Research article

Injective coloring of planar graphs with girth 5

  • Received: 23 February 2023 Revised: 20 April 2023 Accepted: 24 April 2023 Published: 17 May 2023
  • MSC : 05C10, 05C15

  • A k-injective-coloring of a graph G is a mapping c:V(G){1,2,,k} such that c(u)c(v) for any two vertices u and v if u and v have a common vertex. The injective chromatic number of G, denoted by χi(G), is the least k such that G has an injective k-coloring. In this paper, we prove that for planar graph G with g(G)5, Δ(G)20 and without adjacent 5-cycles, χi(G)Δ(G)+2.

    Citation: Yuehua Bu, Qiang Yang, Junlei Zhu, Hongguo Zhu. Injective coloring of planar graphs with girth 5[J]. AIMS Mathematics, 2023, 8(7): 17081-17090. doi: 10.3934/math.2023872

    Related Papers:

    [1] Hongyu Chen, Li Zhang . A smaller upper bound for the list injective chromatic number of planar graphs. AIMS Mathematics, 2025, 10(1): 289-310. doi: 10.3934/math.2025014
    [2] Yanyi Li, Lily Chen . Injective edge coloring of generalized Petersen graphs. AIMS Mathematics, 2021, 6(8): 7929-7943. doi: 10.3934/math.2021460
    [3] Yanping Yang, Yang Wang, Juan Liu . List vertex arboricity of planar graphs without 5-cycles intersecting with 6-cycles. AIMS Mathematics, 2021, 6(9): 9757-9769. doi: 10.3934/math.2021567
    [4] Yuehua Bu, Hongrui Zheng, Hongguo Zhu . On the list injective coloring of planar graphs without a $ {4^ - } $-cycle intersecting with a $ {5^ - } $-cycle. AIMS Mathematics, 2025, 10(1): 1814-1825. doi: 10.3934/math.2025083
    [5] Yueying Zhao, Lianying Miao . Planar graphs without $ \{4, 6, 8\} $-cycles are 3-choosable. AIMS Mathematics, 2021, 6(9): 10253-10265. doi: 10.3934/math.2021593
    [6] Yuehua Bu, Qi Jia, Hongguo Zhu, Junlei Zhu . Acyclic edge coloring of planar graphs. AIMS Mathematics, 2022, 7(6): 10828-10841. doi: 10.3934/math.2022605
    [7] Fugang Chao, Donghan Zhang . Neighbor sum distinguishing total choice number of IC-planar graphs with restrictive conditions. AIMS Mathematics, 2023, 8(6): 13637-13646. doi: 10.3934/math.2023692
    [8] Xiaoxue Hu, Jiangxu Kong . An improved upper bound for the dynamic list coloring of 1-planar graphs. AIMS Mathematics, 2022, 7(5): 7337-7348. doi: 10.3934/math.2022409
    [9] Yunfeng Tang, Huixin Yin, Miaomiao Han . Star edge coloring of $ K_{2, t} $-free planar graphs. AIMS Mathematics, 2023, 8(6): 13154-13161. doi: 10.3934/math.2023664
    [10] Jin Cai, Shuangliang Tian, Lizhen Peng . On star and acyclic coloring of generalized lexicographic product of graphs. AIMS Mathematics, 2022, 7(8): 14270-14281. doi: 10.3934/math.2022786
  • A k-injective-coloring of a graph G is a mapping c:V(G){1,2,,k} such that c(u)c(v) for any two vertices u and v if u and v have a common vertex. The injective chromatic number of G, denoted by χi(G), is the least k such that G has an injective k-coloring. In this paper, we prove that for planar graph G with g(G)5, Δ(G)20 and without adjacent 5-cycles, χi(G)Δ(G)+2.



    All graphs considered in this paper are finite simple graphs. For a planar graph G, we use V(G), E(G), F(G), Δ(G), δ(G), g(G) and dG(u,v) to denote its vertex set, edge set, face set, maximum degree, minimum degree, girth and the distance between u and v in graph G, respectively. For a vertex vV(G), we use k (k+ or k)-vertex to denote a vertex of degree k (at least k or at most k). A k (k+ or k)-face is defined similarly. A k-neighbor of v is a k-vertex adjacent to v. For fF(G), we use b(f) to denote the boundary walk of f and write f=[x1x2xk] if x1,x2,,xk are the vertices of b(f) in the clockwise order.

    An injective k-coloring of a graph G is a vertex coloring c:V(G){1,2,,k} such that c(u)c(v) if u and v have a common neighbor. The injective chromatic number of G, denoted by χi(G), is the least integer k such that G has an injective k-coloring.

    The idea of injective coloring was introduced by Hahn et al. [1]. They proved the inequality Δχi(G)Δ2Δ+1 for any planar graph G with maximum degree Δ. For planar graph G with g(G)4, there are fewer results, Bu et al. [2] proved that χi(G)Δ+6 if Δ20 and 4-cycle and 4-cycle are disjoint. For planar graph G with g(G)5, Bu and Lu [3] proved that χi(G)Δ+7; then Dong and Lin [4] improved this result to χi(G)Δ+6; Lu˘zar et al [5] proved that χi(G)Δ+4 if Δ439; recently, Bu and Huang [6] showed that χi(G)Δ+4 if Δ11; Bu and Ye [7] improved this upper bound to Δ+3 when Δ20. For planar graph G with g(G)6, Dong and Lin [8] proved that χi(G)Δ+2 if Δ9 and χi(G)Δ+1 if Δ17. In [9], Borodin et al proved that for every planar graph G, χi(G)=Δ in each of the following cases (i-iv): (i) g(G)=7 and Δ16; (ii) 8g(G)9 and Δ10; (iii) 10g(G)11 and Δ6; (iv) g(G)13 and Δ=5.

    In this paper, we consider the injective chromatic number of planar graph G with g(G)5 and prove the following theorem.

    Theorem 1.1. If G is a planar graph with g(G)5, Δ(G)20 and without adjacent 5-cycles, then χi(G)Δ(G)+2.

    Let G be a graph. If G can not admit any injective coloring with k colors, but any subgraph of G can, then we call G is injective k-critical. In this section, we assume G is injective k-critical. We give some structural properties of G.

    For convenience, we give out some notations. For a k-vertex v, let N(v)={v1,v2,,vk} with d(v1)d(v2)d(vk), D(v)=ki=1d(vi), N2(v)=1ik(N(vi){v}). We use nk(v) to denote the number k-vertices in N(v). If d(v1)=2, then let N(v1)={v,v1}. A 2-vertex or 4+-vertex v of G is called a heavy vertex or is heavy if D(v)Δ+2+d(v), otherwise v is called a light vertex or is light. A 3-vertex v of G is called a heavy vertex or is heavy if D(v)Δ+2+d(v) and n2(v)=0. Conversely, a 3-vertex v is called a light vertex or is light if D(v)Δ+1+d(v) and n2(v)=0. We use nlk(v) and nhk(v) to denote the number of adjacent light and heavy k-vertices of v, respectively. For a partial vertex coloring c of G, we use F(v) to denote the forbidden colors for v. Let C={1,2,,Δ+2} be a color set. For integers k and n, a k(n)-vertex is a k-vertex adjacent to n2-vertices. Now, we discuss the structures of G.

    Lemma 2.1. Let uvE(G). If D(u)Δ+1+d(u), then D(v)Δ+2+d(v).

    Proof. By contradiction, suppose D(v)Δ+1+d(v). By the minimality of G, Guv admits an injective coloring with Δ+2 colors. Erase the colors on u and v. Since D(u)Δ+1+d(u) and D(v)Δ+1+d(v), we have |F(u)|Δ+1, |F(v)|Δ+1. Let c(u)CF(u) and c(v)CF(v), then G has an injective (Δ+2)-coloring, a contradiction.

    Based on Lemma 2.1, we can obtain the following two lemmas.

    Lemma 2.2. δ(G)2 and G contains no adjacent 2-vertices.

    Lemma 2.3. Let v be a 3(1)-vertex of G with N(v)={v1,v2,v3}, where d(v1)=2, then d(v2)+d(v3)Δ+3.

    Lemma 2.4. G contains no adjacent 3(1)-vertices.

    Proof. By contradiction, suppose that there exists two adjacent 3(1)-vertices u and v. Let u1N(u) and v1N(v) be two 2-vertices. By the minimality of G, Guu1 admits an injective coloring with Δ+2 colors. We erase the colors on u, v1 and u1. First, we can color u since |F(u)|Δ+1. Then |F(v1)|Δ+1, |F(u1)|Δ+1, so we can color v1 and u1 in turn to get an injective (Δ+2)-coloring of G, a contradiction.

    Lemma 2.5. If f=[uv1vv2x], v is a light 3-vertex, v1 and v2 are 3(1)-vertices, d(u)=2, then f is a 6+-face.

    Proof. Suppose that the lemma is not true. Assume that f=[uv1vv2x] is a 5-face, then d(x)=Δ by Lemmas 2.1 and 2.2. By the minimality of G, Gvv2 admits an injective coloring with Δ+2 colors. We erase the colors on v, v2 and u. Since |F(v2)|Δ2+1+2=Δ+1 and |F(u)|Δ1+1=Δ, we can color v2 and u. Finally, |F(v)|Δ+1 since v is a light 3-vertex, we can color v, a contradiction.

    In this section, we prove Theorem 1.1 by contradiction. Let G be a minimal counterexample of Theorem 1.1.

    Applying Eulers formula |V|+|F||E|=2 and the fact vV(G)d(v)=fF(G)d(f)=2|E(G)| for a plane graph, we have

    vV(G)(32d(v)5)+fF(G)(d(f)5)=10.

    Design a weight function ω(x) such that ω(x)=32d(x)5 for each xV(G), ω(x)=d(x)5 for each xF(G). Hence, xV(G)F(G)ω(x)=10. Next, we shall transfer weight. Our discharging procedure has two steps. Now, let's look at the structural properties of G, while keeping the total weight sum constant, we obtain a new weight w(x) for all xVF by transferring weights after the first step. We shall prove that w(x)0 for each xVF after the second step and then get the following contradiction:

    0xV(G)F(G)ω(x)=xV(G)F(G)ω(x)=10.

    This contradiction shows that G does not exist and thus Theorem 1.1 is true.

    Thefirststep

    We define the following discharging rules.

    R10. Every vertex v with d(v)10 sends 325d(v) to each adjacent 9-vertex.

    R11. Let v be a 2-vertex with N(v)={v1,v2}. If d(v1)d(v2)9, then vi(i=1,2) sends 1112 to v. If d(v1)9<d(v2), then v1 sends 13+5d(v2) to v.

    R12. Let v be a 3(1)-vertex with N(v)={v1,v2,v3}. If d(v1)=2, 3d(v2)9, and d(v3)20, then v2 sends 5d(v3)14 to v.

    R13. Every heavy vertex v with 3d(v)9 sends 13 to each adjacent light 3-vertex.

    R14. Each of 8-vertex and 9-vertex sends 16 to every adjacent heavy 3-vertex.

    R15. Every 6+-face sends d(f)5d(f) to each incident 9-vertex.

    Let ω(x) be the new weight of each xVF by applying the above rules. Let v be a k-vertex. Note that k2 by Lemma 2.2. By R10, every 10+-vertex sends at least 1 to each adjacent 9-vertex.

    (1) k=2, w(v)=2.

    If d(v1)d(v2)9, then vi(i=1,2) sends 1112 to v. If d(v1)9<d(v2), then v1 sends 13+5d(v2) to v, v2 sends 325d(v2) to v by R10 and R11. If 10d(v1)d(v2), then vi sends 325d(vi)(i=1,2) to v by R10. By R15, 6+-face sends at least 16 to v. Hence,

    ω(v)2+min{1112+1112,13+5d(v2)+325d(v2),325d(v1)+325d(v2)}+16=0.

    (2) k=3, ω(v)=12.

    Case 1. Suppose that v is a 3(1)-vertex. By Lemma 2.3, d(v2)+d(v3)Δ+323. Note that v is adjacent to at least a 12+-vertex and d(v2)3. If 3d(v2)9, then d(v3)Δ6. By R10, R11, R12 and R15, ω(v)121112+16×2+5d(v3)14+325d(v3)=16 when d(v3)20; ω(v)121112+16×2+325d(v3)528 when d(v3)21. If d(v3)d(v2)10, then ω(v)121112+325d(v2)+325d(v3)+16×2>0 by R10, R11 and R15.

    Case 2. Supose that v is a heavy 3-vertex. Note that D(v)Δ+2+325, v is a (8, 8, 9)-vertex or (d1, 9, 9)-vertex(7d19) or (d1, d2, 10+)-vertex(d1d210). By R13, a heavy 3-vertex sends at most 13 to a light 3-vertex. Hence, ω(v)12+min{16×3,16×2,113×2}+16×2>0 by R10, R14 and R15.

    Case 3. Suppose that v is a light 3-vertex. If v is adjacent to a 3(1)-vertex u, then u is adjacent to a Δ-vertex by Lemma 2.3. If Δ=20, then v sends 5Δ14=0 to u. Otherwise, v sends nothing to u. If v is adjacent to at least one 10+-vertex, then ω(v)12(5Δ14)+1+16×2>0 by R10, R12 and R15. If v is adjacent to at most two 3(1)-vertices and not adjacent to 10+-vertex, then ω(v)12+13+16×2>0 by R12, R13 and R15. Finally, v is adjacent to three 3(1)-vertices. If v is incident with three 6+-faces or a 5-face, two 7+-faces or a 5-face, at least a 8+-face, then ω(v)12+min{16×3,27×2,16+38}>0 by R15. If v is incident with a 5-face, two 6-faces or a 5-face, a 6-face and a 7-face, then ω(v)=16 or ω(v)=112 by R15.

    (3) k=4, ω(v)=1.

    Case 1. n2(v)=0. By Lemma 2.3, the 3(1)-neighbor u of v is adjacent to a (Δ1)+-vertex. If u is adjacent to a 21+-vertex, then v sends 0 to u. Suppose v is adjacent to a l-vertex with 19l20. By R12, v sends at most 51914 to 3(1)-vertex. If D(v)Δ+1+4, then v is light. Therefore, ω(v)14×(51914)+16×2>0 by R15. If D(v)Δ+6, then v is heavy. If nl3(v)=1, then ω(v)12×(51914)13+16×2>0 by R12, R13 and R15. If nl3(v)2, then d(v3)+d(v4)Δ. Hence, v4 is a 10+-vertex. This implies that ω(v)113×3+(325d(v4))+16×2>0 by R10, R13 and R15.

    Case 2. n2(v)=1. If D(v)Δ+6, then v is heavy. If n3(v)=0, then v only sends weight to 2-vertex. By R10, R11 and R15, ω(v)11112+16×2>0. If n3(v)=1, then d(v3)+d(v4)Δ+1. Therefore, v4 is a 11+-vertex. It follows from R10, R11, R13 and R15 that ω(v)1111213+(325d(v4))+16×2>0. If n3(v)=2, then d(v4)Δ2. This together with R10, R11, R13 and R15 implies that ω(v)1111213×2+(325Δ2)+16×2>0. Otherwise, D(v)Δ+5, then v is light and the vertices adjacent to v are all heavy. So the 2-neighbor u of v is adjacent to a Δ-vertex and the 3(1)-neighbor of v is adjacent to a (Δ1)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. This implies that ω(v)1(13+5Δ)3×(5Δ114)+16×2>0 by R11, R12 and R15.

    Case 3. n2(v)=2. If D(v)Δ+6, then v is heavy. It follows from Lemma 2.1 that d(v3)+d(v4)Δ+2, which means that v4 is a 11+-vertex. Therefore, ω(v)11112×2+min{13+325Δ1,32511}+16×2>0 by R10, R11, R13 and R15. Otherwise, D(v)Δ+5, then v is light and the vertices adjacent to v are all heavy. Hence, the 2-neighbor of v is adjacent to a Δ-vertex and the 3(1)-neighbor of v is adjacent to a (Δ1)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. This together with R11, R12 and R15 yields that ω(v)1(13+5Δ)×2(5Δ114)×2+16×2>0.

    Case 4. n2(v)=3. If D(v)Δ+6, then v is heavy and d(v4)=Δ. If the 2-vertices adjacent to v are all light and v is incident with at most a 5-face, then ω(v)1+(325Δ)1112×3+16×30 by R10, R11 and R15. If the 2-vertices adjacent to v are all light and v is incident with two 5-faces, then ω(v)1+325Δ1112×3+16×216 by R10, R11 and R15. If the 2-vertices adjacent to v are all heavy, then ω(v)1+(325Δ)(13+5Δ)×3+16×256 by R10, R11 and R15. If the 2-vertices adjacent to v are not all light, let v1 be a light 2-vertex, then we have that ω(v)1+(325Δ)(13+5Δ)1112×2+16×2>0 by R10, R11 and R15. Otherwise, D(v)Δ+5, then v is light and the vertices adjacent to v are all heavy. Hence, the 2-neighbor of v is adjacent to a Δ-vertex and the 3(1)-neighbor of v is adjacent to a (Δ1)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. This together with R11, R12 and R15 yields that ω(v)1(13+5Δ)×3(5Δ114)+16×249114.

    Case 5. n2(v)=4. Clearly, D(v)Δ+5, which implies that the 2-neighbor of v is adjacent to a Δ-vertex. By R11 and R15, ω(v)1(13+5Δ)×4+16×21.

    (4) k=5, ω(v)=52.

    Case 1. n2(v)2. It follows from R11, R13 and R15 that ω(v)521112×213×3+16×3>0.

    Case 2. n2(v)=3. If D(v)Δ+7, then v is heavy and d(v4)+d(v5)Δ+1. Therefore, v5 is a 11+-vertex. This means that ω(v)521112×3+min{13+(325Δ2),32511}+16×3>0 by R10, R11, R13 and R15. Otherwise, D(v)Δ+6, then v is light and the vertices adjacent to v are all heavy. So the 2-neighbor of v is adjacent to a (Δ1)+-vertex and the 3(1)-neighbor of v is adjacent to a (Δ2)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. By R11, R12 and R15, ω(v)52(13+5Δ1)×3(5Δ214)×2+16×3>0.

    Case 3. n2(v)=4. If D(v)Δ+7, then v is heavy and d(v5)Δ1. By R10, R11 and R15, ω(v)521112×4+325Δ1+16×3>0. Otherwise, D(v)Δ+6, then v is light and the vertices adjacent to v are all heavy. Hence, the 2-neighbor of v is adjacent to a (Δ1)+-vertex and the 3(1)-neighbor of v is adjacent to a (Δ2)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. By R11, R12 and R15, ω(v)52(13+5Δ1)×4(5Δ214)+16×3>0.

    Case 4. n2(v)=5. Clearly, D(v)Δ+6, which means that the 2-neighbor of v is adjacent to a (Δ1)+-vertex. So ω(v)52(13+5Δ1)×5+16×3>0 by R11 and R15.

    (5) k=6, ω(v)=4.

    Case 1. n2(v)4. It is clear that ω(v)41112×413×2+16×3>0 by R11, R13 and R15.

    Case 2. n2(v)=5. If D(v)Δ+8, then v is heavy and d(v6)Δ2. This together with R10, R11 and R15 implies that ω(v)41112×5+325Δ2+16×3>0. Otherwise, D(v)Δ+7, then v is light and the vertices adjacent to v are all heavy. Therefore, the 2-neighbor of v is adjacent to a (Δ2)+-vertex and the 3(1)-neighbor of v is adjacent to a (Δ3)+-vertex. If the 3(1)-neighbor of v is adjacent to a 21+-vertex, then v sends 0 to this 3(1)-vertex. By R11, R12 and R15, ω(v)4(13+5Δ2)×5(5Δ314)+16×3>0.

    Case 3. n2(v)=6. It is easy to see that D(v)Δ+7, which implies that the 2-neighbor of v is adjacent to a (Δ2)+-vertex. By R11 and R15, ω(v)4(13+5Δ2)×6+16×3>0.

    (6) k=7, ω(v)=112.

    Case 1. n2(v)6. By R11, R13 and R15, ω(v)1121112×613+16×4>0.

    Case 2. n2(v)=7. It is clear that D(v)Δ+7, which means that the 2-neighbor of v is adjacent to a (Δ3)+-vertex. By R11 and R15, ω(v)112(13+5Δ3)×7+16×4>0.

    (7) k=8, ω(v)=7. Observe that ω(v)71112×8+16×4>0 by R11, R13 and R15.

    (8) k=9, ω(v)=172. By R11, R13 and R15, ω(v)1721112×9+16×5>0.

    (9) k10, ω(v)=32k5. By R10, ω(v)32k5(325k)×k=0.

    After the first step, ω(x)0 for each xVF except some 3-vertices and 4-vertices. For fF(G), if d(f)5, then ω(f)min{0,d(f)5d(f)5d(f)d(f)}=0 by R15. For convenience, a vertex v is called bad if ω(v)<0. If uvE(G) and u, v are all 10+-vertices, then uv is called special edge.

    There are four bad vertices:

    I-vertex: 3-vertex v is adjacent to three 3(1)-vertices and incident with a 5-face, two 6-faces or a 5-face, a 6-face and a 7-face, ω(v)16.

    II-vertex: 4-vertex v is adjacent to three light 2-vertices, a Δ-vertex and incident with two 5-faces, ω(v)16.

    III-vertex: light 4-vertex v is adjacent to three 2-vertices, ω(v)49114.

    IV-vertex: 4-vertex v is adjacent to four 2-vertices, ω(v)1.

    Obviously, bad vertex is not adjacent to bad vertex.

    Thesecondstep

    R20. Every 10+-vertex v sends 12 to incident face f through its special edge.

    R21. Every 9-vertex sends its remained positive weight averagely to each incident face.

    R22. Every 5+-face sends its remained positive weight averagely to each incident bad vertex.

    (1) v is I-vertex, see Figure 1a. Clearly, v is a light 3-vertex. By Lemma 2.1, D(vi)Δ+5, which means that d(vi)=Δ for 1i3.

    Figure 1.  The configurations of bad vertices. (The degrees of black nodes are the actual degrees in the figure).

    Let f1=v1vv2xy be a 5-face with d(x),d(y){2,Δ}, f2 and f3 be 6+-faces. By Lemma 2.2, at least one of x and y is not 2-vertex. Note that {d(x),d(y)}{2,Δ} by Lemma 2.5. Hence, d(x)=d(y)=Δ, which means that v1v2 is a special edge. By R20, f1 receives 12×2 from v1,v2. Obviously, f1 is only incident with a bad vertex v, which together with R22 shows that ω(v)16+1>0.

    (2) v is II-vertex, see Figure 1b. By Lemma 2.1, D(vi)Δ+2+d(vi). Let f1 and f3 be 5-faces.

    Case 1. If there is a bad vertex in {v1,v2,v3}, then it can not be I, III, IV-vertex. If v1 is a bad vertex, then v2 is a Δ-vertex by Lemma 2.2, which means that v2 is a heavy 2-vertex, a contradiction. Hence, v1 can not be a bad vertex. Similarly, v2 can not be a bad vertex. Next, suppose that v3 is a bad vertex, see Figure 2a. If u is a 2-vertex, then u is a heavy 2-vertex. This is contradict with II-vertex v3 only adjacent to light 2-vertex. If u is a Δ-vertex, then uv4 is a special edge. By R20, f3 gets 1 from uv4. Obviously, f3 is only incident with two bad vertices v and v3, which implies that ω(v)16+12>0 by R22.

    Figure 2.  II-vertex.

    Case 2. If v1,v2,v3 are not bad vertices, then f4 is incident with at most d(f4)32+1 bad vertices.

    Case 2.1. Suppose that f4 is a 7+-face. Since d(v4)=Δ, we can get ω(f4)d(f4)5d(f4). By the first step and R22, ω(v)1+325Δ1112×3+d(f4)5d(f4)+16+d(f4)5d(f4)×1d(f4)32+1>0.

    Case 2.2. Suppose that f2 is a 7+-face. Since d(v4)=Δ, we can get ω(f4)d(f4)5d(f4). By the first step and R22, ω(v)1+325Δ1112×3+d(f4)5d(f4)+27+d(f4)5d(f4)×1d(f4)32+1>0.

    Case 2.3. Suppose that f2 and f4 are all 6-faces, see Figure 2b. It is easy to see that f2 and f3 are only incident with a bad vertex v, f4 is incident with at most two bad vertices.

    If v1 is a 10+-vertex, then ω(f4)16×2 by d(v4)=Δ. By R22, ω(v)16+16×12>0. If v2 is a 10+-vertex, then ω(f2)16×2. By d(v4)=Δ, ω(f4)16. This means that ω(v)16+16×12+16>0 by R22. If v3 is a 10+-vertex, then ω(f2)16. By d(v4)=Δ, ω(f4)16, which implies that ω(v)16+16×12+16>0.

    If v1,v2,v3 are all 9-vertices, then we discuss the classification of d(x). We can obtain that ω(f4)16 since d(v4)=Δ.

    If d(x)10, then xv4 is a special edge. By R20, each of x and v4 sends 12 to f3. Hence, ω(v)16+16×12+1>0 by R22. If 5d(x)9, then ω(x)32d(x)5+(325Δ)1112(d(x)1)+16×d(x)223d(x)176 by R10, R11 and R15. This means that f3 gets (23d(x)176)1d(x)110 by R21. Hence, ω(v)16+16×12+110>0 by R22.

    If d(x)=4, then ω(x)1+(325Δ)1112×2(5Δ114)+16×2=1419 by R10, R11, R12 and R15. This implies that f3 gets 738 by R21. Hence, ω(v)16+16×12+738>0 by R22. If d(x)=3, then ω(x)12+(325Δ)1112+16×2=16 by R10, R11, R12 and R15. This means that f3 gets 118 by R21. Similarly, we consider the degree of u. If d(u)3, then f4 gets at least min{1,110,118}=118. It follows from R22 that ω(v)16+16×12+118+118×12=0. If d(u)=2, then f4 is only incident with a bad vertex v. By R22, ω(v)16+16=0.

    (3) v is III-vertex, see Figure 1c. Observe that d(vi)=Δ (i = 1, 2, 3) by Lemma 2.1.

    Suppose that f1 or f2 is a 5-face. Without loss of generality, let f1 be a 5-face, then v1v2 is a special edge. By R20, f1 gets at least 1. Clearly, f1 is only incident with a bad vertex v. This together with R22 implies that ω(v)49114+1>0. Next, we consider the case when f1 and f2 are 6+-faces.

    Suppose that f3 or f4 is a 5-face. Without loss of generality, let f4 be a 5-face, then f3 is a 6+-face. After the first step, ω(v)1(13+5Δ)×3(5Δ114)+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3) (observe that v4 is adjacent to a (Δ1)+-vertex if v4 is a 3(1)-vertex). Since d(vi)=Δ for 1i3, ω(fj)2(d(fj)5)d(fj) for 1j2, ω(f3)d(f3)5d(f3). It is easy to see that fj is incident with at most d(fj)42+1 bad vertices for 1j2, f3 is incident with at most d(f3)32+1 bad vertices. Claim that d(fi)6, d(fi)5d(fi)+2(d(fi)5)d(fi)×1d(fi)42+113 for i=1,2; d(f3)5d(f3)+d(f3)5d(f3)×1d(f3)32+114. By R22,

    ω(v)1(13+5Δ)×3(5Δ114)+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3)
    +2(d(f1)5)d(f1)×1d(f1)42+1+2(d(f2)5)d(f2)×1d(f2)42+1+d(f3)5d(f3)×1d(f3)32+1>0.

    Suppose d(fi)6 for 1i4. After the first step,

    ω(v)1(13+5Δ)×3(5Δ114)+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3)+d(f4)5d(f4).

    Since d(vi)=Δ for 1i3, we can get ω(fi)2(d(fi)5)d(fi) for 1i2, ω(fk)d(fk)5d(fk) for 3k4. It is clear that fi is incident with at most d(fi)42+1 bad vertices (i = 1, 2), fk is incident with at most d(fk)32+1 bad vertices (k = 3, 4). By R22,

    ω(v)1(13+5Δ)×3(5Δ114)+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3)+d(f4)5d(f4)+2(d(f1)5)d(f1)
    ×1d(f1)42+1+2(d(f2)5)d(f2)×1d(f2)42+1+d(f3)5d(f3)×1d(f3)32+1+d(f4)5d(f4)×1d(f4)32+1>0.

    (4) v is IV-vertex, see Figure 1d. By Lemma 2.1, d(vi)=Δ for 1i4.

    There is exactly one 5-face in f1, f2, f3 and f4. Without loss of generality, let f1 be a 5-face, then v1v2 is a special edge. By R20, f1 gets at least 1. It is easy to see that f1 is only incident with a bad vertex v. By R22, ω(v)1+1=0. Next, we consider fi (1i4) is 6+-face.

    Suppose v is incident with at least a 6-face. Without loss of generality, let f1 be a 6-face. It is clear that f1 is only incident with a bad vertex v. Since d(v1)=d(v2)=d(v3)=Δ, we can get ω(f1)16. We use m6 to denote the number of 6-faces which is incident with v. After the first step and the second step, ω(v)1(13+5Δ)×4+16×m6+27×(4m6)+26×m6=421+314m6521>0.

    Suppose v is incident with four 7+-faces. Note that fi is incident with at most d(fi)42+1(1i4) bad vertices. Since d(vi)=Δ, we can get ω(fi)2(d(fi)5)d(fi) (1i4). After the first step,

    ω(v)1(13+5Δ)×4+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3)+d(f4)5d(f4).

    By R22,

    ω(v)1(13+5Δ)×4+d(f1)5d(f1)+d(f2)5d(f2)+d(f3)5d(f3)+d(f4)5d(f4)+2(d(f1)5)d(f1)×1d(f1)42+1
    +2(d(f2)5)d(f2)×1d(f2)42+1+2(d(f3)5)d(f3)×1d(f3)42+1+2(d(f4)5)d(f4)×1d(f4)42+1>0.

    Now we have checked that the final weight w(x)0 for each xVF. Then,

    0xVFw(x)=xVFw(x)=10,

    which is a contradiction.

    In this paper, we consider the injective chromatic index of planar graphs without adjacent 5-cycles and proved that such graphs have χi(G)Δ(G)+2 if g(G)5 and Δ(G)20. A natural problem in context of our main result is the following: What is the optimal constant c such that χi(G)Δ(G)+2 for every planar graph G with g(G)5 and Δ(G)c.

    This work was supported by National Natural Science Foundations of China (Grant Nos. 11771403, 11871439, 11901243 and 12201569)

    The authors declare no conflicts of interest.



    [1] G. Hahn, J. Kratochvíl, J. ˇSirˊaˇn, D. Sotteau, On the injective chromatic number of graphs, Discrete Math., 256 (2002), 179–192. https://doi.org/10.1016/S0012-365X(01)00466-6 doi: 10.1016/S0012-365X(01)00466-6
    [2] Y. Bu, C. Qi, J. Zhu, T. Xu, Injective coloring of planar graphs, Theoret. Comput. Sci., 857 (2021), 114–122. https://doi.org/10.1016/j.tcs.2021.01.007 doi: 10.1016/j.tcs.2021.01.007
    [3] Y. Bu, K. Lu, List injective coloring of planar graphs with girth 5, 6, 8, Discrete Appl. Math., 161 (2013), 1367–1377. https://doi.org/10.1016/j.dam.2012.12.017 doi: 10.1016/j.dam.2012.12.017
    [4] W. Dong, W. Lin, Injective coloring of planar graphs with girth 5, Discrete Math., 315 (2014), 120–127. https://doi.org/10.1016/j.disc.2013.10.017 doi: 10.1016/j.disc.2013.10.017
    [5] B. Lu˘zar, R. ˘Skrekovski, M. Tancer, Injective colorings of planar graphs with few colors, Discrete Math., 309 (2009), 5636–5649. https://doi.org/10.1016/j.disc.2008.04.005 doi: 10.1016/j.disc.2008.04.005
    [6] Y. Bu, C. Huang, List injective coloring of a class of planar graphs without short cycles, Discrete Math. Algorithms Appl., 10 (2018), 1850068. https://doi.org/10.1142/S1793830918500684 doi: 10.1142/S1793830918500684
    [7] Y. Bu, P. Ye, Upper bounds for the injective coloring of planar graphs with girth at least five, Adv. Math., 47 (2018), 363–372.
    [8] W. Dong, W. Lin, Injective coloring of planar graphs with girth 6, Discrete Math., 313 (2013), 1302–1311. https://doi.org/10.1016/j.disc.2013.02.014 doi: 10.1016/j.disc.2013.02.014
    [9] O. Borodin, A. Ivanova, List injective colorings of planar graphs, Discrete Math., 311 (2011), 154–165. https://doi.org/10.1016/j.disc.2010.10.008 doi: 10.1016/j.disc.2010.10.008
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1316) PDF downloads(69) Cited by(0)

Figures and Tables

Figures(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog