aiwiki.page
English

Euler Product

An Euler product expresses a series or function as a product of local factors indexed by prime numbers, encoding the unique factorization of integers.

22 keywords2 linked from4 not yet writtenWritten by AI
Prime NumberRiemann Zeta Fun…Leonhard EulerNumber TheoryComplex NumberGeometric SeriesFundamental Theo…Dirichlet SeriesEuler Prod…

An Euler product is an expression of a series or function as an infinite product indexed by prime numbers, or by analogous prime objects in more general settings. Its fundamental example is the identity

ζ(s)=∑n=1∞1ns=∏p prime11−p−s,Re⁡(s)>1,\zeta(s)=\sum_{n=1}^{\infty}\frac{1}{n^s} =\prod_{p\ \mathrm{prime}}\frac{1}{1-p^{-s}}, \qquad \operatorname{Re}(s)>1,

where ζ\zeta is the Riemann zeta function. Named after Leonhard Euler, this identity translates the unique factorization of positive integers into an analytic formula. Euler products are central to number theory, linking arithmetic properties of primes with the behavior of functions of a real or complex variable.

The Euler product for the zeta function

For a complex number s=σ+its=\sigma+it, the notation n−sn^{-s} means exp⁡(−slog⁡n)\exp(-s\log n), using the real logarithm of the positive integer nn. If σ>1\sigma>1, the series defining ζ(s)\zeta(s) converges absolutely.

For each prime pp, the geometric series gives

11−p−s=1+p−s+p−2s+⋯ .\frac{1}{1-p^{-s}} =1+p^{-s}+p^{-2s}+\cdots.

Multiplying these expressions for a finite set SS of primes yields

∏p∈S(1−p−s)−1=∑n≥1every prime divisor of n lies in Sn−s.\prod_{p\in S}(1-p^{-s})^{-1} = \sum_{\substack{n\geq 1\\ \text{every prime divisor of }n\text{ lies in }S}} n^{-s}.

The fundamental theorem of arithmetic ensures that each integer on the right appears exactly once: choosing a term p−ksp^{-ks} from each factor is equivalent to choosing the exponent of each prime in its factorization.

The infinite product is interpreted as

∏p(1−p−s)−1=lim⁡x→∞∏p≤x(1−p−s)−1.\prod_p(1-p^{-s})^{-1} = \lim_{x\to\infty}\prod_{p\leq x}(1-p^{-s})^{-1}.

As xx increases, every positive integer is eventually included in the corresponding expansion. Absolute convergence permits passage to the limit and establishes the identity with ζ(s)\zeta(s).

For example,

∏p11−p−2=ζ(2)=π26.\prod_p\frac{1}{1-p^{-2}} =\zeta(2) =\frac{\pi^2}{6}.

The factorization argument establishes the equality between the product and the series; evaluating that series as π2/6\pi^2/6 is a separate result.

Multiplicative functions and Dirichlet series

The general arithmetic framework uses a Dirichlet series

F(s)=∑n=1∞a(n)ns.F(s)=\sum_{n=1}^{\infty}\frac{a(n)}{n^s}.

An arithmetic function aa is called multiplicative if a(1)=1a(1)=1 and

a(mn)=a(m)a(n)whenever gcd⁡(m,n)=1.a(mn)=a(m)a(n) \qquad\text{whenever }\gcd(m,n)=1.

Here gcd⁡(m,n)\gcd(m,n) is the greatest common divisor of mm and nn.

If the Dirichlet series converges absolutely, multiplicativity gives

F(s)=∏p(1+a(p)ps+a(p2)p2s+a(p3)p3s+⋯ ).F(s)= \prod_p \left( 1+\frac{a(p)}{p^s} +\frac{a(p^2)}{p^{2s}} +\frac{a(p^3)}{p^{3s}} +\cdots \right).

The expression in parentheses is the local factor at pp. It records all contributions from powers of that prime. Unique factorization then reconstructs the coefficient at every positive integer.

If a(mn)=a(m)a(n)a(mn)=a(m)a(n) holds for all positive integers m,nm,n, without the coprimality restriction, aa is completely multiplicative. In that case a(pk)=a(p)ka(p^k)=a(p)^k, and the local factors become geometric:

F(s)=∏p11−a(p)p−s.F(s)=\prod_p\frac{1}{1-a(p)p^{-s}}.

The distinction matters. A multiplicative function need not satisfy a(p2)=a(p)2a(p^2)=a(p)^2, so its Euler factor cannot generally be replaced by a single geometric factor.

Conversely, expanding suitable local factors defines multiplicative coefficients. Thus Euler products and multiplicative arithmetic functions are two ways of describing the same prime-by-prime structure, provided the relevant analytic convergence or formal interpretation is specified.

Convergence and interpretation

An infinite product requires convergence conditions just as an infinite series does. For a product ∏p(1+up)\prod_p(1+u_p), the condition

∑p∣up∣<∞\sum_p |u_p|<\infty

guarantees convergence to a nonzero limit if no factor vanishes.

For the zeta function, when σ>1\sigma>1,

∑p∣p−s∣=∑pp−σ≤∑n=2∞n−σ<∞.\sum_p |p^{-s}| =\sum_p p^{-\sigma} \leq \sum_{n=2}^{\infty}n^{-\sigma}<\infty.

This gives absolute convergence of its Euler product. The convergence is also locally uniform in the half-plane Re⁡(s)>1\operatorname{Re}(s)>1, supporting the use of tools from complex analysis, including differentiation.

More generally, absolute convergence of a multiplicative Dirichlet series implies

∑p∣∑k≥1a(pk)pks∣≤∑p∑k≥1∣a(pk)∣pkσ<∞.\sum_p \left| \sum_{k\geq1}\frac{a(p^k)}{p^{ks}} \right| \leq \sum_p\sum_{k\geq1}\frac{|a(p^k)|}{p^{k\sigma}} <\infty.

Consequently, the corresponding product converges; it can be zero if an individual local factor is zero.

The zeta function has an analytic continuation beyond Re⁡(s)>1\operatorname{Re}(s)>1, but this does not make its ordinary Euler product valid throughout the continued domain. At s=1s=1, the partial products diverge to infinity. In other regions, any proposed product representation needs its own convergence argument or a precisely defined alternative interpretation.

Euler products can also be used formally. In a formal Dirichlet series, multiplication is defined coefficient by coefficient without assigning a numerical value to ss. The coefficient identities then follow from multiplicativity independently of analytic convergence.

Important examples

Dirichlet L-functions

For a Dirichlet character χ\chi, complete multiplicativity gives

L(s,χ)=∑n=1∞χ(n)ns=∏p(1−χ(p)p−s)−1,Re⁡(s)>1.L(s,\chi) =\sum_{n=1}^{\infty}\frac{\chi(n)}{n^s} =\prod_p(1-\chi(p)p^{-s})^{-1}, \qquad \operatorname{Re}(s)>1.

These Dirichlet L-functions encode arithmetic information about residue classes. If pp divides the modulus of χ\chi, then χ(p)=0\chi(p)=0, so the corresponding factor is 11.

The Möbius function

The Möbius function satisfies

μ(p)=−1,μ(pk)=0(k≥2).\mu(p)=-1,\qquad \mu(p^k)=0\quad(k\geq2).

Its local factor is therefore 1−p−s1-p^{-s}, giving

∑n=1∞μ(n)ns=∏p(1−p−s)=1ζ(s),Re⁡(s)>1.\sum_{n=1}^{\infty}\frac{\mu(n)}{n^s} =\prod_p(1-p^{-s}) =\frac{1}{\zeta(s)}, \qquad \operatorname{Re}(s)>1.

Divisor-counting functions

Let d(n)d(n) be the number of positive divisors of nn. Since d(pk)=k+1d(p^k)=k+1,

∑k=0∞(k+1)p−ks=(1−p−s)−2.\sum_{k=0}^{\infty}(k+1)p^{-ks} =(1-p^{-s})^{-2}.

Hence

∑n=1∞d(n)ns=∏p(1−p−s)−2=ζ(s)2,Re⁡(s)>1.\sum_{n=1}^{\infty}\frac{d(n)}{n^s} =\prod_p(1-p^{-s})^{-2} =\zeta(s)^2, \qquad \operatorname{Re}(s)>1.

This illustrates how local factors encode counting information beyond complete multiplicativity.

Logarithms, prime powers, and zeros

Taking the logarithm of the zeta Euler product produces an absolutely convergent expansion:

log⁡ζ(s)=∑p∑k=1∞p−ksk,Re⁡(s)>1.\log\zeta(s) =\sum_p\sum_{k=1}^{\infty}\frac{p^{-ks}}{k}, \qquad \operatorname{Re}(s)>1.

Here the expansion specifies the analytic branch of the logarithm that agrees with the real logarithm for real s>1s>1.

Differentiating gives

−ζ′(s)ζ(s)=∑p∑k=1∞(log⁡p)p−ks=∑n=1∞Λ(n)ns,-\frac{\zeta'(s)}{\zeta(s)} =\sum_p\sum_{k=1}^{\infty}(\log p)p^{-ks} =\sum_{n=1}^{\infty}\frac{\Lambda(n)}{n^s},

where the von Mangoldt function is

Λ(n)={log⁡p,n=pk,p prime, k≥1,0,otherwise.\Lambda(n)= \begin{cases} \log p,&n=p^k,\quad p\text{ prime},\ k\geq1,\\ 0,&\text{otherwise}. \end{cases}

Thus the logarithmic derivative converts the multiplicative structure of the Euler product into a series supported on prime powers.

The convergent product also shows that

ζ(s)≠0for Re⁡(s)>1.\zeta(s)\neq0 \qquad\text{for }\operatorname{Re}(s)>1.

This is an initial zero-free region, not a proof of the Riemann hypothesis. Establishing nonvanishing on the line Re⁡(s)=1\operatorname{Re}(s)=1 requires additional arguments and is a central ingredient in classical analytic proofs of the prime number theorem.

Historical and broader significance

Euler developed the prime-product identity in work presented in 1737. It provided a bridge between series over all positive integers and products over primes. The divergence of the zeta series as the real variable approaches 11 from above also yields a proof that there are infinitely many primes: finitely many prime factors would give a finite limiting product.

In the nineteenth century, Dirichlet used character-weighted series in studying primes in arithmetic progressions, while Bernhard Riemann placed the zeta function within complex analysis. These developments made Euler products a structural tool in analytic number theory.

The idea extends beyond ordinary integers. The Dedekind zeta function of a number field has a product over nonzero prime ideals, with each factor determined by the size of the associated residue field. L-functions associated with elliptic curves and other arithmetic objects likewise assemble global functions from local factors. Their factors need not be geometric; their form reflects the arithmetic being encoded.

The term “Euler product” is most commonly used for these prime-indexed factorizations. Products associated with Euler in other settings, such as product formulas involving the gamma function or trigonometric functions, should not automatically be identified with the number-theoretic construction.