S. Krishnamoorthy,T. Dalal
Abstract:J.-P. Serre, in his paper [1], established a sufficient condition on $n$ for the $n$-th coefficient of the series $\eta^{26}$ to vanish. However, the question that whether this is a necessary condition remained unanswered. In this paper, using the theory of Hecke eigenforms explored by Serre, we prove some partial cases for the converse part.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the sufficiency and necessity of the vanishing conditions of the Fourier coefficient \(p_{26}(n)\) of the 26 -th power of the Dedekind η - function (i.e., \(\eta^{26}\)).
Specifically, J. - P. Serre has already established sufficient conditions in his paper for the \(n\)-th coefficient \(p_{26}(n)\) to vanish, but it has not been proven whether these conditions are also necessary. In this paper, by using the Hecke eigenform theory explored by Serre, the converse propositions in some partial cases are proven to be true, that is, under certain specific conditions, if \(p_{26}(n)=0\), then \(n\) must satisfy certain specific properties.
### Summary of the main problems in the paper
1. **Background and motivation**:
- The Dedekind η - function is defined as: \[
\eta(z) := q^{1/24} \prod_{m = 1}^{\infty}(1 - q^m),\quad q = e^{2\pi iz},\quad \text{Im}(z)>0
\]
- For an integer \(r\), \(\eta^r\) can be expressed as an infinite series: \[
\eta^r(z)=q^{r/24}\sum_{n = 0}^{\infty}p_r(n)q^n
\]
- The vanishing of the coefficient \(p_r(n)\) has attracted wide - spread interest. In particular, when \(r\) takes certain values, the set density of \(p_r(n)\neq0\) is zero, and at this time \(\eta^r\) is said to be lacunary.
2. **Known results**:
- Serre has proven that for \(r = 2,4,6,8,10,14\) and \(26\), \(\eta^r\) is lacunary.
- For \(r = 26\), Serre has given sufficient conditions for \(p_{26}(n)=0\), but it has not been proven whether these conditions are also necessary.
3. **Main contributions of this paper**:
- The author attempts to prove the necessary conditions for \(p_{26}(n)=0\).
- Specifically, they have verified some partial cases and finally proven the following theorems:
- **Theorem 1.4**: Let \(n\in\mathbb{N}\) and for every prime \(p\equiv1\pmod{12}\), we have \(\text{ord}_p(12n + 1)\not\equiv4\pmod{5}\). Then \(p_{26}(25n+1)=0\) if and only if there exists a prime \(p\equiv3\pmod{4}\) such that \(\text{ord}_p(12n + 1)\) is odd; and there exists a prime \(p'\equiv2\pmod{3}\) such that \(\text{ord}_{p'}(12n + 1)\) is odd.
- **Theorem 1.5**: Let \(n\in\mathbb{N}\) and for every prime \(p\equiv1\pmod{12}\), we have \(\text{ord}_p(12n + 1)\not\equiv6\pmod{7}\). Then \(p_{26}(49n+3)=0\) if and only if there exists a prime \(p\equiv3\pmod{4}\) such that \(\text{ord}_p(12n + 1)\) is odd; and there exists a prime \(p'\equiv2\pmod{3}\) such that \(\text{ord}_{p'}(12n + 1)\) is odd.
### Summary
In this paper, by using the Hecke eigenform theory, the Dedekind η is further explored.