Demonstration of unpartible entanglement
Philip Held, Laura Ares, Federico Pegoraro, Jonas Lammers, Benjamin Brecht, Jan Sperling, Christine Silberhorn · 2026
Read Demonstration of unpartible entanglement by Philip Held, Laura Ares, Federico Pegoraro, Jonas Lammers, Benjamin Brecht, Jan Sperling, Christine Silberhorn (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-30468.txt.
Introduction
Philip Held and six co-authors show a form of quantum entanglement that survives renaming the parties. Entanglement normally depends on how a system is split into parts. Change the mode basis and a state can lose its entanglement. The mode-independent kind persists under any such change of basis. That toughness suits quantum links over untrusted channels. No experiment had shown it before this 2026 work.
A flexible interferometer generates the states. Measurement-induced nonlinearities help produce heralded two-photon states. The states are entangled in two modes for every orthonormal basis choice. Tailored tomography certifies the claim in the lab. The measured fidelities validate its presence. The correlation is both resilient and operationally useful, they argue. It safeguards such protocols against basis changes. Noise and hostile parties in real networks motivate the pursuit.
Catalog record
| CREATOR | Philip Held, Laura Ares, Federico Pegoraro, Jonas Lammers, Benjamin Brecht, Jan Sperling, Christine Silberhorn |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 22 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.30468
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Demonstration of unpartible entanglement Philip Held, Laura Ares, Federico Pegoraro, Jonas Lammers, Benjamin Brecht, Jan Sperling, Christine Silberhorn Abstract We report on the first experimental verification of mode-independent entanglement. Commonly, the entanglement of a state is firmly based on pre-defined parties that are correlated, and the state might be disentangled when the definition of the parties is changed. Exceeding this party-dependent concept, we realize a type of quantum entanglement that persists even if the parties, in our case modes, are transformed. This safeguards the performance of entanglement in real-world applications, such as quantum communication settings involving noise and untrusted parties. For the state generation, we present an experimental scheme based on a fully reconfigurable temporally multiplexed interferometer with measurement-induced nonlinearities, which generates heralded two-photon states in two modes that are entangled for all choices of orthonormal mode basis. For the certification process, we utilize a tailored quantum-state tomography, achieving fidelities that validate the presence of mode-independent entanglement as a resilient and operationally advantageous quantum correlation. Introduction.— Entanglement neumann ; RevEnt constitutes a key resource for quantum technologies, such as communication E91 , sensing FreqEnt , and computing QuComp . As this quantum feature emerges from the non-separability of parties, it is, by construction, dependent on the choice of the subsystems that constitute a party. Diverse notions of entanglement emerge based on the physical property that defines each party. One example is particle entanglement, as in the case of EPR states EPR , a prime resource for quantum teleportation QuTele and non-locality tests NonLocal . Further examples show up when the parties considered are a system’s different degrees of freedom, leading to, for example, entanglement within single 1PartEnt or multiple particles 2PartEnt , between electromagnetic fields Enk , and between the internal and external degrees of freedom of a quantum walker EntQW . In particular, entanglement between optical modes ModeEnt has been shown to be a powerful resource for quantum technologies N00Nmeas ; singlePhoComm ; singlePho32Time . In mode entanglement, the identification of the subsystems—who is Alice and who is Bob—depends on the chosen mode basis ModeBasis . Thus, the decomposition into subsystems can be readily modified by transformations across modes. As a consequence, such “nonlocal” transformations can profoundly affect whether a given state is entangled or separable Enk . This feature poses a challenge for quantum protocols operating over non–mode-preserving channels RobustEnt , motivating the search for states that remain entangled under arbitrary mode representations, and leading to the definition of mode-independent quantum entanglement (MIQE) theo_orig . This robust form of quantum correlations prevents even partial separability of the states under any linear unitary transformation of the bosonic creation operators, i.e., the choice of the mode basis. Note that the resulting states are thus not maximally entangled in any basis; however, they remain entangled even in the absence of a pre-defined mode basis theo_orig . While the controlled generation of various entangled states has seen significant advances, the reliable certification of entanglement under realistic conditions remains a central challenge EntHard . Beyond the intrinsic NP-hardness of deciding separability LG03 ; SG10 , imperfections in practical implementations, such as noise and finite statistics, severely limit the reliable observation of these correlations GT2009 . For this reason, entanglement validation increasingly relies on witness-based criteria EntWit2 ; SV13 , which are experiment-tailored bounds. These witnesses are sufficient but not necessary criteria for entanglement EntWit1 , yet directly applicable. Experimentally, several strategies are used to measure the quantities required for witnessing entanglement. Examples are quantum state tomography FullTomo , reduced schemes like shadow tomography Shadow , and quantum reservoir computing QuResComp . In this letter, we report on the first certification of mode-independent entanglement theo_orig . To this end, we generate different states within the most fundamental family of states that present this quantum correlation, i.e., involving the minimal number of modes and particles, in a heralded scheme. While this concept applies to all bosonic systems, we implement it in a photonic architecture, a promising platform for a variety of quantum technologies; for a review, see PhoQuTec . The different states are generated using a fully reconfigurable temporally multiplexed interferometer (TMI) CNOT with measurement-induced non-linearities. A tailored quantum-state tomography is employed to analyze the generated states. We witness the presence of MIQE in several of the reconstructed states with up to six standard deviations. Furthermore, we explicitly show the persistence of this quantum correlation under mode transformations for a generated state, opening a path toward realistic mode-entanglement-based quantum technologies. Theoretical framework.— In order to describe MIQE, we consider two orthogonal optical modes a and b represented by the bosonic creation operators \hat{a}^{\dagger} and \hat{b}^{\dagger} , respectively. Transformations between modes are performed by unitary operators, \hat{U}(\theta,\phi)=\begin{pmatrix}\cos(\theta/2)&e^{i\phi}\sin(\theta/2)\\ -e^{-i\phi}\sin(\theta/2)&\cos(\theta/2)\end{pmatrix}. (1) Furthermore, let us recall that a pure state |\psi\rangle with a fixed total photon number n+m is entangled when |\psi\rangle\neq|n\rangle_{a}\otimes|m\rangle_{b}=|n,m\rangle_{a,b} , where n and m represent the number of photons in each respective mode. Before presenting the family of states under study, exhibiting the MIQE property, we contrast two states broadly utilized in photonics to build an intuition for this quantum feature. Combining two indistinguishable photons, |1,1\rangle_{a,b} , one on each input port of a balanced beam splitter (BS), yields a state of the form (|2,0\rangle_{a,b}-|0,2\rangle_{a,b})/\sqrt{2} because of Hong-Ou-Mandel interference (HOMI) N00N ; HOM . As this two-particle N00N state possesses mode entanglement that has been generated from a separable state by linear unitary operations, it is clearly not resilient against changes in the mode representation. Another example worth to mention are Bell states, (|1,0,1,0\rangle_{1a,1b,2a,2b}-|0,1,0,1\rangle_{1a,1b,2a,2b})/\sqrt{2} , where subscripts 1 and 2 refer to the two particles involved. Experimental validation of MIQE for these states is not possible so far since no criteria for certifying it in this class of states have been developed for four-mode states. We now present the family of states involving the minimum number of modes and photons that carry the MIQE property theo_orig , dubbed MIQE states throughout the remainder of this work. Such states are generated when two photons are excited in two non-parallel and non-orthogonal modes, i.e., modes of the form \hat{a}^{\dagger} and \hat{a}^{\dagger}+\lambda\hat{b}^{\dagger} , respectively, with \lambda being a nonzero complex parameter. This construction leads to the family of states |\Psi(\lambda)\rangle=\frac{\sqrt{2}|2,0\rangle_{a,b}+\lambda|1,1\rangle_{a,b}}{\sqrt{2+|\lambda|^{2}}}. (2) A schematic comparison between N00N, Bell, and the family of MIQE states regarding the persistence of entanglement under mode transformations is depicted in Fig. 1 . To certify mode-independent entanglement, we employ the measurable criterion derived in Ref. theo_orig . It is based on the idea of entanglement witnesses: Only entangled states \hat{\rho} can produce expectation values of a certain observable \hat{L} exceeding the maximal value attainable by separable states \hat{\sigma} sigma ; SV13 , thus \mathrm{tr}(\hat{\rho}\hat{L})>\sup_{\hat{\sigma}}\mathrm{tr}(\hat{\sigma}\hat{L}) . Since the observable utilized is the projector onto the target state, its expectation value corresponds to a quantum state fidelity (QSF) \mathcal{F}=\mathrm{tr}(\hat{\rho}|\Psi(\lambda)\rangle\langle\Psi(\lambda)|). (3) Accordingly, the bound constitutes the maximal overlap between Eq. ( 2 ) and a separable state. In order to witness entanglement in different base representations, one applies a mode transformation to the target state, thus \mathcal{F}_{\hat{U}}=\textrm{sup}_{\hat{\sigma}}\left[\textrm{tr}\left(\hat{\sigma}\hat{U}|\Psi(\lambda)\rangle\langle\Psi(\lambda)|\hat{U}^{\dagger}\right)\right] . The bound for mode-independent entanglement is then the maximum over all mode transformations \mathcal{F}_{\mathrm{bound}}=\textrm{sup}_{\hat{U}}\left[\mathcal{F}_{\hat{U}}\right(\lambda)] . With Eqs. ( 1 ) and ( 2 ), this bound reads as theo_orig \mathcal{F}_{\mathrm{bound}}(\lambda)=\textrm{max}\left\{\frac{1}{2}+\frac{\sqrt{1+|\lambda|^{2}}}{2+|\lambda|^{2}},1-\frac{1}{+|\lambda|^{2}}\right\}.