% EJC papers *must* begin with the following two lines. \documentclass[12pt]{article} \usepackage{e-jc} \specs{P1.1}{23(1)}{2016} % Please remove all other commands that change parameters such as % margins or pagesizes. % only use standard LaTeX packages % only include packages that you actually need % we recommend these ams packages %\usepackage{amsthm,amsmath,amssymb} % we recommend the graphicx package for importing figures \usepackage{graphicx} % use this command to create hyperlinks (optional and recommended) \usepackage[colorlinks=true,citecolor=black,linkcolor=black,urlcolor=blue]{hyperref} % use these commands for typesetting doi and arXiv references in the bibliography \newcommand{\doi}[1]{\href{http://dx.doi.org/#1}{\texttt{doi:#1}}} \newcommand{\arxiv}[1]{\href{http://arxiv.org/abs/#1}{\texttt{arXiv:#1}}} % all overfull boxes must be fixed; % i.e. there must be no text protruding into the margins \usepackage{amscd,amsmath,amsthm,amssymb,graphics} \usepackage{pstricks} \usepackage{lmodern,pst-node} \usepackage{pstcol,pst-plot,pst-3d} %\usepackage{geometric} %\usepackage{mathtime} %\usepackage{mathptmx} \usepackage{multicol} \usepackage{epic,eepic} \usepackage{amsfonts,amssymb,amscd,amsmath,enumitem,verbatim} \psset{unit=0.7cm,linewidth=0.8pt,arrowsize=2.5pt 4} % for vertex a circle with radius 0.5 mm \def\vertex{\pscircle[fillstyle=solid,fillcolor=black]{0.03}} % for fat lines \newpsstyle{fatline}{linewidth=1.5pt} \newpsstyle{fyp}{fillstyle=solid,fillcolor=verylight} \definecolor{verylight}{gray}{0.97} \definecolor{light}{gray}{0.9} \definecolor{medium}{gray}{0.85} \definecolor{dark}{gray}{0.6} % Absolute value notation \newcommand{\abs}[1]{\lvert#1\rvert} % Blank box placeholder for figures (to avoid requiring any % particular graphics capabilities for printing this document). \newcommand{\blankbox}[2]{% \parbox{\columnwidth}{\centering % Set fboxsep to 0 so that the actual size of the box will match the % given measurements more closely. \setlength{\fboxsep}{0pt}% \fbox{\raisebox{0pt}[#2]{\hspace{#1}}}% }% } \unitlength=0.7cm \def\Box#1#2#3{\multiput(#1,#2)(1,0){2}{\line(0,1)1} \multiput(#1,#2)(0,1){2}{\line(1,0)1} \put(#1,#2){\makebox(1,1){$#3$}}} \def\LBox#1#2#3#4{\multiput(#1,#2)(#4,0){2}{\line(0,1)1} \multiput(#1,#2)(0,1){2}{\line(1,0){#4}} \put(#1,#2){\makebox(#4,1){$#3$}}} \def\Ci#1#2{\put(#1.5,#2.5){\circle{0.7}}} %------ GENERAL MACROS ----- % % Standard rings and fields, affine and projective space % \def\NZQ{\Bbb} % the font for N,Z,Q,R,C \def\NN{{\NZQ N}} \def\QQ{{\NZQ Q}} \def\ZZ{{\NZQ Z}} \def\RR{{\NZQ R}} \def\CC{{\NZQ C}} \def\AA{{\NZQ A}} \def\PP{{\NZQ P}} \def\FF{{\NZQ F}} \def\GG{{\NZQ G}} \def\HH{{\NZQ H}} \def\KK{{\NZQ K}} % %------------------------------------------------ % Symbols in "Fraktur" % \def\frk{\frak} % font for "Fraktur" \def\aa{{\frk a}} \def\pp{{\frk p}} \def\Pp{{\frk P}} \def\qq{{\frk q}} \def\Qq{{\frk Q}} \def\mm{{\frk m}} \def\Mm{{\frk M}} \def\Phi{{\frk n}} \def\Phi{{\frk N}} % %------------------------------------------------ \def\MI{{\mathcal I}} \def\MJ{{\mathcal J}} \def\MP{{\mathcal P}} \def\MQ{{\mathcal Q}} \def\MR{{\mathcal R}} \def\MV{{\mathcal V}} \def\ME{{\mathcal E}} \def\MK{{\mathcal K}} \def\MF{{\mathcal F}} \def\MT{{\mathcal T}} \def\MH{{\mathcal H}} \def\MS{{\mathcal S}} \def\MM{{\mathcal M}} \def\MG{{\mathcal G}} \def\MA{{\mathcal A}} \def\MN{{\mathcal N}} % Small letters in bold % \def\ab{{\bold a}} \def\bb{{\bold b}} \def\xb{{\bold x}} \def\yb{{\bold y}} \def\zb{{\bold z}} \def\gb{{\bold g}} \def\cb{{\bold c}} \def\db{{\bold d}} \def\fb{{\bold f}} % \def\opn#1#2{\def#1{\operatorname{#2}}} % to make operators %------------------------------------------------ % Numerical invariants of rings, ideals, and modules % \opn\chara{char} \opn\length{\ell} \opn\pd{pd} \opn\rk{rk} \opn\projdim{proj\,dim} \opn\injdim{inj\,dim} \opn\rank{rank} \opn\depth{depth} \opn\grade{grade} \opn\height{height} \opn\embdim{emb\,dim} \opn\codim{codim} \def\OO{{\mathcal O}} \opn\Tr{Tr} \opn\bigrank{big\,rank} \opn\superheight{superheight}\opn\lcm{lcm} \opn\trdeg{tr\,deg}%\emph{ \opn\reg{reg} \opn\lreg{lreg} \opn\ini{in} \opn\lpd{lpd} \opn\size{size}\opn\bigsize{bigsize} \opn\cosize{cosize}\opn\bigcosize{bigcosize} \opn\sdepth{sdepth}\opn\sreg{sreg} \opn\link{link}\opn\fdepth{fdepth} %------------------------------------------------ % Divisors % \opn\div{div} \opn\Div{Div} \opn\cl{cl} \opn\Cl{Cl} % %------------------------------------------------ \let\epsilon\varepsilon \let\phi=\varphi \let\kappa=\varkappa % Subsets of the spectrum of a ring % \opn\Spec{Spec} \opn\Supp{Supp} \opn\supp{supp} \opn\Sing{Sing} \opn\Ass{Ass} \opn\Min{Min}\opn\Mon{Mon} \opn\dstab{dstab} \opn\astab{astab} \opn\Syz{Syz} % %------------------------------------------------ % Standard operations on ideals and modules % \opn\Ann{Ann} \opn\Rad{Rad} \opn\Soc{Soc} % %------------------------------------------------ % Linear algebra and homology, endo- and automorphisms % \opn\Im{Im} \opn\Ker{Ker} \opn\Coker{Coker} \opn\Am{Am} \opn\Hom{Hom} \opn\Tor{Tor} \opn\Ext{Ext} \opn\End{End} \opn\Aut{Aut} \opn\id{id} \def\Frob{{\mathcal F}} \opn\nat{nat} \opn\pff{pf}% \pf exists already \opn\Pf{Pf} \opn\GL{GL} \opn\SL{SL} \opn\mod{mod} \opn\ord{ord} \opn\Gin{Gin} \opn\Hilb{Hilb}\opn\sort{sort} \opn\initial{init} \opn\ende{end} \opn\height{height} \opn\type{type} % %------------------------------------------------ % Convexity % \opn\aff{aff} \opn\con{conv} \opn\relint{relint} \opn\st{st} \opn\lk{lk} \opn\cn{cn} \opn\core{core} \opn\vol{vol} \opn\link{link} \opn\star{star}\opn\lex{lex} %------------------------------------------------ % Graded rings and Rees algebras \opn\gr{gr} \def\Rees{{\mathcal R}} % %------------------------------------------------ % Polynomials and power series % \def\poly#1#2#3{#1[#2_1,\dots,#2_{#3}]} \def\pot#1#2{#1[\kern-0.28ex[#2]\kern-0.28ex]} \def\Pot#1#2#3{\pot{#1}{#2_1,\dots,#2_{#3}}} \def\konv#1#2{#1\langle #2\rangle} \def\Konv#1#2#3{\konv{#1}{#2_1,\dots,#2_{#3}}} % %------------------------------------------------ % Direct and inverse limits % \opn\dirlim{\underrightarrow{\lim}} \opn\inivlim{\underleftarrow{\lim}} % % % Names with a meaning % \let\union=\cup \let\sect=\cap \let\dirsum=\oplus \let\tensor=\otimes \let\iso=\cong \let\Union=\bigcup \let\Sect=\bigcap \let\Dirsum=\bigoplus \let\Tensor=\bigotimes % %------------------------------------------------ % \let\to=\rightarrow \let\To=\longrightarrow \def\Implies{\ifmmode\Longrightarrow \else \unskip${}\Longrightarrow{}$\ignorespaces\fi} \def\implies{\ifmmode\Rightarrow \else \unskip${}\Rightarrow{}$\ignorespaces\fi} \def\iff{\ifmmode\Longleftrightarrow \else \unskip${}\Longleftrightarrow{}$\ignorespaces\fi} \let\gets=\leftarrow \let\Gets=\longleftarrow \let\followsfrom=\Leftarrow \let\Followsfrom=\Longleftarrow \let\:=\colon % declare theorem-like environments \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{lemma}[theorem]{Lemma} \newtheorem{corollary}[theorem]{Corollary} \newtheorem{proposition}[theorem]{Proposition} \newtheorem{fact}[theorem]{Fact} \newtheorem{observation}[theorem]{Observation} \newtheorem{claim}[theorem]{Claim} \theoremstyle{definition} \newtheorem{definition}[theorem]{Definition} \newtheorem{example}[theorem]{Example} \newtheorem{conjecture}[theorem]{Conjecture} \newtheorem{open}[theorem]{Open Problem} \newtheorem{problem}[theorem]{Problem} \newtheorem{question}[theorem]{Question} \theoremstyle{remark} \newtheorem{remark}[theorem]{Remark} \newtheorem{note}[theorem]{Note} %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \newcommand*{\bigtimes}{\mathop{\raisebox{-.5ex}{\hbox{\huge{$\times$}}}}} % % We like the var forms of some greek letters (as taught in German schools) % \let\epsilon\varepsilon %\let\phi=\varphi \let\kappa=\varkappa % % We print on A4 paper % %\textwidth=15cm \textheight=22cm \topmargin=0.5cm %\oddsidemargin=0.5cm \evensidemargin=0.5cm \pagestyle{plain} % % The pf environment of AMSART needs a little help % %\def\qed{\ifhmode\textqed\fi % \ifmmode\ifinner\quad\qedsymbol\else\dispqed\fi\fi} %\def\textqed{\unskip\nobreak\penalty50 % \hskip2em\hbox{}\nobreak\hfil\qedsymbol % \parfillskip=0pt \finalhyphendemerits=0} %\def\dispqed{\rlap{\qquad\qedsymbol}} %\def\noqed{\def\qed{\relax}} % % ------ END OF GENERAL MACROS ------- \opn\dis{dis} \def\pnt{{\raise0.5mm\hbox{\large\bf.}}} \def\lpnt{{\hbox{\large\bf.}}} \opn\Lex{Lex} %-- macro for local cohomology----------------------------- \def\Coh#1#2{H_{\mm}^{#1}(#2)} %-- macro for a complicated condition for the extended %-- Hochster's formula \def\hchst#1{for all $u\in G(I)$ there exists $i\notin {#1}$ such that $\nu_i(u) > a_i \geq 0$} % if needed include a line break (\\) at an appropriate place in the title \title{\bf Bi-Cohen-Macaulay graphs} % input author, affilliation, address and support information as follows; % the address should include the country, and does not have to include % the street address \author{J\"urgen Herzog\\ \small Fachbereich Mathematik\\[-0.8ex] \small Universit\"at Duisburg-Essen\\[-0.8ex] \small Essen, Germany\\ \small\tt juergen.herzog@uni-essen.de\\ \and Ahad Rahimi\thanks{This paper was written during the visit of the second author at Universit\"at Duisburg-Essen, Campus Essen. He is grateful for its hospitality.}\\ \small Department of Mathematics\\[-0.8ex] \small Razi University\\[-0.8ex] \small Kermanshah, Iran \\ \small\tt ahad.rahimi@razi.ac.ir } \date{\dateline{Sep 8, 2015}{Dec 9, 2015}{Jan 11, 2016}\\ \small Mathematics Subject Classifications: 05E40, 13C14} % \date{\dateline{submission date}{acceptance date}\\ % \small Mathematics Subject Classifications: comma separated list of % MSC codes available from http://www.ams.org/mathscinet/freeTools.html} %\date{\dateline{Jan 1, 2012}{Jan 2, 2012}\\ \begin{document} \maketitle \begin{abstract} In this paper we consider bi-Cohen-Macaulay graphs, and give a complete classification of such graphs in the case they are bipartite or chordal. General bi-Cohen-Macaulay graphs are classified up to separation. The inseparable bi-Cohen-Macaulay graphs are determined. We establish a bijection between the set of all trees and the set of inseparable bi-Cohen-Macaulay graphs. % keywords are optional \bigskip\noindent \textbf{Keywords:} Bi-Cohen--Macaulay, Bipartite and chordal graphs, Generic graphs, Inseparability \end{abstract} \section{Introduction} A simplicial complex $\Delta$ is called {\em bi-Cohen-Macaulay} (bi-CM), if $\Delta$ and its Alexander dual $\Delta^\vee$ are Cohen-Macaulay. This concept was introduced by Fl{\o}ystad and Vatne in \cite{FV}. In that paper the authors associated to each simplicial complex $\Delta$ in a natural way a complex of coherent sheaves and showed that this complex reduces to a coherent sheaf if and only if $\Delta$ is bi-CM. The present paper is an attempt to classify all bi-CM graphs. Given a field $K$ and a simple graph on the vertex set $[n]=\{1,2,\ldots,n\}$, one associates with $G$ the edge ideal $I_G$ of $G$, whose generators are the monomials $x_ix_j$ with $\{i,j\}$ an edge of $G$. We say that $G$ is bi-CM if the simplicial complex whose Stanley-Reisner ideal coincides with $I_G$ is bi-CM. Actually, this simplicial complex is the so-called {\em independence complex} of $G$. Its faces are the independent sets of $G$, that is, subsets $D$ of $[n]$ with $\{i,j\}\not\subset D$ for all edges $\{i,j\}$ of~$G$. By its very definition, any bi-CM graph is also a Cohen-Macaulay graph (CM graph). A complete classification of all CM graphs is hopeless if not impossible. However, such a classification is given for bipartite graphs \cite[Theorem 3.4]{HH1} and for chordal graphs \cite{HHZ}. We refer the reader to the books \cite{HH} and \cite{V} for a good survey on edge ideals and its algebraic and homological properties. Based on the classification of bipartite and chordal CM graphs, we provide in Section~\ref{class} a classification of bipartite and chordal bi-CM graphs, see Theorem~\ref{bipartite1} and Theorem~\ref{chordal}. In Section~\ref{properties} we first present various characterizations of bi-CM graphs. By using the Eagon-Reiner theorem \cite{ER}, one notices that the graph $G$ is bi-CM if and only if it is CM and $I_G$ has a linear resolution. Cohen-Macaulay ideals generated in degree 2 with linear resolution are of very special nature. They all arise as deformations of the square of the maximal ideal of a suitable polynomial ring. From this fact arise constraints on the number of edges of the graph and on the Betti numbers of $I_G$. Though a complete classification of all bi-CM graphs seems to be again impossible, a classification of all bi-CM graphs up to separation can be given, and this is the subject of the remaining sections. A {\em separation} of the graph $G$ with respect to the vertex $i$ is a graph $G'$ whose vertex set is $V(G)\union\{i'\}$ having the property that $G$ is obtained from $G'$ by identifying $i$ with $i'$ and such that $x_i-x_{i'}$ is a non-zerodivisor modulo $I_{G'}$. The algebraic condition on separation makes sure that the essential algebraic and homological invariants of $I_G$ and $I_{G'}$ are the same. In particular, $G$ is bi-CM if and only if $G'$ is bi-CM. A graph which does not allow any separation is called {\em inseparable}, and an inseparable graph which is obtained by a finite number of separation steps from $G$ is called an {\em inseparable model} of $G$. Any graph admits inseparable models and the number of inseparable models of a graph is finite. Separable and inseparable graphs from the view point of deformation theory have been studied in~\cite{ABHL}. In Section~\ref{main} we determine all inseparable bi-CM graphs on $n$ vertices. Indeed, in Theorem~\ref{maintheorem} it is shown that for any tree $T$ on the vertex set $[n]$ there exists a unique inseparable bi-CM graph $G_T$ determined by $T$, and any inseparable bi-CM graph is of this form. Furthermore, if $G$ is an arbitrary bi-CM graph and $T$ is the relation graph of the Alexander dual of $I_G$, then $G_T$ is a separable model of $G$. For a bi-CM graph $G$, the Alexander dual $J=(I_G)^\vee$ of $I_G$ is a Cohen-Macaulay ideal of codimension 2 with linear resolution. As described in \cite{BH1}, one attaches to any relation matrix of $J$ a relation tree $T$. Replacing the entries in this matrix by distinct variables with the same sign, one obtains the so-called {\em generic relation matrix} whose ideals of $2$-minors $J_T$ and its Alexander dual has been computed in \cite{N}. This theory is described in Section~\ref{generic}. The Alexander dual of $J_T$ is the edge ideal of graph, which actually is the graph $G_T$ mentioned before and which serves as a separable model of $G$. \section{Preliminaries and various characterizations of Bi-CM graphs} \label{properties} In this section we recall some of the standard notions of graph theory which are relevant for this paper, introduce the bi-CM graphs and present various equivalent conditions of a graph to be bi-CM. The graphs considered here will all be finite, simple graphs, that is, they will have no double edges and no loops. Furthermore we assume that $G$ has no isolated vertices. The vertex set of $G$ will be denoted $V(G)$ and will be the set $[n]=\{1,2,\ldots,n\}$, unless otherwise stated. The set of edges of $G$ we denote by $E(G)$. A subset $F\subset [n]$ is called a {\em clique} of $G$, if $\{i,j\}\in E(G)$ for all $i,j\in F$ with $i\neq j$. The set of all cliques of $G$ is a simplicial complex, denoted $\Delta(G)$. A subset $C\subset [n]$ is called a {\em vertex cover} of $G$ if $C\sect\{i,j\}\neq \emptyset$ for all edges $\{i,j\}$ of $G$. The graph $G$ is called {\em unmixed} if all minimal vertex covers of $G$ have the same cardinality. This concept has an algebraic counterpart. We fix a field $K$ and consider the ideal $I_G\subset S=K[x_1,\ldots,x_n]$ which is generated by all monomials $x_ix_j$ with $\{i,j\}\in E(G)$. The ideal $I_G$ is called the {\em edge ideal} of $G$. Let $C\subset [n]$. Then the monomial prime ideal $P_C=(\{x_i\: i\in C\})$ is a minimal prime ideal of $I_G$ if and only if $C$ is a minimal vertex cover of $G$. Thus $G$ is unmixed if and only if $I_G$ is unmixed in the algebraic sense. A subset $D\subset [n]$ is called an {\em independent set} of $G$ if $D$ contains no set $\{i,j\}$ which is an edge of $G$. Note that $D$ is an independent set of $G$ if and only if $[n]\setminus D$ is a vertex cover. Thus the minimal vertex covers of $G$ correspond to the maximal independent sets of $G$. The cardinality of a maximal independent set is called the {\em independence number} of $G$. It follows that the Krull dimension of $S/I_G$ is equal to $c$, where $c$ is the independence number of $G$. The graph $G$ is called {\em bipartite} if $V(G)$ is the disjoint union of $V_1$ and $V_2$ such that $V_1$ and $V_2$ are independent sets, and $G$ is called {\em disconnected} if $V(G)$ is the disjoint union of $W_1$ and $W_2$ and there is no edge $\{i,j\}$ of $G$ with $i\in W_1$ and $j\in W_2$. The graph $G$ is called {\em connected} if it is not disconnected. A {\em cycle} $C$ (of length $r$) in $G$ is a sequence of edges $\{i_k,j_k\}$ with $k=1,2,\ldots,r$ such that $j_k=i_{k+1}$ for $k=1,\ldots,r-1$ and $j_r=i_1$. A cord of $C$ is an edge $\{i,j\}$ of $G$ with $i,j\in\{i_1,\ldots,i_r\}$ and $\{i,j\}$ is not an edge of $C$. The graph $G$ is called {\em chordal} if each cycle of $G$ of length $\geq 4$ has a chord. A graph which has no cycle and which is connected is called a {\em tree}. \medskip Now we recall the main concept we are dealing with in this paper. Let $I\subset S$ be a squarefree monomial ideal. Then $I=\Sect_{j=1}^m P_{j}$ where each of the $P_j$ is a monomial prime ideal of $I$. The ideal $I^\vee$ which is minimally generated by the monomials $u_{j}=\prod_{x_i\in P_j}x_i$ is called the {\em Alexander dual} of $I$. One has $(I^\vee)^ \vee=I$. In the case that $I=I_G$, each $P_j$ is generated by the variables corresponding to a minimal vertex cover of $G$. Therefore, $(I_G)^\vee$ is also called the {\em vertex cover ideal} of $G$. According to \cite{FV} a squarefree monomial ideal $I\subset S$ is called {\em bi-Cohen-Macaulay} (or simply bi-CM) if $I$ as well as the Alexander dual $I^\vee$ of $I$ is a Cohen-Macaulay ideal. A graph $G$ is called {\em Cohen-Macaulay} or {\em bi-Cohen-Macaulay (over $K$)} (CM or bi-CM for short), if $I_G$ is CM or bi-CM. One important result regarding the Alexander dual that will be used frequently in this paper is the Eagon-Reiner theorem which says that $I$ is a Cohen-Macaulay ideal if and only if $I^\vee$ has a linear resolution. Thus the Eagon-Reiner theorem implies that $I$ is bi-CM if and only if $I$ is a Cohen-Macaulay ideal with linear resolution. From this description it follows that a bi-CM graph is connected. Indeed, if this is not the case, then there are induced subgraphs $G_1, G_2\subset G$ such that $V(G)$ is the disjoint union of $V(G_1)$ and $V(G_2)$. It follows that $I_G=I_{G_1}+I_{G_2}$, and the ideals $I_{G_1}$ and $I_{G_2}$ are ideals in a different set of variables. Therefore, the free resolution of $S/I_G$ is obtained as the tensor product of the resolutions of $S/I_{G_1}$ and $S/I_{G_2}$. This implies that $I_G$ has relations of degree 4, so that $I_G$ does not have a linear resolution. \medskip From now on we will always assume that $G$ is connected, without further mentioning it. \begin{proposition} \label{basic} Let $K$ be an infinite field and $G$ a graph on the vertex set $[n]$ with independence number $c$. The following conditions are equivalent: \begin{enumerate} \item[{\em (a)}] $G$ is a bi-CM graph over $K$; \item[{\em (b)}] $G$ is a CM graph over $K$, and $S/I_G$ modulo a maximal regular sequence of linear forms is isomorphic to $T/\mm_T^2$ where $T$ is the polynomial ring over $K$ in $n-c$ variables and $\mm_T$ is the graded maximal ideal of $T$. \end{enumerate} \end{proposition} \begin{proof} We only need to show that $I_G$ has a linear resolution if and only if condition (b) holds. Since $K$ is infinite and since $S/I_G$ is Cohen-Macaulay of dimension $c$, there exists a regular sequence $\xb$ of linear forms on $S/I_G$ of length $c$. Let $T=S/(\xb)$. Then $T$ is isomorphic to a polynomial ring in $n-c$ variables. Let $J$ be the image of $I_G$ in $T$. Then $J$ is generated in degree $2$ and has a linear resolution if and only if $I_G$ has linear resolution. Moreover, $J$ is $\mm_T$-primary. The only $\mm_T$-primary ideals with linear resolution are the powers of $\mm_T$. Thus, $I_G$ has a linear resolution if and only if $J=\mm_T^2$. \end{proof} \begin{corollary} \label{edges} Let $G$ be a graph on the vertex set $[n]$ with independence number $c$. The following conditions are equivalent: \begin{enumerate} \item[{\em (a)}] $G$ is a bi-CM graph over $K$; \item[{\em (b)}] $G$ is a CM graph over $K$ and $|E(G)|={n-c+1\choose 2}$; \item[{\em (c)}] $G$ is a CM graph over $K$ and the number of minimal vertex covers of $G$ is equal to $n-c+1$; \item[{\em (d)}] $\beta_i(I_G)=(i+1){n-c+1\choose i+2}$ for $i=0,\ldots, n-c-1$. \end{enumerate} \end{corollary} \begin{proof} For the proof of the equivalent conditions we may assume that $K$ is infinite and hence we may use Proposition~\ref{basic}. (a)\iff (b): With the notation of proposition~\ref{basic} we have $J=\mm_T^2$ if and only if the number of generators of $J$ is equal to ${n-c+1\choose 2}$. Since $I_G$ and $J$ have the same number of generators and since the number of generators of $I_G$ is equal to $|E(G)|$, the assertion follows. (b)\iff (c): Since $S/I_G$ is Cohen-Macaulay, the multiplicity of $S/I_G$ is equal to the length $\length(T/J)$ of $T/J$. On the other hand, the multiplicity is also the number of minimal prime ideals of $I_G$ which coincides with the number of minimal vertex covers of $G$. Thus the length of $T/J$ is equal to the number of minimal vertex covers of $G$. Since $J=\mm_T^2$ if and only if $\length(T/J)=n-c+1$, the assertion follows. (a)\implies (d): Note that $\beta_i(I_G)=\beta_i(J)$ for all $i$. Since $J$ is isomorphic to the ideal of $2$-minors of the matrix \[ \begin{pmatrix} y_{1} & y_2 & \ldots & y_{n-c} & 0 \\ 0 & y_1 & \ldots & y_{n-c-1} & y_{n-c} \end{pmatrix} \] in the variables $y_1,\ldots,y_{n-c}$, the Eagon-Northcott complex (\cite{BV}, \cite{E}) provides a free resolution of $J$ and the desired result follows. (d)\implies (a): It follows from the description of the Betti numbers of the ideal $I_G$ that $\projdim S/I_G = n-c$. Thus, $\depth S/I_G=c$. Since $\dim S/I_G=c$, it follows that $I_G$ is a Cohen-Macaulay ideal. Since $|E(G)|=\beta_0(I_G)={n-c+1\choose 2}$, condition (b) is satisfied, and hence $G$ is bi-CM, as desired. \end{proof} Finally we note that $G$ is a bi-CM graph over $K$ if and only if the vertex cover ideal of $G$ is a codimension $2$ Cohen-Macaulay ideal with linear relations. Indeed, let $J_G$ be the vertex cover ideal of $G$. Since $J_G=(I_G)^\vee$, it follows from the Eagon-Reiner theorem $J_G$ is bi-CM if and only if $I_G$ is bi-CM. \section{The classification of bipartite and chordal bi-CM graphs} \label{class} In this section we give a full classification of the bipartite and chordal bi-CM graphs. \begin{theorem} \label{bipartite1} Let $G$ be a bipartite graph on the vertex set $V$ with bipartition $V=V_1\union V_2$ where $V_1=\{v_1,\ldots,v_n\}$ and $V_2=\{w_1,\ldots,w_m\}$. Then the following conditions are equivalent: \begin{enumerate} \item[{\em (a)}] $G$ is a bi-CM graph; \item[{\em (b)}] $n=m$ and $E(G)=\{\{v_i,w_j\}\: 1\leq i\leq j\leq n\}$. \end{enumerate} \end{theorem} \begin{proof} (a)\implies (b): Since $G$ is a bi-CM graph, it is in particular a CM-graph, and so $n=m$, and by \cite[Theorem 9.1.13]{HH} there exists a poset $P = \{p_1,\dots, p_n\}$ such that $G=G(P)$. Here $G(P)$ is the bipartite graph on $V=\{v_1,\ldots,v_n, w_1,\ldots,w_n\}$ whose edges are those $2$-element subset $\{v_i, w_j\}$ of $V$ such that $p_i\leq p_j$. Thus $I_G=I_{G(P)}= H_P^\vee$, where \[ H_P=\bigcap_{p_i\leq p_j}(x_i, y_j) \] is an ideal of $S=K[\{x_i,y_i\}_{p_i\in P}]$, the polynomial ring in $2n$ variables over $K$. Since $G$ is bi-CM, it follows that $H_P$ is Cohen--Macaulay, and hence \[ \projdim S/H_P=2n-\depth S/H_P=2n-\dim S/H_P=\height H_P=2. \] Thus $\projdim H_P=1$, and hence, by \cite[Corollary 2.2]{HH1}, the Sperner number of $P$, i.e., the maximum of the cardinalities of antichains of $P$ equals $1$. This implies that $P$ is a chain, and this yields (b). (b)\implies (a): The graph $G$ described in (b) is of the form $G=G(P)$ where $P$ is a chain. By what is said in (a)\implies (b), it follows that $G$ is bi-CM. \end{proof} The following picture shows a bi-CM bipartite graph for $n=4$. \begin{figure}[hbt] \begin{center} \psset{unit=0.8cm} \begin{pspicture}(-0.2,-2)(2,2) \rput(-0.8,-1){ \rput(0,0){$\bullet$} \rput(0,1.5){$\bullet$} \rput(1.5,0){$\bullet$} \rput(1.5,1.5){$\bullet$} \rput(3, 0){$\bullet$} \rput(3, 1.5){$\bullet$} \rput(4.5, 0){$\bullet$} \rput(4.5, 1.5){$\bullet$} \psline(0,.0)(0, 1.5) \psline(1.5,.0)(1.5,1.5) \psline(3, .0)(3, 1.5) \psline(4.5, .0)(4.5,1.5) \psline(0.0, 1.5)(1.5,.0) \psline(0.0, 1.55)(3,.0) \psline(0.03, 1.5)(4.5,.0) \psline(1.5, 1.5)(3,.0) \psline(1.5, 1.5)(4.5,.0) \psline(3,1.5)(4.5,.0) \rput(0,2){$x_{1}$} \rput(1.5,2){$x_{2}$} \rput(3, 2){$x_{3}$} \rput(4.5,2){$x_{4}$} \rput(0,-.5){$y_{1}$} \rput(1.5,-.5){$y_{2}$} \rput(3, -.5){$y_{3}$} \rput(4.5,-.5){$y_{4}$} } \end{pspicture} \end{center} \caption{A bi-CM bipartite graph.}\label{bipartite} \end{figure} \begin{theorem} \label{chordal} Let $G$ be a chordal graph on the vertex set $[n]$. The following conditions are equivalent: \begin{enumerate} \item[{\em (a)}] $G$ is a bi-CM graph; \item[{\em (b)}] Let $F_1,\ldots, F_m$ be the facets of the clique complex of $G$. Then $m=1$, \\ or $m>1$ and \begin{enumerate} \item[{\em (i)}] $V(G)=V(F_1)\union V(F_2)\union \ldots \union V(F_m)$, and this union is disjoint; \item[{\em (ii)}] each $F_i$ has exactly one free vertex $j_i$; \item[{\em (iii)}] the restriction of $G$ to $[n]\setminus\{j_1,\ldots,j_m\}$ is a clique. \end{enumerate} \end{enumerate} \end{theorem} \begin{proof} Let $I_{n,d}$ be the ideal generated by all squarefree monomials of degree $d$ in $S=K[x_1,\ldots,x_n]$. It is known (and easy to prove) that $I_{n,d}^\vee=I_{n,n-d+1}$, and that all these ideals are Cohen-Macaulay, and hence all bi-CM. If $m=1$, then $I_G=I_{n,2}$ and the result follows. Now let $m>1$. A bi-CM graph is a CM graph. The CM chordal graphs have been classified in \cite{HHZ}: they are the chordal graphs satisfying (b)(i). Thus for the proof of the theorem we may assume that (b)(i) holds and simply have to show that (b)(ii) and (b)(iii) are satisfied if and only if $I_G$ has a linear resolution. Let $P_i$ be the monomial prime ideal generated by the variables $x_k$ with $k\in V(F_i)\setminus \{j_i\}$, and let $G'$ be subgraph of $G$ whose edges do not belong to any $F_i$. It is shown in the proof of \cite[Corollary 2.1]{HHZ} that there exists a regular sequence on $S/I_G$ such that after reduction modulo this sequence one obtains the ideal $J\subset T$ where $T$ is the polynomial ring on the variables $x_k$ with $k\neq j_i$ for $i=1,\ldots,m$ and where \begin{eqnarray} \label{reduction} J=(P_1^2,\ldots, P_m^2, I_{G'}). \end{eqnarray} By Proposition~\ref{basic}, it follows that $I_G$ has a linear resolution if and only if $J=\mm_T^2$, where $\mm_T$ denotes the graded maximal ideal of $T$. So, now suppose first that $I_G$ has a linear resolution, and hence $J=\mm_T^2$. Suppose that some $F_i$ has more than one free vertex, say $F_i$ has the vertex $k$ with $k\neq j_i$. Choose any $F_t$ different from $F_i$ and let $l\in F_j$ with $l\neq j_t$. Then $x_k$ and $x_l$ belong to $T$ but $x_kx_l\not\in J$ as can be seen from (\ref{reduction}). This is a contradiction. Thus (b)(ii) follows. Suppose next that the graph $G''$ which is the restriction of $G$ to $[n]\setminus\{j_1,\ldots,j_m\}$ is not a clique. Then there exist $i,j\in V(G'')$ such that $\{i,j\}\not\in E(G'')$. However, since all $x_k$ with $k\in V(G'')$ belong to $T$ and since $J=\mm_T^2$, it follows $x_ix_j\in J$. Thus, by (\ref{reduction}), $x_ix_j \in P_k^2$ for some $k$ or $x_ix_j\in I_{G'}$. Since (b)(ii) holds, this implies in both cases that $\{i,j\}\in E(G'')$, a contradiction. Thus (b)(iii) follows. Conversely, suppose (b)(ii) and (b)(iii) hold. We want to show that $J=\mm_T^2$. Let $x_i,x_j\in T$. We have to show that $x_ix_j\in J$. It follows from the description of $J$ that $x_k^2\in J$ for all $x_k\in T$. Thus we may assume that $i\neq j$. If $\{i,j\}$ is not an edge of any $F_k$, then by definition it is an edge of $G'$, and hence $x_ix_j\in I_{G'}\subset J$. On the other hand, if $ \{i,j\}$ is an edge of $F_k$ for some $k$, then $i,j\neq i_k$, and hence $x_ix_j\in P_k^2\subset J$. Thus the desired conclusion follows. \end{proof} Let $G$ be a chordal bi-CM graph as in Theorem~\ref{chordal}(b) with $m>1$. We call the complete graph $G''$ which is the restriction of $G$ to $[n]\setminus\{j_1,\ldots,j_m\}$ the {\em center} of $G$. The following picture shows, up to isomorphism, all bi-CM chordal graphs whose center is the complete graph $K_4$ on $4$ vertices: \begin{figure}[hbt] \begin{center} \psset{unit=0.7cm} \begin{pspicture}(0, -1)(2,1) \rput(0.3,-1){ \rput(0,0){$\bullet$} \rput(1.5,0){$\bullet$} \rput(1.5,1.5){$\bullet$} \rput(3, 1.5){$\bullet$} \rput(3, 0){$\bullet$} \rput(-1.5, 1.5){$\bullet$} \rput(0, 1.5){$\bullet$} \rput(-1.5, 0){$\bullet$} \psline(0,0)(0,1.5) \psline(0,0)(1.5,0) \psline(1.5,0)(1.5,1.5) \psline(0, 1.5)(1.5,1.5) \psline(0, 0)(1.5,1.5) \psline(0, 1.5)(1.5,0) \psline(1.5, 1.5)(3,1.5) \psline(1.5, 0)(3,0) \psline(0,0)(-1.5,0) \psline(0,1.5)(-1.5,1.5) } \end{pspicture} \psset{unit=0.7cm} \begin{pspicture}(-4.5, -1)(2,1) \rput(0.5,-1){ \rput(0,0){$\bullet$} \rput(0,1.5){$\bullet$} \rput(1.5,0){$\bullet$} \rput(1.5,1.5){$\bullet$} \rput(3, 1.5){$\bullet$} \rput(3, 0){$\bullet$} \rput(-1.5, .75){$\bullet$} \psline(0,0)(0,1.5) \psline(0,0)(1.5,0) \psline(1.5,0)(1.5,1.5) \psline(0, 1.5)(1.5,1.5) \psline(0, 0)(1.5,1.5) \psline(0, 1.5)(1.5,0) \psline(1.5, 1.5)(3,1.5) \psline(1.5, 0)(3,0) \psline(0,0)(-1.5,.75) \psline(0,1.5)(-1.5,.75) } \end{pspicture} %\end{center} %\caption{aaa}\label{triangle} %\end{figure} %\begin{figure}[hbt] %\begin{center} \psset{unit=0.7cm} \begin{pspicture}(-.5,-1)(3,2) \rput(-.3,-1){ \rput(0,0){$\bullet$} \rput(0,1.5){$\bullet$} \rput(1.5,0){$\bullet$} \rput(1.5,1.5){$\bullet$} %\rput(3, 1.5){$\bullet$} \rput(3, .75){$\bullet$} \rput(-1.5, .75){$\bullet$} \psline(0,0)(0,1.5) \psline(0,0)(1.5,0) \psline(1.5,0)(1.5,1.5) \psline(0, 1.5)(1.5,1.5) \psline(0, 0)(1.5,1.5) \psline(0, 1.5)(1.5,0) \psline(1.5, 1.5)(3,.75) \psline(1.5, 0)(3,.75) \psline(0,0)(-1.5,.75) \psline(0,1.5)(-1.5,.75) } \end{pspicture} \psset{unit=0.7cm} \begin{pspicture}(-2,-1)(2,1) \rput(-1,-0.27){ \rput(0,0){$\bullet$} \rput(1.5,0){$\bullet$} \rput(2.56,1.05){$\bullet$} \rput(2.56,-1.05){$\bullet$} \rput(3.62, 0){$\bullet$} \rput(5.12,0){$\bullet$} \psline(0,0)(1.5,0) \psline(1.5,0)(3.62,0) \psline(1.5,0)(2.56,1.05) \psline(1.5, 0)(2.56,-1.05) \psline(2.56, 1.05)(2.56,-1.05) \psline(2.56, 1.05)(3.62,0) \psline(2.56, -1.05)(3.62,0) \psline(2.56, 1.05)(5.12,0) \psline(2.56, -1.05)(5.12,0) \psline(5.12, 0)(3.62,0) } \end{pspicture} \end{center} \caption{}\label{triangle1} \end{figure} \section{Generic Bi-CM graphs} \label{generic} As we have already seen in the first section, the Alexander dual $J=I^\vee_G$ of the edge ideal of a bi-CM graph $G$ is a Cohen--Macaulay ideal of codimension 2 with linear resolution. The ideal $J$ may have several distinct relation matrices with respect to the unique minimal monomial set of generators of $J$. As shown in \cite{BH1}, one may attach to each of the relation matrices of $J$ a tree as follows: let $u_1,\ldots,u_m$ be the unique minimal set of generators of $J$. Let $A$ be one of the relation matrices of $J$. Because $J$ has a linear resolution, the generating relations of $J$ may be chosen all of the form $x_ku_i-x_lu_j=0$. This implies that in each row of the $(m-1)\times m$-relation matrix $A$ there are exactly two non-zero entries (which are variables with different signs). We call such relations, {\em relations of binomial type}. \begin{example} \label{relation} Consider the bi-CM graph $G$ on the vertex set $[5]$ and edges $\{1,2\}$ $\{2,3\}$, $\{3,1\}$, $\{2,4\}$, $\{3,4\}$, $\{4,5\}$ as displayed in Figure~\ref{a}. \begin{figure}[hbt] \begin{center} \psset{unit=0.7cm} \begin{pspicture}(-2,-3)(3,1) \rput(-0.8,-1){ \rput(0,0){$\bullet$} \rput(1.5,1){$\bullet$} \rput(1.5,-1){$\bullet$} \rput(3,0){$\bullet$} \rput(4.5,0){$\bullet$} \psline(0,0)(1.5,1) \psline(0,0)(1.5,-1) \psline(1.5,1)(1.5,-1) \psline(1.5,1)(3,0) \psline(1.5,-1)(3,0) \psline(3,0)(4.5,0) \rput(-0.5,0){$x_{1}$} \rput(1.5,1.4){$x_{2}$} \rput(1.5,-1.4){$x_{3}$} \rput(3,0.5){$x_{4}$} \rput(4.5, 0.5){$x_{5}$} } \end{pspicture} \end{center} \caption{} \label{a} \end{figure} The ideal $J=I_G^\vee$ is generated by $u_1=x_2x_3x_4$, $u_2=x_1x_3x_4$, $u_3=x_2x_3x_5$ and $u_4=x_1x_2x_4$. The relation matrices with respect to $u_1,u_2,u_3$ and $u_4$ are the matrices \[ A_1= \begin{pmatrix} x_1 & -x_2 & 0 & 0 \\ x_5 & 0 & -x_4 & 0 \\ x_1 & 0 & 0 & -x_3 \end{pmatrix}, \] and \[ A_2= \begin{pmatrix} x_1 &-x_2 & 0 & 0 \\ x_5 & 0 &-x_4 & 0 \\ 0 & x_2 & 0 & -x_3 \end{pmatrix}. \] \end{example} \medskip \noindent Coming back to the general case, one assigns to the relation matrix $A$ the following graph $\Gamma$: the vertex set of $\Gamma$ is the set $V(\Gamma)=\{1,2,\ldots,m\}$, and $\{i,j\}$ is said to be an edge of $\Gamma$ if and only if some row of $A$ has non-zero entries for the $i$th- and $j$th-component. It is remarked in \cite{BH1} and easy to see that $\Gamma$ is a tree. This tree is in general not uniquely determined by $G$. \medskip In our Example~\ref{relation} the relation tree of $A_1$ is \begin{figure}[hbt] \begin{center} \psset{unit=0.8cm} \begin{pspicture}(-1,-2)(4,1.5) \rput(-0.8,-1){ \rput(0,0){$\bullet$} \rput(1.5,0){$\bullet$} \rput(3,0){$\bullet$} \rput(1.5,1.4){$\bullet$} \psline(0,0)(1.5,0) \psline(1.5,0)(3,0) \psline(1.5,0)(1.5,1.4) \rput(0,-0.5){$x_{4}$} \rput(1.5,-0.5){$x_{1}$} \rput(3,-0.5){$x_{2}$} \rput(1.5, 1.9){$x_{3}$} } \end{pspicture} \end{center} \caption{}\label{b} \end{figure} while the relation tree of $A_2$ is \begin{figure}[hbt] \begin{center} \psset{unit=0.8cm} \begin{pspicture}(0,-2)(3,0) \rput(-0.8,-1){ \rput(0,0){$\bullet$} \rput(1.5,0){$\bullet$} \rput(3,0){$\bullet$} \rput(4.5,0){$\bullet$} \psline(0,0)(1.5,0) \psline(1.5,0)(3,0) \psline(3,0)(4.5,0) \rput(0,-0.5){$x_{3}$} \rput(1.5,-0.5){$x_{1}$} \rput(3,-0.5){$x_{2}$} \rput(4.5,-0.5){$x_{4}$} } \end{pspicture} \end{center} \caption{}\label{c} \end{figure} Now let $J$ be any codimension 2 Cohen-Macaulay monomial ideal with linear resolution. Then, as observed in Section~\ref{properties}, $J^\vee=I_G$ where $G$ is a bi-CM graph. Now we follow Naeem \cite{N} and define for any given tree $T$ on the vertex set $[m]=\{1,\ldots,m\}$ with edges $e_1,\ldots,e_{m-1}$ the $(m-1)\times m$-matrix $A_T$ whose entries $a_{kl}$ are defined as follows: we assign to the $k$th edge $e_k=\{i,j\}$ of $T$ with $i