This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

LPS-Type Ramanujan Graphs from Definite Quaternion Algebras over \mathbb{Q} of Class Number One

Jonah Mendel Department of Mathematics, Brown University [email protected]  and  Jiahui Yu Department of Mathematics, MIT [email protected]
Abstract.

In this paper we construct explicit LPS-type Ramanujan graphs from each definite quaternion algebra over \mathbb{Q} of class number 1, extending the constructions of [10] and [1], and answering in the affirmative a question raised in [6]. We do this by showing that for each definite quaternion algebra \mathcal{H} over \mathbb{Q} of class number 1 with maximal order 𝒪\mathcal{O}, if G=×/Z(×)G=\mathcal{H}^{\times}/Z(\mathcal{H}^{\times}) and pp is prime such that G(p)PGL2(p)G(\mathbb{Q}_{p})\cong\operatorname{PGL}_{2}(\mathbb{Q}_{p}) then there exists a congruence pp-arithmetic subgroup of GG which acts simply transitively on the Bruhat-Tits tree of G(p)G(\mathbb{Q}_{p}).

1. Introduction

A kk-regular graph is called a Ramanujan graph if the second largest eigenvalue in absolute value is less than or equal to 2k12\sqrt{k-1}. Ramanujan graphs are optimal expander graphs from a spectral perspective. In their paper [10], Lubotzky, Philips, and Sarnak explicitly construct a family of degree p+1p+1 Ramanujan graphs, for each odd prime p1(mod4)p\equiv 1\pmod{4} (for the case p3(mod4)p\equiv 3\pmod{4} see [2]). The LPS Ramanujan graphs are Cayley graphs of the group PSL(𝔽q)\operatorname{PSL}(\mathbb{F}_{q}) for primes q1(mod4)q\equiv 1\pmod{4}, with respect to a well-chosen generating set coming from the Hamiltonian quaternion algebra over \mathbb{Q}. In [1], Chiu addresses the case when p=2p=2 and gives an explicit construction of a family of 3-regular Ramanujan graphs, by using the definite quaternion algebra over \mathbb{Q} with discriminant 13.

In [6], Jo and Yamasaki generalize the construction of LPS and Chiu’s Ramanujan graphs. More precisely, they consider any definite quaternion algebra over \mathbb{Q} of class number one, make use of Ibukiyama’s [5] explicit construction of maximal orders for all definite quaternion algebras, and construct Cayley graphs of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}) with generating sets coming from these quaternion algebras, which they call LPS-type graphs.

The question of whether these LPS-type graphs are indeed Ramanujan was raised in [6], who were able to answer it only in the special case of the definite quaternion algebra over \mathbb{Q} with discriminant 13. In this paper we answer this question affirmatively, showing that for any definite quaternion algebra over \mathbb{Q} of class number one we can construct from it LPS-type Ramanujan graphs.

Theorem 1.1.

For any P{2,3,5,7,13}P\in\left\{2,3,5,7,13\right\}, let \mathcal{H} be a quaternion algebra over \mathbb{Q} with discriminant PP and let 𝒪\mathcal{O} be the maximal order of \mathcal{H}. For almost any prime pp, there is a specific subset S{α𝒪:N(α)=p}S\subseteq\left\{\alpha\in\mathcal{O}:N(\alpha)=p\right\} of size p+1p+1, such that the LPS-type graphs of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}), for infinitely many primes qq, with respect to the generating set S(modq)S\pmod{q}, are Ramanujan graphs.

For the precise construction of these Ramanujan graphs, see Section 5. The key ingredient in proving Theorem 1.1 is to find congruence pp-arithmetic subgroups of G=×/Z(×)G=\mathcal{H}^{\times}/Z(\mathcal{H}^{\times}), where pdisc()p\nmid\operatorname{disc}(\mathcal{H}), which acts simply transitively on the vertices of the Bruhat-Tits tree of G(p)PGL2(p)G(\mathbb{Q}_{p})\cong\operatorname{PGL}_{2}(\mathbb{Q}_{p}). The fact that it is a congruence subgroup is crucial for the proof that these graphs are Ramanujan, see for example [9, Theorem 7.3.1].

Therefore, our second main result, which might be of independent interest, is the construction of such congruence subgroups which acts simply transitively on Bruhat-Tits trees.

Theorem 1.2.

Let \mathcal{H} be a quaternion algebra over \mathbb{Q} of class number one and let 𝒪\mathcal{O} be the maximal order of \mathcal{H}. For any commutative ring with unity AA, define the group 𝒢(A)=(A𝒪)×/A×\mathcal{G}(A)=(A\otimes_{\mathbb{Z}}\mathcal{O})^{\times}/A^{\times}. Then there exists mm\in\mathbb{N} and H𝒢(/m)H\leq\mathcal{G}(\mathbb{Z}/m\mathbb{Z}) such that for any prime pmdisc()p\nmid m\cdot\operatorname{disc}(\mathcal{H}), the congruence subgroup

Λp={g𝒢([1/p]):g(modm)H}𝒢(p)\Lambda^{p}=\left\{g\in\mathcal{G}(\mathbb{Z}[1/p]):g\pmod{m}\in H\right\}\leq\mathcal{G}(\mathbb{Q}_{p})

acts simply transitively on the vertices of the Bruhat-Tits tree of 𝒢(p)PGL2(p)\mathcal{G}(\mathbb{Q}_{p})\cong\operatorname{PGL}_{2}(\mathbb{Q}_{p}).

Organization of the paper: In section 2, we collect definitions and results on quaternion algebras, their orders and Bruhat-Tits trees. In section 3, we explain the construction of Ramanujan graphs as Cayley graphs of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}). In section 4, we describe the construction of LPS-type graphs and show how Theorem 1.1 is obtained from Theorem 1.2. In section 5, we prove Theorem 1.2, and show that the definite quaternion algebra over \mathbb{Q} of discriminant 3 does not contain a simply transitively congruence subgroup of level mm, for any prime mm.

2. Preliminaries

Notation.

Let pp be a prime number and RR a ring. We will use the following conventions:

  • 𝔽p\mathbb{F}_{p} denotes the field of pp elements.

  • p\mathbb{Z}_{p} denotes the set of pp-adic integers and p\mathbb{Q}_{p} the set of pp-adic rationals.

  • R×R^{\times} denotes the multiplicative group of units in RR.

  • n(R)\mathcal{M}_{n}(R) denotes the ring of n×nn\times n matrices over RR.

  • GLn(R)=n(R)×\operatorname{GL}_{n}(R)=\mathcal{M}_{n}(R)^{\times} and SLn(R)=Ker(det:GLn(R)R×)\operatorname{SL}_{n}(R)=\operatorname{Ker}(\det:\operatorname{GL}_{n}(R)\to R^{\times}).

  • PGL2(R)=GL2(R)/R×\operatorname{PGL}_{2}(R)=\operatorname{GL}_{2}(R)/R^{\times} and PSL2(R)=SL2(R)/{±1}\operatorname{PSL}_{2}(R)=\operatorname{SL}_{2}(R)/\left\{\pm 1\right\}.

  • Sn,AnS_{n},A_{n} are the symmetric and alternating group on nn elements respectively.

  • CnC_{n} is the cyclic group of nn elements.

  • DnD_{n} is the dihedral group of 2n2n elements.

  • {1}\left\{1\right\} is the trivial group.

  • For a finite group GG, CGC_{G} denotes the conjugacy classes of subgroups of GG.

  • Let a group GG act on a set XX. For xXx\in X, StabG(x)={gG:gx=x}\operatorname{Stab}_{G}(x)=\left\{g\in G:gx=x\right\}.

2.1. Quaternion Algebras

Let kk be a field of characteristic 2\neq 2 and a,bk×a,b\in k^{\times}. The quaternion algebra a,b(k)\mathcal{H}_{a,b}(k) is an algebra over kk generated by the elements 1,i,j,ij1,i,j,ij satisfying

ij+ji=0,i2=a,j2=b.ij+ji=0,\quad i^{2}=a,\quad j^{2}=b.

Let α=x+yi+zj+wija,b(k)\alpha=x+yi+zj+wij\in\mathcal{H}_{a,b}(k). We define the conjugate and norm of α\alpha to be

α¯=xyizjwij,andN(α)=αα¯=x2ay2bz2abw2\overline{\alpha}=x-yi-zj-wij,\quad\text{and}\quad N(\alpha)=\alpha\overline{\alpha}=x^{2}-ay^{2}-bz^{2}-abw^{2}

respectively. We write a,b\mathcal{H}_{a,b} to denote quaternion algebras over \mathbb{Q}.

We say that a quaternion algebra a,b(k)\mathcal{H}_{a,b}(k) splits if a,b(k)2(k)\mathcal{H}_{a,b}(k)\cong\mathcal{M}_{2}(k). Otherwise we say that a,b(k)\mathcal{H}_{a,b}(k) ramifies. Similarly, we say that a quaternion algebra a,b\mathcal{H}_{a,b} over \mathbb{Q} splits at a prime pp if a,bp2(p)\mathcal{H}_{a,b}\otimes\mathbb{Q}_{p}\cong\mathcal{M}_{2}(\mathbb{Q}_{p}). Otherwise, we say that a,b\mathcal{H}_{a,b} ramifies at pp. We say that a quaternion algebra a,b\mathcal{H}_{a,b} is definite if a,b\mathcal{H}_{a,b}\otimes\mathbb{R} is a division algebra. The following lemma gives us helpful facts about splitting.

Proposition 2.1.

[13, Corollary 7.1.2, Theorem 14.1.3]

  1. (1)

    A quaternion algebra a,b(k)\mathcal{H}_{a,b}(k) that is not a division ring splits.

  2. (2)

    For any finite field 𝔽q\mathbb{F}_{q}, all quaternion algebras over 𝔽q\mathbb{F}_{q} split.

  3. (3)

    Two definite quaternion algebras a,b\mathcal{H}_{a,b} and a,b\mathcal{H}_{a^{\prime},b^{\prime}} over \mathbb{Q} are isomorphic if and only if they ramify at exactly the same primes.

  4. (4)

    A quaternion algebra \mathcal{H} over \mathbb{Q} ramifies at only finitely many primes.

The product of the finitely many primes at which a,b=a,b()\mathcal{H}_{a,b}=\mathcal{H}_{a,b}(\mathbb{Q}) ramifies at is called the discriminant of a,b\mathcal{H}_{a,b} and is denoted disc(a,b)\operatorname{disc}(\mathcal{H}_{a,b}).

2.2. Orders of Quaternion Algebras

An order of a quaternion algebra a,b\mathcal{H}_{a,b} is a rank 4 module over \mathbb{Z} that is also a subring of a,b\mathcal{H}_{a,b}. An order 𝒪\mathcal{O} is called a maximal order if it is not properly contained in an order of a,b\mathcal{H}_{a,b}. A right ideal of an order 𝒪\mathcal{O} is a module MM satisfying M𝒪=MM\mathcal{O}=M. The class number of an order 𝒪\mathcal{O} is the number of inequivalent right ideals of 𝒪\mathcal{O}, where two ideals M,NM,N are equivalent if M=αNM=\alpha N for some αa,b×\alpha\in\mathcal{H}_{a,b}^{\times}. It turns out that all maximal orders of a quaternion algebra have the same class number [13, Chapter 17] and we define the class number of the quaternion algebra to be the class number of its maximal orders. The following theorem provides a way to calculate the class number of a quaternion algebra.

Theorem 2.2.

[4, Theorem 6] The class number hh of a definite quaternion algebra \mathcal{H} is given by

h=112p|disc()(p1)+14p|disc()(1(4p))+13p|disc()(1(3p))h=\frac{1}{12}\prod_{p|\operatorname{disc}(\mathcal{H})}(p-1)+\frac{1}{4}\prod_{p|\operatorname{disc}(\mathcal{H})}\left(1-\left(\frac{-4}{p}\right)\right)+\frac{1}{3}\prod_{p|\operatorname{disc}(\mathcal{H})}\left(1-\left(\frac{-3}{p}\right)\right)

where (np)\left(\frac{n}{p}\right) is the Legendre symbol defined as

(ap)={1if a is a quadratic redsidue modulo p1if a is not a quadratic residue modulo p0pa.\left(\frac{a}{p}\right)=\begin{cases}1&\text{if $a$ is a quadratic redsidue modulo $p$}\\ -1&\text{if $a$ is not a quadratic residue modulo $p$}\\ 0&p\mid a.\end{cases}
Corollary 2.3.

The class number hh of \mathcal{H} is 11 if and only if disc(){2,3,5,7,13}\operatorname{disc}(\mathcal{H})\in\left\{2,3,5,7,13\right\}.

A class number one quaternion algebra over \mathbb{Q} has only one maximal order up to conjugation. When the quaternion algebra is definite and has class number 1, the following theorem gives us a way to count the number of elements whose norm is a prime power, unique up to units, in the maximal ideal.

Proposition 2.4.

[1, Theorem 3.2] Let \mathcal{H} be a definite quaternion algebra of class number 1 over \mathbb{Q} which splits at pp. Let 𝒪\mathcal{O} be a maximal order of \mathcal{H}. Then,

#{α𝒪:N(α)=pk}/𝒪×=1+p++pk.\#\left\{\alpha\in\mathcal{O}:N(\alpha)=p^{k}\right\}/\mathcal{O}^{\times}=1+p+\cdots+p^{k}.

2.3. The Bruhat-Tits Tree

A p\mathbb{Z}_{p} lattice in pp\mathbb{Q}_{p}\oplus\mathbb{Q}_{p} is a rank 2 p\mathbb{Z}_{p}-submodule of pp\mathbb{Q}_{p}\oplus\mathbb{Q}_{p}. Given two p\mathbb{Z}_{p}-lattices L1,L2L_{1},L_{2} in pp\mathbb{Q}_{p}\oplus\mathbb{Q}_{p}, if L1=cL2L_{1}=cL_{2} for some cp×c\in\mathbb{Q}_{p}^{\times}, then we say that L1L_{1} and L2L_{2} are homothetic, and we write L1L2L_{1}\sim L_{2}. Let V~\widetilde{V} be the set of p\mathbb{Z}_{p}-lattices in pp\mathbb{Q}_{p}\oplus\mathbb{Q}_{p}. It is clear that the homoethetic relation \sim defines an equivalence relation on V~\widetilde{V}. We define the set of homothetic classes of lattices to be V=V~/V=\widetilde{V}/\sim. Let 𝒯p+1\mathcal{T}_{p+1} be the graph whose vertices are elements of VV. Given two vertices [L][L] and [M][M], there is an edge between [L][L] and [M][M] if and only if there exists L1[L]L_{1}\in[L] and L2[M]L_{2}\in[M] such that

pL2L1L2.pL_{2}\subsetneq L_{1}\subsetneq L_{2}.
Theorem 2.5.

[12, Chapter II Theorem 1] The graph 𝒯p+1\mathcal{T}_{p+1} defined above is a tree of degree p+1p+1.

We call 𝒯p+1\mathcal{T}_{p+1} the Bruhat-Tits tree of degree p+1p+1. The following proposition relates the Bruhat-Tits tree to the groups PGL2(p)\operatorname{PGL}_{2}(\mathbb{Q}_{p}) and PGL2(p)\operatorname{PGL}_{2}(\mathbb{Z}_{p}).

Proposition 2.6.

[12, Chapter II, Sections 3,4] The group PGL2(p)\operatorname{PGL}_{2}(\mathbb{Q}_{p}) acts transitively on the Bruhat-Tits tree 𝒯p+1\mathcal{T}_{p+1} and the stabilizer of the root [pp][\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}] is PGL2(p)\operatorname{PGL}_{2}(\mathbb{Z}_{p}).

3. Ramanujan Graphs and Algebraic Groups

3.1. The Algebraic Group 𝒢\mathcal{G}

Fix a prime pp. Let =a,b()\mathcal{H}=\mathcal{H}_{a,b}(\mathbb{Q}) be a definite class number 1 quaternion algebra over \mathbb{Q} which splits at pp. Let 𝒪=ω1ω2ω3\mathcal{O}=\mathbb{Z}\oplus\mathbb{Z}\omega_{1}\oplus\mathbb{Z}\omega_{2}\oplus\mathbb{Z}\omega_{3} be the unique (up to conjugacy) maximal order of \mathcal{H}. Let N:N:\mathcal{H}\to\mathbb{Q} be the norm. Note that N(𝒪)N(\mathcal{O})\subseteq\mathbb{Z} and

𝒪×={α𝒪:(α)=±1}.\mathcal{O}^{\times}=\left\{\alpha\in\mathcal{O}:\mathbb{N}(\alpha)=\pm 1\right\}.

For any abelian ring with unity AA, define

𝒢(A)=𝒢𝒪(A):=(𝒪A)×/A×.\mathcal{G}(A)=\mathcal{G}_{\mathcal{O}}(A):=\left(\mathcal{O}\otimes A\right)^{\times}/A^{\times}.

When the context is clear, we drop the subscript 𝒪\mathcal{O}. Note that 𝒢()=𝒪×/{±1}\mathcal{G}(\mathbb{Z})=\mathcal{O}^{\times}/\left\{\pm 1\right\}. Furthermore, [1/p]×={pk:k}\mathbb{Z}[1/p]^{\times}=\left\{p^{k}:k\in\mathbb{Z}\right\}. Therefore,

(1) 𝒢([1/p]){γ𝒪:N(γ)=pk,k}/{±pk:k}\mathcal{G}(\mathbb{Z}[1/p])\cong\left\{\gamma\in\mathcal{O}:N(\gamma)=p^{k},\ k\in\mathbb{Z}\right\}/\left\{\pm p^{k}:k\in\mathbb{Z}\right\}

The following proposition relates 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) to the Bruhat-Tits tree 𝒯p+1\mathcal{T}_{p+1}.

Proposition 3.1.

The group 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) acts transitively on the vertices of 𝒯p+1\mathcal{T}_{p+1} and 𝒢()\mathcal{G}(\mathbb{Z}) is the stabilizer of [pp]\left[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}\right].

Proof.

Because \mathcal{H} splits at pp, there is an isomorphism ψ:p2(p)\psi:\mathcal{H}\otimes\mathbb{Q}_{p}\rightarrow\mathcal{M}_{2}(\mathbb{Q}_{p}) such that N(α)=detψ(α)N(\alpha)=\det\psi(\alpha) for α\alpha\in\mathcal{H}. The map ψ\psi induces an embedding, 𝒢([1/p])PGL2(p)\mathcal{G}(\mathbb{Z}[1/p])\hookrightarrow\operatorname{PGL}_{2}(\mathbb{Q}_{p}). Matrices of determinant pp map any vertex of 𝒯p+1\mathcal{T}_{p+1} to a neighbor. Thus, elements in 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) of norm pp map any vertex to a neighbor. By Proposition 2.4, there exist p+1p+1 elements in missingG([1/p])\mathcal{\mathcal{missing}}G(\mathbb{Z}[1/p]) that have norm pp. They map each vertex of 𝒯p+1\mathcal{T}_{p+1} to the p+1p+1 neighbors of it. Thus, we can reach any vertex of 𝒯p+1\mathcal{T}_{p+1} by a product of elements of norm pp acting on the origin. Therefore, 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) acts transitively on 𝒯p+1\mathcal{T}_{p+1}.

We have that p×{±pk:k}={±1}\mathbb{Z}_{p}^{\times}\cap\{\pm p^{k}:k\in\mathbb{Z}\}=\left\{\pm 1\right\}. Elements in GL2(p)\operatorname{GL}_{2}(\mathbb{Z}_{p}) have determinants in p×\mathbb{Z}_{p}^{\times}. Thus, 𝒢([1/p])PGL2(p)=𝒢()\mathcal{G}(\mathbb{Z}[1/p])\cap\operatorname{PGL}_{2}(\mathbb{Z}_{p})=\mathcal{G}(\mathbb{Z}). Therefore, when 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) acts on 𝒯p+1\mathcal{T}_{p+1}, the stabilizer of the root [pp]\left[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}\right] is 𝒢()\mathcal{G}(\mathbb{Z}). ∎

We want to find a subgroup Λ\Lambda of 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) which acts simply transitively on 𝒯p+1\mathcal{T}_{p+1}. Let mm be a positive integer. We define the reduction modulo mm map to be

φ¯m:𝒪𝒪/m,w+xω1+yω2+zω3w¯+x¯ω1+y¯ω2+z¯ω3\bar{\varphi}_{m}:\mathcal{O}\to\mathcal{O}\otimes\mathbb{Z}/m\mathbb{Z},\quad w+x\omega_{1}+y\omega_{2}+z\omega_{3}\mapsto\bar{w}+\bar{x}\omega_{1}+\bar{y}\omega_{2}+\bar{z}\omega_{3}

where a¯\bar{a} is the reduction modulo mm of the integer aa. We know from (1) that elements of 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]) are represented by elements of 𝒪\mathcal{O}, so φ¯m\bar{\varphi}_{m} induces a map,

φm:𝒢([1/p])𝒢(/m).\varphi_{m}:\mathcal{G}(\mathbb{Z}[1/p])\to\mathcal{G}(\mathbb{Z}/m\mathbb{Z}).
Definition 3.2.

Let GG be a group and let H,KH,K be subgroups of GG. We call KK a right complement of HH if

HK={1},HK=G.H\cap K=\left\{1\right\},\quad H\cdot K=G.
Definition 3.3.

Let (m,H)(m,H) be a pair where mm is a positive integer and HH is a subgroup of 𝒢(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z}) such that

  1. (1)

    𝒢()\mathcal{G}(\mathbb{Z}) embeds into 𝒢(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z}) via φm\varphi_{m},

  2. (2)

    HH is a right complement of φm(𝒢())\varphi_{m}(\mathcal{G}(\mathbb{Z})) in 𝒢(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z}).

Then we call (m,H)(m,H) a congruence pair of 𝒪\mathcal{O}. If pmp\nmid m, we define the following congruence subgroup of 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]),

Λp=Λ𝒪p(m,H):={γ𝒢([1/p]):φm(γ)H}=φm1(H)\Lambda^{p}=\Lambda^{p}_{\mathcal{O}}(m,H):=\left\{\gamma\in\mathcal{G}(\mathbb{Z}[1/p]):\varphi_{m}(\gamma)\in H\right\}=\varphi_{m}^{-1}(H)

with a subset

𝒮p=𝒮𝒪p(m,H)={αΛp:N(α)=p}.\mathcal{S}^{p}=\mathcal{S}^{p}_{\mathcal{O}}(m,H)=\left\{\alpha\in\Lambda^{p}:N(\alpha)=p\right\}.

When the context is clear, we write Λp\Lambda^{p} and 𝒮p\mathcal{S}^{p}.

Proposition 3.4.

Let (m,H)(m,H) be a congruence pair of 𝒪\mathcal{O}. Then 𝒢()\mathcal{G}(\mathbb{Z}) and Λp\Lambda^{p} are complementary subgroups of 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]),

Proof.

We have that Λp=φm1(H)\Lambda^{p}=\varphi_{m}^{-1}(H), so Λp\Lambda^{p} is a subgroup of 𝒢([1/p])\mathcal{G}(\mathbb{Z}[1/p]). We can deduce that φm1(H)𝒢()={1}\varphi_{m}^{-1}(H)\cap\mathcal{G}(\mathbb{Z})=\left\{1\right\} from the assumptions H𝒢()={1}H\cap\mathcal{G}(\mathbb{Z})=\left\{1\right\} and φm|𝒢()\varphi_{m}|_{\mathcal{G}(\mathbb{Z})} is injective. Let γ𝒢([1/p])\gamma\in\mathcal{G}(\mathbb{Z}[1/p]). We have that 𝒢()H=𝒢(/m)\mathcal{G}(\mathbb{Z})\cdot H=\mathcal{G}(\mathbb{Z}/m\mathbb{Z}), so we can choose α𝒢()𝒢([1/p])\alpha\in\mathcal{G}(\mathbb{Z})\leq\mathcal{G}(\mathbb{Z}[1/p]) and βH\beta\in H be such that αβ=φm(γ)\alpha\beta=\varphi_{m}(\gamma). We have that

φm(α1γ)=φm(α1)φm(γ)=α1(αβ)=βH.\varphi_{m}(\alpha^{-1}\gamma)=\varphi_{m}(\alpha^{-1})\varphi_{m}(\gamma)=\alpha^{-1}(\alpha\beta)=\beta\in H.

Therefore, α1γΛp\alpha^{-1}\gamma\in\Lambda^{p}. Thus, α(α1γ)=γ\alpha(\alpha^{-1}\gamma)=\gamma, so γ𝒢()Λp\gamma\in\mathcal{G}(\mathbb{Z})\cdot\Lambda^{p}. Therefore, 𝒢()Λp=𝒢([1/p])\mathcal{G}(\mathbb{Z})\cdot\Lambda^{p}=\mathcal{G}(\mathbb{Z}[1/p]). ∎

Proposition 3.5.

Let (m,H)(m,H) be a congruence pair of 𝒪\mathcal{O}. Then the group Λp\Lambda^{p} acts simply transitively on the Bruhat-Tits tree 𝒯p+1\mathcal{T}_{p+1}.

Proof.

We first show that Λp\Lambda^{p} acts transtively on 𝒯p+1\mathcal{T}_{p+1}. Let u,vu,v be vertices of 𝒯p+1\mathcal{T}_{p+1}. Let LL be a p\mathbb{Z}_{p}-lattice. There exists γ𝒢([1/p])\gamma\in\mathcal{G}(\mathbb{Z}[1/p]) such that γ[pp]=[L]\gamma\cdot[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]=[L]. By Proposition 3.4, we can write γ=λα\gamma=\lambda\alpha for α𝒢()\alpha\in\mathcal{G}(\mathbb{Z}) and λΛp\lambda\in\Lambda^{p}. We have that

[L]=γ[pp]=(λα)[pp]=λ(α[pp])=λ[pp][L]=\gamma\cdot[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]=(\lambda\alpha)\cdot[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]=\lambda\cdot(\alpha\cdot[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}])=\lambda\cdot[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]

because 𝒢()\mathcal{G}(\mathbb{Z}) is the stabilizer of [Zpp][Z_{p}\oplus\mathbb{Z}_{p}]. Therefore, Λp\Lambda^{p} acts transitively on 𝒯p+1\mathcal{T}_{p+1}. Lastly, Λp\Lambda^{p} acts imply on 𝒯p+1\mathcal{T}_{p+1} because

StabΛp([pp])=ΛpStab𝒢([1/p])([Zpp])=Λp𝒢()={1}.\operatorname{Stab}_{\Lambda^{p}}([\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}])=\Lambda^{p}\cap\operatorname{Stab}_{\mathcal{G}(\mathbb{Z}[1/p])}([Z_{p}\oplus\mathbb{Z}_{p}])=\Lambda^{p}\cap\mathcal{G}(\mathbb{Z})=\left\{1\right\}.

Proposition 3.6.

For any congruence pair (m,H)(m,H), #𝒮p=p+1\#\mathcal{S}^{p}=p+1 and 𝒮p\mathcal{S}^{p} generates Λp\Lambda^{p}.

Proof.

The group Λp\Lambda^{p} acts transitively on the vertices of 𝒯p+1\mathcal{T}_{p+1}. Therefore, there exist at least p+1p+1 elements of Λp\Lambda^{p} which take [pp]\left[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}\right] to each of its p+1p+1 neighbors respectively. Each of these elements have norm pp. Let α\alpha and β\beta be distinct elements of norm pp such that α[pp]=β[pp]\alpha\cdot\left[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}\right]=\beta\cdot\left[\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}\right]. Then αβ1\alpha\beta^{-1} is a nontrivial element of Λp\Lambda^{p} which acts trivially on [pp][\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]. This is a contradiction because Λp\Lambda^{p} acts simply on the vertices of 𝒯p+1\mathcal{T}_{p+1}. Therefore, there are exactly p+1p+1 elements in Λp\Lambda^{p} of norm pp. We now want to show that 𝒮p\mathcal{S}^{p} generates Λp\Lambda^{p}. Fix a vertex [L][L] of 𝒯p+1\mathcal{T}_{p+1} of distance dd from [pp][\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}]. For every neighbor [M][M] of [L][L], there exists a unique element of 𝒮p\mathcal{S}^{p} which takes [L][L] to [M][M]. Therefore, there exists a unique word of symbols in 𝒮p\mathcal{S}^{p} of length dd which takes [Zpp][Z_{p}\oplus\mathbb{Z}_{p}]. Therefore, there is an isomorphism of graphs 𝒯p+1Cay(Λp,𝒮p)\mathcal{T}_{p+1}\cong\operatorname{Cay}(\Lambda^{p},\mathcal{S}^{p}) with [pp][\mathbb{Z}_{p}\oplus\mathbb{Z}_{p}] taken as the identity. Because 𝒯p+1\mathcal{T}_{p+1} is connected, we have that 𝒮p\mathcal{S}^{p} generates Λp\Lambda^{p}. ∎

The following proposition characterizes the structure of 𝒢(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z}). This proposition also helps us find congruence pairs (m,H)(m,H).

Proposition 3.7.

When m=pkm=p^{k} for k1k\geq 1, then

𝒢(/m)PGL2(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/m\mathbb{Z})
Proof.

\mathcal{H} splits at pp. Thus, 𝒪p2(p)\mathcal{O}\otimes\mathbb{Z}_{p}\cong\mathcal{M}_{2}(\mathbb{Z}_{p}) because 𝒪\mathcal{O} is a maximal order of \mathcal{H}. Therefore, we have the following isomorphisms of rings,

𝒪/m𝒪/m𝒪(𝒪p)/m(𝒪p)2(p)/m2(p)2()/mM2()2(/m).\mathcal{O}\otimes\mathbb{Z}/m\mathbb{Z}\cong\mathcal{O}/m\mathcal{O}\cong(\mathcal{O}\otimes\mathbb{Z}_{p})/m(\mathcal{O}\otimes\mathbb{Z}_{p})\cong\mathcal{M}_{2}(\mathbb{Z}_{p})/m\mathcal{M}_{2}(\mathbb{Z}_{p})\cong\mathcal{M}_{2}(\mathbb{Z})/mM_{2}(\mathbb{Z})\cong\mathcal{M}_{2}(\mathbb{Z}/m\mathbb{Z}).

We have that 𝒢(/m)(𝒪/m)×/(/m)×\mathcal{G}(\mathbb{Z}/m\mathbb{Z})\cong(\mathcal{O}\otimes\mathbb{Z}/m\mathbb{Z})^{\times}/(\mathbb{Z}/m\mathbb{Z})^{\times}. Therefore, by showing that 𝒪/m2(/m)\mathcal{O}\otimes\mathbb{Z}/m\mathbb{Z}\cong\mathcal{M}_{2}(\mathbb{Z}/m\mathbb{Z}), we have that 𝒢(/m)PGL2(/m)\mathcal{G}(\mathbb{Z}/m\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/m\mathbb{Z}). ∎

Proposition 3.8.

Let qpq\neq p be a prime such that \mathcal{H} splits at qq. Let

Λp(q)=Ker(φq)Λp,\Lambda^{p}(q)=\operatorname{Ker}(\varphi_{q})\cap\Lambda^{p},

i.e., the kernel of φq\varphi_{q} restricted to Λp\Lambda^{p}. Then we have the following isomorphism of groups,

Λp(q)\Λp{PGL2(/q)(pq)=1PSL2(/q)(pq)=1.\Lambda^{p}(q)\backslash\Lambda^{p}\cong\begin{cases}\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z})&\left(\frac{p}{q}\right)=-1\\ \operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z})&\left(\frac{p}{q}\right)=1.\end{cases}
Proof.

By Proposition 3.7, φq(Λp)PGL2(/q)\varphi_{q}(\Lambda^{p})\leq\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z}). We want to show that PSL2(/q)φq(Λp)\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z})\leq\varphi_{q}(\Lambda^{p}), for which we use Strong Approximation. By [11, Lemma 1.2], the reduction modulo qq map induces a surjection

{α𝒢([1/p]):N(α)=1}{α𝒢(/q):N(α)=1}.\left\{\alpha\in\mathcal{G}(\mathbb{Z}[1/p]):N(\alpha)=1\right\}\to\left\{\alpha\in\mathcal{G}(\mathbb{Z}/q\mathbb{Z}):N(\alpha)=1\right\}.

Let N(β)1(modq)N(\beta)\equiv 1\pmod{q} for β𝒢(/q)PGL2(/q)\beta\in\mathcal{G}(\mathbb{Z}/q\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z}). By [11, Lemma 1.1], we can find (a0,a1,a2,a3)4(a_{0},a_{1},a_{2},a_{3})\in\mathbb{Z}^{4} such that α=a0+a1ω1+a2ω2+a3ω3\alpha=a_{0}+a_{1}\omega_{1}+a_{2}\omega_{2}+a_{3}\omega_{3} is of norm pkp^{k} where pk1(modq)p^{k}\equiv 1\pmod{q} and φq(α)=β\varphi_{q}(\alpha)=\beta. Therefore,

PSL2(/q)Λp(q)\Λpφq(Λp).\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z})\leq\Lambda^{p}(q)\backslash\Lambda^{p}\cong\varphi_{q}(\Lambda^{p}).

Let αΛp\alpha\in\Lambda^{p}. Then, N(α)=pkN(\alpha)=p^{k} for some kk\in\mathbb{Z}. Suppose that (pq)=1\left(\frac{p}{q}\right)=1. Then there exists xx\in\mathbb{Z} such that qxq\nmid x and px2(modq)p\equiv x^{2}\pmod{q}. Let ll\in\mathbb{Z} such that plx1(modq)p^{l}x\equiv 1\pmod{q}. Note that αpklα\alpha\sim p^{kl}\alpha in Λp\Lambda^{p}. Moreover,

N(plα)=p2klN(α)=p2klx2k1(modq).N(p^{l}\alpha)=p^{2kl}N(\alpha)=p^{2kl}x^{2k}\equiv 1\pmod{q}.

Therefore, φq(α)PSL2(/q)\varphi_{q}(\alpha)\in\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}). Now suppose that (pq)=1\left(\frac{p}{q}\right)=-1. Suppose that N(α)=pN(\alpha)=p and φq(α)PSL2(/q)\varphi_{q}(\alpha)\in\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}). Then there exists yy\in\mathbb{Z}, qyq\nmid y, such that N(yα)=y2p=1N(y\alpha)=y^{2}p=1 but this implies pp is a square modulo qq, which is a contradiction. Therefore, φq(α)PSL2(/q)\varphi_{q}(\alpha)\not\in\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}) and φq(Λp)PSL2(/q)\varphi_{q}(\Lambda^{p})\supsetneq\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}). But PSL2(/q)\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}) is an index 2 subgroup of PGL2(/q)\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z}), so Λp(q)\ΛpPGL2(/q)\Lambda^{p}(q)\backslash\Lambda^{p}\cong\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z}). ∎

Let qq be an odd prime such that (pq)=1\left(\frac{p}{q}\right)=1. Let 𝒮p,q=φq(𝒮p)\mathcal{S}^{p,q}=\varphi_{q}(\mathcal{S}^{p}). Note that when φq\varphi_{q} is not injective on 𝒮p\mathcal{S}^{p}, we consider 𝒮p,q\mathcal{S}^{p,q} to be a multi-set where each element has multiplicity of the size of its pre-image in 𝒮p\mathcal{S}^{p}. As shown above, we can consider this set to be a generating set of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}). Therefore, we have the Cayley graph

𝒳p,q=𝒳𝒪p,q(m,H)=Cay(PSL2(𝔽q),𝒮p,q).\mathcal{X}^{p,q}=\mathcal{X}^{p,q}_{\mathcal{O}}(m,H)=\operatorname{Cay}(\operatorname{PSL}_{2}(\mathbb{F}_{q}),\mathcal{S}^{p,q}).
Proposition 3.9.

When qq is an odd prime such that (pq)=1\left(\frac{p}{q}\right)=1, we have an isometry of graphs

𝒳p,qΛp(q)\𝒯p+1.\mathcal{X}^{p,q}\cong\Lambda^{p}(q)\backslash\mathcal{T}_{p+1}.
Proof.

As a subgroup of Λp\Lambda^{p}, the group Λp(q)\Lambda^{p}(q) acts simply on 𝒯p+1\mathcal{T}_{p+1}. By Proposition 3.8, the orbits of the action are in 1-1 correspondence with PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}). We can thus view the vertices of the quotient graph as elements of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}). We see this in the sequence of bijections of sets,

PSL2(𝔽q)Λp(q)\ΛpΛp(q)\PGL2(p)/PGL2(p)Λp(q)\V(𝒯p+1)\operatorname{PSL}_{2}(\mathbb{F}_{q})\cong\Lambda^{p}(q)\backslash\Lambda^{p}\cong\Lambda^{p}(q)\backslash\operatorname{PGL}_{2}(\mathbb{Q}_{p})/\operatorname{PGL}_{2}(\mathbb{Z}_{p})\cong\Lambda^{p}(q)\backslash V(\mathcal{T}_{p+1})

where V(𝒯p+1)V(\mathcal{T}_{p+1}) are the vertices of 𝒯p+1\mathcal{T}_{p+1}. By Proposition 3.6, 𝒮p\mathcal{S}^{p} generates Λp\Lambda^{p} so 𝒮p,q\mathcal{S}^{p,q} generates the quotient graph. ∎

Theorem 3.10.

[9, Theorem 7.3.1] Let 𝒪\mathcal{O} be a maximal order of class number 1 which splits at a prime pp. Let qpq\neq p be a prime such that (pq)=1\left(\frac{p}{q}\right)=1. Let (m,H)(m,H) be a congruence pair of 𝒪\mathcal{O}. Then the Cayley graph 𝒳𝒪p,q(m,H)\mathcal{X}^{p,q}_{\mathcal{\mathcal{O}}}(m,H) is a Ramanujan graph.

Note that #PSL2(𝔽q)=q(q1)(q+1)/2\#\operatorname{PSL}_{2}(\mathbb{F}_{q})=q(q-1)(q+1)/2 so the graph 𝒳𝒪p,q\mathcal{X}^{p,q}_{\mathcal{O}} has q(q1)(q+1)/2q(q-1)(q+1)/2 vertices.

4. Construction of Explicit Maximal Orders and LPS-Type Ramanujan Graphs

In this section, we explain Jo and Yamasaki’s construction of LPS-type graphs as in [6]. In the construction, we need a quaternion algebra of class number 1 and a maximal order. In [5], Ibukiyama explicitly constructs a maximal order for any definite quaternion algebra over \mathbb{Q}. We discuss a special case for when the quaternion algebra only ramifies at one prime.

Proposition 4.1 ([5]).

Let PP be a prime number. Take a prime QQ such that

(2) Q3mod8,(QP)=1 unless P=2.Q\equiv 3\mod 8,\quad\left(\frac{-Q}{P}\right)=-1\text{ unless }P=2.

Let TT be an integer such that T2PmodQT^{2}\equiv-P\mod Q. Then, P,Q\mathcal{H}_{-P,-Q} (i.e. i2=Pi^{2}=-P, j2=Qj^{2}=-Q, ij=ji=kij=-ji=k) is a definite quaternion algebra which only ramifies at PP and the quaternion algebra P,Q\mathcal{H}_{-P,-Q} has a maximal order 𝒪P,Q=ω1ω2ω3\mathcal{O}_{-P,-Q}=\mathbb{Z}\oplus\mathbb{Z}\omega_{1}\oplus\mathbb{Z}\omega_{2}\oplus\mathbb{Z}\omega_{3} where

ω1=1+j2,ω2=i+ij2,ω3=Tj+kQ.\omega_{1}=\frac{1+j}{2},\quad\omega_{2}=\frac{i+ij}{2},\quad\omega_{3}=\frac{Tj+k}{Q}.
Definition 4.2.

Let (P,Q)(P,Q) be a pair of positive integers such that PP is prime and QQ satisfies (2). We call (P,Q)(P,Q) an Ibukiyama Pair.

Note that by quadratic reciprocity, we can find an appropriate integer TT that satisfies the required conditions. By Corollary 2.3, the class number of P,Q\mathcal{H}_{-P,-Q} is 1 if and only if P{2,3,5,7,13}P\in\{2,3,5,7,13\}. Also, if α=x+yω1+zω2+wω3𝒪P,Q\alpha=x+y\omega_{1}+z\omega_{2}+w\omega_{3}\in\mathcal{O}_{-P,-Q} then

(3) N(α)=x2+Q+14y2+P(Q+1)4z2+T2+PQw2+xy+Tyw+Pzw.N(\alpha)=x^{2}+\frac{Q+1}{4}y^{2}+\frac{P(Q+1)}{4}z^{2}+\frac{T^{2}+P}{Q}w^{2}+xy+Tyw+Pzw.

Note that our choice of QQ and TT is irrelevant. We now make the following adjustments to our notation:

𝒢P,Q(A)=𝒢𝒪P,Q(A),ΛpP,Q(A)=Λp𝒪P,Q(A)\mathcal{G}_{-P,-Q}(A)=\mathcal{G}_{\mathcal{O}_{-P,-Q}}(A),\qquad\Lambda^{p}_{-P,-Q}(A)=\Lambda^{p}_{\mathcal{O}_{-P,-Q}}(A)
Proposition 4.3.

[6, Remark V.2] Let qq be a prime, q2q\neq 2, (Pq)=(Qq)=1\left(\frac{-P}{q}\right)=\left(\frac{Q}{q}\right)=1 and (pq)=1\left(\frac{p}{q}\right)=1. The map βq:𝒪P,Q2(𝔽q)\beta_{q}:\mathcal{O}_{-P,-Q}\rightarrow\mathcal{M}_{2}(\mathbb{F}_{q}) induced by the reduction modulo qq map on 𝒪P,Q\mathcal{O}_{-P,-Q} is defined by

x+yω1+zω2+wω312Q(A11A12A21A22)x+y\omega_{1}+z\omega_{2}+w\omega_{3}\mapsto\frac{1}{2Q}\begin{pmatrix}A_{11}&A_{12}\\ A_{21}&A_{22}\end{pmatrix}

where

A11\displaystyle A_{11} =(2Qx+Qy)+QzP\displaystyle=(2Qx+Qy)+Qz\sqrt{-P}
A12\displaystyle A_{12} =Q(Qy+2Tw+(Qz+2w)P)\displaystyle=\sqrt{Q}(Qy+2Tw+(Qz+2w)\sqrt{-P})
A21\displaystyle A_{21} =Q(Qy+2Tw(Qz+2w)P)\displaystyle=-\sqrt{Q}(Qy+2Tw-(Qz+2w)\sqrt{-P})
A22\displaystyle A_{22} =(2Qx+Qy)QzP\displaystyle=(2Qx+Qy)-Qz\sqrt{-P}

gives an isomorphism satisfying det(βq(α))=N(α)\det(\beta_{q}(\alpha))=N(\alpha).

Lemma 4.4.

Let (P,Q)(P,Q) be an Ibukiyama pair. Then

𝒢P,Q(/P)NH\mathcal{G}_{-P,-Q}(\mathbb{Z}/P\mathbb{Z})\cong N\rtimes H

where

N={1+xi+yij:x,y/P},H={x+yj:x,y/P,(x,y)(0,0)}/(/P)×.N=\left\{1+xi+yij:x,y\in\mathbb{Z}/P\mathbb{Z}\right\},\quad H=\left\{x+yj:x,y\in\mathbb{Z}/P\mathbb{Z},\ (x,y)\neq(0,0)\right\}/(\mathbb{Z}/P\mathbb{Z})^{\times}.
Proof.

Using (3), we see that the number of elements α=a0+a1ω1+a2ω2+a3ω3\alpha=a_{0}+a_{1}\omega_{1}+a_{2}\omega_{2}+a_{3}\omega_{3} with 0ai<P0\leq a_{i}<P and N(α)0(modP)N(\alpha)\not\equiv 0\pmod{P} is P2(P21)P^{2}(P^{2}-1). Therefore, #𝒢(/P)=P2(P21)/(P1)=P2(P+1)\#\mathcal{G}(\mathbb{Z}/P\mathbb{Z})=P^{2}(P^{2}-1)/(P-1)=P^{2}(P+1).

Because P2,QP\neq 2,Q, we can break up the ωi\omega_{i} to put the elements of 𝒢P,Q(/m)\mathcal{G}_{-P,-Q}(\mathbb{Z}/m\mathbb{Z}) in terms of i,j,iji,j,ij. It is clear that the elements of NN and HH have nonzero norm. Furthermore, NN and HH are both closed under multiplication. Thus, NN and HH can be identified as subgroups of 𝒢(/P)\mathcal{G}(\mathbb{Z}/P\mathbb{Z}). Observe that NH={1}N\cap H=\left\{1\right\}, #N=P2\#N=P^{2}, #H=(P21)/(P1)=P+1\#H=(P^{2}-1)/(P-1)=P+1. We have computed that #𝒢(/P)=P2(P+1)\#\mathcal{G}(\mathbb{Z}/P\mathbb{Z})=P^{2}(P+1), which shows that 𝒢(/P)=NH\mathcal{G}(\mathbb{Z}/P\mathbb{Z})=N\cdot H. Lastly, NN is normal in 𝒢(/P)\mathcal{G}(\mathbb{Z}/P\mathbb{Z}) because for nNn\in N and hHh\in H, we have that h1=zyjh^{-1}=z-yj and

hnh1=(x+yj)(1+Xi+Yij)(xyj)=(x+(xXyY)i+(xYyX)ij+yj)(xyj)=A+Bi+CijN.hnh^{-1}=(x+yj)(1+Xi+Yij)(x-yj)\\ =(x+(xX-yY)i+(xY-yX)ij+yj)(x-yj)=A+Bi+Cij\in N.

Therefore, 𝒢(/P)NH\mathcal{G}(\mathbb{Z}/P\mathbb{Z})\cong N\rtimes H. ∎

Definition 4.5.

An LPS-type graph is any graph constructed in the following way:

  1. (1)

    Fix a prime pp.

  2. (2)

    Let (P,Q)(P,Q) be an Ibukiyama pair such that P{2,3,5,7,13}P\in\left\{2,3,5,7,13\right\}, PpP\neq p. We have a definite class number 1 rational maximal order 𝒪P,Q\mathcal{O}_{-P,-Q} which splits at pp.

  3. (3)

    Find all elements in 𝒪×P,Q\mathcal{O}^{\times}_{-P,-Q} by solving the norm equation N(α)=1N(\alpha)=1.

  4. (4)

    Find p+1p+1 elements of 𝒪P,Q\mathcal{O}_{-P,-Q} of norm pp, unique up to units, which form a set 𝒮p\mathcal{S}^{p}.

  5. (5)

    Take a prime qq satisfying

    (4) qp,(Pq)=(Qq)=1,(pq)=1.q\neq p,\quad\left(\frac{-P}{q}\right)=\left(\frac{Q}{q}\right)=1,\quad\left(\frac{p}{q}\right)=1.
  6. (6)

    Via the reduction map φq\varphi_{q} then the isomorphism φq\varphi_{q} in Proposition 4.3, realize 𝒮p\mathcal{S}^{p} as a multi-set of elements of PSL2(𝔽q)\operatorname{PSL}_{2}(\mathbb{F}_{q}) and write it as 𝒮p,q\mathcal{S}^{p,q}. Note that each element of 𝒮p,q\mathcal{S}^{p,q} has multiplicity of the size of its pre-image in 𝒮p\mathcal{S}^{p}.

  7. (7)

    Construct the Cayley graph

    𝒳p,q=𝒳P,Qp,q(𝒮p,q)=PSL2(𝔽q,𝒮p,q).\mathcal{X}^{p,q}=\mathcal{X}_{-P,-Q}^{p,q}\left(\mathcal{S}^{p,q}\right)=\operatorname{PSL}_{2}\left(\mathbb{F}_{q},\mathcal{S}^{p,q}\right).

    The graph 𝒳p,q\mathcal{X}^{p,q} is a LPS-type graph.

Theorem 4.6.

Let (P,Q)(P,Q) be an Ibukiyama pair. Let 𝒮p\mathcal{S}^{p} be p+1p+1 elements of norm pp in 𝒪P,Q\mathcal{O}_{-P,-Q}. Let qpq\neq p be a prime such that qq satisifies ()(\ddagger). Let 𝒳p,qP,Q(𝒮p,q)\mathcal{X}^{p,q}_{-P,-Q}\left(\mathcal{S}^{p,q}\right) be the corresponding LPS-type graph. If we can find a congruence pair (m,H)(m,H) of 𝒪P,Q\mathcal{O}_{-P,-Q} such that 𝒮pΛpP,Q(m,H)\mathcal{S}^{p}\subseteq\Lambda^{p}_{-P,-Q}(m,H), then 𝒳p,qP,Q\mathcal{X}^{p,q}_{-P,-Q} is Ramanujan.

Proof.

We have that 𝒪P,Q\mathcal{O}_{-P,-Q} is a rational maximal class number one maximal order which splits at pp. Because 𝒮pΛpP,Q(m,H)\mathcal{S}^{p}\subseteq\Lambda^{p}_{-P,-Q}(m,H) for some congruence pair (m,H)(m,H), we have that 𝒳P,Qp,q(𝒮p,qP,Q)𝒳𝒪P,Qp,q(m,H)\mathcal{X}_{-P,-Q}^{p,q}\left(\mathcal{S}^{p,q}_{-P,-Q}\right)\cong\mathcal{X}_{\mathcal{O}_{-P,-Q}}^{p,q}(m,H). Theorem 3.10 implies that 𝒳P,Qp,q(m,H)\mathcal{X}_{-P,-Q}^{p,q}(m,H) is Ramanujan. ∎

Corollary 4.7.

Theorem 1.2 implies Theorem 1.1.

5. The Ramanujan Property for Certain LPS-Type Graphs

Our goal is to show that we can construct Ramanujan LPS-type graphs from definite class number 1 rational quaternion algebras. That is, we want to prove Theorem 1.2, which we restate more explicitly below in terms of congruence pairs. Note that by Proposition 3.5, we have that Theorem 5.1 is equivalent to Theorem 1.2.

Theorem 5.1.

Let \mathcal{H} be a definite class number 1 quaternion algebra over \mathbb{Q}. Let 𝒪\mathcal{O} be the unique maximal order of \mathcal{H}. There exists a positive integer mm and a subgroup HH of 𝒢𝒪(/m)\mathcal{G}_{\mathcal{O}}(\mathbb{Z}/m\mathbb{Z}) such that (m,H)(m,H) is a congruence pair.

To prove Theorem 5.1, we need the following calculation.

Proposition 5.2.

The unit groups of 𝒪P,Q\mathcal{O}_{-P,-Q} for P{2,3,4,7,13}P\in\left\{2,3,4,7,13\right\} are

𝒢P,Q()={A4P=2,S3P=3,A3P=5,S2P=7,A2P=13.\mathcal{G}_{-P,-Q}(\mathbb{Z})=\begin{cases}A_{4}&P=2,\\ S_{3}&P=3,\\ A_{3}&P=5,\\ S_{2}&P=7,\\ A_{2}&P=13.\end{cases}
Proof.

For P=2P=2, see Lemma 5.3. For P=3P=3, see Lemma 5.5. For P=5P=5, see Lemma 5.7. For P=7P=7, see Lemma 5.10. The case P=13P=13 is proven in [1, Section 4]. Note that A2A_{2} is the trivial group. ∎

Proof of Theorem 5.1.

For P=2P=2, see Proposition 5.4. For P=3P=3, see Proposition 5.6. For P=5P=5, see Proposition 5.9. For P=7P=7, see Proposition 5.11. For P=13P=13, A2A_{2} is the trivial group so 𝒢P,Q([1/p])\mathcal{G}_{-P,-Q}(\mathbb{Z}[1/p]) acts trivially on 𝒯p+1\mathcal{T}_{p+1}. ∎

Note that 𝒪P,Q\mathcal{O}_{-P,-Q} is completely determined by the ramified primes, so the choice of QQ does not matter.

5.1. The Quaternion Algebra Ramifies at 2

Set P=2P=2 and Q=11Q=11. Observe that Q3mod8Q\equiv 3\mod{8} (we don’t need the reciprocity condition because P=2P=2). Set T=3T=3. Observe that T22mod11T^{2}\equiv-2\mod{11}. 2,11\mathcal{H}_{-2,-11} is a quaternion algebra over \mathbb{Q} which is of class number 1 and ramifies only at 2. By Theorem 4.1, 2,11\mathcal{H}_{-2,-11} has a maximal order 𝒪2,11=+ω1+ω2+ω3\mathcal{O}_{-2,-11}=\mathbb{Z}+\mathbb{Z}\omega_{1}+\mathbb{Z}\omega_{2}+\mathbb{Z}\omega_{3}, where

ω1=1+j2,ω2=i+ij2,ω3=3j+ij11.\omega_{1}=\frac{1+j}{2},\quad\omega_{2}=\frac{i+ij}{2},\quad\omega_{3}=\frac{3j+ij}{11}.
Lemma 5.3.

𝒢2,11()A4\mathcal{G}_{-2,-11}(\mathbb{Z})\cong A_{4}

Proof.

For α=x+yω1+zω2+wω3𝒪2,11\alpha=x+y\omega_{1}+z\omega_{2}+w\omega_{3}\in\mathcal{O}_{-2,-11}, it follows from (3) that

N(α)=x2+3y2+6z2+w2+xy+3yw+2zw.N(\alpha)=x^{2}+3y^{2}+6z^{2}+w^{2}+xy+3yw+2zw.

Solving N(α)=1N(\alpha)=1, we get 2424 solutions in 4\mathbb{Z}^{4}, which form a group under multiplication. When we quotient by {±1}\left\{\pm 1\right\}, we get a group which is isomorphic to A4A_{4}. Representations of the elements of 𝒢2,11()\mathcal{G}_{-2,-11}(\mathbb{Z}) are enumerated in Figure 1. ∎

1 ω1\omega_{1} ω2\omega_{2} ω3\omega_{3} 1 ii jj ijij
1 0 0 1 0 0 3/11 1/11
0 1 0 -2 1/2 0 -1/22 -2/11
0 1 0 -1 1/2 0 5/22 -1/11
1 -3 -1 5 -1/2 -1/2 -3/22 -1/22
1 -3 -1 6 -1/2 -1/2 3/22 1/22
1 -2 -1 4 0 -1/2 1/11 -3/22
1 -1 0 1 1/2 0 -5/22 1/11
1 -1 0 2 1/2 0 1/22 2/11
1 0 0 0 1 0 0 0
2 -4 -1 7 0 -1/2 -1/22 3/22
2 -3 -1 5 1/2 -1/2 -3/22 -1/22
2 -3 -1 6 1/2 -1/2 3/22 1/22
Figure 1. The elements of 𝒢2,11()\mathcal{G}_{-2,-11}(\mathbb{Z}). Each row is a distinct element with a representation in terms of ωi\omega_{i} and in terms of ii and jj.
Proposition 5.4.

There exists a subgroup H𝒢2,11(/3)H\leq\mathcal{G}_{-2,-11}(\mathbb{Z}/3\mathbb{Z}) such that HC2H\cong C_{2} and (3,H)(3,H) is a congruence pair of 𝒪2,11\mathcal{O}_{-2,-11}.

Proof.

We see from Figure 1 that 𝒢2,11()\mathcal{G}_{-2,-11}(\mathbb{Z}) embeds in 𝒢2,11(/3)\mathcal{G}_{-2,-11}(\mathbb{Z}/3\mathbb{Z}). We identify 𝒢2,11()\mathcal{G}_{-2,-11}(\mathbb{Z}) as a subgroup of 𝒢2,11(/3)\mathcal{G}_{-2,-11}(\mathbb{Z}/3\mathbb{Z}). Lemma 5.3 tells us that 𝒢3,19()A4\mathcal{G}_{-3,-19}(\mathbb{Z})\cong A_{4}. Proposition 3.7 implies that 𝒢2,11(/3)PGL2(/3)S4\mathcal{G}_{-2,-11}(\mathbb{Z}/3\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/3\mathbb{Z})\cong S_{4}. There is only one copy of A4A_{4} in S4S_{4} and

S4=A4HS_{4}=A_{4}\rtimes H

for HC2H\cong C_{2}. Therefore, (3,H)(3,H) is a congruence pair of 𝒪2,11\mathcal{O}_{-2,-11}. ∎

5.2. The Quaternion Algebra Ramifies at 3

Set P=3P=3 and Q=19Q=19. Observe that Q3mod8Q\equiv 3\mod{8} and (Q3)=1\left(\frac{-Q}{3}\right)=-1. Set T=4T=4. Observe that T23mod19T^{2}\equiv-3\mod{19}. Thus, 3,19\mathcal{H}_{-3,-19} is a quaternion algebra over \mathbb{Q} which is of class number 1 and ramifies only at 3. By Theorem 4.1, 3,19\mathcal{H}_{-3,-19} has a maximal order 𝒪3,19=+ω1+ω2+ω3\mathcal{O}_{-3,-19}=\mathbb{Z}+\mathbb{Z}\omega_{1}+\mathbb{Z}\omega_{2}+\mathbb{Z}\omega_{3}, where

ω1=1+j2,ω2=i+ij2,ω3=4j+ij19.\omega_{1}=\frac{1+j}{2},\quad\omega_{2}=\frac{i+ij}{2},\quad\omega_{3}=\frac{4j+ij}{19}.
Lemma 5.5.

𝒢3,19()S3\mathcal{G}_{-3,-19}(\mathbb{Z})\cong S_{3}

Proof.

For α=x+yω1+zω2+wω3𝒪3,19\alpha=x+y\omega_{1}+z\omega_{2}+w\omega_{3}\in\mathcal{O}_{-3,-19}, it follows from (3) that

N(α)=x2+5y2+15z2+w2+xy+4yw+3zw.N(\alpha)=x^{2}+5y^{2}+15z^{2}+w^{2}+xy+4yw+3zw.

Solving N(α)=1N(\alpha)=1, we get 12 solutions in 4\mathbb{Z}^{4}, which form a group under multiplication. When we quotient by {±1}\left\{\pm 1\right\}, we get a noncommutative group with six elements. There is only one noncommutative group of order 6, S3S_{3}, so 𝒢3,19()S3\mathcal{G}_{-3,-19}(\mathbb{Z})\cong S_{3}. Representations of the elements of 𝒢3,19()\mathcal{G}_{-3,-19}(\mathbb{Z}) are enumerated in Figure 2. ∎

1 ω1\omega_{1} ω2\omega_{2} ω3\omega_{3} 1 ii jj ijij
1 0 0 0 1 0 0 0
2 -4 -1 10 0 -1/2 2/19 1/38
2 -4 -1 9 0 -1/2 -2/19 -1/38
1 -1 0 2 1/2 0 -3/38 2/19
0 1 0 -2 1/2 0 3/38 -2/19
0 0 0 1 0 0 4/19 1/19
Figure 2. Elements of 𝒢3,19()\mathcal{G}_{-3,-19}(\mathbb{Z}).
Proposition 5.6.

There exist subgroups H,H,H𝒢3,19(/4)H,H^{\prime},H^{\prime\prime}\leq\mathcal{G}_{-3,-19}(\mathbb{Z}/4\mathbb{Z}) such that

HC23,HC4×C22,HD4H\cong C_{2}^{3},\quad H^{\prime}\cong C_{4}\times C_{2}^{2},\quad H^{\prime\prime}\cong D_{4}

and (4,H)(4,H), (4,H)(4,H^{\prime}), and (4,H)(4,H^{\prime\prime}) are congruence pairs of 𝒪3,19\mathcal{O}_{-3,-19}.

Proof.

We see from Figure 2 that 𝒢3,19()\mathcal{G}_{-3,-19}(\mathbb{Z}) embeds in 𝒢3,19(/4)\mathcal{G}_{-3,-19}(\mathbb{Z}/4\mathbb{Z}). We identify 𝒢3,19()\mathcal{G}_{-3,-19}(\mathbb{Z}) as a subgroup of 𝒢3,19(/4)\mathcal{G}_{-3,-19}(\mathbb{Z}/4\mathbb{Z}). Proposition 3.7 implies that 𝒢3,19(/4)PGL2(/4)\mathcal{G}_{-3,-19}(\mathbb{Z}/4\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/4\mathbb{Z}). Lemma 5.5 tells us that 𝒢3,19()S3\mathcal{G}_{-3,-19}(\mathbb{Z})\cong S_{3}. Calculation shows that there are two classes [K],[K]CPGL2(/4)[K],[K^{\prime}]\in C_{\operatorname{PGL}_{2}(\mathbb{Z}/4\mathbb{Z})} such that K,KS3K,K^{\prime}\cong S_{3}. Both KK and KK^{\prime} have complementary subgroups isomorphic to C8C_{8}, C4×C22C_{4}\times C_{2}^{2}, and D4D_{4}. ∎

5.3. The Quaternion Algebra Ramifies at 5

Set P=5P=5 and Q=3Q=3. Observe that Q3mod8Q\equiv 3\mod{8} and (Q5)=1\left(\frac{-Q}{5}\right)=-1. Set T=1T=1. Observe that T25mod3T^{2}\equiv-5\mod{3}. 5,3\mathcal{H}_{-5,-3} is a quaternion algebra over \mathbb{Q} which is of class number 1 and ramifies only at 5. By Theorem 4.1, 5,3\mathcal{H}_{-5,-3} has a maximal order 𝒪2,11=+ω1+ω2+ω3\mathcal{O}_{-2,-11}=\mathbb{Z}+\mathbb{Z}\omega_{1}+\mathbb{Z}\omega_{2}+\mathbb{Z}\omega_{3}, where

ω1=1+j2,ω2=i+ij2,ω3=j+ij3.\omega_{1}=\frac{1+j}{2},\quad\omega_{2}=\frac{i+ij}{2},\quad\omega_{3}=\frac{j+ij}{3}.
Lemma 5.7.

𝒢3,19()1,ω1,1ω1A3\mathcal{G}_{-3,-19}(\mathbb{Z})\cong\langle 1,\omega_{1},1-\omega_{1}\rangle\cong A_{3}

Proof.

For α=x+yω1+zω2+wω3𝒪5,3\alpha=x+y\omega_{1}+z\omega_{2}+w\omega_{3}\in\mathcal{O}_{-5,-3}, it follows from (3) that

N(α)=x2+y2+5z2+2w2+xy+yw+5zw.N(\alpha)=x^{2}+y^{2}+5z^{2}+2w^{2}+xy+yw+5zw.

Solving N(α)=1N(\alpha)=1, we get 66 solutions in 4\mathbb{Z}^{4}, which form a group under multiplication. When we quotient by {±1}\left\{\pm 1\right\}, we get a group which is isomorphic to A3A_{3}. ∎

Lemma 5.8.

Let

A={1+xi+yij:x,y/5},B={j+xi+yij:x,y/5},\displaystyle A=\left\{1+xi+yij:x,y\in\mathbb{Z}/5\mathbb{Z}\right\},\quad B=\left\{j+xi+yij:x,y\in\mathbb{Z}/5\mathbb{Z}\right\},
H=AB.\displaystyle H=A\cup B.

The set HH is a group isomorphic with C5D5C_{5}\rtimes D_{5}.

Proof.

We first check that HH contains inverses. Suppose that α=1+xi+yijA\alpha=1+xi+yij\in A. Then

α(1xiyij)=1+x2i2+y2ijij=1.\alpha(1-xi-yij)=1+x^{2}i^{2}+y^{2}ijij=1.

Therefore, α1A\alpha^{-1}\in A. Next suppose that α=j+xi+yijB\alpha=j+xi+yij\in B, then

α2\displaystyle\alpha^{2} =j2+xji+yjij+xij+x2i2+xyi2j+yij2+xyiji+y2ijij\displaystyle=j^{2}+xji+yjij+xij+x^{2}i^{2}+xyi^{2}j+yij^{2}+xyiji+y^{2}ijij
=3xij+3yi+xij3yi\displaystyle=-3-xij+3yi+xij-3yi
=3.\displaystyle=-3.

In particular, every element of BB has order 2. Therefore, α1=α\alpha^{-1}=\alpha for αB\alpha\in B. We now show that HH is a group. We check the following cases.

  1. (1)

    Let α,βA\alpha,\beta\in A. Let α=1+xi+yij\alpha=1+xi+yij and β=1+xi+yij\beta=1+x^{\prime}i+y^{\prime}ij. Then

    αβ\displaystyle\alpha\beta =(1+xiyij)(1+xi+yij)\displaystyle=(1+xi_{y}ij)(1+x^{\prime}i+y^{\prime}ij)
    =1+xi+yij+xi+xxi2+xyi2j+yij+yxiji+yyijij\displaystyle=1+x^{\prime}i+y^{\prime}ij+xi+xx^{\prime}i^{2}+xy^{\prime}i^{2}j+yij+yx^{\prime}iji+yy^{\prime}ijij
    =1+xi+yij+xi5xx5xyj+yij+5yxj15yy\displaystyle=1+x^{\prime}i+y^{\prime}ij+xi-5xx^{\prime}-5xy^{\prime}j+yij+5yx^{\prime}j-15yy^{\prime}
    =1+(x+x)i+(y+y)ij.\displaystyle=1+(x+x^{\prime})i+(y+y^{\prime})ij.

    Therefore, αβA\alpha\beta\in A.

  2. (2)

    Let α,βB\alpha,\beta\in B. Let α=j+xi+yij\alpha=j+xi+yij and β=j+xk+yij\beta=j+x^{\prime}k+y^{\prime}ij. Then

    αβ\displaystyle\alpha\beta =(j+xi+yij)(j+xi+yij)\displaystyle=(j+xi+yij)(j+x^{\prime}i+y^{\prime}ij)
    =j2+xji+yjij+xij+xxi2+xyi2j+yxiji+yyiji\displaystyle=j^{2}+x^{\prime}ji+y^{\prime}jij+xij+xx^{\prime}i^{2}+xy^{\prime}i^{2}j+yx^{\prime}iji+yy^{\prime}iji
    =3+xji+y3i+xij3yi\displaystyle=-3+x^{\prime}ji+y^{\prime}3i+xij-3yi
    =3+3(yy)i+(xx)ij.\displaystyle=-3+3(y^{\prime}-y)i+(x-x^{\prime})ij.

    Therefore, αβA\alpha\beta\in A.

  3. (3)

    Let αA\alpha\in A and βB\beta\in B. Let α=1+xi+yij\alpha=1+xi+yij and β=j+xi+yij\beta=j+x^{\prime}i+y^{\prime}ij. Then

    αβ\displaystyle\alpha\beta =(1+xi+yij)(j+xi+yij)\displaystyle=(1+xi+yij)(j+x^{\prime}i+y^{\prime}ij)
    =j+xi+yij+xij+xxi2+xyi2j+yij2+yxiji+yyijij\displaystyle=j+x^{\prime}i+y^{\prime}ij+xij+xx^{\prime}i^{2}+xy^{\prime}i^{2}j+yij^{2}+yx^{\prime}iji+yy^{\prime}ijij
    =j+xi+yij+xij3yi\displaystyle=j+x^{\prime}i+y^{\prime}ij+xij-3yi
    =j+(x3y)i+(x+y)ij.\displaystyle=j+(x^{\prime}-3y)i+(x+y^{\prime})ij.

    Therefore, αβB\alpha\beta\in B.

  4. (4)

    Let αB\alpha\in B and βA\beta\in A. Let α=j+xi+yij\alpha=j+xi+yij and β=1+xi+yij\beta=1+x^{\prime}i+y^{\prime}ij. Then

    αβ\displaystyle\alpha\beta =(j+xi+yij)(1+xi+yij)\displaystyle=(j+xi+yij)(1+x^{\prime}i+y^{\prime}ij)
    =j+xji+yjij+xi+xxi2+xyi2j+yij+yxiji+yyijij\displaystyle=j+x^{\prime}ji+y^{\prime}jij+xi+xx^{\prime}i^{2}+xy^{\prime}i^{2}j+yij+yx^{\prime}iji+yy^{\prime}ijij
    =jxij+3yi+xi+yij\displaystyle=j-x^{\prime}ij+3y^{\prime}i+xi+yij
    =j+(3y+x)+(yx)ij.\displaystyle=j+(3y^{\prime}+x)+(y-x^{\prime})ij.

    Therefore, αβB\alpha\beta\in B.

Therefore, HH is in fact a group. HH is noncommutative of order 50, and HH contains 2525 elements of order 2. Therefore, HC5D5H\cong C_{5}\rtimes D_{5}. ∎

Proposition 5.9.

There exists a subgroup H𝒢5,3()H\leq\mathcal{G}_{-5,-3}(\mathbb{Z}) such that

HC5D5.H\cong C_{5}\rtimes D_{5}.

and (5,H)(5,H) is a congruence pair of 𝒪5,3\mathcal{O}_{-5,-3}.

Proof.

We see from Lemma 5.9 that 𝒢5,3()A3\mathcal{G}_{-5,-3}(\mathbb{Z})\cong A_{3} embeds in 𝒢5,3(/5)\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z}). We have that

φ5(𝒢5,3())1,1+j,1j.\varphi_{5}(\mathcal{G}_{-5,-3}(\mathbb{Z}))\cong\langle 1,1+j,1-j\rangle.

By Lemma 4.4, #𝒢5,3(/5)=150\#\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z})=150. Let H𝒢5,3(/5)H\subseteq\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z}) be as in Lemma 5.8, so HH is a subgroup of 𝒢5,3(/5)\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z}) isomorphic to C5D5C_{5}\rtimes D_{5}. By construction, H𝒢5,3()={1}H\cap\mathcal{G}_{-5,-3}(\mathbb{Z})=\left\{1\right\} and

#(𝒢5,3()H)=350=150=#𝒢5,3(/5)\#\left(\mathcal{G}_{-5,-3}(\mathbb{Z})\cdot H\right)=3\cdot 50=150=\#\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z})

so 𝒢5,3(/5)=𝒢5,3()H\mathcal{G}_{-5,-3}(\mathbb{Z}/5\mathbb{Z})=\mathcal{G}_{-5,-3}(\mathbb{Z})\cdot H. Therefore, (5,H)(5,H) is a congruence pair of 𝒪5,3\mathcal{O}_{-5,-3}. ∎

5.4. The Quaternion Algebra Ramifies at 7

Set P=7P=7 and Q=11Q=11. Observe that Q3mod8Q\equiv 3\mod{8} and (Q3)=1\left(\frac{-Q}{3}\right)=-1. Set T=2T=2. Observe that T23mod19T^{2}\equiv-3\mod{19}. 7,11\mathcal{H}_{-7,-11} is a quaternion algebra over \mathbb{Q} which is of class number 1 and ramifies only at 7. By Theorem 4.1, 3,19\mathcal{H}_{-3,-19} has a maximal order 𝒪3,19=+ω1+ω2+ω3\mathcal{O}_{-3,-19}=\mathbb{Z}+\mathbb{Z}\omega_{1}+\mathbb{Z}\omega_{2}+\mathbb{Z}\omega_{3}, where

ω1=1+j2,ω2=i+ij2,ω3=2j+ij11.\omega_{1}=\frac{1+j}{2},\quad\omega_{2}=\frac{i+ij}{2},\quad\omega_{3}=\frac{2j+ij}{11}.
Lemma 5.10.

𝒢7,11()1,ω3A2\mathcal{G}_{-7,-11}(\mathbb{Z})\cong\langle 1,\omega_{3}\rangle\cong A_{2}

Proof.

For α=x+yω1+zω2+wω3𝒪7,11\alpha=x+y\omega_{1}+z\omega_{2}+w\omega_{3}\in\mathcal{O}_{-7,-11}, it follows from (3) that

N(α)=x2+3y2+21z2+w2+xy+2yw+7zw.N(\alpha)=x^{2}+3y^{2}+21z^{2}+w^{2}+xy+2yw+7zw.

Solving N(α)=1N(\alpha)=1, we get 44 solutions in 4\mathbb{Z}^{4} which form a group under multiplication. The solutions to N(α)=1N(\alpha)=1 are ±1,±ω3\pm 1,\pm\omega_{3}. When we quotient by {±1}\left\{\pm 1\right\}, we get the group A2A_{2}. ∎

Proposition 5.11.

There exists a subgroup H𝒢7,11(/2)H\leq\mathcal{G}_{-7,-11}(\mathbb{Z}/2\mathbb{Z}) such that HC3H\cong C_{3} and (H,2)(H,2) is a congruence pair of 𝒪7,11\mathcal{O}_{-7,-11}.

Proof.

Lemma 5.10 tells us that 𝒢7,11()A2\mathcal{G}_{-7,-11}(\mathbb{Z})\cong A_{2} embeds in 𝒢7,11(/2)\mathcal{G}_{-7,-11}(\mathbb{Z}/2\mathbb{Z}) and Proposition 3.7 implies that 𝒢7,11(/2)PGL2(/2)S3\mathcal{G}_{-7,-11}(\mathbb{Z}/2\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/2\mathbb{Z})\cong S_{3}. There is only one class [K]CS3[K]\in C_{S_{3}} such that KA2K\cong A_{2}. Furthermore, there exists HS3H\leq S_{3} such that HC3H\cong C_{3} and HH is a complementary subgroup of C2C_{2}. Therefore, (2,H)(2,H) is a congruence pair of 𝒪7,11\mathcal{O}_{-7,-11}. ∎

6. Further Directions for Finding Congruence Pairs

Theorem 4.6 tells us that a LPS-Type graph 𝒳\mathcal{X} is Ramanujan if we can find a congruence pair whose corresponding congruence subgroup contains the generating set of 𝒳\mathcal{X}. Therefore, it is of some interest to know what the congruence pairs of a given class number 1 maximal order are.

Question 6.1.

What are all the possible congruence pairs for class number one quaternion algebras?

We take some steps in answering this question. In the cases P=3P=3 and P=7P=7, we show that there is no odd prime mm for which there exists a congruence pair (m,H)(m,H) of the respective maximal order. For the case P=3P=3, we use Propostion 6.2 to bound the possible mm for congruence pairs (m,H)(m,H) of 𝒪P,Q\mathcal{O}_{-P,-Q}. For a full treatment of Proposition 6.2, see [3, Chapter 12] or [7, Corollary 2.3]. While we only show the case P=3P=3 and P=7P=7, Proposition 6.2 can be used for P=2P=2 and P=5P=5 to bound mm.

Proposition 6.2.

For q>3q>3 an odd prime, the maximal subgroups of PGL2(/q)\operatorname{PGL}_{2}(\mathbb{Z}/q\mathbb{Z}) are as follows:

  1. (1)

    dihedral groups of order 2(q1)2(q-1) for q>5q>5,

  2. (2)

    dihedral groups of order 2(q+1)2(q+1),

  3. (3)

    a group of order q(q1)q(q-1),

  4. (4)

    PSL2(/q)\operatorname{PSL}_{2}(\mathbb{Z}/q\mathbb{Z}),

  5. (5)

    S4S_{4} when q±3(mod8)q\equiv\pm 3\pmod{8}.

Proposition 6.3.

There is no congruence pair (m,H)(m,H) associated with 𝒪3,19\mathcal{O}_{-3,-19} for odd primes mm.

Proof.

We consider the following cases:

  1. (1)

    Let m=3m=3. Let NN and HH be as in Proposition 4.4, i.e.,

    N={1+xi+yij:x,y/3},H={x+yj:x,y/3:(x,y)(0,0)}//3×.N=\{1+xi+yij:x,y\in\mathbb{Z}/3\mathbb{Z}\},\quad H=\{x+yj:x,y\in\mathbb{Z}/3\mathbb{Z}:(x,y)\not=(0,0)\}/\mathbb{Z}/3\mathbb{Z}^{\times}.

    Then, #N=9\#N=9, #H=4\#H=4, and by Lemma 3.7, 𝒢3,19(/3)=HN\mathcal{G}_{-3,-19}(\mathbb{Z}/3\mathbb{Z})=H\rtimes N. Therefore, #𝒢3,19(/3)=36\#\mathcal{G}_{-3,-19}(\mathbb{Z}/3\mathbb{Z})=36. Suppose that we can find K𝒢3,19(/3)K\leq\mathcal{G}_{-3,-19}(\mathbb{Z}/3\mathbb{Z}) that is a complement of 𝒢3,19()\mathcal{G}_{-3,-19}(\mathbb{Z}) in 𝒢3,19(/3)\mathcal{G}_{-3,-19}(\mathbb{Z}/3\mathbb{Z}). We have that #𝒢()\#\mathcal{G}(\mathbb{Z}) has order 6. Thus, #K=6\#K=6, so KK has an element of order 4. We have that 2ω1=1+j𝒢3,19(/3)2\omega_{1}=1+j\in\mathcal{G}_{-3,-19}(\mathbb{Z}/3\mathbb{Z}) has order 4. Therefore, 1+j𝒢3,19()1+j\not\in\mathcal{G}_{-3,-19}(\mathbb{Z}) and 1+jK1+j\not\in K. We have that

    φ3(𝒢3,19())={1,i+2j+2ij,i+j+ij,1+2ij,1+ij,j+ij}.\varphi_{3}(\mathcal{G}_{-3,-19}(\mathbb{Z}))=\left\{1,i+2j+2ij,i+j+ij,1+2ij,1+ij,j+ij\right\}.

    One can check that for all α𝒢3,19()\alpha\in\mathcal{G}_{-3,-19}(\mathbb{Z}), α1(1+j)\alpha^{-1}(1+j) has order 4. Therefore, 1+jαβ1+j\neq\alpha\beta for α𝒢3,19()\alpha\in\mathcal{G}_{-3,-19}(\mathbb{Z}) and βK\beta\in K. This is a contradiction.

  2. (2)

    Let m=5m=5. One can check that every subgroup isomorphic to S3S_{3} of PGL2(/5)S5\operatorname{PGL}_{2}(\mathbb{Z}/5\mathbb{Z})\cong S_{5} does not have a complementary subgroup.

  3. (3)

    Let m>5m>5. Using Proposition 6.2, we see that for any prime m>5m>5, there is no index 6 subgroup of PGL2(/m)\operatorname{PGL}_{2}(\mathbb{Z}/m\mathbb{Z}).

Proposition 6.4.

There is no congruence pair (m,H)(m,H) associated with 𝒪7,11\mathcal{O}_{-7,-11} for odd primes mm.

Proof.

Case 1 (m=7m=7): Let NN and HH be as in Proposition 3.7, i.e.,

N={1+xi+yij:x,y/7},H={x+yj:x,y/7:(x,y)(0,0)}//7×.N=\left\{1+xi+yij:x,y\in\mathbb{Z}/7\mathbb{Z}\right\},\quad H=\left\{x+yj:x,y\in\mathbb{Z}/7\mathbb{Z}:(x,y)\neq(0,0)\right\}/\mathbb{Z}/7\mathbb{Z}^{\times}.

Then #N=49\#N=49, #H=8\#H=8, and by Proposition 3.7, 𝒢7,11(/7)=HN\mathcal{G}_{-7,-11}(\mathbb{Z}/7\mathbb{Z})=H\rtimes N. Therefore, #𝒢7,11(/7)=392\#\mathcal{G}_{-7,-11}(\mathbb{Z}/7\mathbb{Z})=392. Suppose that we can find KK that is the complement of 𝒢7,11()\mathcal{G}_{-7,-11}(\mathbb{Z}) in 𝒢7,11(/7)\mathcal{G}_{-7,-11}(\mathbb{Z}/7\mathbb{Z}). Then #K=196\#K=196. Therefore, KK does not have any element of order 88. Therefore, 1+jK1+j\not\in K but 1+j𝒢()K1+j\in\mathcal{G}(\mathbb{Z})\cdot K. We have that

ω3=2j+ij114j+2ij(mod7)\omega_{3}=\frac{2j+ij}{11}\equiv 4j+2ij\pmod{7}

and we quotient by scalers, so φ7(ω3)=2j+ij\varphi_{7}(\omega_{3})=2j+ij. Therefore,

(2j+ij)1(1+j)=(2j+ij)(1+j)=2211i+2j+ij1+3i2j+ij(mod7).(2j+ij)^{-1}(1+j)=(2j+ij)(1+j)=-22-11i+2j+ij\equiv-1+3i-2j+ij\pmod{7}.

One can check that 1+3i2j+ijK-1+3i-2j+ij\in K has order 8, which is a contradiction. Therefore, it is impossible to find KK that satisfies our desired conditions.

Case 3 (m7m\not=7): Observe that 1ω3(modm)1\not\equiv\omega_{3}\pmod{m}, so 𝒢7,11()\mathcal{G}_{-7,-11}(\mathbb{Z}) embeds into 𝒢7,11(/m)\mathcal{G}_{-7,-11}(\mathbb{Z}/m\mathbb{Z}). By Proposition 3.7, 𝒢7,11(/m)PGL2(/m)\mathcal{G}_{-7,-11}(\mathbb{Z}/m\mathbb{Z})\cong\operatorname{PGL}_{2}(\mathbb{Z}/m\mathbb{Z}). Note that 𝒢7,11()\mathcal{G}_{-7,-11}(\mathbb{Z}) maps to to a cyclic group of order 2 in PSL2(/m)\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}). Suppose that we can find KK that is the complement of 𝒢7,11()\mathcal{G}_{-7,-11}(\mathbb{Z}) in 𝒢7,11(/m)\mathcal{G}_{-7,-11}(\mathbb{Z}/m\mathbb{Z}). Let K¯=KPSL2(/m)\overline{K}=K\cap\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}). Then, 𝒢7,11()K¯=PSL2(/m)\mathcal{G}_{-7,-11}(\mathbb{Z})\cdot\overline{K}=\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}) and 𝒢7,11()K¯={1}\mathcal{G}_{-7,-11}(\mathbb{Z})\cap\overline{K}=\left\{1\right\}. However, this implies that K¯\overline{K} has index 22 as a subgroup of PSL2(/m)\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}) which implies that K¯\overline{K} is normal in PSL2(/m)\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}). This is a contradiction because PSL2(/m)\operatorname{PSL}_{2}(\mathbb{Z}/m\mathbb{Z}) is simple for odd primes m3m\geq 3. ∎

Question 6.5.

Let XX and YY be two Ramanujan graphs constructed from a Quaternion algebra. Suppose that XX and YY have the same degree and same number of vertices. Are XX and YY isomorphic? In particular, let 𝒮p,p𝒪P,Q\mathcal{S}^{p},\mathcal{R}^{p}\subseteq\mathcal{O}_{-P,-Q} be sets with p+1p+1 elements of norm pp which satisfy the conditions of Theorem 4.6. Then does there exist an isomorphism

𝒳P,Qp,q(𝒮p,q)𝒳P,Qp,q(p,q)?\mathcal{X}_{-P,-Q}^{p,q}\left(\mathcal{S}^{p,q}\right)\cong\mathcal{X}_{-P,-Q}^{p,q}\left(\mathcal{R}^{p,q}\right)?
Question 6.6.

In Table 8.2 of [8], Kirschmer and Voight classify all definite class number one Eichler orders, which gives all the class number one maximal orders of quaternion algebras over a number field. For which of these orders can one find a congruence pair?

Acknowledgements

We would like to thank Professor Shai Evra and the Einstein Institute of Mathematics REU for their support during our research project. Professor Evra provided invaluable guidance and feedback, and the REU program’s funding and resources were essential to our success. We are grateful for the experience and knowledge gained from this opportunity.

References

  • [1] Patrick Chiu “Cubic Ramanujan graphs” In Combinatorica 12, 1992, pp. 275–285
  • [2] Giuliana Davidoff, Peter Sarnak and Alain Valette “Elementary Number Theory, Group Theory and Ramanujan Graphs”, London Mathematical Society Student Texts Cambridge: Cambridge University Press, 2003
  • [3] Leonard Eugene Dickson “Linear Groups, with an Exposition of the Galois Field Theory”, 1958
  • [4] Martin Eichler “The Basis Problem for Modular Forms and the Traces of the Hecke Operators” In Proc. of the Conf. on Modular Functions of One Variable V Berlin, Heidelberg: Springer Berlin Heidelberg, 1973, pp. 75–152
  • [5] Tomoyoshi Ibukiyama “On maximal orders of division quaternion algebras over the rational number field with certain optimal embeddings” In Nagoya Math. J. 88, 1982, pp. 181–195
  • [6] Hyungrok Jo and Yoshinori Yamasaki “LPS-type Ramanujan graphs” In 2018 International Symposium on Information Theory and Its Applications (ISITA) IEEE, 2018, pp. 399–403
  • [7] Oliver H. King “The subgroup structure of finite classical groups in terms of geometric configurations” In Surveys in Combinatorics 2005 327 Cambridge: Cambridge University Press, 2005, pp. 29–56
  • [8] Markus Kirschmer and John Voight “Algorithmic Enumeration of Ideal Classes for Quaternion Orders” In SIAM Journal on Computing 39.5, 2010, pp. 1714–1747
  • [9] Alexander Lubotzky “Discrete Groups, Expanding Graphs and Invariant Measures” Birkhäuser Basel, 1994
  • [10] Alexander Lubotzky, Ralph Phillips and Peter Sarnak “Ramanujan graphs” In Combinatorica 8, 1988, pp. 261–277
  • [11] Andrei Rapinchuk “On strong approximation for algebraic groups” In MSRI Publications 61, 2012, pp. 211–252
  • [12] Jean-Pierre Serre “Trees” Springer Berlin Heidelberg, 1980
  • [13] John Voight “Quaternion Algebras” 288, Graduate Texts in Mathematics Springer, 2021