Abstract
For a prime $p$ and an arbitrary finite group $G$, we show that if $p^{2}$ does not divide the size of each conjugacy class of $p$-regular element (element of order not divisible by $p$) in $G$, then the largest power of $p$ dividing the index $|G:\mathbf{O}_{p}(G)|$ is at most $p^{2}$.
Introduction
In the study of finite groups, a central theme is the relationship between the global structure of a group and its internal properties, such as the sizes of its conjugacy classes or the degrees of its irreducible characters. A classic example of this relationship is the Itô-Michler theorem, which connects the divisibility of character degrees by a prime $p$ to the existence of a normal, abelian Sylow $p$-subgroup. Specifically, if $p$ does not divide any irreducible character degree, the $p$-core $\mathbf{O}_p(G)$ is large enough that the index $|G : \mathbf{O}_p(G)|$ is not divisible by $p$.
This paper addresses a dual version of this problem: if we restrict the power of $p$ that can divide the sizes of conjugacy classes, can we provide a sharp bound for the $p$-part of the index of the $p$-core? While previous research established looser bounds, this work provides a definitive, sharp upper bound for the case where $p^2$ is the restricted divisor.
Main Result: Theorem A
The primary contribution of this research is the proof of Theorem A:
For any prime $p$ and finite group $G$, if $p^2 \nmid |x^G|$ for every $p$-regular element $x \in G$, then $|G : \mathbf{O}_p(G)|_p \leq p^2$.
Here, $|x^G|$ denotes the size of the conjugacy class containing $x$, and $x$ is $p$-regular if its order is not divisible by $p$. The notation $|G : \mathbf{O}_p(G)|_p$ refers to the largest power of $p$ dividing the index of the $p$-core. This result is significant because it improves upon the previous bound of $p^4$ established by Yang and Qian in 2014, reaching the optimal theoretical limit.
Methodology
The authors utilize a rigorous theoretical framework to establish this bound, primarily relying on a "minimal counterexample" strategy. The proof involves several key stages:
- Reduction and Induction: The authors assume the existence of a minimal counterexample $G$ and demonstrate that any normal subgroup or factor group of $G$ would also satisfy the hypothesis, leading toward a contradiction.
- Solvable Radical Analysis: By analyzing the solvable radical (the largest normal solvable subgroup), the authors reduce the problem to cases where the group has a trivial solvable radical, allowing them to focus on the structure of non-solvable groups.
- $p$-Block Theory: A critical component of the proof involves modular representation theory. The authors examine the defect groups of $p$-blocks and the heights of characters within those blocks. This allows them to bridge the gap between the combinatorial property of conjugacy class sizes and the algebraic property of character degrees.
- Integration of Recent Results: For the case of odd primes, the authors invoke a recent theorem by Qian that relates character degrees to the $p$-core index, which is essential for resolving the minimal counterexample.
Discussion and Sharpness
A key feature of this result is its sharpness. The authors note that the bound $|G : \mathbf{O}_p(G)|_p \leq p^2$ cannot be improved further, as there exist known groups where the index is exactly $p^2$ under the given conditions. This settles the question for the $a=2$ case (where $p^a$ is the divisor).
The focus on $p$-regular elements is also noteworthy. In modular representation theory, $p$-regular elements often provide more granular information about the $p$-structure of a group than the full set of elements. By showing that constraints on these specific classes are sufficient to bound the $p$-core index, the paper reinforces the importance of $p$-regularity in group-theoretic bounds.
Conclusion and Future Directions
This paper successfully tightens the bound on the $p$-part of the index of the $p$-core to $p^2$, providing a definitive answer for this specific constraint. Future research may look toward generalizing this result for higher powers of $p$. A natural hypothesis would be to investigate if a general bound exists such that if $p^a \nmid |x^G|$, then $|G : \mathbf{O}_p(G)|_p \leq p^{f(a)}$ for some function $f$. Additionally, exploring these relationships in the context of non-complex characters or more general algebraic structures remains an open area of interest.
Summary updated