🔢 mathematics

The Brauer group of a Stein algebra

This paper investigates the Brauer group of the ring of holomorphic functions on a finite-dimensional Stein space by providing a purely topological computation, establishing a comparison theorem between the étale cohomology of its spectrum and the singular cohomology of the space, and proving a purity theorem and index results for nonsingular cases.

Original authors: Olivier Benoist, James Hotchkiss

Published 2026-02-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Olivier Benoist, James Hotchkiss

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to understand the "shape" of a complex, invisible city. In mathematics, there are two main ways to look at this city:

The Algebraic View: You look at the city through a list of rules, equations, and functions (like a blueprint or a code).
The Topological View: You look at the city as a physical shape, a rubber sheet that can be stretched or twisted, focusing on holes, loops, and connectivity.

Usually, these two views are very different. But in the world of Stein spaces (a special type of complex city in mathematics), the authors of this paper, Olivier Benoist and James Hotchkiss, discovered a magical bridge. They proved that for these specific cities, the "rules" (algebra) and the "shape" (topology) are actually telling the exact same story.

Here is a breakdown of their journey, using simple analogies.

1. The Main Discovery: The "Shape" of the Rules

The paper focuses on the Brauer Group. Think of the Brauer Group as a "fingerprint" of a mathematical object. It measures how complicated the "twists" and "knots" are in the structure of that object.

The Problem: For a long time, mathematicians didn't know if the Brauer Group of a Stein space (the algebraic rules) was the same as the Brauer Group of the same space viewed as a shape (the topology).
The Solution: The authors proved they are identical.
- The Analogy: Imagine you have a knotted rope. One person describes the knot by writing down the exact sequence of moves to tie it (Algebra). Another person describes it by just looking at the shape of the knot (Topology). The authors proved that for Stein spaces, if you know the shape, you automatically know the exact sequence of moves, and vice versa. You don't need the complicated math; you just need to look at the shape.

2. The "Hole" Detector (Cohomology)

To prove this, they used a tool called Cohomology, which is like a "hole detector."

If you poke a donut, you find a hole. If you poke a sphere, you don't.
The authors showed that the Brauer Group is essentially counting specific types of "holes" (specifically, 3-dimensional holes) in the space.
The Result: They found a direct translation formula:

Brauer Group = The number of 3D holes (with some specific math rules applied).

This is huge because counting holes is much easier than solving complex algebraic equations.

3. The "Perfect" vs. "Imperfect" Cities

The paper also explores what happens when the city is "perfect" (smooth, no sharp corners) versus "imperfect" (has singularities or sharp points).

The Perfect City (Nonsingular): If the Stein space is smooth, the authors proved a Purity Theorem.
- The Analogy: Imagine a pristine lake. If you drop a stone in the middle, the ripples travel perfectly to the edge. The authors showed that in a smooth Stein space, the "ripples" (mathematical properties) of the whole city are perfectly determined by the ripples of its "shoreline" (the meromorphic functions, or the functions that might have holes in them). You can reconstruct the whole lake just by looking at the water near the edge.
The Imperfect City: If the space has sharp points, things get messy. The "ripples" don't always match up perfectly. The authors showed that in these cases, the algebraic rules and the topological shape can start to disagree in subtle ways.

4. The "Period" vs. The "Index"

Finally, the paper looks at two specific numbers associated with these knots:

Period: How many times you have to twist the knot before it untangles itself.
Index: The smallest size of a "box" (a mathematical structure) needed to hold the knot.

In many mathematical worlds, these two numbers are related by a simple rule (Index divides Period). The authors found that in Stein spaces, this rule usually holds, BUT they also found a counter-example.

The Analogy: They found a specific "knot" in a Stein space where the number of twists needed to untie it (Period) is small, but the box needed to store it (Index) is huge. It's like having a tiny, tightly wound spring that requires a massive, heavy crate to transport. This proves that even in these "perfect" mathematical cities, there are still surprising, complex secrets hiding in the shape.

Summary

In plain English, this paper says:

"We studied a special kind of mathematical city called a Stein space. We discovered that its algebraic rules are completely determined by its physical shape. We can calculate its most complex invariants just by counting its holes. However, even in these well-behaved cities, there are still some knots that are surprisingly difficult to untangle, proving that the relationship between shape and rules is deep, beautiful, but not entirely simple."

This work is significant because it allows mathematicians to use the tools of topology (which are often more intuitive and visual) to solve problems in algebra (which are often abstract and difficult), specifically for a wide class of complex spaces.

1. Problem Statement and Context

The paper investigates the Brauer group of the ring $O(S)$ of holomorphic functions on a finite-dimensional Stein space $S$ . In algebraic geometry, the Brauer group $Br(R)$ of a ring $R$ is well-understood via étale cohomology (specifically, the torsion subgroup of $H^2_{\text{ét}}(\text{Spec}(R), \mathbb{G}_m)$ ). However, Stein algebras are complex-analytic objects that are generally not Noetherian, making standard algebraic techniques difficult to apply directly.

The central questions addressed are:

Can the algebraic Brauer group $Br(O(S))$ be computed purely topologically?
How does the étale cohomology of $\text{Spec}(O(S))$ compare to the singular cohomology of the underlying Stein space $S$ ?
What is the relationship between the Brauer group of holomorphic functions $O(S)$ and the field of meromorphic functions $M(S)$ ?
How do the period and index of Brauer classes behave in this analytic setting?

2. Methodology

The authors employ a hybrid methodology bridging complex analytic geometry, commutative algebra, and algebraic topology.

Analytification and Oka's Principle: The core strategy relies on the equivalence between the category of coherent sheaves on a Stein space $S$ and finitely presented quasi-coherent sheaves on $\text{Spec}(O(S))$ (established in prior work by the authors). They utilize Grauert's Oka Principle to show that holomorphic vector bundles and Azumaya algebras on $S$ are classified by the same topological invariants as their continuous counterparts.
Topological Reduction: The authors reduce algebraic problems to topological ones by proving that $Br(O(S))$ is isomorphic to the topological Brauer group $Br_{\text{top}}(S)$ .
Twisted Sheaves and Gerbes: To handle non-torsion classes in $H^2_{\text{ét}}(\text{Spec}(O(S)), \mathbb{G}_m)$ , the authors utilize the theory of twisted sheaves (following de Jong and Lieblich) and central separable algebras (following Taylor, Raeburn, Heinloth, and Schröer). This allows them to work with non-unital algebras and gerbes that represent cohomology classes not captured by standard Azumaya algebras.
Cohomological Comparison: They use comparison theorems between étale cohomology of the algebra and singular cohomology of the space, often leveraging the fact that Stein spaces have the homotopy type of finite-dimensional CW complexes.

3. Key Contributions and Results

A. Topological Computation of the Brauer Group

Theorem 1.1 & 1.3: The natural morphism from the algebraic Brauer group to the topological Brauer group is an isomorphism:
$Br(O(S)) \xrightarrow{\sim} Br_{\text{top}}(S)$
Furthermore, for any finite-dimensional Stein space $S$ , the topological Brauer group is isomorphic to the torsion part of the third singular cohomology group:
$Br(O(S)) \xrightarrow{\sim} H^3(S, \mathbb{Z})_{\text{tors}}$
Significance: This provides a purely topological formula for an algebraic invariant, bypassing the non-Noetherian nature of $O(S)$ .

B. Comparison Theorems in Degree 2

Theorem 1.4: For any finite-dimensional Stein space $S$ and integer $n \geq 1$ , the comparison map between étale cohomology of the algebra and singular cohomology of the space is an isomorphism in degree 2 with finite coefficients:
$H^2_{\text{ét}}(\text{Spec}(O(S)), \mathbb{Z}/n) \xrightarrow{\sim} H^2(S, \mathbb{Z}/n)$
Theorem 1.5: For invertible coefficients, the map $H^2_{\text{ét}}(\text{Spec}(O(S)), \mathbb{G}_m) \to H^2(S, O_S^\times)$ is injective but not surjective in general.
Significance: The failure of surjectivity occurs because $H^2(S, O_S^\times)$ can contain non-torsion classes (related to $H^3(S, \mathbb{Z})$ ), whereas the image of the algebraic Brauer group is torsion.

C. Purity and Meromorphic Functions

Theorem 1.6 & 1.7: If $S$ is a locally factorial Stein space (e.g., a non-singular manifold), the natural map $Br(O(S)) \to H^2_{\text{ét}}(\text{Spec}(O(S)), \mathbb{G}_m)$ is an isomorphism. Furthermore, the map to the meromorphic function field is injective:
$H^2_{\text{ét}}(\text{Spec}(O(S)), \mathbb{G}_m) \hookrightarrow H^2_{\text{ét}}(\text{Spec}(M(S)), \mathbb{G}_m)$
Theorem 1.8 (Purity Theorem): For a Stein manifold $S$ , there is an exact sequence:
$0 \to Br(O(S)) \to Br(M(S)) \xrightarrow{\text{res}_D} \bigoplus_{D} H^1_{\text{ét}}(\text{Spec}(M(D)), \mathbb{Q}/\mathbb{Z})$
where $D$ runs over irreducible analytic subsets of codimension 1. This is the analytic analogue of Grothendieck's purity conjecture for regular rings.

D. Period-Index Problem

Theorem 1.9 & 1.10: The index of a Brauer class $\alpha \in Br(O(S))$ (the GCD of degrees of representing Azumaya algebras) can be computed topologically. Specifically, $\text{ind}(\alpha) = \text{ind}(\alpha_{\text{top}})$ and $\text{ind}(\alpha) = \text{ind}(\alpha_{M(S)})$ .
Theorem 1.11: There exist connected Stein manifolds $S$ and classes $\alpha \in \text{Im}[Br(O(S)) \to Br(M(S))]$ such that the period and index differ ( $\text{ind}(\alpha) \neq \text{per}(\alpha)$ ).
Significance: This resolves the period-index problem in the Stein context, showing that the obstruction to equality is topological in nature, analogous to results in algebraic geometry but derived via topological K-theory and obstruction theory.

4. Significance and Impact

Bridging Algebra and Topology: The paper successfully establishes a dictionary between the algebraic properties of non-Noetherian Stein algebras and the topological properties of their underlying spaces. This validates the use of topological tools (like CW complex homotopy theory) to solve algebraic problems in complex analysis.
Refining the Brauer Group: It clarifies the limitations of the classical Brauer group (Azumaya algebras) in the analytic setting, necessitating the use of the "bigger" Brauer group (central separable algebras) to capture all cohomology classes in $H^2_{\text{ét}}$ .
Counterexamples and Bounds: The authors provide explicit counterexamples to surjectivity in degree 2 cohomology and demonstrate that the period-index bound in Stein geometry is sharp, relying on the topological period-index problem studied by Antieau and Williams.
Foundational for Stein Geometry: These results deepen the understanding of the cohomological structure of Stein spaces, offering a rigorous framework for comparing étale and singular cohomology in the analytic category.

In summary, Benoist and Hotchkiss demonstrate that for finite-dimensional Stein spaces, the complex algebraic structure of the Brauer group is entirely governed by the underlying topology, specifically the third cohomology group with integer coefficients, while carefully delineating where algebraic and analytic cohomologies diverge.