Integral zeros of Krawtchouk polynomials and quantum entanglements

Introduction

The relationship between combinatorial orthogonal polynomials and quantum information theory has produced several surprising bridges, notably the connection between integral zeros of binary Krawtchouk polynomials and the range criterion for pairs of subspaces in multipartite entanglement. Heo and Kiem (2019) posed a concrete problem: for a quadruple \((m,n,k,\ell)\) with \(m\le n\) and \(k\le \ell\), determine when the associated Krawtchouk polynomial \(K_{m}^{(n)}(x)\) admits an integer root that simultaneously satisfies the range criterion. Their work settled the symmetric case \(k=\ell\), but the asymmetric regime remained largely open.

In the present paper, Heo, Kiem and Lee advance the program by solving the problem for three infinite families of asymmetric parameters and by showing that all other instances reduce to a finite set of sporadic zeros. The result not only completes a combinatorial classification but also clarifies which subspace configurations can generate entangled states that meet the range criterion.

---

1. Krawtchouk Polynomials and Integral Zeros

Binary Krawtchouk polynomials \(K{m}^{(n)}(x)\) arise from the weight distribution of Hamming codes and are defined by \[ K{m}^{(n)}(x)=\sum{j=0}^{m}(-1)^{j}\binom{x}{j}\binom{n-x}{m-j}. \] Their zeros are well‑studied in coding theory, yet the subset that are integral (i.e., integer‑valued) is sparse and highly structured. For a given quadruple \((m,n,k,\ell)\), the authors consider the polynomial \(K{m}^{(n)}(x)\) evaluated at \(x=k\) and \(x=\ell\). An integral zero occurs when \(K_{m}^{(n)}(k)=0\) (or similarly for \(\ell\)). The existence of such a zero is equivalent, via the range criterion, to the existence of a pair of subspaces \(A\subseteq\mathbb{C}^{n}\) and \(B\subseteq\mathbb{C}^{n}\) whose dimensions are \(k\) and \(\ell\) respectively, and whose orthogonal complements intersect in a prescribed way that yields a non‑trivial entangled state.

---

2. Solved Asymmetric Families

The paper treats three parametrised families, each defined by a simple integer parameter \(t\) or \(s\). For each family the authors prove that the associated Krawtchouk polynomial has an integral zero, thereby establishing the existence of the required subspace pair.

2.1 Family A: \((m,n,k,\ell)=(3, t^{2}\! -\!1, \tfrac{t(t-1)}{2}, \tfrac{t(t+1)}{2})\), \(t\ge2\)

Using elementary combinatorial identities, the authors rewrite \(K_{3}^{(t^{2}-1)}\bigl(\tfrac{t(t-1)}{2}\bigr)\) as a product of linear factors that cancel precisely when \(t\) is an integer. A direct induction on \(t\) shows the expression vanishes, confirming the integral zero for every \(t\ge2\).

2.2 Family B: \((m,n,k,\ell)=(4, 3t^{2}+2t-1, \tfrac{3t^{2}-t}{2}, \tfrac{3t^{2}+5t+2}{2})\), \(t\ge2\)

Here the analysis is more involved: the authors employ generating‑function techniques to express \(K_{4}^{(3t^{2}+2t-1)}\) in terms of binomial sums that simplify under the substitution \(k=\tfrac{3t^{2}-t}{2}\). After a careful parity argument, the sum collapses to zero for all admissible \(t\).

2.3 Family C: \((m,n,k,\ell)=(4, 3s^{2}+4s, \tfrac{3s^{2}+s}{2}, \tfrac{3s^{2}+7s+4}{2})\), \(s\ge1\)

The third family mirrors Family B but with a shifted parameter. By exploiting a symmetry between \(k\) and \(\ell\) in the Krawtchouk recurrence, the authors reduce the problem to the already‑solved case of Family B, thereby establishing the integral zero for all \(s\ge1\).

---

3. Sporadic Zeros and Finiteness

Beyond the three infinite families, the authors turn to the remaining parameter space with \(m\ge5\). A combination of computational enumeration (up to a modest bound) and a Diophantine‑type argument shows that any integral zero must satisfy a restrictive quadratic inequality. Consequently, for each fixed \(m\ge5\) there are only finitely many admissible \((n,k,\ell)\) triples, yielding a sporadic set of solutions. The paper lists these exceptional cases explicitly for \(m=5,6\) and proves that no further infinite families can arise.

---

4. Implications for Quantum Entanglement

From the quantum‑information perspective, each integral zero corresponds to a concrete construction of a bipartite subspace pair that meets the range criterion, a sufficient condition for the existence of edge states—entangled states that cannot be distilled by local operations and classical communication. By providing explicit families, the authors supply a systematic method for generating families of edge states with prescribed dimensions. Moreover, the finiteness result for \(m\ge5\) suggests that highly asymmetric configurations are rare, aligning with the intuition that entanglement becomes increasingly constrained as the dimension gap widens.

---

Conclusion

Heo, Kiem and Lee close a notable gap in the interplay between combinatorial polynomial theory and quantum information. Their constructive proofs for three infinite asymmetric families, together with a comprehensive classification of the remaining sporadic cases, give a complete answer to the problem posed in 2019. The work not only enriches the theory of Krawtchouk polynomials but also equips quantum‑information researchers with new families of subspace configurations that guarantee entanglement under the range criterion.