Arithmetic function

In number theory, an arithmetic function (or number-theoretic function) is a real- or complex-valued function defined on the set of positive integers that expresses some arithmetical property of the integer. These functions are fundamental tools in analytic and algebraic number theory, providing a bridge between discrete integer properties and continuous analytical methods.

Definition and Basic Properties

Formally, an arithmetic function is a mapping $f: \mathbb{Z}^+ \to \mathbb{C}$, where $\mathbb{Z}^+$ denotes the set of positive integers. While the codomain is typically the set of complex numbers, many classical arithmetic functions take only integer or real values. The primary objective in studying these functions is to understand their behavior, classify them based on algebraic properties, and determine their asymptotic growth or average orders.

Because the domain is discrete, arithmetic functions can be represented as infinite sequences of numbers. However, unlike arbitrary sequences, arithmetic functions usually encode structural information about the integers, such as divisibility, prime factorization, or the distribution of prime numbers.

Multiplicative and Additive Functions

Arithmetic functions are frequently classified by how they behave with respect to the multiplication of integers. This classification is crucial because it allows the value of the function for any integer to be determined entirely by its values at prime powers.

Multiplicative Functions

An arithmetic function $f$ is called multiplicative if $f(1) = 1$ and $f(mn) = f(m)f(n)$ whenever $m$ and $n$ are coprime (i.e., $\gcd(m, n) = 1$). If the condition $f(mn) = f(m)f(n)$ holds for all positive integers $m$ and $n$, regardless of whether they are coprime, the function is said to be completely multiplicative.

For a multiplicative function, its value at any integer $n = p_1^{a_1} p_2^{a_2} \cdots p_k^{a_k}$ is completely determined by its values at the prime powers: $f(n) = f(p_1^{a_1}) f(p_2^{a_2}) \cdots f(p_k^{a_k})$.

Additive Functions

An arithmetic function $f$ is called additive if $f(1) = 0$ and $f(mn) = f(m) + f(n)$ whenever $\gcd(m, n) = 1$. If $f(mn) = f(m) + f(n)$ for all positive integers $m$ and $n$, the function is completely additive. Similar to multiplicative functions, an additive function is determined by its values at prime powers. Taking the exponential of an additive function yields a multiplicative function, and taking the logarithm of a positive multiplicative function yields an additive function.

Important Examples of Arithmetic Functions

Several arithmetic functions play central roles in number theory due to their connections with prime numbers and divisibility.

• Euler's Totient Function ($\varphi(n)$): Counts the number of positive integers up to $n$ that are coprime to $n$. It is a multiplicative function.
• Möbius Function ($\mu(n)$): Defined based on the prime factorization of $n$. It is $0$ if $n$ has a squared prime factor, $1$ if $n$ is a product of an even number of distinct primes, and $-1$ if $n$ is a product of an odd number of distinct primes. It is multiplicative and central to the Möbius inversion formula.
• Divisor Functions ($\sigma_k(n)$): Defined as the sum of the $k$-th powers of the positive divisors of $n$. The function $\sigma_0(n)$ (often denoted $d(n)$ or $\tau(n)$) counts the number of divisors, while $\sigma_1(n)$ (often denoted $\sigma(n)$) calculates the sum of the divisors. Both are multiplicative.
• Von Mangoldt Function ($\Lambda(n)$): Defined as $\ln p$ if $n = p^k$ for some prime $p$ and integer $k \ge 1$, and $0$ otherwise. It is neither multiplicative nor additive but is fundamentally important in the proof of the Prime Number Theorem.
• Liouville Function ($\lambda(n)$): Defined as $(-1)^{\Omega(n)}$, where $\Omega(n)$ is the total number of prime factors of $n$ counted with multiplicity. It is completely multiplicative.

Dirichlet Convolution

The set of arithmetic functions can be endowed with a ring structure using pointwise addition and Dirichlet convolution. The Dirichlet convolution of two arithmetic functions $f$ and $g$ is defined as:
$$ (f * g)(n) = \sum_{d|n} f(d)g\left(\frac{n}{d}\right) $$
where the sum is taken over all positive divisors $d$ of $n$.

Dirichlet convolution is commutative, associative, and distributive over pointwise addition. The identity element for this convolution is the function $\varepsilon(n)$, which is $1$ if $n=1$ and $0$ otherwise. Under this operation, the multiplicative functions form a subgroup of the group of units (invertible functions) in the ring of arithmetic functions. Many fundamental identities in number theory, such as $\sum_{d|n} \varphi(d) = n$ and the Möbius inversion formula, are elegantly expressed using Dirichlet convolution.

Dirichlet Series and Generating Functions

To apply the tools of complex analysis to arithmetic functions, mathematicians use generating functions, most notably the Dirichlet series. For an arithmetic function $f$, its Dirichlet series is defined as:
$$ F(s) = \sum_{n=1}^{\infty} \frac{f(n)}{n^s} $$
where $s$ is a complex variable.

A key property of Dirichlet series is that the Dirichlet series of a convolution $f * g$ is the product of the individual Dirichlet series of $f$ and $g$. For example, the Dirichlet series for the constant function $1$ is the Riemann zeta function $\zeta(s)$. The Dirichlet series for the Möbius function $\mu(n)$ is $1/\zeta(s)$. This algebraic correspondence allows researchers to study the analytic properties of the generating function (such as poles and zeros) to deduce asymptotic information about the arithmetic function.

Asymptotic Behavior and Analytic Number Theory

A major focus in the study of arithmetic functions is determining their average order and asymptotic behavior, as their exact values can fluctuate wildly. For instance, while the divisor function $d(n)$ can be arbitrarily large, its average order is $\ln n$.

Analytic number theory relies heavily on estimating sums of the form $\sum_{n \le x} f(n)$. Techniques such as Perron's formula, contour integration, and the study of the Riemann zeta function are employed to find precise asymptotic expansions for these sums. The Prime Number Theorem, which states that the number of primes less than $x$ is asymptotically $x / \ln x$, is equivalent to stating that the average order of the von Mangoldt function $\Lambda(n)$ is $1$. Understanding the error terms in these asymptotic formulas remains one of the most profound and active areas of mathematical research, intimately connected to the Riemann Hypothesis.