Research article Special Issues

On the general atom-bond sum-connectivity index

  • This paper is concerned with a generalization of the atom-bond sum-connectivity (ABS) index, devised recently in [A. Ali, B. Furtula, I. Redžepović, I. Gutman, Atom-bond sum-connectivity index, J. Math. Chem., 60 (2022), 2081-2093]. For a connected graph G with an order greater than 2, the general atom-bond sum-connectivity index is represented as ABSγ(G) and is defined as the sum of the quantities (12(dx+dy)1)γ over all edges xy of the graph G, where dx and dy represent the degrees of the vertices x and y of G, respectively, and γ is any real number. For 10γ10, the significance of ABSγ is examined on the data set of octane isomers for predicting six selected physicochemical properties of the mentioned compounds; promising results are obtained when the approximated value of γ belongs to the set {3,1,3}. The effect of the addition of an edge between two non-adjacent vertices of a graph under ABSγ is also investigated. Moreover, the graphs possessing the maximum value of ABSγ, with γ>0, are characterized from the set of all connected graphs of a fixed order and a fixed (ⅰ) vertex connectivity not greater than a given number or (ⅱ) matching number.

    Citation: Abeer M. Albalahi, Zhibin Du, Akbar Ali. On the general atom-bond sum-connectivity index[J]. AIMS Mathematics, 2023, 8(10): 23771-23785. doi: 10.3934/math.20231210

    Related Papers:

    [1] Zhen Wang, Kai Zhou . On the maximum atom-bond sum-connectivity index of unicyclic graphs with given diameter. AIMS Mathematics, 2024, 9(8): 22239-22250. doi: 10.3934/math.20241082
    [2] Usman Babar, Haidar Ali, Shahid Hussain Arshad, Umber Sheikh . Multiplicative topological properties of graphs derived from honeycomb structure. AIMS Mathematics, 2020, 5(2): 1562-1587. doi: 10.3934/math.2020107
    [3] Ali N. A. Koam, Ali Ahmad, Azeem Haider, Moin A. Ansari . Computation of eccentric topological indices of zero-divisor graphs based on their edges. AIMS Mathematics, 2022, 7(7): 11509-11518. doi: 10.3934/math.2022641
    [4] Tariq A. Alraqad, Igor Ž. Milovanović, Hicham Saber, Akbar Ali, Jaya P. Mazorodze, Adel A. Attiya . Minimum atom-bond sum-connectivity index of trees with a fixed order and/or number of pendent vertices. AIMS Mathematics, 2024, 9(2): 3707-3721. doi: 10.3934/math.2024182
    [5] Alaa Altassan, Muhammad Imran, Bilal Ahmad Rather . On $ ABC $ energy and its application to anticancer drugs. AIMS Mathematics, 2023, 8(9): 21668-21682. doi: 10.3934/math.20231105
    [6] Fei Yu, Hifza Iqbal, Saira Munir, Jia Bao Liu . M-polynomial and topological indices of some transformed networks. AIMS Mathematics, 2021, 6(12): 13887-13906. doi: 10.3934/math.2021804
    [7] Xiaodi Song, Jianping Li, Jianbin Zhang, Weihua He . Trees with the second-minimal ABC energy. AIMS Mathematics, 2022, 7(10): 18323-18333. doi: 10.3934/math.20221009
    [8] Ali Al Khabyah . Mathematical aspects and topological properties of two chemical networks. AIMS Mathematics, 2023, 8(2): 4666-4681. doi: 10.3934/math.2023230
    [9] Jianwei Du, Xiaoling Sun . On the graph connectivity and the variable sum exdeg index. AIMS Mathematics, 2021, 6(1): 607-622. doi: 10.3934/math.2021037
    [10] Sumaira Hafeez, Rashid Farooq . On generalized inverse sum indeg index and energy of graphs. AIMS Mathematics, 2020, 5(3): 2388-2411. doi: 10.3934/math.2020158
  • This paper is concerned with a generalization of the atom-bond sum-connectivity (ABS) index, devised recently in [A. Ali, B. Furtula, I. Redžepović, I. Gutman, Atom-bond sum-connectivity index, J. Math. Chem., 60 (2022), 2081-2093]. For a connected graph G with an order greater than 2, the general atom-bond sum-connectivity index is represented as ABSγ(G) and is defined as the sum of the quantities (12(dx+dy)1)γ over all edges xy of the graph G, where dx and dy represent the degrees of the vertices x and y of G, respectively, and γ is any real number. For 10γ10, the significance of ABSγ is examined on the data set of octane isomers for predicting six selected physicochemical properties of the mentioned compounds; promising results are obtained when the approximated value of γ belongs to the set {3,1,3}. The effect of the addition of an edge between two non-adjacent vertices of a graph under ABSγ is also investigated. Moreover, the graphs possessing the maximum value of ABSγ, with γ>0, are characterized from the set of all connected graphs of a fixed order and a fixed (ⅰ) vertex connectivity not greater than a given number or (ⅱ) matching number.



    A property of a graph that is preserved by the graph isomorphism is called a graph invariant [9]. Topological indices are widely used to refer to real-valued graph invariants. We refer the reader to the books [5,9,23] for (chemical-)graph-theory terminology and notations.

    Among the much-investigated and applied topological indices, the connectivity index [10] (also known as the Randić index, which was invented in [19] under the name branching index) has secured a prominent place. According to [10], the connectivity index is presumed to be the topological index that has been predominantly examined, both theoretically and practically. The following number associated with a graph G is the connectivity index of G:

    R(G)=stE(G)1dsdt,

    where E(G) represents the edge set of G and dv denotes the degree of a vertex v in G. (If two or more graphs are being considered at a time, then we use dv(G) to represent the degree of v in G to avoid confusion.) The survey papers [13,18], books [11,12] and related articles referred therein provide further information regarding the investigation of the connectivity index.

    The popularity of the connectivity index has led to the introduction of several variants of this index in the literature. Among the extensively researched variants of the connectivity index are the sum-connectivity (SC) index [24] and the atom-bond connectivity (ABC) index [6,7] (see also [20,21]), which are defined by

    SC(G)=stE(G)1ds+dt

    and

    ABC(G)=stE(G)ds+dt2dsdt,

    respectively. A new version of the ABC index, known as the atom-bond sum-connectivity (ABS) index, was developed in [2] using the basic concept of the SC index. For a graph G, the ABS index is defined as follows:

    ABS(G)=stE(G)12ds+dt.

    Some extremal problems regarding the ABS index of (molecular) tree graphs and general graphs were solved in [2]. In [3], not only was an extremal problem regarding the ABS index for unicyclic graphs solved, but also chemical applications of the ABS index were reported. The trees, with a given number of vertices of degree 1 and with a given order, having the least ABS index were independently investigated in preprints [4,16]; the maximal version of this problem was solved recently in [17].

    The general ABS index [3] is defined by

    ABSγ(G)=stE(G)(12ds+dt)γ,

    where γ may take any real number, provided that if γ<0, then the graph G satisfies the inequality ds+dt>2 for every stE(G). We make a note here that the general ABS index (and hence the ABS index) is a particular form of a more general topological index of this kind, as introduced in [22]. For a recent study on the general ABS index, we refer the reader to [1].

    In the next section, the significance of ABSγ is examined on the data set of octane isomers for predicting six selected physicochemical properties of the mentioned compounds for 10γ10; promising results are obtained when the approximated value of γ belongs to the set {3,1,3}. Section 3 is devoted to investigating the effect of the addition of an edge in a non-complete graph under ABSγ, where a non-complete graph is a graph different from the complete graph. In Section 4, we characterize the graphs possessing the maximum value of ABSγ, with γ>0, in the set of all connected graphs of a fixed order and a fixed (i) vertex connectivity not greater than a given number or (ii) matching number.

    In [1], the chemical applicability of the ABSγ, with the constraint 10γ10, was tested on the data set of 25 benzenoid hydrocarbons for predicting their enthalpy of formation; it was concluded that the predictive ability of ABSγ for the considered property of the examined hydrocarbons is comparable to other existing general indices of this kind. In the present section, the significance of ABSγ is examined on the data set of octane isomers for predicting six selected physicochemical properties of the mentioned compounds for 10γ10. These six properties are the following: enthalpy of vaporization, boiling point, acentric factor, enthalpy of formation, entropy and standard enthalpy of vaporization. The experimental data for these selected properties can be found in [15].

    The positive values of the correlation r (between the selected properties of the considered compounds and ABSγ), with γ[10,10], are depicted in Figures 16. The maximum positive values of the correlation r (between the selected properties of the considered compounds and ABSγ), with γ[10,10] and with γ[3,3], are given in Tables 1 and 2, respectively.

    Figure 1.  The positive values of the correlation r (between the boiling point of the considered compounds and ABSγ) when γ[10,10].
    Figure 2.  The positive values of the correlation r (between the entropy of the considered compounds and ABSγ) when γ[10,10].
    Figure 3.  The positive values of the correlation r (between the enthalpy of vaporization of the considered compounds and ABSγ) when γ[10,10].
    Figure 4.  The positive values of the correlation r (between the standard enthalpy of vaporization (DHVAP) of the considered compounds and ABSγ) when γ[10,10].
    Figure 5.  The positive values of the correlation r (between the enthalpy of formation (HFORM) of the considered compounds and ABSγ) when γ[10,10].
    Figure 6.  The positive values of the correlation r (between the acentric factor of the considered compounds and ABSγ) when γ[10,10].
    Table 1.  The maximum positive values of the correlation r, between the selected properties of octane isomers and ABSγ, when γ[10,10].
    r γ
    Boiling point 0.8511 -6.5275
    Enthalpy of vaporization 0.9558 -6.1399
    Standard enthalpy of vaporization 0.9600 -3.3760
    Entropy 0.8903 5.9019
    Acentric factor 0.8801 1.3381
    Enthalpy of formation 0.8679 -4.2677

     | Show Table
    DownLoad: CSV
    Table 2.  The maximum positive values of the correlation r, between the selected properties of octane isomers and ABSγ, when γ[3,3].
    r γ
    Boiling point 0.8478 -3.0000
    Enthalpy of vaporization 0.9529 -3.0000
    Standard enthalpy of vaporization 0.9600 -3.0000
    Entropy 0.8884 3.0000
    Acentric factor 0.8801 1.3381
    Enthalpy of formation 0.8674 -3.0000

     | Show Table
    DownLoad: CSV

    In Tables 1 and 2, corresponding to every listed property, we observe that there is only a slight difference between the two values of r. (For example, in the case of boiling point, the mentioned difference is 0.0033.) However, there is a considerable difference between the corresponding two values of γ. In view of these observations, we conclude that the approximated values of γ concerning any promising results belong to the set {3,1,3}. These findings suggest that the following three particular versions of ABSγ deserve to be examined further:

    (ⅰ) The topological index ABS1 is useful in predicting the acentric factor of octane isomers. It seems to be interesting to note that

    ABS1(G)=stE(G)(12ds+dt)=|E(G)|H(G),

    where H(G) is the harmonic index, first appeared in [8].

    (ⅱ) The topological index ABS3 is useful in predicting the enthalpy of vaporization, boiling point, enthalpy of formation and standard enthalpy of vaporization of octane isomers.

    (ⅲ) The topological index ABS3 is useful in predicting the entropy of octane isomers.

    For a graph G and stE(G), the graph generated from G by inserting the edge st is represented by G+st. In this section, the difference ABSγ(G+st)ABSγ(G) is investigated. We start with the following known result.

    Corollary 3.1. [1] For a graph G, if stE(G) such that max{ds(G),dt(G)}1, then,

    ABSγ(G+st)>ABSγ(G)for γ0.

    Proposition 3.1. For a graph G, if stE(G) such that ds(G)=1 and dt(G)=0, then,

    ABSγ(G+st)>ABSγ(G)for every γ.

    Proof. If γ0, then the required conclusion follows from Corollary 3.1. In the remaining proof, suppose that γ<0. By the definition of ABSγ, G contains no component isomorphic to K2 and it holds that

    ABSγ(G+st)ABSγ(G)=(ds(G)ds(G)+2)γ(ds(G)1ds(G)+1)γ+(13)γ, (3.1)

    where s is the unique vertex adjacent with s in G. Since the degree of s is 1 in G (and γ<0), by the definition of ABSγ we must have ds(G)>1 (for otherwise G contains a component isomorphic to K2). Note that the function ϕγ, with the following definition, is strictly increasing for α>1:

    ϕγ(α)=(αα+2)γ(α1α+1)γ,

    because its derivative function ϕγ is

    ϕγ(α)=2γ(fγ(α+1)fγ(α)),

    where

    fγ(α)=(α1)γ1(α+1)γ+1andfγ(α)=2(γα)(α1)γ2(α+1)γ+2<0,for γ<0 and α>1.

    Consequently,

    (dsds+2)γ(ds1ds+1)γ+(13)γ=ϕγ(ds)+(13)γϕγ(2)+(13)γ>0,

    which together with Eq (3.1) give ABSγ(G+st)ABSγ(G)>0.

    For a graph G and wV(G), let NG(w)={xV(G):xwE(G)}.

    Proposition 3.2. For a graph G, if stE(G) such that ds(G)=dt(G)=1, then,

    ABSγ(G+st)>ABSγ(G)for γ>1.

    Proof. If γ0, then the required conclusion follows from Corollary 3.1. In the following, we assume that 1<γ<0. Take NG(s)={s} and NG(t)={t}. Since ds(G)=1=dt(G) and 1<γ<0, by the definition of ABSγ, we must have min{ds(G),dt(G)}>1 (for otherwise G contains a component isomorphic to K2). Note that the function ϕγ, with the following definition, is strictly increasing for α>1 and γ<0 (see the proof of Proposition 3.1):

    ϕγ(α)=(αα+2)γ(α1α+1)γ.

    Thus, for 1<γ<0, we have

    ABSγ(G+st)ABSγ(G)=(ds(G)ds(G)+2)γ(ds(G)1ds(G)+1)γ+(dt(G)dt(G)+2)γ(dt(G)1dt(G)+1)γ+(12)γ=ϕγ(ds)+ϕγ(dt)+(12)γ2ϕγ(2)+(12)γ>0.

    Although the next result's proof is similar to the proof of Proposition 3.2, we include it here for completeness.

    Proposition 3.3. For a graph G, if stE(G) such that both the vertices s,t have degree 1 (in G) and that one of their unique neighbors has a degree greater than 3 (in G) and the other unique neighbor has a degree greater than 2 (in G), then,

    ABSγ(G+st)>ABSγ(G)for every γ.

    Proof. If γ0, then the required conclusion follows from Corollary 3.1. In the following, we assume γ<0. Take NG(s)={s} and NG(t)={t}. By the given constraints, min{ds(G),dt(G)}>2 and max{ds(G),dt(G)}>3. Note that the function ϕγ, with the following definition, is strictly increasing for α>1 and γ<0 (see the proof of Proposition 3.1):

    ϕγ(α)=(αα+2)γ(α1α+1)γ.

    Thus, for γ<0, we have

    ABSγ(G+st)ABSγ(G)=(ds(G)ds(G)+2)γ(ds(G)1ds(G)+1)γ+(dt(G)dt(G)+2)γ(dt(G)1dt(G)+1)γ+(12)γ=ϕγ(ds)+ϕγ(dt)+(12)γϕγ(3)+ϕγ(4)+(12)γ>0.

    In the current section, we characterize the graphs possessing the maximum value of ABSγ, with γ>0, in the set of all connected graphs of a fixed order and a fixed (ⅰ) vertex connectivity not greater than a given number or (ⅱ) matching number. The following lemma is very crucial in proving the first main result of this section.

    Lemma 4.1. Define the functions Φ1 and Φ2 as

    Φ1(x)=(t+x2t+x1)γx(x1)2

    and

    Φ2(x)=(x+n+t4x+n+t2)γtx

    with the constraint 1xnt2, where γ is a fixed positive real number greater than zero, while t and n are fixed positive integers. Define Φ(x)=Φ1(x)+Φ2(x). If either of the following two conditions holds:

    (i) t2;

    (ii) 2xnt2 and t=1,

    then, the inequality

    Φ(ntx)+Φ(x)Φ(nt1)+Φ(1)

    holds with equality if and only if x=1.

    Proof. First, we assume that t2. The second-derivative function Φ1 of Φ1 is given as

    Φ1(x)=Ψ1(x)2(t+x2)2(t+x1)2(t+x2t+x1)γ,

    where

    Ψ1(x)=γ(2t2(2x1)+2t(x(3x7)+3)+(x1)(x(2x7)+4))+2(t+x2)2(t+x1)2+γ2(x1)x.

    In the expression of Ψ1, note that the coefficient of γ attains its minimum value at t=2 and hence is positive (because t2 and x1). Thus, Ψ1(x)>0 (as γ>0), and therefore Φ1(x)>0. Additionally, the second-derivative function Φ2 of Φ2 is given as

    Φ2(x)=4γtΨ2(x)(n+t+x4)2(n+t+x2)2(n+t+x4n+t+x2)γ,

    where

    Ψ2(x)=n2+n(2t+x6)+t(t+x6)+γx3x+8.

    Here, Ψ2(x)=n+t+γ3>0 because n4, t2 and γ>0. Since x1, we have

    Ψ2(x)Ψ2(1)=(n4)(n1)+t(2n+t5)+γ+1>0.

    Therefore, it holds that Φ2(x)>0. Because Φ(x)=Φ1(x)+Φ2(x), we conclude that the first-derivative function Φ of Φ is strictly increasing. Since xntx, we have

    ddx(Φ(x)+Φ(ntx))=Φ(x)Φ(ntx)0,

    where the equality

    ddx(Φ(x)+Φ(ntx))=0

    holds if and only if x=ntx. Consequently, we deduce that the expression Φ(x)+Φ(ntx) attains its maximum possible value only at x=1.

    The desired inequality when 2xnt2 and t=1 remains to be proved. In what follows, it is assumed that 2xnt2 and t=1. The second-derivative function Φ1 of Φ1 is given as

    Φ1(x)=(x1)γ1xγ12(2x(x1)+γ(2x1)+γ2),

    which is certainly positive (because γ>0 and x2). Additionally, the second-derivative function Φ2 of Φ2 is given as

    Φ2(x)=4γ(n+x3n+x1)γn(n4)+x(n2)+γx+3(n+x3)2(n+x1)2,

    which is positive too because n5, x2 and γ>0. Thus, in the case under consideration, the expression Φ(x)+Φ(n1x) attains its maximum possible value only at x=2. Thus, in order to complete the proof, it is enough to show that

    Φ(2)+Φ(n3)Φ(1)Φ(n2)<0.

    Take Θ(n,γ)=Φ(2)+Φ(n3)Φ(1)Φ(n2). Then, we have

    Θ(n,γ)=[(12)γ(n2n)γ]+(n4)(n3)2[(n4n3)γ(n3n2)γ]+[2(n1n+1)γ(n2)(2n52n3)γ]. (4.1)

    Since

    (n1n+1)γ<(2n52n3)γfor n5 and γ>0,

    it holds that

    2(n1n+1)γ<(n2)(n1n+1)γ<(n2)(2n52n3)γ for n5 and γ>0.

    Additionally, for n5 and γ>0, we have

    (12)γ<(n2n)γ

    and

    (n4n3)γ<(n3n2)γ.

    Therefore, Eq (4.1) yields Θ(n,γ)<0, as desired.

    Theorem 4.1. In the set of all n-order connected graphs having the vertex connectivity at most t (being a positive integer satisfying 1tn2), the graph K(t)n uniquely possesses the largest value of ABSγ for γ>0, where n5 and K(t)n is the graph formed by joining a new vertex (through edges) to exactly t vertices of the complete graph Kn1. The mentioned maximum value is as follows:

    ABSγ(K(t)n)=(n2n1)γt(t1)2+t(t+n3t+n1)γ+(nt2)(nt1)2(n3n2)γ+(nt1)t(2n52n3)γ.

    Proof. First, we consider a positive integer s less than t. Observe that K(t)n can be formed by adding some edge(s) in the graph K(s)n. Thus, by Corollary 3.1, the inequality ABSγ(K(s)n)<ABSγ(K(t)n) holds for γ>0. Consequently, it is adequate to prove the theorem only for the n-order connected graphs having the vertex connectivity t.

    Let G be a graph possessing the largest value of ABSγ in the set of all n-order connected graphs having the vertex connectivity t for γ>0, where n5 and 1tn2. Since the vertex connectivity of G is t, there exists a subset A of the vertex set of G such that |A|=t and GA consists of at least two components, where GA is the graph formed from G by removing all the vertices (and their incident edges) of A. If the graph GA has more than two components, then adding an edge connecting the vertices lying in two different components of GA increases the value of ABSγ(G) (by Corollary 3.1); however, the vertex connectivity of G remains the same, which is antithetical to the maximality of ABSγ(G). Thereby, the graph GA must have only two components; we name them as C1 and C2. Additionally, by Corollary 3.1, the graphs C1, C2 and G[A] are complete, and every vertex of both the components C1,C2, is adjacent to every vertex of the set A in G, where G[A] is the induced subgraph of G formed on the vertices of A. For i{1,2}, let ci be the order of Ci and suppose that c1c2. Then, t+c1+c2=n and c1nt2. Note that the degree of every vertex belonging to Ci is ci1+t in G, where i=1,2. Moreover, the degree of every vertex belonging to A is n1 in G. By utilizing the formula of ABSγ, we obtain the following:

    ABSγ(G)=(t+c12t+c11)γc1(c11)2+(n+t+c14n+t+c12)γc1t+(t+c22t+c21)γc2(c21)2+(n+t+c24n+t+c22)γc2t+(n2n1)γt(t1)2.

    By making use of the definition of the function Φ defined in Lemma 4.1, we obtain the following:

    ABSγ(G)=Φ(c1)+Φ(c2)+(n2n1)γt(t1)2.

    Utilizing the fact that c2=nc1t, we arrive at the following:

    ABSγ(G)=Φ(c1)+Φ(ntc1)+(n2n1)γt(t1)2.

    By making use of Lemma 4.1 and the definition of G, we conclude that

    ABSγ(G)=Φ(c1)+Φ(ntc1)+(n2n1)γt(t1)2=Φ(1)+Φ(nt1)+(n2n1)γt(t1)2,

    which implies that c1=1 and thereby the graph G is isomorphic to K(t)n. Hence,

    ABSγ(K(t)n)=Φ(1)+Φ(nt1)+(n2n1)γt(t1)2.

    or

    ABSγ(K(t)n)=(n2n1)γt(t1)2+t(t+n3t+n1)γ+(nt2)(nt1)2(n3n2)γ+(nt1)t(2n52n3)γ.

    A component C of a graph is said to be an odd component if the order of C is odd. Let CGn,β be the set of all connected n-order graphs having a matching number β, where 1βn/2 (this condition is imposed because the matching number of any n-order connected graph cannot be greater than n/2). Note that the n-order complete graph Kn has the matching number n/2. Thus, by Corollary 3.1, Kn uniquely possesses the largest value of ABSγ over CGn,n/2 for γ>0. Thereby, in the next result, we consider the case when 1βn/21.

    Theorem 4.2. In the set CGn,β, the graph ¯Knβ+Kβ uniquely possesses the largest value of ABSγ for γ>0, where ¯Knβ denotes the complement of the complete graph Knβ, "+" denotes the graph-operation join, 1βn/21 and n5. Additionally, the mentioned maximum value is as follows:

    ABSγ(¯Knβ+Kβ)=(n2n1)ββ(β1)2+(β+n3β+n1)α(nβ)β.

    Proof. Let G be a graph possessing the maximum value of ABSγ in the set CGn,β for γ>0, provided that 1βn/21 and n5. The Tutte-Berge formula (see [14]) confirms that the vertex set V(G) has a subset W such that

    n2β=o(GW)|W|, (4.2)

    where o(GW) is the number of odd components of the graph GW (which is the graph formed by dropping all the vertices (and the edges incident with them) of W from G). Note that

    no(GW)+|W|,

    which together with (4.2) imply that β|W|.

    If |W|=0, then from Eq (4.2), we deduce that either n=2β or n=2β+1 because G is connected. Whether n=2β or n=2β+1, one has n/2=β, which is impossible under the given constraints. Thus, β|W|1, and from Eq (4.2), we get o(GW)3 because βn/21.

    Let G1,,Gq be odd components of the graph GW, where q=o(GW)3. For i=1,,q, suppose that Gi has ri vertices and assume that rqrq1r1. Now, we show that the graph GW does not have any even component. On the contrary, suppose that Gq+1 is an even component of the graph GW. Construct a new graph G from G by adding an edge w1wq+1 such that wq+1V(Gq+1) and w1V(G1). Then, certainly we have β(G)β(G). Moreover, by the Tutte-Berge formula, we have the following:

    o(GW)|W|n2β(G). (4.3)

    Note that o(GW)=o(GW), which together with Eq (4.2) implies that

    o(GW)|W|=n2β(G);

    this last equation and (4.3) yield β(G)β(G). Thus, β(G)=β(G). However, by Corollary 3.1, it holds that ABSγ(G)<ABSγ(G) for γ>0, which is antithetical to the maximality of ABSγ(G). Therefore, the graph GW does not possess any even component.

    By Corollary 3.1, each of the graphs Gi (for i=1,,q) and G[W] is complete, and every vertex of W is adjacent to every vertex of all the graphs G1,,Gq. Therefore, G=(Kr1Krq)+Kp, where 1p=|W|β; o(GW)=q3; qi=1ri+p=n; n2β=qp and each of the numbers r1,r2,,rq is odd, with rqrq1r1.

    Case 1. When rq=1.

    In this case, rq=rq1==r1=1 and p=β. Since n2β=qp, we have G=¯Knβ+Kβ, as desired.

    Case 2. When rq3 and rq1=1.

    In this case, rq1==r1=1. From the equations qi=1ri+p=n and n2β=qp, we get rq=12p+2β (which implies that β>p because rq3). Thus,

    G=(¯Kq1Krq)+Kp=(¯Kn+p2β1K12p+2β)+Kp. (4.4)

    Observe that (¯Kn+p2β1K12p+2β)+Kp is a spanning subgraph of the graph ¯Knβ+Kβ and thus by making use of Corollary 3.1, we deduce that

    ABSγ(G)=ABSγ((¯Kn+p2β1K12p+2β)+Kp)<ABSγ(¯Knβ+Kβ)

    for γ>0. This is antithetical to the maximality of G. Therefore, this case is not possible.

    Case 3. When rq13.

    In this case, rq3 as rqrq1. Take

    Θ=ABSγ((Kr1Krq2K1Krq+rq11)+Kp)ABSγ(G). (4.5)

    After elementary computations, we obtain

    Θ=(p+n3p+n1)γp+(p+rq+rq13p+rq+rq12)γ(rq+rq11)(rq+rq12)2+(p+rq+rq1+n5p+rq+rq1+n3)γ(rq+rq11)p(p+rq12p+rq11)γrq1(rq11)2+(p+rq1+n4p+rq1+n2)γrq1p(p+rq2p+rq1)γrq(rq1)2+(p+rq+n4p+rq+n2)γrqp=Φ(1)+Φ(rq+rq11)Φ(rq1)Φ(rq),

    where the definition of the function Φ is given in Lemma 4.1. Set rq+rq1+β=n. Then rq1nβ2 because rq1rq. Thus,

    Θ=Φ(1)+Φ(nβ1)Φ(rq1)Φ(nβrq1),

    and by Lemma 4.1, the right-handed expression of this equation is greater than 0. Therefore, Eq (4.5) yields a contradiction to the maximality of ABSγ(G). Consequently, this case is also impossible.

    In this paper, the chemical usefulness and several mathematical aspects of the index ABSγ have been considered and studied. For 10γ10, the significance of ABSγ is examined on the data set of octane isomers for predicting six selected physicochemical properties of the mentioned compounds; promising results are obtained when the approximated value of γ belongs to the set {3,1,3}. (The selected six properties are the following: enthalpy of vaporization, boiling point, acentric factor, enthalpy of formation, entropy, and standard enthalpy of vaporization.) The value γ=1 corresponds to the index that can be written in the form of an existing index, namely the harmonic index. The findings of Section 2 indicate that the topological indices ABS1 and ABS3 are useful in predicting the acentric factor and the entropy of octane isomers, respectively; the index ABS3 can be utilized to predict the remaining four selected properties (enthalpy of vaporization, boiling point, enthalpy of formation, and standard enthalpy of vaporization) of octane isomers. The effect of the addition of an edge between two non-adjacent vertices of a graph under ABSγ has also been investigated. Moreover, the graphs possessing the maximum value of ABSγ, with γ>0, are characterized from the set of all connected graphs of a fixed order and a fixed (ⅰ) vertex connectivity not greater than a given number or (ⅱ) matching number. From the above-mentioned set of three values of γ, the choices γ=3,3, yield two new indices with promising chemical usefulness. Thus, these two indices, namely ABS3 and ABS3, deserve to be examined further.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    This research was funded by Deputy for Research & Innovation, Ministry of Education through Initiative of Institutional Funding at University of Ha'il-Saudi Arabia through project number IFP-22 005.

    The authors declare that they have no conflicts of interest.



    [1] A. M. Albalahi, E. Milovanović, A. Ali, General atom-bond sum-connectivity index of graphs, Mathematics, 11 (2023), 1–15. https://doi.org/10.3390/math11112494 doi: 10.3390/math11112494
    [2] A. Ali, B. Furtula, I. Redžepović, I. Gutman, Atom-bond sum-connectivity index, J. Math. Chem., 60 (2022), 2081–2093. https://doi.org/10.1007/s10910-022-01403-1 doi: 10.1007/s10910-022-01403-1
    [3] A. Ali, I. Gutman, I. Redžepović, Atom-bond sum-connectivity index of unicyclic graphs and some applications, Electron. J. Math., 5 (2023), 1–7. https://doi.org/10.47443/ejm.2022.039 doi: 10.47443/ejm.2022.039
    [4] T. A. Alraqad, I. Ž. Milovanović, H. Saber, A. Ali, J. P. Mazorodze, Minimum atom-bond sum-connectivity index of trees with a fixed order and/or number of pendent vertices, 2022, arXiv: 2211.05218.
    [5] J. A. Bondy, U. S. R. Murty, Graph theory, Springer, 2008.
    [6] E. Estrada, L. Torres, L. Rodríguez, I. Gutman, An atom-bond connectivity index: Modelling the enthalpy of formation of alkanes, Indian J. Chem. Sect. A, 37 (1998), 849–855.
    [7] E. Estrada, Atom-bond connectivity and the energetic of branched alkanes, Chem. Phys. Lett., 463 (2008), 422–425. https://doi.org/10.1016/j.cplett.2008.08.074 doi: 10.1016/j.cplett.2008.08.074
    [8] S. Fajtlowicz, On conjectures of Graffiti-Ⅱ, Congr. Numer., 60 (1987), 187–197.
    [9] J. L. Gross, J. Yellen, Graph theory and its applications, 2 Eds., New York: Chapman & Hall/CRC, 2005. https://doi.org/10.1201/9781420057140
    [10] I. Gutman, Degree-based topological indices, Croat. Chem. Acta, 86 (2013), 351–361. http://dx.doi.org/10.5562/cca2294 doi: 10.5562/cca2294
    [11] I. Gutman, B. Furtula, Recent results in the theory of Randić index, Kragujevac: University of Kragujevac, 2008.
    [12] X. L. Li, I. Gutman, Mathematical aspects of Randić-type molecular structure descriptors, Kragujevac: University of Kragujevac, 2006.
    [13] X. L. Li, Y. T. Shi, A survey on the Randić index, MATCH Commun. Math. Comput. Chem., 59 (2008), 127–156.
    [14] L. Lovász, M. D. Plummer, Matching theory, North Holland, 1986.
    [15] Molecular descriptors. Available from: https://web.archive.org/web/20180912171255if_/http://www.moleculardescriptors.eu/index.htm
    [16] V. Maitreyi, S. Elumalai, S. Balachandran, The minimum ABS index of trees with given number of pendent vertices, 2022, arXiv: 2211.05177.
    [17] S. Noureen, A. Ali, Maximum atom-bond sum-connectivity index of n-order trees with fixed number of leaves, Discrete Math. Lett., 12 (2023), 26–28. https://doi.org/10.47443/dml.2023.016 doi: 10.47443/dml.2023.016
    [18] M. Randić, The connectivity index 25 years after, J. Mol. Graph. Model., 20 (2001), 19–35. https://doi.org/10.1016/S1093-3263(01)00098-5 doi: 10.1016/S1093-3263(01)00098-5
    [19] M. Randić, On characterization of molecular branching, J. Am. Chem. Soc., 97 (1975), 6609–6615. https://doi.org/10.1021/ja00856a001 doi: 10.1021/ja00856a001
    [20] A. Shabbir, M. F. Nadeem, Computational analysis of topological index-based entropies of carbon nanotube Y-junctions, J. Stat. Phys., 188 (2022), 31. https://doi.org/10.1007/s10955-022-02955-x doi: 10.1007/s10955-022-02955-x
    [21] X. D. Song, J. P. Li, J. B. Zhang, W. H. He, Trees with the second-minimal ABC energy, AIMS Math., 7 (2022), 18323–18333. https://doi.org/10.3934/math.20221009 doi: 10.3934/math.20221009
    [22] Y. F. Tang, D. B. West, B. Zhou, Extremal problems for degree-based topological indices, Discrete Appl. Math., 203 (2016), 134–143. https://doi.org/10.1016/j.dam.2015.09.011 doi: 10.1016/j.dam.2015.09.011
    [23] S. Wagner, H. Wang, Introduction to chemical graph theory, Boca Raton: CRC Press, 2018.
    [24] B. Zhou, N. Trinajstić, On a novel connectivity index, J. Math. Chem., 46 (2009), 1252–1270. https://doi.org/10.1007/s10910-008-9515-z doi: 10.1007/s10910-008-9515-z
  • This article has been cited by:

    1. Akbar Ali, Tomislav Došlić, Zahid Raza, On trees of a fixed maximum degree with extremal general atom-bond sum-connectivity index, 2024, 1598-5865, 10.1007/s12190-024-02275-1
    2. Xuewu Zuo, Akbar Jahanbani, Hajar Shooshtari, On the atom-bond sum-connectivity index of chemical graphs, 2024, 1296, 00222860, 136849, 10.1016/j.molstruc.2023.136849
    3. Kannan Aarthi, Suresh Elumalai, Selvaraj Balachandran, Sourav Mondal, On difference between atom-bond sum-connectivity index and Randić index of graphs, 2025, 1598-5865, 10.1007/s12190-024-02339-2
    4. Xiujun Zhang, Sahar Aftab, Sadia Noureen, Adnan Aslam, Sobhy M. Ibrahim, Topological descriptors and connectivity analysis of coronene fractal structures: insights from atom-bond sum-connectivity and Sombor indices, 2025, 0932-0784, 10.1515/zna-2024-0286
    5. Akbar Ali, Tomislav Došlić, Abdulaziz M. Alanazi, Taher S. Hassan, Adel A. Attiya, Minimizing the general atom-bond sum-connectivity Index of unicyclic graphs with a given maximum degree, 2025, 1598-5865, 10.1007/s12190-025-02503-2
  • 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(2045) PDF downloads(97) Cited by(5)

Figures and Tables

Figures(6)  /  Tables(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog