Three-Photon and Hybrid Coherent-Fock Interference in a Two-Phase Six-Port Mach-Zehnder Interferometer

AI summary

Overview

Research area: Quantum optics / photonic quantum information — multiport (SU(3)) interference, three-photon bosonic statistics, and hybrid discrete/continuous-variable state engineering.

Technical level: Advanced. The paper is purely theoretical and relies on permanent/DFT algebra, second quantization, and density-matrix formalism, though its conclusions are stated in terms of measurable detection probabilities and mean photon numbers.

Scope (one sentence): An exact analytic treatment of a six-port Mach-Zehnder interferometer built from two cascaded tritters and two independent phase modulators, covering both a three-single-photon Fock input and a hybrid coherent-coherent-Fock input.

What This Paper Is About

Multiport interferometers generalize the Hong-Ou-Mandel beam splitter to three or more paths, but the two-phase version of the six-port Mach-Zehnder interferometer (6p-MZI) had not been analyzed in full: earlier work treated a single phase modulator and did not derive the three-photon Fock case. The authors derive the exact transfer matrix of the two-phase device, sort its nine elements into symmetry classes, and compute complete output statistics for two different input states. The goal is to show how the second, independent phase makes the device a programmable platform rather than a fixed interference circuit.

Key Contributions

1. Exact two-phase transfer matrix and symmetry classification. The matrix $\hat{U}{\rm MZI} = \hat{U}T \hat{U}\phi \hat{U}T$ is derived in closed form (Eq. 5), and its nine elements are shown to collapse into exactly three symmetry triplets ($u{11}=u{23}=u_{32}$; $u_{12}=u_{21}=u_{33}$; $u_{13}=u_{31}=u_{22}$), governed by the discrete Fourier transform structure of the tritter and the conjugate relations $\omega^{}=\omega^{-1}$, $\omega^{2}=\omega$. The matrix is symmetric, consistent with reciprocity.

2. Complete three-photon output distributions over the full 2D phase space. For the input $|1\rangle_1|1\rangle_2|1\rangle_3$, closed-form probabilities are given for the even distribution $P_{[111]}$, the fully bunched NOON-like class $P_{[{300}]}$, and the partially bunched class $P_{[{210}]}$, expressed through a pairwise sum $\mathcal{P}=\cos(3\phi_1)+\cos(3\phi_2)+\cos(3(\phi_1-\phi_2))$ and a cyclic sum $\mathcal{C}=\cos(\phi_1+\phi_2)+\cos(\phi_2-2\phi_1)+\cos(\phi_1-2\phi_2)$.

3. Density-matrix treatment of the hybrid coherent-Fock input. For $|\alpha\rangle_1|\alpha\rangle_2|1\rangle_3$, the output factorizes into three local displacement operators acting on a single-photon superposition, with output coherent amplitudes $\xi_k = (u_{k1}+u_{k2})\alpha$ given explicitly for all three ports.

4. A phase-trajectory-dependent dark-to-bright threshold. The threshold amplitude is not a fixed constant: it sits at $|\alpha|^2 = 2$ along the single-phase cut $\phi_2=0$ but drops to $|\alpha|^2 = 1/2$ in the full two-phase configuration, and the paper maps this threshold surface and the associated amplitude-dependent phase steering at the outer ports.

Main Findings

• Perfect Hong-Ou-Mandel suppression at the first tritter. The partially bunched class vanishes identically at the intermediate stage because its amplitude sums $1+\omega+\omega^{-1}=0$. The surviving intermediate state is $-\frac{1}{\sqrt{3}}|111\rangle + \frac{\sqrt{2}}{3}(|300\rangle+|030\rangle+|003\rangle)$, with a negative sign on $|111\rangle$ reflecting destructive interference from bosonic symmetry.

• Closed-form class probabilities. $P_{[111]} = \frac{7}{27} + \frac{8}{81}\mathcal{P} + \frac{4}{27}\mathcal{C}$; $P_{[{300}]} = \frac{8}{27} + \frac{4}{81}\mathcal{P} - \frac{4}{27}\mathcal{C}$; $P_{[{210}]} = \frac{4}{27}(3-\mathcal{P})$. Normalization holds analytically for all $(\phi_1,\phi_2)$.

• 100% visibility in the single-phase limit. Setting $\phi_2=0$ recovers the single-modulator result: $P_{[111]}$ reaches unity at $\phi_1 = 2m\pi$ (integer $m$), the signature of perfect photon indistinguishability. At $\phi_1=0$, $P_{[111]}=1$ while the other two classes are zero.

• Frequency tripling. $P_{[{210}]}$ carries a pure $\cos(3\phi_1)$ dependence in the single-phase limit, a threefold frequency multiplication relative to a standard two-port Mach-Zehnder interferometer.

• Partial bunching is blind to global phase. $P_{[{210}]}$ depends only on the pairwise sum $\mathcal{P}$, i.e., on $\phi_1$, $\phi_2$, and $\phi_1-\phi_2$; the global phase $\Phi=\phi_1+\phi_2$ drops out entirely. This insensitivity to a uniform phase shift on both arms is invisible in a single-phase treatment.

• Strict complementarity of even and bunched channels. $P_{[{111}]}$ and $P_{[{300}]}$ carry the same cyclic sum $\mathcal{C}$ with opposite signs, so any phase configuration that drives coalescence suppresses the even distribution and vice versa.

• The antisymmetric cut decouples the partial-bunching channel. Setting $\phi_2 = -\phi_1$ makes $\mathcal{C}=0$ identically, leaving $P_{[{210}]}$ as the only phase-dependent observable, fully decoupled from the other two channels. The symmetric cut $\phi_2=\phi_1$ instead reduces the system to a one-dimensional slice.

• No coherent-Fock cross term. The mean photon number at port $k$ is $\langle n_k\rangle = |\xi_k|^2 + |u_{k3}|^2$ — the coherent and single-photon contributions are strictly additive at every $(\phi_1,\phi_2)$. Total photon number is conserved: $\langle n_1\rangle+\langle n_2\rangle+\langle n_3\rangle = 2|\alpha|^2+1$.

• Exact port-exchange identity. $\langle n_3\rangle(\phi_1,\phi_2) = \langle n_1\rangle(\phi_2,\phi_1)$, following from $\omega^{*}=\omega^{-1}$. The two outer ports therefore carry the same two-dimensional interference information related by a fixed coordinate swap, not two independent functions.

• Threshold set by a single common factor. All cosine coefficients in $\langle n_k\rangle$ are built from $(2-4|\alpha|^2)$ up to a port-dependent sign, which changes sign at $|\alpha|^2_{\rm th} = 1/2$, independent of which port is observed. This contrasts with the single-phase case, where only port 2 shows a transition and the critical amplitude is four times larger at $|\alpha|^2 = 2$.

• Threshold behavior along the symmetric cut. $\langle n_2\rangle|_{\phi_2=\phi_1} = \frac{1}{9}[(8|\alpha|^2+5)+(4-8|\alpha|^2)\cos\phi]$, which has twice the modulation depth of the corresponding single-phase expression. Exactly at threshold, $\langle n_2\rangle = 1$ identically and independent of $\phi$.

• Richer profile along the antisymmetric cut. $\langle n_2\rangle|_{\phi_2=-\phi_1}$ contains both a $\cos\phi$ term that inverts sign at threshold and a $\cos(2\phi)$ term that persists unchanged across it, producing a two-frequency interference profile.

• Qualitative reorganization of the 2D landscape. Below threshold the $\langle n_2\rangle$ density map shows diagonal stripes parallel to $\phi_1=\phi_2$; above threshold it reorganizes into a single elliptical bright lobe centered at $(\pi,\pi)$, a reshaping rather than a simple rescaling.

Methodology in Plain English

The authors model the device as three successive optical elements: a tritter, a phase-shifting section, and a second identical tritter. The tritter is an ideal lossless 3×3 mixer whose entries are powers of the cube root of unity, so a single photon entering any port exits any output with probability 1/3. Rather than tracking wavefunctions by hand, they track how photon creation operators transform, which automatically enforces bosonic normalization and lets them read off which output patterns can and cannot occur.

For the three-photon case, they propagate one photon per input port through the first tritter, apply the phase shifts according to how many photons sit in each arm, then propagate through the second tritter. Squaring the resulting amplitudes gives detection probabilities, which they organize into three classes (even, fully bunched, partially bunched) and reduce to compact expressions in two cosine sums. Two special phase settings — symmetric and antisymmetric — are then examined as one-dimensional slices.

For the hybrid case, they exploit a property of linear optics: displacement operators transform into displacement operators, so the two coherent beams simply emerge as three new coherent amplitudes while the single photon splits across the outputs. Because the resulting state is a displaced single-photon state, the mean photon number separates cleanly into a coherent part plus a photon part with no interference cross term. They then locate where the oscillating coefficients change sign and plot the resulting landscapes as density maps and line cuts.

Why This Matters

Impact on research. The paper supplies the exact analytic reference that multiport interferometer experiments currently lack for the two-phase configuration: complete three-photon coincidence statistics in two dimensions, an explicit symmetry classification of the transfer matrix, and a demonstration that a device parameter (the dark-to-bright threshold) can be tuned by choosing a phase trajectory rather than by changing hardware. It also extends the boundary between discrete- and continuous-variable quantum optics, showing how a single injected photon reshapes a classical interference pattern.

Real-world applications (as framed in the paper):

• Coherent amplitude sensing — the amplitude-dependent phase steering at ports 1 and 3 provides a readout whose operating point moves with the phase trajectory.
• Programmable quantum state preparation — the two-phase configuration engineers the interference landscape rather than merely sweeping through it, enabling on-demand routing into the even, fully bunched, or partially bunched classes.
• Quantum key distribution — the dark-to-bright transition is conceptually tied to the intensity discrimination between a macroscopic reference pulse and a sub-single-photon signal in B92 QKD, here generalized to three output modes with a continuously tunable balance point.
• Boson sampling and photonic information processing — exact multi-photon statistics in tritter networks support the classically hard computational primitive that motivates multiport interference.

Industry relevance. The architecture is built from components already realized in integrated waveguides and single-mode fiber, and the directionally unbiased version with two independent modulators has been demonstrated experimentally. That makes the results directly relevant to photonic chip designers who need calibration-free analytic models for programmable multiport meshes, and to sensing/metrology groups looking for tunable operating points rather than fixed 50:50 conditions.

Future Directions

• Experimental confirmation of the threshold surface. The prediction that the critical amplitude falls from $|\alpha|^2=2$ to $|\alpha|^2=1/2$ between the single-phase and two-phase configurations is stated analytically and plotted at $|\alpha|^2=0.3$ and $1.0$, but the paper reports no experimental data.

• Imperfect and asymmetric devices. The analysis assumes ideal, lossless, symmetric tritters; how fabrication asymmetry, propagation loss, and finite indistinguishability shift the symmetry triplets and the threshold is not addressed.

• Extension beyond three photons and three ports. The symmetry-class structure follows from the cube roots of unity, so it is natural to ask how the classification generalizes to larger DFT-based mixers and to higher photon numbers.

• Full characterization of port 3. The article relies on the port-exchange identity to infer port 3 behavior from port 1, but the subsection on amplitude-dependent steering at the outer ports is truncated in the available text, leaving the complete two-parameter steering conditions at port 3 not fully reported here.

Target Audience

Quantum optics theorists and photonic experimentalists working with multiport interferometers, integrated waveguide circuits, or three-photon interference; graduate students who need a worked, fully analytic example of permanent/DFT-based multi-photon statistics; and quantum sensing or QKD engineers interested in tunable discrimination thresholds in SU(3)-based devices. Readers without a background in second quantization and linear-optics transfer matrices will find the derivations heavy, though the final probability and photon-number expressions are directly usable.

Authors’ abstract

We present a unified theoretical analysis of three-photon quantum interference in a Six-Port Mach-Zehnder Interferometer (6p-MZI) constructed from two cascaded tritters, with two independent phase modulators placed between the tritter arms. We analytically derive the transfer matrix of the 6p-MZI and show how they organize into three symmetry classes, governed by the discrete Fourier transform (DFT) structure of the tritter and the conjugate relations. Furthermore, we analyze two input regimes: First, three indistinguishable single photons are injected into the tritter, and the output probability distributions $P_{[111]}$, $P_{[\{300\}]}$, and $P_{[\{210\}]}$ are derived as functions of the two relative phases $(φ_1, φ_2)$. At $φ_2 = 0$, the single-phase limit is recovered, which confirms 100\% visibility of the even-distribution fringe. Second, a hybrid coherent-Fock input $|α\rangle_1|α\rangle_2|1\rangle_3$ is analyzed via the density matrix formalism. The average photon number at each output port exhibits amplitude-dependent phase shifts. Our results establish the 6p-MZI as a programmable platform for tripartite quantum state manipulation and coherent amplitude sensing.