Sei sulla pagina 1di 14

Chapter 2

Generalized derivations in rings with


involution

2.1 Introduction
An additive mapping x → x∗ on a ring R is called an involution on R if
(xy)∗ = y ∗ x∗ and (x∗ )∗ = x hold for all x, y ∈ R. A ring equipped with an
involution is called a ring with involution or ∗-ring. A ring with an involution
‘*’ is said to be ∗-prime if aRb = aRb∗ = {0} or a∗ Rb = aRb = {0} implies
that either a = 0 or b = 0. It is obvious that every prime ring with an
involution ‘*’ is ∗-prime but the converse need not hold in general. An
example due to Oukhtite [65] justifies the above fact that is, let R be a prime
ring. Consider S = R × Ro , where Ro is the opposite ring of R. Define
involution ∗ on S as (x, y)∗ = (y, x). Since (0, x)S(x, 0) = {0}, it follows
that S is not prime. Further, it can be easily seen that if (a, b)S(c, d) =
(a, b)S(c, d)∗ = {0}, then either (a, b) = 0 or (c, d) = 0. Hence S is ∗-prime
but not prime. The set of symmetric and skew-symmetric elements of a ∗-
ring will be denoted by S∗(R) i.e., S∗ (R) = {x ∈ R | x∗ = ±x}.
In Section 2.2 it is shown that a ∗-Lie ideal U of ∗-prime rings R is central
if it admits a generalized derivation F with associated derivation d satisfying
any one of the following identities on ∗-Lie ideal U of R: (i)F [u, v] = [F (u), v],
(ii)F (u ◦ v) = F (u) ◦ v, (iii)F [u, v] = [F (u), v] + [d(v), u], (iv)F (u ◦ v) =
F (u) ◦ v + d(v) ◦ u, (v)F (uv) ± uv = 0 and (vi)d(u)F (v) ± uv = 0 for all
u, v ∈ U.
Section 2.3 deals with the commutativity of ∗-prime ring R admitting
generalized derivations F and G associated with non-zero derivations d and
g respectively satisfying any one of the following identities on a non-zero ∗-
Jordan ideal J of R: (i)[F (u), u] = 0,(ii)F [u, v] = u ◦ v, (iii)F (u ◦ v) = [u, v],
(iv)F (uv) = F (u)F (v) , (v)F [u, v] = [G(u), v], (vi)F (u)G(v) ± uv = 0 for all
u, v ∈ J.
We begin Section 2.4 by defining generalized (α, β)∗ -derivation and gen-

11
eralized (α, β)∗ - reverse derivation on a ring R. The main result of this
section states that if a semi-prime ring with involution ‘*’ admits a general-
ized (α, β)∗ -derivation F on R such that α is surjective, then F maps R into
Z(R). Various other theorems concerning prime ring are also obtained.

2.2 Generalized derivations on ∗-Lie ideal


A Lie ideal of a ring R is said to a ∗-Lie ideal if U ∗ = U . In the year 1981,
Bergen et. al [18] proved that for a Lie ideal U of a prime ring R (i) if
d(U ) ⊆ Z(R) then U ⊆ Z(R), (ii) if d2 (U) ⊆ Z(R) then U ⊆ Z(R). These
results were further extended for ∗-Lie ideal U of ∗-prime ring R. Also,
commutativity of ∗-prime rings have been investigated involving various
identities for derivations and generalized derivations (for refrences see [67],
[68]).

We begin our discussion by stating some results of Oukhtite et. al which


are also used in proving our results.

Lemma 2.2.1 ([68], Lemma 3) Let U be a non-zero ∗-Lie ideal of a 2-


torsion free ∗-prime ring R. If [U, U]
= {0}, then there exists a non-zero
∗-ideal M of R such that [M, R] ⊆ U and [M, R]
⊆ Z(R).

Lemma 2.2.2 ([68], Lemma 4) If U


⊆ Z(R) is a ∗-Lie ideal of a 2-torsion
free ∗-prime ring and a, b ∈ R such that aUb = {0} = a∗U b then either a = 0
or b = 0.

Lemma 2.2.3 ([67], Lemma 2.3) Let U be a non-zero ∗-Lie ideal of a 2-


torsion free ∗-prime ring R. If [U, U ] = {0}, then U ⊆ Z(R).

Lemma 2.2.4 ([67], Theorem 1.1) Let R be a 2-torsion free ∗-prime ring, U
a non-zero Lie ideal of R and d a non-zero derivation of R which commutes
with ∗. If d2(U ) = {0}, then U ⊆ Z(R).

Lemma 2.2.5 ([67], Lemma 2.4) Let U be a Lie ideal of a 2-torsion free
∗-prime ring R and d(
= 0) be a derivation of R which commutes with ∗. If
d(U ) ⊆ Z(R), then U ⊆ Z(R).

Lemma 2.2.6 ([67], Lemma 2.5) Let d(


= 0) be derivation of a 2-torsion
free ∗-prime ring R which commutes with ∗. Let U
⊆ Z(R) be a ∗-Lie ideal
of R. If t ∈ R satisfies td(U) = {0} or d(U )t = {0} then t = 0.

We shall now prove the following :

12
Lemma 2.2.7 Let R be a 2-torsion free ∗-prime ring and U be a ∗-Lie ideal
of R. If a ∈ S∗ (R) ∩ R such that [a, U ] ⊆ Z(R) then either U ⊆ Z(R) or
a ∈ Z(R).

Proof Let U
⊆ Z(R). The given hypothesis can be written as Ia (U) ⊆ Z(R)
where Ia is the inner derivation determined by a. Hence using Lemma 2.3.5,
Ia = 0 and this gives that a ∈ Z(R).

Lemma 2.2.8 Let R be a 2-torsion free ∗-prime ring and d be a non-zero


derivation of R which commutes with ∗. If U
⊆ Z(R) is a ∗-Lie ideal of R
such that [a, d(U)] = {0} for some a ∈ S∗(R) ∩ R, then a ∈ Z(R).

Proof Replacing u by [a, u] in [a, d(U)] = {0} we have,

0 = [a, d[a, u]] = [a, [a, d(u)]] + [a, [d(a), u]]


= [a, [d(a), u]] for all u ∈ U.
Hence, 0 = d[a, [d(a), u]] = [d(a), [d(a), u]] + [a, d[d(a), u]] for all u ∈ U. Now
using the hypothesis 0 = [d(a), [d(a), u]] for all u ∈ U. Let Id(a) be the inner
2
derivation determined by d(a) then we have Id(a) (u) = 0 hence by Lemma
2.3.4, d(a) ∈ Z(R). Therefore, d[a, u] = [d(a), u] + [a, d(u)] = 0. Replacing u
by [a2 , u] in [a, d(U)] = {0} we obtain,

0 = [a, d[a2 , u]]


= [a, d(a[a, u] + [a, u]a)]
= [a, d(a)[a, u] + ad[a, u] + d[a, u]a + [a, u]d(a)]
= [a, d(a)[a, u]] + [a, [a, u]d(a)]
= d(a)[a, [a, u]] + [a, d(a)][a, u] + [a, u][a, d(a)] + [a, [a, u]]d(a)
= d(a)[a, [a, u]] + [a, [a, u]]d(a)
= 2d(a)[a, [a, u]].

Since R is 2-torsion free, d(a)[a, [a, u]] = 0 for all u ∈ U. Hence

d(a)[a, [a, u]] = d(a)U[a, [a, u]] = 0 for all u ∈ U.

Since, a ∈ S∗ (R) ∩ R so, d(a) ∈ S∗ (R) ∩ R. Thus, {0} = d(a)U [a, [a, u]] =
(d(a))∗U [a, [a, u]] for all u ∈ U . Therefore, either [a, [a, u]] = 0 for all u ∈ U
or d(a) = 0. If [a, [a, u]] = 0 for all u ∈ U then a ∈ Z(R). If d(a) = 0, using
Lemma 2.3.1 there exists an ∗-ideal M of R, let [va, u] ∈ U where v ∈ [M, R].
Hence, we obtain

0 = [a, d[va, u]]


= [a, d(v)[a, u] + vd[a, u] + d[v, u]a + [v, u]d(a)]
= [a, d(v)[a, u] + [v, u]d(a)]
= d(v)[a, [a, u]] + [a, d(v)][a, u] + [v, u][a, d(a)] + [a, [v, u]]d(a)
= d(v)[a, [a, u]] for all v ∈ [M, R], u ∈ U.

13
Therefore, d[M, R][a, [a, u]] = {0} for all u ∈ U. Using Lemma 2.3.6,
[a, [a, u]] = 0 for all u ∈ U and again using Lemma 2.3.4 we get the required
result.

We facilitate our discussion by proving the following theorem


Theorem 2.2.9 Let R be a 2-torsion free ∗-prime ring and F : R → R be a
generalized derivation with associated non-zero derivation d which commutes
with ∗. If U is a ∗-Lie ideal of R such that F [u, v] = [F (u), v] for all u, v ∈ U
then U ⊆ Z(R).
Proof Replacing u by [u, ru] in F [u, v] = [F (u), v] for all u, v ∈ U we have

F [[u, r]u, v] = [F ([u, r]u), v] for all u, v ∈ U, r ∈ R.

This implies that F ([u, r][u, v] + [[u, r], v]u) = [F [u, r]u + [u, r]d(u), v] for all
u, v ∈ U, r ∈ R. Using the hypothesis we obtain [u, r]d[u, v] = [u, r][d(u), v]
for all u, v ∈ U, r ∈ R. This gives us [u, r][u, d(v)] = 0 for all u, v ∈ U, r ∈ R.
Replacing r by rs for some s in R we get

[u, R]R[u, d(v)] = {0} for all u, v ∈ U. (2.2.1)

If u ∈ S∗ (R) ∩ U, then [u, R]R[u, d(U)] = [u, R]∗ R[u, d(U)]. Thus, for some
u ∈ S∗ (R) ∩ U either [u, R] = {0} or [u, d(U )] = {0}. But for any u ∈ U,
u −u∗ , u+u∗ ∈ S∗(R) ∩U. Therefore, for some u ∈ U either [u − u∗ , R] = {0}
or [u − u∗, d(U)] = {0}. If [u − u∗, R] = {0} then from equation (2.2.1) we
obtain that [u, R]R[u, d(U )] = [u, R]∗R[u, d(U )] = {0} for all u ∈ U hence
either [u, R] = {0} or [u, d(U )] = {0}. Let L = {u ∈ U | [u, R] = {0}} and
K = {u ∈ U | [u, d(U )] = {0}}. Then it can be seen that L and K are
two additive subgroups of U whose union is U. Using Brauer’s trick we have
either L = U or K = U . If L = U , then [u, R] = {0} for all u ∈ U that is
U ⊆ Z(R) and if K = U , then [u, d(U)] = {0} for all u ∈ U, which implies
that U ⊆ Z(R) by Lemma 2.2.8. If [u − u∗ , d(U)] = {0}, then again by
equation (2.2.1) we obtain that [u, R]R[u, d(U )] = [u, R]R[u, d(U)]∗ = {0}
for all u ∈ U . This gives us either [u, R] = {0} or [u, d(U)] = {0}. For
each such u, the set satisfying these two properties form additive subgroups
of U whose union is U. Therefore by Brauers’s trick we find that either
[u, R] = {0} for all u ∈ U or [u, d(U)] = {0} for all u ∈ U. If [u, d(U)] = {0}
for all u ∈ U then using Lemma 2.2.7 we obtain U ⊆ Z(R). Hence in any
U ⊆ Z(R).

Theorem 2.2.10 Let R be a 2-torsion free ∗-prime ring and F : R → R be a


generalized derivation with associated non-zero derivation d which commutes
with ∗. If U is a ∗-Lie ideal of R such that F (u◦v) = F (u)◦v for all u, v ∈ U
then U ⊆ Z(R).

14
Proof Replacing u by [u, ru] in F (u ◦ v) = F (u) ◦ v for all u, v ∈ U, r ∈ R
we have

F ([u, r]u ◦ v) = F ([u, r]u) ◦ v for all u, v ∈ U, r ∈ R.


This yields that,

F (([u, r]◦v)u+[u, r][u, v]) = F [u, r]u◦v +[u, r]d(u)◦v for all u, v ∈ U, r ∈ R.

Thus we obtain,
F ([u, r] ◦ v)u + ([u, r] ◦ v)d(u) + F [u, r][u, v] + [u, r]d[u, v]

= (F [u, r] ◦ v)u + F [u, r][u, v] + ([u, r] ◦ v)d(u)


+[u, r][d(u), v] for all u, v ∈ U, r ∈ R.
Using our hypothesis we find that [u, r]d[u, v] = [u, r][d(u), v] for all u, v ∈
U, r ∈ R. Hence, we obtain, [u, r][u, d(v)] = 0 for all u, v ∈ U, r ∈ R. Replac-
ing r by rs for some s in R we get [u, R]R[u, d(v)] = {0} for all u, v ∈ U.
This leads to equation (2.2.1). Hence, proceeding on the same way as above,
we obtain that U ⊆ Z(R).

Theorem 2.2.11 Let R be a 2-torsion free ∗-prime ring and F : R → R be a


generalized derivation with associated non-zero derivation d which commutes
with ∗. If U is a ∗-Lie ideal of R such that F [u, v] = [F (u), v] + [d(v), u] for
all u, v ∈ U then U ⊆ Z(R).

Proof We have F [u, v] = [F (u), v] + [d(v), u] for all u, v ∈ U. Now replacing


u by [u, ru] we get

F [[u, r]u, v] = [F ([u, r]u), v] + [d(v), [u, r]u] for all u, v ∈ U, r ∈ R.

This gives us
F ([u, r][u, v] + [[u, r], v]u) = [F [u, r]u + [u, r]d(u), v] + [d(v), [u, r]u]
for all u, v ∈ U, r ∈ R.
Hence,
F [u, r][u, v] + [u, r]d[u, v] + F [[u, r], v]u + [[u, r], v]d(u)

= F [u, r][u, v] + [F [u, r], v]u + [u, r][d(u), v] + [[u, r], v]d(u)
+[u, r][d(v), u] + [d(v), [u, r]]u for all u, v ∈ U, r ∈ R.
Using the hypothesis we obtain , [u, r]d[u, v] = [u, r][d(v), u] + [u, r][d(u), v]
for all u, v ∈ U, r ∈ R. This gives us [u, r][u, d(v)] = 0 for all u, v ∈ U, r ∈ R.
This is same as equation (2.2.1) hence, continuing in the same manner as
above we obtain that U ⊆ Z(R).

15
Theorem 2.2.12 Let R be a 2-torsion free ∗-prime ring and F : R → R be a
generalized derivation with associated non-zero derivation d which commutes
with ∗. If U is a ∗-Lie ideal of R such that F (u ◦ v) = F (u) ◦ v + d(v) ◦ u for
all u, v ∈ U then U ⊆ Z(R).

Proof Replacing u by [u, ru] in F (u ◦ v) = F (u) ◦ v + d(v) ◦ u for all u, v ∈ U


we get, F ([u, r]u ◦ v) = F ([u, r]u) ◦ v + d(v) ◦ [u, r]u for all u, v ∈ U, r ∈
R. This gives us, F (([u, r] ◦ v)u + [u, r][u, v]) = F [u, r]u ◦ v + [u, r]d(u) ◦ v +
d(v) ◦ [u, r]u for all u, v ∈ U, r ∈ R. Thus,

F ([u, r] ◦ v)u + ([u, r] ◦ v)d(u) + F [u, r][u, v] + [u, r]d[u, v]

= (F [u, r] ◦ v)u + F [u, r][u, v] + ([u, r] ◦ v)d(u) + [u, r][d(u), v]


+(d(v) ◦ [u, r])u + [u, r][d(v), u] for all u, v ∈ U, r ∈ R.

Using our hypothesis [u, r]d[u, v] = [u, r][d(u), v] + [u, r][d(v), u] for all u, v ∈
U, r ∈ R. This gives us [u, r][u, d(v)] = 0 for all u, v ∈ U, r ∈ R. Replacing r
by rs for some s in R we get [u, R]R[u, d(v)] = {0} for all u, v ∈ U which is
equation (2.2.1). Therefore, proceeding in the same way as above we obtain
that U ⊆ Z(R).

In view of Lemma 2.2.8 one can easily prove the following:

Theorem 2.2.13 Let R be a 2-torsion free ∗-prime ring and U a ∗-Lie ideal
of R. If d and g are any two derivations such that one of them is non-zero
which commute with ∗. If [g(U), d(U)] = {0} then U ⊆ Z(R).

Theorem 2.2.14 Let R be a 2-torsion free ∗-prime ring and U a ∗-Lie ideal
of R. If F is a generalized derivation with associated non-zero derivation d
which commutes with ∗ such that F (uv) ± uv = 0 for all u, v ∈ U then
U ⊆ Z(R).

Proof We have F (uv) ± uv = 0 for all u, v ∈ U. This can be rewritten as

F (u)v + ud(v) ± uv = 0 for all u, v ∈ U. (2.2.2)

Replacing u by [u, ru] in (2.2.2) and using (2.2.2) we get

[u, r]ud(v) = 0 for all u, v ∈ U, r ∈ R. (2.2.3)

Substituting rs in place of r for some s ∈ R in (2.2.3) we get [u, R]Rud(U) =


{0} for all u ∈ U. If u ∈ S∗(R) ∩ U, then either [u, R] = {0} or ud(U) = {0}
for each fixed u ∈ S∗(R) ∩ U . For any v ∈ U we have v − v∗ ∈ S∗(R) ∩ U and
v + v∗ ∈ S∗(R) ∩ U, thus 2v ∈ S∗ (R) ∩ U. Thus for some fixed v ∈ U, either

16
[2v, R] = {0} or 2vd(U ) = {0}. As R is 2-torsion free we have for some fixed
v ∈ U , either [v, R] = {0} or vd(U ) = {0}. Let A = {v ∈ U | [v, R] = {0}}
and B = {v ∈ U | vd(U ) = {0}}. It can be easily seen that A and B are two
additive subgroups of U whose union is U. Thus using Brauer’s trick we get
A = U or B = U. If A = U, then U ⊆ Z(R). If B = U, then using Lemma
2.3.6 we obtain that U ⊆ Z(R) or U = {0}. Thus in every case we obtain
that U ⊆ Z(R).

Theorem 2.2.15 Let R be a 2-torsion free ∗-prime ring and U a ∗-Lie ideal
of R. If F is a generalized derivation with associated non-zero derivation d
which commutes with ∗ such that d(u)F (v) ± uv = 0 for all u, v ∈ U then
U ⊆ Z(R).

Proof We have d(u)F (v) ± uv = 0 for all u, v ∈ U . Replacing v by [v, rv],


r ∈ R we obtain
d(u)[v, r]d(v) = 0 for all u, v ∈ U, r ∈ R. (2.2.4)
Using Lemma 2.3.6 we find that, [v, r]d(v) = 0 for all v ∈ U, r ∈ R. Substi-
tuting rs for r where s ∈ R we have
[v, R]Rd(v) = {0} for all v ∈ U. (2.2.5)
If v ∈ S∗ (R) ∩ U, then either [v, R] = {0} or d(v) = 0. For any u ∈ U,
u − u∗ ∈ S∗(R) ∩ U. Thus from above [u − u∗ , R] = {0} or d(u − u∗ ) = 0
that is either [u, R] = [u, R]∗ or d(u) = (d(u))∗ . If [u, R] = [u, R]∗ then from
(2.2.5) we have [u, R]Rd(u) = [u, R]∗ Rd(u) = {0} for all u ∈ U. Thus either
[u, R] = {0} or d(u) = 0. Now if d(u) = (d(u))∗ , then again equation (2.2.5)
yields that [u, R] = {0} or d(u) = 0. Using Brauer’s trick we find that either
[u, R] = {0} for all u ∈ U or d(u) = 0 for all u ∈ U. Using Lemma 2.3.5 in
the latter case we find that U ⊆ Z(R). Thus either way we conclude that
U ⊆ Z(R).

2.3 Generalized derivations on ∗-Jordan ideals


A Jordan ideal of a ring equipped with an involution ‘*’ is said to be a ∗-
Jordan ideal if J = J ∗. Awtar [15] showed that for any Jordan ideal J of a
ring R, 4j 2r, 4rj 2 ∈ J for all j ∈ J, r ∈ R. In the same paper he established
that any Jordan ideal J of a prime ring R which is also a subring of R is
central if R admits a centralizing derivation d on J. Very recently Mamouni
and Oukhtite (see [62], [63]) obtained the commutativity of ∗-prime ring R
satisfying various differential identities on ∗-Jordan ideal of R. In this section
we continue with similar investigation and obtain commutativity of ∗-prime
ring satisfying certain identities involving generalized derivations acting on
∗-Jordan ideal of R.

17
In order to develop the proofs of our results we shall require the following
known lemmas;

Lemma 2.3.1 ([76], Lemma 2.4 ) If R is a ring and J a non-zero Jordan


ideal of R, then 2[R, R]J ⊆ J and 2[J, J]R ⊆ J.

Lemma 2.3.2 ([69], Lemma 2 ) Let R be a 2-torsion free ∗-prime ring and
J a non-zero ∗-Jordan ideal of R. If a, b ∈ R such that aJb = {0} = a∗ Jb
then either a = 0 or b = 0.

Lemma 2.3.3 ([62], Lemma 3 ) Let R be a 2-torsion free ∗-prime ring and
J a non-zero ∗-Jordan ideal of R. If d is a non-zero derivation such that
d(x2 ) = 0 for all x ∈ J, then R is commutative.

Lemma 2.3.4 ([63], Lemma 3 ) Let R be a 2-torsion free ∗-prime ring and
J a non-zero ∗-Jordan ideal of R. If R admits a non-zero derivation d such
that [[r, s], y]Jd(y 2 ) = {0} for all r, s ∈ R and y ∈ J, then J ∩ Z(R)
= {0}.

Before we begin with our lemma it is worth mentioning here that

(i) If aJ = {0} or Ja = {0} then a = 0.

(ii) Every ∗-prime ring is semi-prime ring.

Lemma 2.3.5 Let R be a 2-torsion free ∗-prime ring and J be a non-zero


commutative ∗-Jordan ideal of R. Then R is commutative.

Proof We have [u, v] = 0 for all u, v ∈ J. Replacing u by 2u[r, s] we get


u[[r, s], v] = 0 for all u, v ∈ J and r, s ∈ R . This gives us that [[r, s], v] = 0
for all v ∈ J and r, s ∈ R. Replacing r by rs we obtain [r, s][s, v] = 0 for
all v ∈ J and r, s ∈ R. Again on replacement of r by rt for any t ∈ R we
have [r, s]R[s, v] = {0} for all v ∈ J and r, s ∈ R. If s ∈ S∗(R), then we
get [R, s]R[s, v] = {0} for all v ∈ J and s ∈ S∗ (R) and [R, s]∗ R[s, v] = {0}
for all v ∈ J, s ∈ S∗(R). For any s ∈ R, s + s∗ , s − s∗ ∈ S∗(R). Thus for
any, s ∈ R, 2s ∈ S∗(R). Since R is 2-torsion free, replacing s by 2s in the
above equation we conclude that [R, s]R[s, v] = {0} for all v ∈ J and s ∈ R
and [R, s]∗ R[s, v] = {0} for all v ∈ J and s ∈ R. This shows either R is
commutative or [s, v] = 0 for all v ∈ J, s ∈ R. If [s, v] = 0 for all v ∈ J, s ∈ R
then proceeding similarly as above we obtain the desired result.

Lemma 2.3.6 Let R be a 2-torsion free ∗-prime ring and J a non-zero ∗-


Jordan ideal of R. If [u2 , v] = 0 for all u, v ∈ J, then R is commutative.

18
Proof Adopting the similar procedure as in the above Lemma 2.3.5 we easily
obtain that either [u2 , r] = 0 for all u ∈ J or [r, s] = 0 for all s ∈ R. Let
L = {r ∈ R | [u2 , r] = 0 for all u ∈ J} and K = {r ∈ R | [r, s] = 0 for
all s ∈ R} . Then it can be seen that L and K are two additive subgroups
of R whose union is R. Using Brauer’s trick we have either L = R or
K = R. If K = R then we are done. Assume L = R, then [u2 , r] = 0
for all u ∈ J and r ∈ R i.e., u2 ∈ Z(R). Linearizing on u we obtain
0 = [uv + vu, r] = u[v, r] + [u, r]v + v[u, r] + [v, r]u. Replacing u by 2u2 and
using the fact that u2 ∈ Z(R) and R is 2-torsion free we get

u2R[v, r] = {0}, for all u, v ∈ J and r ∈ R. (2.3.1)

Since, J is a ∗-Jordan ideal we get

u2 R[v, r]∗ = {0}, for all u, v ∈ J and r ∈ R. (2.3.2)

Thus equations (2.3.1) and (2.3.2) gives us that either u2 = 0 for all u ∈ J or
[v, R] = {0} for all v ∈ J. If u2 = 0 for all u ∈ J, then on linearizing we obtain
uw + wu = 0 for all u, w ∈ J. Left multiplication by u yields uJu = {0} for
all u ∈ J which gives that J = {0}, a contradiction. Hence, [v, R] = {0} for
all v ∈ J. Thus using Lemma 2.3.5 we get that R is commutative.

Theorem 2.3.7 Let R be a 2-torsion free ∗-prime ring and J a non-zero


∗-Jordan ideal of R. If R admits a generalized derivation F with associated
non-zero derivation d such that F is commuting on J, i.e., [F (u), u] = 0 for
all u ∈ J then R is commutative.

Proof We have [F (u), u] = 0 for all u ∈ J. Linearizing on u we obtain

[F (u), v] + [F (v), u] = 0 for all u, v ∈ J. (2.3.3)

Replacing u by 4uv2 in the equation (2.3.3) and using (2.3.3) we obtain that
u[d(v2 ), v] + [u, v]d(v2 ) = 0 for all u, v ∈ J. Again replacing u by 2[r, s]u for
any r, s ∈ R we get

[[r, s], v]Jd(v 2) = {0} for all v ∈ J and r, s ∈ R. (2.3.4)

Lemma 2.3.4 implies that J ∩ Z(R)


= {0}. Let 0
= z ∈ J ∩ Z(R). Substi-
tuting 4u2 z in place of u in equation (2.3.3) we have [u2 , v]d(z) = 0 for all
u, v ∈ J. As d(z) ∈ Z(R)

[u2 , v]Jd(z) = {0} for all u, v ∈ J. (2.3.5)

Since J is a ∗-Jordan ideal,

[u2 , v]∗Jd(z) = {0} for all u, v ∈ J. (2.3.6)

19
From equations (2.3.5) and (2.3.6) we obtain that either [u2 , v] = 0 for all
u, v ∈ J or d(z) = 0 for any z ∈ J ∩ Z(R). If d(z) = 0 for any z ∈ J ∩ Z(R).
Then replacing u by 4rz 2 in (2.3.3) we get ([F (r), v] + [F (v), r])z 2 = 0 for all
r ∈ R, v ∈ J and z ∈ J ∩ Z(R). This implies that

([F (r), v] + [F (v), r])Jz 2 = {0} for all r ∈ R, v ∈ J and z ∈ J ∩ Z(R).


(2.3.7)
Also since, 0
= z ∗ ∈ J ∩ Z(R) proceeding in the same manner as above we
have

([F (r), v] + [F (v), r])J(z 2 )∗ = {0} for all r ∈ R, v ∈ J and z ∈ J ∩ Z(R).


(2.3.8)
Thus equations (2.3.7) and (2.3.8) implies that [F (r), v]+[F (v), r] = 0 for all
r ∈ R and v ∈ J. Substituting rv in place of r yields r[d(v), v]+[r, v]d(v) = 0
for all r ∈ R and v ∈ J. Again for rt in place of r for any t ∈ R we
get [r, v]Rd(v) = {0} for all r ∈ R and v ∈ J. If v ∈ J ∩ S∗ (R) then,
[r, v]∗ Rd(v) = {0} for all r ∈ R. Since for any v ∈ J, v+v ∗ , v−v ∗ ∈ J ∩S∗ (R).
Hence for any v ∈ J, 2v ∈ J ∩ S∗ (R). As R is 2-torsion free, from the above
equations we conclude that [r, v]Rd(v) = {0} for all r ∈ R and v ∈ J and
[r, v]∗ Rd(v) = {0} for all r ∈ R and v ∈ J. This gives us that either [r, v] = 0
for all r ∈ R and v ∈ J or d(v) = 0, v ∈ J. Using Brauer’s trick we obtain
either [r, v] = 0 for all r ∈ R, v ∈ J or d(v) = 0 for all v ∈ J. If d(v) = 0 for
all v ∈ J then on replacing v by 2v[r, s] we have [d(r), s] + [r, d(s)] = 0 for all
r, s ∈ R. Substituting rt in place of r we obtain d(r)[t, s] + r[d(t), s] = 0 for
all s, t, r ∈ R. Again replacing r by rp we finally get d(r)R[t, s] = {0} for all
r, s, t ∈ R. Also, d(r)R[t, s]∗ = {0} for all r, s, t ∈ R. Since d is a non-zero
derivation on R we find that R is commutative.

Theorem 2.3.8 Let R be a 2-torsion free ∗-prime ring and J a non-zero


∗-Jordan ideal of R. If R admits a generalized derivation F with associated
non-zero derivation d such that any one of the following holds:
(i) F [u, v] = u ◦ v, for all u, v ∈ J,
(ii) F (u ◦ v) = [u, v], for all u, v ∈ J,
(iii) For all u, v ∈ J either F [u, v] = u ◦ v or F (u ◦ v) = [u, v].
Then R is commutative.

Proof The proof of (i) and (ii) follows on a similar basis as above.
(iii) For each fixed u ∈ J, let A = {v ∈ J | F [u, v] = u ◦ v} and B = {v ∈ J |
F (u ◦ v) = [u, v]}. Then A and B both are additive subgroups of J such that
A ∪ B = J. By Brauer’s trick either A = J or B = J. Further by similar
arguments as above we have J = {u ∈ J | A = J} and J = {u ∈ J | B = J}.
Then by (i) and (ii) we obtain the desired result.

20
Theorem 2.3.9 Let R be a 2-torsion free ∗-prime ring and J a non-zero
∗-Jordan ideal of R. If R admits a generalized derivation F with associated
non-zero derivation d such that F (uv) = F (u)F (v) holds for all u, v ∈ J then
R is commutative.

Proof From our hypothesis we obtain that

F (u)v + ud(v) = F (u)F (v) holds for all u, v ∈ J. (2.3.9)

Replacing v by 4vw2 in the above equation we get

(u − F (u))vd(w2 ) = 0 for all u, v, w ∈ J. (2.3.10)

If w ∈ J ∩ S∗ (R) then (u − F (u))vd(w2 )∗ = 0 for all u, v, w ∈ J. Proceeding


in a similar manner as in previous results and replacing w by 2w we obtain
(u − F (u))vd(w 2) = 0 for all u, v, w ∈ J and (u − F (u))vd(w2 )∗ = 0 for
all u, v, w ∈ J. Thus either u = F (u) for all u ∈ J or d(w2 ) = 0 for all
w ∈ J. If u = F (u) for all u ∈ J then equation (2.3.9) yields that ud(v) = 0
for all u, v ∈ J. This implies that d(v)Jd(v) = {0} for all v ∈ J and
d(v)∗Jd(v) = {0} for all v ∈ J. This gives us that d(v) = 0 for all v ∈ J.
Proceeding in a similar fashion as in the end of the proof of the Theorem
2.3.7 we obtain the required result.

Theorem 2.3.10 Let R be a 2-torsion free ∗-prime ring and J a non-zero


∗-Jordan ideal of R. If R admits generalized derivations F and G with asso-
ciated non-zero derivations d and g respectively such that F [u, v] = [G(u), v]
for all u, v ∈ J, then R is commutative.

Proof Using similar arguments as used in the proof of the Theorem 2.3.7 we
obtain the required result.

Theorem 2.3.11 Let R be a 2-torsion free ∗-prime ring and J a non-zero


∗-Jordan ideal of R. If R admits generalized derivations F and G with asso-
ciated non-zero derivations d and g respectively such that F (u)G(v) ± uv = 0
for all u, v ∈ J, then R is commutative.

Proof We have F (u)G(v)±uv = 0 for all u, v ∈ J. Substituting 4vw2 in place


of v we obtain 0 = F (u)G(vw2 ) ± uvw2 = F (u)G(v)w2 + F (u)vg(w 2) − uvw2
for all u, v, w ∈ J. Using the hypothesis we get F (u)vg(w2 ) = 0 for all
u, v ∈ J. Using the same arguments as in the previous results, we get either
F (u) = 0 or g(w2 ) = 0. If F (u) = 0 then our hypothesis yields that uv = 0
for all v ∈ J. Consequently, we obtain u = 0. Therefore, g(w2 ) = 0 for all
w ∈ J. Implementing Lemma 2.3.3 we get that R is commutative.

21
2.4 Generalized (α, β)∗ -derivation
Let R be a ∗-ring. An additive mapping d : R → R is called a ∗-derivation
(resp. ∗- reverse derivation) if d(xy) = d(x)y∗ +xd(y) (resp. d(xy) = d(y)x∗ +
yd(x)) holds for all x, y ∈ R. Following [2], an additive mapping F : R → R
is called a generalized ∗-derivation (resp. generalized ∗- reverse derivation) if
there exists a ∗-derivation (resp. ∗- reverse derivation) d such that F (xy) =
F (x)y∗ + xd(y) (resp. F (xy) = F (y)x∗ + yd(x)) holds for all x, y ∈ R. Let
α, β be endomorphisms of R. An additive mapping d : R → R is called
a (α, β)∗-derivation (resp. (α, β)∗ -reverse derivation) if d(xy) = d(x)α(y ∗) +
β(x)d(y) (resp. d(xy) = d(y)α(x) + β(y)d(x)) holds for all x, y ∈ R. Now we
extend the concepts of generalized ∗-derivation (resp. generalized ∗-reverse
derivation) in the following way: an additive mapping F : R → R is called a
generalized (α, β)∗ -derivation (resp. generalized (α, β)∗ - reverse derivation) if
there exists a (α, β)∗-derivation (resp. (α, β)∗ -reverse derivation)d such that
F (xy) = F (x)α(y ∗ )+β(x)d(y) (resp. F (xy) = F (y)α(x)+β(y)d(x)) holds for
all x, y ∈ R. Note that for IR (the identity map on R), generalized (IR , IR )∗ -
derivation (resp. generalized (IR , IR )∗ -reverse derivation) is a generalized
∗-derivation (resp. generalized ∗- reverse derivation). Thus, the concept
of generalized (α, β)∗-derivation covers the concepts of (α, β)∗ -derivations.
Moreover, generalized (α, β)∗ -derivation with d = 0 covers the concept of
left α∗ -multipliers i.e., additive maps F satisfying F (xy) = F (x)α(y ∗) for
all x, y ∈ R. In [25], Bresar and Vukman proved that if a prime ring R
with involution ‘*’ admits a ∗-derivation (resp. ∗- reverse derivation) d,
then either R is commutative or d = 0. Further, Ashraf and Shakir in [6]
extended the above mentioned result for semi-prime ring with involution ‘*’
in the setting of (α, β)∗ -derivations. Recently, Shakir [2] proved that if a
semi-prime ring with involution ‘*’ admits a generalized ∗-derivation (resp.
generalized ∗-reverse derivation) F, then F maps R into Z(R). More precisely,
he established that if there exists a generalized ∗-derivation on a semi-prime
ring with involution ‘*’ , then F maps R into Z(R). The aim of the present
section is to extend these results for generalized (α, β)∗-derivations in semi-
prime ring with involution ‘*’ . We begin by proving our first result.
Theorem 2.4.1 Let R be a semi-prime ring with involution ‘*’ and α, β be
the endomorphisms of R. If F : R → R is a generalized (α, β)∗ -derivation
such that α is surjective, then F maps R into Z(R).
Proof We are given that F (xy) = F (x)α(y ∗ ) + β(x)d(y) for all x, y ∈ R. Let
us first consider for all x, y, z ∈ R
F (xyz) = F (x(yz)) = F (x)α(z ∗ y∗ ) + β(x)d(y)α(z ∗ ) + β(xy)d(z). (2.4.1)
On the other hand for all x, y, z ∈ R,
F (xyz) = F ((xy)z) = F (x)α(y ∗z ∗ ) + β(x)d(y)α(z ∗ ) + β(xy)d(z). (2.4.2)

22
Comparing (2.4.1) and (2.4.2) we obtain F (x)[α(y∗ ), α(z ∗ )] = 0. Replacing y
by y∗ and z by z ∗ and using the fact that α is surjective, we get

F (x)[y, z] = 0 for all x, y, z ∈ R. (2.4.3)

Now replacing y by yF (x) in the above relation we obtain

F (x)y[F (x), z] = 0 for all x, y, z ∈ R. (2.4.4)


Substituting zy for y in equation ( 2.4.4) we have

F (x)zy[F (x), z] = 0 for all x, y, z ∈ R. (2.4.5)


Again left multiplying equation (2.4.3) by z we obtain

zF (x)y[F (x), z] = 0 for all x, y, z ∈ R. (2.4.6)


On combining the relations (2.4.5) and (2.4.6) we find that
[F (x), z]y[F (x), z] = 0 for all x, y, z ∈ R. Semiprimeness of R forces
us that [F (x), z] = 0 for all x, z ∈ R. Hence we conclude that F maps R
into Z(R).

Theorem 2.4.2 Let R be a semi-prime ring with involution ‘*’ and α, β be


endomorphisms of R. If F : R → R is a generalized (α, β)∗ -reverse derivation
with associated (α, β)∗-reverse derivation d such that β is surjective then d
maps R into Z(R).

Proof By our hypothesis we have F (xy) = F (y)α(x∗ ) + β(y)d(x) holds for


all x, y ∈ R. On one hand we have for all x, y, z ∈ R

F (xyz) = F (x(yz)) = F (z)α(y ∗x∗ ) + β(z)d(y)α(x∗ ) + β(yz)d(x). (2.4.7)

Next consider for all x, y, z ∈ R

F (xyz) = F ((xy)z) = F (z)α(y ∗x∗ ) + β(z)d(y)α(x∗ ) + β(zy)d(x). (2.4.8)

Comparing the equations (2.4.7) and (2.4.8) we obtain that [β(y), β(z)]d(x) =
0 for all x, y, z ∈ R. Since β is surjective we get

[y, z]d(x) = 0 holds for all x, y, z ∈ R.

Proceeding in a similar way as in the proof of the Theorem 2.4.1 we obtain


the required result.

Theorem 2.4.3 Let R be a non-commutative prime ring with involution ‘*’


and α, β be endomorphisms of R. If F : R → R is a generalized (α, β)∗-
derivation with associated (α, β)∗-derivation d such that α is surjective then
F = 0.

23
Proof In view of equation (2.4.3) we can directly have F (x)[y, z] =
0 for all x, y, z ∈ R. Replacing y by yr we get F (x)y[r, z] = 0 for all x, y, z ∈
R. Since R is non-commutative prime ∗-ring, we obtain that F (x) = 0 for all
x ∈ R.
Using the similar arguments as above with necessary variations, we can
prove the next theorem.

Theorem 2.4.4 Let R be a non-commutative prime ring with involution ‘*’


and α, β be the endomorphisms of R. If F : R → R is a generalized (α, β)∗-
reverse derivation with associated (α, β)∗-reverse derivation d such that β is
surjective then F is left α∗ -multipliers on R .

Following are the immediate consequences of the above theorems.

Corollary 2.4.5 ([2], Theorem 2.2) Let R be a semi-prime ring with invo-
lution ‘*’ , and α be the endomorphism of R such that α is surjective. If
T : R −→ R is an additive mapping such that T (xy) = T (x)α(y∗ ) for all
x, y ∈ R, then T maps R into Z(R).

Corollary 2.4.6 ([2], Theorem 2.3) Let R be a semi-prime ring with invo-
lution ‘*’ . If R admits a generalized ∗-derivation F with associated non-zero
∗ -derivation d, then F maps R into Z(R).

The following example shows that primeness of ring is crucial in proving the
Theorem 2.4.3.
 0 a b  
Example 2.4.7 S be any ring. Let R = 0 0 c | a, b, c ∈ S .
0 0 0
 0 a b   0 a −b 
Define α, β, ∗ : R → R such that α 0 0 c = 0 0 −c ,
0 0 0 0 0 0
 0 a b   0 −a −b   0 a b ∗  0 c b 
β 0 0 c = 0 0 c and 0 0 c = 0 0 a .
0 0 0 0 0 0 0 0 0 0 0 0
 0 a b   0 0 a 
Let F, d be the mappings on R such that F 0 0 c = 0 0 0
0 0 0 0 0 0
 0 a b   0 0 c 
and d 0 0 c = 0 0 0 .
0 0 0 0 0 0
It is obvious to see that F is a non-zero generalized (α, β)∗ -derivation with
associated (α, β)∗-derivation d.

24

Potrebbero piacerti anche