International Mathematical Forum, 2, 2007, no.

3, 103 - 108

The Commutativity in Prime Gamma Rings with Left Derivation

Mustafa ASCI Gazi University, Science and Art Faculty Department of Mathematics, Ankara, Turkey masci@gazi.edu.tr Sahin CERAN Pamukkale University, Science and Art Faculty Department of Mathematics, Denizli, Turkey sceran@pamukkale.edu.tr

Abstract. Let M be a ring, U an ideal of M, charM = 2, 3 and 0 = d : M M be a left derivation. In this paper we have proved the following results. (i) If d(U ) U and d2 (U) Z then M is commutative where Z = {c M : cm = mc : m M, } is the center of M. (ii) Let d1 , d2 be non zero left and right derivations of M and d2 (U) U. If d1 d2 (U) Z then M is commutative. Mathematics Subject Classication: Primary 16Y30, 16W25; Secondary 16U80 Keywords: Left derivation, commutativity, prime rings

1. Introduction The gamma ring is dened in [1] as follows. Let M and be additive Abelian groups. If for a, b, c M and , the following conditions are satised. (G1 ) ab M, (G2 ) (a + b)c = ac + bc, a( + )b = ab + ab, a(b + c) = ab + ac (G3 ) (ab)c = a(bc) Then M is called a ring (in the sense of Barnes). Every ring is a ring. A right (left) ideal of a ring M is an additive subgroup U of M such that UM U(MU U). If U is both a right and left ideal, then we say that U is an ideal of M. If the following conditions hold for a gamma ring M then M

104

Mustafa ASCI and Sahin CERAN

is called a prime gamma ring [2]. For a, b M aMb = 0 a = 0 or b = 0 In this paper we will prove some results in gamma rings which are known in ring theory. Throughout this paper M will be a prime ring, Z will be the center of M. Let M be a ring and D : M M be an additive map. D is called a left derivation if for all x, y M, D(xy) = xD(y) + yD(x) A right derivation is dened similarly. A ring M is called commutative if xy = yx for any x, y M and . An element m M is called a right nonzero divisor if xm = 0 implies x = 0 for any . A left nonzero divisor is dened similarly.

2. Main Results Lemma 1. Let U be a non-zero right ideal of M and d : M M left derivation (right derivation). (i) If d(U ) = 0 then d = 0 (ii) Let U be a non-zero ideal of M and d : M M right derivation, d(M ) left non-zero divisor. If ad(U ) = 0 then a = 0 (iii) Let U be a non-zero ideal of M and d : M M left derivation, D(M) right non-zero divisor. If d(U )a = 0 then a = 0 (iv) Let U be a non-zero ideal of M and charM = 2, d : M M right derivation. If d2 (U) = 0 then d = 0 (v) Let U be a non-zero ideal of M, charM = 2, d1 : M M left derivation and d2 : M M right derivation. If d2 (U) U and d1 d2 (U) = 0 then d1 = 0 or d2 = 0 Proof. (i) For any u U, and m M we have 0 = d(um) = ud(m) + md(u) = = ud(m) That is, Ud(M ) = 0. This implies that d(M ) = 0, by [[4] Lemma 2 (ii)] and so d = 0 (ii) Let , , u U and m M. Then 0 = ad(um) = ad(u)m + ad(m)u = ad(m)u

The commutativity in prime gamma rings

105

is obtained. That is, ad(M)U = 0 Hence ad(M) = 0 by [[4] Lemma 2 (ii)] and 0 = ad(m)d1 (m) = ad(m)d1 (m) + amm = aMM [[4] Lemma 2 (ii)] a = 0 (iii) Let , , u U and m M. Then 0 = d(um)a = ud(m)a + md(u)a = ud(m)a is obtained. That is, Ud(M )a = 0 hence d(M )a = 0, [[4] by Lemma 2 (ii)] and d(md1 (m))a = mma + d1 (m)d(m)a => MMa = 0 [[4] Lemma 2 (ii)], a = 0 (iv) Let charM = 2 and d2 (U) = 0. For any u U and we have 0 = d2 (uu) = d(u)d(u). Therefore, d(U )d(U ) = 0 and so d(U ) = 0 by (ii). Hence d = 0 by (i) (v) If d1 d2 (U) = 0 and d2 (U) U then for all u, v U and 0 = d1 d2 (uv) = d1 (d2 (u)v + d2 (v)u) = d2 (u)d1 (v) + vd1 d2 (v) + d2 (v)d1(u) + ud1d2 (v) = d2 (u)d1 (v) + d2 (v)d1 (u) If we replace u by d2 (u) in the last equation we get d2 (U)d1 (U) = 0 By (ii) 2 and (iii), we obtain d1 = 0 or d2 = 0 In the rest of this paper U will be a nonzero ideal of M. Lemma 2. Let d be a non-zero right derivation of M and charM = 2. If d(U ) Z then M is commutative. Proof. For all u U, y M, z Z and 0 = [d(uz), y] = [d(u)z + d(z)u, y] = d(z)[u, y] This implies that (2.1) [U, M] = 0 or d(Z) = 0 by [[4] Lemma 1(vi)] for all . If d(Z) = 0 then d2 (U) = 0, which implies that d = 0 by [[4] Lemma 1 (iv)]. This is a contradiction. Thus U Z is obtained and so M is commutative by [[4] Lemma 2 (i)].

106

Mustafa ASCI and Sahin CERAN

Theorem 3. Let d be a non-zero right derivation of M and charM = 2, 3. If d2 (U) Z and d(U ) U then M is commutative. Proof. For any u U, , and y M we have, 0 = [d2 (d(u)d(u)), y] = 2[d(d2 (u)d(u)), y] = 2[d3 (u)d(u) + d2 (u)d2 (u), y] = 2d3 (u)[d(u), y] and so by [[4]Lemma 1(v)] (2.2) 0 = = = = = d3 (u) = 0 or d(u) Z [d2 (ud(u)), y] [d(d(u)d(u) + d2 (u)u), y] [2d2 (u)d(u) + d3 (u)u + d(u)d2(u), y] 2d2 (u)[d(u), y] + [d(u), y] d2 (u) 3d2 (u)[d(u), y] d2 (u) = 0 or d(u) Z the sets K = {u U : d2 (u) = 0} and L = {u U : d(u) Z} are additive subgroups of U and U = K L If U = K then d = 0 by Lemma 1 (iv). This is a contradiction. Hence we have U = L by Brauer trick and so M is commutative by Lemma 2. Lemma 4. Let 0 = d1 : M M left derivation and 0 = d2 : M M right derivation such that d1 d2 (U) Z and d2 (U) U, charM = 2, 3. If d1 d2 (U) = 0 then M is commutative. 2 Proof. For any u U and we have for all y M. 0 = [d1 (d2 (d2 (u)d2(u))), y] = 2[d1 (d2 (u)d2(u)), y] 2 = 2[d2 (u)d1d2 (u) + d2 (u)d1d2 (u), y] 2 2 2 = 2[d2 (u)d1d2 (u), y] Since charM = 2, if we use [lemma 1(iv)] we get (2.3) d1 d2 (u) = 0 or d2 (u) Z 2 Let K = {u U : d1 d2 (u) = 0} and L = {u U : d2 (u) Z}. In this case K 2 and L are subgroups of U and U = K L. If U = K then d1 = 0 or d2 = 0 by Lemma 1 (v). This is a contradiction. Therefore U = L by Brauer trick, which implies that M is commutative by Theorem 1. is obtained. If d3 (U) = 0 then

Considering charM = 3 and using [[4] Lemma 1 (v)] we get for all u U

The commutativity in prime gamma rings

107

Lemma 5. Let 0 = d1 : M M left derivation and 0 = d2 : M M right derivation such that d1 d2 (U) Z, d2 (U) U, charM = 2 and a M. If [d1 (U), a] = 0 then [d2 (U), a] = 0 for all . Proof. For any x U and we have 0 = [d1 d2 (x)d2 (x), a] = 2[d2 (x)d1 d2 (x), a] = 2[d2 (x), a] d1 d2 (x) By [[4] Lemma 1 (iv)] for all x U, (2.4) d1 d2 (x) = 0 or [d2 (x), a] = 0 is obtained. Let K = {x U : d1 d2 (x) = 0} and L = {x U : [d2 (x), a] = 0}. Then K and L are subgroups of U and U = K L. Since d1 = 0 and d2 = 0, U can not equal to K by Lemma 1 (v). Hence U = L by Brauer trick. That is [d2 (U), a] = 0 for all , is obtained. Lemma 6. Let d1 be non-zero left derivation of M and d2 be non-zero right derivation of M such that d1 d2 (U) Z, d2 (U) U and a M. If [d1 (U), a] = 0 for all then a Z. Proof. By Lemma 4 we have [d2 (U), a] = 0 for all . Hence for all x U and 0 = = = = And so (2.5) d1 d2 (x) = 0 or [x, a] = 0 [[4] Lemma 1 (iv)]. If we apply to (2.5) Brauer trick and use Lemma 1 (v) we have a Z. Theorem 7. Let d1 be non-zero left derivation of M and d1 d2 (U) Z and d2 (U) U and a M, charM = 2, 3. If [d1 (U), a] Z for all then a Z. Proof. If Z = 0 d1 d2 (U) = 0 and so d1 = 0or d2 = 0 by Lemma 1. It is a contradiction. Hence we can assume that Z = 0. For any x U, z Z and , we have Z [d1 (xz), a] = [xd1 (z) + zd1 (x), a] = [x, a] d1(z) + z[d1 (x), a] [U, a] d1 (z) Z [d1 d2 (xx), a] 2[d1 (d2 (x)x), a] 2[d2 (x)d1 (x) + xd1 d2 (x), a] 2[xd1 d2 (x), a]

That is, for all , (2.6)

108

Mustafa ASCI and Sahin CERAN

By [[4] Lemma 1 (v)] and (2.6) we obtain for all (2.7) [U, a] Z or d1 (z) = 0 If d1 (z) = 0 then [U, a] Z, for all . Now, for any m M, , we have [[U, a] , m] = 0. Let us take uu instead of u in the last equation. Then we get, 0 = [[uu, a] , m] = [u[u, a] + [u, a] u, m] = 2[u[u, a] , m] Since M is a prime gamma ring, then for all m M, , , u U we have (2.8) [u, m] [u, a] = 0 [u, m] = 0 or [u, a] = 0 That is, for all (2.9) [U, a] = 0 If we consider Lemma 5 and (2.9) we get a Z. If d1 (z) = 0 then d1 d1 d2 (U) = 0. Thus for any x U and we have 0 = = = = and so for all x U (2.10) d1 d2 (x) = 0 or d1 d2 (x) = 0 2 If we say that K = {x U : d1 d2 (x) = 0} and L = {x U : d1 d2 (x) = 0} then 2 K and L are subgroups of U and U = K L. Since d1 = 0 and d2 = 0, we have U = K by Lemma 1. Thus if we consider Brauer trick we get U = L. Thus it d1 d2 (U) = 0. Now, d1 d2 (U) = 0, d1 d2 (U) Z, d2 (U) U and charM = 2, 3 2 2 implies that M is commutative by Lemma 3 and a Z. References
[1] Barnes W. E., On the -rings of Nobusawa, Pacic J. of Math.Vol. 18, No 3 , (1966). [2] Kyum S. , On prime gamma rings, Pacic J. of Math.Vol. 25, No 1 , (1978). [3] Feng Jie Jing, On derivations of -rings; Qu fu shifon Daxue Xuebeo Ziran Kexue Ban 13 No 4, 159 - 161 (1987) [4] Soyt rk M., The commutativity in prime gamma rings with derivation, Tr. J.Of u Math.Vol. 18. 149 - 155 , (1994).

This implies that, [[4] by Lemma 1 (iv)] for all m M, ,

d1 d1 d2 (d2 (x)d2 (x)) d1 d1 (2d2 (x)d2 (x)) 2 2 2d1 (d2 (x)d1 d2 (x) + d2 (x)d1 d2 (x)) 2 4d1 d2 (x)d1 d2 (x) 2

Received: May 30, 2006