Abstract: We developed a quantum theory for degenerate $\chi^{(3)}$ two-photon state generated from optical fiber, and compared the theory predictions with an experimental result which exhibits a Hong-Ou-Mandel dip visibility of around 94%. Excellent agreement between theory and experiment has been achieved, and we attribute the missing 6% visibility mainly to spatial mode mismatch between signal and idler photons at the beamsplitter.
Introduction
In the field of quantum optics, the generation of identical photon pairs is a cornerstone for advancing quantum computing and secure communication protocols. A prevalent method for producing these pairs is through four-wave mixing (FWM) within optical fibers, frequently utilizing a Sagnac-loop configuration to generate "degenerate" photons—those sharing the same central frequency.
In an ideal scenario, when two perfectly identical photons arrive simultaneously at a 50/50 beam splitter, they undergo quantum interference and always exit together through the same output port. This phenomenon, known as the Hong-Ou-Mandel (HOM) effect, results in a characteristic "dip" in coincidence counts between two detectors. Theoretically, this dip should reach zero, representing 100% visibility. However, experimental results often plateau at approximately 94% visibility. This research provides a comprehensive quantum mechanical model to bridge this gap and identify the physical mechanisms responsible for the missing 6% visibility.
Experimental Setup and Methodology
The theoretical framework begins with the interaction Hamiltonian for four-wave mixing in a $\chi^{(3)}$ nonlinear medium. To validate the theory, the researchers employed a Sagnac interferometer setup designed to produce high-purity photon pairs.
Key aspects of the methodology include:
- First-Order Perturbation Theory: Utilized to calculate the state vector of the two-photon state at the fiber output.
- Dispersion Modeling: A Taylor expansion of the wave-vector ($k$) up to the second-order dispersion term ($\beta_2$) was used to account for the group velocity dispersion within the 300-meter fiber.
- Filter Modeling: Because experimental setups rely on Optical Bandpass Filters (OBPFs) to isolate signal photons, the authors derived analytical expressions for coincidence rates using both Gaussian and super-Gaussian filter profiles.
- Spatial Mode Analysis: To investigate visibility degradation, Fourier-optic calculations modeled how small angular mismatches between photon paths at the final beam splitter affect spatial mode overlap.
Key Results and Theoretical Agreement
The study achieved a high degree of correlation between the developed quantum theory and experimental data. The experimental coincidence counts, measured over 60-second intervals at varying time delays ($\delta\tau$), were compared against theoretical curves.
The experimental HOM dip visibility was measured at approximately 94.3%. As shown in the figure above, the theoretical models for both Gaussian and super-Gaussian filters provided an excellent fit for the experimental data points. While dispersion influences the specific shape and width of the dip, the analysis confirmed that higher-order dispersion is not the primary factor limiting maximum visibility in this system.
Identifying the Source of Visibility Loss
The most significant insight from this work is the extreme sensitivity of the system to spatial alignment. By solving the overlap integral for the photon modes, the authors demonstrated that a minute angular mismatch of just 30 microradians ($3 \times 10^{-5}$ rad) is sufficient to reduce visibility from the theoretical 100% to the observed 94.3%. This almost imperceptible tilt in the beam splitter alignment has a far greater impact on quantum interference than spectral asymmetries or filter imperfections.
Conclusion
This research provides a robust roadmap for improving fiber-based quantum sources. By attributing the majority of visibility loss to spatial mode mismatch rather than fundamental spectral issues, the study suggests that future improvements in quantum interference experiments should prioritize ultra-precise spatial alignment and spatial mode cleaning. While other noise sources like Raman scattering exist, the spatial overlap remains the dominant hurdle in achieving perfect HOM visibility in $\chi^{(3)}$ systems.
Summary updated