147
Views
0
CrossRef citations to date
0
Altmetric
Research Article

Tight toughness bounds for path-factor critical avoidable graphs

&
Received 12 Oct 2022, Accepted 29 Jan 2024, Published online: 26 Feb 2024

Abstract

Given a graph G and an integer k2, a spanning subgraph H of G is called a Pk-factor of G if every component of H is a path with at least k vertices. A graph G is Pk-factor avoidable if for every edge eE(G), G has a Pk-factor excluding e. A graph G is said to be (Pk,n)-factor critical avoidable if the graph GV is Pk-factor avoidable for any VV(G) with |V|=n. Here we study the sharp bounds of toughness and isolated toughness conditions for the existence of (P3,n)-factor critical avoidable graphs. In view of graph theory approaches, this paper mainly contributes to verify that (i) An (n+2)-connected graph is (P3,n)-factor critical avoidable if its toughness τ(G)>n+24; (ii) An (n+2)-connected graph is (P3,n)-factor critical avoidable if its isolated toughness I(G)>n+64.

Mathematics Subject Classification:

1 Introduction

All graphs considered here are finite and simple. We refer to [Citation3] for the notations and terminologies not defined here. Let G be a graph with vertex set V(G) and edge set E(G). Given a vertex vV(G), let dG(v) denote the degree of v. If dG(v)=0 for some vertex v in G, then v is said to be an isolated vertex in G. Let i(G) be the number of isolated vertices of G. For any subset SV(G), let G[S] denote the subgraph of G induced by S, and GS:=G[V(G)S] is the resulting graph after deleting the vertices of S from G. The number of connected components of a graph G is denoted by ω(G). We write κ(G) for the vertex connectivity of G.

A subgraph H of G is called a spanning subgraph of G if V(H)=V(G) and E(H)E(G). For a family of connected graphs F, a spanning subgraph H of a graph G is called an F-factor of G if each component of H is isomorphic to some graph in F. A spanning subgraph H of a graph G is called a Pk-factor of G if every component of H is isomorphic to a path of order at least k, where k2 is an integer. For example, a P3-factor means a graph factor in which every component is a path of order at least three. Since Tutte proposed the well known Tutte 1-factor theorem [16], there are many results on graph factors [Citation2, Citation4, Citation6, Citation8, Citation9, Citation11, Citation12, Citation14, Citation15, Citation17, Citation18, Citation23, Citation26] and factor covered graphs [Citation7, Citation10, Citation20, Citation22, Citation25, Citation27]. More results on graph factors of graphs refer to the books [Citation1, Citation21].

In order to characterize a graph possessing a P3-factor, Kaneko [Citation13] put forward the concept of a sun as follows. A graph R is called factor-critical if Rv has a 1-factor for each vV(R). Let R be a factor-critical graph and V(R)={v1,v2,,vn}. By adding new vertices {u1,u2,,un} together with new edges {viui:1in} to R, the resulting graph is called a sun. Note that, according to Kaneko [Citation13], we regard K1 and K2 also as a sun, respectively. Usually, the suns other than K1 and K2 are also called big suns. It is called a sun component of G if the component of G is isomorphic to a sun. We denote by sun(G) the number of sun components in G.

Theorem 1.1.

(Kaneko [Citation13]) A graph G has a P3-factor if and only if sun(GX)2|X| for all XV(G).

Next, we introduce two parameters for a graph, namely, toughness and isolated toughness. The toughness of G was first introduced by Chvátal [Citation5] as τ(G)=min{|X|ω(GX):XV(G),ω(GX)2}, if G is not complete; otherwise, τ(G)=+. The isolated toughness of G was defined by Yang et al. in [Citation19] as I(G)=min{|X|i(GX):XV(G),i(GX)2}, if G is not complete; otherwise, I(G)=+.

A graph G is Pk-factor avoidable if for every edge eE(G), G has a Pk-factor that excludes e. A graph G is (Pk,n)-factor critical avoidable if for any subset VV(G) with |V|=n, the graph GV is Pk-factor avoidable. More recently, Zhou [Citation24] explored the relationship between toughness (resp. isolated toughness) and (Pk,n)-factor critical avoidable graphs. In the same paper, they gave toughness and isolated toughness conditions for (P3,n)-factor critical avoidable graphs, respectively. However, we find the toughness bound and isolated toughness bound obtained by Zhou [Citation24] are not sharp and have a lot of space to improve them.

Inspired by finding the sharpness bounds, in this study, we continue to study the relationship between different graph parameters and the existence of (P3,n)-factor critical avoidable graphs. Our main results are listed as follows which will be proved in the next section. Note that the conditions in Theorem 1.2 and 1.3 are sharp, which improve the results of Zhou [Citation24].

Theorem 1.2.

Let G be an (n+2)-connected graph, where n0 is an integer. If its toughness τ(G)>n+24, then G is a (P3,n)-factor critical avoidable graph.

Theorem 1.3.

Let G be an (n+2)-connected graph, where n is a positive integer. If its isolated toughness I(G)>n+64, then G is a (P3,n)-factor critical avoidable graph.

2 Proof of main results

Our goal in this section is to present the detailed proofs of main theorems. Additionally, we will use some examples to explain all the conditions in Theorems 1.2 and 1.3 are best possible in some sense.

2.1 Proof of Theorem 1.2

If G is a complete graph, then it is easily seen that G is a (P3,n)-factor critical avoidable graph by κ(G)n+2. Next, we consider that G is a non-complete graph. For any VV(G) with |V|=n, we write G=GV. To verify Theorem 1.2, it suffices to prove that H:=Ge contains a P3-factor for any e=xyE(G). On the contrary, we assume that H has no P3-factor. Then by Theorem 1.1, there exists a subset XV(H) such that sun(HX)>2|X|. In terms of the integrality of sun(HX), we obtain that (1) sun(HX)2|X|+1.(1)

Claim 2.1. X.

Proof.

On the contrary, we assume that X=. It follows from (1) that (2) sun(H)=sun(HX)2|X|+1=1.(2)

Note that H is connected with at least three vertices since κ(H)κ(G)|V|11. Combining this with (2), we have 1sun(H)ω(H)=1, that is, (3) sun(H)=ω(H)=1.(3)

It follows from (3) that H is a big sun, and |V(H)|6. Let R be the factor-critical graph of H. Then |V(R)|3 and there exists wV(R) such that ω(G{w})=ω(H{w})=2. Obviously, w is a cut-vertex of G, which is a contradiction since κ(G)κ(G)|V|2. Hence, Claim 2.1 is verified. □

Claim 2.2. |X|2.

Proof.

On the contrary, by Claim 2.1, we assume that |X|=1. Since G is (n+2)-connected, κ(G)κ(G)n=2. It follows that GX is a connected graph, i.e., ω(GX)=1. Then by the definition of H, we have that (4) ω(HX)=ω(GXe)ω(GX)+1=2.(4)

On the other hand, according to (1), ω(HX)sun(HX)2|X|+13,

which is a contradiction to (4). Hence, Claim 2.2 is verified. □

It follows immediately from (1) and Claim 2.2 that sun(HX)2|X|+15. Then, we obtain that ω(G(VX))=ω(GX)ω(GXe)1=ω(HX)1sun(HX)14.

Combining this with (1), Claim 2.2 and the definition of τ(G), we have τ(G)|VX|ω(G(VX))n+|X|sun(HX)1n+|X|2|X|12+n2|X|12+n4=n+24, which contradicts that τ(G)>n+24 in Theorem 1.2.

Remark 2.1.

Next, we show that the toughness condition τ(G)>n+24 in Theorem 1.2 is sharp. We construct a graph G=Kn+2(4K2). Then it is easily seen that τ(G)=n+24 and κ(G)=n+2. Set eE(4K2),VV(Kn+2) with |V|=n. Let G=GV and H=Ge. For X=V(Kn+2)V, we admit sun(HX)=5>4=2|X|. According to Theorem 1.1, H does not contain P3-factor. Hence, G is not a P3-factor avoidable graph. Combining this with G=GV, G is not a (P3,n)-factor critical avoidable graph.

2.2 Proof of Theorem 1.3

Theorem 1.3 obviously holds for a complete graph. Next, we assume that G is not complete. Let VV(G) with |V|=n, and G=GV. Let e=xyE(G) and H=Ge. Since G is (n+2)-connected, thus H is connected. To prove Theorem 1.3, it suffices to show that H admits a P3-factor. On the contrary, suppose that H has no P3-factor. Then by Theorem 1.1, there exists a subset XV(H) such that (5) sun(HX)2|X|+1.(5)

Claim 2.3. X.

Proof.

Assume X=. Then by (5), we have that (6) sun(H)=sun(HX)2|X|+1=1.(6)

Note that H is connected with at least three vertices since κ(H)κ(G)|V|11. This together with (6) implies sun(H)=ω(H)=1. It follows that H is a big sun with |V(H)|6. Let R be the factor-critical graph of H, then i(HV(R))=|V(R)|3, and thus G=H+e has a vertex with degree one. This is a contradiction, since κ(G)κ(G)|V|2. Hence, Claim 2.3 holds. □

Claim 2.4. |X|2.

Proof.

On the contrary, by Claim 2.3, we assume that |X|=1. Since G is (n+2)-connected, κ(G)κ(G)n=2. It follows that GX is a connected graph, i.e., ω(GX)=1. Then by the definition of H, we have that (7) ω(HX)=ω(GXe)ω(GX)+1=2.(7)

On the other hand, according to (5), ω(HX)sun(HX)2|X|+13,

which is contradiction to (7). Hence, Claim 2.4 is verified. □

Let Sun(HX) denote the union of sun components of HX, which consists of a isolated vertices, b K2-components and c big sun components S1,S2,,Sc. Let Ri be the factor-critical subgraph of Si for 1ic, and write Z=i=1cV(Ri). We select one vertex from every K2 component of HX, and the set of such vertices is denoted by Y. Clearly, |Y|=b. Then i(H(XYZ))=a+b+|Z| and it follows from (5) and Claim 2.3 that (8) a+b+c=sun(HX)2|X|+15.(8)

Claim 2.5. x,yV(bK2)V(S1)V(Sc).

Proof.

On the contrary, without loss of generality, we assume that xV(bK2)V(S1)V(Sc). Then there exists zV(bK2)V(S1)V(Sc) such that xzE(G) and there is at least one vertex of {x, z} with degree one in the subgraph bK2S1Sc. Thus, we obtain (9) i(G(VX(YZ{z}){x}))=a+b+3c5(9) by (8), c0 and |Z|3c. Combining(9) with the definition of I(G) and I(G)>n+64, we obtain n+64<I(G)|VX(YZ{z}){x}|i(G(VX(YZ{z}){x}))n+|X|+b+|Z|a+b+|Z|1+n+|X|a+b+|Z|1+n+|X|a+b+c1+12+na+b+c32+n5=2n+1510,

which is a contradiction. Hence, Claim 2.5 is verified. □

Claim 2.6. 0a1.

Proof.

On the contrary, we assume that a2. In terms of (8), c0 and |V(Ri)|3, we obtain i(G(VXYZ{x}))=i(G(XYZ{x})e)=i(H(XYZ{x}))i(H(XYZ))1=a+b+|Z|1a+b+c14.

Combining this with (8), a2,c0 and the definition of I(G), we derive n+64<I(G)|VXYZ{x}|i(G(VXYZ{x}))n+|X|+b+|Z|+1a+b+|Z|11+n+|X|+2aa+b+|Z|11+n+|X|a+b+c11+12+na+b+c11+12+n4=n+64,

which is a contradiction. Hence, Claim 2.6 holds. □

By Claim 2.6, we easily see that xV(aK1) or yV(aK1) since 0a1. This together with Claim 2.5 implies that there is at least one vertex in {x, y} such that the vertex does not belong V(aK1)V(bK2)V(S1)V(Sc). Without loss of generality, we let xV(aK1)V(bK2)V(S1)V(Sc), then xV(G)(VV(aK1)V(bK2)V(S1)V(Sc)). Next, we will distinguish two cases below to complete the proof of Theorem 1.3.

Case 1.

yV(aK1).

In terms of (8), c0 and |Z|3c, we can easily deduce that (10) i(G(VXYZ))a+b+|Z|a+b+3c5.(10)

Combining this with (10) and the definition of I(G), we derive (11) I(G)|VXYZ|i(G(VXYZ))n+|X|+b+|Z|a+b+|Z|.(11)

It follows from (8), (11), a0,c0,|Z|3c and I(G)>n+64 that 0(I(G)1)(a+b+|Z|)+an|X|(I(G)1)(a+b+3c)n|X|(I(G)1)(a+b+c)n|X|(I(G)1)(2|X|+1)n|X|=(2|X|+1)I(G)n3|X|1, which implies (12) I(G)3|X|+n+12|X|+1.(12)

In terms of (12), n1 and |X|2, we have I(G)3|X|+n+12|X|+1=32+n122|X|+132+n125=n+75, which contradicts I(G)>n+64.

Case 2.

yV(aK1).

In this case, a1. In terms of (8), c0 and |Z|3c, we can derive that (13) i(G(VXYZ{x}))a+b+|Z|a+b+3c5.(13)

According to (13) and the definition of I(G), we derive (14) I(G)|VXYZ{x}|i(G(VXYZ{x}))n+|X|+b+|Z|+1a+b+|Z|.(14)

It follows from (8), (14), a1,c0,|Z|3c,|X|2 and I(G)>n+64 that n+64<I(G)n+|X|+b+|Z|+1a+b+|Z|=1+n+|X|+1aa+b+|Z|1+n+|X|a+b+3c1+n+a+b+c12a+b+c1+12+n12a+b+c32+n125=n+75, which is a contradiction.

This finishes the proof of Theorem 1.3.

Remark 2.2.

The isolated toughness condition I(G)>n+64 in Theorem 1.3 is sharp. We construct a graph G=Kn+2(4K2). Then it is easily seen that I(G)=n+64 and κ(G)=n+2. Set eE(4K2),VV(Kn+2) with |V|=n. Let H=GVe. For X=V(Kn+2)V, we admit sun(HX)=5>4=2|X|. According to Theorem 1.1, H does not contain P3-factor. Hence, G is not a (P3,n)-factor critical avoidable graph.

Additional information

Funding

This work was supported by Natural Science Foundation of Jiangsu Province (No. BK20230388) and National Natural Science Foundation of China (No. 12301441).

References

  • Akiyama, J., Kano, M. (2011). Factors and Factorizations of Graphs. Lecture Notes in Mathematics, Vol. 2031. Berlin: Springer, pp. 1–347.
  • Ando, K., Egawa, Y., Kaneko, A., Kawarabayashi, K. I., Matsuda, H. (2002). Path factors in claw-free graphs. Discrete Math. 243: 195–200.
  • Bondy, J. A., Murty, U. S. R. (1982). Graph Theory with Applications. NewYork-Amsterdam-Oxford: North-Holland.
  • Chen, Y., Dai, G. (2022). Binding number and path-factor critical deleted graphs. AKCE Int. J. Graphs Comb. 19(3): 197–200.
  • Chvátal, V. (1973). Tough graphs and Hamiltonian Circuits. Discrete Math. 5: 215–228.
  • Dai, G. (2022). Remarks on component factors in graphs. RAIRO-Oper. Res. 56: 721–730.
  • Dai, G. (2023). The existence of path-factor covered graphs. Discussiones Math. Graph Theory. 43: 5–16.
  • Dai, G., Hang, Y., Zhang, X., Zhang, Z., Wang, W. (2022). Sufficient conditions for graphs with {P2,P5} -factor. RAIRO-Oper. Res. 56: 2895–2901.
  • Dai, G., Hu, Z. (2020). P3-factors in the square of a tree. Graphs Comb. 36: 1913–1925.
  • Dai, G., Zhang, Z., Hang, Y., Zhang, X. (2021). Some degree conditions for P≥k -factor covered graphs. RAIRO-Oper. Res. 55: 2907–2913.
  • Egawa, Y., Furuya, M., Ozeki, K. (2018). Sufficient conditions for the existence of a path-factor which are related to odd components. J. Graph Theory. 89: 327–340.
  • Kaneko, A., Kelmans, A., Nishimura, T. (2001). On packing 3-vertex paths in a graph. J. Graph Theory. 36: 175–197.
  • Kaneko, A. (2003). A necessary and sufficient condition for the existence of a path factor every component of which is a path of length at least two. J. Combin. Theory Ser. B. 88: 195–218.
  • Kano, M., Lee, C., Suzuki, K. (2008). Path and cycle factors of cubic bipartite graphs. Discussiones Math. Graph Theory. 28: 551–556.
  • Liu, H (2022). Binding number for path-factor uniform graphs. Proc. Romanian Acad. Ser. A: Math. Phys. Techn. Sci. Inf. Sci. 23(1): 25–32.
  • Tutte, W. T. (1952). The factors of graphs. Can. J. Math. 4: 314–328.
  • Wang, S., Zhang, W. (2022). Isolated toughness for path factors in networks. RAIRO-Oper. Res. 56(4): 2613–2619.
  • Wu, J. (2022). Path-factor critical covered graphs and path-factor uniform graphs. RAIRO-Oper. Res. 56(6): 4317–4325.
  • Yang, J., Ma, Y., Liu, G. (2001). Fractional (g, f)-factors in graphs. Appl. Math. J. Chinese Univ. Ser. A. 16: 385–390.
  • Yu, Q., Chen, C. (1987). On tree-factor covered graphs. J. Combin. Math. Combin. Comput. 2: 211–218.
  • Yu, Q., Liu, G. (2009). Graph Factors and Matching Extensions. Beijing: Higher Education Press.
  • Zhang, H., Zhou, S. (2009). Characterizations for P≥2 -factor and P≥3 -factor covered graphs, Discrete Math. 309: 2067–2076.
  • Zhou, S. (2023). Path factors and neighborhoods of independent sets in graphs. Acta Math. Appl. Sinica, English Ser. 39(2): 232–238.
  • Zhou, S. (2023). Some results on path-factor critical avoidable graphs. Discuss. Math. Graph Theory. 43: 233–244.
  • Zhou, S., Wu, J., Bian, Q. (2022). On path-factor critical deleted (or covered) graphs. Aequationes Math. 96: 795–802.
  • Zhou, S., Wu, J., Xu, Y. (2022). Toughness, isolated toughness and path factors in graphs. Bull. Austral. Math. Soc. 106(2): 195–202.
  • Zhou, S., Wu, J., Zhang, T. (2017). The existence of P≥3 -factor covered graphs. Discuss. Math. Graph Theory. 37: 1055–1065.