Photon pair antibunching and second-order correlations between pair events

AI summary

Overview

Research area: Quantum optics — nonclassicality criteria and photon-pair (biphoton) source characterization, sitting at the intersection of quantum state theory and experimental photon-correlation measurement.

Technical level: Intermediate. The mathematics is modest (normally ordered moments, Cauchy–Schwarz inequality, Gaussian moment factoring), but the conceptual argument about what counts as a "pair event" requires comfort with second-order coherence and phase-space distributions.

Scope (one sentence): The paper introduces a second-order correlation function built from the biphoton operator $P^\dagger = a^\dagger b^\dagger$, proves that all classical two-mode fields obey $g^{(2)}_{\textrm{pairs}} \geq 1$, and evaluates the quantity for coherent, squeezed-vacuum, and truncated pair sources.

What This Paper Is About

Standard nonclassicality tests — heralded single-mode $g^{(2)}s$ and signal–idler cross-correlation $g^{(2)}{si}$ — describe the statistics of individual modes or of photon numbers across two modes, but not the statistics of the pair-generation process itself. The authors ask whether one can define an antibunching criterion for genuine biphoton emission, and answer by defining $g^{(2)}_{\textrm{pairs}} = \langle (P^\dagger)^2 P^2\rangle / \langle P^\dagger P\rangle^2$ with $P^\dagger = a^\dagger b^\dagger$. The goal is a rigorous, loss-tolerant, experimentally measurable witness of nonclassical pair statistics that needs no phase reference or state tomography.

Key Contributions

1. Definition of a pair-level correlation function. The paper defines $g^{(2)}_{\textrm{pairs}} = \langle (P^\dagger)^2 P^2 \rangle / \langle P^\dagger P \rangle^2$ through the pair operator $P^\dagger = a^\dagger b^\dagger$, with $>1$, $=1$, and $<1$ corresponding to pair bunching, Poissonian pair statistics, and pair antibunching respectively.

2. A classical bound proved from the Cauchy–Schwarz inequality. For any two-mode field with a positive Glauber–Sudarshan $P$-function, the normally ordered moments reduce to moments of a classical pair amplitude $z = \alpha\beta$, so $\langle |z|^4 \rangle \geq \langle |z|^2 \rangle^2$ and therefore $g^{(2)}_{\textrm{pairs}} \geq 1$. Pair antibunching is thus classically forbidden.

3. State-by-state evaluation. The function is computed for a two-mode coherent state (exactly 1), the two-mode squeezed vacuum (an explicit closed form in $\bar{n}$), the single-pair truncated state (exactly 0), and finite-order truncations of the squeezed vacuum, including a general formula for truncation order $N$.

4. An explicit contrast with heralded correlations, including loss behavior. The authors show that $g^{(2)}_{\textrm{pairs}}$ is invariant under uniform linear loss while heralded $g^{(2)}_s$ depends on conditioning efficiency, and they describe a coincidence-counting scheme for measuring it.

Main Findings

• The classical threshold is unity. Values of $g^{(2)}{\textrm{pairs}}$ greater than, equal to, or less than one mean pair bunching, Poissonian pair statistics, and pair antibunching; the bound $g^{(2)}{\textrm{pairs}} \geq 1$ holds for every classical two-mode field and survives arbitrary linear attenuation.

• A product coherent state saturates the bound. For $|\alpha\rangle \otimes |\beta\rangle$, all normally ordered moments factorize and $g^{(2)}_{\textrm{pairs}} = 1$ exactly — the reference point for statistically independent joint detection events.

• The two-mode squeezed vacuum always bunches, however weak. For TMSV with $\lambda = \tanh r$ and $\bar{n} = \sinh^2 r$, the result is $g^{(2)}{\textrm{pairs}} = 4(6\bar{n}^2 + 6\bar{n} + 1)/(2\bar{n}+1)^2$, giving $g^{(2)}{\textrm{pairs}} \to 4$ as $\bar{n} \to 0$ and $g^{(2)}{\textrm{pairs}} \to 6$ as $\bar{n} \to \infty$. The paper states the exact TMSV gives $g^{(2)}{\textrm{pairs}} > 4$ for any nonzero squeezing, no matter how weak.

• Maximal pair antibunching requires truncation. The state $\sqrt{1-p},|0,0\rangle + \sqrt{p},|1,1\rangle$ (the lowest, $N=1$, truncation) has vanishing two-pair probability and therefore $g^{(2)}_{\textrm{pairs}} = 0$, the strongest possible violation of the classical bound.

• Truncation order changes the qualitative behavior. All finite-order truncations with $N \geq 2$ share the weak-squeezing limit $g^{(2)}{\textrm{pairs}} \to 4$. For $N=2$ and $N=3$ the pair correlation crosses over from bunching to antibunching as the state parameter grows, reaching $g^{(2)}{\textrm{pairs}} = 12/25 < 1$ for $N=2$ in the large-squeezing limit (with $\bar{n} \to 1$ there); for $N=4$ the statistics remain bunched over the entire range.

• Heralded antibunching and pair antibunching are logically independent. Because heralded $g^{(2)}s$ post-selects on an idler click while $g^{(2)}{\textrm{pairs}}$ probes the unconditional pair statistics, a state can show antibunching in one and bunching in the other. The paper makes this concrete for TMSV at detection efficiencies $\eta = 1$ and $\eta = 0.5$: the heralded curve can show antibunching in the weak-pumping regime $\bar{n} \ll 1$, while the pair curve does not.

• Uniform loss cancels out. Both numerator and denominator of $g^{(2)}_{\textrm{pairs}}$ scale with detection efficiency, so the curves for $\eta = 1$ and $\eta = 0.5$ overlap; heralded correlations instead depend explicitly on conditioning efficiency.

• Scope caveat stated by the authors. Pair antibunching certifies nonclassicality of the joint two-mode field, not correlated biphoton generation specifically, since product states containing a nonclassical single-mode component can also violate the bound.

Methodology in Plain English

The authors take a normally ordered intensity-correlation formula and replace the single-mode operators with a "pair operator" $P^\dagger = a^\dagger b^\dagger$, so that $\langle P^\dagger P\rangle$ counts coincidences and $\langle (P^\dagger)^2 P^2\rangle$ counts double-pair coincidences, both normalized by the square of the single-pair rate. To establish a classical boundary, they use the fact that any classical two-mode field with a non-negative Glauber–Sudarshan $P$-function can be written as an average over random complex amplitudes $\alpha$ and $\beta$; the pair quantity then becomes a moment of the single random variable $z = \alpha\beta$, and the ordinary Cauchy–Schwarz inequality for random variables forces the ratio to be at least one. They then substitute specific states — a product coherent state, the squeezed vacuum, and manually truncated versions of it — and evaluate the required moments, using Gaussian moment factoring (Wick's theorem) for the squeezed vacuum, whose eight-operator expectation value reduces to products of two-operator correlations. Finally, they translate the abstract operator expression into a measurement recipe: quadruple coincidences for the numerator, squared double coincidences for the denominator.

Why This Matters

The paper supplies a missing rung in the hierarchy of nonclassicality witnesses: single-photon antibunching has long been the standard hallmark, but no analogous criterion existed for the pair-generation process itself. By tying the witness to a threshold that uniform loss cannot move, it offers an experimentally practical diagnostic for multipair suppression in biphoton sources — the regime that limits most photonic quantum technologies.

Real-world applications:

• Heralded single-photon sources — pair antibunching would signal suppressed multi-pair emission, the analogue of photon antibunching in single-photon sources.
• Quantum communication and entanglement distribution — quantitative characterization of how far a source deviates from ideal pair emission.
• Biphoton source metrology — a characterization figure of merit for spontaneous parametric down-conversion and four-wave mixing sources that needs no phase stabilization and no state reconstruction.
• Detector and coincidence-chain validation — the measurement recipe (photon-number-resolving or multiplexed threshold detectors with time-tagged coincidence counting) doubles as a test of fourfold coincidence infrastructure.

Industry relevance: developers of quantum photonic hardware, squeezed-light sources for sensing, and integrated biphoton sources gain a loss-robust, calibration-light benchmark that can be quoted alongside conventional $g^{(2)}$ figures — though the authors do not report any device-level demonstration or measured values.

Future Directions

• Experimental realization. The paper gives the measurement recipe but reports no data; the natural next step is a fourfold-coincidence measurement on a real source, which the authors state is feasible with existing detector technology.

• Finding physical sources that actually antibunch. Since the exact TMSV always bunches and only truncation produces $g^{(2)}_{\textrm{pairs}} < 1$, an open question is which physically realizable, genuinely pair-correlated sources can reach the antibunched regime without simply removing the multipair sector by hand.

• Behavior under non-uniform loss and detector imperfections. The invariance proof covers uniform linear loss; how $g^{(2)}_{\textrm{pairs}}$ behaves under unbalanced loss, unequal detection efficiencies between signal and idler, or dark counts is not treated.

• Relation to higher-order pair correlations and other witnesses. The paper analyzes only the second-order pair function; whether higher-order pair moments, or combinations with $g^{(2)}_{si}$ and heralded $g^{(2)}_s$, sharpen the discrimination between classical and quantum pair statistics remains open.

Target Audience

Quantum optics researchers and graduate students working on nonclassicality criteria, biphoton source characterization, or squeezed-light experiments; experimentalists building coincidence-counting setups who need a rigorous but easily measured witness; and theorists interested in phase-space arguments for operator-level (rather than mode-level) correlation functions. Readers should already be comfortable with second-order coherence, the Glauber–Sudarshan $P$-function, and Gaussian moment factoring.

Authors’ abstract

We introduce the pair second-order correlation function $g_{\textrm{pairs}}^{\left(2\right)}=\left\langle \left(P^{\dagger}\right)^{2}P^{2}\right\rangle /\left\langle P^{\dagger}P\right\rangle ^{2}$, defined through the pair operator $P^{\dagger}=a^{\dagger}b^{\dagger}$, to characterize second-order correlations and pair bunching and antibunching in photon-pair creation processes. This quantity directly probes correlations between pair-generation events within a single two-mode quantum state, providing access to the intrinsic pair-generation process beyond conventional single-mode or heralded second-order correlations, which do not directly capture correlations between pair events. Values of $g_{\textrm{pairs}}^{\left(2\right)}$ greater than, equal to, or less than unity correspond respectively to pair bunching, Poissonian pair statistics, and pair antibunching. Using the Cauchy--Schwarz inequality, we further show that all classical two-mode fields described by a positive Glauber--Sudarshan \ensuremath{P}-function satisfy $g_{\textrm{pairs}}^{\left(2\right)}\geq1$, so that pair antibunching is classically forbidden and constitutes an unambiguous signature of nonclassicality. We evaluate $g_{\textrm{pairs}}^{\left(2\right)}$ for several representative quantum states and show, in particular, that even arbitrarily weak two-mode squeezed vacuum states exhibit pair bunching. Comparison with heralded second-order correlations highlights the complementary information provided by these observables. The proposed correlation function is experimentally accessible via standard coincidence measurements, requires no phase reference or state reconstruction, and remains invariant under uniform loss.