c Applied Mathematics E-Notes, 1(2001), 24-30 ° Available free at mirror sites of http://math2.math.nthu.edu.tw/»amen/ Some Results for Generalized Lie Ideals in Prime Rings with Derivation II ¤ Ä K^az³m Kayay, Oznur GÄolba»siz, Ne»set Aydinx Received 18 December 2000 Abstract Let R be a prime ring of characteristic di®erent from two, d : R ! R a nonzero derivation, and M a non-zero left ideal of R: We prove the following results: (1) if a 2 R and [d(R); a]¾;¿ = 0; then ¾(a) + ¿ (a) 2 Z; the center of R; (2) if d([R; a]¾;¿ ) = 0; then ¾(a) + ¿ (a) 2 Z; (3) if ([R; M ]¾;¿ ; a)¾;¿ = 0; then a 2 Z; (4) (d(R); a) = 0 if, and only if, d((R; a)) = 0: Let R be a ring and ¾; ¿ be two mappings from R into itself. We write [x; y]; (x; y); [x; y]¾;¿ ; (x; y)¾;¿ for xy ¡yx; xy +yx; x¾(y)¡¿ (y)x and x¾(y)+¿ (y)x respectively, and will make extensive use of the following basic commutator identities: (xy; z) = x[y; z] + (x; z)y = x(y; z) ¡ [x; z]y; [xy; z]¾;¿ = x[y; z]¾;¿ + [x; ¿ (z)]y = x[y; ¾(z)] + [x; z]¾;¿ y. An additive mapping D : R ! R is called a derivation if D(xy) = D(x)y + xD(y) holds for all x; y 2 R. A derivation D is inner if there exists an a 2 R such that D(x) = [a; x] for all x 2 R: For subsets A; B ½ R; let [A; B] ([A; B]¾;¿ ) be the additive subgroup generated by all [a; b] ([a; b]¾;¿ ) for all a 2 A and b 2 B: We recall that a Lie ideal L is an additive subgroup of R such that [R; L] ½ L: We ¯rst introduce the generalized Lie ideal in [4] as following. Let U be an additive subgroup of R, and let ¾; ¿ : R ! R be two mappings. Then (i) U is a (¾; ¿ )-right Lie ideal of R if [U; R]¾;¿ ½ U; (ii) U is a (¾; ¿ )-left Lie ideal of R if [R; U ]¾;¿ ½ U; (iii) if U is both a (¾; ¿ )-right Lie ideal and (¾; ¿ )-left Lie ideal of R; then U is a (¾; ¿ )-Lie ideal of R: Every Lie ideal of R is a (1; 1)-Lie ideal of R; where 1 : R ! R is the identity map. As an example, let I be the set of integers, ½µ ¶ ¾ x y R= : x; y; z; t 2 I ; z t ½µ ¶ ¾ x y U= : x; y 2 I ½ R; 0 x µ ¶ 1 ¡1 and ¿ : R ! R the mapping de¯ned by ¿ (x) = bxb, where b = 2 R. Then 0 ¡1 U is a (1; ¿ )-left Lie ideal but not a Lie ideal of R: ¤ Mathematics Subject Classi¯cations: 16N10, 16W25, 16U80. Onsekiz mart Univ., Math. Department, C » anakkale, Turkey z Cumhuriyet Univ., Math. Department, 58140 Sivas, Turkey x Mersin Univ., Math. Department, 33342 Mersin, Turkey yC » anakkale 24 25 Kaya et al. Some algebraic properties of (¾; ¿ )-Lie ideals are considered in [1], [2] and [5], where further references can be found. Let R be a prime ring of characteristic di®erent from two, d : R ! R a nonzero derivation, Z the center of R and a 2 R. Lee and Lee in [6] proved that if [d(R); d(R)] ½ Z; then R is commutative. In the present paper, we generalize this result for generalized Lie ideal. In [3], Herstein proved that in a prime ring of characteristic di®erent from two, [d(R); a] = 0 implies a 2 Z: We shall extend Herstein's theorem by proving that [d(R); a]¾;¿ = 0 implies ¾(a) + ¿ (a) 2 Z: Throughout this note, R will be a prime ring with characteristic di®erent from 2, Z the center of R; d a non-zero derivation of R and U is (¾; ¿ )-left Lie ideal of R. LEMMA 1. If a 2 R and [d(R); a]¾;¿ = 0; then ¾(a) + ¿ (a) 2 Z: PROOF. If a 2 Z then the proof of the theorem is obvious. So we assume that a2 = Z: By hypothesis, we have for all x 2 R, 0 = [d(x¾(a)); a]¾;¿ = [d(x)¾(a) + xd(¾(a)); a]¾;¿ = d(x)[¾(a); ¾(a)] + [d(x); a]¾;¿ ¾(a) + x[d(¾(a)); a]¾;¿ + [x; ¿ (a)]d(¾(a)): Hence we obtain [x; ¿ (a)]d(¾(a)) = 0; x 2 R (1) Replacing x by xy; y 2 R in (1) and using (1), we get [R; ¿ (a)]Rd(¾(a)) = 0: Since R is prime ring and a 2 = Z, we obtain d(¾(a)) = 0: Now let us consider the following mappings on R : D(x) = [x; ¾(a)] and H(x) = [x; a]¾;¿ ; where D is a non-zero derivation of R such that Hd(x) = 0: For any x; y 2 R; we have H(xy) = [xy; a]¾;¿ = x[y; ¾(a)] + [x; a]¾;¿ y: Hence we get H(xy) = H(x)y + xD(y) (2) But this can also be calculated in a di®erent way. Indeed, H(xy) = [xy; a]¾;¿ = x[y; a]¾;¿ + [x; ¿ (a)]y and so one obtains H(xy) = [x; ¿ (a)]y + xH(y) (3) For any r 2 R; 0 = [d(r); a]¾;¿ = d(r)¾(a) ¡ ¿ (a)d(r); and so, 0 = = d(0) = d(d(r)¾(a) ¡ ¿ (a)d(r)) d2 (r)¾(a) + d(r)d(¾(a)) ¡ d(¿ (a))d(r) ¡ ¿ (a)d2 (r) [d2 (r); a]¾;¿ ¡ d(¿ (a))d(r): This implies that d(¿ (a))d(r) = 0 for all r 2 R: Using [7, Lemma 1], we obtain d(¿ (a)) = 0: On the other hand, for any x 2 R; dH(x) = = d([x; a]¾;¿ ) = d(x¾(a) ¡ ¿ (a)x) d(x)¾(a) + xd(¾(a)) ¡ d(¿ (a))x ¡ ¿ (a)d(x) = [d(x); a]¾;¿ = 0: 26 Generalized Ideals in Prime Rings Thus we get, dH(R) = 0: (4) In view of (2), (3) and (4), one obtains, for any x; y 2 R; 0 = Hd(xH(y)) = H(d(x)H(y) + xdH(y)) = H(d(x)H(y)) = Hd(x)H(y) + d(x)DH(y) = d(x)DH(y); and so, d(R)DH(R) = 0: By using [7, Lemma 1] we arrive at DH(x) = 0: That is, [[x; a]¾;¿ ; ¾(a)] = 0; x 2 R: (5) Taking ¿ (a)x; instead of x in (5), we get 0 = [[¿ (a)x; a]¾;¿ ; ¾(a)] = [¿ (a)[x; a]¾;¿ + [¿ (a); ¿ (a)]x; ¾(a)] = [¿ (a)[x; a]¾;¿ ; ¾(a)] = ¿ (a)[x; a]¾;¿ ; ¾(a)] + [¿ (a); ¾(a)][x; a]¾;¿ ; and so, [¿ (a); ¾(a)][x; a]¾;¿ = 0; x 2 R: (6) Replacing x by xy; y 2 R in (6) and using (6), we obtain [¿ (a); ¾(a)]R[R; ¾(a)] = 0: Since R is a prime ring and a 2 = Z, we get [¿ (a); ¾(a)] = 0: (7) Now, expanding (5) and using (7) one obtains 0 = [[x; a]¾;¿ ; ¾(a)] = [x¾(a) ¡ ¿ (a)x; ¾(a)] = [x; ¾(a)]¾(a) ¡ ¿ (a)[x; ¾(a)] = [[x; ¾(a)]; a]¾;¿ ; that is, HD(x) = 0; x 2 R: (8) Linearizing (8), we get 0 = HD(xy) = H(D(x)y + xD(y)) = HD(x)y + D(x)D(y) + xHD(y) + [x; ¿ (a)]D(y) = D(x)D(y) + [x; ¿ (a)]D(y) = [x; ¾(a)]D(y) + [x; ¿ (a)]D(y); that is, [x; ¾(a) + ¿ (a)]D(y) = 0; x; y 2 R: Since D is a non-zero derivation of the prime ring R, using [7, Lemma 1], we obtain ¾(a) + ¿ (a) 2 Z: 27 Kaya et al. COROLLARY 2. Let R be a prime ring of characteristic di®erent from two, d be a nonzero derivation of R and U is (¾; ¿ )-left Lie ideal of R: If [d(R); U ]¾;¿ = 0; then ¾(u) + ¿ (u) 2 Z for all u 2 U: THEOREM 3. Let R be a prime ring of characteristic di®erent from two, d a nonzero derivation of R and a 2 R: If d([R; a]¾;¿ ) = 0; then ¾(a) + ¿ (a) 2 Z: PROOF. For all r 2 R; 0 = d([¿ (a)r; a]¾;¿ ) = d(¿ (a)[r; a]¾;¿ ) = d(¿ (a))[r; a]¾;¿ + ¿ (a)d([r; a]¾;¿ ) and so, d(¿ (a))[r; a]¾;¿ = 0; r 2 R: (9) Taking rs; s 2 R instead of r in (9), we have 0 = d(¿ (a))[rs; a]¾;¿ = d(¿ (a))r[s; ¾(a)] + d(¿ (a))[r; a]¾;¿ s: Using (9), we get d(¿ (a))R[R; ¾(a)] = 0: Since R is a prime ring, we see that d(¿ (a)) = 0 or a 2 Z: If a 2 Z; then ¾(a)+¿ (a) 2 Z: The proof is then complete. Therefore, let us assume that a 2 = Z. For r 2 R; 0 = d([r¾(a); a]¾;¿ ) = d([r; a]¾;¿ ¾(a)) = d([r; a]¾;¿ )¾(a) + [r; a]¾;¿ d(¾(a)) and so, [r; a]¾;¿ d(¾(a)) = 0; r 2 R: Replacing r by rs; s 2 R in (10), we get 0 = [rs; a]¾;¿ d(¾(a)) = r[s; a]¾;¿ d(¾(a)) + [r; ¿ (a)]sd(¾(a)) = [r; ¿ (a)]sd(¾(a)): That is, [r; ¿ (a)]Rd(¾(a)) = 0; r 2 R: Since R is a prime ring, and since we have assumed that a 2 = Z; we obtain d(¾(a)) = 0: Now, from our hypothesis, we have, for any r 2 R; 0 = d([r; a]¾;¿ ) = d(r¾(a) ¡ ¿ (a)r) = d(r)¾(a) + rd(¾(a)) ¡ d(¿ (a))r ¡ ¿ (a)d(r): Since d(¾(a)) = 0 and d(¿ (a)) = 0, we get [d(r); a]¾;¿ = 0; r 2 R: By Lemma 1, we have ¾(a) + ¿ (a) 2 Z: (10) 28 Generalized Ideals in Prime Rings COROLLARY 4. Let R be a prime ring of characteristic di®erent from two, d a nonzero derivation of R and U is (¾; ¿ )-left Lie ideal of R: If d([R; U ]¾;¿ ) = 0; then ¾(u) + ¿ (u) 2 Z for all u 2 U: EXAMPLE. Let ½µ ¶ ¾ a b R= : a; b; c; d 2 I c d and µ ¶ µ ¶ µ ¶ µ ¶ a b a ¡b ¾ = ;¿ = c d ¡c d µ ¶ 1 0 be two autumorphisms of R and a = 2 = Z: If we de¯ne d : R ! R by 0 2 µ ¶ µ ¶ a b 0 ¡b d = ; then d is a derivation of R such that d([R; a]¾;¿ ) = 0; but c d c 0 ¾(a) + ¿ (a) 2 Z: THEOREM 6. Let R be a prime ring of characteristic di®erent from two, M a non-zero ideal of R and a 2 R: If ([R; M ]¾;¿ ; a)¾;¿ = 0; then a 2 Z: PROOF. Let m be a noncentral element of M . Consider the following mappings on R : d1 (r) = [r; m]¾;¿ and d2 (r) = (r; a)¾;¿ : Then for any r 2 R; we have d2 d1 (r) = d2 ([r; m]¾;¿ ) = ([r; m]¾;¿ ; a)¾;¿ = 0: That is, a b c d d ¡b ¡c a d2 d1 (r) = 0; r 2 R: (11) For r 2 R; d1 (r¾(m)) = = [r¾(m); m]¾;¿ r[¾(m); ¾(m)] + [r; m]¾;¿ ¾(m) = d1 (r)¾(m); and so we get, d1 (r¾(m)) = d1 (r)¾(m); r 2 R: (12) Let us consider (11) together with (12), we obtain 0 = = d2 d1 (r¾(m)) = d2 (d1 (r)¾(m)) = (d1 (r)¾(m); a)¾;¿ d1 (r)[¾(m); ¾(a)] + (d1 (r); a)¾;¿ ¾(m); and so we have d1 (R)[¾(m); ¾(a)] = 0: (13) Putting in (13) rs; s 2 R for r we obtain 0 = d1 (rs)[¾(m); ¾(a)] = rd1 (s)[¾(m); ¾(a)] + [r; ¿ (m)]s[¾(m); ¾(a)]; and using (13) we get [r; ¿ (m)]R[¾(m); ¾(a)] = 0; r 2 R: Since R is a prime ring, we have m 2 Z or ¾([m; a]) = 0: If m 2 Z; then ¾([m; a]) = 0: That is, [M; a] = 0 since ¾ is automorphism of R: For all x 2 R and m 2 M; 0 = 29 Kaya et al. [xm; a] = x[m; a] + [x; a]m = [x; a]m: That is, [x; a]RM = (0); the zero ideal of R. By the primeness of R we have a 2 Z: THEOREM 7. Let R be a prime ring of characteristic di®erent from two, d is a nonzero derivation of R and a 2 R: Then (d(R); a) = 0 if, and only if, d((R; a)) = 0: PROOF. Suppose (d(R); a) = 0: In this case we claim that d(a) = 0: If a = 0 then d(a) = 0. So we assume that a 6= 0. For any x 2 R; the relation (d(x); a) = 0 gives 0 = (d(xa); a) = (d(x)a + xd(a); a) = d(x)[a; a] + (d(x); a)a + x(d(a); a) ¡ [x; a]d(a) and so, [x; a]d(a) = 0; x 2 R: (14) If we take xy, y 2 R instead of x in (14), we obtain [R; a]Rd(a) = (0) Since R is a prime ring we have a 2 Z or d(a) = 0: Now, if a 2 Z then 0 = (d(a); a) = d(a)a + ad(a) = 2d(a)a. Then d(a)a = 0: Since we assumed that 0 6= a and R is a prime ring, we get d(a) = 0: Thus, we conclude that d(a) = 0: Hence for any r 2 R we have d((r; a)) = (d(r); a) + (r; d(a)); and so, d((R; a)) = 0: Conversely, for x 2 R; 0 = = d((ax; a)) = d(a(x; a) + [a; a]x) d(a(x; a)) = d(a)(x; a) + ad((x; a)): Hence we have d(a)(x; a) = 0; x 2 R: (15) In (15) replace x by xy and use (15), we get 0 = d(a)(xy; a) = d(a)x[y; a] + d(a)(x; a)y = d(a)x[y; a]; that is, d(a)R[R; a] = 0: Since R is a prime ring, we have d(a) = 0 or a 2 Z: If d(a) = 0; then for any r 2 R we have 0 = d((r; a)) = (d(r); a) + (r; d(a)) = (d(r); a); and so (d(R); a) = 0: If a 2 Z; then 0 = d((a; a)) = 4d(a)a: Since the characteristic of R is di®erent from 2; d(a)a = 0: Since a 2 Z, we get d(a)Ra = 0; and so, d(a) = 0 or a = 0: Thus, d(a) = 0 is obtained. Finally, (d(R); a) = 0 as required. References [1] N. Ayd³n and H. Kandamar, (¾; ¿ )-Lie ideals in prime rings, Doģa Tr. J. of Math., 18(2)(1994), 143-148. [2] N. Ayd³n and M. SoytÄ urk, (¾; ¿ )-Lie ideals in prime rings with derivations,.Doģa Tr. J. of Math., 19(1993), 239-244. [3] I. N. Herstein, A note on derivations II, Canad. Math. Bull., 22(4)(1979), 509-511. [4] K. Kaya, (¾; ¿ )-Lie ideals in prime rings, An. Univ. Timisoara Stiinte Math., 30(23)(1992), 251-255. 30 Generalized Ideals in Prime Rings [5] K. Kaya and N. Ayd³n, Some results in generalized Lie ideal, Albasapr Sci. J. Issued by Jordan University for Woman, 3(1)(1999), 53-61. [6] P. H. Lee and T. K. Lee, On derivations of prime rings, Chinese J. Math., 9(2)(1981), 107-110. [7] E. C. Posner, Derivations in prime rings, Proc. Amer. Math. Soc., 8(1957), 10931100.