Derivations of rings of infinite matrices
Extracted source text ยท arXiv 2310.11509
%\documentclass[sn-mathphys]{sn-jnl}
\documentclass[12pt,a4paper]{amsart}
\usepackage{graphics}
\usepackage{epsfig}
\usepackage[all]{xy}
\usepackage{graphicx}
\theoremstyle{plain}
\usepackage{amssymb}
\usepackage{cite}
\usepackage[english]{babel}
\advance\hoffset-20mm \advance\textwidth40mm
\newcommand{\anm}[1]{\marginpar[\hfill {\bf !}]{\bf !}\footnote{#1}}
\newtheorem{theorem}{Theorem}
\newtheorem{lemma}{Lemma}
\renewcommand{\labelenumi}{\arabic{enumi}.}
% Makros
% Makros
%
\newcommand{\cs}{$\clubsuit$}
\newcommand{\sd}{$\spadesuit$}
\newcommand\qmatrix[2][1]{\left(\renewcommand\arraystretch{#1}
\begin{array}{*{20}r}#2\end{array}\right)}
\renewcommand{\baselinestretch}{1.4}
\author{O.Bezushchak}
\title{Derivations of rings of infinite matrices}
\def\baselinestretch{1.1}
\begin{document}
\maketitle
\begin{abstract} We describe derivations of several important associative and Lie rings of infinite matrices over general rings of coefficients.
\end{abstract}
\keywords{\emph{Key words and phrases.} Ring of infinite matrices; derivation.} % Keywords
% Math. Subj. Class. codes
\subjclass{2020 \emph{Mathematics Subject Classification.} 15B30, 16W25.}
\section*{Introduction}
Let $R$ be an associative ring with unit $1_R,$ and let $I$ be an infinite set.
Consider the ring $M(I,R)$ of $(I\times I)$-matrices over the ring $R$ having finitely many nonzero entries in each column. The ring $M(I,R)$ is isomorphic to the ring of endomorphisms of a free $R^{op}$-module of rank $\text{card}(I).$ Here, $R^{op}$ is a ring that is anti-isomorphic to $R;$ and $\text{card}(I)$ is the cardinality of the set $I.$
Consider also the subring $M_{\infty}(I,R)<M(I,R)$ of all $(I\times I)$-matrices over $R$ having finitely many nonzero entries, and the subring $M_{rcf}(I,R)<M(I,R)$ of all $(I\times I)$-matrices over $R$ having finitely many nonzero entries in each row and in each column.
Recall that an additive mapping $d:R \to R $ is called a \textit{derivation} if $$ d(ab)= d(a) b + a d(b) \quad \text{for arbitrary elements} \quad a, \ b \in R . $$ Let $V$ be a bi-module over a ring $R.$ An additive mapping $d:R \to V$ is called a \textit{derivation} or a $1$-\textit{cocycle} if $ d(ab)= d(a) b + a d(b) $ for arbitrary elements $ a, b \in R .$ For an element $v\in V$ the mapping $d_v : R\to V,$ $d_v(a)=av-va$ is a derivation.
The purpose of this paper is to determine derivations of the rings $M_{\infty}(I,R),$ $M_{rcf}(I,R),$ $M(I,R).$ Recall that all derivations of a ring form a Lie ring; see \cite{6Jacobson}.
Every derivation of the ring $R$ gives rise to a derivation of the ring $M(I,R)$ that leaves $M_{rcf}(I,R)$ and $M_{\infty}(I,R)$ invariant. Hence, the Lie ring $\text{Der}(R)$ lies in each of the Lie rings $\text{Der}(M(I,R)),$ $\text{Der}(M_{rcf}(I,R)),$ $\text{Der}(M_{\infty}(I,R)).$
For an element $a\in M(I,R),$ let $$\text{ad}(a):x\mapsto [a,x]=ax -xa $$ denote the \textit{inner derivation}. Since $M_{\infty}(I,R)$ is a two-sided ideal in $M_{rcf}(I,R),$ it follows that $M_{\infty}(I,R)$ is invariant under any inner derivation $\text{ad}(a),$ $a\in M_{rcf}(I,R).$
\begin{theorem}\label{Th1} \
\begin{enumerate}
\item[\text{(a)}] An arbitrary derivation $d$ of the ring $M_{\infty}(I, R)$ is of the type $$d=\emph{ad} (a) + u, \quad \text{where} \quad a\in M_{rcf}(I,R), \quad u \in \text{\emph{Der}}(R);$$
\item[\text{(b)}] an arbitrary derivation $d$ of the ring $M(I,R)$ $($resp. $M_{rcf}(I,R))$ is of the type $$d=\emph{ad} (a) + u, \quad \text{where} \quad a\in M(I,R) \quad (\text{resp.} \ a\in M_{rcf}(I,R)), \quad u \in \text{\emph{Der}}(R).$$
\end{enumerate}
\end{theorem}
In \cite{1Bezushchak}, we proved Theorem~\ref{Th1} in the case when $R$ is a field and the derivation $d$ is linear. W.~Ho{\l}ubowski and S.~\.Zurek proved Theorem \ref{Th1} under the assumptions that $(i)$ the set $I$ is countable, $(ii)$ the ring $R$ is commutative, and $(iii)$ the derivation $d$ is $R$-linear; see \cite{5Hol_Zurek}.
An arbitrary associative ring $R$ gives rise to the Lie ring $$R^{(-)}=(R, \ [a,b]=ab-ba).$$
Using the proof of Herstein's Conjectures by K.I.~Beidar, M.~Bre\v{s}ar, M.A.~Chebotar and W.S.~Martindale (see \cite{2Bei_Bre_Cheb_Mart_1, 3Bei_Bre_Cheb_Mart_2, 4Bei_Bre_Cheb_Mart_3}) and Theorem~\ref{Th1}, we obtain descriptions of derivations of Lie rings $$\mathfrak{gl}_{rcf}(I,R)=M_{rcf}(I,R)^{(-)}, \quad \mathfrak{gl}(I,R) =M(I,R)^{(-)}, \quad \mathfrak{sl}_{\infty}(I,R)=[\mathfrak{gl}_{\infty}(I, R),\mathfrak{gl}_{\infty}(I, R)]$$ under the assumption that $\frac{1}{2}\in R.$
\begin{theorem}\label{Th2} \
\begin{enumerate}
\item[\text{(a)}] An arbitrary derivation $d$ of the Lie ring $\mathfrak{sl}_{\infty}(I,R)$ is of the type $$d=\emph{ad} (a) + u, \quad \text{where} \quad a\in \mathfrak{gl}_{rcf}(I,R), \quad u \in \text{\emph{Der}}(R);$$
\item[\text{(b)}] an arbitrary derivation $d$ of the Lie ring $\mathfrak{gl}(I,R)$ $($resp. $\mathfrak{gl}_{rcf}(I,R))$ is of the type $$d=\emph{ad} (a) + u, \quad \text{where} \quad a\in \mathfrak{gl}(I,R) \quad (\text{resp.} \ a\in \mathfrak{gl}_{rcf}(I,R)), \quad u \in \text{\emph{Der}}(R).$$
\end{enumerate}
\end{theorem}
For an element $a\in R$ and indices $ i,j\in I$ let $e_{ij}(a)$ denote the $(I\times I)$-matrices having the element $a$ at the intersection of the $i$-th row and the $j$-th column, all other entries are equal to zero.
The following lemma is straightforward.
\begin{lemma}\label{lemma1} Let $A=(a_{ij})_{i,j\in I} \in M(I,R),$ $a_{ij}\in R.$ Then $$ A=e_{kk}(1_R)A + Ae_{kk}(1_R)$$ if and only if $ a_{ij}=0 $ whenever $ i \neq k, $ $ j \neq k$ or $ i=j=k.$ \end{lemma}
Let $\mathbb{Z}$ be the ring of integers. Then $\mathbb{Z} \cdot 1_R$ is a subring of the ring $R.$ The ring $M(I,R)$ is a bimodule over the ring $M_{\infty} (I, \mathbb{Z}\cdot 1_R).$
\begin{lemma}\label{lemma2} For an arbitrary derivation $d:M_{\infty} (I, \mathbb{Z}\cdot 1_R) \to M (I, R)$ there exists a matrix $v\in M_{\infty}(I,R)$ such that $$ \big(d-d_v\big) \big(e_{kk}(1_R)\big)=0 $$ for an arbitrary index $k\in I.$ \end{lemma}
\begin{proof} Let $$ d\big(e_{kk}(1_R)\big) =(a_{ij}^{(k)})_{i,j\in I}, \quad k\in I, \quad a_{ij}^{(k)} \in R. $$ We have $$ d\big(e_{kk}(1_R)\big)= d\big(e_{kk}(1_R)\big) e_{kk}(1_R) + e_{kk}(1_R)d\big(e_{kk}(1_R)\big) .$$ By Lemma \ref{lemma1}, $a_{ij}^{(k)}=0$ whenever $ i \neq k, $ $ j \neq k$ or $ i=j=k.$ Choose distinct indices $p,q \in I.$ Then $e_{pp}(1_R)e_{qq}(1_R) =0$ implies $$ (a_{ij}^{(p)})_{i,j\in I} \ e_{qq}(1_R)+ e_{pp}(1_R)\ (a_{kl}^{(q)})_{k,l\in I}=0. $$ The $(p,q)$-entry of the matrix on the left hand side is $$ a_{pq}^{(p)}+ a_{pq}^{(q)}=0. $$ Let $X=(x_{ij})_{i,j\in I},$ $x_{ij} \in R.$ The $(p,q)$-entry of the matrix $[X, e_{kk}(1_R)]$ is equal to $0$ if $p\neq k , $ $q\neq k;$ is equal to $-x_{pq}$ if $k=p;$ and to $x_{pq}$ if $k=q.$ The diagonal entries of the matrix $ [X, e_{kk}(1_R)] $ are equal to zero.
Define $a_{ij} =a_{ij}^{(j)}$ for $i \neq j;$ $a_{ii}=0;$ and $v=(a_{ij})_{i,j\in I}.$ By the above, $$ d\big(e_{kk}(1_R)\big) =(a_{ij}^{(k)})_{i,j\in I}= \big[v, e_{kk}(1_R)\big] $$ for an arbitrary index $k\in I.$
The $j$-th column of the matrix $v$ is equal to the $j$-th column of the matrix $ d\big(e_{kk}(1_R)\big) \in M(I,R).$ Hence, only finitely many entries in the $j$-th column are different from zero, hence $v\in M(I,R).$ This completes the proof of the lemma.
\end{proof}
\begin{lemma}\label{lemma3} Let $v\in M(I,R).$ Then
\begin{enumerate}
\item[\text{(a)}] $\big[ v, M_{\infty}(I, R)\big]\subseteq M_{rcf}(I,R)$ if and only if $v\in M_{rcf}(I,R);$
\item[\text{(b)}] if $v\in M_{rcf}(I,R),$ then in fact $\big[ v, M_{\infty}(I, R) \big] \subseteq M_{\infty}(I,R).$
\end{enumerate} \end{lemma}
\begin{proof} The assertion (b) immediately follows from the fact that $M_{\infty}(I, R)$ is a two-sided ideal in the ring $M_{rcf}(I, R).$
Let $$ v=(a_{ij})_{i,j\in I}\in M(I,R), \quad \big[ v,e_{kk}(1_R) \big] \in M_{rcf}(I,R) \quad \text{for any} \quad k\in I.$$ All entries in the $k$-th row of the matrix $v,$ except for the diagonal one, are negatives of the corresponding entries in the $k$-th row of the matrix $[v,e_{kk}(1_R)].$ Since $\big[ v, e_{kk}(1_R) \big] \in M_{rcf}(I,R),$ it follows that only finitely many entries of the $k$-th row of the matrix $v$ are different from zero. Hence, $v\in M_{rcf}(I,R).$ This completes the proof of the lemma.
\end{proof}
Now, we are ready to prove Theorem \ref{Th1}.
\begin{proof} Let $d$ be a derivation of the ring $M_{\infty}(I,R).$ By Lemma \ref{lemma2}, there exists a matrix $v\in M(I,R)$ such that $d\,' =d-d_v$ maps all matrix units $e_{kk}(1_R)$ to zero. This implies that $$ d\,'\big(e_{ij}(R)\big)\subseteq e_{ij}(R) $$ for all indices $i, j \in I.$
Define the additive mapping $d_{ij}:R\to R$ via $$ d\,' \big( e_{ij}(a) \big) =e_{ij}\big( d_{ij}(a)\big) .$$ Clearly, $d_{ij}(1_R)=0.$ Let $a,b\in R;$ $i, j, k \in I.$ We have $ e_{ij}(ab)=e_{ik}(a)e_{kj}(b).$ Hence, $$ d_{ij}(ab)=d_{ik}(a) \cdot b + a \cdot d_{kj}(b). $$ Let $b=1_R.$ Then $d_{ij}(a)=d_{ik}(a).$ This implies that $d_{ij}$ does not depend on $i,$ $j.$
Define $d=d_{ij};$ $i, j \in I.$ The mapping $d$ is a derivation of the ring $R.$ Hence, $d\,' \in \text{Der}(R).$ All derivations from $\text{Der}(R)$ map $M_{\infty}(I,R)$ into itself. Therefore, $$ d_v\big( M_{\infty}(I,R) \big) \subseteq M_{\infty}(I,R).$$ By Lemma \ref{lemma3}(a), it implies that $v\in M_{rcf}(I,R)$ and completes the proof of the first part of the theorem.
Let $d$ be a derivation of the ring $M(I,R).$ Arguing as above, we find a matrix $v\in M(I,R)$ such that $u=d- d_v \in \text{Der}(R).$ If $d$ is a derivation of the ring $ M_{rcf}(I,R),$ then the mapping $d_v$ maps $M_{\infty}(I,R)$ to $M_{rcf}(I,R).$ By Lemma \ref{lemma3}(a), $v\in M_{rcf}(I,R).$ This completes the proof of the theorem.
\end{proof}
\begin{thebibliography}{99}
\bibitem{2Bei_Bre_Cheb_Mart_1} Beidar K.I., Bre\v{s}ar M., Chebotar M.A., Martindale W.S. 3rd. \emph{On Herstein's Lie map conjectures I}. Trans. Amer. Math. Soc. 2001, \textbf{353}, 4235--4260. DOI:10.1090/S0002-9947-01-02731-3
\bibitem{3Bei_Bre_Cheb_Mart_2} Beidar K.I., Bre\v{s}ar M., Chebotar M.A., Martindale W.S. 3rd. \emph{On Herstein's Lie map conjectures II}. J. Algebra 2001, \textbf{238}, 239--264. DOI:10.1006/jabr.2000.8628
\bibitem{4Bei_Bre_Cheb_Mart_3} Beidar K.I., Bre\v{s}ar M., Chebotar M.A., Martindale W.S. 3rd. \emph{On Herstein's Lie map conjectures III}. J. Algebra 2002, \textbf{249}, 59--94. DOI:10.1006/jabr.2001.9076
\bibitem{1Bezushchak} Bezushchak O. \emph{Automorphisms and derivations of algebras of infinite matrices}. Linear Algebra Appl. 2022, \textbf{650}, 42--59. DOI:10.1016/j.laa.2022.05.020
\bibitem{5Hol_Zurek} Ho{\l}ubowski W., \.Zurek S., Lie algebras of infinite matrices: ideals and derivations, 2021. \textbf{arXiv}:1806.01099v2
\bibitem{6Jacobson} Nathan Jacobson, Lie algebras, Dover Publ., Inc, New Yourk (1979). Replication of the 1962 original.
\end{thebibliography}
\end{document}
Lie algebra of column-finite infinite matrices: Ideals and derivations https://doi.org/10.1016/j.jalgebra.2022.11.032
Journal of Algebra
Volume 619, 1 April 2023, Pages 517-537