Arithmetic and Asymptotic Properties of Restricted Totient Sums
Es-said En-naoui · 2025
Read Arithmetic and Asymptotic Properties of Restricted Totient Sums by Es-said En-naoui (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2509-07004.txt.
Introduction
Es-said En-naoui, writing alone in a 2025 mathematics paper, studies restricted sums of Euler's totient function. The totient counts numbers up to a given value that share no common factor with it, and its summatory behavior is a classical topic in analytic number theory. This article extends the author's previous study of those sums.
Two complementary restricted sums are investigated: one over integers coprime to a prime, and one over integers divisible by that prime, which together decompose the full totient sum. Recurrence formulas, congruence relations, and generating function identities are established. A congruence result holds for every prime, and an asymptotic expansion with an error term is derived. The paper also studies average orders and connections with divisor structures, refining understanding of totients in arithmetic progressions.
Catalog record
| CREATOR | Es-said En-naoui |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-07004
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Arithmetic and Asymptotic Properties of Restricted Totient Sums Es-said En-naoui Abstract This article extends our previous study [ 5 ] on the summatory behavior of Euler’s totient function φ ( n ) \varphi(n) . We investigate two complementary restricted sums, Υ ( x , p ) = ∑ k ≤ x gcd ( k , p ) = 1 φ ( k ) , Δ ( x , p ) = ∑ k ≤ x p | k φ ( k ) , \Upsilon(x,p)=\sum_{\begin{subarray}{c}k\leq x\\ \gcd(k,p)=1\end{subarray}}\varphi(k),\qquad\Delta(x,p)=\sum_{\begin{subarray}{c}k\leq x\\ p\mid k\end{subarray}}\varphi(k), which satisfy the decomposition Ψ ( x ) = Υ ( x , p ) + Δ ( x , p ) \Psi(x)=\Upsilon(x,p)+\Delta(x,p) , where Ψ ( x ) = ∑ k ≤ x φ ( k ) \Psi(x)=\sum_{k\leq x}\varphi(k) . We establish recurrence formulas, congruence relations, and generating function identities for Δ ( x , p ) \Delta(x,p) . In particular, we prove that Δ ( x , p ) ≡ 0 ( mod p − 1 ) \Delta(x,p)\equiv 0\pmod{p-1} for every prime p p , and derive the asymptotic expansion Δ ( x , p ) = 3 π 2 ( p + 1 ) x 2 + O ( x log x ) . \Delta(x,p)=\frac{3}{\pi^{2}(p+1)}\,x^{2}+O(x\log x). Furthermore, we study average orders, connections with ω ( n ) \omega(n) , and relations with divisor structures. These results refine the analytic understanding of totients in arithmetic progressions and complement the asymptotic theory of Ψ ( x ) \Psi(x) . A Preprint 1 Introduction Euler’s totient function \varphi(n) is one of the central objects in elementary number theory [ 1 ] . Defined by \varphi(n)=\#\{\,1\leq k\leq n:\gcd(k,n)=1\,\}, it measures the number of integers up to n that are coprime to n . The multiplicative nature of \varphi , together with its deep connections to prime factorization, has made it a cornerstone in arithmetic and analytic investigations. A classical object of study is the summatory function \Psi(x)=\sum_{n\leq x}\varphi(n), whose asymptotic behavior, \Psi(x)=\frac{3}{\pi^{2}}x^{2}+O(x\log x), was established in the early 20th century and plays a crucial role in understanding the distribution of totients. In this article we extend this framework by introducing two complementary restricted sums associated with a fixed prime p : \Upsilon(x,p)=\sum_{\begin{subarray}{c}n\leq x\\ \gcd(n,p)=1\end{subarray}}\varphi(n),\qquad\Delta(x,p)=\sum_{\begin{subarray}{c}n\leq x\\ p\mid n\end{subarray}}\varphi(n). These satisfy the natural decomposition \Psi(x)=\Upsilon(x,p)+\Delta(x,p), which allows us to filter the contribution of \varphi(n) according to divisibility by a prime. The main goal of this paper is to investigate arithmetic and analytic properties of \Delta(x,p) , including congruence relations, recurrence identities, generating functions, and asymptotic behavior. In particular, we prove that \Delta(x,p)\equiv 0\pmod{p-1} for all primes p , and establish the asymptotic formula : \Delta(x,p)=\frac{3}{\pi^{2}(p+1)}\,x^{2}+O(x\log x). These results provide a refined understanding of how Euler’s totient values are distributed among multiples of a fixed prime, and they complement the classical theory of \Psi(x) . The methods combine elementary number-theoretic arguments with analytic techniques, and suggest further connections with divisor sums and multiplicative functions. 2 Preliminaries and Basic Properties In this section we recall some standard facts about Euler’s totient function and establish the basic framework for our study of restricted sums. 2.1 Euler’s Totient Function Definition 2.1 . For n\in\mathbb{N} , Euler’s totient function is defined by \varphi(n)=\#\{\,1\leq k\leq n:\gcd(k,n)=1\,\}. It is well known that \varphi is multiplicative, i.e., [ 4 ] \gcd(m,n)=1\;\;\Rightarrow\;\;\varphi(mn)=\varphi(m)\varphi(n). Moreover, if n has prime factorization n=\prod_{i=1}^{r}p_{i}^{\alpha_{i}} , then \varphi(n)=n\prod_{i=1}^{r}\left(1-\frac{1}{p_{i}}\right). (1) 2.2 Summatory Totient Function A central object in analytic number theory is the summatory function \Psi(x)=\sum_{n\leq x}\varphi(n). Lemma 2.1 (Dirichlet, 19th century) . For x\geq 1 , the following asymptotic holds: [ 2 ] [ 3 ] \Psi(x)=\frac{3}{\pi^{2}}x^{2}+O(x\log x). This formula shows that \Psi(x) grows quadratically with x , with error controlled by O(x\log x) . It plays a key role in our later analysis. 2.3 Restricted Totient Sums Fix a prime p . We introduce the following restricted versions of \Psi(x) : Definition 2.2 . For x\geq 1 and a prime p , define \Upsilon(x,p)=\sum_{\begin{subarray}{c}n\leq x\\ \gcd(n,p)=1\end{subarray}}\varphi(n),\qquad\Delta(x,p)=\sum_{\begin{subarray}{c}n\leq x\\ p\mid n\end{subarray}}\varphi(n). Clearly, the two functions satisfy the decomposition \Psi(x)=\Upsilon(x,p)+\Delta(x,p). (2) Thus \Upsilon(x,p) measures the contribution of totients coprime to p , while \Delta(x,p) isolates the contribution of totients supported on multiples of p . 2.4 First Examples We illustrate these definitions with a small example. Let x=10 and p=2 . \Psi(10)=\varphi(1)+\cdots+\varphi(10)=32. The odd terms contribute \Upsilon(10,2)=\varphi(1)+\varphi(3)+\varphi(5)+\varphi(7)+\varphi(9)=16, while the even terms contribute \Delta(10,2)=\varphi(2)+\varphi(4)+\varphi(6)+\varphi(8)+\varphi(10)=16. As expected, the decomposition identity ( 2 ) holds: \Psi(10)=\Upsilon(10,2)+\Delta(10,2)=32 . Lemma 2.2 . Let n a positive integer and p\in\mathbb{P} , and put \Delta_{p}(n)=\sum\limits_{\underset{p|k}{k=1}}^{n}\phi(k) , Then we have : \Delta_{p}(n)=(p-1)\sum_{\alpha=1}^{\big[\frac{ln(n)}{ln(p)}\big]}\Psi\bigg(\big[\frac{n}{p^{\alpha}}\big]\bigg) (3) Proof. (See : [ 5 ] ) ∎ 3 Congruence Properties of \Delta(x,p) In this section we prove several elementary but useful congruence and arithmetic properties of the function \Delta(x,p)=\sum_{\begin{subarray}{c}n\leq x\\ p\mid n\end{subarray}}\varphi(n) for a fixed prime p . The main result is that \Delta(x,p) is always divisible by p-1 . We also record a simple finite-difference identity which will be used later. Proposition 3.1 . For every integer n\geq 1 , We have : \Delta(n+1,p)-\Delta(n,p)=\begin{cases}\varphi(n+1),&p\mid(n+1),\\[4.0pt] 0,&p\nmid(n+1).\end{cases} Proof. By definition \Delta(t,p)=\sum_{\begin{subarray}{c}1\leq k\leq t\\ p\mid k\end{subarray}}\varphi(k) . Therefore \Delta(n+1,p)-\Delta(n,p)=\sum_{\begin{subarray}{c}1\leq k\leq n+1\\ p\mid k\end{subarray}}\varphi(k)-\sum_{\begin{subarray}{c}1\leq k\leq n\\ p\mid k\end{subarray}}\varphi(k). All terms with k\leq n cancel, so only the term k=n+1 (if any) remains. If p\mid(n+1) then that remaining term equals \varphi(n+1) ; otherwise there is no remaining term and the difference is 0 . This yields the stated case distinction. ∎ Proposition 3.2 . For any prime p and any integer n\geq 1 , We have : \Delta(n,p)\equiv 0\pmod{p-1}. Proof. Write the definition \Delta(n,p)=\sum_{m=1}^{\lfloor n/p\rfloor}\varphi(pm). Fix m\geq 1 and factor m=p^{a}r with a\geq 0 and \gcd(r,p)=1 . Then pm=p^{a+1}r and, by multiplicativity of \varphi on coprime integers, \varphi(pm)=\varphi\bigl(p^{a+1}\bigr)\varphi(r). For a prime power we have \varphi(p^{a+1})=p^{a}(p-1) , hence \varphi(pm)=p^{a}(p-1)\varphi(r), which is divisible by p-1 . Because every term \varphi(pm) in the sum defining \Delta(n,p) is divisible by p-1 , the whole sum is divisible by p-1 . Thus \Delta(n,p)\equiv 0\pmod{p-1} . ∎ 4 Connections of \Delta(n,p) with Other arithmetic Functions In this section we record several additional identities satisfied by the restricted summatory function \Delta(n,p) . These identities connect \Delta with arithmetic functions such as \omega(n) , the number of distinct prime divisors, and reveal combinatorial and generating function structures. Proposition 4.1 . Let n\geq 1 be an integer and for each prime p . we have : \sum_{p}\Delta(n,p)=\sum_{k=1}^{n}\varphi(k)\,\omega(k), where \omega(k) denotes the number of distinct prime divisors of k . Proof. Start from the definition and sum over primes p\leq n : \sum_{p\leq n}\Delta(n,p)=\sum_{p\leq n}\sum_{\begin{subarray}{c}1\leq k\leq n\\ p\mid k\end{subarray}}\varphi(k). Both sums are finite, so we may interchange the order of summation: \sum_{p\leq n}\sum_{\begin{subarray}{c}1\leq k\leq n\\ p\mid k\end{subarray}}\varphi(k)=\sum_{k=1}^{n}\sum_{\begin{subarray}{c}p\leq n\\ p\mid k\end{subarray}}\varphi(k). Fix an integer k with 1\leq k\leq n . For such a fixed k the inner condition \{\,p\leq n\text{ prime}:\ p\mid k\,\} is equivalent to \{\,p\text{ prime}:\ p\mid k\,\} because any prime divisor p of k satisfies p\leq k\leq n . Hence \sum_{\begin{subarray}{c}p\leq n\\ p\mid k\end{subarray}}\varphi(k)=\sum_{\begin{subarray}{c}p\ \text{prime}\\ p\mid k\end{subarray}}\varphi(k). The value \varphi(k) does not depend on p , so the inner sum is \varphi(k) multiplied by the number of distinct prime divisors of k , which is \omega(k)