomorphism ring of right R -module from M R to N R is written as
Hom R (M , N ). Consider Hom R (M , N ) is additive abelian group of mapping. Furthermore, End R (M , M ) is the set of all
homomorphism ring of right R -module from M R to M R over addition and composition as multiplication mapping. In general, if two right R -
modules are isomorphic then their endomorphism rings End R (M , M ) are isomorphic, but antithetical is not true. The Baer-Kaplansky theorem
states that two reduced torsion abelian group are isomorphic iff their endomorphism rings are isomorphic as rings [1-3]. May and Toubassi explored

whether this correspondence can be extended beyond reduced torsion groups to groups with torsion subgroups and they investigated how the torsion

subgroups of abelian groups influence the structure of the endomorphism ring [4]. The work of Wolfson investigates the relationship between the structure

of torsion-free modules over a complete discrete valuation ring (DVR) and their endomorphism rings [5]. By employing methods similar to those of

Kaplansky [1], Wolfson establishes a theorem that links isomorphisms of the endomorphism rings of such modules with the modules themselves.

Wolfson's Theorem got: “Let M and N be torsion-free modules over a complete discrete valuation ring R . Then each ring isomorphism between
End R (M ) and End R (N ) is induced by a module isomorphism between M and N ." Wolfson divides the proof into two distinct cases based
on the properties of the torsion-free module M where case 1: M is torsion-free but not divisible and case 2: M is torsion-free and divisible. May
extends the classical Baer-Kaplansky theorem to address mixed modules over complete valuation domains [6]. This generalization considers reduced

modules that are neither purely torsion nor torsion-free. Ivanov demonstrates that there exists a class of modules over triangular matrix rings that behaves

analogously to torsion abelian groups in the Baer-Kaplansky theorem [7]. To establish this result, Ivanov builds on key properties of torsion groups: Every
 Z ( p∞ ) .
n n
indecomposable p-group is isomorphic to either Z (p ) (The cyclic group of order p or The endomorphism ring

n ∞
End R (Z ( p )) is isomorphic to Z /( pn ), the ring of integer modulo pn. The endomorphism ring End R (Z ( p )) is isomorphic to p-
adic integers Z p . These results, detailed in works by Fuchs [2] and Kaplansky [1], form the foundation for Ivanov's generalization. Ivanov's introduction
of IP-isomorphism’s (Isomorphism’s that Preserve Indecomposable direct summands) further refines the relationship between endomorphism rings and

module structure. This notion arises in the context of classifying modules with the finite embedding property and highlights how the preservation of

module decomposition under endomorphism ring isomorphism’s guarantees module isomorphism. Breaz in [8] extends the classical Baer-Kaplansky

theorem to a class of modules over principal ideal domains (PIDs). Leveraging the structure of p-adic integers and the isomorphism

End R ( Z ( p ) ) ≅ Z p. This work generalizes the original theorem, traditionally confined to torsion abelian groups, to modules over broader
∞

algebraic settings.

In this paper, a ring will be a ring with associative unity and all modules are unitary right R -module except in special circumstance. Here we used
definition of module, category, ring, and others are used refer to [9].

2 Methodology

In this section we will introduce decomposable direct summand, bi-modules, factor modules and factor rings.

Direct Sum:

Let
M be a right
R -module and let { A i , i∈ I } be a family of submodules of M . Then the sum ∑ A i is called a direct sum if for any
i∈

j ∈I, ∑ A i ⋂ A j=0. In this case, we denote it by ⨁

i ∈I A i . Let { A i , i∈ I } be a family of submodules of M R . Then the
i ∈I , i ≠ j

following conditions are equivalent:

(i) The sum ∑ A i is direct.

i∈

n n n
x=∑ ai , x=∑ ai =∑ a'i ,
(ii) can be expressed as uniquely in the form by the meaning where
x k k k
k=1 k=1 k=1

'
a i , a i ∈ Ai , then
k k k

'
a i =a i , for all a i .
k k k

Bi-modules: Let R , S be two rings and let M be an abelian group. Then M is called R - S bi-module if M is a left R -module, right S-left
module, and if for any r ∈ R , m∈ M , s ∈ S , we have r ( ms )= ( rm ) s .
Factor modules and Factor rings:

Let M
M be a right R -module and let X ⊂ ≻ M . Then the relation xRy ⇔ x− y ∈ X , is an equivalence relation, and we consider ,
R
the quotient set.

M M where M M
Denote
≡ , = { m=m+ X|m ∈ M } . On , we defined
X R X X

M
m+m '=m+m ' and m r=mr ∀ m, m' ∈ M , ∀ r ∈ R .Then becomes a right
R -module, called a factor module of M
X
by X

In this section we will solve some problem which are related to above definition.

2.1 Let S= End R (M ) be the set of all R - homomorphism from M R to M R . Prove that with the two operations +¿ and ∘ , for any
f , g ∈ S , and for any

x ∈ M : ( f + g ) ( x )=f ( x ) + g(x ) and ( f ∘ g ) ( x ) =f ( g ( x ) ) . Then S becomes an associative ring with identity. Finally, M
becomes an S- R -bimodule.

Proof: Denote Hom R ( M , N )= { f : M → N|f is an R -homorphism}. We must prove that Hom R ( M , N ) is an abelian group
with the binary operation addition. Then for any

f , g ∈ Hom R ( M , N ) , ( f + g )( x )=f ( x )+ g ( x ) . So addition is well defined.

Now for any f , g ∈ Hom R ( M , N ) , consider the map

φ : Hom R ( M , N ) × Hom R ( M , N ) → Hom R ( M , N ) is defined by

( f , g ) ↦ f + g , ( f + g ) ( x )=f ( x ) + g ( x ) . Then the map φ is well defined because for any

( f , g ) ∈ Hom R ( M , N ) × Hom R ( M , N ) , we have only one f (x) and only one g(x ) and +¿ is the addition on N , so
(f + g)(x) is well defined. Therefore,

( f , g ) ↦ f + g is defined.
Also, ( f , g ) =( f ' , g ' ) ⇒ f =f ' , g=g' ⇒ f ( x )=f ' ( x) and

g ( x )=g' ( x ) ∀ x ∈ M ⇒ ( f + g ) ( x )= ( f ' + g' ) ( x ) ⇒ f + g=f ' + g ' . So, the map φ is well defined.

Moreover, f +g is an R -homomorphism. For any x , y ∈M and r∈R, we have (by definition)

( f +g ) ( x + y ) =f ( x + y ) + g(x + y )=f ( x ) +f ( y ) + g ( x ) + g( y ) (because f ,g are R -homomorphism’s)

¿ ( f ( x ) + g ( x ) ) + ( f ( y )) + g ( y )¿because +¿ is commutative in N ¿= ( f + g ) ( x )+(f + g)( y) (by definition).

Also, ( f +g ) ( xr )=f ( xr )+ g ( xr ) =f ( x ) r+ g ( x)r (because f , g are R -homomorphism’s ¿ [f ( x ) + g(x )]r (because

N is a right R -module)

¿ [(f + g)(x )]r (by definition). So f +g ∈ Hom R ( M , N ) .

Take any f , g ∈ Hom R ( M , N ) . Then for any x ∈ M , ( f + g ) ( x ) =f ( x ) + g(x ) (by denition) ¿ g ( x )+ f (x ) (because
N is abelian ) (g+ f )(x) (by definition). Hence

f +g=g+ f .

The addition on Hom R ( M , N ) is associative. In fact, let f , g , h ∈ Hom R ( M , N ) . Then for any

x ∈ M , [ ( f + g ) +h ] ( x )=( f +g ) ( x ) +h ( x ) (by definition)

¿ [ f ( x ) + g ( x ) ] + h(x) (by definition) ¿ f ( x ) +[g ( x ) +h( x )] (because N is an abelian group) f ( x ) +(g +h)(x)(by
definition) ¿ [f +(g+ h)](x) (by definition). Hence ( f +g ) +h=f +(g+h).

Define O:M→N by O ( x )=0 for any x∈ M . We can see that O is an R -homomorphism. For any

( f +O )( x )=f ( x )+ O ( x )=f ( x )+ 0=f ( x ) .

Hence f +O=O+ f =f . So O is the zero element of Hom R ( M , N ) . We denote O by 0.

Let f ∈ HomR ( M , N ) . Define −f as follows:

for any x ∈ M , (−f ) ( x ) =−f ( x ) . Then −f ∈ HomR ( M , N ) . We have

[ f + (−f ) ] ( x )=f ( x ) + (−f ) ( x )=f ( x )−f ( x )=0=O (x) for any x ∈ M . Hence f + (−f )=O . This shows that
Hom R ( M , N ) is an abelian group.

Let S= End R ( M ) =HomR ( M , M ) and f , g ∈ S . Then for any

x ∈ M , ( f + g ) ( x ) =f ( x ) + g(x ) and ( f ∘ g ) ( x ) =f (g (x)). We must prove that S is an associative ring with identity.

(a) We can see that Hom R ( M , N ) is an abelian group. Take f , g ∈ S . Then f ∘ g∈ S .

(b) For any f , g , h ∈ S , ( f ∘ g ) ∘ h=f ∘(g ∘ h) and f ∘ ( g+h ) =f ∘ g+ f ∘h ,
( g+h ) ∘ f =g ∘ f + h∘ f .
(c) Finally, We must prove that if 1 : M → M , then 1 M ∈ S and for any f ∈ S ,we can write f ∘ 1=1 ∘ f =f . Since f , g
maps from M to M , so is f ∘ g and
∀ x, y ∈M ,r∈ R, we have (by definition) ( f ∘ g ) ( x + y ) =f ( g ( x+ y ) )
¿=¿=f ( g ( x )+ g ( y ) ) =f ( g ( x ) ) +f ( g ( y ) ) =( f ∘ g )( x ) +(f ∘ g)( y ) and

( f ∘ g ) ( xr )=f ( g ( xr )) =f ( g ( x ) r )=f ( g ( x ) ) r =[ ( fog ) ( x ) r ] . Hence f ∘ g∈ S .

In fact f , g , h ∈ S . Then for any x ∈ M , we have (by definition)

[ ( f ∘ g ) ∘ h ] ( x )=( f ∘ g ) ( h ( x ) ) =f ( g ( h ( x ) ) ) =f ( ( g ∘ f )( x ) )=[ f ∘ ( g ∘h ) ] ( x ) .

Hence f ∘ ( g+h ) =f ∘ g+ f ∘h . Similarly, for any x ∈ M , we have

[ ( g+h ) ∘ f ] ( x )=( g+h ) ( f ( x ) ) =g ( f ( x ) ) +h ( f ( x ) ) =( g ∘ f )( x ) + ( h ∘ f ) ( x ) .

Hence ( g+h ) ∘ f =g ∘ f + f ∘h .

Let 1 : M → M , then 1 M ∈ S and for any f ∈ S , we have f ∘ 1M =1M ∘ f =f ,

1 M ( x + y )=x + y=1 M ( x )+1 M ( y ) and 1 M ( xr ) =xr=1 M ( x ) r . Hence 1M ∈ S . Also,

(f ∘ 1¿¿ M ) ( x )=f ( 1 M ( x )) =f (x) ¿ implies that f ∘ 1M =f and

(1¿¿ M ∘ f ) ( x )=1M (f ( x ))=f ( x )¿ implies that 1 M ∘ f =f . This ring S is not commutative in general.
We finally show that M is an S- R -bimodule. Let M R be a right R -module. We first show that M is a left S-module.

Define ψ :M ×S→ M by ψ ( mα )=α ( m ) . Then for any α ,β ∈S and for any

'
m,m ∈M , we have

α ( m+ m' ) =α ( m+m' )=α ( m )+ α ( m' )=αm+αm ' because α ∈ S . Also,

( α + β ) . m=( α + β ) ( m )=α ( m ) + β ( m )=α . m+ β . m and 1 M . m=1 M ( m )=m (unitary). Now,

∀ α ∈ S , ∀ m∈ M , ∀ r ∈ R , we have α ( mr ) =( α . m ) r=α ( m) r .

2.2 Let M be a right R -module and let S= End R ( M ) , the endomorphism ring of M R . Prove that a submodule A of M is a direct

summand of M if and only if there exist a idempotent f ∈ S (i.e. f 2=f ) such that A=f ( M ) .

Proof: Assume that A is a direct summand of M . Then there exists a submodule B⊆ M such that M = A ⨁ B , which means every
m ∈ M can be uniquely written as

m=a+b where a ∈ A and b ∈ B .

Define f : M → M by f ( m) =a+b with a ∈ A and b ∈ B . Since f ( m) ∈ A , So

ℑ ( f )= A . To check f is R -homomorphism: ∀ m1 , m2 ∈ M , ∀ a1 , a2 ∈ A , ∀ b1 ,b 2 ∈ M

we have f ( m1 +m2 )=f ( ( a1+ b1 ) + ( a2 +b2 ) ) =f ( ( a1 +a 2 ) + ( b1 +b 2 ) )

¿ a 1+ a2=f ( m1 ) + f ( m2 ) and f ( mr ) =f ( ( a+ b ) r )=f ( ar +br )=ar =f ( m ) r Since A and B are submodules. We

f =f . Take m ∈ M . Then
2
now show that

m=a+b , by hypothesis where a ∈ A , b∈ B . Also we have f ( m) =a . So,

f ( m )=f ( f ( m ) ) =f ( a )=a=f ( m ) , because a=a+ 0. Thus f 2=f .

2

Conversely, suppose f is an idempotent, so

2
f =f ∈ S and A=f ( M ) . We must show that A is a direct summand of M . Define
B=Ker ( f )={ m∈ M |f ( m )=0 } . We claim that M = A ⨁ B . Now, we check direct sum property: For any m ∈ M , let
a=f (m) and b=m−f ( m ) . Then: m=a+b , where a=f ( m ) ∈ A , and b=m−f ( m ) ∈ B

(since f ( b )=f ( m−f ( m ) ) =f ( m )−f ( m )=0 ¿ .

Uniqueness: If m=a+b with a ∈ A and b ∈ B , then:

f ( m) =f ( a+ b )=f ( a ) + f ( b )=a+0=a . Thus a=f (m) and

b=m−f ( m ) . Showing uniqueness.

Now, we will check intersection property A ∩ B= {0 } : If x ∈ A ∩B , then

x ∈ A=f ( M ) and x ∈ B=Ker ( f ) . Thus x=f ( x ) =0.

Therefore, M = A ⨁ B and A is a direct summand of M .

3 Results & Discussion

3.1 A submodule X of M R is called fully invariant if for any f ∈ S=End R ( M ) , we have f ( X ) ⊂ X . Let M be a right R -module
and let S= End R ( M ) , the endomorphism ring of M R . Prove that an abelian subgroup X of M R is a bi-submodule of S- R -bimodule if
and only if X is a fully invariant submodule of M R , Where M R is called a duo module if every submodule is fully invariant.

Proof:

First part, if X is a bi-submodule of S- R -bimodule, then we need to prove that X is fully invariant.

If X ⊂ ≻ M R ,then X ∈ Mod - R and the endomorphism ring S= End R ( M ) has a natural left action on M , making M as an S-
R bimodule. A submodule X of M is called a bi-submodule if it is a submodule of M both as a right R -module and as a left S-module. This
means sX ⊆ X for all s ∈ S . By definition of S= End R , each s ∈ S is a right R -module homomorphism s : M → M . Since X
is called under left multiplication by element of S , we have f (X )⊆ X for every f ∈ S . This exactly matches the definition of a fully invariant
submodule of M R : A submodule X of M R is called fully invariant if for any

f ∈ S=End R ( M ) , we have f ( X ) ⊂ X . Proving that X is fully invariant.

Second part, if X is fully invariant, then we have to show that X is bi-submodule of

S- R -bimodule.
Suppose, X is fully invariant, meaning that f (X )⊆ X for all f ∈ S=End R ( M ) . By definition of a bi-submodule, we need to show that
X is closed under the left action of S , i.e., sX ⊆ X for all s ∈ S . But this is precisely the definition of full invariance: for every s ∈ S , we
have s ( X ) ⊆ X . Since X is a already a right submodule (as a submodule of M R ,), it follows that X is a bi-submodule of the S- R -bimodule.

3.2 Let M be a right R -module and let S= End R ( M ) , the endomorphism ring of M R . Prove that endomorphism of a right R -module
isomorphic to R and endomorphism of a left R -module also isomorphic to R .

Proof: Case 1:

Consider S consists all R -module endomorphism as a right R -module. That is:

S= { f : R → R|f ( x + y )=f ( x ) + f ( y ) , f ( xr )=f ( x ) r ∀ x , y ∈ R ,r ∈ R } . For any

f ∈ S , consider a=f ( 1 ) ∈ R , where 1 is the multiplicative identity of R . We claim that,

f ( x )=ax ∀ x ∈ R .

To check the following property:

Linearity of f : Since f is R -linear,

so f ( x + y )=f ( x )+ f ( y) and f ( xr )=f ( x ) r ∀ x , y ∈ R ,r ∈ R .

Behavior on arbitrary x ∈ R : Let x ∈ R . We can write x=1. x . Then:

f ( x )=f ( 1. x )=f ( 1 ) x =ax . Thus, f is determined entirely by a=f ( 1 ) .

The correspondence f ↔ a : Define a map ϕ :S → R by ϕ ( f )=f ( 1 ) . This map is: Injective: If ϕ ( f )=ϕ ( g ) ,then f ( 1 ) =g (1 ) .
For any x ∈ R , we have

f ( x )=f ( 1 ) x=g ( 1 ) x=g ( x ) , so f =g .

Surjective: For any a ∈ R , define f a : R → R by f a ( x ) =ax . This f a ∈ S satisfies

ϕ ( f a )=a . Hence ϕ is bijection. This implies that S is isomorphic to R .

Case II:

Since S consists all R -module endomorphism as a left R -module. That is:

S= { f : R → R|f ( x + y )=f ( x ) + f ( y ) , f ( rx )=rf ( x ) ∀ x , y ∈ R ,r ∈ R } . For any

f ∈ S , consider a=f ( 1 ) ∈ R . We claim that, f ( x )=xa ∀ x ∈ R .

To check the following property:

Linearity of f : Since f is R -linear,

so f ( x + y )=f ( x )+ f ( y) and f ( rx )=rf ( x ) ∀ x , y ∈ R ,r ∈ R .

Behavior on arbitrary x ∈ R : Let x ∈ R . We can write x=x .1. Then:

f ( x )=f ( x .1 )= xf ( 1 ) =xa . Thus, f is determined entirely by a=f ( 1 ) .

The correspondence f ↔ a : Define a map ϕ :S → R by ϕ ( f )=f ( 1 ) . This map is: Injective: If ϕ ( f )=ϕ ( g ) ,then f ( 1 ) =g (1 ) .
For any x ∈ R , we have

f ( x )=xf ( 1 ) =xg ( 1 )=g ( x ) , so f =g .

Surjective: For any a ∈ R , define f a : R → R by f a ( x ) =xa . This f a ∈ S satisfies

ϕ ( f a )=a . Hence ϕ is bijection.

Therefore S is isomorphic to R.

In this section we will introduce von Neumann regular ring, M -cyclic submodule and solve some exercise.

Von Neumann regular ring:

A ring R is called a von Neumann regular ring if every principal right ideal generated by an idempotent. This means that for any a∈R , there is an

2
idempotent e=e ∈ R such that aR=eR . This also means that every principal right ideal is a direct summand.
M -cyclic submodule:

A submodule X of M is called an M -cyclic submodule if X =s (M ) for some

s ∈ S=End R ( M ) .

3.3 A ring R is von Neumann regular if and only if every finitely generated right ideal is a direct summand if and only if every finitely generated left
ideal is a direct summand.

Proof: We prove this by showing the equivalences:

i. R is von Neumann regular ⟺ Every finitely generated right ideal is a direct summand.
ii. R is von Neumann regular ⟺ Every finitely generated left ideal is a direct summand.

Since the definitions of right and left ideals are symmetric, proving one direction will imply the other:

Step 1: Suppose R is von Neumann regular. Show that every finitely generated right ideal is a direct summand. By definition R is a von Neumann regu-
lar if for every a ∈ R there exists x ∈ R such that a=axa . Consider a finitely generated right ideal I of R , say

I =R a1 + R a 2+ …+ R an .

For each a i , there exist x i such that a i=ai x i a i .

Define f :I →I by: f ( ∑ r i ai ) =∑ r i ai x i . This is a right R -module homomorphism satisfying f ( ai )=ai x i . Since

a i=ai x i a i , it follows that f ( f ( a i ) )=f ( a i) , making f an idempotent map. The image of f , denoted J=ℑ(f ) is a direct summand of

I , meaning that I =J ⨁ Ker (f ). Thus, every finitely generated right ideal is a direct summand.

Step 2: Suppose every finitely generated right ideal is a direct summand, show that R is von Neumann regular.

Consider a principal right ideal I =Ra for some a ∈ R . By assumption, I is a direct summand of R , so there exists another right ideal J such
that R=Ra ⊕ J . This means there exists an idempotent module homomorphism f : R → Ra such that f ( a )=a . Since f is right R -lin-
ear, there exists x ∈ R such that f ( r )=rx for all r ∈ R and specifically: a=ax . Applying this to ax , we get: ax=axx . So setting
x '=x , we get a=ax ' a , proving that R is von Neumann regular.

The proof for left ideals follows the same reasoning, showing that every finitely generated left ideal is a direct summand if and only if R is von Neumann
regular.

3.4 Let M be a right R -module and let S= End R ( M ) , the endomorphism ring of M R . if S is von Neumann regular, then every M -
cyclic submodule of M is a direct summand of M .

Proof: Step 1: Understanding the Given Conditions

The endomorphism ring S= End R ( M ) consists of all R -module homomorphism’s

f :M →M .

S is von Neumann regular, meaning that for every f ∈ S , there exists g ∈ S such that: f =fgf .

An M -cyclic submodule is a submodule of M of the form mS for some m ∈ M . This means: mS={ms ∣ s ∈ S }.

Step 2: Constructing an Idempotent Projection

Consider an M -cyclic submodule X =mS , where m ∈ M . Define f : M → M by: f ( ms )=ms ∀ s ∈ S . This is a well-defined
S-module homomorphism. This equation implies that f is idempotent: f ( f ( m ) )=f ( m ) . The image of an idempotent endomorphism in a mod-
ule is a direct summand. Since f is idempotent and projects onto X , we can write M = X ⨁ Ker ( f ) . This shows that every M -cyclic sub-
module is a direct summand of M.

Thus, we have proved that if S= End R ( M ) is von Neumann regular, then every M -cyclic submodule is a direct summand of M .

4 Conclusion:

In this section we will prove some theory related to discussion on R -module homomorphism, Cyclic module, finitely generated module, idempotent mod-
ule and them.

Ring homomorphism:
Let M R , N R be two right R -modules. Then a map f : M → N is called an R -homomorphism if for any m , m' ∈ M and any r ∈ R ,
We have

f ( m+m ' ) =f ( m ) +f ( m ' ) , f ( mr )=f ( m) r .Then map f is isomorphism if it is one-one and onto. The set

Ker ( f )={m∈ M ∨f ( m) =0 } is called the kernel of f and the set ℑ ( f )= { f ( m )∨m ∈ M } is called the image of f .

Cyclic modules:

Let M be a right R -module and let m ∈ M . Then the set

mR={mr∨r ∈ R } is a submodule of M , called the cyclic submodule of M . We can see that M = ∑ mR . A right R -module
m∈ M

M is called a cyclic right R -module if M is generated by a single element, i.e., there exists m ∈ M such that M =mR .

Finitely generated module: A right R -module M is called a finitely generated right

R -module if M is generated by a finite number of elements, i.e., there exists

k
M =|{ m1 , m2 , … . , mk } ) =∑ mi R .
such that
m1 ,m2 , … ., mk ∈ M
i=1

A cyclic right ideal is called a principal right ideal.

Idempotent module: Let M be a right R -module and e be an idempotent element in End R ( M ) . Then 1−e ∈ End R ( M ) . Furthermore,
e and 1−e orthogonal and M has decomposition M =eM ⨁ (1−e ) M . Lemma 2 will give description of idempotent element e .

4.1 Theorem: Let be an -homomorphism. Then image of denoted as is isomorphic to M where

f : M R→ N R R f, ℑ( f ) ,
ker ( f )
ker ( f )= { m∈ M |f ( m )=0 } .

Proof: Define quotient map:

The Kernel of f , denoted as ker ( f ) ⊂ ≻ M .

Define the canonical projection map M by
ϕ:M→ ϕ ( m )=m+ker ( f ) .
ker (f )

This is an R -module homomorphism.

Define a Well-Defined Map on the Quotient:

Define a function M
f: → ℑ(f ) by f ( m+ ker ( f ) ) =f ( m ) .
ker (f )

Well defined: If m+ker ( f )=m' +ker ( f ) , then m−m' ∈ ker ( f ) , so

f ( m−m' ) =0 , which gives f ( m) =f ( m' ) , meaning f is well defined.

Homomorphism: For any a , b ∈ M and r ∈ R , we have

f ( ( a+b )+ ker ( f ) )=f ( a+b )=f ( a )+ f ( b )

¿ f ( a+ker ( f ) ) + f ( b+ker ( f ) ).

Similarly for scalar multiplication:

f ( ( mr )+ ker ( f ) ) =f ( mr )=f ( m ) r =f ( m+ ker ( f ) ) r .

Thus, f is an R -module homomorphism.

Now, we need to prove f is an isomorphism.

Injectivity: Suppose f ( m+ ker ( f ) ) =0 , meaning f ( m) =0. By definition,

m ∈ker ( f ) , so m+ker ( f )=0 in the quotient module. Hence, f is injective.

Surjectivity: By definition, the image of f is ℑ(f ). Thus, f is surjective.

Since f is a bijective R-module homomorphism, it is an isomorphism.

4.2 Theorem: Let M R be a right module, and let X ,Y be submodules of M such that X ⊂Y , then there exists an isomorphism:

M
≅
( X)
M
.
Y
(X)
Y

Proof: Define the Natural Projection Maps:

Consider the canonical projection M be defined by

ϕx: M → ϕ x ( m )=m+ X .
X

This is an R -module homomorphism with kernel X , so ker ( ϕ x )= X .

Similarly, we define the canonical projection M be defined by

ϕY : M → ϕ Y ( m) =m+ Y .
Y

This is an R -module homomorphism with kernel Y , so ker ( ϕ Y ) =Y .

Define the Induced Map on the Quotient: Since we can define a quotient module M where
X Y M
X ⊆Y , , ={ y + X| y ∈ Y } ⊆ .
Y X X
X

We define a function: by
φ : M /Y →(M / X )/(Y / X) φ ( m+Y ) =( m+ X ) + ( YX ) .
This function takes elements of maps them into quotient M
X
M /Y .
Y
X

Verify φ is well defined:

If
' '
m+Y =m +Y , then m−m ∈ Y , so m+ X =m + X
'
(in M / X ).

This implies
( m+ X ) + ( YX )= (m + X )+( YX ).
'
Thus, φ ( m+Y ) =φ ( m' + Y ) , proving φ is well defined.

Verify φ is an R -module homomorphism:

For any m , m' ∈ M and r ∈ R , we have addition property:

φ ( ( m+ m' ) +Y )=( ( m+m' ) + X ) + ( YX )= ( m+ X ) +(m + X )+( YX ) .

'

Since (M / X )/(Y / X) is a module, this equals: φ ( m+Y ) + φ ( m' +Y ) .

Hence, φ preserves addition.

Scalar Multiplication:

φ ( ( mr ) +Y ) =( ( mr ) + X ) + ( YX )=( m+ X ) r +( YX ).
Since module operations respect quotients, this equals: φ (m+ Y )r .

Hence, φ preserves scalar multiplication. Thus φ is an R -module homomorphism.

To prove φ is injective: Suppose φ ( m+Y ) =0 in (M / X )/(Y / X).

This means that Y i.e.,

m+ X ∈ , m+ X = y + X for some y ∈Y .
X

This implies m− y ∈ X , meaning that m ∈ y + X ⊆ Y .

Thus, M proving that is injective.

m+Y =0∈ , φ
Y

To prove φ is surjective: Every element in (M / X )/(Y / X) is of the form ( m+ X ) +(Y / X )

for some
m∈M .
But by the definition,
φ ( m+Y ) =( m+ X ) + ( YX ) . Thus every element of
(M / X )/(Y / X)
has a preimage

under φ , proving surjectivity.

Since

φ
is bijective module homomorphism, we conclude

M
≅
( X)
M
.
Y
(X)
Y

4.3 Theorem: Let X , Y be submodules of a right R -module M . Then the following module isomorphism’s holds:

( X +Y ) X and ( X +Y ) Y
≅ ≅ .
Y (X ∩Y ) X (X ∩Y )

Proof:

Case 1: Define the Homomorphism: Consider the canonical projection map:

φ : X →( X +Y )/Y be define by φ ( x )=( x +Y ) .

This map takes elements of X and sends them to their cosets in (X +Y )/Y .

Check Well Defined-ness: If x 1+ Y =x 2+ Y , so φ is well defined.

Check Homomorphism Property: For any x , x ' ∈ X and r ∈ R , We check

φ ( x + x ' )=( x + x ' ) +Y =( x +Y ) + ( x ' +Y )=φ ( x )+ φ ( x ' ) and

φ ( xr )=( xr ) +Y =φ ( x ) r .

Thus, φ is an R -module homomorphism.

Check Kernel: The kernel of φ consists of all x ∈ X such that φ ( x )=( x +Y ) =0+Y .

This means x ∈ Y , i.e., x ∈ ( X ∩Y ) . Hence, ker ( φ )= X ∩Y .

Check Surjective: Since every element (X +Y )/Y is of the form ( x + y ) +Y =x +Y , where x ∈ X , the mapφ is surjective.

Since the First Isomorphism Theorem gives: X (X +Y )

ker ( φ )= X ∩Y , ≅ .
( X ∩Y ) Y
Thus, we have proved: ( X +Y ) X
≅ .
Y X ∩Y

Case 2: Define the Homomorphism: Consider the canonical projection map:

ψ :Y →( X +Y )/ X be define by ψ ( y )=( y + X ) .

This map takes elements of Y and sends them to their cosets in (X +Y )/ X .

Check Well Defined-ness: If y 1 + X = y 2+ X , then y 1= y 2 . so ψ is well defined.

Check Homomorphism Property: For any y , y ' ∈ Y and r ∈ R , We check

ψ ( y + y ' )=( y + y ' ) + X=( y + X )+ ( y ' + X )=ψ ( y ) +ψ ( y ' ) and

ψ ( yr )=( yr )+ X=ψ ( y ) r .

Thus, ψ is an R -module homomorphism.

Check Kernel: The kernel of ψ consists of all y ∈Y such that ψ ( y )=( y + X )=0+ X .

This means y ∈ X , i.e., y ∈ ( X ∩Y ) . Hence, ker ( ψ ) =X ∩Y .

Check Surjective: Since every element (X +Y )/ X is of the form ( x + y ) + X = y+ X , where y ∈Y , the mapψ is surjective.

By the First Isomorphism Theorem, we get Y ( X +Y )

≅ .
(X ∩Y ) X

Thus we have proved: ( X +Y ) Y

≅ .
X (X ∩Y )

4.4 Lemma: Let M R be a finitely generated module.

(i) Then for any submodule M

X of M , written as X ⊂ ≻ M , prove that factor module is also finitely generated.
X
 (ii) If M
X and are finitely generated then so is
M.
X

Proof: Case i: Let M be a finitely generated right R -module. Then there exists a finite set of generator {m1 , m2 , … . , mn} such that

M =SpanR {m1 , m2 , … . , mn }.

That is, every element of M can be written as: m=r 1 m1+ r 2 m2 +…+ r n mn , for some r i ∈ R . Since X ⊂ ≻ M , and we need
to show that M / X is finitely generated.

Consider the cosets of the generators M i.e.,

m1 ,m2 , … ., mn in ,
X

M
m1 + X , m2 + X , …. , mn + X . We claim that this element generate .
X

Generating any element of M Take any element M Since

: m+ X ∈ . M is generated by
X X

{m1 , m2 , … . , mn} , we can write m=r 1 m1+ r 2 m2 +…+ r n mn , for some r i ∈ R .

M
Now, considering the coset of
m in
:m+ X=( r 1 m1 +r 2 m2+ …+r n mn )+ X . Using coset properties:
X
M
m+ X =r 1 ( m1 + X ) +r 2 ( m2+ X ) +…+ r n ( mn + X ) . Thus every element of can be expressed as
R -linear combination of
X
m1 + X , m2 + X , …. ,

M
mn + X , meaning that these cosets generate .
X

Since M we conclude that M

{m1 + X , m2 + X , …. , mn + X } is a finite generating set for , is finitely generated as
R -module.
X X
Hence, we have proved that if M.
M is finitely generated, then is
X

Case ii: Since M is finite generated by

{m1 + X , m2 + X , …. ,
mn + X }, every element
m∈M can be written as:
X
m=r 1 m1+ r 2 m2 +…+ r n mn + x , for some r i ∈ R and
x ∈ X . Since X is generated by { x 1 , x 2 , … ., x k } , we write:

x=s1 x 1+ s 2 x 2+ …+ sk x k for some s j ∈ R . Substituting back, we get

m=r 1 m1+ r 2 m2 +…+ r n mn + s1 x 1+ s 2 x 2+ …+ sk x k .

This shows that M is generated by the finite set:{m1 , m2 , … . , mn , x 1 , x 2 , … . , x k }.

Since this is a finite generating set for M , we conclude that M is finitely generated.

4.5 Lemma: Let M be a right R -module and let e be a nonzero idempotent element in End R ( M ) , meaning e 2=e . Then the direct summand
eM is decomposable if and only if e is the sum of two nonzero orthogonal idempotent in End R ( M ) .

That is, eM =A ⨁ B (for some nonzero submodules A , B) if and only if e=e1 +e 2 with
2
e 1=e 1 ,
2
e 2=e 2 , e 1 e2=e2 e1 =0.

Proof: We have to prove both directions of the equivalence.

Suppose eM is decomposable, so we can write it as a direct sum: eM =A ⨁ B for some nonzero submodules A,B. Consider the

projections e 1 : eM → eM onto A along B and e 2 : eM → eM onto B along A .

2 2
These projections satisfy: e 1=e 1 , e 2=e 2 , e 1 e2=e2 e1 =0.

Also since eM =A ⨁ B , we have e 1+ e 2=e . Clearly e 1 ,e 2 ∈ End R ( M ) .

Thus, e decomposes into the sum of two nonzero orthogonal idempotents e 1 ,e 2.

Conversely, e can decomposes as e=e1 +e 2, where e 1 ,e 2 are orthogonal idempotent.

Consider the submodules A=e 1 M , B=e2 M . Since e 1 e2=e2 e1 =0 , we have

e 1 M ∩e 2 M =0.

Every element in eM can be written as:

em= ( e1 + e2 ) m=e 1 m+e 2 m∈ e1 M + e2 M .

This shows that eM =A ⨁ B , proving that eM is decomposable.

Acknowledgement: Author must acknowledge to Professor Dr Khandker Farid Uddin Ahmed, Department of Mathematics, BUET for his valuable advice

and suggestion along this research.

Conflict of Interest: All authors state that there is no conflict of interest.

References:

[1] Kaplansky, I., 1954. Infinite Abelian Group. University of Michigan Press, Ann Arbor.

[2] Fuchs, L., 1973. Infinite Abelian Groups Vol II. Academic Press, New York.

[3] Faticony, T. G., 1991. A new proof of the Bayer-Kaplansky Theorem. Communication in Algebra, vol 19, pp. 3119-3123.

[4] May, W. and Toubassi, E., 1976. Isomorphisms abelian groups and the theorem of Bayer and Kaplansky. Journal of Algebra, Vol-43, pp. 1-13.

[5] Wolfson, K. G., 1962. Isomorphisms of the endomorphism rings of torsion-free modules. Proceeding of the mathematical society, vol 13, pp. 712-714.

[6] May, W., 1995. The Theorem of Bayer and Kaplansky for Mixed Modules, Journal of Algebra, Vol-177, pp. 255-263.

[8] Ivanov, G., 1998. Generalizing the Bayer-Kaplansky theorem for modules over principal ideal domains. Journal of Pure and Applied Algebra, Vol 133,

pp 107-115.

[9] Anderson, F. W. and Fuller, K. R., 1992. Rings and Categories of Modules. Springer-Verlag, New York.