1 Primary ideals and their defining property

1.1 Definition and equivalent formulations

Let \(R\) be a commutative ring with \(1\). An ideal \(Q\subseteq R\) is primary if \(Q\neq R\) and the following holds: whenever \(ab\in Q\) for \(a,b\in R\), then either \(a\in Q\) or \(b^n\in Q\) for some integer \(n\ge 1\).

Several equivalent forms are commonly used. One convenient variant uses the contrapositive: if \(a\notin Q\), then for every \(b\in R\), the condition \(ab\in Q\) forces \(b^n\in Q\) for some \(n\). Another reformulation is module-theoretic and states that \(Q\) is primary precisely when \(R/Q\) has a “one-sided” zero-divisor behavior: zero divisors in \(R/Q\) are nilpotent, in a sense compatible with the ideal-theoretic condition above.

1.2 The radical of a primary ideal

If \(Q\) is primary, its radical \(\sqrt{Q}=\{r\in R: r^m\in Q \text{ for some } m\ge 1\}\) is a prime ideal. Intuitively, the defining property ensures that “factors that fail to land in \(Q\)” can only do so up to nilpotent error, and that error collapses to a prime condition after passing to radicals.

As a consequence, the radical \(\sqrt{Q}\) is often viewed as the “prime shadow” of \(Q\). The ideal \(Q\) itself can be strictly larger than \(\sqrt{Q}\) when it contains nilpotent elements mod its radical.

1.3 Examples illustrating “primary but not prime”

A standard source of primary ideals comes from powers of prime ideals. Let \(P\) be a prime ideal. Then \(P^k\) is typically primary for \(k\ge 1\) (and is prime only when \(k=1\) in many familiar settings, especially in reduced rings).

A simple illustration: in \(k[x,y]\), consider \(Q=(x^2,xy)\). One checks that \(Q\) is primary with radical \(\sqrt{Q}=(x)\). However, \(Q\) is not prime: \(x\notin Q\) while \(x\cdot y\in Q\), and the prime test fails. The primary condition is rescued by the fact that \(y^n\in Q\) for suitable \(n\) once the product lands in \(Q\).

1.4 Comparison with prime ideals

Prime ideals satisfy a stronger requirement: if \(ab\in P\), then \(a\in P\) or \(b\in P\), with no need for powers. Every prime ideal is primary, because the “raise to a power” clause becomes unnecessary.

The distinction can be summarized as follows: primary ideals permit a controlled failure of primality, where one factor may not belong to the ideal itself but must become nilpotent modulo the ideal. This makes primary ideals a natural refinement of prime ideals that records both geometric (prime/radical) and infinitesimal (nilpotent thickness) information.

Primary ideals are tied to prime ideals through radicals and through the behavior of annihilators in modules such as \(R/Q\). This gives rise to the notion of associated primes, which act as bookkeeping devices for the prime components that occur inside a primary decomposition.

2.1.1 The associated prime of a primary ideal

For a primary ideal \(Q\), the associated prime corresponding to \(Q\) is \(\sqrt{Q}\). In the language of modules, \(\sqrt{Q}\) is the unique prime ideal associated to \(R/Q\). Equivalently, \(\sqrt{Q}\) is the prime that measures where the “main” zero-divisor behavior concentrates.

2.1.1.1 Properties of \(\sqrt{Q}\) when \(Q\) is primary

When \(Q\) is primary:

  • \(\sqrt{Q}\) is prime.
  • \(Q\) is \(\sqrt{Q}\)-primary, meaning that every element outside \(Q\) behaves like a non–zero divisor away from \(\sqrt{Q}\).
  • The nilpotent part of the quotient \(R/Q\) is governed by the difference between \(Q\) and \(\sqrt{Q}\).

These properties make \(\sqrt{Q}\) the central prime invariant attached to \(Q\).

2.2 Primary components versus associated primes

Given a decomposition of an ideal \(I\) as an intersection of primary ideals, \[ I=\bigcap_{i} Q_i, \] each \(Q_i\) contributes its radical \(\sqrt{Q_i}\), and these radicals are primes that typically appear among the associated primes of \(R/I\). In favorable finiteness contexts (notably Noetherian rings), one can relate the set of associated primes of \(R/I\) to the radicals of the primary components in a primary decomposition.

However, not every primary component corresponds to a distinct associated prime: different components can share the same radical, and irredundancy conditions clarify which components genuinely contribute new information.

2.3 Uniqueness aspects in primary decompositions

Primary decompositions are rarely unique term-by-term, but certain features are rigid. Under standard hypotheses (e.g., Noetherian rings), the set of associated primes of \(R/I\) is uniquely determined by \(I\), and the radicals \(\sqrt{Q_i}\) appearing in a minimal decomposition are uniquely determined up to order.

This partial uniqueness explains why associated primes are so central: they capture the “prime-level” invariants that persist across different decompositions.

3 Primary decomposition of ideals

3.1 Motivation and statement of the decomposition idea

The goal of primary decomposition is to break an arbitrary ideal into simpler pieces whose structure is controlled by one prime at a time. Prime decomposition splits an ideal according to prime ideals; primary decomposition refines this by allowing nilpotent thickening around each prime.

In geometric terms (over suitable classes of rings), radicals correspond to underlying varieties, while primary components record additional multiplicity and embedded structure.

3.2 Decompositions into primary ideals

A primary decomposition expresses an ideal \(I\) as an intersection of primary ideals: \[ I=\bigcap_{i=1}^t Q_i \] with each \(Q_i\) primary. The components are often grouped by their radicals \(\sqrt{Q_i}\), which correspond to primes associated with \(R/I\).

Decompositions can also appear in dual forms using sums/products depending on conventions, but the intersection form is the standard for recording “simultaneous containment” of functions in multiple primary pieces.

3.3 Existence conditions (Noetherian context)

Existence of primary decompositions is guaranteed in many practical settings. A canonical one is:

  • If \(R\) is Noetherian and \(I\subseteq R\) is an ideal, then \(I\) admits a primary decomposition into finitely many primary ideals.

The Noetherian condition ensures that chains of ideals stabilize, preventing infinite refinement and enabling inductive arguments on dimension or length.

3.4 Uniqueness and irredundancy

A decomposition \[ I=\bigcap_{i=1}^t Q_i \] is called irredundant if no \(Q_i\) can be omitted while keeping the intersection unchanged. Under Noetherian hypotheses, irredundant decompositions have strong uniqueness properties:

  • The set of primes \(\sqrt{Q_i}\) appearing is uniquely determined.
  • Components whose radicals are distinct correspond uniquely (up to order) to the associated prime structure.
  • Embedded components (those associated primes that are not minimal over \(I\)) require care, but their existence is still constrained by the associated primes of \(R/I\).

These principles make primary decomposition a robust structural tool rather than merely a convenient decomposition procedure.

4 Operations and behavior under ideal-theoretic constructions

4.1 Intersections of primary ideals

Intersections of primary ideals need not be primary, but they are central to the existence theory of primary decomposition. In general, an intersection can preserve enough nilpotent structure to remain tightly controlled by the radicals of the involved primary ideals.

Two useful practical points are:

  • The radical of an intersection satisfies \(\sqrt{\bigcap_i Q_i}=\bigcap_i \sqrt{Q_i}\).
  • Associated primes of the intersection can often be recovered from those of the components, especially in Noetherian rings.

4.2 Sums, products, and containments

For primary ideals \(Q_1,Q_2\), operations like sums and products typically change primariness and can create more complicated decompositions. Still, containments among ideals can be analyzed using radicals:

  • \(\sqrt{Q_1Q_2}=\sqrt{Q_1\cap Q_2}\) often holds in standard situations, and the relationship helps track radicals.
  • If radicals are comaximal, intersections and products relate more simply, mirroring Chinese-remainder-type behavior.

In computations, one often reduces questions about sums/products to statements about radicals and then uses primary components to recover finer detail.

4.3 Localization and contraction of primary ideals

Localization is a key construction in commutative algebra. If \(S\subseteq R\) is multiplicative, one forms \(R_S\) and extends ideals by \[ S^{-1}Q=\{q/s : q\in Q,\, s\in S\}. \] Localization interacts with primariness in a predictable way:

  • If \(Q\) is \(P\)-primary with \(P=\sqrt{Q}\), then the localized ideal \(S^{-1}Q\) is either the whole ring (when \(P\cap S\neq\varnothing\)) or a primary ideal in \(R_S\) whose radical is \(S^{-1}P\).

Contraction of an ideal from \(R_S\) back to \(R\) can also reflect primariness, though it may not preserve it in a one-to-one fashion without additional conditions.

4.4 Quotients and their induced primary structures

Passing to a quotient \(R/I\) transforms ideals via the correspondence between ideals of \(R\) containing \(I\) and ideals of \(R/I\). If \(Q\supseteq I\) is primary in \(R\), then the image \(Q/I\) behaves like a primary ideal in \(R/I\) under suitable hypotheses; its radical corresponds to the image of \(\sqrt{Q}\).

In particular, primary decomposition of \(I\) can be transported to information about the structure of \(R/I\), including associated primes, embedded components, and nilpotent behavior.

5 Characterizations and algebraic tests

5.1 Zero divisors and torsion viewpoint

The defining condition of a primary ideal has a natural interpretation in terms of zero divisors in the quotient \(R/Q\). If \(Q\) is primary with radical \(P\), then elements of \(R/Q\) that act as zero divisors are precisely those whose annihilation forces nilpotence governed by \(P\).

Equivalently, torsion phenomena in modules over \(R/Q\) concentrate at the prime \(P\). This viewpoint connects primary ideals to homological and module-theoretic invariants.

5.2 Module-theoretic reformulation

The ideal-theoretic definition can be reframed using annihilators. In module language, \(Q\) is primary if and only if \(R/Q\) is an example of a module whose “associated prime” is unique. More precisely, one can characterize primariness by the condition that \(\operatorname{Ann}(m)\) has a controlled form for nonzero elements \(m\in R/Q\), and that the radical of \(Q\) governs these annihilators.

This reformulation is often the gateway to generalizations in the theory of primary submodules.

5.3 Colon ideal criteria for primariness

Colon ideals provide practical tests. For an element \(a\in R\), the colon ideal \((Q:a)=\{r\in R: ra\in Q\}\) measures how multiplication by \(a\) pushes elements into \(Q\). In many treatments, a criterion characterizes primariness via stabilization properties of \((Q:a)\) when \(a\notin Q\), reflecting the “some power lands in \(Q\)” condition from the definition.

Such tests are valuable in computations and in proofs, since colon ideals behave well under localization and exact sequences.

5.4 Stabilization of powers and annihilators

A further characterization uses the behavior of repeated multiplication. If \(Q\) is primary and \(a\notin Q\), then the set of \(b\) for which \(ab\in Q\) collapses into those \(b\) that eventually enter \(Q\) after enough powers. This can be expressed by the stabilization of sequences of annihilators or colon ideals such as \[ (Q:a)\subseteq (Q:a^2)\subseteq (Q:a^3)\subseteq \cdots, \] which become constant in Noetherian settings. The “eventual” nature of inclusion mirrors the power condition in the definition of primary.

6 Special settings and standard examples

6.1 Primary ideals in principal ideal domains

In a principal ideal domain \(R\), ideals are generated by a single element. Primary ideals correspond closely to powers of primes: if \((p)\) is a prime ideal, then \((p^k)\) is primary for \(k\ge 1\), with radical \((p)\). Since PIDs have simple ideal structure, the classification is straightforward and illustrates the general theme: a primary ideal is a prime ideal thickened by nilpotent multiplicity.

6.2 Primary ideals in polynomial rings

Polynomial rings \(k[x_1,\dots,x_n]\) over a field provide a rich supply of primary ideals. Many appear as powers of ideals generated by linear or irreducible polynomials, or as monomial ideals with support concentrated on one set of variables.

In geometric interpretations, radicals correspond to prime varieties, while primary ideals encode how functions vanish to higher order along those varieties.

6.3 Monomial ideals and computational-friendly descriptions

Monomial ideals in polynomial rings admit combinatorial descriptions: membership, radicals, and sometimes primary decomposition can be approached via exponent sets and faces of associated monoids.

For monomial ideals, one can often detect primariness by analyzing the behavior of exponent vectors under addition and the structure of supports. This makes monomial primary decomposition a practical computational target in Gröbner basis and related algorithmic frameworks.

6.4 Simple worked examples of decompositions

A typical elementary example is the decomposition of an ideal like \((x^2y,\, xy^2)\) in a polynomial ring \(k[x,y]\). One can express it as an intersection of primary ideals corresponding to the primes \((x)\) and \((y)\), with each component capturing a different axis behavior of vanishing.

Another standard pattern is that ideals involving products of relatively prime elements factor into intersections of primary components supported at each prime divisor. While the exact algebra depends on the ring and generators, these examples illustrate how radicals track which “directions” (primes) the ideal supports, while the powers capture multiplicity.

7 Applications and conceptual roles

7.1 Relation to irreducible decompositions

Primary ideals sit close to irreducible ideals in the sense of ideal decomposition: an irreducible ideal is one that cannot be written as a proper intersection of two larger ideals. Primary ideals often refine irreducible decompositions, and in many Noetherian contexts, primary components can be chosen to be irreducible when the associated prime satisfies suitable minimality conditions.

Thus, primary decomposition can be viewed as a structured way to reach irreducible-like building blocks while preserving information about radicals and nilpotent thickness.

7.2 Depth and dimension heuristics via associated primes

Associated primes influence invariants related to depth and dimension. Informally, minimal associated primes tend to reflect the “main” geometric dimension, while embedded primes often correspond to lower-dimensional phenomena and torsion that affect depth.

These heuristics do not replace formal definitions, but they guide intuition: examining which primes occur as radicals of primary components helps predict where regular sequences can exist and how modules behave under localization.

7.3 Connections to primary decomposition algorithms

Algorithmic approaches to primary decomposition frequently proceed by:

  1. computing radical and candidate associated primes,
  2. forming or guessing primary components compatible with those radicals,
  3. verifying that the intersection of components equals the original ideal.

In polynomial rings, tools such as Gröbner bases support these steps, enabling symbolic computation of primary decompositions for many classes of ideals, especially monomial ideals and ideals in well-behaved Noetherian algebras.

7.4 Using primary ideals to study quotient rings

A central conceptual benefit of primary ideals is that they translate structural questions about \(R/I\) into ideal-theoretic data. For instance, nilpotent behavior, the location of zero divisors, and the prime spectrum of the quotient can be read off from the primary components.

Because radicals of primary ideals are prime, each primary component isolates how a specific prime controls torsion in the quotient. This decomposition-based viewpoint is a cornerstone for analyzing the internal structure of rings and modules, particularly when questions about singularities, multiplicities, or torsion localization arise.