Loading [MathJax]/jax/output/SVG/jax.js
Research article Special Issues

Additivity of nonlinear higher anti-derivable mappings on generalized matrix algebras

  • In this article, we proved that each nonlinear higher anti-derivable mapping on generalized matrix algebras is automatically additive. As for its applications, we find a similar conclusion on triangular algebras, full matrix algebras, unital prime rings with a nontrivial idempotent, unital standard operator algebras and factor von Neumann algebras respectively.

    Citation: Xiuhai Fei, Haifang Zhang. Additivity of nonlinear higher anti-derivable mappings on generalized matrix algebras[J]. Electronic Research Archive, 2023, 31(11): 6898-6912. doi: 10.3934/era.2023349

    Related Papers:

    [1] Xiuhai Fei, Cuixian Lu, Haifang Zhang . Nonlinear Jordan triple derivable mapping on $ * $-type trivial extension algebras. Electronic Research Archive, 2024, 32(3): 1425-1438. doi: 10.3934/era.2024066
    [2] Fei Ma, Min Yin, Yanhui Teng, Ganglian Ren . Nonlinear generalized semi-Jordan triple derivable mappings on completely distributive commutative subspace lattice algebras. Electronic Research Archive, 2023, 31(8): 4807-4817. doi: 10.3934/era.2023246
    [3] Hongliang Chang, Yin Chen, Runxuan Zhang . A generalization on derivations of Lie algebras. Electronic Research Archive, 2021, 29(3): 2457-2473. doi: 10.3934/era.2020124
    [4] Chao Yang, Juntao Wu, Zhengyang Qiao . An improved fixed-time stabilization problem of delayed coupled memristor-based neural networks with pinning control and indefinite derivative approach. Electronic Research Archive, 2023, 31(5): 2428-2446. doi: 10.3934/era.2023123
    [5] Shasha Bian, Yitong Pei, Boling Guo . Numerical simulation of a generalized nonlinear derivative Schrödinger equation. Electronic Research Archive, 2022, 30(8): 3130-3152. doi: 10.3934/era.2022159
    [6] Daochang Zhang, Dijana Mosić, Liangyun Chen . On the Drazin inverse of anti-triangular block matrices. Electronic Research Archive, 2022, 30(7): 2428-2445. doi: 10.3934/era.2022124
    [7] Weiwei Qi, Yongkun Li . Weyl almost anti-periodic solution to a neutral functional semilinear differential equation. Electronic Research Archive, 2023, 31(3): 1662-1672. doi: 10.3934/era.2023086
    [8] Jiarui Chen, Aimin Tang, Guanfeng Zhou, Ling Lin, Guirong Jiang . Walking dynamics for an ascending stair biped robot with telescopic legs and impulse thrust. Electronic Research Archive, 2022, 30(11): 4108-4135. doi: 10.3934/era.2022208
    [9] Kailash C. Misra, Sutida Patlertsin, Suchada Pongprasert, Thitarie Rungratgasame . On derivations of Leibniz algebras. Electronic Research Archive, 2024, 32(7): 4715-4722. doi: 10.3934/era.2024214
    [10] Mingliang Song, Dan Liu . Common fixed and coincidence point theorems for nonlinear self-mappings in cone $ b $-metric spaces using $ \varphi $-mapping. Electronic Research Archive, 2023, 31(8): 4788-4806. doi: 10.3934/era.2023245
  • In this article, we proved that each nonlinear higher anti-derivable mapping on generalized matrix algebras is automatically additive. As for its applications, we find a similar conclusion on triangular algebras, full matrix algebras, unital prime rings with a nontrivial idempotent, unital standard operator algebras and factor von Neumann algebras respectively.



    In past decades, research of derivations and nonlinear derivable mappings on algebras has attracted the attention of many mathematicians.

    Definition 1.1 Let R be a commutative ring with identity and A a unital algebra over R, and N the set of non-negative integers, i,j,k,nN.

    (1) If Δ is an additive mapping such that

    Δ(XY)=Δ(X)Y+XΔ(Y) (1.1)

    for all X,YA, then Δ is said to be a derivation. If Δ is not necessarily additive and Eq (1.1) hold for all X,YA, then Δ is said to be a nonlinear derivable mapping.

    (2) For any X,YA, call XY:=XY+YX the Jordan product of X,Y. If Δ is an additive mapping such that

    Δ(XY)=Δ(X)Y+XΔ(Y)

    for all X,YA, then Δ is said to be a Jordan derivation. Nonlinear Jordan derivation is defined similarly to the nonlinear derivable mapping.

    (3) Let D={dn}nN be a sequence of additive mappings (resp., without assumption of additivity) on A with d0=idA the identity mapping on A such that

    dn(XY)=i+j=ndi(X)dj(Y)

    for all nN, and X,YA, then D is said to be a higher derivation (resp., nonlinear higher derivable mapping).

    Obviously, every additive derivation is an additive Jordan derivation, and every additive higher derivation is an additive Jordan higher derivation. However, the inverse statement is not true in general. It is natural to ask the following two questions:

    Problem 1 Under what conditions is a Jordan (higher) derivation is a (higher) derivation?

    Problem 2 Under what conditions is a nonlinear (Jordan, higher) derivation is a (Jordan, higher) derivation?

    There are many works that consider Problem 1. For example, see [1,2,3,4,5]. In this paper, we focus on Problem 2. Rickart [6] proved that, under certain conditions, any one-to-one and multiplicative mapping from a ring into another ring is necessarily additive. Martindale [7] obtained the result that each multiplicative bijective mapping on an arbitrary algebra which contains a nontrivial idempotent is automatically additive. For other similar results about additivity of multiplicative mappings on rings or algebras, we refer the readers to [8,9,10,11] and references therein for more details. Daif [12] showed that, under certain conditions, any multiplicative derivation is additive. Later, Daif [13] extended this result to the case of multiplicative generalized derivation. Lu [14] proved that, under some conditions, every multiplicative Jordan derivation on a prime ring is an additive derivation. For more similar results about additivity of nonlinear Jordan derivable on rings or algebras, see [15,16] and references therein. Fu and Xiao [17] and Ashraf and Jabeen [18] showed that all nonlinear Jordan higher derivable mappings and nonlinear Jordan higher triple derivable mappings on triangular algebras is an additive higher derivation, respectively.

    In [1], Benkovič defined anti-derivations on algebras as the following.

    Definition 1.2 Let C be a commutative ring with unity, A an algebra over C and M an A-bimodule. Let δ:AM be a linear map. If

    δ(ab)=δ(b)a+bδ(a)

    for all a,bA, then δ is said an anti-derivation. For more results about anti-derivation on rings or algebras, see [19,20] and references therein.

    Motivated by the above definition, we introduce the following higher anti-derivation.

    Definition 1.3 Let C be a commutative ring with unity, and A be an algebra over C. Let D={δn}nN be a sequence of additive maps from A into itself with δ0=idA. If

    δn(ab)=i+j=nδi(b)δj(a)

    for all a,bA and all nN, then D is called a higher anti-derivation. If δn is not necessarily additive, then D is called a non-linear higher anti-derivable mapping.

    Our main purpose in this paper is to show that every nonlinear higher anti-derivable mapping on a generalized matrix algebra is additive. In the following section, we introduce some basic concepts and the properties of generalized matrix algebras we require. Generalized matrix algebra is a particular structure of generalized n-matrix rings (see for example [21]), if we do not consider the scalar multiplication.

    Let R be a commutative ring with identity, A and B be two unital R-algebras, and 1A and 1B be the unit elements of A and B respectively. Let M be a faithful (A,B)-bimodule (i.e., for any AA,BB, if AM=0, then A=0; if MB=0, then B=0), and N be a (not necessarily faithful) (B,A)-bimodule. Suppose that there are two bimodule homomorphisms ΦMN:MBNA and ΨNM:NAMB satisfying the following associativity conditions: (MN)M=M(NM) and (NM)N=N(MN) for all M,MM,N,NN, where MN=ΦMN(MBN) and NM=ΨNM(NAM). Then

    G(A,M,N,B)=(AMNB)={(AMNB):AA,MM,NN,BB}

    is an R-algebra under the usual matrix-like addition, and the following multiplication:

    (AMNB)(AMNB)=(AA+ΦMN(MN)AM+MBNA+BNBB+ΨNM(NM))

    for all A,AA,M,MM,N,NN and B,BB, where at least one of the two bimodules M and N is distinct from zero. Such an R-algebra is called a generalized matrix algebra. This type of algebra was first introduced by Morita [22]. In the following, we simply write G(A,M,N,B) as G. For any associative algebra A, if A is unital with the identity 1A, and has a non-trivial idempotent P (P2=P, P0 and P1A), then the Peirce decomposition of A corresponding to P is A=PAP+PAQ+QAP+QGQ, where Q=1AP. With respect to this decomposition, A is a generalized matrix algebra, and we then know that any an associative algebra containing a non-trivial idempotent is a generalized algebra.

    Consider a generalized matrix algebra G, let 1 be the unit of G. Set

    P1=(1A000), P2=1P1=(0001B)

    and Gij=PiGPj(1i,j2). Then, G can be represented as

    G=G11+G12+G21+G22,

    where G11 is a subalgebra of G isomorphic to A, G22 is a subalgebra of G isomorphic to B, G12 is a (G11,G22)-bimodule isomorphic to M, and G21 is a (G22,G11)-bimodule isomorphic to N. Thus, G12 is a faithful (G11,G22)-bimodule. Furthermore, for any AG, A can be represented as A=A11+A12+A21+A22, where AijGij (1ij2).

    In this section, our main result is the following Theorem 2.1. In [12], Daif proved every multiplicative derivation on a ring having an idempotent element which satisfies some conditions is additive. It is not hard to see an anti-derivation on a generalized matrix algebra G is a derivation from G into its anti-algebra. However, Theorem 2.1 is not a direct corollary of the theorem in [12]. This is because there are no idempotent elements in a generalized matrix algebra satisfying the conditions in [12]. Further, in [23], Ferreira and Sandhu showed that multiplicative anti-derivations are additive on generalized n-matrix rings. When n=2, a generalized n-matrix ring is just the generalized matrix ring. However, these results are not the same as the following results.

    Theorem 2.1 Let G be a generalized matrix algebra, and φ be a mapping of G (without assumption of additivity). If φ satisfies

    φ(XY)=φ(Y)X+Yφ(X) (2.1)

    for all X,YG, then φ is additive.

    In order to prove Theorem 2.1, we introduce Lemmas 2.1–2.4, and then prove that Lemmas 2.1–2.4 hold.

    Lemma 2.1 If φ is an nonlinear anti-derivable mapping on G, then

    (ⅰ) φ(0)=0;

    (ⅱ) φ(P1)=P1φ(P1)P2+P2φ(P1)P1;

    (ⅲ) φ(P2)=P1φ(P2)P2+P2φ(P2)P1;

    (ⅳ) φ(P1)=φ(P2).

    Proof (ⅰ) Taking X=Y=0 in Eq (2.1), we have φ(0)=φ(0)0+0φ(0)=0, and so φ(0)=0.

    (ⅱ) Taking X=P1,Y=P1 in Eq (2.1), we get φ(P1)=φ(P1)P1+P1φ(P1), which implies that P1φ(P1)P1=P2φ(P1)P2=0. Hence, we obtain that φ(P1)=P1φ(P1)P2+P2φ(P1)P1. Similarly, we can show (ⅲ) holds.

    (ⅳ) Taking X=P1,Y=P2 in Eq (2.1), we get

    0=φ(P1P2)=φ(P2)P1+P2φ(P1)=P2φ(P2)P1+P2φ(P1)P1.

    Similarly, we get

    0=φ(P2P1)=φ(P1)P2+P1φ(P2)=P1φ(P1)P2+P1φ(P2)P2.

    Adding the above two equations, it follows from Lemma 2.1 (ⅱ) and (ⅲ) that

    0=P1φ(P1)P2+P2φ(P1)P1+P1φ(P2)P2+P2φ(P2)P1=φ(P1)+φ(P2).

    Therefore, φ(P1)=φ(P2). The proof is completed.

    Lemma 2.2 If φ is an nonlinear anti-derivable mapping on G, then for all A11G11,A12G12,A21G21,A22G22,

    (ⅰ) φ(A12)=P2φ(A12)P1;

    (ⅱ) φ(A21)=P1φ(A21)P2;

    (ⅲ) φ(A11)=P1φ(A11)P2+P2φ(A11)P1;

    (ⅳ) φ(A22)=P1φ(A22)P2+P2φ(A22)P1;

    (ⅴ) φ(P1)A12=φ(P2)A12=A12φ(P1)=A12φ(P2)=0;

    (ⅵ) φ(P1)A21=φ(P2)A21=A21φ(P1)=A21φ(P2)=0.

    Proof (ⅰ) For any A12G12, taking X=P1,Y=A12 in Eq (2.1), we have

    φ(A12)=φ(P1A12)=φ(A12)P1+A12φ(P1). (2.2)

    This yields from P2φ(P1)P2=0 that

    P1φ(A12)P2=P2φ(A12)P2=0.

    Similarly, we have

    φ(A12)=φ(A12P2)=φ(P2)A12+P2φ(A12). (2.3)

    This implies that

    P1φ(A12)P1=0.

    Therefore, we get φ(A12)=P2d1(A12)P1. Similarly, we can show that (ⅱ) holds.

    (ⅲ) For any A11G11,A12G12, taking X=A12,Y=A11 in Eq (2.1), then by Lemma 2.2 (ⅰ), we have

    0=φ(A12A11)=φ(A11)A12+A11φ(A12)=φ(A11)A12.

    This yields from the faithfulness of G12 that

    P1φ(A11)P1=0.

    Similarly, taking X=P2,Y=A11 in Eq (2.1), we have

    0=φ(P2A11)=φ(A11)P2+A11φ(P2).

    This implies that

    P2φ(A11)P2=0.

    Therefore, we obtain that φ(A11)=P1φ(A11)P2+P2φ(A11)P1. Similarly, we can show that (ⅳ) holds.

    (ⅴ) For any A12G12, it follows from Eqs (2.2)–(2.3) and φ(A12)=P2φ(A12)P1 that

    P1φ(A12)=0=A12φ(P1)  and  φ(A12)P2=0=φ(P2)A12.

    Therefore, we obtain from φ(P1)=φ(P2) that φ(P1)A12=φ(P2)A12=A12φ(P1)=A12φ(P2)=0. Similarly, we can show that (ⅵ) holds.

    Lemma 2.3 If φ is a nonlinear anti-derivable mapping on G, then for all A11,B11G11,A12,B12G12,A21,B21G21,A22,B22G22,

    (ⅰ) φ(A11+B11)=φ(A11)+φ(B11);

    (ⅱ) φ(A22+B22)=φ(A22)+φ(B22);

    (ⅲ) φ(A11+A12)=φ(A11)+φ(A12);

    (ⅳ) φ(A12+A22)=φ(A12)+φ(A22);

    (ⅴ) φ(A21+A22)=φ(A21)+φ(A22);

    (ⅵ) φ(A12+B12)=φ(A12)+φ(B12);

    (ⅶ) φ(A21+B21)=φ(A21)+φ(B21).

    Proof (ⅰ) For any A11,B11G11, taking X=A11,Y=P1 in Eq (2.1), we have

    P2φ(A11)P1=φ(P1)A11.

    Taking X=P2,Y=A11 in Eq (2.1), we have 0=φ(P2A11)=φ(A11)P2+A11φ(P2), this yields from φ(P1)=φ(P2) that

    P1φ(A11)P2=A11φ(P2)=A11φ(P1).

    Hence, we can get from above two equations and φ(A11)=P1φ(A11)P2+P2φ(A11)P1 that

    φ(A11)=A11φ(P1)+φ(P1)A11.

    Similarly, we get

    φ(B11)=B11φ(P1)+φ(P1)B11.

    And,

    φ(A11+B11)=(A11+B11)φ(P1)+φ(P1)(A11+B11).

    Therefore, it follows from above three equations that φ(A11+B11)=φ(A11)+φ(B11). Similarly, we can show (ⅱ) holds.

    (ⅲ) For any A11G11,A12G12, taking X=A11+A12,Y=P1 in Eq (2.1), we get that

    φ(A11)=φ((A11+A12)P1)=φ(P1)(A11+A12)+P1φ(A11+A12)=φ(P1)A11+P1φ(A11+A12).

    Similarly, taking X=A11+A12,Y=P2 in Eq (2.1), by Lemma 2.2 (ⅴ), we have

    φ(A12)=φ((A11+A12)P2)=φ(P2)(A11+A12)+P2φ(A11+A12)=φ(P2)A11+P2φ(A11+A12).

    Adding the above two equations, we then obtain from φ(P1)=φ(P2) that φ(A11+A12)=φ(A11)+φ(A12). Similarly, we can show (ⅳ) and (ⅴ) hold.

    (ⅵ) For any A12,B12G12, taking X=A12,Y=B12 in Eq (2.1), it follows from A12B12=0 that

    0=φ(A12B12)=φ(B12)A12+B12φ(A12). (2.4)

    Since A12+B12=(P1+A12)(P2+B12), we take X=P1+A12,Y=P2+B12 in Eq (2.1), and then we get from Lemma 2.2, Lemma 2.3(ⅰ), (ⅴ), Lemma 2.4(ⅲ)-(ⅳ) and Eq (2.4) that

    φ(A12+B12)=φ((P1+A12)(P2+B12))=φ(P2+B12)(P1+A12)+(P2+B12)φ(P1+A12)=(φ(P2)+φ(B12))(P1+A12)+(P2+B12)(φ(P1)+φ(A12))=φ(P2)P1+φ(P2)A12+φ(B12)P1+φ(B12)A12+P2φ(P1)+P2φ(A12)+B12φ(P1)+B12φ(A12)=P2φ(A12)+φ(B12)P1=P2φ(A12)P1+P2φ(B12)P1=φ(A12)+φ(B12).

    Similarly, we can show (vii) holds. The proof is completed.

    Lemma 2.4 If φ is a nonlinear anti-derivable mapping on G, then φ(A11+A12+A21+A22)=φ(A11)+φ(A12)+φ(A21)+φ(A22) for all A11G11,A12G12,A21G21 and A22G22.

    Proof For any A11G11,A12G12,A21G21,A22G22, taking X=P1,Y=A11+A12+A21+A22 in Eq (2.1), we get from Lemma 2.3 (ⅲ), (ⅴ) and Lemma 2.2(ⅴ)-(ⅵ) that

    φ(A11)+φ(A12)=φ(P1(A11+A12+A21+A22))=φ(A11+A12+A21+A22)P1+(A11+A12+A21+A22)φ(P1)=φ(A11+A12+A21+A22)P1+(A11+A22)φ(P1).

    Similarly, we obtain that

    φ(A21)+φ(A22)=φ(P2(A11+A12+A21+A22))=φ(A11+A12+A21+A22)P2+(A11+A12+A21+A22)φ(P2)=φ(A11+A12+A21+A22)P2+(A11+A22)φ(P2).

    Adding the above two equations, using φ(P1)=φ(P2), we get φ(A11+A12+A21+A22)=φ(A11)+φ(A12)+φ(A21)+φ(A22). The proof is completed.

    Now, we complete the proof of Theorem 2.1.

    Proof of Theorem 2.1 For any X,YG, set X=A11+A12+A21+A22 and Y=B11+B12+B21+B22, where Aij,BijGij(1ij2), then by Lemmas 2.3 and 2.4, we obtain that

    φ(X+Y)=φ((A11+A12+A21+A22)+(B11+B12+B21+B22))=φ((A11+B11)+(A12+B12)+(A21+B21)+(A22+B22))=φ(A11+B11)+φ(A12+B12)+φ(A21+B21)+φ(A22+B22)=φ(A11)+φ(B11)+φ(A12)+φ(B12)+φ(A21)+φ(B21)+φ(A22)+φ(B22)=φ(A11+A12+A21+A22)+φ(B11+B12+B21+B22)=φ(X)+φ(Y).

    Therefore, φ is an additive mapping on G. The proof is completed.

    Next, we will give the second main result.

    Theorem 3.1 Let G be a generalized matrix algebra and D={dn}nN be a sequence mapping from G into itself (without assumption of additivity) such that

    dn(XY)=i+j=ndi(Y)dj(X) (3.1)

    for any nN,X,YG, then D is an additive mapping on G.

    In the following, to prove Theorem 3.1, we will introduce Lemmas 3.1–3.3, and then use mathematical induction to prove that Lemmas 3.1–3.3 hold. We assume that G is a generalized matrix algebra, and D={dn}nN is a higher anti-derivable mapping on G. Let N be the set of non-negative integers, N+ be the set of positive integers, and i,j,k,p,q,nN. For any X,YG, A11G11,A12G12,A21G21,A22G22. We say a map f:GG satisfies the set of properties L, if

    (ⅰ) f(X+Y)=f(X)+f(Y);

    (ⅱ) f(0)=0,f(P1)=f(P2)M+N;

    (ⅲ) f(A12)=P2f(A12)P1;

    (ⅳ) f(P1)A12=f(P2)A12=A12f(P1)=A12f(P2)=0;

    (ⅴ) f(A21)=P1f(A21)P2;

    (ⅵ) f(P1)A21=f(P2)A21=A21f(P1)=A21f(P2)=0;

    (ⅶ) f(A11)=P1f(A11)P2+P2f(A11)P1;

    (ⅷ) f(A22)=P1f(A22)P2+P2f(A22)P1.

    It is known from Theorem 2.1 that d1 satisfies the set of properties L. Now, for any X,YG, A11G11,A12G12,A21G21,A22G22, we assume that dk(1k<n) satisfies the set of properties L. In the following, we show dn satisfies the set of properties L.

    Lemma 3.1 For any nN+,A11G11,A12G12,A21G21,A22G22, dn satisfies the set of properties L.

    Proof (ⅰ) For any nN+, taking X=Y=0 in Eq (3.1), it follows from the set of properties L(ⅱ) that

    dn(0)=i+j=ndi(0)dj(0)=i+j=n,1i,jdi(0)dj(0)+dn(0)0+0dn(0)=0.

    For any n,i,jN+(i,j<n), since di(P1),dj(P1),di(P2),dj(P2)M+N, and so by the set of properties L (ⅳ) and (ⅵ), we get that

    di(P1)dj(P1)=di(P2)dj(P2)=di(P1)dj(P2)=di(P2)dj(P1)=0. (3.2)

    Taking X=P1,Y=P1 in Eq (3.1), by Eq (3.2), we get

    dn(P1)=i+j=ndi(P1)dj(P1)=i+j=n,1i,jdi(P1)dj(P1)+dn(P1)P1+P1dn(P1)=dn(P1)P1+P1dn(P1).

    This implies that

    P1dn(P1)P1=P2dn(P1)P2=0. (3.3)

    Similarly, we have

    P1dn(P2)P1=P2dn(P2)P2=0. (3.4)

    Taking X=P1,Y=P2 in Eq (3.1), by Eq (3.2), we get

    0=i+j=ndi(P2)dj(P1)=i+j=n,1i,jdi(P1)dj(P2)+dn(P2)P1+P2dn(P1)=dn(P2)P1+P2dn(P1).

    Therefore, we get

    P2dn(P2)P1=P2dn(P1)P1. (3.5)

    Similarly, we obtain that

    P1dn(P2)P2=P1dn(P1)P2. (3.6)

    Therefore, by Eqs (3.3)–(3.6), we get that dn(P1)=dn(P2)M+N.

    (ⅱ)-(ⅲ) For any nN+, A12G12, taking X=P1,Y=A12 in Eq (3.1), it follows from (ⅲ) and (ⅵ) of the set of properties L that

    dn(A12)=dn(P1A12)=i+j=ndi(A12)dj(P1)=i+j=n,1i,jP2di(A12)P1dj(P1)+dn(A12)P1+A12dn(P1)=dn(A12)P1+A12dn(P1).

    This yields from P2dn(P1)P2=0 that

    P2dn(A12)P2=P1dn(A12)P2=A12dn(P1)=0. (3.7)

    Similarly, we get

    dn(A12)=dn(A12P2)=i+j=ndi(P2)dj(A12)=i+j=n,1i,jdi(P2)P2dj(A12)P1+dn(P2)A12+P2dn(A12)=dn(P2)A12+P2dn(A12).

    This yields that

    P1dn(A12)P1=dn(P2)A12=0. (3.8)

    Therefore, by dn(P1)=dn(P2) and Eqs (3.7) and (3.8), we get (ⅱ) and (ⅲ). Similarly, we can show that (ⅳ) and (ⅴ) hold.

    (ⅵ) For any nN+,A11G11,A12G12, taking X=A11,Y=A12 in Eq (3.1), it follows from the set of properties L (ⅱ) and Lemma 3.1 (ⅱ) that

    0=dn(A12A11)=i+j=ndi(A11)dj(A12)=i+j=n,1i,jdi(A11)(P2dj(A12)P1)+dn(A11)A12+A11dn(A12)=i+j=n,1i,jP1di(A11)P2dj(A12)P1+dn(A11)A12.

    This implies that P1dn(A11)P1A12=0, and so by the faithfulness of G12, we get

    P1dn(A11)P1=0 (3.9)

    Taking X=A11,Y=P1 in Eq (3.1), we get from (ⅳ), (ⅵ) and (vii) of the set of properties L and Lemma 3.1 (vii) that

    dn(A11)=dn(A11P1)=i+j=ndi(P1)dj(A11)=i+j=n,1i,jdi(P1)(P1dj(A11)P2+P2dj(A11)P1)+dn(P1)A11+P1dn(A11)=dn(P1)A11+P1dn(A11).

    This yields that

    P2dn(A11)P2=0  and  P2dn(A11)P1=dn(P1)A11. (3.10)

    Similarly, taking X=P2,Y=A11 in Eq (3.1), we get from (ⅳ), (ⅵ) and (vii) of the set of properties L and Lemma 3.1 (vii) that

    0=dn(P2A11)=i+j=ndi(A11)dj(P2)=i+j=n,1i,j(P1di(A11)P2+P2di(A11)P1)dj(P2)+dn(A11)P2+A11dn(P2)=dn(A11)P2+A11dn(P2).

    This yields that

    P1dn(A11)P2=A11dn(P2)=A11dn(P1). (3.11)

    Therefore, we get from Eqs (3.9)–(3.11) that

    dn(A11)=P1dn(A11)P2+P2dn(A11)P1=A11dn(P1)+dn(P1)A11. (3.12)

    Similarly, we can show that (viii) holds. The proof is completed.

    Lemma 3.2 For any nN+, A11,B11G11,A12,B12G12,A21,B21G21,A22,B22G22, then

    (ⅰ) dn(A11+B11)=dn(A11)+dn(B11);

    (ⅱ) dn(A22+B22)=dn(A22)+dn(B22);

    (ⅲ) dn(A11+A12)=dn(A11)+dn(A12);

    (ⅳ) dn(A12+A22)=dn(A12)+dn(A22);

    (ⅴ) dn(A21+A22)=dn(A21)+dn(A22);

    (ⅵ) dn(A12+B12)=dn(A12)+dn(B12);

    (vii) dn(A21+B21)=dn(A21)+dn(B21).

    Proof (ⅰ) For any nN+,A11,B11G11, we get from Eq (3.12) that

    dn(A11+B11)=(A11+B11)dn(P1)+dn(P1)(A11+B11)=(A11dn(P1)+dn(P1)A11)+(B11dn(P1)+dn(P1)B11)=dn(A11)+dn(B11)

    Similarly, we show that (ⅱ) holds.

    (ⅲ) For any nN+,A11G11,A12G12, taking X=A11+A12,Y=P1 in Eq (3.11), we get from the set of properties L (ⅰ) and Lemma 3.1 that

    dn(A11)=dn((A11+A12)P1)=i+j=ndi(P1)dj(A11+A12)=i+j=n,1i,jdi(P1)(dj(A11)+dj(A12))+dn(P1)(A11+A12)+P1dn(A11+A12)=i+j=n,1i,jdi(P1)dj(A11)+dn(P1)A11+P1dn(A11+A12)=i+j=n,1i,jdi(P1)(P1dj(A11)P2+P2dj(A11)P1)+dn(P1)A11+P1dn(A11+A12)=dn(P1)A11+P1dn(A11+A12).

    Thus, we get

    dn(A11)=dn(P1)A11+P1dn(A11+A12).

    Similarly, taking X=A11+A12,Y=P2 in Eq (3.1), we obtain that

    dn(A12)=dn(P2)A11+P2dn(A11+A12).

    Adding the above two equations, we obtain from dn(P1)=dn(P2) that dn(A11+A12)=dn(A11)+dn(A12). Similarly, we can show (ⅳ) and (ⅴ) hold.

    (ⅵ) For any A12,B12G12, taking X=A12,Y=B12 in Eq (3.1), then it follows from A12B12=0 that

    0=dn(A12B12)=i+j=ndi(B12)dj(A12). (3.13)

    Since A12+B12=(P1+A12)(P2+B12), we take X=P1+A12,Y=P2+B12 in Eq (3.1), and then we get from Lemma 3.1, Lemma 3.2(ⅲ)-(ⅳ), and Eq (3.13) that

    dn(A12+B12)=dn((P1+A12)(P2+B12))=i+j=ndi(P2+B12)dj(P1+A12)=i+j=n,1i,jdi(P2+B12)dj(P1+A12)+dn(P2+B12)(P1+A12)+(P2+B12)dn(P1+A12)=i+j=n,1i,jdi(P2)(dj(P1)+dj(A12))+i+j=n,1i,jdi(B12)(dj(P1)+dj(A12))+(dn(P2)+dn(B12))(P1+A12)+(P2+B12)(dn(P1)+dn(A12))=i+j=n,1i,jdi(B12)dj(A12)+dn(P2)P1+dn(B12)P1+dn(B12)A12+P2dn(P1)+P2dn(A12)+B12dn(A12)=i+j=ndi(B12)dj(A12)P2dn(P1)P1+P2dn(P1)P1+P2dn(A12)P1+P2dn(B12)P1=dn(A12)+dn(B12).

    Similarly, we can show (vii) holds. The proof is completed.

    Lemma 3.3 For any nN+, A11G11,A12G12,A21G21 and A22G22, then dn(A11+A12+A21+A22)=dn(A11)+dn(A12)+dn(A21)+dn(A22).

    Proof For any nN+, A11G11,A12G12,A21G21,A22G22, taking X=P1,Y=A11+A12+A21+A22 in Eq (3.1), we obtain from the set of properties L (ⅰ), Lemma 3.2 (ⅲ) and (ⅴ) that

    dn(A11)+dn(A12)=dn(A11+A12)=dn(P1(A11+A12+A21+A22))=i+j=ndi(A11+A12+A21+A22)dj(P1)=i+j=n,1i,jdi(A11+A12+A21+A22)dj(P1)+dn(A11+A12+A21+A22)P1+(A11+A12+A21+A22)dn(P1).

    Similarly, we take X=P2,Y=A11+A12+A21+A22 in Eq (3.1), then we obtain that

    dn(A21)+dn(A22)=dn(A21+A22)=dn(P2(A11+A12+A21+A22))=i+j=ndi(A11+A12+A21+A22)dj(P2)=i+j=n,1i,jdi(A11+A12+A21+A22)dj(P2)+dn(A11+A12+A21+A22)P2+(A11+A12+A21+A22)dn(P2).

    Adding the above two equations, by dn(P1)+dn(P2)=0, we get that dn(A11+A12+A21+A22)=dn(A11)+dn(A12)+dn(A21)+dn(A22). The proof is completed.

    Now, we complete the proof of Theorem 3.1.

    Proof of Theorem 3.1 For any nN+, X,YG, set X=A11+A12+A21+A22 and Y=B11+B12+B21+B22, where Aij,BijGij(1ij2), then, by Lemmas 3.2 and 3.3, we can obtain that

    dn(X+Y)=dn((A11+A12+A21+A22)+(B11+B12+B21+B22))=dn((A11+B11)+(A12+B12)+(A21+B21)+(A22+B22))=dn(A11+B11)+dn(A12+B12)+dn(A21+B21)+dn(A22+B22)=dn(A11)+dn(B11)+dn(A12)+dn(B12)+dn(A21)+dn(B21)+dn(A22)+dn(B22)=dn(A11+A12+A21+A22)+dn(B11+B12+B21+B22)=dn(X)+dn(Y).

    Therefore, D={dn}nN is an additive mapping on G. The proof is completed.

    In the following, we give some applications of Theorem 3.1.

    Because triangular algebras and full matrix algebras are two special classes of generalized matrix algebras, we can get Corollaries 3.1–3.5 immediately. For the definition of triangular algebra, we refer readers to [24]. It is worth pointing out that, in [25], Ferreira showed that under certain conditions, every m-multiplicative derivation on a triangular n-matrix ring is additive.

    Corollary 3.1 Let A and B be unital algebras, M be a unital (A,B)-bimodule, which is faithful as both a left A-module and a right B-module, and U=Tri(A,M,B)={(am0b):aA,mM,bB} be a triangular algebra. If D={dn}nN is a nonlinear higher anti-derivable mapping on U, then D={dn}nN is additive.

    Corollary 3.2 Let A be an unital algebra, and Mn(A)(2n) be the full matrix algebras of all n×n matrices over A. If D={dn}nN is a nonlinear higher anti-derivable mapping on Mn(A), then D={dn}nN is additive.

    Corollary 3.3 Let R be a unital prime ring with a nontrivial idempotent P, and I be the unit of R. If D={dn}nN is a nonlinear higher anti-derivable mapping on U, then D={dn}nN is additive.

    Proof of Corollary 3.3 Suppose Q=IP. Since R a prime ring, it follows that PRQ is a faithful (PRP,QRQ)-bimodule. Then, R is isomorphic to the generalized matrix algebra

    (PRPPRQQRPQRQ)

    Therefore, by Theorem 3.1, we know that D={dn}nN is additive.

    Since standard operator algebras and factor von Neumann algebras are prime algebras with nontrivial idempotents, by Corollary 3.3, we obtain Corollary 3.4 and Corollary 3.5 as follows.

    Corollary 3.4 Let X be a Banach space over number field F, and A(X) be an unital standard operator algebra over X. If D={dn}nN is a nonlinear higher anti-derivable mapping on A(X), then D={dn}nN is additive.

    Corollary 3.5 Let H be a Hilbert space over number field F, and V be a factor von Neumann algebra over H. If D={dn}nN is a nonlinear higher anti-derivable mapping on V, then D={dn}nN is additive.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    This work was supported by Talent Project Foundation of Yunnan Province (No.202105AC160089), Natural Science Foundation of Yunnan Province(No.202101BA070001198), and Basic Research Foundation of Yunnan Education Department(Nos.2020J0748, 2021J0635).

    The authors declare there is no conflicts of interest.



    [1] D. Benkovič, Jordan derivations and anti-derivations on triangular matrices, Linear Algebra Appl., 397 (2005), 235–244. https://doi.org/10.1016/j.laa.2004.10.017 doi: 10.1016/j.laa.2004.10.017
    [2] J. H. Zhang, W. Y. Yu, Jordan derivations of triangular algebras, Linear Algebra Appl., 419 (2006), 251–255. https://doi.org/10.1016/j.laa.2006.04.015 doi: 10.1016/j.laa.2006.04.015
    [3] Z. K. Xiao, F. Wei, Jordan higher derivations on triangular algebras, Linear Algebra Appl., 432 (2010), 2615–2622. https://doi.org/10.1016/j.laa.2009.12.006 doi: 10.1016/j.laa.2009.12.006
    [4] W. L. Fu, Z. K. Xiao, X. K. Du, Nonlinear Jordan higher derivations on triangular algebras, Commun. Math. Res., 31 (2015), 119–130.
    [5] H. R. E. Vishki, M. Mirzavaziri, F. Moafian, Jordan higher derivations on trivial extension algebras, Commun. Korean Math. Soc., 31 (2016), 247–259. https://doi.org/10.4134/CKMS.2016.31.2.247 doi: 10.4134/CKMS.2016.31.2.247
    [6] C. E. Rickart, One-to-one mappings of rings and lattices, Bull. Amer. Math. Soc., 54 (1948), 758–764. https://doi.org/10.1090/S0002-9904-1948-09073-5 doi: 10.1090/S0002-9904-1948-09073-5
    [7] W. S. Martindale III, When are multiplicative mappings additive?, Proc. Amer. Math. Soc., 21 (1969), 695–698. https://doi.org/10.1090/S0002-9939-1969-0240129-7 doi: 10.1090/S0002-9939-1969-0240129-7
    [8] Y. Wang, The additivity of multiplicative maps on rings, Commun. Algebra, 37 (2009), 2351–2356. https://doi.org/10.1080/00927870802623369 doi: 10.1080/00927870802623369
    [9] Y. Li, Z. Xiao, Additivity of maps on generalized matrix algebras, Electron. J. Linear Algebra, 22 (2011), 743–757. https://doi.org/10.13001/1081-3810.1471 doi: 10.13001/1081-3810.1471
    [10] J. M. Ferreira, B. L. M. Ferreira, Additivity of n-multiplicative maps on alternative rings, Commun. Algebra, 44 (2016), 1557–1568. https://doi.org/10.1080/00927872.2015.1027364 doi: 10.1080/00927872.2015.1027364
    [11] R. C. Shaheen, On additivity of Jordan higher mappings on generalized matrix algebras, Iraqi J. Sci., 62 (2021), 1334–1343. https://doi.org/10.24996/ijs.2021.62.4.29 doi: 10.24996/ijs.2021.62.4.29
    [12] M. N. Daif, When is a multiplicative derivation additive?, Int. J. Math. Math. Sci., 14 (1991), 615–618. https://doi.org/10.1155/S0161171291000844 doi: 10.1155/S0161171291000844
    [13] M. N. Daif, M. S. Tammam-El-Sayiad, Multiplicative generalized derivations which are additive, East-West J. Math., 9 (2007), 31–37.
    [14] F. Lu, Jordan derivable maps of prime rings, Comm. Algebra, 12 (2010), 4430–4440. https://doi.org/10.1080/00927870903366884 doi: 10.1080/00927870903366884
    [15] P. Ji, Y. Lai, E. Hou, Multiplicative Jordan derivations on Jordan algebras, Acta Math, Sinica (Chin. Ser.), 53 (2010), 571–578.
    [16] J. Wu, F. Lu, Additivity of Jordan (triple) derivations on rings, Commun. Algebra, 40 (2012), 2700–2719. https://doi.org/10.1080/00927872.2011.584927 doi: 10.1080/00927872.2011.584927
    [17] W. Fu, Z. Xiao, X. Du, Nonlinear Jordan higher derivations on triangular algebras, Commun. Math. Res., 31 (2015), 119–130.
    [18] M. Ashraf, A. Jabeen, Nonlinear Jordan triple higher derivable mappings of triangular algebras, Southeast Asian Bull. Math., 42 (2018), 503–520.
    [19] D. A. Abulhamil, F. B. Jamjoom, A. M. Peralta, Linear maps which are anti-derivable at zero, Bull. Malays. Math. Sci. Soc., 43 (2020), 4315–4334. https://doi.org/10.1007/s40840-020-00918-7 doi: 10.1007/s40840-020-00918-7
    [20] H. Ghahramani, M. N. Ghosseiri, T. Rezaei, Anti-derivations on triangular rings, Khayyam J. Math., 9 (2023), 30–37. https://doi.org/10.22034/KJM.2022.367135.2696 doi: 10.22034/KJM.2022.367135.2696
    [21] B. L. M. Ferreira, A. Jabeen, Multiplicative maps on generalized n-matrix rings, arXiv preprint, (2022), arXiv: 2205.15728. https://doi.org/10.48550/arXiv.2205.15728
    [22] K. Morita, Duality for modules and its applications to the theory of rings with minimum condition, Rep. Tokyo Kyoiku Diagaku Sect., 6 (1958), 83–142.
    [23] B. L. M. Ferreira, G. S. Sandhu, Multiplicative anti-derivations of generalized n-matrix rings, J. Algebra Appl., (2022), 2450079. https://doi.org/10.1142/S0219498824500798
    [24] W. S. Cheung, Mappings on triangular algebras, Ph.D Thesis, University of Victoria, 2000. Available from: http://dspace.library.uvic.ca/bitstream/handle/1828/9349/Cheung_Wai-Shun_PhD_2000.pdf?sequence = 1 & isAllowed = y
    [25] B. L. M. Ferreira, Multiplicative maps on triangular n-matrix rings, Int. J. Math. Game Theor. Algebra, 23 (2014), 1–14.
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1167) PDF downloads(53) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog