{"title":"The Trace and Integrable Commutators of the Measurable Operators Affiliated to a Semifinite von Neumann Algebra","authors":"A. M. Bikchentaev","doi":"10.1134/s0037446624030030","DOIUrl":null,"url":null,"abstract":"<p>Assume that <span>\\( \\tau \\)</span> is a faithful normal semifinite trace\non a von Neumann algebra <span>\\( {\\mathcal{M}} \\)</span>, <span>\\( I \\)</span> is the unit of <span>\\( \\mathcal{M} \\)</span>,\n<span>\\( S({\\mathcal{M}},\\tau) \\)</span> is the <span>\\( * \\)</span>-algebra of <span>\\( \\tau \\)</span>-measurable operators,\nand <span>\\( L_{1}({\\mathcal{M}},\\tau) \\)</span> is the Banach space of <span>\\( \\tau \\)</span>-integrable operators.\nWe present a new proof of the following generalization\nof Putnam’s theorem (1951):\nNo positive self-commutator\n<span>\\( [A^{*},A] \\)</span>\nwith\n<span>\\( A\\in S({\\mathcal{M}},\\tau) \\)</span>\nis invertible in <span>\\( {\\mathcal{M}} \\)</span>.\nIf <span>\\( \\tau \\)</span>\nis infinite\nthen no positive self-commutator\n<span>\\( [A^{*},A] \\)</span>\nwith\n<span>\\( A\\in S({\\mathcal{M}},\\tau) \\)</span>\ncan be of the form\n<span>\\( \\lambda I+K \\)</span>,\nwhere <span>\\( \\lambda \\)</span>\nis a nonzero complex number and <span>\\( K \\)</span>\nis a <span>\\( \\tau \\)</span>-compact operator.\nGiven\n<span>\\( A,B\\in S({\\mathcal{M}},\\tau) \\)</span>\nwith\n<span>\\( [A,B]\\in L_{1}({\\mathcal{M}},\\tau) \\)</span>\nwe seek for the conditions that\n<span>\\( \\tau([A,B])=0 \\)</span>.\nIf\n<span>\\( X\\in S({\\mathcal{M}},\\tau) \\)</span>\nand\n<span>\\( Y=Y^{3}\\in{\\mathcal{M}} \\)</span>\nwith\n<span>\\( [X,Y]\\in L_{1}({\\mathcal{M}},\\tau) \\)</span>\nthen\n<span>\\( \\tau([X,Y])=0 \\)</span>.\nIf\n<span>\\( A^{2}=A\\in S({\\mathcal{M}},\\tau) \\)</span>\nand\n<span>\\( [A^{*},A]\\in L_{1}({\\mathcal{M}},\\tau) \\)</span>\nthen\n<span>\\( \\tau([A^{*},A])=0 \\)</span>.\nIf a partial isometry <span>\\( U \\)</span>\nlies in <span>\\( {\\mathcal{M}} \\)</span>\nand\n<span>\\( U^{n}=0 \\)</span>\nfor some integer\n<span>\\( n\\geq 2 \\)</span>\nthen <span>\\( U^{n-1} \\)</span>\nis a commutator\nand\n<span>\\( U^{n-1}\\in L_{1}({\\mathcal{M}},\\tau) \\)</span>\nimplies that\n<span>\\( \\tau(U^{n-1})=0 \\)</span>.</p>","PeriodicalId":0,"journal":{"name":"","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"","FirstCategoryId":"100","ListUrlMain":"https://doi.org/10.1134/s0037446624030030","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Assume that \( \tau \) is a faithful normal semifinite trace
on a von Neumann algebra \( {\mathcal{M}} \), \( I \) is the unit of \( \mathcal{M} \),
\( S({\mathcal{M}},\tau) \) is the \( * \)-algebra of \( \tau \)-measurable operators,
and \( L_{1}({\mathcal{M}},\tau) \) is the Banach space of \( \tau \)-integrable operators.
We present a new proof of the following generalization
of Putnam’s theorem (1951):
No positive self-commutator
\( [A^{*},A] \)
with
\( A\in S({\mathcal{M}},\tau) \)
is invertible in \( {\mathcal{M}} \).
If \( \tau \)
is infinite
then no positive self-commutator
\( [A^{*},A] \)
with
\( A\in S({\mathcal{M}},\tau) \)
can be of the form
\( \lambda I+K \),
where \( \lambda \)
is a nonzero complex number and \( K \)
is a \( \tau \)-compact operator.
Given
\( A,B\in S({\mathcal{M}},\tau) \)
with
\( [A,B]\in L_{1}({\mathcal{M}},\tau) \)
we seek for the conditions that
\( \tau([A,B])=0 \).
If
\( X\in S({\mathcal{M}},\tau) \)
and
\( Y=Y^{3}\in{\mathcal{M}} \)
with
\( [X,Y]\in L_{1}({\mathcal{M}},\tau) \)
then
\( \tau([X,Y])=0 \).
If
\( A^{2}=A\in S({\mathcal{M}},\tau) \)
and
\( [A^{*},A]\in L_{1}({\mathcal{M}},\tau) \)
then
\( \tau([A^{*},A])=0 \).
If a partial isometry \( U \)
lies in \( {\mathcal{M}} \)
and
\( U^{n}=0 \)
for some integer
\( n\geq 2 \)
then \( U^{n-1} \)
is a commutator
and
\( U^{n-1}\in L_{1}({\mathcal{M}},\tau) \)
implies that
\( \tau(U^{n-1})=0 \).