betaFeedback

WorksMathematics

Graded version of local-global principle for Steinberg and Chevalley groups

Abstract

In this article, an analog of a graded version of Quillen-Suslin's local-global principle was established for Steinberg groups St(Φ,−)St(Φ,-) and Chevalley groups G(Φ,−)G(Φ,-) over a reduced irreducible root system ΦΦ.

Introduction

In the context of classical K-theory, the local-global principle was introduced by Quillen [10] and Suslin [14] as part of their solution to Serre’s problem on projective modules over polynomial extensions of a field. This principle was initially applied to prove Serre’s conjecture. The solution to Serre’s problem ultimately relied on two key analogs: the K2K_2-analog of Quillen’s local-global principle, which deals with projective modules, and the K1K_1-analog of Suslin’s local-global principle, which deals with elementary matrices.

Recall that for a commutative ring RR, denote G(Φ,R)G(\Phi,R) as a simply connected Chevalley group (see [13]) and St⁡(Φ,R)\operatorname{St}(\Phi,R) as a Steinberg group (see Definition 2.1) with a reduced irreducible root system Φ\Phi. The subgroup of G(Φ,R)G(\Phi,R) generated by the elementary root unipotents eα(r)e_\alpha(r) for all α∈Φ\alpha\in\Phi and r∈Rr \in R is called the elementary subgroup of G(Φ,R)G(\Phi,R) and is denoted by E(Φ,R)E(\Phi,R). There is a well-defined map ϕ ⁣:St⁡(Φ,R)→G(Φ,R)\phi\colon\operatorname{St}(\Phi,R) \to G(\Phi,R) sending xα(r)↦eα(r)x_\alpha(r) \mapsto e_\alpha(r), and the image of ϕ\phi is E(Φ,R)E(\Phi,R), which is a normal subgroup of G(Φ,R)G(\Phi,R) when rank⁡(Φ)≥2\operatorname{rank}(\Phi) \ge2. The groups K1(Φ,R)K_1(\Phi,R) and K2(Φ,R)K_2(\Phi,R) are defined as the cokernel and kernel of ϕ\phi, respectively

1⟶K2(Φ,R)⟶St⁡(Φ,R)→ϕG(Φ,R)⟶K1(Φ,R)⟶1.1 \longrightarrow K_2(\Phi,R) \longrightarrow\operatorname{St}(\Phi,R) \xrightarrow{\phi} G(\Phi,R) \longrightarrow K_1(\Phi,R) \longrightarrow1.

The K1K_1-analog of the local-global principle was established for Chevalley groups of type AnA_n by Suslin [14]. Subsequent work extended it to type DnD_n by Suslin–Kopeiko [15], to type CnC_n by Grunewald–Mennicke–Vaserstein [4], and to all Chevalley groups by Abe [1] and Wendt [18]. Turning to the K2K_2-analog of the local-global principle, it was proven for Chevalley groups of type AnA_n by Tulenbaev [16]. Recently, it was extended to type EnE_n by Sinchuk [11], to type DnD_n by Lavrenov-Sinchuk [9], and to type CnC_n by Lavrenov [8]. The present article aims to revisit these results in the setting of commutative N\mathbb{N}-graded rings.

The graded version of the K1K_1-analog of the local-global principle has been studied for classical groups in [3, 5]. Meanwhile, for Chevalley groups and Steinberg groups, the graded versions of both the K1K_1-analog and K2K_2-analog are expected to hold. We therefore establish the graded versions of the local-global principle for Steinberg groups and Chevalley groups. These results constitute the main theorems of the present article and are stated below.

Theorem 1.1 (Theorem 3.4). Let R=⨁i≥0RiR = \bigoplus_{i \ge0} R_i be a graded ring and Φn\Phi_n be a either root system of type simply-laced or CnC_n of rank ≥3\ge3. Suppose θ∈St⁡(Φn,R,R+)\theta\in\operatorname{St}(\Phi_n,R,R^+). Then

2020 Mathematics Subject Classification. 20G35, 19C20, 13A02.

Key words and phrases. Steinberg group, Chevalley group, Local-global principle, Graded ring.

θ=1 if and only if λm∗(θ)=1∈St⁡(Φn,Rm) for all maximal ideals m of R0.(1)\theta= 1 \text{ if and only if } \lambda_{\mathfrak{m}}^{*}(\theta) = 1 \in\operatorname{St}(\Phi_n,R_{\mathfrak{m}}) \text{ for all maximal ideals } \mathfrak{m} \text{ of } R_0. \tag*{(1)}
θ∈Im⁡(lR) if and only if λm∗(θ)∈Im⁡(lRm) for all maximal ideals m of R0.(2)\theta\in\operatorname{Im}(l_R) \text{ if and only if } \lambda_{\mathfrak{m}}^{*}(\theta) \in\operatorname{Im}(l_{R_{\mathfrak{m}}}) \text{ for all maximal ideals } \mathfrak{m} \text{ of } R_0. \tag*{(2)}

Theorem 1.2 (Theorem 4.4). Let R=⨁i≥0RiR = \bigoplus_{i \ge0} R_i be a graded ring and Φ\Phi be a root system of rank ≥2\ge2. Assume that RR has no residue fields of 2 elements if Φ=C2\Phi= C_2 or G2G_2. Suppose θ∈G(Φ,R,R+)\theta\in G(\Phi,R,R^{+}). Then θ∈E(Φ,R)\theta\in E(\Phi,R) if and only if λm∗(θ)∈E(Φ,Rm)\lambda_{\mathfrak{m}}^{*}(\theta) \in E(\Phi,R_{\mathfrak{m}}) for all maximal ideals m\mathfrak{m} of R0R_0.

The centrality of K2K_2 (see [9]) follows from the K2K_2-analog of Quillen’s local-global principle, just as the normality theorem for the elementary subgroup is derived from the K1K_1-analog of Suslin’s local-global principle. In other words, the local-global principle plays a foundational role in establishing both the normality of E(Φ,R)E(\Phi,R) and the centrality of St⁡(Φ,R)\operatorname{St}(\Phi,R). To extend these results from the polynomial ring to the graded ring, one must employ a combination of three key techniques: the methodology of Quillen-Suslin’s argument, the dilation principle, and Swan–Weibel’s homotopy technique, which allows one to connect polynomial rings to graded rings via a homotopy map.

Preliminaries and Notations

Throughout the article, all rings are assumed to be commutative with unity, and all root systems are reduced and irreducible of rank ≥2\ge2 unless otherwise stated.

Definition 2.1. Let AA be a ring and Φ\Phi be a root system. The Steinberg group St⁡(Φ,A)\operatorname{St}(\Phi,A) is the group generated by symbols xα(u)x_\alpha(u), where α∈Φ\alpha\in\Phi and u∈Au \in A, subject to the following relations:

(St1) xα(u)xα(v)=xα(u+v),u,v∈R, α∈Φx_\alpha(u)x_\alpha(v) = x_\alpha(u+v), \qquad u,v \in R,\ \alpha\in\Phi;

(St2) [xα(u),xβ(v)]=∏iα+jβ∈Φxiα+jβ(Ci,j,α,βuivj),u,v∈R, α,β∈Φ, α≠−β[x_\alpha(u),x_\beta(v)] = \prod_{i\alpha+j\beta\in\Phi} x_{i\alpha+j\beta}(C_{i,j,\alpha,\beta}u^iv^j), \qquad u,v \in R,\ \alpha,\beta\in\Phi,\ \alpha\ne-\beta,

where [p,q]=pqp−1q−1[p,q] = pqp^{-1}q^{-1}, the product is taken over all positive integers i,ji,j such that iα+jβ∈Φi\alpha+j\beta\in\Phi, and Ci,j,α,β∈ZC_{i,j,\alpha,\beta} \in\mathbb{Z} are the Chevalley structure constants (depending only on Φ\Phi). More information on Chevalley groups, the structure constants, and various types of root systems Φ\Phi can be found in [13, 17].

Definition 2.2. A root system Φ\Phi is called simply-laced if the Dynkin diagram of Φ\Phi does not contain double bonds (i.e., Φ\Phi has type AnA_n, DnD_n and E6,7,8E_{6,7,8}).

For an ideal II of AA, the subgroup St⁡(Φ,I)\operatorname{St}(\Phi,I) is spanned by symbols xα(s)x_\alpha(s), where α∈Φ\alpha\in\Phi and s∈Is \in I. An ideal II is called a splitting ideal if the canonical projection ring homomorphism π:A→A/I\pi: A \to A/I splits, i.e., there exists σ:A/I→A\sigma: A/I \to A such that πσ=1\pi\sigma= 1. It is noted that TA[T]TA[T] is a splitting ideal of A[T]A[T].

Definition 2.3. For a splitting ideal II, the relative Steinberg group St⁡(Φ,A,I)\operatorname{St}(\Phi,A,I) is defined as the kernel Ker⁡(π∗:St⁡(Φ,A)→St⁡(Φ,A/I))\operatorname{Ker}(\pi^{*} : \operatorname{St}(\Phi,A) \to\operatorname{St}(\Phi,A/I)). Obviously, St⁡(Φ,A,I)\operatorname{St}(\Phi,A,I) coincides with the normal closure of St⁡(Φ,I)\operatorname{St}(\Phi,I) in St⁡(Φ,A)\operatorname{St}(\Phi,A).

In the case of an arbitrary ideal II, the relative Steinberg group St⁡(Φ,A,I)\operatorname{St}(\Phi,A,I) can be defined as a central extension of Ker⁡(π∗)\operatorname{Ker}(\pi^{*}). For the purposes of the present text, we concentrate on the relative Steinberg groups corresponding to splitting ideals. For more information regarding the general case, we refer the reader to [11].

Definition 2.4. (Universal Property) Let F(X)\mathcal{F}(X) denote the free group on the generating set XX. If GG is any group with a generating set XX and certain defining relations ui=viu_i=v_i, there is an isomorphism F(X)/D≅G\mathcal{F}(X)/\mathcal{D}\cong G taking xˉ↦x\bar{x}\mapsto x for each x∈Xx\in X, where D\mathcal{D} is a normal subgroup of F(X)\mathcal{F}(X) generated by the conjugates of the relators uivi−1u_iv_i^{-1}. Then F(X)\mathcal{F}(X) has the following universal property: If HH is any group and each function f:X→Hf:X\to H extends uniquely to a group homomorphism φ:F(X)→H\varphi:\mathcal{F}(X)\to H that makes the diagram commute.

X→fH↓↑f∗F(X)→φH↓F(X)/D\begin{CD} X @>{f}>> H \\ @VVV @AA{f^*}A \\ \mathcal{F}(X) @>{\varphi}>> H \\ @VVV \\ \mathcal{F}(X)/\mathcal{D} \end{CD}

Now, to define a group homomorphism f∗:G→Hf^*:G\to H, it suffices to choose a function f:X→Hf:X\to H for which, when xx is replaced by f(x)f(x) and x−1x^{-1} by f(x)−1f(x)^{-1}, the defining relations ui=viu_i=v_i for GG become true equations in HH. Indeed, the homomorphism φ\varphi extending ff takes D\mathcal{D} to 1H1_H; thus, it induces a homomorphism f∗f^* taking xˉ\bar{x} to f(x)f(x) for x∈Xx\in X. Thus, the composite map f∗:F(X)/D→Hf^*:\mathcal{F}(X)/\mathcal{D}\to H is a homomorphism taking x∈Xx\in X to f(x)∈Hf(x)\in H. For more details on universal property and classical K-theory, we refer the reader to [6].

Lemma 2.5. If f:A→Bf:A\to B is a ring homomorphism, then ff induces a group homomorphism f∗:St⁡(Φ,A)→St⁡(Φ,B)f^*:\operatorname{St}(\Phi,A)\to\operatorname{St}(\Phi,B).

Proof. Since any element of St⁡(Φ,A)\operatorname{St}(\Phi,A) can be written as a finite product of elements of the form xα(u)x_\alpha(u), α∈Φ\alpha\in\Phi, u∈Au\in A, it is enough to define a map on generators. We define a map f∗(xα(u))=xα(f(u))f^*(x_\alpha(u))=x_\alpha(f(u)) where α∈Φ\alpha\in\Phi. Due to the universal property 2.4, we need to verify that this map preserves the Steinberg relations, i.e., replacing uu by f(u)f(u) and vv by f(v)f(v) in the defining relations (St1) and (St2) for St⁡(Φ,A)\operatorname{St}(\Phi,A) yields equations in St⁡(Φ,B)\operatorname{St}(\Phi,B). However, this is indeed true because ff is a ring homomorphism. ■

One of the key components for handling the graded situation is the well-known “Swan–Weibel’s homotopy trick”. Let R=R0⊕R1⊕R2⊕⋯R=R_0\oplus R_1\oplus R_2\oplus\cdots be a graded ring and R+=R1⊕R2⊕⋯R^+=R_1\oplus R_2\oplus\cdots denote the irrelevant ideal of RR. An element r∈Rr\in R will be denoted by r=r0+r1+r2+⋯r=r_0+r_1+r_2+\cdots where ri∈Rir_i\in R_i for each ii, and all but finitely many rir_i are zero.

Definition 2.6. For a graded ring R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i, the Swan–Weibel homotopy map h:R→R[T]h:R\to R[T] is defined as

h(r0+r1+⋯+rm)=r0+r1T+⋯+rmTm,(ri∈Ri).h(r_0+r_1+\cdots+r_m)=r_0+r_1T+\cdots+r_mT^m,\quad(r_i\in R_i).

For any R0R_0-algebra AA and any a∈Aa\in A, consider an evaluation map ev⁡a:R[T]→A\operatorname{ev}_a:R[T]\to A, which is the R0R_0-algebra homomorphism sending T↦aT\mapsto a.

We denote the image rr in AA by rh(a):=(ev⁡a∘h)(r)r^h(a):=(\operatorname{ev}_a\circ h)(r). If A=RA=R, then it is easy to observe that ev⁡0∘h:R→R\operatorname{ev}_0\circ h:R\to R is precisely the projection onto the degree-zero part and ev⁡1∘h:R→R\operatorname{ev}_1\circ h:R\to R is precisely the identity map. For a fixed element c∈R0c\in R_0, similarly denote rh(c):=h(r)(c)=(ev⁡c∘h)(r)r^h(c):=h(r)(c)=(\operatorname{ev}_c\circ h)(r), and note that rh(0)=r0r^h(0)=r_0 and rh(1)=rr^h(1)=r. In order to emphasize the choice of AA, we denote by ev⁡[R[T]→AT↦a]\operatorname{ev}\left[\frac{R[T]\to A}{T\mapsto a}\right] for R0R_0-algebra homomorphism ev⁡a\operatorname{ev}_a.

Remark 2.7. In view of Lemma 2.5, the map h:R→R[T]h : R \to R[T] (resp. ev⁡a:R[T]→A\operatorname{ev}_{a} : R[T] \to A) induces a group homomorphism h∗:St⁡(Φ,R)→St⁡(Φ,R[T])h^{*} : \operatorname{St}(\Phi, R) \to\operatorname{St}(\Phi, R[T]) (resp. ev⁡a∗:St⁡(Φ,R[T])→St⁡(Φ,A)\operatorname{ev}_{a}^{*} : \operatorname{St}(\Phi, R[T]) \to\operatorname{St}(\Phi, A)) that sends the generators xα(r)x_{\alpha}(r) to xα(h(r))x_{\alpha}(h(r)) (resp. xα(p(T))x_{\alpha}(p(T)) to xα(p(a))x_{\alpha}(p(a))) where α∈Φ\alpha\in\Phi. For θ∈St⁡(Φ,R)\theta\in\operatorname{St}(\Phi, R) and c∈R0c \in R_{0}, analogously, θh∗(c):=(ev⁡c∗∘h∗)(θ)\theta^{h^{*}}(c) := (\operatorname{ev}_{c}^{*} \circ h^{*})(\theta) is defined. For more details on Swan–Weibel’s homotopy trick, we refer the reader to [7 Chapter V].

Notations: For a ring AA, we denote by λa:A→Aa\lambda_{a} : A \to A_{a} and λm:A→Am\lambda_{m} : A \to A_{m} the homomorphisms of localization at an element a∈Aa \in A and at a maximal ideal mm of AA respectively. If f:A→Bf : A \to B is a ring homomorphism, its corresponding Steinberg group homomorphism is denoted by f∗:St⁡(Φ,A)→St⁡(Φ,B)f^{*} : \operatorname{St}(\Phi, A) \to\operatorname{St}(\Phi, B). If Φn\Phi_{n} is an irreducible root system of rank nn, we denote by ιA:St⁡(Φn−1,A)→St⁡(Φn,A)\iota_{A} : \operatorname{St}(\Phi_{n-1}, A) \to\operatorname{St}(\Phi_{n}, A) the homomorphism of Steinberg groups. For visual understanding, it is noted that we have the following commutative diagram

St⁡(Φn−1,A)→ιASt⁡(Φn,A)f∗↓↓f∗St⁡(Φn−1,B)→ιBSt⁡(Φn,B).\begin{CD} \operatorname{St}(\Phi_{n-1}, A) @>{\iota_{A}}>> \operatorname{St}(\Phi_{n}, A) \\ @V{f^{*}}VV @VV{f^{*}}V \\ \operatorname{St}(\Phi_{n-1}, B) @>{\iota_{B}}>> \operatorname{St}(\Phi_{n}, B). \end{CD}

Local-Global Principle for Stienberg Groups

In this section, we assume that Φn\Phi_{n} is a root system either of type simply-laced (i.e., An≥3,Dn≥3A_{n\geq3}, D_{n\geq3} or En=6,7,8E_{n=6,7,8}) or Cn≥3C_{n\geq3}. We begin with the following observation.

Lemma 3.1. Let R=⨁i≥0RiR = \bigoplus_{i\geq0} R_{i} be a graded ring and ‘overbar’ denote the reduction modulo the ideal R+R^{+}. For θ∈St⁡(Φ,R)\theta\in\operatorname{St}(\Phi, R) one has θh∗(0)=θˉ\theta^{h^{*}}(0) = \bar{\theta} and θh∗(1)=θ\theta^{h^{*}}(1) = \theta.

Proof. An element θ∈St⁡(Φ,R)\theta\in\operatorname{St}(\Phi, R) has the form ∏i=1kxαi(ri)\prod_{i=1}^{k} x_{\alpha_{i}}(r_{i}), where αi∈Φ\alpha_{i} \in\Phi and ri=⨁j=0mirijr_{i} = \bigoplus_{j=0}^{m_{i}} r_{ij} with rij∈Rjr_{ij} \in R_{j}. Then

h∗(θ)=∏i=1kh∗(xαi(ri))h^{*}(\theta) = \prod_{i=1}^{k} h^{*}\bigl(x_{\alpha_{i}}(r_{i})\bigr)
=∏i=1kxαi(h(⨁j=0mirij))(by xref‘Remark 2.7‘)= \prod_{i=1}^{k} x_{\alpha_{i}}\left(h\left(\bigoplus_{j=0}^{m_{i}} r_{ij}\right)\right) \quad\text{(by {xref}`Remark 2.7`)}
=∏i=1kxαi(⨁j=0mirijTj).= \prod_{i=1}^{k} x_{\alpha_{i}}\left(\bigoplus_{j=0}^{m_{i}} r_{ij}T^{j}\right).

Now, it is easy to observe that θh∗(0)=(ev⁡0∗∘h∗)(θ)=∏i=1kxαi(ri0)=θˉ\theta^{h^{*}}(0) = (\operatorname{ev}_{0}^{*} \circ h^{*})(\theta) = \prod_{i=1}^{k} x_{\alpha_{i}}(r_{i0}) = \bar{\theta} and θh∗(1)=(ev⁡1∗∘h∗)(θ)=θ\theta^{h^{*}}(1) = (\operatorname{ev}_{1}^{*} \circ h^{*})(\theta) = \theta. ■

The following lemma is known as the dilation principle for Steinberg groups.

Lemma 3.2. Let AA be an arbitrary ring and a∈Aa \in A. Suppose θ(Z)∈St⁡(Φn,A[Z],ZA[Z])\theta(Z) \in\operatorname{St}(\Phi_{n}, A[Z], ZA[Z]).

  1. If λa∗(θ(Z))=1\lambda_{a}^{*}(\theta(Z)) = 1, then for sufficiently large kk one has θ(akZ)=1\theta(a^{k}Z) = 1.

  2. If λa∗(θ(Z))∈Im⁡(ιAa[Z])\lambda_{a}^{*}(\theta(Z)) \in\operatorname{Im}(\iota_{A_{a}[Z]}), then for sufficiently large kk one has θ(akZ)∈Im⁡(ιA[Z])\theta(a^{k}Z) \in\operatorname{Im}(\iota_{A[Z]}).

Proof. (1) [9 Proposition 2.2] for Φn=An,Dn,E6,7,8\Phi_{n} = A_{n},D_{n},E_{6,7,8} and [8 Lemma 3.1] for Φn=Cn\Phi_{n} = C_{n}.

(2) [12 Lemma 5.1] for Φn=An,Dn,E6,7,8\Phi_{n} = A_{n},D_{n},E_{6,7,8} and [8 Lemma 3.1] for Φn=Cn\Phi_{n} = C_{n}. ■

The following lemma is a crucial step towards the proof of Theorem 1.1.

Lemma 3.3. Let R=⨁i≥0RiR = \bigoplus_{i\geq0} R_i be a graded ring and a,b∈R0a,b \in R_0 be two co-maximal elements. Suppose θ∈St⁡(Φn,R,R+)\theta\in\operatorname{St}(\Phi_n,R,R^+).

(1) If λa∗(θ)\lambda_a^*(\theta) and λb∗(θ)\lambda_b^*(\theta) are equal to 11, then θ=1\theta=1.

(2) If λa∗(θ)∈Im⁡(lRa)\lambda_a^*(\theta) \in\operatorname{Im}(l_{R_a}) and λb∗(θ)∈Im⁡(lRb)\lambda_b^*(\theta) \in\operatorname{Im}(l_{R_b}), then θ∈Im⁡(lR)\theta\in\operatorname{Im}(l_R).

Proof. Consider an element ξ:=ξ(X,Y)=θh∗(X)⋅(θh∗(X+Y))−1∈St⁡(Φn,A[Y])\xi:= \xi(X,Y)=\theta^{h^*}(X)\cdot(\theta^{h^*}(X+Y))^{-1}\in\operatorname{St}(\Phi_n,A[Y]), where A=R[X]A=R[X]. This means that

ξ(X,Y)=(ev⁡[R[T]→A[Y]T↦X]∗∘h∗)(θ)⋅(ev⁡[R[T]→A[Y]T↦X+Y]∗∘h∗)(θ−1).\xi(X,Y)=\left(\operatorname{ev}\left[\frac{R[T]\to A[Y]}{T\mapsto X}\right]^*\circ h^*\right)(\theta)\cdot\left(\operatorname{ev}\left[\frac{R[T]\to A[Y]}{T\mapsto X+Y}\right]^*\circ h^*\right)(\theta^{-1}).

Now, it is easy to observe that ξ(X,0)=1\xi(X,0)=1. Hence, ξ(X,Y)\xi(X,Y) belongs to the kernel of St⁡(Φn,A[Y])→St⁡(Φn,A)\operatorname{St}(\Phi_n,A[Y])\to\operatorname{St}(\Phi_n,A). Thus, we can assume that ξ(X,Y)∈St⁡(Φn,A[Y],YA[Y])\xi(X,Y)\in\operatorname{St}(\Phi_n,A[Y],YA[Y]). The image of ξ(X,Y)\xi(X,Y) in the ring St⁡(Φn,Aa[Y])\operatorname{St}(\Phi_n,A_a[Y]) is

λa∗(ξ)=(λa∗(ev⁡[R[T]→A[Y]T↦X]∗∘h∗))(θ)⋅(λa∗(ev⁡[R[T]→A[Y]T↦X+Y]∗∘h∗))(θ−1)=(ev⁡[Ra[T]→Aa[Y]T↦X]∗∘ha∗)(λa∗(θ))⋅(ev⁡[Ra[T]→Aa[Y]T↦X+Y]∗∘ha∗)(λa∗(θ−1))=(λa∗(θ))ha∗(X)⋅((λa∗(θ))ha∗(X+Y))−1.\begin{aligned} \lambda_a^*(\xi)=\left(\lambda_a^*\left(\operatorname{ev}\left[\frac{R[T]\to A[Y]}{T\mapsto X}\right]^*\circ h^*\right)\right)(\theta)\cdot\left(\lambda_a^*\left(\operatorname{ev}\left[\frac{R[T]\to A[Y]}{T\mapsto X+Y}\right]^*\circ h^*\right)\right)(\theta^{-1}) \\ &=\left(\operatorname{ev}\left[\frac{R_a[T]\to A_a[Y]}{T\mapsto X}\right]^*\circ h_a^*\right)(\lambda_a^*(\theta))\cdot\left(\operatorname{ev}\left[\frac{R_a[T]\to A_a[Y]}{T\mapsto X+Y}\right]^*\circ h_a^*\right)(\lambda_a^*(\theta^{-1})) \\ &=(\lambda_a^*(\theta))^{h_a^*(X)}\cdot\left((\lambda_a^*(\theta))^{h_a^*(X+Y)}\right)^{-1}. \end{aligned}

(1) Since λa∗(ξ)=1∈St⁡(Φn,Aa[Y])\lambda_a^*(\xi)=1\in\operatorname{St}(\Phi_n,A_a[Y]) and analogously λb∗(ξ)=1∈St⁡(Φn,Ab[Y])\lambda_b^*(\xi)=1\in\operatorname{St}(\Phi_n,A_b[Y]), by Lemma 3.2(1), for some natural number kk, we have ξ(X,akY)=1\xi(X,a^kY)=1 and ξ(X,bkY)=1\xi(X,b^kY)=1. We choose u,v∈R0u,v\in R_0 such that uak+vbk=1ua^k+vb^k=1. This implies ξ(X,−uakY)=1\xi(X,-ua^kY)=1 and ξ(X,−vbkY)=1\xi(X,-vb^kY)=1. From Lemma 3.1, we have θh∗(0)=1\theta^{h^*}(0)=1. Then

θ=θh∗(1)(θh∗(0))−1=θh∗(1)(θh∗(uak))−1θh∗(uak)(θh∗(0))−1=ξ(1,−vbk)ξ(uak,−uak)=1.\begin{aligned} \theta=\theta^{h^*}(1)(\theta^{h^*}(0))^{-1} \\ &=\theta^{h^*}(1)(\theta^{h^*}(ua^k))^{-1}\theta^{h^*}(ua^k)(\theta^{h^*}(0))^{-1} \\ &=\xi(1,-vb^k)\xi(ua^k,-ua^k) \\ &=1. \end{aligned}

(2) The hypothesis implies that λa∗(ξ)∈Im⁡(lAa[Y])\lambda_a^*(\xi)\in\operatorname{Im}(l_{A_a[Y]}) and λb∗(ξ)∈Im⁡(lAb[Y])\lambda_b^*(\xi)\in\operatorname{Im}(l_{A_b[Y]}). By Lemma 3.2(2), both ξ(X,−uakY)\xi(X,-ua^kY) and ξ(X,−vbkY)\xi(X,-vb^kY) lie in the image of lA[Y]l_{A[Y]} as in the case (1). Then the identities θ=ξ(1,−vbk)ξ(uak,−uak)\theta=\xi(1,-vb^k)\xi(ua^k,-ua^k) implies that θ∈Im⁡(lR)\theta\in\operatorname{Im}(l_R). ■

We now prove Theorem 1.1, which we restate here for the reader’s convenience.

Theorem 3.4. Let R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i be a graded ring. Suppose θ∈St⁡(Φn,R,R+)\theta\in\operatorname{St}(\Phi_n,R,R^+). Then

(1) θ=1\theta=1 if and only if λm∗(θ)=1∈St⁡(Φn,Rm)\lambda_{\mathfrak m}^*(\theta)=1\in\operatorname{St}(\Phi_n,R_{\mathfrak m}) for all maximal ideals m\mathfrak m of R0R_0.

(2) θ∈Im⁡(lR)\theta\in\operatorname{Im}(l_R) if and only if λm∗(θ)∈Im⁡(lRm)\lambda_{\mathfrak m}^*(\theta)\in\operatorname{Im}(l_{R_{\mathfrak m}}) for all maximal ideals m\mathfrak m of R0R_0.

Proof. We first prove that the following set is an ideal of R0R_0 (called the Quillen ideal).

Q(R0)={{a∈R0∣λa∗(θ)=1}for (1),{a∈R0∣λa∗(θ)∈Im⁡(lRa)}for (2).\mathcal{Q}(R_0)= \begin{cases} \{a\in R_0\mid\lambda_a^*(\theta)=1\} & \text{for (1)},\\ \{a\in R_0\mid\lambda_a^*(\theta)\in\operatorname{Im}(l_{R_a})\} & \text{for (2)}. \end{cases}

Clearly 0∈Q(R0)0\in\mathcal{Q}(R_0). If a,b∈Q(R0)a,b\in\mathcal{Q}(R_0) and c∈aR0+bR0c\in aR_0+bR_0, then clearly ac,bc∈Q(R0)ac,bc\in\mathcal{Q}(R_0). Note that aa and bb are comaximal elements of (R0)c(R_0)_c. Since λc∗(θ)∈St⁡(Φn,Rc,Rc+)\lambda_c^*(\theta)\in\operatorname{St}(\Phi_n,R_c,R_c^+), apply Lemma 3.3 to the ring RcR_c, we get λc∗(θ)=1\lambda_c^*(\theta)=1 (resp. λc∗(θ)∈Im⁡(lRc)\lambda_c^*(\theta)\in\operatorname{Im}(l_{R_c})), and hence c∈Q(R0)c\in\mathcal{Q}(R_0). Therefore, Q(R0)\mathcal{Q}(R_0) is an ideal of R0R_0. It suffices to show if part of the statement. To get this we need to prove that Q(R0)=R0\mathcal{Q}(R_0)=R_0, i.e., 1∈Q(R0)1\in\mathcal{Q}(R_0). If otherwise, there exists a maximal ideal m\mathfrak{m} of R0R_0 such that Q(R0)⊆m\mathcal{Q}(R_0)\subseteq\mathfrak{m}. Since λm∗(θ)=1\lambda_{\mathfrak{m}}^*(\theta)=1 (resp. λm∗(θ)∈Im⁡(lRm)\lambda_{\mathfrak{m}}^*(\theta)\in\operatorname{Im}(l_{R_{\mathfrak{m}}})), there exists a∈R0∖ma\in R_0\setminus\mathfrak{m} such that λa∗(θ)=1\lambda_a^*(\theta)=1 (resp. λm∗(θ)∈Im⁡(lRa)\lambda_{\mathfrak{m}}^*(\theta)\in\operatorname{Im}(l_{R_a})). Then a∈Q(R0)⊆ma\in\mathcal{Q}(R_0)\subseteq\mathfrak{m} which is a contradiction. ■

Local-Global Principle for Chevalley Groups

In this section, we deduce an analog of the local-global principle for the Chevalley group over graded rings, and we assume that RR has no residue fields of 2 elements when Φ=C2\Phi=C_2 or G2G_2. We first recall the definition of a relative elementary subgroup.

Definition 4.1. For an ideal II of a ring AA, the congruence subgroup G(Φ,A,I)G(\Phi,A,I) is the kernel of the canonical map G(Φ,A)→G(Φ,A/I)G(\Phi,A)\to G(\Phi,A/I). The subgroup E(Φ,I)E(\Phi,I) of G(Φ,A,I)G(\Phi,A,I) is spanned by elementary root unipotents eα(s)e_\alpha(s) where s∈Is\in I and the relative elementary subgroup E(Φ,A,I)E(\Phi,A,I) is a normal closure of E(Φ,I)E(\Phi,I) in E(Φ,A)E(\Phi,A).

Let R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i be a graded ring. The Swan-Weibel’s homotopy map h:R→R[T]h:R\to R[T] induced a group homomorphism h∗:G(Φ,R)→G(Φ,R[T])h^*:G(\Phi,R)\to G(\Phi,R[T]). As in Remark 2.7, if θ∈G(Φ,R)\theta\in G(\Phi,R), for c∈R0c\in R_0, we denote by θh∗(c):=(ev⁡c∗∘h∗)(θ)\theta^{h^*}(c):=(\operatorname{ev}_c^*\circ h^*)(\theta).

The following lemma is similar to Lemma 2.5. Hence, we skip the proof.

Lemma 4.2. If f:A→Bf:A\to B is a ring homomorphism, then ff induces a group homomorphism f∗:E(Φ,A)→E(Φ,B)f^*:E(\Phi,A)\to E(\Phi,B) sending the generators eα(a)e_\alpha(a) to eα(f(a))e_\alpha(f(a)).

The following lemma generalizes the result of Suslin’s [14 Lemma 3.4], where he proved it over a polynomial ring for an elementary linear group.

Lemma 4.3. Let R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i be a graded ring. Suppose θ∈G(Φ,R)\theta\in G(\Phi,R), and s∈R0s\in R_0. If λs∗(θ)∈E(Φ,Rs)\lambda_s^*(\theta)\in E(\Phi,R_s), then there exists an integer kk such that, for two indeterminates X,YX,Y, we have

θh∗(X+skY)⋅θh∗(X)−1∈E(Φ,R[X,Y],(Y)).\theta^{h^*}(X+s^kY)\cdot\theta^{h^*}(X)^{-1}\in E(\Phi,R[X,Y],(Y)).

In particular, for any c,d∈R0c,d\in R_0 with c≡d(modskR0)c\equiv d\pmod{s^kR_0}, we have θh∗(c)⋅θh∗(d)−1∈E(Φ,R)\theta^{h^*}(c)\cdot\theta^{h^*}(d)^{-1}\in E(\Phi,R).

Proof. Denote Γ:=Γ(Z)=θh∗(Z)=(ev⁡R[T]→R[Z]T↦Z∗∘h∗)(θ)∈G(Φ,R[Z])\Gamma:=\Gamma(Z)=\theta^{h^*}(Z)=\left(\operatorname{ev}_{\frac{R[T]\to R[Z]}{T\mapsto Z}}^*\circ h^*\right)(\theta)\in G(\Phi,R[Z]). Since λs∗(θ)∈E(Φ,Rs)\lambda_s^*(\theta)\in E(\Phi,R_s), by Lemma 4.2, we have λs∗(Γ)=(ev⁡Rs[T]→Rs[Z]T↦Z∗∘h∗)(λs∗(θ))∈E(Φ,Rs[T])\lambda_s^*(\Gamma)=\left(\operatorname{ev}_{\frac{R_s[T]\to R_s[Z]}{T\mapsto Z}}^*\circ h^*\right)(\lambda_s^*(\theta))\in E(\Phi,R_s[T]). Consider the following element in G(Φ,R[X,Y,Z],(Y))G(\Phi,R[X,Y,Z],(Y)),

Γ′(X,Y,Z)=Γ((X+Y)Z)⋅Γ(XZ)−1=θh∗((X+Y)Z)⋅θh∗(XZ)−1=ev⁡R[T]→R[X,Y,Z]T↦(X+Y)Z∗(h∗(θ))⋅ev⁡R[T]→R[X,Y,Z]T↦XZ∗(h∗(θ−1)).\begin{aligned} \Gamma'(X,Y,Z)&=\Gamma((X+Y)Z)\cdot\Gamma(XZ)^{-1}\\ &=\theta^{h^*}((X+Y)Z)\cdot\theta^{h^*}(XZ)^{-1}\\ &=\operatorname{ev}_{\frac{R[T]\to R[X,Y,Z]}{T\mapsto(X+Y)Z}}^*(h^*(\theta))\cdot\operatorname{ev}_{\frac{R[T]\to R[X,Y,Z]}{T\mapsto XZ}}^*(h^*(\theta^{-1})). \end{aligned}

Since λs∗(Γ)∈E(Φ,Rs[Z])\lambda_s^*(\Gamma)\in E(\Phi,R_s[Z]), we have λs∗(Γ′)∈E(Φ,Rs[X,Y,Z])\lambda_s^*(\Gamma')\in E(\Phi,R_s[X,Y,Z]). Applying [2 Lemma 1.4] to the ring Rs[X,Y,Z]R_s[X,Y,Z] and the splitting ideal YRs[X,Y,Z]Y R_s[X,Y,Z], we have

G(Φ,Rs[X,Y,Z],(Y))∩E(Φ,Rs[X,Y,Z])=E(Φ,Rs[X,Y,Z],(Y)).G(\Phi,R_s[X,Y,Z],(Y))\cap E(\Phi,R_s[X,Y,Z])=E(\Phi,R_s[X,Y,Z],(Y)).

Therefore, λs∗(Γ′)∈E(Φ,Rs[X,Y,Z],(Y))\lambda_s^*(\Gamma')\in E(\Phi,R_s[X,Y,Z],(Y)). Then, by [2 Theorem 6.3], there exists an integer kk such that Γ′(X,skY,Z)∈E(Φ,R[X,Y,Z],(Y))\Gamma'(X,s^kY,Z)\in E(\Phi,R[X,Y,Z],(Y)). Hence, substituting Z=1Z=1, we have that

Γ′(X,skY):=θh∗(X+skY)⋅θh∗(X)−1∈E(Φ,R[X,Y],(Y)).\Gamma'(X,s^kY):=\theta^{h^*}(X+s^kY)\cdot\theta^{h^*}(X)^{-1}\in E(\Phi,R[X,Y],(Y)).

::: For any c,d∈R0c,d \in R_0 with c=d+skbc=d+s^k b for some b∈R0b \in R_0 and specializing X=dX=d and Y=bY=b, we get

Γ∗(d,skb):=θh∗(c)⋅θh∗(d)−1∈E(Φ,R).\Gamma^*(d,s^k b):=\theta^{h^*}(c)\cdot\theta^{h^*}(d)^{-1}\in E(\Phi,R).

This gives the desired conclusion.

Theorem 4.4. Let R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i be a graded ring. Suppose θ∈G(Φ,R)\theta\in G(\Phi,R) with θh∗(0)=1\theta^{h^*}(0)=1. Then θ∈E(Φ,R)\theta\in E(\Phi,R) if and only if λm∗(θ)∈E(Φ,Rm)\lambda^*_m(\theta)\in E(\Phi,R_m) for all maximal ideals mm of R0R_0.

Proof. It suffices to show if part of the statement. We shall first prove that if λa∗(θ)∈E(Φ,Ra)\lambda^*_a(\theta)\in E(\Phi,R_a) and λb∗(θ)∈E(Φ,Rb)\lambda^*_b(\theta)\in E(\Phi,R_b) with a,ba,b are two co-maximal elements of R0R_0, then θ∈E(Φ,R)\theta\in E(\Phi,R).

By Lemma 4.3, we can choose kk such that θh∗(c)θh∗(d)−1∈E(Φ,R)\theta^{h^*}(c)\theta^{h^*}(d)^{-1}\in E(\Phi,R) whenever c≡d(modakR0)c\equiv d\pmod{a^kR_0} or c≡d(modbkR0)c\equiv d\pmod{b^kR_0}. Note that aku+bkv=1a^ku+b^kv=1 for some u,v∈R0u,v\in R_0. Since θh∗(0)=1\theta^{h^*}(0)=1, we write

θ=θh∗(1)(θh∗(uak))−1θh∗(uak)(θh∗(0))−1.\theta=\theta^{h^*}(1)(\theta^{h^*}(ua^k))^{-1}\theta^{h^*}(ua^k)(\theta^{h^*}(0))^{-1}.

Then θ∈E(Φ,R)\theta\in E(\Phi,R) since aku≡1(modbkR0)a^ku\equiv1\pmod{b^kR_0} and aku≡0(modakR0)a^ku\equiv0\pmod{a^kR_0}.

As in Theorem 3.4, it is easy to verify that the set Q(θ)={a∈R0∣λa∗(θ)∈E(Φ,Ra)}Q(\theta)=\{a\in R_0\mid\lambda^*_a(\theta)\in E(\Phi,R_a)\} is an ideal of R0R_0. To show θ∈E(Φ,R)\theta\in E(\Phi,R) we need to prove that 1∈Q(R0)1\in Q(R_0). If otherwise, there exists a maximal ideal mm such that Q(R0)⊆mQ(R_0)\subseteq m. Since λm∗(θ)∈E(Φ,Rm)\lambda^*_m(\theta)\in E(\Phi,R_m), there exists a∈R0∖ma\in R_0\setminus m such that λa∗(θ)∈E(Φ,Ra)\lambda^*_a(\theta)\in E(\Phi,R_a). Then a∈Q(R0)⊆ma\in Q(R_0)\subseteq m which is a contradiction.

Corollary 4.5. Let R=⨁i≥0RiR=\bigoplus_{i\geq0}R_i be a graded ring and θ∈G(Φ,R)\theta\in G(\Phi,R). If λm∗(θ)∈G(Φ,(R0)m)⋅E(Φ,Rm)\lambda^*_m(\theta)\in G(\Phi,(R_0)_m)\cdot E(\Phi,R_m) for all maximal ideals mm of R0R_0, then θ∈G(Φ,R0)⋅E(Φ,R)\theta\in G(\Phi,R_0)\cdot E(\Phi,R).

Proof. Let ‘overbar’ denote the reduction modulo the ideal R+R^+. After left-multiplying θ\theta by θ‾−1\overline{\theta}^{-1}, we can assume that θh∗(0)=1\theta^{h^*}(0)=1. For all maximal ideals mm of R0R_0, we have the decomposition λm∗(θ)=η⋅γ\lambda^*_m(\theta)=\eta\cdot\gamma where η∈G(Φ,(R0)m)\eta\in G(\Phi,(R_0)_m) and γ∈E(Φ,Rm)\gamma\in E(\Phi,R_m). Now, reduction modulo R+R^+, we get I=η‾⋅γ‾I=\overline{\eta}\cdot\overline{\gamma}, and hence η=γ‾−1\eta=\overline{\gamma}^{-1}. Thus

λm∗(θ)=γ‾−1⋅γ∈E(Φ,Rm).\lambda^*_m(\theta)=\overline{\gamma}^{-1}\cdot\gamma\in E(\Phi,R_m).

Then, by Theorem 4.4, we get θ∈E(Φ,R)\theta\in E(\Phi,R). Therefore θ∈G(Φ,R0)⋅E(Φ,R)\theta\in G(\Phi,R_0)\cdot E(\Phi,R) since the original θ\theta has already been modified by a left factor from G(Φ,R0)G(\Phi,R_0).

Acknowledgments: The author thanks Prof. Manoj Kumar Keshari for suggesting the problem. This material is based on work supported by IIT Bombay Postdoctoral Research Fellowship.

References

Department of Mathematics, Indian Institute of Technology Bombay, Powai, Mumbai 400076, India

Email address: [email protected], [email protected]

References

  1. [1]Abe, E.: Whitehead groups of Chevalley groups over polynomial rings. Comm. Algebra 11(12), 1271–1307 (1983)DOI
  2. [2]Apte, H., Stepanov, A.: Local-global principle for congruence subgroups of Chevalley groups. Central Eur. J. Math. 12, 801–812 (2014)
  3. [3]Basu, R., Mishra, M.K.: Quillen-Suslin theory for classical groups: revisited over graded rings. Contemp. Math. 751, 5–18 (2020)DOI
  4. [4]Grunewald, F., Mennicke, J., Vaserstein, L.N.: On symplectic groups over polynomial rings. Math. Zeitschrift 206, 35–56 (1991)DOI
  5. [5]Gubeladze, J.: The elementary action on unimodular rows over a monoid ring. J. Algebra 148, 135–161 (1992)DOI
  6. [6]Hahn, A.J., O’Meara, O.T.: The classical groups and K-theory. Springer, Berlin (1989)DOI
  7. [7]Lam, T.Y.: Serre’s problem on projective modules. Springer Monographs in Mathematics, Springer-Verlag, Berlin (2006)DOI
  8. [8]Lavrenov, A.: A local-global principle for symplectic K₂. Doc. Math. 23, 653–675 (2018)DOI
  9. [9]Lavrenov, A., Sinchuk, S.: On centrality of even orthogonal K₂. J. of Pure and Appl. Algebra 221(2), 1134–1145 (2017)DOI
  10. [10]Quillen, D.: Projective modules over polynomial rings. Invent. Math. 36, 167–171 (1976)DOI
  11. [11]Sinchuk, S.: On centrality of K₂ for Chevalley groups of type Eₗ. J. of Pure and Appl. Algebra 220(2), 857–875 (2016)DOI
  12. [12]Sinchuk, S.: On the A¹-invariance of K₂ of Chevalley groups of simply-laced type. Eur. J. Math. 11(14), 1–42 (2025)
  13. [13]Steinberg, R.: Lectures on Chevalley groups. Yale University (1968)
  14. [14]Suslin, A.A.: On the structure of the special linear group over polynomial rings. Math. USSR-Izv. 11, 221–238 (1977)DOI
  15. [15]Suslin, A.A., Kopeiko, V.I.: Quadratic modules and orthogonal group over polynomial rings. J. Sov. Math. 20(6), 2665–2691 (1982)DOI
  16. [16]Tulenbaev, M.S.: The Steinberg group of a polynomial ring. Math. USSR Sb. 45(1), 139–145 (1983)DOI
  17. [17]Vavilov, N.A., Plotkin, E.: Chevalley groups over commutative rings: I. Elementary calculations. Acta Appl. Math. 45(1), 73–113 (1996)DOI
  18. [18]Wendt, W.: A¹-homotopy of Chevalley groups. J. K-Theory 5, 245–287 (2010)

Paper details

vlnc:W_fajeknj8
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 9

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents