• 検索結果がありません。

IDENTITIES WITH DERIVATIONS AND AUTOMORPHISMS ON SEMIPRIME RINGS

N/A
N/A
Protected

Academic year: 2022

シェア "IDENTITIES WITH DERIVATIONS AND AUTOMORPHISMS ON SEMIPRIME RINGS"

Copied!
8
0
0

読み込み中.... (全文を見る)

全文

(1)

ON SEMIPRIME RINGS

JOSO VUKMAN

Received 31 May 2004 and in revised form 17 January 2005

The purpose of this paper is to investigate identities with derivations and automorphisms on semiprime rings. A classical result of Posner states that the existence of a nonzero centralizing derivation on a prime ring forces the ring to be commutative. Mayne proved that in case there exists a nontrivial centralizing automorphism on a prime ring, then the ring is commutative. In this paper, some results related to Posner’s theorem as well as to Mayne’s theorem are proved.

We investigate identities with derivations and automorphisms on semiprime rings. We prove, for example, that in case there exist a derivationD:RRand an automorphism α:RR, whereRis a 2-torsion-free semiprime ring, such that [D(x)x+xα(x),x]=0 holds for allxR, thenDandαI, whereIdenotes the identity mapping, mapRinto its center. Throughout this paper,Rwill represent an associative ring with centerZ(R).

A ringRis 2-torsion-free in case 2x=0 implies thatx=0 for anyxR. As usual, we write [x,y] forxyyxand make use of the commutator identities [xy,z]=[x,z]y+ x[y,z], [x,yz]=[x,y]z+y[x,z],x,y,zR. We denote byI the identity mapping of a ringR. An additive mappingD:RRis called a derivation ifD(xy)=D(x)y+xD(y) holds for all pairsx,yR. Let αbe an automorphism of a ring R. An additive map- pingD:RRis called anα-derivation ifD(xy)=D(x)α(y) +xD(y) holds for all pairs x,yR.Note that the mappingD=αI is anα-derivation. Of course, the concept of α-derivation generalizes the concept of derivation, sinceI-derivation is a derivation. We denote byCthe extended centroid of a semiprime ringRand byQMartindale ring of quotients. For the explanation of the extended centroid of a semiprime ringRand the Martindale ring of quotients, we refer the reader to [1]. A mapping f ofRinto itself is called centralizing onRif [f(x),x]Z(R) holds for allxR; in the special case when [f(x),x]=0 holds for allxR, the mappingf is said to be commuting onR. The history of commuting and centralizing mappings goes back to 1955 when Divinsky [5] proved that a simple Artinian ring is commutative if it has a commuting nontrivial automor- phism. Two years later, Posner [8] has proved that the existence of a nonzero centralizing derivation on a prime ring forces the ring to be commutative (Posner’s second theorem).

Luh [6] generalized the Divinsky result, we have just mentioned above, to arbitrary prime

Copyright©2005 Hindawi Publishing Corporation

International Journal of Mathematics and Mathematical Sciences 2005:7 (2005) 1031–1038 DOI:10.1155/IJMMS.2005.1031

(2)

rings. Mayne [7] proved that in case there exists a nontrivial centralizing automorphism on a prime ring, then the ring is commutative (Mayne’s theorem). A result of Breˇsar [2], which states that every additive commuting mapping of a prime ringRis of the form xλx+ζ(x) whereλis an element ofCandζ:RCis an additive mapping, should be mentioned. A mappingf :RRis called skew-centralizing onRiff(x)x+x f(x)Z(R) holds for allxR; in particular, if f(x)x+x f(x)=0 holds for allxR, then it is called skew-commuting onR.Breˇsar [3] has proved that ifRis a 2-torsion-free semiprime ring and f :RRis an additive skew-commuting mapping onR, thenf =0.

First, we list three lemmas which will be needed in the sequel.

Lemma1 [11, Lemma 1]. LetRbe a semiprime ring. Suppose that the relationaxb+bxc=0 holds for allxRand somea,b,cR. In this case,(a+c)xb=0is satisfied for allxR.

Lemma2 [12, Lemma 1.3]. LetRbe a semiprime ring. Suppose that there existsaRsuch thata[x,y]=0holds for all pairsx,yR.In this case,aZ(R).

Lemma3. LetRbe a semiprime ring and let f :RRbe an additive mapping. If either f(x)x=0orx f(x)=0holds for allxR, thenf =0.

Proof. We can restrict our attention to the case

f(x)x=0, xR, (1)

because of left-right symmetry. The linearization of the above relation gives

f(x)y+f(y)x=0, x,yR. (2) The substitution ofy2foryin the above relation gives

f(x)y2+ fy2x=0, x,yR. (3) Right multiplication of (2) byygives

f(x)y2+f(y)xy=0, x,yR. (4)

Subtracting (4) from (3), we obtain

fy2xf(y)xy=0, x,yR. (5)

Putting in the above relationx f(y) forx, we obtain, because of (1),

fy2x f(y)=0, x,yR. (6) Right multiplication of (5) by f(y) gives, because of the above relation,f(y)xy f(y)= 0,x,yR, which leads toy f(y)xy f(y)=0,x,yR, whence it follows that

x f(x)=0, xR. (7)

(3)

Right multiplication of the relation (2) by f(x) gives, because of the above relation, f(x)y f(x)=0,x,yR, whence it follows thatf =0, which completes the proof.

We are ready for our first result.

Theorem4. LetRbe a semiprime ring. Suppose that there exist a derivationD:RRand an automorphismα:RR,such that the mappingxD(x) +α(x)is commuting onR.In this case,DandαImapRintoZ(R).

At the end of the proof of the result above, we will need the result below.

Theorem5 [9, Proposition 2.3]. LetRbe a semiprime ring and letD:RRbe a com- mutingα-derivation onR.In this case,DmapsRintoZ(R).

We will need also the result below which is a special case of [2, Proposition 3.1].

Theorem 6. Let Rbe a 2-torsion-free semiprime ring and let f :RR be an additive centralizing mapping onR. In this case, f is commuting onR.

Proof ofTheorem 4. The linearization of the relation

D(x) +α(x),x=0, xR (8) gives

D(x) +α(x),y+D(y) +α(y),x=0, x,yR. (9) Putting in the above relationyxforyand applying the relation (8), we obtain

0=

D(x) +α(x),yx+D(y)x+yD(x) +α(y)α(x),x

=

D(x) +α(x),yx+D(y),xx+ [y,x]D(x) +yD(x),x +α(y),xα(x) +α(y)α(x),x, x,yR.

(10)

We therefore have

D(x) +α(x),yx+D(y),xx

+ [y,x]D(x) +yD(x),x+α(y),xα(x) +α(y)α(x),x=0, x,yR. (11) According to relations (8) and (9), one can replace in the above relationy[D(x),x] by

y[α(x),x] and [D(x) +α(x),y]x+ [D(y),x]xby[α(y),x]xwhich gives

α(y),xG(x) +G(y)α(x),x+ [y,x]D(x)=0, x,yR, (12) whereG(x) denotesα(x)x.Putting in the above relationxyfory, we obtain, after some calculation,

α(x),xα(y)G(x) +α(x)α(y),xG(x) +G(x)α(y)α(x),x

+xG(y)α(x),x+x[y,x]D(x)=0, x,yR. (13)

(4)

Multiplying the relation (12) from the left side byx, subtracting the relation so ob- tained from the above relation and replacingα(y) byy, we obtain (note that [α(x),x]= [G(x),x],xR)

G(x),xyG(x) +G(x)[y,x]G(x) +G(x)yG(x),x=0, x,yR, (14) which reduces to

xG(x)yG(x) +G(x)yG(x)x=0, x,yR. (15) ApplyingLemma 1, the above relation gives

G(x),xyG(x)=0, x,yR. (16) Putting in the above relationyxfory, then multiplying the relation (16) from the right side byx, then subtracting the relations so obtained one from another, we obtain [G(x), x]y[G(x),x]=0,x,yR, which gives

G(x),x=0, xR. (17)

We therefore have [α(x),x]=0,xR, which gives, together with the relation (8),

D(x),x=0, xR. (18)

We have therefore proved thatGandDare both commuting onR.NowTheorem 5

completes the proof of the theorem.

Corollaries7and8are related to Posner’s second theorem as well as to Mayne’s theo- rem.

Corollary7. LetRbe a2-torsion-free semiprime ring. Suppose that there exist a deriva- tionD:RRand an automorphismα:RR,such that the mappingxD(x) +α(x)is centralizing onR.In this case,DandαImapRintoZ(R).

Proof. The proof is an immediate consequence of Theorems6and4.

Corollary8. LetRbe a noncommutative prime ring of characteristic different from two.

Suppose that there exist a derivationD:RRand an automorphismα:RR,such that the mappingxD(x) +α(x)is centralizing onR. In this case,D=0andα=I.

Thaheem [9] has proved that in case we have derivationsD,G:RR, whereRis a semiprime ring, satisfying the relationD(x)x+xG(x)=0, for allxR, then both deriva- tions mapRinto its center andD= −G(see also [10]).

In the same paper, Thaheem raised the question about a solution of the equation f(x)x+xg(x)=0, xR, (19) where f andgare additive mappings of a semiprime ringRinto itself.

The result below proved by Breˇsar [4] gives the answer to Thaheem’s question in case Ris a prime ring.

(5)

Theorem9 [4, Corollary 4.9]. LetRbe a prime ring and let f,g:RRbe additive map- pings satisfying the relation (19) for allxR. In this case, there existaQand an additive mappingσ:RCsuch that

f(x)=xa+σ(x), g(x)= −axσ(x) (20)

for allxR.

We point out that the identity (19) generalizes both concepts, the concept of commut- ing and the concept of skew-commuting mappings.

Our next result is related to Thaheem’s question mentioned above.

Theorem10. LetRbe a semiprime ring. Suppose that there exist a derivationD:RR and an automorphismα:RR,such thatD(x)x+x(α(x)x)=0holds for allxR.In this case,D=0andα=I.

Proof. We have the relation

D(x)x+xG(x)=0, xR, (21) whereG(x) stands forα(x)x.The linearization of the above relation gives

D(x)y+D(y)x+xG(y) +yG(x)=0, x,yR. (22) Putting in the above relationyxforyand applying (21), we obtain

D(x)yx+D(y)x2+xG(y)α(x) +xyG(x)=0, x,yR. (23) Right multiplication of the relation (22) byxgives

D(x)yx+D(y)x2+xG(y)x+yG(x)x=0, x,yR. (24) Subtracting the above relation from the relation (23), we obtain

xG(y)G(x) +xyG(x)yG(x)x=0, x,yR. (25) Putting in the above relationxyforyand applying (25), we obtain

0=xG(x)α(y)G(x) +x2G(y)G(x) +x2yG(x)xyG(x)x

=xG(x)α(y)G(x), x,yR. (26)

We therefore havexG(x)yG(x)=0,x,yR, which givesxG(x)yxG(x)=0,x,yR, whence it follows that

xG(x)=0, xR. (27)

(6)

From the above relation, one obtains according toLemma 3thatG(x)=0, xR.In other wordsα=I, which proves a part of the theorem. Now the relation (21) reduces to D(x)x=0, whence it follows, applying againLemma 3, thatD=0, which completes the

proof of the theorem.

We are ready for our last result.

Theorem11. LetRbe a2-torsion-free semiprime ring. Suppose that there exist a derivation D:RRand an automorphismα:RR,such that the mappingxD(x)x+xα(x)is commuting onR.In this case,DandαImapRintoZ(R).

Proof. We have the relation

D(x)x+xα(x),x=0, xR. (28) From the above relation, one easily obtains

A(x),y+D(x)y+D(y)x+xα(y) +yα(x),x=0, x,yR, (29) whereA(x) stands forD(x)x+xα(x).Let in the relation aboveybeyx.Then we have

0=

A(x),yx+D(x)yx+D(y)x2+yD(x)x+xα(y)α(x) +yxα(x),x

=

A(x),yx+D(x)y+D(y)xx,x+yA(x),x+xα(y)α(x),x

=

A(x),yx+D(x)y+D(y)x,xx+ [y,x]A(x) +α(y)α(x),x, x,yR.

(30)

We therefore have

A(x),yx+D(x)y+D(y)x,xx+ [y,x]A(x) +xα(y)α(x),x=0, x,yR. (31) According to (29), one can replace in the above relation [A(x),y]x+ [D(x)y+D(y)x, x]xby[xα(y) +yα(x),x]x.Thus we have

0= −

xα(y) +yα(x),xx+ [y,x]A(x) +xα(y)α(x),x

=xα(y),xx[y,x]α(x)xyα(x),xx+ [y,x]A(x) +xα(y),xα(x) +xα(y)α(x),x, x,yR.

(32)

We therefore have

xα(y),xG(x) + [y,x]B(x)yα(x),xx+xα(y)α(x),x=0, x,yR, (33) whereG(x) andB(x) denoteα(x)xandD(x)x+ [x,α(x)], respectively. The substitution xyforyin the above relation gives

xα(x),xα(y)G(x) +xα(x)α(y),xG(x) +x[y,x]B(x)xyα(x),xx

+xα(x)α(y)α(x),x=0, x,yR. (34) Left multiplication of the relation (33) byxgives

x2α(y),xG(x) +x[y,x]B(x)xyα(x),xx+x2α(y)α(x),x=0, x,yR. (35)

(7)

Subtracting the above relation from the relation (34) and replacingα(y) by y, one obtains (note that [α(x),x]=[G(x),x],xR)

xG(x),xyG(x) +xG(x)[y,x]G(x) +xG(x)yG(x),x=0, x,yR. (36) Collecting terms, the above relation we can write as

x2G(x)yG(x) +xG(x)yG(x)x=0, x,yR. (37) The substitution ofyxforyin the above relation gives

x2G(x)yxG(x) +xG(x)yxG(x)x=0, x,yR. (38) ApplyingLemma 1, one obtains, from the above relation,

xG(x),xyxG(x)=0, x,yR. (39) Putting first in the above relationyxfory, then multiplying the relation (39) from the right side byx, and then subtracting the relations so obtained one from another, we arrive atx[G(x),x]yx[G(x),x]=0,x,yR, whence it follows that

xα(x),x=0, xR. (40)

Combining the above relation with the relation (28), one obtains

D(x),xx=0, xR. (41)

From the relation (40), one easily obtains

xα(x),y+xα(y),x+yα(x),x=0, x,yR. (42) The substitutionxyforyin the above relation gives

0=xα(x),xy+xα(x)α(y),x+xyα(x),x

=x2α(x),y+xα(x)α(y),x+xyα(x),x, x,yR. (43) We therefore have

x2α(x),y+xα(x)α(y),x+xyα(x),x=0, x,yR. (44) Multiplying the relation (42) from the left side byx and subtracting the relation so obtained from the above relation, we obtainxG(x)[α(y),x]=0, x,yR, which means that we have

xG(x)[y,x]=0, x,yR. (45) Putting in the above relationyzfory, we arrive at

xG(x)y[z,x]=0, x,y,zR. (46)

(8)

From the above relation, one obtains

xG(x)y[z,w] +xG(w)y[z,x] +wG(x)y[z,x]=0, x,y,zR. (47) Putting in the above relation [z,w]yxG(x) fory and applying the relation (45), we obtain (xG(x)[z,w])y(xG(x)[z,w])=0,x,y,z,wR, whence it follows that

xG(x)[z,w]=0, x,z,wR. (48) From the above relation, it follows, according toLemma 3, that

G(x)[z,w]=0, x,z,wR. (49)

From the above relation andLemma 2, one can conclude thatG(x)Z(R) for any xR.In other words,αImapsRintoZ(R), which completes part of the proof. Using a similar approach, one can prove thatDmapsRintoZ(R) starting from the relation

(41). The proof of the theorem is complete.

Acknowledgments

The author thanks the referees for helpful comments and suggestions. This research has been supported by the Research Council of Slovenia.

References

[1] K. I. Beidar, W. S. Martindale III, and A. V. Mikhalev,Rings with Generalized Identities, Mono- graphs and Textbooks in Pure and Applied Mathematics, vol. 196, Marcel Dekker, New York, 1996.

[2] M. Breˇsar,Centralizing mappings and derivations in prime rings, J. Algebra156(1993), no. 2, 385–394.

[3] ,On skew-commuting mappings of rings, Bull. Austral. Math. Soc.47(1993), no. 2, 291–

296.

[4] ,On generalized biderivations and related maps, J. Algebra172(1995), no. 3, 764–786.

[5] N. Divinsky,On commuting automorphisms of rings, Trans. Roy. Soc. Canada. Sect. III. (3)49 (1955), 19–22.

[6] J. Luh,A note on commuting automorphisms of rings, Amer. Math. Monthly77(1970), 61–62.

[7] J. H. Mayne,Centralizing automorphisms of prime rings, Canad. Math. Bull.19(1976), no. 1, 113–115.

[8] E. C. Posner,Derivations in prime rings, Proc. Amer. Math. Soc.8(1957), 1093–1100.

[9] A. B. Thaheem and M. S. Samman,A note onα-derivations on semiprime rings, Demonstratio Math.34(2001), no. 4, 783–788.

[10] J. Vukman, Identities with derivations on rings and Banach algebras, to appear in Glasnik Matematicki.

[11] ,Centralizers on semiprime rings, Comment. Math. Univ. Carolin.42(2001), no. 2, 237–245.

[12] B. Zalar,On centralizers of semiprime rings, Comment. Math. Univ. Carolin.32(1991), no. 4, 609–614.

Joso Vukman: Department of Mathematics, Faculty of Education, University of Maribor, Koroˇska cesta 160, 2000 Maribor, Slovenia

E-mail address:[email protected]

参照

関連したドキュメント

In this paper, we extend the concept of a zero-divisor of a commutative ring R to that of a k-zero-divisor and investigate the interplay between the ring-theoretic properties of R

We remarked at the end of the proof of Theorem 3.6 that each residue map on a two- dimensional, complete, normal local ring is continuous with respect to the adic topology on the

In the past fifteen years, there has been an ongoing interest in derivations on prime or semiprime rings; and many of the results have involyed commutativity.. THE

a If R is a semiprime ring, the center of a nonzero one-sided ideal is contained in the center of R, in particular, any commutative one-sided ideal is contained in the center of R..

Thus a natural extension of the notion of a dependent element of mappings on a C ∗ -algebra or von Neumann algebras is the study of this notion in the context of semiprime rings

Tehranian, Science and Research Branch Islamic Azad University, Tehran, Iran E-mail : [email protected]. Yassemi, Center of Excellence in Biomathematics School of

Let R be an associative ring not necessarily with identity element. Every derivation is obviously a Jordan derivation and the converse is in general not true [1, Example 3.2.1].

An example will be given below of a connected ring that is a lattice under the Bordalo order and fails to be an RON -ring...