2025-11-15T04:22:11.347744

Sárközy's theorem for shifted primes with restricted digits

Burgin
For a base $b\geq 2$ and a set of digits $\mathcal{A}\subset \{0,...,b-1\}$, let $\mathcal{P}$ denote the set of prime numbers with digits restricted to $\mathcal{A}$, when written in base-$b$. We prove that if $A\subset \mathbb{N}$ has positive upper Banach density, then there exists a prime $p\in \mathcal{P}$ and two elements $a_1,a_2\in A$ such that $a_2=a_1+p-1$. The key ingredients are the Furstenberg correspondence principle and a discretized Hardy-Littlewood circle method used by Maynard. As a byproduct of our work, we prove a Dirichlet-type theorem for the distribution of $\mathcal{P}$ in residue classes, and a Vinogradov-type theorem for the decay of associated exponential sums. These estimates arise from the unique structure of associated Fourier transforms, which take the form of Riesz products.
academic

Sárközy's theorem for shifted primes with restricted digits

Basic Information

  • Paper ID: 2510.13076
  • Title: Sárközy's theorem for shifted primes with restricted digits
  • Author: Alex Burgin (Georgia Institute of Technology)
  • Classification: math.NT (Number Theory)
  • Publication Date: October 15, 2025
  • Paper Link: https://arxiv.org/abs/2510.13076

Abstract

This paper investigates Sárközy's theorem for shifted primes with restricted digits. For a base b2b \geq 2 and a digit set A{0,...,b1}\mathcal{A} \subset \{0,...,b-1\}, let P\mathcal{P} denote the set of primes whose base-bb digits are restricted to A\mathcal{A}. The author proves that if ANA \subset \mathbb{N} has positive upper Banach density, then there exist a prime pPp \in \mathcal{P} and two elements a1,a2Aa_1, a_2 \in A such that a2=a1+p1a_2 = a_1 + p - 1. The main technical tools are the Furstenberg correspondence principle and the discretized Hardy-Littlewood circle method employed by Maynard. As a byproduct, the paper also establishes a Dirichlet-type theorem for the distribution of P\mathcal{P} in residue classes and a Vinogradov-type theorem on the decay of exponential sums.

Research Background and Motivation

Problem Background

  1. Sárközy-type Problems: A central problem in arithmetic combinatorics investigating forbidden difference patterns in integer sets. For a set SS, if AZA \subset \mathbb{Z} satisfies a1a2Sa_1 - a_2 \notin S for all a1,a2Aa_1, a_2 \in A, what is the structure of AA?
  2. Intersective Sets: Many sets SS (such as perfect squares, shifted primes {p±1:pP}\{p \pm 1 : p \in P\}, etc.) are intersective, meaning that if AA avoids all differences in SS, then AA is small, satisfying limN#(A{N,...,N})2N+1=0\lim_{N\to\infty}\frac{\#(A \cap \{-N,...,N\})}{2N+1} = 0.
  3. Primes with Restricted Digits: Inspired by Maynard's results on primes with restricted digits, this work studies sets satisfying two simultaneous constraints: being both shifted primes and having restricted digits.

Research Motivation

This paper aims to prove that the set of shifted primes with restricted digits PC1P_C - 1 (where CC is the restricted digit set) retains the intersective property, thereby generalizing Sárközy's classical result to sparser subsets of primes.

Core Contributions

  1. Main Theorem: Proves that for a set AA with positive upper Banach density, there exist a prime pp with restricted digits and a1,a2Aa_1, a_2 \in A such that a2=a1+p1a_2 = a_1 + p - 1.
  2. Dirichlet-type Theorem: Establishes asymptotic formulas for the distribution of restricted-digit primes in residue classes, generalizing Maynard's results.
  3. Vinogradov-type Theorem: Proves decay properties of exponential sums over restricted-digit primes.
  4. van der Corput Sets: Demonstrates that PC1P_C - 1 is a van der Corput set, which is stronger than being intersective.

Methodology Details

Problem Setup

Let b2b \geq 2 be a base and A:={0,...,b1}{d1,...,ds}A := \{0,...,b-1\} \setminus \{d_1,...,d_s\} be the set of allowed digits. Define the restricted digit set: C:={i=0Nnibi:niA,NN0}C := \left\{\sum_{i=0}^N n_i b^i : n_i \in A, N \in \mathbb{N}_0\right\}

The goal is to prove that if ANA \subset \mathbb{N} has positive upper Banach density, then there exist a prime pPCp \in P_C (primes in CC) and a1,a2Aa_1, a_2 \in A such that a1+p1=a2a_1 + p - 1 = a_2.

Core Technical Framework

1. Furstenberg Correspondence Principle

Through dynamical systems theory, the combinatorial problem is transformed into an ergodic-theoretic problem. The key result is:

Proposition 7: For any measure-preserving system (X,B,μ,T)(X,\mathcal{B},\mu,T) and fL(X,B,μ)f \in L^{\infty}(X,\mathcal{B},\mu) with f0,f≢0f \geq 0, f \not\equiv 0, the set {n>0:fTnfdμ>0}\{n > 0 : \int f \cdot T^n f d\mu > 0\} contains elements from PC1P_C - 1.

2. Exponential Sum Estimates

Precise estimates are established for the Fourier transform C^bN(t)\hat{C}_{b^N}(t) of the restricted digit set:

C^bN(x)=i=0N1(cAe(bicx))\hat{C}_{b^N}(x) = \prod_{i=0}^{N-1}\left(\sum_{c \in A} e(b^i c x)\right)

Key estimates include:

  • L1L^1 Bound: supxRabNC^bN(x+a/bN)(C0blogb)N\sup_{x \in \mathbb{R}} \sum_{a \leq b^N} |\hat{C}_{b^N}(x + a/b^N)| \leq (C_0 b \log b)^N
  • Large Sieve Estimates and Mixed Estimates to control sums at different scales
  • LL^{\infty} Bound: Exponential decay for poorly approximable rationals

3. Circle Method Decomposition

Uses the discretized Hardy-Littlewood circle method:

Proposition 28: For α<1/5\alpha < 1/5, n<bN1C(n)Λ(n)e(nθ)=d<logA(bN)μ(d)ϕ(d)(Z/dZ)C^bN(θ+d)+OA((bs)NlogA(bN))\sum_{n<b^N} 1_C(n)\Lambda(n)e(n\theta) = \sum_{d<\log^{A'}(b^N)} \frac{\mu(d)}{\phi(d)} \sum_{\ell \in (Z/dZ)^*} \hat{C}_{b^N}\left(\theta + \frac{\ell}{d}\right) + O_A\left(\frac{(b-s)^N}{\log^A(b^N)}\right)

Technical Innovations

  1. Riesz Product Structure: The Fourier transform of the restricted digit set exhibits a special product structure allowing precise analytic control.
  2. bb-adic Methods: Exploits the periodicity properties of bb-adic expansions to decompose the problem into subproblems at different scales.
  3. Condition Optimization: By requiring bs>(k+1)b4/5+ϵb - s > (k+1)b^{4/5+\epsilon}, the main term dominates.

Main Results

Theorem 3 (Main Theorem)

Let ANA \subset \mathbb{N} have positive upper Banach density, and C=C(b,A)C = C(b,\mathcal{A}) satisfy:

  • (I) 1A1 \in \mathcal{A}
  • (II) The forbidden digit set {d1,...,ds}=i=1kIi\{d_1,...,d_s\} = \bigsqcup_{i=1}^k I_i (disjoint union of intervals)
  • (III) bs>(k+1)b4/5+ϵb - s > (k+1)b^{4/5+\epsilon}, bb sufficiently large

Then there exist a prime pCp \in C and a1,a2Aa_1, a_2 \in A such that a1+p1=a2a_1 + p - 1 = a_2.

Theorem 4 (Dirichlet-type Theorem)

For any m1m \geq 1 and tZ/mZt \in \mathbb{Z}/m\mathbb{Z}, 0n<bNnt(modm)1C(n)Λ(n)=κm,t(bs)N+OC((bs)NlogC(bN))\sum_{\substack{0 \leq n < b^N \\ n \equiv t \pmod{m}}} 1_C(n)\Lambda(n) = \kappa_{m,t}(b-s)^N + O_C\left(\frac{(b-s)^N}{\log^C(b^N)}\right)

where κm,t\kappa_{m,t} is an explicit constant.

Theorem 6 (Vinogradov-type Theorem)

For any irrational θ\theta, 0n<bN1C(n)Λ(n)e(nθ)=o((bs)N)\sum_{0 \leq n < b^N} 1_C(n)\Lambda(n)e(n\theta) = o((b-s)^N)

Technical Details

Core Lemmas for Fourier Estimates

Lemma 12 (L1L^1 Bound): If C0:=k+1+2(bs)blogbC_0 := k + 1 + \frac{2(b-s)}{b}\log b, then supxRabNC^bN(x+abN)(C0blogb)N\sup_{x \in \mathbb{R}} \sum_{a \leq b^N} \left|\hat{C}_{b^N}\left(x + \frac{a}{b^N}\right)\right| \leq (C_0 b \log b)^N

Lemma 20 (LL^{\infty} Bound): Let 1<d<bN/31 < d < b^{N/3}, Z\ell \in \mathbb{Z} such that bi/dZb^i\ell/d \notin \mathbb{Z} for all i1i \geq 1, and ϵ<(2b2N/3)1|\epsilon| < (2b^{2N/3})^{-1}. Then C^bN(d+ϵ)(bs)Nexp(cN/logd)\left|\hat{C}_{b^N}\left(\frac{\ell}{d} + \epsilon\right)\right| \leq (b-s)^N \exp(-cN/\log d)

Inversion Theorem

Proposition 24: For θ,xT\theta, x \in \mathbb{T} and sufficiently large BB, η<logB(bN)bNx+ηZC^bN(θ+x+ηbN)k=0bN1e(kηbN)=bNC^bN(θ+x)+O(bN(bs)NlogA(bN))\sum_{\substack{|\eta| < \log_B(b^N) \\ b^N x + \eta \in \mathbb{Z}}} \hat{C}_{b^N}\left(\theta + x + \frac{\eta}{b^N}\right) \sum_{k=0}^{b^N-1} e\left(-\frac{k\eta}{b^N}\right) = b^N \hat{C}_{b^N}(\theta + x) + O\left(\frac{b^N(b-s)^N}{\log^A(b^N)}\right)

Proof Strategy

Step 1: Establish Exponential Sum Estimates

Analyze the product structure of C^bN(t)\hat{C}_{b^N}(t) to establish precise upper bounds in different regions.

Step 2: Apply Circle Method

Decompose the exponential sum into major arcs (near rationals with small denominators) and minor arcs (remaining regions), handling major arcs via the inversion theorem and controlling minor arcs through Fourier estimates.

Step 3: Furstenberg Correspondence

Use the spectral theorem and mean ergodic theorem to reduce the dynamical systems problem to the asymptotic behavior of exponential sums.

  1. Sárközy (1978): Proved the intersective property of shifted primes, initiating this research direction.
  2. Green (2024): Achieved breakthrough power-saving results on Sárközy's problem for shifted primes.
  3. Maynard (2021): Established asymptotic formulas for primes with restricted digits, providing technical foundations for this work.
  4. Furstenberg (1977): Established the correspondence between combinatorics and ergodic theory.

Conclusions and Discussion

Main Conclusions

This paper successfully generalizes Sárközy's theorem to shifted primes with restricted digits, proving that the intersective property persists even under this dual constraint.

Limitations

  1. The condition bs>(k+1)b4/5+ϵb - s > (k+1)b^{4/5+\epsilon} is quite restrictive, limiting the scope of applicability.
  2. The exponent 4/5 stems from known bounds on exponential sums over primes and may not be optimal.
  3. The results are qualitative without quantified power-saving estimates.

Future Directions

  1. Quantified Estimates: Can power-saving bounds be obtained as Green achieved for shifted primes?
  2. Exponent Improvement: Can the 4/5 exponent condition be relaxed?
  3. More General Patterns: Extend to more complex additive combinatorial patterns.

In-Depth Evaluation

Strengths

  1. Technical Innovation: Cleverly combines analytic number theory (circle method), ergodic theory (Furstenberg correspondence), and Fourier analysis.
  2. Deep Results: Establishes nontrivial additive combinatorial structures on extremely sparse prime subsets.
  3. General Methodology: The developed techniques have potential value for studying other number-theoretic problems with restricted digits.
  4. Completeness: Not only proves the main result but also establishes corresponding distribution theorems and exponential sum estimates.

Weaknesses

  1. Restrictive Conditions: Technical conditions are relatively stringent, limiting practical applicability.
  2. Non-constructive: The proof is existential without providing algorithms for finding specific examples.
  3. Complex Constants: The dependence relationships among various constants are intricate, making numerical estimation difficult.

Impact

This paper makes important contributions at the intersection of analytic number theory and additive combinatorics, opening new technical pathways for studying prime distribution under restricted conditions. The developed methodology has significant reference value for related problems.

Applicable Scenarios

The method applies to studying additive properties of prime subsets under various restrictions, particularly those involving digit restrictions and congruence conditions.

References

  1. Sárközy, A. (1978). On difference sets of sequences of integers. III. Acta Math. Acad. Sci. Hungar.
  2. Green, B. (2024). On Sárközy's theorem for shifted primes. J. Amer. Math. Soc.
  3. Maynard, J. (2021). Primes and Polynomials with restricted digits. International Mathematics Research Notices.
  4. Furstenberg, H. (1977). Ergodic behavior of diagonal measures and a theorem of Szemerédi on arithmetic progressions. Anal. Math.

This paper demonstrates considerable technical depth and innovation, successfully extending the classical Sárközy theorem to a more refined and challenging setting, making meaningful contributions to the development of analytic number theory and additive combinatorics.