Russian Mathematical Surveys
RUS  ENG    JOURNALS   PEOPLE   ORGANISATIONS   CONFERENCES   SEMINARS   VIDEO LIBRARY   PACKAGE AMSBIB  
General information
Latest issue
Archive
Impact factor
Submit a manuscript

Search papers
Search references

RSS
Latest issue
Current issues
Archive issues
What is RSS



Uspekhi Mat. Nauk:
Year:
Volume:
Issue:
Page:
Find






Personal entry:
Login:
Password:
Save password
Enter
Forgotten password?
Register


Russian Mathematical Surveys, 2022, Volume 77, Issue 4, Pages 762–765
DOI: https://doi.org/10.4213/rm10065e
(Mi rm10065)
 

This article is cited in 1 scientific paper (total in 1 paper)

Brief Communications

Monomial non-Golod face rings and Massey products

I. Yu. Limonchenkoa, T. E. Panovbac

a National Research University Higher School of Economics
b Lomonosov Moscow State University
c Institute for Information Transmission Problems of the Russian Academy of Sciences
References:
Funding agency Grant number
HSE Basic Research Program
Contest «Young Russian Mathematics»
This study was funded within the framework of the HSE University Basic Research Program. The first author is a Young Russian Mathematics award winner, and he would like to thank the sponsors and jury of the contest.
Received: 05.04.2022
Russian version:
Uspekhi Matematicheskikh Nauk, 2022, Volume 77, Issue 4(466), Pages 203–204
DOI: https://doi.org/10.4213/rm10065
Bibliographic databases:
Document Type: Article
MSC: 13F55, 55S30
Language: English
Original paper language: Russian

In this paper a criterion for the Golodness of the face ring $\Bbbk[K]$ of a simplicial complex $K$ over the field $\Bbbk$ is obtained. A similar criterion was proposed in [4], but one of the assertions there depended on the main result of [1], which was shown to be false in [5]. Our proof fills this gap. We also construct an example of a minimally non-Golod complex $K$ such that the cohomology of the corresponding moment-angle complex $\mathcal Z_K$ has a trivial cup product and a non-trivial triple Massey product.

Let $K$ be a simplicial complex on the vertex set $[m]=\{1,2,\dots,m\}$. The face ring $\Bbbk[K]:=\Bbbk[v_1,\dots,v_m]/(v_{i_1}\cdots v_{i_r}\mid\{i_1,\dots,i_r\} \notin K)$ is said to be Golod (over $\Bbbk$) if the product and all higher Massey products in the Koszul complex $(\Lambda[u_1,\dots,u_m]\otimes\Bbbk[K],d)$ are trivial. By [3], $\Bbbk[K]$ is a Golod ring if and only if Serre’s inequality, relating the Hilbert series of $\operatorname{Ext}_{\Bbbk[K]}(\Bbbk,\Bbbk)$ and $\operatorname{Tor}_{\Bbbk[v_1,\dots,v_m]}(\Bbbk,\Bbbk[K])$, turns to equality. If $\Bbbk[K]$ is not Golod, but $\Bbbk[K_{[m]\setminus\{i\}}]$ is Golod for each $i\in [m]$, then $\Bbbk[K]$ is called minimally non-Golod (over $\Bbbk$).

Given a topological pair $(X,A)$, its polyhedral product $(X,A)^K$ is defined as $\bigcup_{\sigma\in K}(X,A)^{\sigma}$ for $(X,A)^{\sigma}:=\prod_{i\in [m]}X_i$, where $X_i=X$ if $i\in\sigma$ and $X_i=A$ otherwise. Recall that $\mathcal Z_K:=(\mathbb D^2,\mathbb S^1)^K$ and $\mathit{DJ}(K):=(\mathbb{C}\mathbb P^\infty,\ast)^K$. The Koszul complex $(\Lambda[u_1,\dots,u_m]\otimes\Bbbk[K],d)$ is quasi-isomorphic to the cellular cochains of $\mathcal Z_K$ with an appropriate diagonal approximation ([2], Lemma 4.5.3); in particular, $H^*(\mathcal Z_K;\Bbbk)\cong \operatorname{Tor}_{\Bbbk[v_1,\dots,v_m]}(\Bbbk,\Bbbk[K])$.

Theorem 1. Let $\Bbbk$ be a field. Then the following conditions are equivalent:

(a) $\Bbbk[K]$ is a Golod ring over $\Bbbk$;

(b) the cup product and all Massey products in $H^{+}(\mathcal Z_K;\Bbbk)$ are trivial;

(c) $H_{*}(\Omega\mathcal Z_K;\Bbbk)$ is a graded free associative algebra;

(d) the Hilbert series satisfy the identity

$$ \begin{equation*} \operatorname{Hilb}(H_{*}(\Omega\mathcal Z_K;\Bbbk);t) =\frac1{1-\operatorname{Hilb}(\Sigma^{-1}\widetilde{H}^{*}(\mathcal Z_K;\Bbbk);t)}. \end{equation*} \notag $$

Proof. The equivalence (a) $\Leftrightarrow$ (b) follows from [2], Theorem 4.5.4.

For (a) $\Leftrightarrow$ (d), a theorem of Golod [3] asserts that $\Bbbk[K]$ is a Golod ring if and only if the following identity holds for the Hilbert series:

$$ \begin{equation*} \operatorname{Hilb}\bigl(\operatorname{Ext}_{\Bbbk[K]}(\Bbbk,\Bbbk);t\bigr) =\frac{(1+t)^m}{1-\sum_{i,j>0}\beta^{-i,2j}(\Bbbk[K])t^{-i+2j-1}}\,, \end{equation*} \notag $$
where
$$ \begin{equation*} \beta^{-i,2j}(\Bbbk[K])=\dim\operatorname{Tor}^{-i,2j}_{\Bbbk[v_1,\dots,v_m]}(\Bbbk,\Bbbk[K]). \end{equation*} \notag $$
By [2], Proposition 8.4.10, there is an algebra isomorphism $H_{*}(\Omega\mathit{DJ}(K);\Bbbk)\cong\operatorname{Ext}_{\Bbbk[K]}(\Bbbk,\Bbbk)$. The loop space decomposition
$$ \begin{equation*} \Omega\mathit{DJ}(K)\simeq\Omega\mathcal Z_K\times\mathbb{T}^m \end{equation*} \notag $$
([2], (8.15)) implies that
$$ \begin{equation*} \operatorname{Hilb}(H_{*}(\Omega\mathit{DJ}(K);\Bbbk);t) =\operatorname{Hilb}(H_{*}(\Omega\mathcal Z_K;\Bbbk);t)\cdot (1+t)^m. \end{equation*} \notag $$
Also,
$$ \begin{equation*} \operatorname{Hilb}(\Sigma^{-1}\widetilde{H}^{*}(\mathcal Z_K;\Bbbk);t) =\sum_{i,j>0}\beta^{-i,2j}(\Bbbk[K])t^{-i+2j-1} \end{equation*} \notag $$
by [2], Theorem 4.5.4. Substituting this in yields the identity in (d).

We prove that (c) $\Rightarrow$ (d). Let

$$ \begin{equation*} Q=H_{>0}(\Omega\mathcal Z_K;\Bbbk)/(H_{>0}(\Omega\mathcal Z_K;\Bbbk)\cdot H_{>0}(\Omega\mathcal Z_K;\Bbbk)) \end{equation*} \notag $$
be the space of indecomposables. By assumption $H_{*}(\Omega\mathcal Z_K;\Bbbk)=T\langle Q\rangle$, where $T\langle Q\rangle$ is the free associative algebra on the graded $\Bbbk$-module $Q$. The Milnor–Moore (bar) spectral sequence has the $E_2$-term $E_{2}^{b}= \operatorname{Tor}_{H_{*}(\Omega\mathcal Z_K;\Bbbk)}(\Bbbk,\Bbbk)$ and converges to $\Sigma^{-1}H_{*}(\mathcal Z_K;\Bbbk)$. By assumption $E_{2}^{b}\cong\operatorname{Tor}_{H_{*}(T\langle Q\rangle)}(\Bbbk,\Bbbk) \cong\Bbbk\oplus Q$ (as $\Bbbk$-modules), so
$$ \begin{equation*} \operatorname{Hilb}(\Sigma^{-1}\widetilde{H}_{*}(\mathcal Z_K;\Bbbk);t)= {\operatorname{Hilb}(E_{\infty}^{b};t)-1}\leqslant \operatorname{Hilb}(E_{2}^{b};t)-1=\operatorname{Hilb}(Q;t). \end{equation*} \notag $$
In particular,
$$ \begin{equation*} \operatorname{Hilb}(T\langle\Sigma^{-1} \widetilde{H}_{*}(\mathcal Z_K;\Bbbk)\rangle;t)\leqslant \operatorname{Hilb}(T\langle Q\rangle;t)= \operatorname{Hilb}(H_{*}(\Omega\mathcal Z_K;\Bbbk);t). \end{equation*} \notag $$
The Adams (cobar) spectral sequence has $E_{2}^{c}=\operatorname{Cotor}_{H_{*}(\mathcal Z_K;\Bbbk)}(\Bbbk;\Bbbk)$ and converges to $H_{*}(\Omega\mathcal Z_K;\Bbbk)$. We have
$$ \begin{equation*} \operatorname{Hilb}(H_{*}(\Omega\mathcal Z_K;\Bbbk);t)= \operatorname{Hilb}(E_{\infty}^{c};t)\leqslant \operatorname{Hilb}(E_{2}^{c};t)\leqslant \operatorname{Hilb}(T\langle\Sigma^{-1} \widetilde{H}_{*}(\mathcal Z_K;\Bbbk)\rangle;t), \end{equation*} \notag $$
where the last inequality follows from the cobar construction (it turns to equality when all differentials in the cobar construction on $H_{*}(\mathcal Z_K;\Bbbk)$ vanish). By combining the two inequalities we obtain
$$ \begin{equation*} \begin{aligned} \, \operatorname{Hilb}\bigl(H_{*}(\Omega\mathcal Z_K;\Bbbk);t\bigr) &=\operatorname{Hilb}\bigl(T\langle\Sigma^{-1} \widetilde{H}_{*}(\mathcal Z_K;\Bbbk)\rangle;t\bigr) \\ &=\frac1{1-\operatorname{Hilb}(\Sigma^{-1}\widetilde{H}^{*}(\mathcal Z_K;\Bbbk);t)}\,. \end{aligned} \end{equation*} \notag $$

To prove that (d) $\Rightarrow$ (c), observe that the identity in (d) is equivalent to

$$ \begin{equation*} \operatorname{Hilb}(H_{*}(\Omega\mathcal Z_K;\Bbbk);t)= \operatorname{Hilb} (T\langle\Sigma^{-1}\widetilde{H}_{*}(\mathcal Z_K;\Bbbk)\rangle;t). \end{equation*} \notag $$
This implies that all differentials in the Adams cobar construction on $H_{*}(\mathcal Z_K;\Bbbk)$ are trivial. Thus $H_{*}(\Omega\mathcal Z_K;\Bbbk)$ is a free associative algebra (on $\Sigma^{-1}\widetilde{H}_{*}(\mathcal Z_K;\Bbbk)$). $\Box$

In the case of flag $K$ it was proved in [4] that $\Bbbk[K]$ is Golod if and only if $\operatorname{cup}(\mathcal Z_K)=1$, and if $\Bbbk[K]$ is minimally non-Golod, then $\operatorname{cup}(\mathcal Z_K)=2$. In general, for minimally non-Golod $\Bbbk[K]$ we have an upper bound $\operatorname{cup}(\mathcal Z_K)\leqslant2$, which follows easily from Baskakov’s description of the product in $H^*(\mathcal Z_K;\Bbbk)$: see [2], Theorem 4.5.4. Building upon a construction in [5] we give an example of a minimally non-Golod complex $\mathcal K$ such that $\mathrm{cup}(\mathcal Z_\mathcal K)=1$ and $H^*(\mathcal Z_\mathcal K)$ has a non-trivial triple Massey product.

Theorem 2. Let $\mathcal K$ be given by its minimal non-faces $(1,2,3)$, $(4,5,6)$, $(7,8,9)$, $(1,4,7)$, $(1,2,4,5)$, $(5,6,7,8)$, $(2,3,7,8)$, $(2,3,5,6,7)$, $(1,2,4,6,8,9)$, $(1,3,4,5,8,9)$, $(1,3,5,6,7,9)$, $(2,3,4,5,7,9)$, $(2,3,4,5,8,9)$, $(2,3,4,6,7,9)$, $(2,3,5,6,8,9)$. Then $\mathcal K$ is a $4$-dimensional minimally non-Golod complex such that $\operatorname{cup}(\mathcal Z_{\mathcal K})=1$ and there exists a non-trivial indecomposable triple Massey product of $5$-dimensional Koszul cohomology classes in $H^{14}(\mathcal Z_\mathcal K)$:

$$ \begin{equation*} \langle[v_1v_2u_3],[v_5v_6u_4],[v_7v_8u_9]\rangle=\{[v_1v_2v_5v_7v_8u_3u_4u_6u_9]\}. \end{equation*} \notag $$

Proof. The description of the cup product for $\mathcal Z_\mathcal K$ ([2], Theorem 4.5.4) implies that $\operatorname{cup}(\mathcal Z_\mathcal K)=1$. The full subcomplex $\mathcal K_{[m]\setminus\{i\}}$ is Golod for any $i\in [m]$ by [5], Theorem 6.3, (5), so $\mathcal K$ is minimally non-Golod. It remains to show that the triple Massey product above is defined, non-trivial, and indecomposable. It is defined and single-valued due to [6], Lemma 3.3, since
$$ \begin{equation*} \widetilde{H}^*(\mathcal K_{\{1,2,3,4,5,6\}})=\widetilde{H}^*(\mathcal K_{\{4,5,6,7,8,9\}})=0. \end{equation*} \notag $$
It is non-trivial since $[v_1v_2v_5v_7v_8u_3u_4u_6u_9]$ corresponds to a non-zero class in $H^4(\mathcal K)$ and $\dim\mathcal K=4$. It is indecomposable since
$$ \begin{equation*} \widetilde{H}^*(\mathcal K_{[m]\setminus\{1,2,3\}})\cong \widetilde{H}^*(\mathcal K_{[m]\setminus\{4,5,6\}})\cong \widetilde{H}^*(\mathcal K_{[m]\setminus\{7,8,9\}})=0 \end{equation*} \notag $$
and $\widetilde{H}^p(\mathcal K_{[m]\setminus\{1,4,7\}})\cong\Bbbk$ for $p=4$ and is zero otherwise, whereas $\widetilde{H}^q(\mathcal K_{\{1,4,7\}})\cong\Bbbk$ for $q=1$ and is zero otherwise. $\Box$

It follows directly from [5], Theorem 6.3, (5), (7), that that if $K$ is a minimally non-Golod simplicial complex such that $\operatorname{cup}(\mathcal Z_K)=1$ and there exists a non-trivial Massey product in $H^{*}(\mathcal Z_K)$, then $\dim(K)\geqslant\dim(\mathcal K)=4$ and $f_0(K)\geqslant f_0(\mathcal K)=9$.

We are grateful to Victor Buchstaber for fruitful discussions and his interest to this work.


Bibliography

1. A. Berglund and M. Jöllenbeck, J. Algebra, 315:1 (2007), 249–273  crossref  mathscinet  zmath
2. V. M. Buchstaber and T. E. Panov, Toric topology, Math. Surveys Monogr., 204, Amer. Math. Soc., Providence, RI, 2015, xiv+518 pp.  crossref  mathscinet  zmath
3. E. S. Golod, Dokl. Akad. Nauk SSSR, 144:3 (1962), 479–482  mathnet  mathscinet  zmath; English transl. in Soviet Math. Dokl., 3 (1962), 745–749
4. J. Grbić, T. Panov, S. Theriault, and J. Wu, Trans. Amer. Math. Soc., 368:9 (2016), 6663–6682  crossref  mathscinet  zmath
5. L. Katthän, J. Algebra, 479 (2017), 244–262  crossref  mathscinet  zmath
6. I. Yu. Limonchenko, Algebraic topology, combinatorics, and mathematical physics, Tr. Mat. Inst. Steklova, 305, Stekolv Mathematical Institute, Moscow, 2019, 174–196  mathnet  crossref  mathscinet  zmath; English transl. in Proc. Steklov Inst. Math., 305 (2019), 161–181  crossref

Citation: I. Yu. Limonchenko, T. E. Panov, “Monomial non-Golod face rings and Massey products”, Uspekhi Mat. Nauk, 77:4(466) (2022), 203–204; Russian Math. Surveys, 77:4 (2022), 762–765
Citation in format AMSBIB
\Bibitem{LimPan22}
\by I.~Yu.~Limonchenko, T.~E.~Panov
\paper Monomial non-Golod face rings and Massey products
\jour Uspekhi Mat. Nauk
\yr 2022
\vol 77
\issue 4(466)
\pages 203--204
\mathnet{http://mi.mathnet.ru/rm10065}
\crossref{https://doi.org/10.4213/rm10065}
\mathscinet{http://mathscinet.ams.org/mathscinet-getitem?mr=4461390}
\adsnasa{https://adsabs.harvard.edu/cgi-bin/bib_query?2022RuMaS..77..762L}
\transl
\jour Russian Math. Surveys
\yr 2022
\vol 77
\issue 4
\pages 762--765
\crossref{https://doi.org/10.4213/rm10065e}
\isi{https://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=Publons&SrcAuth=Publons_CEL&DestLinkType=FullRecord&DestApp=WOS_CPL&KeyUT=000992300700007}
\scopus{https://www.scopus.com/record/display.url?origin=inward&eid=2-s2.0-85165391171}
Linking options:
  • https://www.mathnet.ru/eng/rm10065
  • https://doi.org/10.4213/rm10065e
  • https://www.mathnet.ru/eng/rm/v77/i4/p203
  • This publication is cited in the following 1 articles:
    Citing articles in Google Scholar: Russian citations, English citations
    Related articles in Google Scholar: Russian articles, English articles
    Успехи математических наук Russian Mathematical Surveys
    Statistics & downloads:
    Abstract page:339
    Russian version PDF:24
    English version PDF:62
    Russian version HTML:144
    English version HTML:101
    References:67
    First page:14
     
      Contact us:
     Terms of Use  Registration to the website  Logotypes © Steklov Mathematical Institute RAS, 2024