On the genus of the annhilator graph of a commutative ring

Let R be a commutative ring and Z(R)* be its set of non-zero zero-divisors. The annihilator graph of a commutative ring R is the simple undirected graph AG(R) with vertices Z(R)*, and two distinct vertices x and y are adjacent if and only if ann(xy)≠ann(x)∪ann(y). The notion of annihilator graph has...

Full description

Saved in:
Bibliographic Details
Published in:Algebra and Discrete Mathematics
Date:2017
Main Authors: Chelvam, T.T., Selvakumar, K.
Format: Article
Language:English
Published: Інститут прикладної математики і механіки НАН України 2017
Online Access:https://nasplib.isofts.kiev.ua/handle/123456789/156637
Tags: Add Tag
No Tags, Be the first to tag this record!
Journal Title:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Cite this:On the genus of the annhilator graph of a commutative ring / T.T. Chelvam, K. Selvakumar // Algebra and Discrete Mathematics. — 2017. — Vol. 24, № 2. — С. 191-208. — Бібліогр.: 26 назв. — англ.

Institution

Digital Library of Periodicals of National Academy of Sciences of Ukraine
id nasplib_isofts_kiev_ua-123456789-156637
record_format dspace
spelling Chelvam, T.T.
Selvakumar, K.
2019-06-18T18:18:04Z
2019-06-18T18:18:04Z
2017
On the genus of the annhilator graph of a commutative ring / T.T. Chelvam, K. Selvakumar // Algebra and Discrete Mathematics. — 2017. — Vol. 24, № 2. — С. 191-208. — Бібліогр.: 26 назв. — англ.
1726-3255
2010 MSC:05C99, 05C15, 13A99.
https://nasplib.isofts.kiev.ua/handle/123456789/156637
Let R be a commutative ring and Z(R)* be its set of non-zero zero-divisors. The annihilator graph of a commutative ring R is the simple undirected graph AG(R) with vertices Z(R)*, and two distinct vertices x and y are adjacent if and only if ann(xy)≠ann(x)∪ann(y). The notion of annihilator graph has been introduced and studied by A. Badawi [7]. In this paper, we determine isomorphism classes of finite commutative rings with identity whose AG(R) has genus less or equal to one
This work was supported by the UGC-BSR One-time grant and UGC Major Research Project (F. No. 42-8/2013(SR)) of University Grants Commission, Government of India through first and second authors respectively.
en
Інститут прикладної математики і механіки НАН України
Algebra and Discrete Mathematics
On the genus of the annhilator graph of a commutative ring
Article
published earlier
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
collection DSpace DC
title On the genus of the annhilator graph of a commutative ring
spellingShingle On the genus of the annhilator graph of a commutative ring
Chelvam, T.T.
Selvakumar, K.
title_short On the genus of the annhilator graph of a commutative ring
title_full On the genus of the annhilator graph of a commutative ring
title_fullStr On the genus of the annhilator graph of a commutative ring
title_full_unstemmed On the genus of the annhilator graph of a commutative ring
title_sort on the genus of the annhilator graph of a commutative ring
author Chelvam, T.T.
Selvakumar, K.
author_facet Chelvam, T.T.
Selvakumar, K.
publishDate 2017
language English
container_title Algebra and Discrete Mathematics
publisher Інститут прикладної математики і механіки НАН України
format Article
description Let R be a commutative ring and Z(R)* be its set of non-zero zero-divisors. The annihilator graph of a commutative ring R is the simple undirected graph AG(R) with vertices Z(R)*, and two distinct vertices x and y are adjacent if and only if ann(xy)≠ann(x)∪ann(y). The notion of annihilator graph has been introduced and studied by A. Badawi [7]. In this paper, we determine isomorphism classes of finite commutative rings with identity whose AG(R) has genus less or equal to one
issn 1726-3255
url https://nasplib.isofts.kiev.ua/handle/123456789/156637
citation_txt On the genus of the annhilator graph of a commutative ring / T.T. Chelvam, K. Selvakumar // Algebra and Discrete Mathematics. — 2017. — Vol. 24, № 2. — С. 191-208. — Бібліогр.: 26 назв. — англ.
work_keys_str_mv AT chelvamtt onthegenusoftheannhilatorgraphofacommutativering
AT selvakumark onthegenusoftheannhilatorgraphofacommutativering
first_indexed 2025-11-24T04:43:57Z
last_indexed 2025-11-24T04:43:57Z
_version_ 1850841729613168640
fulltext “adm-n4” 22:47 page #13 Algebra and Discrete Mathematics RESEARCH ARTICLE Volume 24 (2017). Number 2, pp. 191–208 © Journal “Algebra and Discrete Mathematics” On the genus of the annihilator graph of a commutative ring∗ T. Tamizh Chelvam and K. Selvakumar Communicated by V. V. Kirichenko Abstract. Let R be a commutative ring and Z(R)∗ be its set of non-zero zero-divisors. The annihilator graph of a commu- tative ring R is the simple undirected graph AG(R) with vertices Z(R)∗, and two distinct vertices x and y are adjacent if and only if ann(xy) 6= ann(x) ∪ ann(y). The notion of annihilator graph has been introduced and studied by A. Badawi [7]. In this paper, we determine isomorphism classes of finite commutative rings with identity whose AG(R) has genus less or equal to one. 1. Introduction The study of algebraic structures, using the properties of graphs, became an exciting research topic in the past twenty years, leading to many fascinating results and questions. I. Beck[8] began the study of associating a graph called the zero-divisor graph Γ0(R) to a commutative ring R and was mainly interested in the coloring of the zero-divisor graph. For a commutative ring R, the zero-divisor graph is the simple graph with R as the vertex set and two distinct elements x and y are adjacent if and only if xy = 0 [8]. D. F. Anderson and P. S. Livingston[3] modified and ∗This work was supported by the UGC-BSR One-time grant and UGC Major Research Project (F. No. 42-8/2013(SR)) of University Grants Commission, Government of India through first and second authors respectively. 2010 MSC: 05C99, 05C15, 13A99. Key words and phrases: commutative ring, annihilator graph, genus, planar, local rings. “adm-n4” 22:47 page #14 192Genus of the annihilator graph of a commutative ring studied the zero-divisor graph Γ(R) as the graph with vertex set as the nonzero zero-divisors Z(R)∗ of R and two distinct elements x, y ∈ Z∗(R) are adjacent if and only if xy = 0. Thereafter, several graphs have been associated with commutative rings. These graphs exhibit the interplay between the algebraic properties of R and graph theoretical properties of the associated graph. For a ∈ R, let ann(a) = {d ∈ R : da = 0} be the annihilator of a ∈ R. In 2014, A. Badawi [7] introduced the annihilator graph AG(R) as the simple graph with vertex set Z(R)∗, and two distinct vertices x and y are adjacent if and only if ann(xy) 6= ann(x)∪ann(y). One can see that the zero-divisor graph Γ(R) is a subgraph of the annihilator graph AG(R). The main objective of topological graph theory is to embed a graph into a surface. Let Sk denote the sphere with k handles, where k is a nonnegative integer, that is, Sk is an oriented surface of genus k. The genus of a graph G, denoted g(G), is the minimal integer n such that the graph can be embedded in Sn. Intuitively, G is embedded in a surface if it can be drawn in the surface so that its edges intersect only at their common vertices. A graph G with genus 0 is called a planar graph, whereas a graph G with genus 1 is called a toroidal graph. Further note that if H is a subgraph of a graph G, then g(H) 6 g(G). For details on the notion of embedding a graph in a surface, see [26]. Many research articles have appeared on the genus of zero divisor graphs of commutative rings. In particular, there are many papers [1,2,9, 15,24], where the planarity of zero-divisor graphs has been discussed. The question addressed in these papers is this:For which finite commutative rings R is Γ(R) planar? A partial answer was given in [1], but the question remained open for local rings of order 32. In [15], and independently in both [9] and [24], it was shown that no local ring of order 32 has the planar zero divisor graph. Furthermore, Smith [15] gave a complete list of finite planar rings; this list included the 2 infinite families Z2 × F and Z3 × F , where F is any finite field, and the 42 other isomorphism classes of finite planar rings. H.J. Wang determined rings of the forms Zp α1 1 ×· · ·×Zp αr r and Zn[x] 〈xm〉 that have genus at most one [24, Theorems 3.5 and 3.11]. Further H.J. Wang and N.O. Smith obtained all commutative rings whose zero divisor graph has genus at most one [23, Theorem 3.6.2]. Note that the zero divisor graph Γ(R) is a subgraph of AG(R). In [7], it has been shown that for any reduced ring R that is not an integral domain, AG(R) = Γ(R) if and only if R has exactly two distinct minimal prime ideals [7, Theorem 3.6]. Note that using the proof of this result, “adm-n4” 22:47 page #15 T. Tamizh Chelvam, K. Selvakumar 193 x1 x2 x3 x4 x5x6x7 x8x9 x10 x11 Figure 1. Graph G. one can establish that for any reduced ring, AG(R) is complete if and only if, Γ(R) is complete if and only if, R ∼= Z2 × Z2. By a graph G = (V, E), we mean an undirected simple graph with vertex set V and edge set E. A graph in which each pair of distinct vertices is joined by the edge is called a complete graph. We use Kn to denote the complete graph with n vertices. An r-partite graph is one whose vertex set can be partitioned into r subsets so that no edge has both ends in any one subset. A complete r-partite graph is one in which each vertex is joined to every vertex that is not in the same subset. The complete bipartite graph (2-partite graph) with part sizes m and n is denoted by Km,n. If G = K1,n where n > 1, then G is a star graph. Pn denotes the path of length n for n > 1. A graph G is said to be unicyclic if it contains a unique cycle. Given a connected graph G, we say that a vertex v of G is a cut vertex if G − v is disconnected. For a subset S of vertices of G, the induced subgraph of G is the subgraph with vertex set S together with edges whose both ends are in S and is denoted by < S >. A block is a maximal connected subgraph of G having no cut vertices. A result of Battle, Harary, Kodama, and Youngs states that the genus of a graph is the sum of the genera of its blocks[6]. For example, the graph G in Figure 1 has two blocks, both isomorphic to K3,3, and so has genus 2 [25, C. Wickham]. Throughout this paper, we assume that R is a finite commutative ring with identity, Z(R) its set of zero-divisors and Nil(R) its set of nilpotent elements, R× its group of units, Fq denotes the field with q elements, and R∗ = R − {0}. The following results are useful for further reference in this paper. “adm-n4” 22:47 page #16 194Genus of the annihilator graph of a commutative ring Theorem 1. [23, Theorem 3.5.1] Let (R,m) be a finite local ring which is not a field. Then Γ(R) is planar if and only if R is isomorphic to one of the following 29 rings: Z4, Z2[x] 〈x2〉 , Z9, Z3[x] 〈x2〉 , Z8, Z2[x] 〈x3〉 , Z4[x] 〈x3, x2 − 2〉 , Z2[x, y] 〈x2, xy, y2〉 , Z4[x] 〈2x, x2〉 , F4[x] 〈x2〉 , Z4[x] 〈x2 + x + 1〉 , Z25, Z5[x] 〈x2〉 , Z16, Z2[x] 〈x4〉 , Z4[x] 〈x2 − 2, x4〉 , Z4[x] 〈x3 − 2, x4〉 , Z4[x] 〈x3 + x2 − 2, x4〉 , Z2[x, y] 〈x3, xy, y2 − x2〉 , Z4[x] 〈x3, x2 − 2x〉 , Z8[x] 〈x2 − 4, 2x〉 , Z4[x, y] 〈x3, x2 − 2, xy, y2 − 2, y3〉 , Z4[x] 〈x2〉 , Z4[x, y] 〈x2, y2, xy − 2〉 , Z2[x, y] 〈x2, y2〉 , Z27, Z3[x] 〈x3〉 , Z9[x] 〈x2 − 3, x3〉 , Z9[x] 〈x2 + 3, x3〉 . One can have the following theorem from Theorem 3.7 [15]. Theorem 2. [15, Theorem 3.7] Let R = F1 × · · · × Fn be a finite ring, where each Fi is a field and n > 2. Then Γ(R) is planar if and only if R is isomorphic to one of the following rings: Z2 × F, Z3 × F, Z2 × Z2 × Z2, Z2 × Z2 × Z3, where F is a finite field. Theorem 3. [23, Theorem 3.5.2] Let (R,m) be a finite local ring which is not a field. Then g(Γ(R)) = 1 if and only if R is isomorphic to one of the following 17 rings: Z49, Z7[x] 〈x2〉 , Z2[x, y] 〈x3, xy, y2〉 , Z4[x] 〈x3, 2x〉 , Z4[x, y] 〈x3, x2 − 2, xy, y2〉 , Z8[x] 〈x2, 2x〉 , F8[x] 〈x2〉 , Z4[x] 〈x3 + x + 1〉 , Z4[x, y] 〈2x, 2y, x2, xy, y2〉 , Z2[x, y, z] 〈x, y, z〉2 , Z32, Z2[x] 〈x5〉 , Z4[x] 〈x3 − 2, x5〉 , Z4[x] 〈x4 − 2, x5〉 , Z8[x] 〈x2 − 2, x5〉 , Z8[x] 〈x2 − 2x + 2, x5〉 , Z8[x] 〈x2 + 2x − 2, x5〉 . Theorem 4. [11, 12, Theorem 6.3] Let G be a connected graph. Then G is a split graph if and only if G contains no induced subgraph isomorphic to 2K2, C4, C5. “adm-n4” 22:47 page #17 T. Tamizh Chelvam, K. Selvakumar 195 Theorem 5. [3, Lemma 2.12] Let R be a finite commutative ring. If Γ(R) has exactly one vertex adjacent to every other vertex and no other adjacent vertices, then either R ∼= Z2 × F , where F is a finite field with |F | > 3, or R is local with maximal ideal m satisfying R m ∼= Z2, m3 = 0 and |m2| 6 2. Thus |Γ(R)| is either pn or 2n − 1 for some prime p and integer n > 1. Theorem 6. [3, Theorem 2.10] Let R be a finite commutative ring. If Γ(R) is complete, then either R ∼= Z2 × Z2 or R is local with char R = p or p2, and |Γ(R)| = pn − 1, where p is prime and n > 1. Theorem 7. [3, Theorem 2.13] Let R be a finite commutative ring with |Γ(R)| > 4. Then Γ(R) is a star graph if and only if R ∼= Z2 × F, where F is a finite field. In particular, if Γ(R) is a star graph, then Γ(R) = pn for some prime p and integer n > 0. Conversely, each star graph of order pn can be realized as Γ(R). Theorem 8. [7, Theorem 3.6] Let R be a reduced commutative ring that is not an integral domain. Then AG(R) = Γ(R) if and only if R has exactly two distinct minimal prime ideals. Theorem 9. [7, Theorem 3.10] Let R be a nonreduced commutative ring with | Nil(R)∗| > 2 and let AGN (R) be the (induced) subgraph of AG(R) with vertices Nil(R)∗. Then AGN (R) is complete. 2. Basic properties of annihilator graph In this section, we state some basic observations of the annihilator graph. Especially, we identify the annihilator ideal graph of small order and in particular we list out all local rings R with |R| 6 7, for which the annihilator graph AG(R) is complete. Remark 1. Let R = F1 × F2 where F1 and F2 are finite fields. Then R is reduced with exactly two distinct minimal prime ideals. Hence, by Theorem 8, AG(R) ∼= Γ(R) = K|F ∗ 1 |,|F ∗ 2 |. Remark 2. Let R 6= Z2 × Z2 be a local ring that is not a field and |Z(R)∗| > 3. By Theorem 9, AG(R) is complete and hence gr(AG(R)) = 3. On the other hand, let R be a finite commutative ring with identity but not a field. Since R is finite, R ∼= R1 × · · · × Rn, where each Ri is a local ring. If n > 3, then (1, 0, . . . , 0)− (0, 1, 0, . . . , 0)− (0, 0, 1, 0, . . . , 0)− (1, 0, . . . , 0) is a cycle in AG(R) and hence gr(AG(R) = 3. “adm-n4” 22:47 page #18 196Genus of the annihilator graph of a commutative ring |Z(R)∗| Local ring R AG(R) 1 Z4, Z2[x] 〈x2〉 K1 2 Z9, Z3[x] 〈x2〉 K2 3 Z8, Z2[x] 〈x3〉 , Z4[x] 〈2x,x2−2〉 K3 3 F4[x] 〈x2〉 , Z4[x] 〈x2+x+1〉 K3 3 Z4[x] <x,2>2 , Z2[x,y] <x,y>2 K3 4 Z25, Z5[x] 〈x2〉 K4 Table 1. Remark 3. Note that Γ(R) is a subgraph of AG(R). D.F. Anderson et al., [2] gave all zero-divisor graphs of order 6 4. Using this, we give in Table 1, all commutative local rings R for which |Z(R)∗| 6 4 and AG(R) is complete. One can note that there are only two rings Z2 × F4 and Z3 × Z3 with |Z(R)∗| 6 4 which are not mentioned in Table 1 since Γ(Z2 × F4) = K1,3 and Γ(Z3 × Z3) = K2,2 (refer Remark 1). S. P. Redmond [13,14] has given all local commutative rings R with |Z(R)∗| 6 7. Using the list given in [13, 14], Table 2 provides all finite commutative local rings, for which 6 6 |Z(R)∗| 6 7 and AG(R) is com- plete. |Z(R)∗| Local Ring R AG(R) 6 Z49, Z7[x] 〈x2〉 K6 7 Z16, Z2[x] 〈x4〉 , Z4[x] 〈x2+2〉 , Z4[x] 〈x2+2x+2〉 K7 7 Z4[x] 〈x3−2,2x2,2x〉 , Z2[x,y] 〈x3,xy,y2〉 , Z8[x] 〈2x,x2〉 K7 7 Z4[x] 〈x3,2x2,2x〉 , Z4[x,y] 〈x2−2,xy,y2,2x,2y〉 , Z4[x] 〈x2+2x〉 K7 7 Z8[x] 〈2x,x2+4〉 , Z2[x,y] 〈x2,y2−xy〉 , Z4[x,y] 〈x2,y2−xy,xy−2,2x,2y〉 K7 7 Z4[x,y] 〈x2,y2,xy−2,2x,2y〉 , Z2[x,y] 〈x2,y2〉 , Z4[x] 〈x2〉 , K7 7 Z2[x,y,z] 〈x,y,z〉2 , Z4[x,y] 〈x2,y2,xy,2x,2y〉 , F8[x] 〈x2〉 , Z4[x] 〈x3+x+1〉 K7 Table 2. “adm-n4” 22:47 page #19 T. Tamizh Chelvam, K. Selvakumar 197 Theorem 10. Let R be a finite commutative ring with identity that is not a field. Then AG(R) is a tree if and only if R is isomorphic to one of the following 5 rings: Z4, Z2[x] 〈x2〉 , Z9, Z3[x] 〈x2〉 or Z2 × F, where F is a finite field. Proof. Since R is finite, R ∼= R1 × · · · × Rn, where each Ri is a local ring. Suppose AG(R) is a tree. In view Remark 2, n 6 2. Suppose R ∼= R1 × R2. If Z(R1)∗ 6= {0}, then there exist x1, y1 ∈ Z(R1)∗ such that x1y1 = 0 and |R× 1 | > 2. Let x = (0, 1), y = (x1, 0) z = (y1, 1) and w = (1, 0). Then x, y, z, w ∈ Z(R)∗ and (x1, 1) ∈ ann(zw) where as (x1, 1) is neither in ann(z) nor in ann(w). Hence x−y−z −w−x is a cycle in AG(R), a contradiction. Hence R1 and R2 are fields and so AG(R) ∼= K|R1|−1,|R2|−1. Since AG(R) is tree, |R1| = 2 or |R2| = 2 and so R ∼= Z2 × F , where F is a finite field. Suppose R ∼= R1. Since R is not a field, Z(R)∗ 6= {0}. Here R = R1 is a local ring and so AG(R) is complete. Hence by the assumption viz., AG(R) is a tree, we get that |Z(R)∗| 6 2. Hence R ∼= Z4, Z2[x] 〈x2〉 , Z9, or Z3[x] 〈x2〉 . The converse can be ascertained from Table 1 and Remark 1. Theorem 11. Let R be a finite commutative ring with identity that is not a field. Then AG(R) is unicyclic if and only if R is isomorphic to one of the following 8 rings: Z8, Z2[x] 〈x3〉 , Z4[x] 〈2x, x2 − 2〉 , F4[x] 〈x2〉 , Z4[x] 〈x2 + x + 1〉 , Z4[x] 〈x, 2〉2 , Z2[x, y] 〈x, y〉2 or Z3 × Z3. Proof. Sufficient part follows from Table 1 and Remark 1. Conversely, assume that AG(R) contains a unique cycle of length ℓ > 3. Since R is finite, R ∼= R1 × · · · × Rn, where each Ri is a lo- cal ring. Suppose n > 3. Let x1 = (1, 0, 0, . . . , 0), x2 = (0, 1, 0, . . . , 0), x3 = (0, 0, 1, 0, . . . , 0), y1 = (0, 1, 1, 0 . . . , 0), y2 = (1, 0, 1, 0 . . . , 0). Then x1, x2, x3, y1, y2 ∈ Z(R)∗ and x1−x2−x3−x1 as well as x1−y1−y2−x2−x1 are two distinct cycles in AG(R), a contradiction. Hence n 6 2. Case 1. Suppose n = 2. If Z(R1) 6= {0}, then there exist x, y ∈ Z(R1)∗ such that xy = 0. Since R1 is local, R1 contains at least one unit u1 ∈ R× 1 “adm-n4” 22:47 page #20 198Genus of the annihilator graph of a commutative ring apart from the identity. Then (x, 0) − (y, 1) − (1, 0) − (0, 1) − (x, 0) and (x, 0)− (y, 1)− (u1, 0)− (0, 1)− (x, 0) are cycles in AG(R), a contradiction. Hence R1 and R2 are fields and so AG(R) ∼= K|R1|−1,|R2|−1. Since AG(R) is unicyclic, R1 ∼= Z3 and R2 ∼= Z3. Case 2. Suppose n = 1. Here R is a local ring but not a field. By Theorem 9, AG(R) is complete. Since AG(R) is unicyclic, |Z(R)∗| = 3 and by Table 1, R is isomorphic to one of the following rings: Z8, Z2[x] 〈x3〉 , Z4[x] 〈2x,x2−2〉 , F4[x] 〈x2〉 , Z4[x] 〈x2+x+1〉 , Z4[x] 〈x,2〉2 , Z2[x,y] 〈x,y〉2 . Note that any complete graph is a split graph with any single vertex as an independent set and all the other vertices induce a clique. Hence, if R is a local ring, then AG(R) is a split graph. Now, we characterize all nonlocal rings R for which AG(R) is a split graph. Theorem 12. Let R be a finite commutative nonlocal ring with identity and |Z(R)∗| > 2. Then AG(R) is a split graph if and only if R ∼= Z2 × F , where F is a finite field. Proof. Suppose R = Z2 × F , where F is a finite field. By Remark 1, AG(R) = K1,|F ∗| and hence AG(R) is a split graph. Conversely, suppose AG(R) is a split graph. Since R is finite, R ∼= R1 × · · · × Rn where each Ri is local for 1 6 i 6 n and n > 2. If n > 3, then (1, 0, 0, . . . , 0) − (0, 1, 0, . . . , 0) − (1, 0, 1, 0 . . . , 0) − (0, 1, 1, 0, . . . , 0) − (1, 0, 0, . . . , 0) is a cycle of length 4 in AG(R) and by Theorem 4, AG(R) is not split, a contradiction. Hence n = 2. If Z(R1) 6= {0}, then there exists an element x ∈ Z(R1)∗ such that xy = 0 for some y ∈ Z(R1)∗ and so (x, 0) − (0, 1) − (1, 0) − (y, 1) − (x, 0) is a cycle of length 4 in AG(R), a contradiction. Hence R1, R2 are fields and so AG(R) ∼= K|R1|−1,|R2|−1. Since AG(R) is split, either |R1| − 1 = 1 or |R2| − 1 = 1 and so R ∼= Z2 × F where F is a finite field. Theorem 13. Let R be a finite commutative ring with identity that is not a field. Then (i) gr(AG(R)) = ∞ if and only if R ∼= Z4, Z2[x] 〈x2〉 , Z9, Z3[x] 〈x2〉 , or Z2 × F , where F is a finite field; (ii) gr(AG(R)) = 4 if and only if R ∼= Z4 × Z2, Z2[x] 〈x2〉 × Z2, or F1 × F2, where F1, F2 are finite fields with |F1| > 3 and |F2| > 3; (iii) gr(AG(R)) = 3 if and only if R is not isomorphic to the rings in (i) and (ii). Proof. (i) Suppose gr(AG(R)) = ∞. Then AG(R) contains no cycles and so AG(R) is a tree. Remaining part of the proof follows from Theorem 10. “adm-n4” 22:47 page #21 T. Tamizh Chelvam, K. Selvakumar 199 (ii) Suppose gr(AG(R)) = 4. By Remark 2, R cannot be local. Also |Z(R)∗| > 4. Let R ∼= R1 × · · · × Rn where each Ri is a local ring. If n > 3, by Remark 2, gr(AG(R)) = 3, a contradiction and hence n = 2. Suppose R is not reduced. Then |Z(Ri) ∗| 6= {0} for some i. If |Z(R1)∗| > 2, then there exist x, y ∈ Z(R1)∗ such that x 6= y and xy = 0. From this, we get that (x, 0) − (y, 0) − (0, 1) − (x, 0) is a cycle of length 3 in AG(R), a contradiction. Thus |Z(Ri) ∗| 6 1 for i = 1, 2. If |Z(Ri) ∗| = 1 for i = 1, 2, then Ri ∼= Z4 or Z2[x] 〈x2〉 . Suppose R ∼= Z4 × Z4, then (2, 0)− (0, 1)− (2, 2) is a cycle in AG(R). Note that all the remaining cases produce the same AG(R). Hence in all the cases gr(AG(R)) = 3, a contradiction. This shows that |Z(R1)∗| = 1 or |Z(R2)∗| = 1. If |Z(R1)∗| = 1 and |Z(R2)∗| = 0, then R1 ∼= Z4 or Z2[x] 〈x2〉 and R2 is a field. If |R2| > 3, then (2, 0) − (2, 1) − (2, x) − (2, 0) for some 1 6= x ∈ R∗ 2 is a cycle of length three in AG(R), a contradiction. Hence |R2| = 2 and so R2 ∼= Z2 and so R ∼= Z4 × Z2 or Z2[x] 〈x2〉 × Z2. Suppose R is reduced. Then R1 and R2 are fields and so AG(R) ∼= K|R1|−1,|R2|−1. Since gr(AG(R)) = 4, |R1| > 3 and |R2| > 3. Converse part of (ii) is trivial. Part (iii) now follows directly from the above two cases. Corollary 1. Let R be a finite commutative ring with identity that is not a field. Then AG(R) is a complete bipartite graph if and only if R is isomorphic to one of the following rings: Z9, Z3[x] 〈x2〉 , Z4 × Z2, Z2[x] 〈x2〉 × Z2 or F1 × F2, where F1 and F2 are finite fields. Proof. We only need to prove the necessary part. Suppose AG(R) is a complete bipartite graph. Then AG(R) does not contains a cycle of odd length and so girth of AG(R) cannot be odd. By Theorem 13, gr(AG(R) = 4 or gr(AG(R)) = ∞. By the assumption that AG(R) is complete bipartite, AG(R) ≇ K1. Hence R is isomorphic to one of the following rings: Z9, Z3[x] 〈x2〉 , Z4 × Z2, Z2[x] 〈x2〉 × Z2, or F1 × F2, where F1 and F2 are fields. 3. Planar annihilator graphs In this section, we characterize finite commutative rings R for which AG(R) is planar. The following are known regarding the genus. “adm-n4” 22:47 page #22 200Genus of the annihilator graph of a commutative ring Lemma 1 ([26]). A connected graph G is planar if and only if G contains no subdivision of either K5 or K3,3. Lemma 2 ([26]). Let n be a positive integer and for real number x, ⌈x⌉ is the least integer that is the greater than or equal to x. Then g(Kn) = ⌈ (n−3)(n−4) 12 ⌉ if n > 3. In particular, g(Kn) = 1 if n = 5, 6, 7. Lemma 3 ([26]). Let m, n be positive integers and for real number x, ⌈x⌉ is the least integer that is the greater than or equal to x. Then g(Km,n) = ⌈ (m−2)(n−2) 4 ⌉ if m, n > 2. In particular, g(K4,4) = g(K3,n) = 1 if n = 3, 4, 5, 6 and g(K5,4) = g(K6,4) = g(Km,4) = 2 if m = 7, 8, 9, 10. Lemma 4 ([26, Euler formula]). If G is a finite connected graph with n vertices, m edges, and genus g, then n − m + f = 2 − 2g, where f is the number of faces created when G is minimally embedded on a surface of genus g. Theorem 14. Let (R,m) be a finite commutative local ring with identity. Then AG(R) is planar if and only if R is isomorphic to one of the following 13 rings: Z4, Z2[x] 〈x2〉 , Z9, Z3[x] 〈x2〉 , Z8, Z2[x] 〈x3〉 , Z4[x] 〈2x, x2 − 2〉 , F4[x] 〈x2〉 , Z4[x] 〈x2 + x + 1〉 , Z4[x] 〈x, 2〉2 , Z2[x, y] 〈x, y〉2 , Z25 or Z5[x] 〈x2〉 . Proof. By Lemma 1, AG(R) is planar if and only if AG(R) contains no subdivision of either K5 or K3,3. Hence |Z(R)∗| 6 4. Now proof follows from Table 1. Theorem 15. Let R = R1 × · · · × Rn be a finite commutative nonlocal ring, where each Ri is a local ring and n > 2. Then AG(R) is planar if and only if R is isomorphic either of the following rings: Z2 × F, Z3 × F, Z4 × Z2, Z2[x] 〈x2〉 × Z2 or Z2 × Z2 × Z2, where F is a finite field. Proof. Note that AG(Z2 ×F ) = K1,|F ∗| and AG(Z3 ×F ) = K2,|F ∗| and so by Lemma 3, they are planar. Since AG(Z4×Z2) = AG(Z2[x] 〈x2〉 × Z2) ∼= K2,3, they are also planar. As per the embedding given in Figure 2, AG(Z2 × Z2 × Z2) is planar. “adm-n4” 22:47 page #23 T. Tamizh Chelvam, K. Selvakumar 201 (0, 1, 1) (1, 0, 0) (0, 0, 1) (0, 1, 0) (1, 1, 0) (1, 0, 1) Figure 2. AG(Z2 × Z2 × Z2). Conversely assume that AG(R) is planar. Suppose n > 4. Let x1 = (1, 0, 0, 0, . . . , 0), x2 = (0, 1, 0, 0, . . . , 0), x3 = (1, 1, 0, 0, . . . , 0), y1 = (0, 0, 1, 0, . . . , 0), y2 = (0, 0, 0, 1, 0, . . . , 0), y3 = (0, 0, 1, 1, 0, . . . , 0). Then Ω = {x1, x2, x3, y1, y2, y3} ⊆ Z(R)∗ and the subgraph of AG(R) induced by Ω contains K3,3 as a subgraph, a contradiction. Hence n 6 3. Case 1. n = 2. Suppose mi 6= {0} for all i = 1, 2. Then |Ri| > 4 and |R× i | > 2. Let ai, bi be two distinct elements in R∗ i other than identity for i = 1, 2. Let d1 = (1, 0), d2 = (a1, 0), d3 = (b1, 0), g1 = (0, 1), g2 = (0, a2), g3 = (0, b2) ∈ Z(R)∗. Then Ω1 = {d1, d2, d3, g1, g2, g3} ⊆ Z(R)∗ and the subgraph of AG(R) induced by Ω1 contains K3,3 as a subgraph, a contradiction. Hence mi = {0} for some i. Without loss of generality, we assume that m2 = {0}. Then R2 is a field. Suppose m1 6= {0}. We claim that |m∗ 1| = 1. If not, |m∗ 1| > 2 and |R× 1 | > 3. Note that |R2| > 2. For a, b ∈ m ∗ 1 with ab = 0, and two distinct units u1, u2 ∈ R× 1 , let z1 = (a, 1), z2 = (b, 1), z3 = (0, 1), w1 = (1, 0), w2 = (u1, 0), w3 = (u2, 0) ∈ Z(R)∗. Then Ω2 = {z1, z2, z3, w1, w2, w3} ⊆ Z(R)∗ with z3wi = 0 in R for i = 1, 2, 3. Clearly z2 ∈ ann(z1w1), z2 /∈ ann(z1) ∪ ann(w1), z2 ∈ ann(z1w2), z2 /∈ ann(z1) ∪ ann(w2) and so z1 is adjacent to both w1 and w2 in AG(R). Further z1 ∈ ann(z2w1), z1 /∈ ann(z2) ∪ ann(w1), z1 ∈ ann(z2w2), z1 /∈ ann(z2) ∪ ann(w2) and so z2 is adjacent to both w1 and w2 in AG(R). From this, we observe that K3,3 is a subgraph of AG(R), a contradiction. Hence |m∗ 1| = 1 and so R1 ∼= Z4 or Z2[x] 〈x2〉 . Suppose |R2| > 3. For d ∈ m ∗ 1 with d2 = 0 and, 1 6= v1 ∈ R× 1 and 1 6= v2 ∈ R∗ 2, let s1 = (d, 1), s2 = (d, v2), s3 = (0, 1), p1 = (d, 0), p2 = “adm-n4” 22:47 page #24 202Genus of the annihilator graph of a commutative ring (v1, 0), p3 = (1, 0). Then Ω3 = {s1, s2, s3, p1, p2, p3} ⊆ Z(R)∗ and, p1s1 = p1s2 = 0 in R and s3pi = 0 in R for i = 1, 2, 3. Clearly s2 ∈ ann(s1p2), s2 /∈ ann(s1) ∪ ann(p2), s2 ∈ ann(s1p3), s2 /∈ ann(s1) ∪ ann(p3) and so s1 is adjacent to both p2 and p3 in AG(R). Also s1 ∈ ann(s2p2), s1 /∈ ann(s2) ∪ ann(p2), s1 ∈ ann(s2p3), s1 /∈ ann(s2) ∪ ann(p3) and so s2 is adjacent to both p2 and p3 in AG(R). From this, we get that K3,3 is a subgraph of AG(R), a contradiction. Hence |R2| = 2 and so R2 ∼= Z2. Suppose m1 = {0}. Then R1 is a field and by Remark 1, AG(R) ∼= K|R1|−1,|R2|−1. By Lemma 3, R ∼= Z2 × F or Z3 × F , where F is a finite field. (1, 0, 0) (0, 1, 0) (1, 0, 1) (1, 1, 0) (0, 0, 1) (0, 0, 2) (0, 1, 1) (1, 0, 2) (0, 1, 2) Figure 3. AG(Z2 × Z2 × Z3). Case 2. n = 3. Suppose |Ri| > 4 for some i. Without loss of generality, we assume that |R1| > 4. Let u1, u2, u3 be three distinct nonzero elements in R1. Let h1 = (u1, 0, 0), h2 = (u2, 0, 0), h3 = (u3, 0, 0), t1 = (0, 1, 0), t2 = (0, 0, 1), t3 = (0, 1, 1) ∈ Z(R)∗. Then hitj = 0 for all i, j = 1, 2, 3 and so K3,3 is a subgraph of AG(R), a contradiction. Hence |Ri| 6 3 is field for i = 1, 2, 3 and so Ri ∼= Z2 or Z3 for i = 1, 2, 3. Since Γ(R) is a subgraph of AG(R) and AG(R) is planar, by Theorem 2, the possibilities for R are Z2 × Z2 × Z2 or Z2 × Z2 × Z3. Note that the edges in dark lines of Figure 3 constitute a subdivision of K3,3 and so by Lemma 1, AG(Z2 × Z2 × Z3) is not planar. Thus R ∼= Z2 × Z2 × Z2. 4. Genus of AG(R) In this section, we characterize isomorphism classes of finite commuta- tive rings R whose AG(R) has genus one. First, let us characterize finite commutative local rings R for which genus of AG(R) is one. Theorem 16. Let (R,m) be a finite commutative local ring with identity that is not a field. Then g(AG(R)) = 1 if and only if R is isomorphic to “adm-n4” 22:47 page #25 T. Tamizh Chelvam, K. Selvakumar 203 one of the following 22 rings: Z49, Z7[x] 〈x2〉 , Z16, Z2[x] 〈x4〉 , Z4[x] 〈x4, x2 − 2〉 , Z2[x] 〈x3 − 2, x4〉 , Z4[x] 〈x4, x3 + x2 − 2〉 , Z2[x] 〈x3, x2 − 2x〉 , Z2[x, y] 〈x3, xy, y2 − x2〉 , Z8[x] 〈x2 − 4, 2x〉 , Z4[x, y] 〈x3, xy, x2 − 2, y2 − 2, y3〉 , Z4[x] 〈x2〉 , Z4[x, y] 〈x2, y2, xy − 2〉 , Z2[x, y] 〈x2, y2〉 , Z2[x, y] 〈x2, y2, xy〉 , Z4[x] 〈x3, 2x〉 , Z4[x, y] 〈x3, x2 − 2, xy, y2〉 , Z8[x] 〈x2, 2x〉 , F8[x] 〈x2〉 , Z4[x] 〈x3 + x + 1〉 , Z4[x, y] 〈2x, 2y, x2, y2, xy〉 , Z2[x, y, z] 〈x, y, z〉2 . Proof. By Theorem 9, AG(R) is complete. By Lemma 2, 5 6 |Z(R)∗| 6 7. Note that there is no local ring R with |Z(R)∗| = 5. Now the proof follows from Table 2. Remark 4. Note that if R ∼= R1 × · · · × Rn is a commutative ring with identity, where each (Ri,mi) is a local ring with mi 6= {0} and n > 2, then K5,6 is a subgraph of AG(R) and hence g(AG(R)) > 2. Thus if g(AG(R)) = 1, then one of the components Ri must be a field. Theorem 17. Let R = R1 × · · · × Rn be a finite commutative nonlocal ring, where each Ri is a local ring and n > 2. Then g(AG(R)) = 1 if and only if R is isomorphic to one of the following 7 rings: F4 × F4, F4 × Z5, Z5 × Z5, F4 × Z7, Z4 × Z3, Z2[x] 〈x2〉 × Z3 or Z2 × Z2 × Z3. Proof. Assume that g(AG(R)) = 1. Suppose n > 4. Let x1 = (1, 1, 0, 0, . . . , 0), x2 = (0, 1, 0, 1, . . . , 1), x3 = (0, 0, 1, 0, . . . , 0), x4 = (0, 1, 1, 0, . . . , 0), x5 = (1, 0, 1, . . . , 1), x6 = (1, 1, 0, 1, . . . , 1), x7 = (1, 1, 1, 0, . . . , 0), x8 = (1, 0, 1, 0, . . . , 0), x9 = (0, 0, 0, 1, . . . , 1), x10 = (1, 0, 0, 1, . . . , 1), x11 = (0, 1, . . . , 1). Then Ω = {x1, . . . , x11} ⊆ Z(R)∗ with x1x3 = x2x3 = x3x6 = x1x9 = x7x9 = x8x9 = 0. Clearly x5 ∈ ann(x1x4), x5 /∈ ann(x1) ∪ ann(x4), x11 ∈ ann(x1x5), x11 /∈ ann(x1)∪ann(x5), x5 ∈ ann(x2x4), x5 /∈ ann(x2)∪ “adm-n4” 22:47 page #26 204Genus of the annihilator graph of a commutative ring ann(x4), x7 ∈ ann(x2x5), x7 /∈ ann(x2) ∪ ann(x5), x4 ∈ ann(x5x6), x4 /∈ ann(x5) ∪ ann(x6), x5 ∈ ann(x6x4), x5 /∈ ann(x6) ∪ ann(x4), x11 ∈ ann(x1x10), x11 /∈ ann(x1) ∪ ann(x10), x5 ∈ ann(x1x11), x5 /∈ ann(x1) ∪ ann(x11), x11 ∈ ann(x8x10), x11 /∈ ann(x8) ∪ ann(x10), x6 ∈ ann(x11x8), x6 /∈ ann(x11) ∪ ann(x8), x10 ∈ ann(x11x7), x10 /∈ ann(x11) ∪ ann(x7), x11 ∈ ann(x10x7) and x11 /∈ ann(x10) ∪ ann(x7). From all the above ob- servations, Note that the subgraph induced by Ω in AG(R) contains G given in Figure 1 as a subgraph and so g(AG(R)) > 2. Hence n 6 3. Case 1. n = 3. Suppose R1 and R2 are not fields. Then as mentioned in Remark 4, K5,6 is a subgraph of AG(R) and so g(AG(R)) > 2, a contradiction. Hence at least two of the components Ri, (1 6 i 6 3) must be fields. Without loss of generality, let us assume that R2 and R3 are fields. Then |R2| > 2 and |R3| > 2. Suppose R1 is not a field. Then |m1| > 2 and |R× 1 | > 2. For z ∈ m ∗ 1 with ann(z) = m1 and u ∈ R× 1 and 1 6= u, let x1 = (z, 1, 0), x2 = (0, 1, 0), x3 = (z, 0, 0), x4 = (1, 0, 0), x5 = (u, 0, 0), x6 = (0, 0, 1, ), x7 = (0, 1, 1), x8 = (z, 1, 1), x9 = (1, 0, 1), x10 = (z, 0, 1), x11 = (u, 0, 1) and Ω1 = {x1, . . . , x11}. Without much difficulty, one can check that the subgraph induced by Ω1 in AG(R) contains G given in Figure 1 and so g(AG(R)) > 2, a contradiction. Hence R1 must be a field. Since AG(R) is non-planar, by Theorem 15, R ≇ Z2 × Z2 × Z2. Suppose at least two of the components Ri(i 6 i 6 3) contain at least three elements. Without loss of generality, assume that |R2| > 3 and |R3| > 3. For 1 6= a ∈ R∗ 3 and 1 6= b ∈ R∗ 2, let x1 = (1, 1, 0), x2 = (0, 1, 0), x3 = (0, 0, 1), x4 = (0, 0, a), x5 = (1, 0, 1), x6 = (0, 1, 0), x7 = (1, b, 0), x8 = (1, 0, 0), x9 = (0, 1, a), x10 = (0, 1, 1), x11 = (0, b, 1) and Ω2 = {x1, . . . , x11}. Then the subgraph of AG(R) induced by Ω2 contains G in Figure 1 and so g(AG(R)) > 2, a contradiction. Hence two of the components Ri must be Z2. Without loss of generality, we assume that R1 = R2 = Z2 and |R3| > 3. Suppose |R3| > 4. Let u and v be two distinct non-zero elements in R∗ 3 other than identity. Let x1 = (1, 0, 0), x2 = (1, 0, u), x3 = (0, 1, 1), x4 = (0, 1, v), x5 = (0, 1, u), x6 = (1, 0, v), x7 = (1, 1, 0), x8 = (0, 1, 0), x9 = (0, 0, 1), x10 = (0, 0, u), x11 = (1, 0, v) and Ω3 = {x1, . . . , x11}. Then the subgraph of AG(R) induced by Ω3 contains G given in Figure 1 and so g(AG(R)) > 2, a contradiction. Hence R3 = Z3 and so R ∼= Z2 ×Z2 ×Z3. Case 2. n = 2. If R1 and R2 are not fields, then as mentioned in Remark 4, K5,6 is a subgraph of AG(R) and so g(AG(R)) > 2, a contradiction. Hence one of the components Ri must be a field. Without loss generality, we assume that R2 is a field and so |R2| > 2. By Theorem 15, R ≇ Z4 × Z2 and Z3[x] 〈x2〉 × Z2. “adm-n4” 22:47 page #27 T. Tamizh Chelvam, K. Selvakumar 205 We claim that, if |m∗ 1| > 2, then g(AG(R)) > 2. Suppose |m∗ 1| = 2. By the facts given in Table 1, R1 ∼= Z9 or Z3[x] 〈x3〉 and hence |R× 1 | = 6. Let m ∗ 1 = {a, b}, R× 1 = {u1, . . . , u6} and y1 = (a, 1), y2 = (b, 1), y3 = (0, 1), x1 = (a, 0), x2 = (b, 0), x3 = (u1, 0), x4 = (u2, 0), x5 = (u3, 0), x6 = (u4, 0), x7 = (u5, 0), x8 = (u6, 0). Then Ω4 = {y1, y2, y3, x1, . . . , x8} ⊆ Z(R)∗ and the subgraph of AG(R) induced by Ω4 contains K3,8. By Lemma 3, g(AG(R)) > 2, a contradiction. Suppose |m∗ 1| > 3. Then |R× 1 | > 4. Let d, e, f ∈ m ∗ 1 such that de = df = 0 and {v1, . . . , v4} ⊆ R× 1 . Consider Ω5 = {z1, . . . , z7, w1, . . . , w4} ⊆ Z(R)∗, where z1 = (d, 0), z2 = (e, 0), z3 = (f, 0), z4 = (v1, 0), z5 = (v2, 0), z6 = (v3, 0), z7 = (v4, 0), w1 = (d, 1), w2 = (e, 1), w3 = (f, 1), w4 = (0, 1). Then the subgraph induced by Ω5 of AG(R) contains K4,7 and by Lemma 3, g(AG(R)) > 2, a contradiction. Hence we conclude that |m1| 6 2. From this R1 ∼= Z4 or Z2[x] 〈x2〉 when |m1| = 2 and R1 must be a field when |m1| = 1. By Theorem 15, R2 ≇ Z2 and so |R2| > 3. Suppose |m1| = 2 and |R2| > 5. For a ∈ m ∗ 1 with a2 = 0 and u1, u2 ∈ R× 1 , ej ∈ R∗ 2, let x1 = (a, 0), x2 = (u1, 0), x3 = (u2, 0), x4 = (a, e1), x5 = (a, e2), x6 = (a, e3), x7 = (a, e4), x8 = (0, e1), x9 = (0, e2), x10 = (0, e3), x11 = (0, e4) and Ω6 = {x1, . . . , x11} ⊂ Z(R)∗. Then the subgraph induced by Ω6 of AG(R) contains K3,8 and by Lemma 3, g(AG(R)) > 2, a contradiction. Hence R2 is isomorphic to either Z3 or F4. (2, 1) (2, ω) (2, ω2) (1, 0) (2, 0) (3, 0) (0, 1) (0, ω) (0, ω2) Figure 4. AG(Z4 × F4) ∼= AG ( Z2[x] 〈x 2〉 × F4 ) . Consider the case that R2 ∼= F4 = {0, 1, ω, ω2}. One can observe from Figure 4 that K3,6 is a subgraph of AG(R). By Lemma 3, g(K3,6) = 1 and hence one can fix an embedding of K3,6 on the surface of torus. By Euler’s formula, there are 9 faces in the embedding of K3,6, say {S1, . . . , S9}. Let ni be the length of the face Si. Note that 9 ∑ i=1 ni = 36 and ni > 4 for “adm-n4” 22:47 page #28 206Genus of the annihilator graph of a commutative ring every i. Thus ni = 4 for every i. Let U = {(2, 1), (2, ω), (2, ω2)} ⊂ V (K3,6). Further, the subgraph G of AG(R) induced by the vertices in U is K3, E(G)∩E(K3,6) = ∅. Since K3 cannot be embedded in the torus along with an embedding with only rectangles as faces, one cannot have an embedding of G and K3,6 together in the torus. This implies that g(AG(R)) > 2. Hence R2 ≇ F4 and so R is isomorphic to either Z4 × Z3 or Z2[x] 〈x2〉 × Z3. Suppose |m1| = 1 and in this case both R1 and R2 are fields. Note that AG(R) ∼= K|R1|−1,|R2|−1. Since g(AG(R)) = 1 and by Lemma 3, R is isomorphic to one of the following rings F4 × F4, F4 × Z5, Z5 × Z5 or F4 × Z7. Converse follows from Remark 1, Lemma 3, Figure 5 and Figure 6. (1, 0) (0, 2) (3, 0) (2, 2) (1, 0) (2, 0) (2, 1) c c d d e e Figure 5. Embedding of AG(Z4 × Z3) ∼= AG ( Z2[x] 〈x 2〉 × Z3 ) in S1. (1, , 1, 0) (1, 0, 1) (1, 0, 0) (0, 0, 1) (0, 1, 1) (0, 1, 0) (0, 0, 2) (1, 0, 2) (1, 0, 2) (0, 1, 2) (0, 1, 2) a a c c b b Figure 6. Embedding of AG(Z2 × Z2 × Z3) in S1. “adm-n4” 22:47 page #29 T. Tamizh Chelvam, K. Selvakumar 207 Acknowledgments The authors thank the anonymous referees for their comments which improved the presentation of the paper in many places. References [1] S. Akbari, H.R. Maimani and S. Yassemi, When a zero-divisor graph is planar or a complete r-partite graph, J. Algebra 270 (2003), 169–180. [2] D. F. Anderson, A. Frazier, A. Lauve, and P. S. Livingston, The zero-divisor graph of a commutative ring II, Lecture Notes in Pure and Appl. Math. 202 (2001), Marcel Dekker, New York, 61–72. [3] D. F. Anderson and P. S. Livingston, The zero-divisor graph of a commutative ring, J. Algebra 217 (1999), 434-447. [4] T. Asir and T. Tamizh Chelvam, On the genus of generalized unit and unitary Cayley graphs of a commutative ring, Acta Math. Hungar. 142 (2014), no. 2, 444–458, DOI: 10.1007/s10474-013-0365-1. [5] M. F. Atiyah and I. G. Macdonald, Introduction to Commutative Algebra, Addison- Wesley Publishing Company, 1969. [6] J. Battle, F. Harary, Y. Kodama, J. W. T. Youngs, Additivity of the genus of a graph, Bull. Amer. Math. Soc. 68 (1962), 565–568. [7] A. Badawi, On the annihilator graph of a commutative ring, Comm. Algebra 42 (2014), no. 1, 108–121. [8] I. Beck, Coloring of commutative rings, J. Algebra 116 (1988), 208–226. [9] R. Belshoff, J. Chapman, Planar zero-divisor graphs, J. Algebra 316 (2007), no. 1, 471–480. [10] G. Chartrand and L. Lesniak, Graphs and Digraphs, Wadsworth and Brooks/ Cole, Monterey, CA, 1986. [11] S. Földers and P.L. Hammer, Split Graphs, Proc. 8th Southeastern Conf. on Combinatorics, Graph Theory and Computing (F. Koffman et al. eds.), Louisiana State Univ., Baton Rouge, Louisianna, (1977), 311–315. [12] M.C. Golumbic, Algorithmic Graph Theory and Perfect Graphs, Second Edition, Elsevier B.V., Amsterdam, 2004. [13] S. P. Redmond, On zero-divisor graphs of small finite commutative rings, Discrete Math. 307 (2007), 1155–1166. [14] S. P. Redmond, Corrigendum to “On zero-divisor graphs of small finite commuta- tive rings”, Discrete Math. 307 (2007), 2449–2452. [15] N. O. Smith, Planar zero-divisor graphs, Int. J. Comm. Rings 2 (2002), no. 4, 177–188. [16] T. Tamizh Chelvam and T. Asir, On the genus of the total graph of a commutative ring, Comm. Algebra 41 (2013), 142–153. [17] T. Tamizh Chelvam and S. Nithya, Crosscap of the ideal based zero-divisor graph, Arab J. Math. Sci. 2 (2015), http://dx.doi.org/10.1016/j.ajmsc. 2015.01.003. “adm-n4” 22:47 page #30 208Genus of the annihilator graph of a commutative ring [18] T. Tamizh Chelvam and K. Selvakumar, Central sets in annihilating-ideal graph of a commutative ring, J. Combin. Math. Combin. Comput. 88 (2014), 277–288. [19] T. Tamizh Chelvam and K. Selvakumar, Domination in the directed zero-divisor graph of ring of matrices, J. Combin. Math. Combin. Comput. 91 (2014), 155–163. [20] T. Tamizh Chelvam and K. Selvakumar, On the intersection graph of gamma sets in the zero-divisor, Dis. Math. Alg. Appl. 7 (2015), no. 1, # 1450067, doi: 10.1142/S1793830914500670. [21] T. Tamizh Chelvam and K. Selvakumar, On the connectivity of the annihilating- ideal graphs, Discussiones Mathematicae General Algebra and Applications 35 (2015), no. 2, 195 – 204, doi:10.7151/dmgaa.1241. [22] T. Tamizh Chelvam, K. Selvakumar and V. Ramanathan, On the planarity of the k- zero-divisor hypergraphs, AKCE International Journal of Graphs and Combinatorics 12 (2015), no. 2, 169–176. [23] H. J. Wang and Neal O. Smith, Commutative rings with toroidal zero-divisor graphs, Houston J. Math. 36 (2010), no. 1, 1–31. [24] H. J. Wang, Zero-divisor graphs of genus one, J. Algebra 304 (2006), no. 2, 666–678. [25] C. Wickham, Classification of rings with genus one zero-divisor graphs, Comm. Algebra 36 (2008), 325–345. [26] A. T. White, Graphs, Groups and Surfaces, North-Holland, Amsterdam, 1973. Contact information T. Tamizh Chelvam, K. Selvakumar Department of Mathematics Manonmaniam Sundaranar University Tirunelveli 627 012, Tamil Nadu, India E-Mail(s): tamche59@gmail.com, selva_158@yahoo.co.in Received by the editors: 06.10.2015 and in final form 17.07.2016.