🔢 mathematics

On the non-existence of skew-Hadamard difference sets in certain non-abelian groups

This paper establishes the first general structural restrictions for skew-Hadamard difference sets in non-abelian groups by proving, via rational group algebra methods that avoid character theory, that any nilpotent group admitting such a set must be a $p$ -group.

Original authors: Vitor Araujo Garcia

Published 2026-06-12✓ Author reviewed ⓘ

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Vitor Araujo Garcia

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 by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a master architect trying to build a very specific, perfect structure called a Skew-Hadamard Difference Set (SHDS). This structure isn't made of bricks, but of numbers and relationships within a mathematical "group" (a collection of elements that can be combined in specific ways).

For a long time, mathematicians knew that if you want to build this structure, the "land" you build on (the group) has some very strict rules. If the land is Abelian (meaning the order in which you combine elements doesn't matter, like adding numbers), we know the land must be a specific type of "prime-numbered" territory. But what if the land is Non-Abelian (where the order of operations does matter, like putting on socks before shoes vs. shoes before socks)? Until this paper, that was a huge mystery.

Here is what the author, Vitor Araujo Garcia, discovered, explained through simple analogies:

1. The Problem: The "Order" Matters

In the world of Abelian groups, the rules for building this structure are well-known. But in the chaotic, non-Abelian world, mathematicians were stuck. They tried to use a tool called "character tables" (like a complex map of the land's DNA), but that map only works for the orderly Abelian lands. It breaks down completely for the messy, non-Abelian ones.

2. The New Tool: The "Rational Group Algebra"

Instead of using the broken map, the author invented a new way to look at the land. He used something called the Rational Group Algebra.

The Analogy: Imagine you have a giant, complex machine (the group). Instead of trying to trace every single wire (the characters), you look at the machine's "shadow" or its "skeleton" when projected onto a simpler screen. This screen is the Abelianization of the group (essentially, the part of the group where you ignore the order of operations and just look at the basic ingredients).
By looking at this simplified shadow, the author could derive rules that apply to the whole machine, even if the machine itself is chaotic.

3. The Big Discovery: The "Prime-Only" Rule

The paper proves a major new rule for building these structures in non-Abelian groups:

The Finding: If a group is Nilpotent (a type of group that is "almost" Abelian, or can be built up from simple layers) and it admits an SHDS, then that group must be a p-group.
The Translation: A "p-group" is a land where every single element's size is a power of a single prime number (like 3, 7, or 11). You cannot have a mix of different prime numbers (like a land with both 3s and 5s) if you want to build this structure.
Why it matters: This is the first time anyone has proven a general structural rule for these sets in non-Abelian groups. Before this, we only knew this for the orderly Abelian groups. Now we know that even in the messy, non-Abelian world, if the group is "nilpotent," it still has to be a single-prime territory.

4. The "Square Root" Test

How did the author prove this?

The Analogy: Imagine you have a magic equation that says, "To build this structure, you must be able to take the square root of a negative number related to the size of your land."
The author showed that if your land has a mix of different prime numbers (like having both 3s and 5s in its size), the math breaks. You end up trying to take a square root of a number that simply doesn't exist in the mathematical "neighborhood" you are looking at.
Therefore, the land must be made of only one type of prime number to make the math work.

5. What We Still Don't Know

The paper is very careful to say what it doesn't prove.

The Conjecture: The author suspects that any group (even those that aren't "nilpotent") that admits this structure must be a p-group.
The Gap: However, the paper admits this is still unproven for certain tricky groups (like a specific mix of a 49-cycle and a 3-cycle). The author says, "We don't know yet if these specific tricky groups can hold the structure."

Summary

Think of this paper as a new set of building codes for a very exclusive club.

Old Rule: We knew the rules for the "Orderly Club" (Abelian groups).
New Rule: We now know that even for the "Chaos Club" (Non-Abelian groups), if the club is "mostly orderly" (Nilpotent), they still have to follow the Single-Prime Rule. You can't mix different prime numbers in your membership if you want to build the special structure.

The author didn't just guess this; they built a new mathematical lens (using rational group algebras) that allowed them to see these rules clearly for the first time, without needing the old, broken tools.

Technical Summary: On the non-existence of skew-Hadamard difference sets in certain non-abelian groups

Problem Statement
The paper addresses the existence of skew-Hadamard difference sets (SHDS) in finite groups. An SHDS in a group $G$ of order $v$ is a subset $D$ of size $k$ such that $G$ is the disjoint union of $D$ , $D^{(-1)}$ (the set of inverses), and the identity element $\{1\}$ . This structure implies specific parameters ( $v=4n-1, k=2n-1, \lambda=n-1$ ) and satisfies the identity $DD^{(-1)} = n + \lambda G$ in the rational group algebra $\mathbb{Q}[G]$ .

While the abelian case is well-studied, with known structural constraints (e.g., $G$ must be a $p$ -group for some prime $p \equiv 3 \pmod 4$ ) and a conjecture that $G$ must be elementary abelian, the non-abelian case remains largely unexplored. Existing proofs for the abelian case rely heavily on group characters and character tables, a method that does not generalize well to non-abelian groups. The paper seeks to establish necessary conditions for the order and structure of finite groups admitting an SHDS, specifically targeting the non-abelian setting without relying on character theory.

Methodology
The author employs a purely algebraic approach using the structure of the rational group algebra $\mathbb{Q}[G]$ , deliberately avoiding the use of group characters. The methodology relies on:

Decomposition of Group Algebras: Utilizing the Wedderburn-Artin theorem (Theorem 1.1) to decompose $\mathbb{Q}[G]$ into a direct sum of simple components (matrix rings over division rings).
Abelianization: Analyzing the relationship between $\mathbb{Q}[G]$ and the group algebra of the abelianization $G/G'$ , where $G'$ is the commutator subgroup.
Central Idempotents: Constructing specific central idempotents $e_1 = \widehat{G'}(1 - \widehat{G})$ and $e_2 = 1 - e_1$ to isolate components of the algebra.
Algebraic Number Theory: Applying properties of cyclotomic fields $\mathbb{Q}(\zeta_d)$ and quadratic subfields to derive contradictions regarding the existence of square roots of negative integers within these fields.

Key Derivations and Results
The core of the paper establishes a necessary condition for the existence of an SHDS based on the order of the group and its abelianization.

Key Identity: For a group $G$ admitting an SHDS $D$ , the element $x = D - D^{(-1)}$ satisfies $x^2 = G - |G|$ in $\mathbb{Q}[G]$ .
Theorem 2.2: If $G$ admits an SHDS, then $\sqrt{-|G|}$ must belong to the cyclotomic field $\mathbb{Q}(\zeta_d)$ for every integer $d > 1$ such that $G/G'$ contains a cyclic subgroup of order $d$ .
Corollary 2.3: If a prime factor $\pi_1$ $π_{1}$ of $|G|$ $∣ G ∣$ has an odd multiplicity in the primary decomposition of $|G|$ $∣ G ∣$ , and the order of the abelianization $|G/G'|$ $∣ G / G^{'} ∣$ contains a distinct prime factor $\pi_2$ $π_{2}$ , then $G$ $G$ cannot admit an SHDS.
- Implication: If an SHDS exists, $G/G'$ must be a $p$ -group for some prime $p \equiv 3 \pmod 4$ , and the multiplicity of $p$ in $|G|$ must be odd.
Corollaries 2.5–2.7: These results eliminate groups where:
- Primes $q \equiv 1 \pmod 4$ appear with odd multiplicity in $|G|$ .
- Multiple distinct primes $p \equiv 3 \pmod 4$ appear with odd multiplicity in $|G|$ .
Corollary 2.8 (Main Structural Result): If $G$ $G$ is a finite nilpotent group admitting an SHDS, then $G$ $G$ must be a $p$ -group.
- Proof Logic: A nilpotent group is a direct product of its Sylow subgroups. If $|G|$ had multiple prime factors, the abelianization would reflect this, violating the conditions derived in Corollary 2.3.

Significance and Claims
The paper claims to provide the first general structural restrictions for SHDSs in non-abelian groups.

Generalization: It generalizes the known result that abelian groups admitting SHDSs are $p$ -groups to the broader class of nilpotent groups.
Methodological Shift: The work demonstrates that structural constraints can be derived using the rational group algebra and abelianization, bypassing the limitations of character theory which are difficult to apply in non-abelian contexts.
Conjecture: The results support Conjecture 2.9, which posits that any finite group admitting an SHDS must be a $p$ -group. The author notes that while this is now settled for nilpotent groups, the conjecture remains open for other non-abelian classes (e.g., metacyclic groups like $C_{49} \rtimes C_3$ ).

The paper concludes by acknowledging that while these necessary conditions eliminate broad classes of groups, finding specific exponent bounds or proving the conjecture for all non-abelian groups remains a difficult open problem, particularly given the existence of SHDSs in non-abelian $p$ -groups of order $p^3$ .