Extended Symbolic Rees Ring

• Extended Symbolic Rees Ring is a Laurent-graded enlargement of the symbolic Rees algebra that packages symbolic powers of an ideal into a graded object.
• The construction links key algebraic tools such as blow-up algebras, Cox rings, and projective module splitting, and exhibits a dimension jump (d+1) similar to classical Rees algebras.
• Finite generation and Noetherianity of these rings depend on delicate geometric and combinatorial conditions, impacting asymptotic invariants and singularity analysis.

Searching arXiv for papers on extended symbolic Rees rings and related symbolic/extended Rees algebras. An extended symbolic Rees ring is a Laurent-graded enlargement of the symbolic Rees algebra built from symbolic powers of an ideal. For an ideal $I$ in a commutative Noetherian ring $R$, one definition used in the literature is

$$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$

together with the extended symbolic form

$$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$

Recent work also uses the notation $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$, and, in the study of classical varieties, an “extended” viewpoint based on recursively generated symbolic Rees algebras for families of ideals. These constructions package symbolic powers into a single graded object and connect blow-up algebras, Cox rings, projective modules, and asymptotic invariants of ideals (Bhaumik et al., 2023, Kurano, 6 Aug 2025, Kumar et al., 2024).

1. Definitions and terminological conventions

The literature uses closely related but not identical conventions for the term “extended symbolic Rees ring.” In all cases, the construction records symbolic powers in graded form, but the ambient category and the indexing convention vary.

Usage	Formula	Setting
Laurent symbolic extension	$R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$	Ideal $I\subset R$, with $I^{(n)}=R$ for $n\le 0$ (Bhaumik et al., 2023)
Inverted symbolic Rees ring	$R'_s(I)=R_s(I)[T^{-1}]\subset A[T^{\pm1}]$	Symbolic powers defined via minimal primes of $R$0 (Kurano, 6 Aug 2025)
Family-based recursive form	$R$1	Ordered families of ideals on classical varieties (Kumar et al., 2024)

For prime ideals, symbolic powers are often written as $R$2. For non-prime ideals, one formulation in the cited literature defines $R$3 as the intersection of $R$4 over the minimal primes $R$5 of $R$6 (Sannai et al., 2017, Kurano, 6 Aug 2025). The family-based convention is structurally different: it does not merely adjoin negative degrees, but organizes an entire chain of related ideals into a recursively generated symbolic Rees algebra (Kumar et al., 2024).

A plausible implication is that statements about “the” extended symbolic Rees ring must be read together with the convention fixed by the source, especially when comparing results on Noetherianity, Cox rings, or singularity invariants.

2. Algebraic structure and dimension

For a Noetherian domain $R$7 of dimension $R$8 and a nonzero ideal $R$9, the symbolic Rees algebra and its extended version have the same dimension jump as ordinary Rees algebras: $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$0 These formulas appear as Lemma 2.7 and Lemma 2.8 in the study of projective modules over symbolic and extended symbolic Rees algebras (Bhaumik et al., 2023). Because symbolic Rees algebras can be non-Noetherian, that paper also introduces the valuation dimension $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$1, defined as the supremum of the dimensions of valuation overrings of $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$2, to control dimension-theoretic arguments beyond the Noetherian setting (Bhaumik et al., 2023).

In a different direction, the paper on symbolic powers of classical varieties gives explicit recursive generators for symbolic Rees algebras in the “main setting”: $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$3 For generic determinantal ideals, minors of generic symmetric matrices, generic extended Hankel matrices, and ideals of pfaffians of skew-symmetric matrices, the symbolic Rees algebra is Noetherian and generated by this recursive family; Bruns–Vetter and Jeffries–Montano–Varbaro are cited there for these finite-generation results (Kumar et al., 2024).

The structural consequence of these formulas is twofold. First, in the non-Noetherian direction, the extended symbolic Rees ring remains amenable to dimension-theoretic and patching arguments. Second, in the finitely generated direction, the recursive presentation makes symbolic powers accessible to explicit computation of invariants such as least degrees, Waldschmidt constants, and resurgence (Kumar et al., 2024).

3. Finite generation, non-Noetherian phenomena, and effective criteria

Finite generation of symbolic and extended symbolic Rees rings is highly sensitive to the underlying geometry. It is not guaranteed in general: the literature cited in work on classical varieties names Cowsik, Huneke, and Roberts as sources of counterexamples, and Roberts is explicitly referenced for non-finite generation of symbolic Rees algebras (Kumar et al., 2024, Bhaumik et al., 2023).

A sharp negative result is that for the polynomial ring over an arbitrary field with twelve variables, there exists a prime ideal whose symbolic Rees algebra is not finitely generated (Sannai et al., 2017). In that construction, finite generation of the symbolic Rees algebra is equivalent to finite generation of a multisection ring and of the Cox ring of a blow-up, via Proposition 2.14 (Sannai et al., 2017). This ties non-Noetherianity directly to birational geometry.

For space monomial curves, Kurano and Nishida study the symbolic Rees ring $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$4 of the defining ideal $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$5 of $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$6. Under characteristic $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$7, three-generator hypotheses, and the assumption that the minimal-degree homogeneous element in $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$8 defines a negative curve with $$R_s(I)=\bigoplus_{n\geq 0} I^{(n)}x^n \subset R[x], \qquad I^{(n)}=\bigcap_{p\in \mathrm{Ass}(R/I)}(I^nR_p\cap R),$$9, they prove that $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$0 is Noetherian if and only if there exists a homogeneous element in $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$1 satisfying Huneke’s condition together with the negative-curve element $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$2 (Kurano et al., 2017). They also introduce the combinatorial condition EU; for $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$3, it is necessary and sufficient for finite generation, and if EU fails while GK holds, the symbolic Rees ring is infinitely generated (Kurano et al., 2017).

A more explicit criterion appears in the 2022 paper on weighted projective blow-ups. If there exists a negative curve $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$4 with $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$5 and $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$6, then the symbolic Rees ring

$$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$7

is Noetherian if and only if the EMU condition holds for $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$8 (Inagawa et al., 2022). In that criterion, if $$R_s(I,x^{-1})=\bigoplus_{n\in \mathbb{Z}} I^{(n)}x^n \subset R[x,x^{-1}], \qquad I^{(n)}=R \text{ for } n\leq 0.$$9 is the sorted list of certain lattice-point counts attached to a triangle, then EMU requires

$R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$0

The same paper states that $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$1 is finitely generated over $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$2 if and only if $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$3 is finitely generated, and that the Cox ring of $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$4 is isomorphic to an extended symbolic Rees ring (Inagawa et al., 2022).

The existence of a negative curve is not, by itself, sufficient for finite generation. Kurano and Nishida provide an infinitely generated example in which the minimal-degree element in $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$5 is a negative curve, showing that the further orthogonality and Huneke-type conditions are essential (Kurano et al., 2017).

Characteristic dependence can be equally sharp. For a toric surface $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$6 determined by a triangle and its blow-up $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$7 at $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$8, the Cox ring is identified with an extended symbolic Rees ring $R'_s(I)=R_s(I)[T^{-1}] \subset A[T^{\pm1}]$9. In the explicit example

$R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$0

the ideal $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$1 is not prime, $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$2, $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$3, and $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$4 is Noetherian if and only if the characteristic of $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$5 is $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$6 or $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$7 (Kurano, 6 Aug 2025).

4. Projective modules and Serre-type splitting

A major structural theorem for extended symbolic Rees algebras concerns projective modules. Let $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$8 be a commutative Noetherian domain of dimension $R_s(I,x^{-1})=\bigoplus_{n\in\mathbb Z} I^{(n)}x^n$9, let $I\subset R$0 be an ideal, and let

$I\subset R$1

If $I\subset R$2 is a finitely generated projective $I\subset R$3-module of rank $I\subset R$4, then $I\subset R$5 has a unimodular element; equivalently, $I\subset R$6 for some $I\subset R$7 (Bhaumik et al., 2023). Since $I\subset R$8, this extends the familiar “rank $I\subset R$9 dimension” splitting phenomenon to symbolic and extended symbolic Rees algebras, even though these rings may be non-Noetherian (Bhaumik et al., 2023).

The proof strategy is based on sheaf patching. The argument localizes at the multiplicative set of non-zero-divisors, reduces to polynomial or Laurent polynomial situations where unimodular elements are already known, and then glues local data through fiber-product constructions and surjective homomorphisms. Lemmas 3.1 and 3.2 identify localizations of symbolic Rees algebras with more classical rings, while Lemma 3.3 supplies the gluing step; Heitmann’s results are used in the dimension-theoretic and patching arguments (Bhaumik et al., 2023).

The significance of this theorem is that no additional obstruction theory is needed in the high-rank range for symbolic and extended symbolic Rees algebras. In the setting covered there, projective modules of sufficiently large rank still split off a free summand, despite the possible failure of finite generation for the ring itself (Bhaumik et al., 2023).

5. Singularity-theoretic and canonical-module methods around extended Rees constructions

Several papers study the ordinary extended Rees algebra

$I^{(n)}=R$0

and then state consequences or extensions to symbolic and extended symbolic Rees settings.

For test modules in positive characteristic, the decomposition

$I^{(n)}=R$1

implies in particular that the degree-zero component is

$I^{(n)}=R$2

The same paper states that this “principalization” reduces the computation of test modules for non-principal ideals to the principal case and gives explicit control over $I^{(n)}=R$3-singularities in extended Rees algebras and, consequently, via the associated graded, for symbolic and extended symbolic Rees rings (Ajit et al., 1 Sep 2025).

In characteristic $I^{(n)}=R$4, multiplier modules admit the analogous decomposition

$I^{(n)}=R$5

with degree-zero piece $I^{(n)}=R$6. That work explicitly states that the same techniques and formulas apply, via analogous graded structures, to symbolic and extended symbolic Rees algebras (Ajit, 24 Oct 2025). Budur, Mustaţă, and Saito are cited there for earlier smooth-case results obtained via $I^{(n)}=R$7-filtration theory (Ajit, 24 Oct 2025).

A filtration-theoretic quasi-Gorenstein criterion also reaches symbolic powers. For a Hilbert filtration $I^{(n)}=R$8, the extended Rees algebra

$I^{(n)}=R$9

is quasi-Gorenstein, under the paper’s depth and finiteness hypotheses, if and only if a specific length equality between local cohomology modules of $n\le 0$0 and the associated graded ring $n\le 0$1 holds. Symbolic powers, integral closures, and Ratliff–Rush closures are listed there as examples to which the theory applies when the associated Rees algebra is Noetherian (Endo, 4 Feb 2025).

These developments do not redefine the extended symbolic Rees ring, but they show that the homological toolkit for extended Rees algebras—test modules, multiplier modules, and quasi-Gorenstein criteria—has already been formulated in ways that transfer to symbolic filtrations or to analogous symbolic graded structures.

6. Geometric realizations, Cox rings, and asymptotic invariants

The extended symbolic Rees ring frequently appears as a Cox ring or multisection ring of a blow-up. For the defining ideal $n\le 0$2 of a space monomial curve, the symbolic Rees ring coincides with the Cox ring of the blow-up $n\le 0$3 of the weighted projective surface $n\le 0$4 at a typical point (Inagawa et al., 2022). For blow-ups of projective space along a smooth connected subvariety, the equivalence between finite generation of the symbolic Rees algebra, a multisection ring, and the Cox ring is made explicit in Proposition 2.14 (Sannai et al., 2017). For toric blow-ups at $n\le 0$5, the Cox ring of $n\le 0$6 is identified with $n\le 0$7, even when $n\le 0$8 is not prime; when $n\le 0$9 is torsion-free, $R'_s(I)=R_s(I)[T^{-1}]\subset A[T^{\pm1}]$0 is the defining ideal of a space monomial curve (Kurano, 6 Aug 2025).

This geometric interpretation explains why finite generation is often recast as a Mori dream space problem, and why negative curves, nef but non-semiample divisors, and divisor classes disjoint from exceptional curves govern the algebraic behavior of symbolic Rees rings (Sannai et al., 2017, Inagawa et al., 2022, Kurano, 6 Aug 2025).

In the Noetherian recursive setting of classical varieties, the symbolic Rees algebra provides explicit asymptotic data. The paper on symbolic powers of classical varieties computes, under its hypotheses,

$R'_s(I)=R_s(I)[T^{-1}]\subset A[T^{\pm1}]$1

and shows stronger Chudnovsky- and Demailly-type bounds. The examples include generic determinantal ideals, minors of generic symmetric matrices, generic extended Hankel matrices, ideals of pfaffians, and also star configurations (Kumar et al., 2024).

Taken together, these results place the extended symbolic Rees ring at a crossroads of symbolic power theory and birational geometry. It is simultaneously a receptacle for symbolic powers, a test object for finite generation, a Cox ring in many geometric realizations, and a graded algebra on which module-theoretic and singularity-theoretic techniques can still be deployed effectively, even when Noetherianity fails in general.