Vertex-primitive s-arc-transitive digraphs of linear groups

https://doi.org/10.1016/j.jpaa.2019.04.009Get rights and content

Abstract

We study G-vertex-primitive and (G,s)-arc-transitive digraphs for almost simple groups G with socle PSLn(q). We prove that s2 for such digraphs, which provides the first step in determining an upper bound on s for all the vertex-primitive s-arc-transitive digraphs.

Introduction

A digraph Γ is a pair (V,) with a set V (of vertices) and an antisymmetric irreflexive binary relation → on V. For a non-negative integer s, an s-arc of Γ is a sequence v0,v1,,vs of vertices with vivi+1 for each i=0,,s1. A 1-arc is also simply called an arc. For a subgroup G of Aut(Γ), we say Γ is (G,s)-arc-transitive if G acts transitively on the set of s-arcs of Γ. An (Aut(Γ),s)-arc-transitive digraph Γ is said to be s-arc-transitive. Note that a vertex-transitive (s+1)-arc-transitive digraph is necessarily s-arc-transitive. A transitive permutation group G on a set Ω is said to be primitive if G does not preserve any nontrivial partition of Ω. For a subgroup G of Aut(Γ), we say Γ is G-vertex-primitive if G is primitive on the vertex set. An Aut(Γ)-vertex-primitive digraph Γ is said to be vertex-primitive. All digraphs and groups considered in this paper will be finite.

It appears that vertex-primitive s-arc-transitive digraphs with large s are very rare. Indeed, the existence of vertex-primitive 2-arc-transitive digraphs besides directed cycles was only recently determined [8] and no vertex-primitive 3-arc-transitive examples are known. In [9] the authors asked the following question:

Question 1.1

Is there an upper bound on s for vertex-primitive s-arc-transitive digraphs that are not directed cycles?

A group G is said to be almost simple if G has a unique minimal normal subgroup T and T is a nonabelian simple group. These are precisely the groups lying between a nonabelian simple group T and its automorphism group Aut(T). A systematic investigation of the O'Nan-Scott types of primitive groups has reduced Question 1.1 to almost simple groups by showing that an upper bound on s for vertex-primitive s-arc-transitive digraphs Γ with Aut(Γ) almost simple will be an upper bound on s for all vertex-primitive s-arc-transitive digraphs [9, Corollary 1.6]. This paper provides the first step in determining such an upper bound by studying vertex-primitive s-arc-transitive digraphs whose automorphism group is an almost simple linear group. Our main result is as follows.

Theorem 1.2

Let Γ be a G-vertex-primitive (G,s)-arc-transitive digraph, where G is almost simple with socle PSLn(q). Then s2.

We remark that an infinite family of G-vertex-primitive (G,2)-arc-transitive digraphs with G=PSL3(p2) for each prime p>3 such that p±2(mod5) was constructed in [8]. These digraphs have vertex stabilizer A6 and arc-stabilizer A5, and are the only known examples of G-vertex-primitive (G,2)-arc-transitive digraphs such that G is almost simple. A complete classification of G-vertex-primitive (G,2)-arc-transitive digraphs for almost simple groups G, even for those with Soc(G)=PSLn(q), seems out of reach at this stage, though would be achievable for small values of n.

Note that if Soc(G)=PSLn(q) then either GPΓLn(q) or G has an index 2 subgroup contained in PΓLn(q) and G contains an element that acts on the projective space associated with G by interchanging the set of 1-spaces and the set of hyperplanes. For any G-vertex-primitive (G,s)-arc-transitive digraph Γ, the vertex stabilizer Gv for any vertex v of Γ is maximal in G. We prove Theorem 1.2 by analyzing the maximal subgroups of G according to the classes provided by Aschbacher's theorem [1]. The classes C1, C2, …, C8 are discussed in Sections 4–6, while the remaining class C9 is dealt with in Section 3. We actually prove that there is no G-vertex-primitive (G,2)-arc-transitive digraph with Gv from classes C3, …, C6 (Theorem 5.6) though the possibility for an example with Gv from classes C1, C2, C7 or C8 remains open. The examples in [8] have Gv from the class C9. At the end of Section 6 we give a proof of Theorem 1.2.

Section snippets

Notation

For a group X, denote by Soc(X) the socle of X (that is, the product of all minimal normal subgroups of X), F(G) the Fitting subgroup of G, Rad(X) the largest soluble normal subgroup of X, and X() the smallest normal subgroup of X such that X/X() is soluble.

For a group X and a prime p, denote by Op(X) the largest normal p-subgroup of X, and Ωp(X) the subgroup of X generated by the elements of order p in X.

For any integer n and prime number p, denote by np the p-part of n (that is, the largest

Homogeneous factorizations

From Lemma 2.11 we see that the s-arc-transitivity of digraphs can be characterized by group factorizations. If a group G acts s-arc-transitively on a digraph Γ with s2 and v0v1vs1vs is an s-arc of Γ, then Lemma 2.11 implies that Gv1vi=Gv0v1viGv1vivi+1 for all 1is1. In addition, since G acts i-arc-transitively on Γ, the two factors Gv0v1vi and Gv1vivi+1 are conjugate in G and hence isomorphic. This motivates the following definition:

Definition 3.1

A factorization G=AB is called a homogeneous

C1 and C2-subgroups

Hypothesis 4.1

Let Γ be a G-vertex-primitive (G,2)-arc-transitive digraph of valency at least 3, where G is almost simple with socle L=PSLn(q) and q=pf for some prime p. Take an arc uv of Γ. Let g be an element of L such that ug=v and let w=vg. Then uvw is a 2-arc in Γ. Let X=SLn(q) acting naturally on V=Fqn, φ be the projection from X to L, and g be a preimage of g under φ.

Under Hypothesis 4.1, if in addition Γ is (G,3)-arc-transitive, then Lemma 2.12 asserts that Γ is (L,2)-arc-transitive so that Lv=L

C3, C4, C5 and C6-subgroups

We recall Hypothesis 4.1 and observe that Gv=GuvGvw with GuvGvw and so π(Gv)=π(Guv)=π(Gvw).

Lemma 5.1

If Hypothesis 4.1 holds then Gv is not a C3-subgroup of G.

Proof

Suppose that Hypothesis 4.1 holds and Gv is a C3-subgroup of G. Then by [13, Proposition 4.3.6], either n is prime, or Gv/Rad(Gv) is almost simple with socle PSLn/r(qr) for some prime divisor r of n.

First assume that n is a prime with ppd(p,nf). If (n,q)=(2,8) then Gv=D18 or C9C6, where G=PSL2(8) or PΓL2(8) respectively, but a Magma [4]

C7 and C8-subgroups

In this section we need to consider the stronger hypothesis that Γ is (G,3)-arc-transitive so that we only need to consider the structure of Lv instead of Gv.

Hypothesis 6.1

Let Γ be a G-vertex-primitive (G,3)-arc-transitive digraph of valency at least 3, where G is almost simple with socle L=PSLn(q) and q=pf for some prime p. Then by Lemma 2.12, Γ is (L,2)-arc-transitive. Take an arc uv of Γ. Let g to be an element of L such that ug=v and let w=vg. Then uvwwg is a 3-arc in Γ.

Under Hypothesis 6.1, it

Acknowledgements

This research was supported by Australian Research Council grant DP150101066. The third author's work on this paper was done when he was a research associate at the University of Western Australia. The second author acknowledges the support of NNSFC grant 11771200. The authors would like to thank the anonymous referee for careful reading and valuable suggestions.

References (20)

There are more references available in the full text version of this article.

Cited by (7)

  • On edge-primitive 3-arc-transitive graphs

    2021, Journal of Combinatorial Theory. Series B
    Citation Excerpt :

    The following results allow us to eliminate certain families of maximal subgroups. The first is essentially proved in [9, Lemma 2.14], and we omit the proof. Tutte's 8-cage

View all citing articles on Scopus
View full text