arXiv:2610.28020
Enhanced orbital-angular-momentum dimensionality and spatial-spectral decoupling in thin-film SPDC
Dmitrii Akatev, Jakub Lewandowski, Maria Chekhova
Abstract
Thin-film spontaneous parametric down-conversion, due to the relaxed phase-matching condition, generates photon pairs that are extremely broadband in frequency and angle and highly entangled in time/frequency and position/momentum. Here we show that it also provides a very broad orbital angular momentum (OAM) spectrum, which, in contrast to the case of bulk crystals, is decoupled from the frequency spectrum. To demonstrate it experimentally, we generate photon pairs in a $7~μ\mathrm{m}$-thick lithium niobate film and compare them with those from a $500~μ\mathrm{m}$-thick periodically poled lithium niobate crystal under identical pump conditions. The thin-film source exhibits a substantially broader OAM spectrum, with the azimuthal Schmidt number almost 9 times larger than that of the bulk source. On this highly entangled state, we perform quantum-state tomography in selected two- and four-dimensional OAM subspaces, demonstrating high-purity maximally entangled states. These results highlight thin nonlinear films as a promising platform for high-dimensional quantum-state generation with weak spatial--spectral coupling.
AI Summary
The paper asks whether relaxed longitudinal phase matching in thin-film SPDC enables broader orbital-angular-momentum (OAM) distributions and weaker spatial–spectral coupling than bulk SPDC. It compares a $7\,\mu\mathrm{m}$ lithium-niobate film with a $500\,\mu\mathrm{m}$ periodically poled lithium-niobate crystal, using the same 775-nm continuous-wave pump and $37.5\,\mu\mathrm{m}$ pump waist, with collection optics optimized separately. Calculations predict azimuthal participation numbers $K_{\mathrm{az}}=1/\sum_\ell P_\ell^2$ of 59.4 and 6.66, respectively: an 8.9-fold increase for the film. Phase-flattening coincidence measurements, corrected for OAM-dependent detection efficiency, show broader OAM populations consistent with these predictions. Measurements reach only $|\ell|\leq12$ because SLM diffraction efficiency decreases at higher orders; the full predicted dimensionality is therefore not directly measured. Tomography establishes coherence within selected subspaces: Bell-state fidelities exceed 0.98 for $\ell=\pm1,\pm3,\pm7$, while a four-dimensional state has fidelity $0.916\pm0.014$ and purity $0.956\pm0.025$. Frequency-resolved calculations predict an approximately frequency-independent OAM distribution for the film, unlike the bulk source. Time-of-flight spectroscopy using 3 km of dispersive fiber finds nearly unchanged spectra for OAM orders up to $|\ell|=10$, providing an experimental indication of weak spatial–spectral coupling rather than a full characterization of spatial–spectral separability. The accessible spectral range is narrow relative to the source bandwidth, limited by weak fiber dispersion near 1310 nm and SLM operation up to 1600 nm. Mode-dependent diffraction and coupling, cross-talk, and counting fluctuations are identified as limitations on reconstruction quality.
Central claim may not hold: The reported K_az = 59.4 and almost ninefold enhancement are model-derived values supported by a limited measurement window, rather than direct reconstructions of the full experimental OAM distributions. Experimental evidence for weak OAM–frequency coupling covers only a limited spectral range, while broadband decoupling rests on the theoretical model. Frequency-independent OAM probabilities do not establish spatial–spectral quantum-state factorization because frequency-dependent phases and radial correlations can remain undetected. The marginal OAM inverse participation ratio does not by itself certify approximately 59-dimensional OAM entanglement after radial modes are traced out.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 10 pages, 5 figures, 1 table