Beam-Splitter State in Quantum Optics

• Beam-splitter state is a quantum state produced by applying a two-mode SU(2) unitary operation on inputs, yielding entangled and coherent outputs.
• They are characterized using binomial decomposition and phase basis methods, quantifying entanglement via Schmidt coefficients and entropy metrics.
• Beam-splitter states underpin applications in entanglement distribution, quantum state engineering, and Bell inequality tests in quantum optics.

A beam-splitter state refers to the quantum state or class of quantum states resulting from the action of a beam splitter—a linear, two-mode SU(2) unitary—on a set of input states, typically Fock, coherent, squeezed, or generalized phase states. The physical and mathematical structure of these states underpins many fundamental phenomena in quantum optics, such as entanglement generation, quantum coherence, Bell-inequality violations, and quantum state engineering. The canonical beam-splitter state forms when a well-defined quantum input, such as a number state, interacts with vacuum at a balanced (50:50) beam splitter, but the notion generalizes to multiport networks and various classes of nonclassical inputs.

1. Beam Splitter Unitary and Mode Transformations

The beam splitter is mathematically represented by a two-mode unitary operator acting on annihilation operators $a$, $b$ as

$$U_{BS}(\theta) \begin{pmatrix} a \ b \end{pmatrix}U_{BS}^\dagger(\theta) = \begin{pmatrix} a_{out} \ b_{out} \end{pmatrix} = \begin{pmatrix} \cos\theta & -\sin\theta \ \sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix} a \ b \end{pmatrix}$$

with transmissivity $T = \cos^2\theta$, reflectivity $R = \sin^2\theta$, and $T+R=1$ (Shringarpure et al., 2019, Gagatsos et al., 2013). A 50:50 beam splitter corresponds to $\theta = \pi/4$.

On the level of quantum states, this implements an SU(2) rotation in the Fock space, preserving the total excitation number. For multiport or temporally multiplexed beam splitters, composite unitaries are constructed by layering such elementary rotations across different input/output mode pairs, as in beam-splitter arrays (Delgado et al., 2023, Okuno et al., 2024).

2. Fock, Coherent, and Phase Basis Structure of Beam-Splitter States

A paradigmatic construction is the evolution of a Fock state $|n\rangle_a$ and vacuum $|0\rangle_b$ through a balanced beam splitter (Shringarpure et al., 2019, Gagatsos et al., 2013): $$U_{BS}|n\rangle_a|0\rangle_b = \sum_{k=0}^n \sqrt{\binom n k}\,2^{-n/2}\,|k\rangle_a|n-k\rangle_b$$ This binomial (Schmidt) decomposition specifies the output in the number basis and directly quantifies bipartite entanglement. In the phase basis, leveraging the conjugacy between number and phase, the same state can be represented as an equal-amplitude superposition with continuously correlated phases in both outputs: $$U_{BS}|n\rangle_a|0\rangle_b = \frac{1}{2\pi}\int_0^{2\pi} d\phi\,e^{-in\phi}|\tfrac{Re^{i\phi}}{\sqrt{2}}\rangle_a |\tfrac{Re^{i\phi}}{\sqrt{2}}\rangle_b$$ where $R = \sqrt{n}$, revealing the state as a generalized, continuous-variable Schrödinger cat state with maximal phase correlations (Shringarpure et al., 2019).

When a single-mode coherent-state superposition (cat state) is injected, the output at a balanced splitter is the entangled "Bell-cat" state: $$|{\Phi^{Bell}_{\alpha,\alpha}}\rangle = \frac{1}{\sqrt{N_{\alpha}^{Bell}}}\Big(|\alpha,\alpha\rangle + |-\alpha,-\alpha\rangle\Big)$$ with $N_{\alpha}^{Bell} = 2(1+e^{-4|\alpha|^2})$ (Slaoui et al., 2023). Beam splitters also transform temporally stable phase states—eigenstates of a generalized phase operator—into entangled outputs, with the resulting entanglement analytically dependent on dimension and temporal parameter (Daoud et al., 2012).

3. Quantification of Entanglement and Coherence

Beam-splitter states are archetypal for bipartite entanglement generation. The binomially correlated output for Fock inputs yields a pure bipartite state with explicit Schmidt coefficients $\sqrt{p_k}$,

$p_k = \binom{n}{k} 2^{-n}$

with entropy of entanglement

$S_E = -\sum_{k=0}^n p_k \log_2 p_k$

approaching $\frac{1}{2}\log_2(\pi e n / 2)$ for large $n$ (Shringarpure et al., 2019, Gagatsos et al., 2013). Negativity is

$$\mathcal{N} = \frac{(\sum_{k=0}^n \sqrt{p_k})^2 - 1}{2}$$

Majorization theory applies: increasing input photon number $k$ monotonically increases entanglement, and the entropy of entanglement is maximized for balanced beam splitters ($\theta = \pi/4$). Infinitesimal majorization holds up to a crossover angle, beyond which output probability vectors become incomparable, but catalysis by ancillary entangled states can restore convertibility (Gagatsos et al., 2013).

Beam splitters universally increase quantum coherence (as measured by the $l_1$-norm and relative entropy of coherence); the maximal gain in coherence is achieved for two-mode squeezed vacuum inputs and balanced splitting (Ares et al., 2022).

4. Bell Inequalities, Nonlocality, and Quantum Randomness

Phase-entangled beam-splitter states enable violation of Bell-type inequalities. For the binomial state generated from $|n\rangle_a|0\rangle_b$, nonlocal correlations are demonstrated using Mach–Zehnder interferometers with nonlinearity (Kerr media), phase shifters, single-photon detection, and homodyne-postselected measurement. The correlations appear in the form

$$P(x_{1M}, x_{2M}; \theta_A, \theta_B) \propto 1 + y \cos(\theta_A + \theta_B)$$

and the CHSH parameter $|S|$ exceeds the local realism bound for all $n\geq 1$, reaching the Tsirelson bound $2\sqrt{2}$ (Shringarpure et al., 2019). Crucially, fair sampling is not needed if the homodyne outcome is obtained before setting the phase, eliminating a common loophole in nonlocality tests.

For classical, fully mixed initial states propagating through networks of sequential beam splitters with appropriately chosen reflectivities, one can purify a maximally mixed Fock-diagonal state into a pure NOON state—a maximally entangled path state—without postselection, deterministically (V. et al., 2020).

5. Multimode, Multiphysics, and Programmable Generalizations

Beam-splitter states are not limited to spatially two-mode, temporally static, or lossless scenarios. Multiport generalizations (e.g., tritters and larger interferometric arrays) extend the formalism; in three-mode systems, appropriate input state preparation and postselection enables the generation of multipartite entangled states (e.g., W, GHZ, and G states), as demonstrated experimentally by fidelity up to $87.3\%$ (Kumar et al., 2023).

The formal multimode theory describes beam splitters as mode-mixing unitaries parametrized by physical quantities such as aperture geometry, enabling diffraction and spatial masking to produce entangled states, number-path entanglement, and path-interference phenomena analogous to the Hong-Ou-Mandel effect (Xiao et al., 2017). Programmable time-domain beam splitters implement dynamically controlled SU(2) coupling between modes, allowing non-Gaussian state storage and manipulation with high coherence and controllable Wigner negativity (Okuno et al., 2024).

Novel conditional beam splitters, such as those capable of separating bright (symmetric) and dark (antisymmetric) field components via atom–cavity interactions, further extend the class of attainable beam-splitter states, enabling programmable operations based on collective modes and atomic state control (Solak et al., 2024).

6. Null Entanglement and Special Input States

Despite the universality of entanglement generation, specific input state classes are mapped by beam splitters to separable states. States with identical single-mode squeezing magnitudes and phase relations, unpolarized two-mode mixtures, and convex mixtures of such Gaussian product states remain separable at the output for any splitting ratio (Goldberg et al., 2017). This underlines that nonclassicality is necessary but not sufficient for output entanglement: entanglement generation depends sensitively on input state structure and phase relationships.

7. Experimental Significance and Applications

Beam-splitter states underpin quantum technological protocols including entanglement distribution, state engineering, quantum metrology, and foundational tests of quantum mechanics. Entanglement generated via beam splitters enables dense coding, teleportation, and continuous-variable quantum information tasks. Hybrid protocols distribute entanglement by acting on nonentangled yet classically correlated states, which is operationally significant for robust quantum communication (Croal et al., 2015).

Sensitivity to loss is a consistent limitation for macroscopic superpositions; protocols involving pre-attenuation and re-amplification, as well as correct measurement ordering, are required to retain Bell violations and minimize decoherence (Shringarpure et al., 2019).

---

References:

(Shringarpure et al., 2019, Gagatsos et al., 2013, Kumar et al., 2023, Slaoui et al., 2023, Daoud et al., 2012, Ares et al., 2022, V. et al., 2020, Goldberg et al., 2017, Croal et al., 2015, Xiao et al., 2017, Okuno et al., 2024, Solak et al., 2024)