Subfields of Cyclotomic Fields

The act of adjoining a root of unity, $\zeta_n$, to the rational numbers $\mathbb{Q}$ creates vast number systems known as cyclotomic fields. While seemingly simple in their construction, these fields possess a profoundly intricate and ordered internal structure. The central question this article addresses is: how can we systematically explore and map the universe of smaller fields—the subfields—that lie hidden within a given cyclotomic field? Answering this question reveals not just algebraic curiosities, but fundamental principles that connect various branches of mathematics.

This article provides a comprehensive guide to the subfields of cyclotomic fields. In the first chapter, "Principles and Mechanisms," we will introduce the elegant framework of Galois theory, which serves as our primary tool for discovering and classifying subfields. We will explore how this "map of symmetries" allows us to pinpoint specific structures, such as quadratic fields and real subfields, and culminates in the magnificent Kronecker-Weber theorem. Following this, the "Applications and Interdisciplinary Connections" chapter will demonstrate the power of these concepts, showing how cyclotomic subfields solve problems in number theory, settle ancient geometric puzzles, and reveal surprising connections in analysis and dynamics.

Imagine you've discovered a beautiful, intricate crystal. At first, you admire its overall shape. But soon, you notice it's not just a single, uniform block. It has internal planes of symmetry, smaller repeating patterns, and substructures within it. How do you map this inner world? You could shine a light through it and see how the light is altered, or you could study its symmetries—the rotations and reflections that leave its appearance unchanged.

The world of cyclotomic fields, like $\mathbb{Q}(\zeta_n)$, is much like that crystal. Adjoining a simple root of unity, $\zeta_n = \exp(2\pi i / n)$, to the familiar rational numbers $\mathbb{Q}$ creates a vast and beautifully structured universe of new numbers. But its true beauty lies not in its size, but in its profound internal order. The key to unlocking this inner world is a powerful idea from the brilliant mind of Évariste Galois: the ​​Galois correspondence​​.

The Galois correspondence is one of the most elegant and powerful theorems in mathematics. It provides a perfect dictionary, a Rosetta Stone, for translating between the language of fields and the language of group theory. For a cyclotomic field like $K = \mathbb{Q}(\zeta_n)$, its "group of symmetries" is the set of all automorphisms that shuffle the numbers in $K$ around while leaving the base field $\mathbb{Q}$ completely untouched. This is the ​​Galois group​​, denoted $\text{Gal}(K/\mathbb{Q})$. For cyclotomic fields, this group is always isomorphic to $(\mathbb{Z}/n\mathbb{Z})^\times$, the group of integers modulo $n$ that have a multiplicative inverse—a group that is, crucially, always ​​abelian​​ (commutative).

The correspondence states that there is a perfect, one-to-one, order-reversing relationship between the ​​subfields​​ of $K$ and the ​​subgroups​​ of its Galois group, $G$.

• To every subfield $E$ (where $\mathbb{Q} \subseteq E \subseteq K$), there corresponds a unique subgroup $H$ of $G$. This subgroup consists of all the symmetries in $G$ that leave every single element of $E$ fixed. We call $E$ the ​​fixed field​​ of $H$.
• Conversely, to every subgroup $H$ of $G$, there corresponds a unique fixed field $E$.

The "order-reversing" part is wonderfully counter-intuitive at first, but perfectly logical upon reflection. A smaller subfield is a simpler structure, meaning it's easier to leave it unchanged. Therefore, it will have a larger group of symmetries that fix it. A tiny subfield like $\mathbb{Q}$ itself is fixed by every symmetry in $G$. The entire field $K$, being the most complex, is only fixed by the identity symmetry (the one that does nothing). This beautiful duality is our map. To find hidden subfields, we need only search for subgroups of symmetries.

Let's use our map to go treasure hunting. The simplest, most interesting treasures to look for are the ​​quadratic fields​​—subfields of the form $\mathbb{Q}(\sqrt{d})$ for some square-free integer $d$. A quadratic field is an extension of degree 2 over $\mathbb{Q}$. In our Galois dictionary, this corresponds to finding a subgroup $H$ whose index, $[G:H] = |G|/|H|$, is 2. This means we're looking for subgroups of half the size of the full Galois group.

Let's start with the 5th cyclotomic field, $K = \mathbb{Q}(\zeta_5)$. Its Galois group is $G \cong (\mathbb{Z}/5\mathbb{Z})^\times = \{1,2,3,4\}$, a cyclic group of order 4. A cyclic group of order 4 has exactly one subgroup of order 2. In this case, it's the subgroup $H = \{1, 4\}$, corresponding to the automorphisms $\sigma_1(\zeta_5) = \zeta_5$ (the identity) and $\sigma_4(\zeta_5) = \zeta_5^4 = \zeta_5^{-1}$ (complex conjugation).

To find the field fixed by this subgroup, we need to find a number that is unchanged by complex conjugation. A natural candidate is the sum of an element and its conjugate: $\alpha = \zeta_5 + \zeta_5^{-1}$. A little algebraic manipulation, using the fact that $1 + \zeta_5 + \zeta_5^2 + \zeta_5^3 + \zeta_5^4 = 0$, reveals that $\alpha$ satisfies the equation $\alpha^2 + \alpha - 1 = 0$. The solutions are $\frac{-1 \pm \sqrt{5}}{2}$. Voila! The field generated by $\alpha$ is none other than $\mathbb{Q}(\sqrt{5})$. We have found our first jewel: the unique quadratic subfield of $\mathbb{Q}(\zeta_5)$ is $\mathbb{Q}(\sqrt{5})$.

This is not a coincidence. This phenomenon is perfectly general. For any odd prime $p$, the cyclotomic field $\mathbb{Q}(\zeta_p)$ contains exactly one quadratic subfield. This subfield can be found by constructing a remarkable object called a ​​Gauss sum​​, which is ingeniously designed to transform in a simple way under the action of the Galois group. The astonishing result is that the unique quadratic subfield of $\mathbb{Q}(\zeta_p)$ is $\mathbb{Q}(\sqrt{d})$ where $d = (-1)^{(p-1)/2} p$. This beautiful formula links the prime $p$ to the sign of its square root subfield, depending on whether $p \equiv 1 \pmod 4$ or $p \equiv 3 \pmod 4$. For $p=5$, we get $d = (-1)^{(5-1)/2} 5 = 5$, just as we found! For the field $\mathbb{Q}(\zeta_3)$, we get $d = (-1)^{(3-1)/2} 3 = -3$, so the quadratic subfield is $\mathbb{Q}(\sqrt{-3})$.

The world within a cyclotomic field is far richer than just quadratic fields. Our Galois map can guide us to all sorts of interesting territories.

The Real World Inside

One of the most natural subfields is the ​​maximal real subfield​​, denoted $\mathbb{Q}(\zeta_n)^+$. This is the set of all numbers within $\mathbb{Q}(\zeta_n)$ that are real. It is the fixed field of the subgroup containing just the identity and the complex conjugation automorphism (which maps $\zeta_n \mapsto \zeta_n^{-1}$). A generator for this field is the number $\zeta_n + \zeta_n^{-1} = 2\cos(2\pi/n)$. These real subfields, like $\mathbb{Q}(\cos(2\pi/7))$, are themselves well-behaved "Galois" extensions of $\mathbb{Q}$. They represent the part of the cyclotomic crystal that lies perfectly on the real number line.

Fields Within Fields

Sometimes, a subfield is itself a cyclotomic field. Consider the sprawling field $\mathbb{Q}(\zeta_{45})$. Its Galois group is isomorphic to $(\mathbb{Z}/45\mathbb{Z})^\times$. Using the Chinese Remainder Theorem, we can decompose this group of symmetries: $(\mathbb{Z}/45\mathbb{Z})^\times \cong (\mathbb{Z}/9\mathbb{Z})^\times \times (\mathbb{Z}/5\mathbb{Z})^\times$. What if we look for the subfield fixed only by symmetries that correspond to the identity in the second component? That is, the subgroup $H$ of automorphisms $\sigma_a$ where $a \equiv 1 \pmod 5$. Galois theory tells us this subfield must exist. When we perform the calculation, we find this fixed field is precisely $\mathbb{Q}(\zeta_5)$. The larger, more complex structure contains a smaller, simpler cyclotomic structure, perfectly revealed by analyzing the decomposition of its symmetries.

This also works in reverse. If we know a subfield $E$ is inside a larger field $L$, then the Galois group of the subfield, $\text{Gal}(E/\mathbb{Q})$, can be seen as a "quotient" or a simplified image of the larger group $\text{Gal}(L/\mathbb{Q})$. For instance, since $\zeta_5 = \zeta_{15}^3$, the field $E = \mathbb{Q}(\zeta_5)$ is a subfield of $L = \mathbb{Q}(\zeta_{15})$. The Galois group of $E$, which has order $\phi(5)=4$, is a quotient of the Galois group of $L$, which has order $\phi(15)=8$.

We have journeyed inside cyclotomic fields, exploring their internal geography. Now, we zoom out to see their place in the entire cosmos of numbers. The result is one of the crowning achievements of 19th-century number theory: the ​​Kronecker-Weber Theorem​​.

The theorem makes a claim that is as breathtaking as it is profound:

Every finite abelian extension of the rational numbers $\mathbb{Q}$ is a subfield of some cyclotomic field $\mathbb{Q}(\zeta_n)$.

Let's pause to appreciate this. An "abelian extension" is a Galois extension whose Galois group is abelian (commutative). The theorem says that any such number field, no matter how it is constructed, can be found living inside one of the cyclotomic fields we've been exploring. The roots of unity, $\zeta_n$, are the fundamental building blocks for the entire universe of abelian extensions of $\mathbb{Q}$. The union of all cyclotomic fields, $\bigcup_{n \ge 1} \mathbb{Q}(\zeta_n)$, forms a colossal field known as the ​​maximal abelian extension of $\mathbb{Q}$​​, denoted $\mathbb{Q}^{ab}$.

It is crucial to understand the nuances of this powerful statement. First, the theorem guarantees inclusion, not equality. As we saw, $\mathbb{Q}(\sqrt{5})$ is an abelian extension, and it lives inside $\mathbb{Q}(\zeta_5)$, but it is not equal to it. In fact, most abelian extensions are not cyclotomic fields themselves, but rather subfields of them.

Second, the "abelian" condition is absolutely essential. What about extensions whose Galois groups are non-abelian, like the violent, non-commutative symmetries of the symmetric group $S_3$? Could they also be found inside a cyclotomic field? The answer is a definitive ​​no​​. Our Galois correspondence provides a beautiful, simple proof. The Galois group of any cyclotomic field is abelian. If a non-abelian extension $K/\mathbb{Q}$ were a subfield of $\mathbb{Q}(\zeta_n)$, its Galois group, $\text{Gal}(K/\mathbb{Q})$, would have to be a quotient of the abelian group $\text{Gal}(\mathbb{Q}(\zeta_n)/\mathbb{Q})$. But a quotient of an abelian group is always abelian! This is a contradiction. Therefore, non-abelian Galois extensions, such as the splitting fields of the polynomials $x^3 - 2$ or $x^4 - 2$, cannot be contained in any cyclotomic field.

The Kronecker-Weber theorem thus draws a magnificent line in the sand. The orderly, commutative world of abelian extensions is entirely described by the arithmetic of roots of unity. The wild, non-commutative world lies beyond. It tells us that the simple act of dividing a circle into $n$ equal parts, an idea known since ancient Greece, holds the secret to all commutative field symmetries, unifying geometry, algebra, and number theory in a single, beautiful tapestry.

Having journeyed through the elegant principles and mechanisms that govern cyclotomic fields and their subfields, we might feel like we've been admiring the intricate gears and springs of a masterfully crafted watch. Now, it's time to see what this beautiful machine can do. We are about to witness how these abstract algebraic structures are not idle curiosities but powerful tools that unlock profound secrets across the mathematical landscape, from the ancient puzzles of geometry to the deepest questions of modern number theory. The story of their applications is a story of unexpected unity, revealing a hidden symphony that connects disparate-looking ideas.

Perhaps the most astonishing discovery is the ​​Kronecker-Weber theorem​​. In essence, it tells us that if you want to build any "well-behaved" number system as an extension of the rational numbers $\mathbb{Q}$—specifically, any finite abelian extension—you don't need to search far and wide. Every single one of them can be found living inside a cyclotomic field $\mathbb{Q}(\zeta_n)$. This theorem isn't just a statement of existence; it's a map to an entire universe of number fields. The subfields of cyclotomic fields are the complete set of building blocks for the abelian world over $\mathbb{Q}$.

This grand principle has immediate, practical consequences. Imagine you're an algebraic engineer and you have a blueprint for a number field: you need its Galois group over $\mathbb{Q}$ to be a specific abelian group, say the product of a cyclic group of order 2 and one of order 4, $C_2 \times C_4$. How do you build it? The Kronecker-Weber theorem tells you to go shopping in the cyclotomic aisle! The Galois group of $\mathbb{Q}(\zeta_n)$ over $\mathbb{Q}$ is famously isomorphic to the group of invertible integers modulo $n$, $(\mathbb{Z}/n\mathbb{Z})^\times$. Our task is thus transformed into a puzzle in elementary number theory: find an integer $n$ such that the group $(\mathbb{Z}/n\mathbb{Z})^\times$ is isomorphic to $C_2 \times C_4$. A systematic search reveals several candidates, including $n=15$, $n=16$, and $n=20$. By taking the field $\mathbb{Q}(\zeta_{16})$, for example, we can construct a world of numbers with precisely the Galois structure we desired. The abstract inverse Galois problem, for abelian groups, is thereby solved constructively. We can not only prove such fields exist, we can point to them. This same logic allows us to realize any finite abelian group as a Galois group over $\mathbb{Q}$ by finding the right cyclotomic field (or one of its subfields) that hosts it.

The connection flows both ways. Given an abelian extension, like the biquadratic field $K = \mathbb{Q}(\sqrt{17}, \sqrt{-42})$, we know it must live inside some $\mathbb{Q}(\zeta_n)$. But which one? Can we find its "minimal cyclotomic address"? This address is a specific number, the ​​conductor​​ of the field. And beautifully, there's a concrete algorithm to find it. The conductor of a composite field is the least common multiple of the conductors of its constituent subfields. For a quadratic field $\mathbb{Q}(\sqrt{d})$, the conductor is simply the absolute value of its fundamental discriminant—a number which depends delicately on whether $d$ is congruent to $1$ modulo $4$. By applying these simple rules, we can compute that the smallest integer $n$ for which $\mathbb{Q}(\sqrt{17}, \sqrt{-42})$ is a subfield of $\mathbb{Q}(\zeta_n)$ is precisely $n=2856$. An abstract question of field containment is reduced to a deterministic calculation, showcasing the powerful predictive nature of the theory.

This "conductor" is more than just an address; it's a magic number that governs the arithmetic within the field. This leads us to one of the crown jewels of number theory: ​​reciprocity laws​​. A fundamental question is how a prime number $p$ from $\mathbb{Q}$ behaves in an extension field $K$. Does it remain prime, or does it factor into a product of primes in $K$? For abelian extensions, the answer is breathtakingly simple and is governed by the ​​Artin symbol​​. For a prime $p$ that doesn't divide the conductor $n$, its factorization pattern is completely determined by the residue class of $p$ modulo $n$. For instance, the field $\mathbb{Q}(\sqrt{2})$ can be viewed as the maximal real subfield of $\mathbb{Q}(\zeta_8)$. The conductor is $8$. The behavior of a prime $p$ in $\mathbb{Q}(\sqrt{2})$ depends on whether $p \equiv 1, 3, 5,$ or $7 \pmod{8}$. The Artin symbol provides a concrete map from this modular arithmetic to the Galois group of the extension, turning a deep question about prime factorization into a simple calculation. This is the heart of class field theory: the intricate dance of primes is choreographed by the simple rhythm of modular arithmetic.

The story does not end with number theory. A crucial observation is that the maximal real subfield of $\mathbb{Q}(\zeta_n)$, denoted $\mathbb{Q}(\zeta_n)^+$, is generated by the element $\zeta_n + \zeta_n^{-1} = 2\cos(2\pi/n)$. This simple trigonometric identity forms a bridge between the discrete, algebraic world of Galois theory and the continuous, analytic world of geometry and functions.

This connection shines a bright light on the famed impossibility of trisecting an angle with a compass and straightedge. Why is it impossible to trisect a $60^\circ$ angle? The problem boils down to constructing the number $\cos(20^\circ)$. Using the triple-angle identity, $4\cos^3\theta - 3\cos\theta = \cos(3\theta)$, one finds that $\cos(20^\circ)$ is a root of the irreducible cubic polynomial $8x^3 - 6x - 1 = 0$. Compass and straightedge constructions can only produce numbers that live in field extensions of degree $2^k$ over $\mathbb{Q}$. Since $\cos(20^\circ)$ generates an extension of degree 3, it cannot be constructed. But our theory tells us more. We can positively identify the field where this construction is possible: it is precisely the maximal real subfield of $\mathbb{Q}(\zeta_{18})$. The "impossible" number $\cos(20^\circ)$ has a home, and its home is a real cyclotomic subfield.

The surprises continue. Consider the ​​Chebyshev polynomials​​, $T_n(x)$, which are ubiquitous in approximation theory and the solution of differential equations. They are defined by the innocent-looking property $T_n(\cos\theta) = \cos(n\theta)$. What is the algebraic structure of these polynomials? Their roots are of the form $\cos(\frac{(2k-1)\pi}{2n})$. It turns out that the splitting field generated by all these roots—the smallest field containing them all—is none other than the maximal real subfield of $\mathbb{Q}(\zeta_{4n})$. This reveals a hidden, rigid algebraic skeleton beneath the analytic flesh of these special functions.

An even more startling connection appears in the study of dynamical systems. Consider the simple quadratic map $x \mapsto x^2 - 2$. If we take a number $\alpha$ and repeatedly apply this map, we generate a sequence $\alpha, \alpha^2-2, (\alpha^2-2)^2-2, \dots$. Now, suppose $\alpha$ is the root of an irreducible polynomial over $\mathbb{Q}$, and we impose the strange condition that this entire sequence of iterates must also be roots of the same polynomial. Since there are only finitely many roots, the sequence must eventually repeat. What does this dramatic restriction imply about $\alpha$? The substitution $\alpha = u + u^{-1}$ works wonders here. The map $\alpha \mapsto \alpha^2-2$ transforms into the much simpler map $u+u^{-1} \mapsto u^2+u^{-2}$. For the orbit to be finite, $u$ must be a root of unity, say $\zeta$. This forces every such $\alpha$ to be of the form $\zeta + \zeta^{-1}$! The roots of any such polynomial are real numbers from cyclotomic fields. A condition from dynamics forces our solution into the world of real cyclotomic subfields, unifying algebra, number theory, and dynamics in a single, elegant argument.

Beyond the structure of the fields themselves, cyclotomic subfields serve as the perfect laboratory for studying their deeper arithmetic: the properties of their "integers," their units, and the failure of unique factorization measured by the class number.

For a number field $K = \mathbb{Q}(\alpha)$, it's a fundamental and often difficult question whether the simple ring $\mathbb{Z}[\alpha]$ constitutes all the algebraic integers in $K$. The answer is encoded in the field's ​​discriminant​​. In a remarkable display of unity, we can compute this discriminant in two different ways. For the field $K = \mathbb{Q}(2\cos(2\pi/7))$, which is the maximal real subfield of $\mathbb{Q}(\zeta_7)$, we can compute the discriminant from its minimal polynomial, $x^3+x^2-2x-1=0$, and find it to be $49$. On the other hand, class field theory gives us the powerful ​​conductor-discriminant formula​​, which relates the discriminant to the conductor of the field. Since we identified the conductor as $7$, the formula gives the discriminant as $7^{3-1} = 49$. The two calculations match perfectly, which proves that the index $[\mathcal{O}_K : \mathbb{Z}[2\cos(2\pi/7)]]$ must be $1$. The abstract machinery of class field theory provides a tool to answer a very concrete question about the integral basis of the field.

Similarly, ​​Dirichlet's Unit Theorem​​ describes the structure of the group of invertible integers in a number field. For the real cyclotomic subfield $K = \mathbb{Q}(\zeta_p)^+$, we can easily compute the rank of its unit group. More profoundly, we can explicitly construct a special family of units called ​​cyclotomic units​​, built from expressions like $1-\zeta_p^a$. These units are, in a sense, the "obvious" ones. The truly amazing fact, first glimpsed by Kummer, is that these obvious units are almost all there are. The index of the group of cyclotomic units inside the full group of units is finite, and this index is miraculously equal to the ​​class number​​ of the field, $h_K$. For $K = \mathbb{Q}(\zeta_7)^+$, this class number is $1$, meaning the cyclotomic units generate the entire unit group (up to roots of unity). This connects the multiplicative structure of units to the ideal structure of the ring of integers, a deep and fruitful relationship at the heart of modern number theory.

The story of cyclotomic fields and their subfields over $\mathbb{Q}$ is one of remarkable completeness. They provide the raw material for all abelian extensions. This beautiful picture, however, raises a tantalizing question: what happens if we change our base field? What if we start not with $\mathbb{Q}$, but with, say, an imaginary quadratic field like $K = \mathbb{Q}(\sqrt{-5})$? This is the essence of ​​Hilbert's 12th Problem​​, which sought to find the functions that would play the role for other fields that the exponential function $z \mapsto \exp(2\pi i z)$ plays for $\mathbb{Q}$.

It turns out that simply adjoining roots of unity to $K$ is not enough. The world of abelian extensions over $K$ is far richer. The Hilbert class field of $\mathbb{Q}(\sqrt{-d})$, for instance, which is the maximal unramified abelian extension, is not generated by roots of unity. Instead, it is generated by special values of modular functions, a theory known as ​​Complex Multiplication​​. These extensions, while abelian over $K$, can be non-abelian over $\mathbb{Q}$, and thus cannot possibly be subfields of any cyclotomic field.

This shows that the theory we have explored is both complete and special. It is the first, simplest, and most elegant chapter in the grander saga of class field theory. The subfields of cyclotomic fields are not just a collection of examples; they are the Rosetta Stone that allowed us to first decipher the laws governing abelian extensions, providing a blueprint and an inspiration for exploring the vast, uncharted territories of the mathematical universe.