350
Views
2
CrossRef citations to date
0
Altmetric
Original Article

Nilpotent graphs with crosscap at most twoFootnote

&
Pages 229-237 | Received 19 Apr 2017, Accepted 22 Nov 2017, Published online: 10 Jun 2020

Abstract

Let R be a commutative ring with identity. The nilpotent graph of R, denoted by ΓN(R), is a graph with vertex set ZN(R), and two vertices x and y are adjacent if and only if xy is nilpotent, where ZN(R)={xR:xyis nilpotent, for someyR}. In this paper, we characterize finite rings (up to isomorphism) with identity whose nilpotent graphs can be embedded in the projective plane or Klein bottle. Also, we classify finite rings whose nilpotent graphs are ring graph or outerplanarity index 1,2.

1 Introduction

Throughout, rings are commutative with 10. The zero-divisor of a ring R is denoted by Z(R) and Z(R)=Z(R){0}. The zero-divisor graph of R, denoted by Γ(R), is an undirected graph whose vertices are elements of Z(R) and two distinct vertices a and b are adjacent if and only if ab=0. The zero-divisor graph of a commutative ring was first investigated by Beck [Citation1]. He was interested in coloring of the graph Γ(R).

Later, it was modified as in present form by Anderson and Livingston [Citation2]. This paper deals with the nilpotent graph of rings. Recall that the nilpotent graph of R, denoted by ΓN(R), is a graph with vertex set ZN(R), and two vertices x and y are adjacent if and only if xy is nilpotent, where ZN(R)={xR:xyis nilpotent,for someyR}. It was first investigated by Chen [Citation3]. He consoders all the elements of ring R as the vertices of the graph and two vertices x and y are adjacent if and only if xy is nilpotent. However, in 2010, Li [Citation4] modified and studied the nilpotent graph ΓN(R) of R. Note that the usual zerodivisor graph Γ(R) is a subgraph of the graph ΓN(R). If RF1×F2××Fn,where each Fi is a field and n2, then ΓN(R)Γ(R).

The focus of this paper is on finding non-orientable genus(crosscap) of the graph ΓN(R). This work is motivated by the following result of Hsieh [Citation5] who completely determined the finite commutative rings whose zero-divisor graphs are projective planar. This paper is organized as follows: In Section 2, we characterize finite rings (up to isomorphism) with identity whose nilpotent graphs can be embedded in the projective plane or Klein bottle. In Section 3, we classify finite rings whose nilpotent graphs are ring graph or outerplanar.

One can refer to [Citation6] for the following: We denote the ring of integers modulo n by Zn and the field with q elements by Fq. If S is a subset of R,we denote S=S{0},R×=U(R)= set of units of R and Spec(R) is the set of all prime ideals of R. If R is a direct product of local rings, then 0R denotes the zero element of R. We use Kn for the complete graph with n vertices and Km,n for the complete bipartite graph.

One can refer to [Citation7] for the following: By a surface, we mean a connected two-dimensional real manifold, i.e., a connected topological space such that each point has a neighborhood homeomorphic to an open disc. Every non-orientable compact surface Nk is a connected sum of finite copies of k projective planes (see [Citation7]). The number of copies of projective planes is called the crosscap number (non-orientable genus) of this surface. The crosscap of a graph G, denoted γ̄(G), is the smallest integer k such that the graph can be embedded in Nk. For example, the projective plane is of crosscap one and the Klein bottle is of crosscap two. A non-planar graph is said to be projective if it can be embedded into a projective plane. The following results are used in the subsequent section.

Lemma 1

[Citation7]

Let m,n be integers and for a real number x, let x denote the smallest integer that is greater than or equal to x. Then

(i)

γ¯(Kn)=(n3)(n4)6ifn3andn73ifn=7.

(ii)

γ¯(Km,n)=(m2)(n2)2, where m,n2.

Theorem 1

[Citation8, Kuratowski’s]

The graph G is planar if and only if it contains no subgraph isomorphic to K5 or K3,3 or a subdivision of one of these.

Theorem 2

[Citation9, Theorem 3.1]

Let (R,m) be a finite local ring with |R|=pn for some prime p and some positive integer n2. Then ΓN(R) is planar if and only if R is isomorphic to one of the following rings: Z4,Z2[x]x2,Z9 or Z3[x]x2.

Theorem 3

[Citation9, Theorem 3.2]

Let RR1××Rn×F1××Fm be a finite commutative ring with identity, where each (Ri,mi) is a local ring and Fj is a field, m,n1 and m+n2. Then ΓN(R) is planar if and only if R is isomorphic to one of the following rings: Z4×Z2 or Z2[x]x2×Z2.

Theorem 4

[Citation9]

Let RF1×F2××Fn be a finite ring, where each Fi is a field and n2. 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 or Z2×Z2×Z3, where F is a finite field.

Theorem 5

[Citation5]

Let R be a finite ring such that |Spec(R)|5. Then γ¯(Γ(R))6.

Theorem 6

[Citation9]

Let (R,m) be a finite local ring with |R|=pn for some prime p and some positive integer n2. Then ΓN(R)K|m|+K|R×|¯, where R×=Rm is the set of all units in R.

Theorem 7

[Citation5]

Let R be a finite ring such that |Spec(R)|=2. Then γ¯(Γ(R))=1 if and only if R is isomorphic to one of the following 16 rings: F4×F4,F4×Z5,Z2×F4[x]x2,Z2×Z4[x]x2+x+1,Z2×Z2[x,y]x2,xy,y2,Z2×Z4[x]2x,x2,Z3×Z8,Z3×Z2[x]x3,Z3×Z4[x]x22,x3,Z4×Z5,Z2[x]x2×Z5,F4×Z4,F4×Z2[x]x2,Z4×Z4,Z4×Z2[x]x2,Z2[x]x2×Z2[x]x2.

Theorem 8

[Citation5]

Let R be a finite ring such that |Spec(R)|=3. Then γ¯(Γ(R))=1 if and only if R is isomorphic to one of the following 6 rings: Z2×Z2×F4,Z2×Z2×Z4,Z2×Z2×Z2[x]x2,Z2×Z2×Z5,Z2×Z3×Z3,Z3×Z3×Z3.

Theorem 9

[Citation5]

Let R be a finite ring such that |Spec(R)|=4. Then γ¯(Γ(R))=1 if and only if R is isomorphic to Z2×Z2×Z2×Z2.

Theorem 10

[Citation10]

Suppose m0,n0, and mn1. The non-orientable genus of K¯m+Kn is given by γ¯(K¯m+Kn)=0ifn2(m2)(n2)2ifn2and(m,n)(4,5),4if(m,n)=(4,5).

Theorem 11

[Citation11]

The complete bipartite graph K3,5 admits the unique embedding into the Klein bottle, up to congruence, as shown in .

Fig. 1 The unique embedding of K3,5 on the Klein bottle.

2 Crosscap of ΓN(R)

In this section, we classify all finite rings (up to isomorphism) with identity whose nilpotent graphs can be embedded in a projective plane or Klein bottle.

The following Lemma gives the general formula for the crosscap of the graph ΓN(R) in the case of local ring.

Lemma 2

If (R,m) is a finite local ring with |R|=α and |m|=β4, then γ¯(ΓN(R))=(αβ2)(β3)2.

Proof

By Theorem 6, ΓN(R)K|R×|¯+K|m|. Also |R×||m|12 and (|R×|,|m|)(4,5). By Theorem 10, γ¯(ΓN(R))=γ¯(K|R×|¯+K|m|)=γ¯(K|R×|,|m|)=(αβ2)(β3)2.

Lemma 3

If RR1×R2××Rn(n2) is a finite commutative ring with identity, where Ri’s are local rings which are not fields, then γ¯(ΓN(R))5.

Proof

Since ΓN(Z4×Z4) is a subgraph of ΓN(R), ΓN(R) contains a copy of K4,7 which is a subgraph of the graph induced by the set of vertices [(R1××m2)(m1×R2×)(m1×m2)]{0R}. By Lemma 1 (ii), γ¯(ΓN(R))γ¯(K4,7)5. Thus the result follows.□

We consider the finite ring R and so RR1×R2××Rn where Ri is a local ring. On the observation of Lemma 3, at least one Ri is a field, for some i.

Lemma 4

Let RF1××Fn be a finite commutative ring with identity, where each Fj is a field, n2 and |F1||F2||Fn|. Then γ¯(ΓN(R))3 if one of the following holds:

(i)n5

(ii)n=4 and |F4|3

(iii)n=3,|F1|3, and either F2Z3 or F3Z3

(iv)n=3,|F1|=2=|F2| and |F3|8

(v)n=3,|F1|=2,|F2|=3 and |F3|5

(vi)n=3,|F1|=2 and |F2|4.

Proof

(i)

Follows from Theorem 5.

(ii)

Suppose n=4 and |F4|3. Then since ΓN(Z2×Z2×Z2×Z3) is a subgraph of ΓN(R), ΓN(R) contains a copy of K3,5 which is a subgraph of a graph induced by set of vertices [(Z2×Z2×0×0)(0×0×Z2×Z3)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,5)2. We claim that γ¯(ΓN(R))2. By Theorem 11, K3,5 has a unique embedding in N2. The embedding of K3,5 in N2 has six 4-faces and one 6 face. Consider the face f of length six. Let Vi be the partition of K3,5, for i=1,2. Then f contains exactly 3 vertices from Vi for all i. In order to embed the set of vertices S={(0,1,1,0),(1,0,0,1),(1,0,0,2)}, we have to choose the face f of length 6. Clearly N(S)={(1,0,0,0),(0,1,0,0),(0,0,0,1),(0,0,0,2),(0,0,1,0)}. If f contains all the vertices from N(S), then the vertices of N(S) that are adjacent to the vertices (0,1,1,0),(1,0,0,1) and (1,0,0,2) make two edge crossings where N(S) denote the set of vertices adjacent to the vertices of S. If f does not contain all the elements of N(S), then we cannot insert the vertices of S without crossing. Thus we cannot embed ΓN(R) in N2. Hence γ¯(ΓN(R))3.

(iii)

Suppose n=3,|F1|3, and either F2Z3 or F3Z3. Then ΓN(Z3×Z3×F4) is a subgraph of ΓN(R). So ΓN(R) contains a copy of K3,8 with vertex set [(0×0×F4)(Z3×Z3×0)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,8)3. By Theorem 8, γ¯(ΓN(Z3×Z3×Z3))=1.

(iv)

Suppose |F1|=2=|F2|,|F3|8 and n=3. Then since ΓN(Z2×Z2×F8) is a subgraph of ΓN(R), ΓN(R) contains a copy of K3,7 with vertex set [(Z2×Z2×0)(0×0×F8)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,7)3.

(v)

Suppose |F1|=2,|F2|=3,|F3|5 and n=3. Then since ΓN(Z2×Z3×Z5) is a subgraph of ΓN(R), ΓN(R) contains a copy of K4,5 with vertex set [(Z2×Z3×0)(0×0×Z5)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K4,5)3.

(vi)

Suppose |F1|=2,|F2|4 and n=3. Then since ΓN(Z2×F4×F4) is a subgraph of ΓN(R), ΓN(R) contains a copy of K3,7 with vertex set [(0×0×F4)(Z2×F4×0)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,7)3.□

The following Theorems characterize the ring R for which γ¯(ΓN(R))=1 and γ¯(ΓN(R))=2.

Theorem 12

Let R be a finite ring. Then γ¯(ΓN(R))=1 if and only if R is isomorphic to one of the following rings:

Z8,Z2[x]x3,Z4[x]2x,x22,Z2[x,y]x,y2,Z4[x]2,x2,Z4×Z3,Z2[x]x2×Z3,F4×F4,F4×Z5,Z2×Z2×F4,Z2×Z2×Z5,Z2×Z3×Z3,Z3×Z3×Z3 or Z2×Z2×Z2×Z2.

Proof

Assume that γ¯(ΓN(R))=1. Then by Theorem 5, |Spec(R)|4. Since R is finite, RR1×R2××Rn, where (Ri,mi) is a local ring for each i. Without loss of generality, assume that |R1||R2||Rn|. If |Spec(R)|=4, then by Lemma 4 (ii), RZ2×Z2×Z2×Z2.

Case 1: |Spec(R)|=3.

If Ri is a field for all i, then ΓN(R)Γ(R) and so by Theorem 8, R is isomorphic to one of the following rings: Z2×Z2×F4,Z2×Z2×Z5,Z2×Z3×Z3,Z3×Z3×Z3. If not, without loss of generality, let R1 be a local ring but not a field. Then ΓN(Z4×Z2×Z2) is a subgraph of ΓN(R). So ΓN(R) contains a copy of K3,6 which is the subgraph of a graph induced by the set of vertices [(Z4×0×0)(m1×Z2×Z2)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,6)2, a contradiction.

Case 2: |Spec(R)|=2.

If Ri is a field, then ΓN(R)Γ(R) and so by Theorem 7, RF4×F4 or F4×Z5. If not, without loss of generality, let R1 be a local ring but not a field. We claim that |m1|=1 and |R2|3.

Suppose that |m1|2. Then let z1=(b1,0),z2=(b2,0),z3=(u1,0),z4=(u2,0),z5=(u3,0),z6=(b1,1),z7=(b2,1) and z8=(0,1), where b1,b2m1,uiR1× for 1i3. Then ΓN(R) contains a copy of K3,5 which is the subgraph of the graph induced by set of vertices {z1,,z8} and so γ¯(ΓN(R))γ¯(K3,5)2, a contradiction. Hence |m1|=1 and so |R1|=4.

Suppose that |R2|4. Then let zm1 and y1=(z,0),y2=(u1,0),y3=(u2,0),y4=(0,v1),y5=(0,v2),y6=(0,v3),y7=(z,v1),y8=(z,v2),y9=(z,v3) where u1,u2R1× and viR2 for all i. Then ΓN(R) contains a copy of K3,6 which is the subgraph of the graph induced by set of vertices {y1,,y9}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,6)2, a contradiction. Hence |R2|3.

Thus we get RZ4×Z2,Z4×Z3,Z2[x]x2×Z2,Z2[x]x2×Z3. By Theorem 3, ΓN(R) is planar for RZ4×Z2,Z2[x]x2×Z2. Hence we conclude that RZ4×Z3 or Z2[x]x2×Z3.

Case 3: |Spec(R)|=1.

If |m1|3, then by Theorem 2, ΓN(R) is planar. Hence |m1|4. Now by Lemma 2, γ¯(ΓN(R))=(αβ2)(β3)2, where |R|=α and |m|=β. Since γ¯(ΓN(R))=1, |R|=8 and |m|=4. So R is isomorphic to one of the following rings: Z8,Z2[x]x3,Z4[x]2x,x22,Z2[x,y]x,y2 or Z4[x]2,x2.

Conversely, suppose R is isomorphic to one of the following rings: Z8,Z2[x]x3,Z4[x]2x,x22,Z2[x,y]x,y2 or Z4[x]2,x2. Then by Lemma 2, γ¯(ΓN(R))=1.

If RZ4×Z3 or Z2[x]x2×Z3, then by Theorem 3, ΓN(R) is non-planar. Moreover shows one explicit embedding of ΓN(R) in the projective plane. Thus γ¯(ΓN(R))=1.

Suppose R is isomorphic to one of the following rings: F4×F4,F4×Z5,Z2×Z2×F4,Z2×Z2×Z5,Z2×Z3×Z3,Z3×Z3×Z3 or Z2×Z2×Z2×Z2. Then by Theorems 7 9, γ¯(ΓN(R))=1. This completes the proof.□

Fig. 2 Embedding of ΓN(Z4×Z3)ΓN(Z2[x]x2×Z3) in N1.

Theorem 13

Let R be a finite ring. Then γ¯(ΓN(R))=2 if and only if R is isomorphic to one of the following rings: Z4×F4,Z2[x]x2×F4,F4×Z7,Z5×Z5,Z2×Z2×Z7 or Z2×Z3×F4.

Proof

Assume that γ¯(ΓN(R))=2. Then by Theorem 5, |Spec(R)|4. Since R is finite, RR1×R2××Rn, where (Ri,mi)’s are local ring for each i. Without loss of generality, assume that |R1||R2||Rn|.

Case 1: |Spec(R)|=4.

Then by Lemma 4 (ii), RZ2×Z2×Z2×Z2. Also, by Theorem 12, γ¯(ΓN(R))=1. Thus there is no ring available in this case.

Case 2: |Spec(R)|=3.

If Ri is a field for all i, then by Lemma 4 (iii)(vi), R is isomorphic to one of the following rings: Z2×Z2×Z2,Z2×Z2×Z3,Z2×Z2×F4,Z2×Z2×Z5,Z2×Z2×Z7,Z2×Z3×Z3,Z2×Z3×F4,Z3×Z3×Z3.

(i) If R is isomorphic to Z2×Z2×Z2, then ΓN(R)Γ(R) and so by Theorem 4, ΓN(R) is planar.

(ii) If R is isomorphic to one of the following rings: Z2×Z2×Z3,Z2×Z2×F4,Z2×Z2×Z5,Z2×Z3×Z3,Z3×Z3×Z3, then by Theorem 12, γ¯(ΓN(R))=1.

Thus R is isomorphic to the following rings: Z2×Z2×Z7,Z2×Z3×F4.

Suppose that at least one Ri is not a field. Without loss of generality, let R1 be a local ring but not a field. Then ΓN(Z4×Z2×Z2) is a subgraph of ΓN(R). So ΓN(R) contains a copy of K3,6 which is the subgraph of a graph induced by the set of vertices [(Z4×0×0)(m×Z2×Z2)]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,6)2. We claim that γ¯(ΓN(R))2. By Theorem 11, K3,5 has a unique embedding in N2 and it has six 4-faces and one 6-face. Hence we can find a unique embedding of K3,6 in N2 that contains nine 4-faces. Let Vi be a partition of K3,6 for i=1,2. Then each face contains exactly two vertices from Vi. Also each v from V2 lie in exactly 3 faces. So there may be at most one face that contains the vertices {(0,1,0),(2,0,0),(2,1,0)}. Suppose there exists a face that contains the vertices {(0,1,0),(2,0,0),(2,1,0)}. Since (1,0,1) and (3,0,1) are adjacent to these vertices in ΓN(R), there must be one edge crossing in the embedding of ΓN(R). Suppose not. Then we cannot insert that vertices without crossing. Hence we cannot have the embedding of ΓN(R) in the non-orientable surface of crosscap 2. Thus γ¯(ΓN(R))3, a contradiction.

Case 3: |Spec(R)|=2.

If Ri is a field, then ΓN(R)K|F1|,|F2|. Since γ¯(ΓN(R))=2,|F1|=4,|F2|=7 or |F1|=5=|F2|. Hence RF4×Z7 or Z5×Z5. If not, then at least one Ri is not a field. Without loss of generality, let R1 be a local ring but not a field. We claim that |m1|=1 and |R2|4.

Suppose that |m1|=2. Then R1Z9 or Z3[x]x2. Let b1,b2m1. Let z1=(b1,0),z2=(b2,0),z3=(u1,0),z4=(u2,0),z5=(u3,0),z6=(b1,1),z7=(b2,1),z8=(0,1),z9=(u4,0),z10=(u5,0)and z11=(u6,0), where uiR1× for 1i6. Then ΓN(R) contains a copy of K5,6 which is the subgraph of the graph induced by set of vertices {z1,,z11} and so γ¯(ΓN(R))γ¯(K5,6)6, a contradiction. Hence |m1|=1 or |m1|3.

Suppose that |m1|3. Then ΓN(R) contains a copy of K4,7 which consists of vertices of [{(a,0):aR1×}{(b,1),(b,0):bm1}]{0R}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K4,7)5, which contradicts our assumption. Therefore |m1|=1 and so |R1|=4.

Suppose that |R2|5. Then let zm1 and y1=(z,0),y2=(u1,0),y3=(u2,0),y4=(0,v1),y5=(0,v2),y6=(0,v3),y7=(0,v4),y8=(z,v1),y9=(z,v2),y10=(z,v3),y11=(z,v4) where u1,u2R1× and viR2 for all i. Then ΓN(R) contains a copy of K3,8 which is the subgraph of the graph induced by set of vertices {y1,,y11}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,8)3, a contradiction. Hence |R2|4.

Thus we get R is isomorphic to one of the following rings: Z4×Z2,Z4×Z3,Z4×F4,Z2[x]x2×Z2,Z2[x]x2×Z3,Z2[x]x2×F4. By Theorem 2, ΓN(R) is planar for RZ4×Z2,Z2[x]x2×Z2. Also, by Theorem 12, γ¯(ΓN(R))=1 for RZ4×Z3,Z2[x]x2×Z3. Hence we conclude that RZ4×F4 or Z2[x]x2×F4.

Conversely, suppose that RF4×Z7 or Z5×Z5. Then since ΓN(R)K|F1|,|F2|, by Lemma 1 (ii), γ¯(ΓN(R))=2.

Assume that R is isomorphic to one of the following rings: Z4×F4 or Z2[x]x2×F4.

Then since ΓN(Z4×F4)ΓN(Z2[x]x2×F4), ΓN(R) contains a copy of K3,6 which is the subgraph of the graph induced by the set of vertices [(Z4×0)(m×F4)]{0R}. By Lemma 1 (ii), γ¯(ΓN(R))γ¯(K3,6)2. Moreover shows one explicit embedding of ΓN(R) in N2. Thus γ¯(ΓN(R))=2.

Assume that R is isomorphic to Z2×Z2×Z7. Then let ri=(0,0,i),r6+i=(0,1,i),r12+i=(1,0,i),r19=(0,1,0),r20=(1,0,0) and r21=(1,1,0) for 1i6. Then ΓN(R) contains a copy of K3,6 which consists of vertices {r1,,r6,r19,r20,r21}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,6)2. Moreover shows the embedding in the non-orientable surface of crosscap 2. Thus γ¯(ΓN(R))=2.

Suppose that RZ2×Z3×F4. Then let wi=(0,0,ai),w3+i=(0,1,ai),w6+i=(0,2,ai),w9+i=(1,0,ai),w13=(0,1,0),w14=(0,2,0),w15=(1,0,0),w16=(1,1,0),w17=(1,2,0) for 1i3. Then ΓN(R) contains a copy of K3,5 which consists of vertices {w1,w2,w3,w13,w14,w15,w16,w17}. It then follows from Lemma 1 (ii) that γ¯(ΓN(R))γ¯(K3,5)2. Moreover shows one explicit embedding of ΓN(R) in the non-orientable surface of crosscap 2. Thus γ¯(ΓN(R))=2. This completes the proof.□

Fig. 3 Embedding of ΓN(Z4×F4)ΓN(Z2[x]x2×F4) in N2.
Fig. 4 Embedding of ΓN(Z2×Z2×Z7) in N2.
Fig. 5 Embedding of ΓN(Z2×Z3×F4) in N2.

3 Outerplanarity of ΓN(R)

Let G be a graph with n vertices and q edges. A chord is any edge of G joining two nonadjacent vertices in a cycle of G. Let C be a cycle of G. We say C is a primitive cycle if it has no chords. Also, a graph G has the primitive cycle property (PCP) if any two primitive cycles intersect in at most one edge. The number frank(G) is called the free rank of G and it is the number of primitive cycles of G. Also, the number rank(G)=qn+r is called the cycle rank of G, where r is the number of connected components of G. The cycle rank of G can be expressed as the dimension of the cycle space of G. By [Citation12, Proposition 2.2], we have rank(G)frank(G). A graph G is called a ring graph if it satisfies one of the following equivalent conditions (see [Citation9]).

(i) rank(G) = frank(G),

(ii) G satisfies the PCP and G does not contain a subdivision of K4 as a subgraph.

The following concepts were defined in [Citation13]. An embedding ϕ of a planar graph is 1-outerplanar if it is outerplanar (i.e.) all vertices are incident on the outerface in ϕ. An embedding of a planar graph is k-outerplanar if removing all vertices on the outerface (together with their incident edges) yields a (k1)-outerplanar embedding. A graph is k-outerplanar if it has a k-outerplanar embedding. The outerplanarity index of a graph G is the smallest k such that G is k-outerplanar. There is a characterization for outerplanar graphs that says a graph is outerplanar if and only if it does not contain a subdivision of K4 or K2,3. Clearly, every outerplanar graph is a ring graph or every ring graph is a planar graph. In this section, we characterize all rings R such that ΓN(R) is a ring graph and outerplanar graph with index 1,2.

Theorem 14

Let R be a finite ring. Then ΓN(R) is a ring graph if and only if R is isomorphic to one of the following rings:

Z4,Z2[x]x2,Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2,Z2×F,Z3×Z3 or Z2×Z2×Z2.

Proof

Assume that ΓN(R) is a ring graph. Then since every ring graph is planar, it is enough to check rings in Theorems 2–4, whether it is a ring graph or not. Now the nilpotent graph of rings Z4,Z2[x]x2,Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2,Z2×F,Z3×Z3,Z2×Z2×Z2,we have that rank(ΓN(R))=frank(ΓN(R)), and so they are ring graphs. Also for rings Z2×Z2×Z3,Z3×F, where |F|4, ΓN(R) contains a subdivision of K4 or does not satisfy PCP (see ) and so they are not ring graphs. The converse is obvious.□

short-legendFig. 6

Theorem 15

Let R be a finite ring. Then ΓN(R) is an outerplanar if and only if R is isomorphic to one of the following rings: Z4,Z2[x]x2,Z2×F,Z3×Z3 or Z2×Z2×Z2.

Proof

Assume that ΓN(R) is an outerplanar. Then since ΓN(R) is a ring graph, by Theorem 14, we need to check rings Z4,Z2[x]x2,Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2,Z2×F,Z3×Z3,Z2×Z2×Z2are outerplanar or not. By , we can find the subdivision of K2,3 in ΓN(R) for rings Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2 and so are not outerplanar. Also, the nilpotent graph for rings Z4,Z2[x]x2,Z2×F,Z3×Z3,Z2×Z2×Z2, is P3,K1,q1,C4 or G (see ) respectively and so are outerplanar. The converse is obvious.□

short-legendFig. 7

Theorem 16

Let R be a finite ring. Then ΓN(R) has an outerplanarity index 2 if and only if R is isomorphic to one of the following rings:

Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2,Z3×F,Z2×Z2×Z3,|F|4.

Proof

Assume that ΓN(R) has an outerplanarity index 2. Then ΓN(R) is planar, by Theorems 2–4, we need only to check rings: Z4,Z2[x]x2,Z9,Z3[x]x2,Z4×Z2,Z2[x]x2×Z2,Z2×F,Z3×F,Z2×Z2×Z2,Z2×Z2×Z3. are 2-outerplanar and is minimum or not. Since ΓN(R) is not 1-outerplanar, by Theorem 15, R is not isomorphic to the following rings: Z4,Z2[x]x2,Z2×F,Z3×Z3,Z2×Z2×Z2. By Figs. , , the nilpotent graph for rings Z9,Z3[x]x2,Z4×Z2,Z2[x]x2,Z3×F,Z2×Z2×Z3,|F|4, are 2-outerplanar, as the removal of vertices in the outer face remains the graph P3 or totally disconnected. The converse is obvious.□

Corollary 1

Let R be a finite ring. Then ΓN(R) has an outerplanarity index at most two.

Notes

Peer review under responsibility of Kalasalingam University.

References

  • Beck I. Coloring of commutative rings J. Algebra 116 1988 208 226
  • Anderson D.F. Livingston P.S. The zero-divisor graph of a commutative ring J. Algebra 217 1999 434 447
  • Chen P.W. A kind of graph structure of rings Algebra Colloq. 10 2 2003 229 238
  • Li A. Li Q. A kind of graph of structure on von-Neumann regular rings Int. J. Algebra 4 6 2010 291 302
  • Chiang-Hsieh H.J. Classification of rings with projective zero-divisor graphs J. Algebra 319 2008 2789 2802
  • Atiyah M.F. Macdonald I.G. Introduction to Commutative Algebra 1969 Addison–Wesley Reading, MA
  • White T. Graphs, Groups and Surfaces North-Holland Mathematics Studies vol. 188 (1984) North-Holland. Amsterdam.
  • Chartrand G. Lesniak L. Graphs and Digraphs 1986 Wadsworth and Brooks/Cole Monterey, CA
  • Kala R. Kavitha S. On nilpotent graph of genus one Discrete Math. Algorithms Appl. 63 2014 1450037(1–10)
  • Ellingham M.N. Stephens C. The nonorientable genus of joins of complete graphs with large edgeless graphs J. Combin. Theory Ser. B 97 2007 827 845
  • Negami Seiya Suzuki Yusuke The 2-extendability of 5-connected graphs on the Klein bottle Discrete Math. 310 2010 2510 2518
  • Gitler I. Reyes E. Villarreal R.H. Ring graphs and complete intersection toric ideals Discrete Math. 310 3 2010 430 441
  • Kammer Frank Determining the smallest k such that G is k-outerplanar Lecture Notes in Comput. Sci. 4698 2007 359 370