Abstract
Let be a self-adjoint standard operator algebra on a real or complex Hilbert space of dimension , and let . The k-skew commutator for is defined by and . Assume that is a map whose range contains all rank-one projections. In this paper, we prove that is strong k-skew-commutativity preserving, that is, for all if and only if one of the following statements holds: (i) is either the identity map or the negative identity map whenever ; (ii) is the identity map whenever .
Keywords:
bounded linear operators; self-standard operator algebras; k-skew commutators; preservers; Hilbert spaces MSC:
47B49
1. Introduction
Let X be a real or complex Banach space and denote by the algebra of all bounded linear operators on X. For a positive integer k and any elements , the k-commutator or k-Lie product of A and B is defined inductively by , and . The k-commutators appear naturally in the famous Baker–Campbell–Hausdorff formula
which is used in quantum physics frequently and has been investigated extensively.
Let be a ring. A map is said to be strong k-commutativity preserving if preserves k-commutators invariant, that is, for all . The problem of characterizing strong k-commutativity-preserving maps has been studied intensively on various rings and operator algebras. For example, the readers can see [1,2,3,4,5,6] and the references therein.
If is a real or complex Hilbert space, then is a ∗-algebra with ∗ operation the usual adjoint of operators. Thus, the skew Lie product or the skew commutator of A and B may be defined as (Ref. [7], where is denoted by ). For any positive integer , like the concept of the k-commutator, the k-skew commutator of a pair in is naturally defined inductively by with , . Note that the k-skew commutators appear in so-called skew Baker–Campbell–Hausdorff formula in a natural way:
The concept of k-skew Lie product is found playing an important role in some research topics, and their study has attracted many authors’ attention (Refs. [8,9,10,11,12,13,14,15,16,17,18]).
Recall that a standard operator algebra on H is said to be self-adjoint if implies . The concept of the k-skew commutator can be introduced for any ring with involution ∗. Similar to the commutator case stated in the previous paragraph, a map is said to be strong k-skew-commutativity preserving if keeps the k-skew commutators invariant, i.e., = for any . For the problem of how to characterize the strong k-skew-commutativity-preserving maps, it was shown in [11] that a nonlinear surjective map on a factor von Neumann algebra is strong skew-commutativity preserving if and only if it has the form , where is a strong skew-commutativity-preserving bijective linear map and is a functional with . As usual, and stand, respectively, for the real number field and the complex number field. In [16], a more precise form of the above surjective map was achieved: , where , the set of all self-adjoint elements in the center of . In a pure algebraic frame, it is also shown in [16] that if is a prime ∗-ring which contains a nontrivial symmetric idempotent, then every strong skew commutativity preserving surjective map on has the form for all , where and . This result later was generalized in [15] and [14] to unital ∗-rings which contain a nontrivial symmetric idempotent P satisfying “” and “”. Recently, ref. [12] has sharpened the results in [14,15,16] by proving that the symmetric center valued map f is 0 and then gives further a characterization of the strong 2-skew-commutativity-preserving maps; that is, a surjective map is strong 2-skew commutativity preserving if and only if for all , where with and is the symmetric extended centroid of . This result then is applied to some operator algebras such as prime C∗-algebras, factor von Neumann algebras, the indefinite self-adjoint standard operator algebras, and symmetric standard operator algebras.
Suppose that is a self-adjoint standard operator algebra on a complex Hilbert space of dimension . Without the surjectivity assumption on the maps, it is conjectured in [18] that a map on with range containing all rank-one projections is a strong k-skew-commutativity-preserving map if and only if it has the form , where with . However, no approach has been found to provide a proof valid for all cases of dimension and the value of k. In [18], the authors gave a proof of the above result for the case when and , which is not valid for the rest of the cases or . Thus, the proofs for the cases that or still need to be presented.
One of the purposes of this paper is to prove the main result announced in [18] for the cases and . But, much more can be said. Though we do not know whether or not the announced result in [18] is still true for the case that the Hilbert spaces are real, we consider here not only the complex but also the real Hilbert spaces H with . Under the assumption that the range contains only all rank-one projections, we give a complete characterization of the maps on self-adjoint standard operator algebras that are strong k-skew-commutativity preserving for . Recall that an operator P on H is called a projection if , and the rank of an operator is the dimension of its range.
The following is our main result.
Theorem 1.
Let , H be a real or complex Hilbert space with and be a self-adjoint standard operator algebra. Assume that is a map whose range contains all rank-one projections. Then, is strong k-skew-commutativity preserving if and only if is the identity map or the negative identity map whenever ; is the identity map whenever .
This paper is organized as follows. Section 2 gives some lemmas that are needed to prove the main result. Some of these lemmas reveal properties of k-skew commutativity and are interesting themselves. Section 3, Section 4, and Section 5 are, respectively, devoted to proving Theorem 1 for the case and 3.
2. Preliminaries and Lemmas
To prove Theorem 1, we need following lemmas.
Assume that X is a Banach space and . In [19], the authors proved that if holds for all and () are linearly independent, then are depend linearly on . By this result, the following lemma is obvious.
Lemma 1.
Let be a standard operator algebra on a Banach space, and let , , , such that for all rank-one operators . If are linearly independent, then each is a linear combination of . Similarly, if are linearly independent, then each is a linear combination of .
Bear in mind that
where , and, particularly,
and
The following lemma comes from [18]. Notice that the lemma holds automatically if .
Lemma 2
([18]). Let H be a real or complex Hilbert space, be an integer and . Then, for all rank-one projections P on H if and only if for some scalar α.
Let k be a positive integer and S be a bounded linear operator on a complex Hilbert space H. It was shown in [18] that for all rank-one projections P if and only if for some and . Here, stands for the integer part of a real number t and . Our next lemma says that this result also holds for real case if .
Lemma 3.
Let , H be a real or complex Hilbert space. Then, for , the following statements are equivalent.
(1) for all .
(2) for all rank-one projections P on H.
(3) for some real number λ and some nilpotent operator with .
Proof.
(3)⇒(1): Note that if , and if . So, if S has the form in (3), then whenever or 2. In this case it is easily checked that, for any , and . If , then with , and so for any , we have
The implication (1)⇒(2) is obvious.
Let us check (2)⇒(3). Assume that holds for every rank-one projection P.
Case 1. . For any unit vector , is a rank-one projection. Thus, . It follows that, for any vector x, there exists some scalar depending on x such that . This implies for some scalar . Moreover, since , is real.
Case 2. . If for some scalar , then, picking a rank-one projection P, entails that is real and thus (3) is true.
So, we may assume that S is linearly independent of I and then induce a contradiction. For any , define
Then, is an elementary operator on which is linear and continuous under the weak operator topology (WOT, briefly). Since for all rank-one projections P by (2), we see that also vanishes at every finite rank operator. Note that the set of all finite rank operators is dense in under WOT. So, , that is,
holds for all . By Lemma 1, as , there exist scalars such that . Thus, the linear independence of S and I implies that and hence, for some scalars and . Substituting this into Equation (1) gives
for all . However, by Lemma 1, this implies that for some scalar , contradicting to the fact that is linearly independent to I.
Case 3. . For any , define
Then, is a linear transformation from into and is continuous under WOT. Since for any rank-one projection P, holds for every , that is,
If there exists some such that , then Equation (2) reduces to
for any A, which implies for any A. Hence, is real and S has the required form in (3).
In the sequel, we assume that S is linearly independent of I. In this case, applying Lemma 1 to Equation (2), there exist , , , such that
As S and I are linearly independent, by Lemma 1 again, one obtains
Thus, we must have
On the other hand, by , we obtain
Equations (3) and (4) together entail and . Let . It is easily checked that and has the required form. □
In the following lemmas, we always assume that is a self-adjoint standard operator algebra on a real or complex Hilbert space H, and is a strong k-commutativity-preserving map whose range contains all rank-one projections.
Note that, as the surjectivity of is not supposed, one can not assert that is additive as that in [12]. What we can say in the present situation is the following.
Lemma 4.
Φ is almost additive, that is, for any , there exists a scalar depending on such that .
Proof.
For any , we have
As the range of contains all rank-one projections, Lemma 2 implies that for some scalar . □
Similarly, is not symmetric without the surjectivity assumption. But, we have
Lemma 5.
For any , we have for some scalar depending on A.
Proof.
In fact, for any ,
holds. As the range of contains all rank-one projections, by use of Lemma 2, one obtains for some scalar , a scalar multiple of the identity. □
Lemma 6.
For , if and only if . Particularly, for the cases , if and only if .
Proof.
For any , by Lemma 3, holds for all . As the range of contains all rank-one projections, Lemma 3 also implies that . Conversely, if , still by Lemma 3, holds for all , which entails that .
The last assertion holds because if . □
Lemma 7.
For any projection , we have
for all .
Proof.
Note that, if P is a projection, we have for any . Then, for any ,
and
So, for all rank-one projections Q by the assumption on the range of . Now, it is easily checked that holds for all operators . □
3. The Proof of Theorem 1:
In this section, we will check the “only if” part of Theorem 1 for the case .
Let be a self-adjoint standard operator algebra on a Hilbert space H of dimension ≥2 over the field with or . Let be a map whose range contains all rank one projections. Assume that is strong skew-commutativity preserving, that is,
holds for any . We have to show that or , where id stands for the identity map on . This will be performed in several steps.
Step 3.1. For any nontrivial projection , we have for some with .
Write , . As P is nontrivial, , . Write with , .
Note that, by Lemma 7, for any Thus, for any , we have
So, holds for all , which is equivalent to that holds for all . By Lemma 1, we must have . Note that for some by Lemma 5. We see that , and thus It follows that for any , . Since can be regarded as a self-adjoint standard operator algebra on ; by Lemma 3, we obtain for some , . Let and ; then , as desired. Clearly, . Otherwise, , and Lemma 6 implies that , a contradiction.
Step 3.2. Assume that . Then, there exists a nonzero real number such that, for any rank-one projection , for some .
Let be two rank-one projections on H so that (i.e., ). Let . Then, R is a nontrivial projection as . By Lemma 4 and Step 3.1, we have
which implies that , and consequently, . Now, for any rank-one projections P and Q, the assumption ensures that there exists a rank-one projection Z such that and . Then, by what we just proved above, we have and . Hence, still holds. This means that there exists a nonzero so that for every rank-one projection P.
Step 3.3. Assume that . Then, and for every .
Take rank-one projections so that . Then
gives . So .
For any , note that holds for all rank one projections P. By Lemma 2, this implies that . As , and one sees that for some scalar .
We claim that for each A. Let Q be a rank one idempotent operator with and consider
One obtains
Since dim and , we must have . This implies that . It follows that
which gives . Hence .
Step 3.4. Assume that . Then, , too.
If , then can be identified with , the algebra of all matrices over . So we can assume that is a map on . Denote by
the matrix units for . By Step 3.1, there exist real numbers such that
For any , denote by . Then, Equation (5) implies that
which gives and .
On the other hand, since
we obtain and . Thus, we have proved that
Similarly, replacing by in Equations (6) and (7), one obtains
So, by the equation
one obtains , and thus .
For any , denote . Then,
implies that . Thus, . Write . As , one obtains and . It follows that
and therefore, .
Now, for , by a similar argument to that of , one can show that
for all , where . Note that
This yields , and so .
Finally, for any , writing , Lemma 4 ensures that
for some scalar . In addition, by
it is easily checked that for some . Thus, as
we obtain . So for all A, where .
This completes the proof of Theorem 1 for .
4. The Proof of Theorem 1:
In this section, we will check the “only if” part of Theorem 1 for case .
Assume that is strong 2-skew-commutativity preserving, that is,
for any , of which the range contains all rank-one projections. We have to show that is the identity map.
Step 4.1. For any nontrivial projection , there exists some such that .
Let be any nontrivial projection. Write , and , . Then, there are so that .
Taking and applying Lemma 7, we have
which implies that by Lemma 1. Then, by Lemma 5. Hence, . Applying Lemma 7 again, one obtains that
holds for all . That is, for all . By Lemma 3, for some real number .
Analogously, there exists some real number such that . It follows that
where .
We claim that . If not, then , which implies that by Lemma 6, a contradiction as .
Now, take any nontrivial projection Q so that . Then,
which gives . Similarly, . So , and consequently, . The proof of the step is completed.
Step 4.2. holds for any .
Let be arbitrary. For any rank-one projection ; by Step 4.1, we have
By Lemma 2, we see that for some scalar .
Thus, we have shown that there exists a functional such that
In the following, let us prove by considering two cases.
Case 4.2.1..
We first claim that for any rank-one idempotent Q with . In fact, by Equation (8),
for some scalars , that is,
Note that the rank of is not greater than 2 as Q is of rank one. Since , we must have , which implies . Thus,
which entails that as .
Now, for any , taking a non self-adjoint rank one idempotent Q and by using of Equation (8), we have
So , as desired.
Case 4.2.2..
In this case we can identify with . Let and denote . Then,
So, and , which entail . Now, by a similar argument to that of the corresponding part in Case 4.2.1, one can show that for each .
The proof of Theorem 1 for is completed.
5. The Proof of Theorem 1:
In this section, we assume that H is a real or complex Hilbert space with , is a self-adjoint standard operator algebra, and is a map of which the range contains all rank-one projections. To complete the proof of Theorem 1 for the case , one only needs to show that if is strong 3-skew-commutativity preserving, that is, if satisfies
for any , then is the identity map or the negative identity map. We will perform this in several steps.
Step 5.1. For any rank-one projection , there exist real scalars with such that .
Let P be any nontrivial projection in . Still, denote , and , . Then, is a self-adjoint standard operator algebra on , .
Thus, there are so that
As , by Lemma 5, we have for some scalar , which implies that for .
We have to consider the real case and the complex case separately.
Case 5.1.1. .
In this case, we in fact show that Step 5.1 holds for any nontrivial projection P. Write as in Equation (9).
Assume that in Equation (9). Obviously, as .
Taking any and applying Lemma 7, we have
It follows that
Then, by Lemma 1, there exists some such that
and consequently, .
Similarly, it follows from the relation that there exists a real number such that
Next, taking and applying Lemma 7, we have which, together with Equations (10) and (11), implies that
holds for all . Since , by Lemma 1 again, there exists some such that
and
Similarly, for any , by the relation , there exists some such that
and
Note that . It follows from Equation (12) that
Hence,
Similarly, it follows from Equation (14) that
If , then (). So, by Equation (10), , which implies , contradicting to the assumption. Therefore, we must have , and thus . This yields , . Then, Equation (13) becomes to
Particularly, we have by Equations (15) and (16). Multiplying by from the left of Equation (12) gives
Comparing this with Equation (16), one obtains
By using Equation (10) twice for the above equation, we obtain As , we have . Let ; then . Combining Equations (10)–(14), one obtains
So . However, this and Lemma 6 imply , a contradiction. Therefore, we must have , and thus .
Now, for any , by Lemma 7, we have
As is also a self-adjoint standard operator algebra, it follows from Lemma 3 that
Replacing by , a similar argument ensures that
Hence,
where , as . Since , one sees that , which entails and as . Consequently, . Finally, we claim . Otherwise, , which, together with Lemma 6, entails , a contradiction.
This completes the proof of Step 5.1 for the real case.
Case 5.1.2. .
Assume that is of rank one; then, and every has rank ≤2. Thus, by Lemma 7, for any , . Particularly, for each unit vector , applying the above fact to the rank-one projection , we see that the rank of
is not greater than 2. Denote with as in Equation (9). Above equation implies that is a linearly dependent set for each . By the Kaplansky theorem, S is an algebraic operator of degree ; that is, there exists a polynomial such that . So the spectra of S consists of at most three eigenvalues.
If there exists a with , then for some unit vector x and
which induces a contradiction . So, we must have . This, together with the fact by Lemma 5, gives and . Note that has at least two points. Otherwise, as S is self-adjoint, we have , which is impossible by Lemma 6.
We next claim that the one dimensional subspace is invariant under S. To see this, take and unit vectors so that , . Then,
So, and . As and cannot be 0 at the same time, we see that and ; and . Without loss of generality, say and . Then, is invariant under S. Consequently, and . In fact, . Assume, on the contrary, that there exists . Then, for some unit vector . However,
implies , a contradiction. Hence, , and thus with and .
So far we have proved that, for any rank-one projection P, there are real scalars such that
that is, Step 5.1 is also true for complex case.
Step 5.2. If , then there exists a such that holds for every rank-one projection .
Let be any rank-two projection. Fix a unit vector and take a unit vector so that . Then, for any scalars with , we have
Denote by and . By Lemma 4 and Step 5.1, one has
for some scalars . Obviously, this implies and .
Thus, we in fact have shown that, for any rank-one projections , whenever .
Since , for any rank-one projections , there exists a rank-one projection so that . By what we just proved above, one sees that . So, there exists a real number such that
Now, pick two rank-one projections so that . Then, by
one obtains . It follows that as is real.
Step 5.3. If , then for all .
Let . For any rank-one projection ; by Step 5.2, we have
It follows from Lemma 2 that for some scalar , and so,
What remains is to check . Assume for some A. By Step 5.1, for any rank-one projection P. We claim further that for any rank one idempotent operator Q. Note that, if Q is an idempotent, then . Thus, by Equation (17), we have
for some scalars , that is,
Since and , we must have . So
which entails that whenever . Hence we have shown that
It follows that holds for any .
This completes the proof of Theorem 1 for the case and .
In the following Steps 5.4–5.6, we assume that . In this situation, we have , and so we can identify as the two by two-matrix algebra , where or . Denote by
the matrix units for . By Step 5.1, there exist real numbers and such that
Step 5.4. If , then, for any , there exists such that
Here, we only give the details of the proof for the case . The other case is dealt with similarly.
Write . By Equation (18), a simple computation reveals that
So and ; that is,
On the other hand, by Lemma 6, for some real scalar and N with . Comparing this with Equation (19) gives and , as required.
Step 5.5. , and, for any , we have , .
Let . Since
we see that .
Now let us check that, for or , it is true that
In fact, if , then, by Lemma 4, for some . Note that, by Equation (18) and Step 5.4,
We see that and . If is not real, then and , which leads to a contradiction . The case can be dealt with similarly. So Equation (20) holds.
These imply that and . So holds for any .
Similarly, one can prove that holds for any .
Finally, since
we obtain and . Therefore, and . The fact can be checked similarly.
Step 5.6. for every .
For any , denote by . By Lemma 4 and Step 5.5, there is a scalar so that
Furthermore, by Equation (20) and Step 5.5, we have
Hence, , and for every . This completes the proof of the “only if” part for the case and .
The proof of Theorem 1 for is finished.
Author Contributions
Conceptualization, T.Z. and X.Q.; methodology, X.Q.; formal analysis, X.Q.; writing—original draft preparation, T.Z.; writing—review and editing, X.Q.; supervision, X.Q.; project administration, X.Q.; funding acquisition, X.Q. All authors have read and agreed to the published version of the manuscript.
Funding
This work is partially supported by the Natural Science Foundation of China (12171290, 12301152) and the Fundamental Research Program of Shanxi Province (202203021222018).
Data Availability Statement
Data sharing is not applicable to this article as no new data were created or analyzed in this study.
Acknowledgments
The authors wish to give their many thanks to the anonymous referees for helpful comments and suggestions that improved the original manuscript.
Conflicts of Interest
The authors declare no conflicts of interest.
References
- Brešar, M.; Miers, C.R. Strong commutativity preserving maps of semiprime ring. Can. Math. Bull 1994, 37, 457–460. [Google Scholar] [CrossRef] [Scilit]
- Hou, J.; Qi, X. Strong k-commutativity preservers on complex standard operator algebras. Linear Multilinear Algebra 2018, 66, 902–918. [Google Scholar] [CrossRef] [Scilit]
- Lin, J.; Liu, C. Strong commutativity preserving maps on Lie ideals. Linear Algebra Appl. 2008, 428, 1601–1609. [Google Scholar] [CrossRef] [Scilit]
- Liu, L. Strong commutativity preserving maps on von Neumann algebras. Linear Multilinear Algebra 2015, 63, 490–496. [Google Scholar] [CrossRef] [Scilit]
- Qing, D.; Mohammad, A. On strong commutativity preserving maps. Results Math. 1996, 30, 259–263. [Google Scholar]
- Qi, X.; Hou, J. Strong 3-commutativity preserving maps on prime rings. Linear Multilinear Algebra 2017, 65, 2153–2169. [Google Scholar] [CrossRef] [Scilit]
- Molnár, L. A condition for a subspace of to be an ideal. Linear Algebra Appl. 1996, 235, 229–234. [Google Scholar] [CrossRef] [Scilit]
- Šemrl, P. Quadratic and quasi-quadratic functionals. Proc. Amer. Math. Soc. 1993, 119, 1105–1113. [Google Scholar] [CrossRef] [Scilit]
- Cui, J.; Hou, J. Linear maps preserving elements annihilated by a polymnomial XY − YX†. Studia Math. 2006, 174, 183–199. [Google Scholar] [CrossRef] [Scilit]
- Cui, J.; Li, C.-K. Maps preserving product XY − YX∗ on factor von Neumann algebras. Linear Algebra Appl. 2009, 431, 833–842. [Google Scholar] [CrossRef] [Scilit]
- Cui, J.; Park, C. Maps preserving strong skew lie product on factor von Neumann algebras. Acta Math. Sci. 2012, 32, 531–538. [Google Scholar]
- Hou, J.; Wang, W. Strong 2-skew commutativity preserving maps on prime rings with involution. Bull. Malays. Math. Sci. Soc. 2019, 42, 33–49. [Google Scholar] [CrossRef] [Scilit]
- Li, C.; Lu, F.; Fang, X. Nolinear ξ-Jordan ∗-dervations on von Neumann algebras. Linear Multilinear Algebra 2014, 2, 466–473. [Google Scholar] [CrossRef] [Scilit]
- Li, C.; Chen, Q. Strong skew commutativity preserving maps on rings with involution. Acta Math. Sin. 2016, 32, 745–752. [Google Scholar] [CrossRef] [Scilit]
- Liu, L. Strong skew commutativity preserving maps on rings. J. Aust. Math. Soc. 2016, 100, 78–85. [Google Scholar] [CrossRef] [Scilit]
- Qi, X.; Hou, J. Strong skew commutativity preserving maps on von Neumann algebras. J. Math. Anal. Appl. 2013, 397, 363–370. [Google Scholar] [CrossRef] [Scilit]
- Šemrl, P. Jordan *-derivations of standard operator algebras. Proc. Amer. Math. Soc. 1994, 120, 515–518. [Google Scholar] [CrossRef] [Scilit]
- Wang, W.; Hou, J. Strong k-skew commutativity preserving maps on self-adjoint standard operator algebras. Acta Math. Sin. (Ser. A) 2017, 60, 39–52. [Google Scholar]
- Fong, C.-K.; Sourour, A.R. On the operator equation AiXBi ≡ 0. Canad. J. Math. 1979, 31, 845–857. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).