Genuine Hybrid Number–Polarization Entanglement

• The paper introduces a fidelity witness showing that macroscopic Bell states possess genuine hybrid entanglement that cannot be reproduced by convex mixtures of number- or polarization-separable states.
• The approach employs a four-mode Fock space framework, merging photon-number and polarization degrees of freedom via spontaneous parametric down-conversion and squeezing operations.
• It demonstrates that standard CV and DV witnesses are insufficient, necessitating specialized hybrid witnesses based on fidelity bounds to certify coherent inter-sector entanglement.

Searching arXiv for the core paper and closely related hybrid-entanglement references. Genuine hybrid number–polarization entanglement denotes an optical form of entanglement whose irreducible structure involves both photon-number correlations and polarization correlations, and which therefore cannot be described adequately either by the standard continuous-variable notion of mode entanglement or by the standard discrete-variable notion of polarization entanglement at fixed photon number. In the formulation introduced for macroscopic Bell states, the relevant four-mode Fock space is built from Alice’s polarization modes $A_H,A_V$ and Bob’s polarization modes $B_H,B_V$, and the key distinction is between states that are separable in number, separable in polarization, or genuinely hybrid in the sense of lying outside the convex hull of those partially separable classes (Schiffer et al., 26 May 2026). This concept sharpens earlier discussions of “hybrid entanglement,” many of which concern unlike encodings such as polarization–OAM, polarization–frequency, or polarization–cat-state entanglement rather than number–polarization hybridity proper (Nagali et al., 2011).

1. Definition and conceptual scope

The modern formulation of genuine hybrid number–polarization entanglement is given for a four-mode optical Hilbert space

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$

with basis states

$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$

Within this space, correlations may reside simultaneously in photon-number structure and in the polarization label $H/V$. The central claim is that some optical states, especially macroscopic Bell states generated by spontaneous parametric down-conversion, are not merely entangled in one of these senses but have entanglement that intrinsically uses both (Schiffer et al., 26 May 2026).

The relevant partially separable sets are defined as

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$

$$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$

A state is genuinely hybrid-entangled when it is not in

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$

so it cannot be written as a convex mixture of a number-separable state and a polarization-separable state (Schiffer et al., 26 May 2026).

This definition separates genuine hybridity from the looser observation that a state may display both number-like and polarization-like features. The distinction parallels the logic of genuine multipartite entanglement: simultaneous presence of multiple correlation types is not sufficient if the state remains decomposable into mixtures that lack one of them. A classical mixture such as

$$\rho = \sum_{N=0}^\infty w_N \ket{\psi_N^-}\bra{\psi_N^-}$$

may retain polarization entanglement sector by sector while losing the cross-sector coherence required for genuine number–polarization hybridity (Schiffer et al., 26 May 2026).

A persistent source of confusion is terminological. Several earlier optical “hybrid entanglement” proposals and demonstrations involve inequivalent local encodings, but not number–polarization entanglement in this strict sense. For example, polarization–OAM entanglement is genuinely hybrid across unlike degrees of freedom, yet its local spaces are polarization and orbital angular momentum rather than number and polarization (Nagali et al., 2011). Likewise, polarization–frequency and polarization–cat-state states are hybrid in other senses but do not realize the present notion.

2. Macroscopic Bell states as the canonical example

The principal exemplary states are the macroscopic Bell states generated by the squeezing operators

$B_H,B_V$0

$B_H,B_V$1

with $B_H,B_V$2. The resulting states are

$B_H,B_V$3

For $B_H,B_V$4, the macroscopic singlet is

$B_H,B_V$5

Reorganizing by local photon number $B_H,B_V$6 gives

$B_H,B_V$7

where

$B_H,B_V$8

Each $B_H,B_V$9 is a singlet of two effective spin-$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$0 systems, and within each fixed-$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$1 sector the reduced state satisfies

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$2

Thus every sector is maximally polarization-entangled, while the full state also preserves coherent superposition across infinitely many photon-number sectors (Schiffer et al., 26 May 2026).

This layered structure is what makes the macroscopic Bell states paradigmatic. They are not exhausted by the continuous-variable description of two-mode squeezing, because within each $$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$3-sector they possess finite-dimensional polarization-singlet structure. Nor are they exhausted by a discrete-variable account of fixed-$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$4 polarization entanglement, because the sector amplitudes

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$5

link infinitely many sectors coherently. The state is therefore neither “just CV” nor “just DV” in the conventional sense (Schiffer et al., 26 May 2026).

The same point can be restated by comparison with the ordinary two-mode squeezed vacuum,

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$6

which is the standard number-entangled continuous-variable benchmark. The macroscopic Bell state supplements such number correlations with polarization singlets at every fixed local photon number. This suggests a unification of squeezing-based and spin-like optical entanglement within a single state family (Schiffer et al., 26 May 2026).

3. Why standard CV and DV witnesses are insufficient

The reason earlier entanglement criteria fail to certify genuine hybrid number–polarization entanglement is structural. Standard polarization witnesses based on Stokes observables commute with total photon number and therefore only access the block-diagonal fixed-$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$7 content of the density matrix. The Stokes operators are given as

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$8

With total spin

$$\mathcal{H}=\mathcal{F}_{A_H}\otimes \mathcal{F}_{A_V}\otimes \mathcal{F}_{B_H}\otimes \mathcal{F}_{B_V},$$9

the macroscopic singlet satisfies

$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$0

This establishes polarization anticorrelation, but because the observables preserve total photon number they cannot distinguish a coherent superposition across $\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$1-sectors from an incoherent mixture having identical sectorwise polarization content (Schiffer et al., 26 May 2026).

Accordingly, a mixed state

$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$2

can reproduce the same polarization-based violations as the coherent macroscopic Bell state. Such a state lies in $\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$3: it may remain polarization-entangled in each fixed-$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$4 block, yet it lacks the number coherence required for genuine hybridity. Older witnesses therefore detect $\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$5, but not $\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$6 (Schiffer et al., 26 May 2026).

This limitation clarifies why superficially related notions should not be conflated. A 2021 proposal on “hybrid entanglement between optical discrete polarizations and continuous quadrature variables” constructs a state between a true single-photon polarization qubit and a cat/quadrature qubit,

$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$7

with

$\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$8

Because even and odd cat states have definite photon-number parity, the state has a parity interpretation, but the continuous subsystem remains a cat-state/quadrature qubit rather than a finite-dimensional number qubit. It is therefore best described as polarization–quadrature or polarization–cat-state hybrid entanglement, not genuine finite-dimensional number–polarization entanglement (Wen et al., 2021).

A different kind of nonstandard number–polarization nonseparability was analyzed in terms of total intensity $\ket{N_H^A,N_V^A,N_H^B,N_V^B}.$9 and polarization-like variables such as $H/V$0. Because the corresponding Hilbert-space decomposition does not factor natively as $H/V$1, that work characterized the effect as “proto-entanglement,” not ordinary subsystem entanglement. This distinction remains essential: genuine hybrid number–polarization entanglement in the modern sense is a statement about the optical four-mode state itself, not merely about commuting collective variables of a single two-mode system (Sanchidrián-Vaca et al., 2018).

4. Operational witness and fidelity bounds

The principal operational result is a fidelity witness constructed with the target macroscopic singlet: $H/V$2 The witness must satisfy

$H/V$3

so that

$H/V$4

certifies genuine hybrid number–polarization entanglement (Schiffer et al., 26 May 2026).

Defining

$H/V$5

$H/V$6

$H/V$7

one has the bound

$H/V$8

Hence a valid threshold is

$H/V$9

For number-separable states, the maximal overlap is bounded by the largest population of a fixed photon-number sector: $$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$0 where

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$1

This reflects the fact that a number-separable state may concentrate entirely in the best-matching sector and be maximally polarization-entangled there, but cannot reproduce the coherent superposition over sectors (Schiffer et al., 26 May 2026).

For polarization-separable states, the bound is

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$2

The sectorwise ingredient is the overlap constraint

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$3

valid for polarization-separable $$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$4. Consequently the central witness inequality takes the form

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$5

and genuine hybrid number–polarization entanglement is certified whenever

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$6

This is a sufficient criterion: failure to violate it does not imply absence of hybrid entanglement (Schiffer et al., 26 May 2026).

The witness philosophy is consistent with wider developments in hybrid-entanglement certification. An implementable DV–CV witness for $$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$7 versus cat-state encodings showed how heterogeneous entanglement can be tested using a small number of experimentally accessible observables rather than full tomography, although that work addressed vacuum/one-photon–cat entanglement rather than number–polarization hybridity (Masse et al., 2020). A later Bell-analysis of the state

$$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$8

showed that hybrid entanglement between a single-rail qubit and a cat-state qubit can even be certified through CHSH inequalities with no postselection, but again in a number–cat rather than number–polarization setting (Moradi et al., 2024). These works reinforce the methodological point that genuine hybridity demands certification tools matched to the mixed structure of the encoded subsystems.

5. Experimental implementation and practical regime

The proposed implementation of the genuine hybrid number–polarization witness requires two ingredients: knowledge of the squeezing parameter $$\mathrm{Sep} \coloneq \{ \rho~|~\rho \text{ n-separable } \land \text{ p-separable}\},$$9 and an estimate of the fidelity with the target macroscopic Bell state $$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$0. The squeezing parameter can be obtained from second moments of the involved quadratures, while the fidelity requires reconstruction of the Fock-space density matrix, for which tomography via pattern functions is outlined (Schiffer et al., 26 May 2026).

Because only a finite photon-number window is experimentally accessible, the witness can be evaluated in a truncated subspace. Since local filtering on photon number is an LOCC operation, truncation cannot create entanglement, so a renormalized distribution $$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$1 in the accessible subspace gives a sufficient criterion for the conditioned state in that subspace. The resulting certification is operationally meaningful but does not by itself establish a statement about the full infinite-dimensional state (Schiffer et al., 26 May 2026).

A practical issue is vacuum contamination. In the full Fock space the polarization-separable bound becomes

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$2

which is attained by

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$3

Because the vacuum sector can inflate the overlap, it is advantageous to work in the non-vacuum subspace $$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$4, with renormalized probabilities

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$5

Then

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$6

For approximately $$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$7, a value stated for macroscopic Bell-state generation in the cited discussion, the required fidelity to certify hybrid entanglement is about $$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$8 (Schiffer et al., 26 May 2026).

This implementation strategy contrasts with neighboring hybrid platforms. In integrated polarization–frequency entanglement, the state is reconstructed in a restricted hybrid polarization-frequency subspace from a joint spectral intensity and Hong–Ou–Mandel measurement, not by Fock-space coherence reconstruction. The generated state,

$$\mathrm{N} \coloneq \{ \rho~|~\rho \text{ n-separable}\},$$9

is directly generated on chip without postmanipulation, but the hybridity there is polarization–frequency, not number–polarization (Francesconi et al., 2022). The comparison underscores that the witness design depends critically on which observables jointly define the hybrid structure.

6. Relation to other forms of hybrid optical entanglement

The term “hybrid entanglement” has been used for several physically distinct architectures. Genuine hybrid number–polarization entanglement should therefore be located within a broader typology rather than treated as a synonym for all cross-encoding optical entanglement.

A clear precedent for interparticle hybridity is polarization–OAM entanglement. Starting from a polarization singlet and coherently transferring one photon’s polarization into the OAM subspace $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$0, one obtains

$$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$1

with reported tomographic figures $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$2, $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$3, $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$4, and CHSH value $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$5 (Nagali et al., 2011). This is genuinely hybrid across unlike local degrees of freedom, but not number–polarization entanglement.

A second line concerns polarization entanglement with structured vector-beam modes. In that setting the state is of the form

$$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$6

where $$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$7 are orthogonal structured single-photon spin-orbit modes. This again qualifies as hybrid entanglement between unlike encodings, but the second subsystem is a vector/spin-orbit mode, not number (Fickler et al., 2013).

A third line comprises polarization–cat or polarization–coherent-state proposals. One such scheme targets

$$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$8

with the polarization qubit defined as

$$\mathrm{P} \coloneq \{ \rho~|~\rho \text{ p-separable}\}.$$9

That construction is a genuine hybrid DV–CV state whose discrete side is a true single-photon polarization qubit, but the continuous side is a coherent-state mode. It is thus best characterized as single-photon-polarization–coherent-state entanglement rather than the macroscopic number–polarization hybridity of macroscopic Bell states (Kwon et al., 2014).

A different axis of development concerns states that are not merely hybrid but genuinely non-Gaussian in a stronger resource-theoretic sense. A recent certification framework defined genuine non-Gaussian entanglement by the impossibility of writing a pure state as

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$0

for any Gaussian unitary $\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$1 acting on a separable input, and proposed fidelity-threshold tests for both entangled Fock superpositions and hybrid number–cat states of the form

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$2

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$3

Although polarization was not treated explicitly, this framework suggests a broader classification in which genuine hybrid number–polarization entanglement may also be asked whether it lies beyond Gaussian-separable generation, albeit in a more complex multimode setting (Lachman et al., 24 Apr 2026).

7. Interpretive issues, limitations, and open directions

A common misconception is that any state correlating polarization with parity or with a nontrivial Fock decomposition is already a genuine number–polarization hybrid state. This is not generally correct. In polarization–cat states, for example, even and odd cats do correlate polarization with even- and odd-photon sectors, but the second subsystem remains an infinite-dimensional coherent-state superposition naturally defined in phase space, not a finite-dimensional number qubit (Wen et al., 2021). Genuine number–polarization hybridity, as defined for macroscopic Bell states, instead concerns the inseparability of number-sector coherence and polarization-sector entanglement within the same four-mode optical state (Schiffer et al., 26 May 2026).

A second misconception is that any nonseparability between number-like and polarization-like observables suffices. The analysis of total intensity versus polarization in two-mode coherent states showed that such variable-based nonseparability can be operationally meaningful and even convertible into ordinary entanglement by ancilla coupling, yet because the original Hilbert space does not factor naturally into number and polarization subsystems, the phenomenon was explicitly classified as proto-entanglement rather than ordinary bipartite entanglement (Sanchidrián-Vaca et al., 2018). Genuine hybrid number–polarization entanglement, by contrast, is defined directly at the state level through exclusion from $\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$4 (Schiffer et al., 26 May 2026).

The present witness is deliberately narrow. It is fidelity-based and tailored to states close to $\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$5, so it is not a universal detector for all hybrid-entangled states. It is sufficient but not necessary, depends explicitly on the squeezing parameter $\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$6, and in practice often applies to a truncated non-vacuum subspace rather than to the full infinite-dimensional state. These are methodological limitations, not contradictions of the concept itself (Schiffer et al., 26 May 2026).

Open directions follow naturally. One is a broader theory of hybrid entanglement beyond macroscopic Bell states and polarization labels. Another is extension to other labeling degrees of freedom such as time, energy, or spectral mode. A further direction is classification of states that simultaneously exhibit CV-type and DV-type structure but do not fit neatly into existing categories. The discussion of states like

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$7

and

$\mathrm{conv}(\mathrm{N}\cup\mathrm{P}),$8

suggests that number–polarization hybridity may be a special case of a broader landscape of genuine hybrid quantum correlations (Schiffer et al., 26 May 2026).

In this sense, genuine hybrid number–polarization entanglement is both a concrete property of experimentally relevant SPDC states and a proposal for reorganizing optical entanglement theory. It identifies states whose entanglement is carried not merely by continuous-variable mode correlations or by discrete-variable polarization qudits alone, but by coherent coupling of both structures in a way that cannot be reduced to mixtures of the two.