Design blueprint derived from the source paper (DOI 10.5281/zenodo.24159577): Converting a single photon from one wavelength to another in a periodically poled waveguide is governed by a sinusoidal exchange with the pump field. The first maximum of that exchange sets the pump power and length a design should aim for; overshooting it converts the photon back. This study computes the quantities that follow from stated assumptions for a periodically poled lithium niobate waveguide converting 780 nm to 1550 nm, normalised efficiency 100 percent per watt per square centimetre in the stated convention, using the relations set out in the Methods. The headline results are pump full at 3cm W = 0.2742, pump full at 1cm W = 2.467, pump full at 5cm W = 0.0987, conversion detuned dkL 2 = 0.7081, conversion on phase match = 1. The central question is stated as follows: For a stated normalised efficiency, what pump power reaches full conversion in waveguides of 1 to 5 cm, and how much conversion remains when the phase matching is detuned by a stated amount? The answer obtained here is conditional on the assumptions listed in the paper and in the deposited calculation; changing an assumption changes the numbers in the way the sensitivity discussion describes. Waveguide converters report internal conversion from tens of percent to near unity, with external values reduced by coupling and filtering losses. The work is a computation, not a laboratory measurement, and it is labelled accordingly. Where the computed result is unfavourable to the device or effect under study, that outcome is reported in the same terms as a favourable one, because a bound that only appears when it is welcome is not a bound at all. All numerical statements in this abstract are reproduced by the deposited script from the stated inputs. Status: DESIGN BLUEPRINT — SIMULATED, NOT MEASURED — NOT BUILT. All numbers are simulated values from the source paper's computation, deposited with this blueprint as compute.py and results.json. AI-use disclosure: Dave, an AI research system, assisted with the computation and drafting under the author's direction; author of record William Alexander Aldred.
Design blueprint derived from the source paper (DOI 10.5281/zenodo.24159577): Converting a single photon from one wavelength to another in a periodically poled waveguide is governed by a sinusoidal exchange with the pump field. The first maximum of that exchange sets the pump power and length a design should aim for; overshooting it converts the photon back. This study computes the quantities that follow from stated assumptions for a periodically poled lithium niobate waveguide converting 780 nm to 1550 nm, normalised efficiency 100 percent per watt per square centimetre in the stated convention, using the relations set out in the Methods. The headline results are pump full at 3cm W = 0.2742, pump full at 1cm W = 2.467, pump full at 5cm W = 0.0987, conversion detuned dkL 2 = 0.7081, conversion on phase match = 1. The central question is stated as follows: For a stated normalised efficiency, what pump power reaches full conversion in waveguides of 1 to 5 cm, and how much conversion remains when the phase matching is detuned by a stated amount? The answer obtained here is conditional on the assumptions listed in the paper and in the deposited calculation; changing an assumption changes the numbers in the way the sensitivity discussion describes. Waveguide converters report internal conversion from tens of percent to near unity, with external values reduced by coupling and filtering losses. The work is a computation, not a laboratory measurement, and it is labelled accordingly. Where the computed result is unfavourable to the device or effect under study, that outcome is reported in the same terms as a favourable one, because a bound that only appears when it is welcome is not a bound at all. All numerical statements in this abstract are reproduced by the deposited script from the stated inputs. Status: DESIGN BLUEPRINT — SIMULATED, NOT MEASURED — NOT BUILT. All numbers are simulated values from the source paper's computation, deposited with this blueprint as compute.py and results.json. AI-use disclosure: Dave, an AI research system, assisted with the computation and drafting under the author's direction; author of record William Alexander Aldred.
Converting a single photon from one wavelength to another in a periodically poled waveguide is governed by a sinusoidal exchange with the pump field. The first maximum of that exchange sets the pump power and length a design should aim for; overshooting it converts the photon back. This study computes the quantities that follow from stated assumptions for a periodically poled lithium niobate waveguide converting 780 nm to 1550 nm, normalised efficiency 100 percent per watt per square centimetre in the stated convention, using the relations set out in the Methods. The headline results are pump full at 3cm W = 0.2742, pump full at 1cm W = 2.467, pump full at 5cm W = 0.0987, conversion detuned dkL 2 = 0.7081, conversion on phase match = 1. The central question is stated as follows: For a stated normalised efficiency, what pump power reaches full conversion in waveguides of 1 to 5 cm, and how much conversion remains when the phase matching is detuned by a stated amount? The answer obtained here is conditional on the assumptions listed in the paper and in the deposited calculation; changing an assumption changes the numbers in the way the sensitivity discussion describes. Waveguide converters report internal conversion from tens of percent to near unity, with external values reduced by coupling and filtering losses. The work is a computation, not a laboratory measurement, and it is labelled accordingly. Where the computed result is unfavourable to the device or effect under study, that outcome is reported in the same terms as a favourable one, because a bound that only appears when it is welcome is not a bound at all. All numerical statements in this abstract are reproduced by the deposited script from the stated inputs.
Design blueprint derived from the source paper (DOI 10.5281/zenodo.24159701): On a photonic chip, a ring resonator stores light for a time set by its circumference and its propagation loss. That storage time, together with the free spectral range, decides whether the ring can act as a filter, a delay, or a photon-pair source enhancement. This study computes the quantities that follow from stated assumptions for silicon nitride waveguide rings, group index 2.0, wavelength 1550 nm, propagation loss 0.1, 0.5 and 1.0 dB per centimetre, using the relations set out in the Methods. The headline results are Q at 0.1 dBcm = 3.521e+06, lifetime at 0.1 dBcm s = 2.897e-09, Q at 1 dBcm = 3.521e+05, lifetime at 1 dBcm s = 2.897e-10, FSR at 100 um Hz = 2.386e+11. The central question is stated as follows: For silicon nitride rings of radius 20 to 200 micrometres with stated propagation losses, what quality factor, photon lifetime and free spectral range follow, and which combination reaches a lifetime of one nanosecond? The answer obtained here is conditional on the assumptions listed in the paper and in the deposited calculation; changing an assumption changes the numbers in the way the sensitivity discussion describes. Integrated rings are reported with quality factors from ten thousand to above ten million depending on material and loss, so the computed table spans the practical interval. The work is a computation, not a laboratory measurement, and it is labelled accordingly. Where the computed result is unfavourable to the device or effect under study, that outcome is reported in the same terms as a favourable one, because a bound that only appears when it is welcome is not a bound at all. All numerical statements in this abstract are reproduced by the deposited script from the stated inputs. Status: DESIGN BLUEPRINT — SIMULATED, NOT MEASURED — NOT BUILT. All numbers are simulated values from the source paper's computation, deposited with this blueprint as compute.py and results.json. AI-use disclosure: Dave, an AI research system, assisted with the computation and drafting under the author's direction; author of record William Alexander Aldred.
The principle of all-optical wavelength conversion based on the cascaded sum and difference frequency generation (SFG+DFG) in quasi-phase-matched (QPM) periodically poled LiNbO3 (PPLN) waveguides is introduced. The conversion efficiency of the wavelength converter is calculated. The influences of the powers and span of two pumps on the conversion efficiency are analyzed. Bandwidth of one pump wavelength is discussed. The tuning is about 0.5 nm. The signal pulse width is compressed about 0.68 times.
Design blueprint derived from the source paper (DOI 10.5281/zenodo.24159685): An off-resonant Raman memory trades optical depth and control energy for bandwidth: it can accept short photons that a narrowband memory cannot, at the price of a noise floor set by spontaneous scattering from the control field. This study computes the quantities that follow from stated assumptions for a warm caesium vapour Raman memory, photon duration 1 ns, detuning 1 to 5 GHz, control energies spanning an order of magnitude, using the relations set out in the Methods. The headline results are efficiency E4 D2 = 0.8347, noise E4 D2 = 0.08, efficiency E10 D5 = 0.5132, noise E10 D5 = 5.120e-03, bandwidth E4 D2 GHz = 1.5. Further computed values are operating points meeting both = 1. The central question is stated as follows: For stated control pulse energies and detunings, what acceptance bandwidth, storage efficiency and noise photons per pulse follow, and is there an operating point with efficiency above one half and noise below one hundredth of a photon? The answer obtained here is conditional on the assumptions listed in the paper and in the deposited calculation; changing an assumption changes the numbers in the way the sensitivity discussion describes. Raman memories have demonstrated gigahertz bandwidths with efficiencies of tens of percent and noise levels that remain the main obstacle to single-photon operation. The work is a computation, not a laboratory measurement, and it is labelled accordingly. Where the computed result is unfavourable to the device or effect under study, that outcome is reported in the same terms as a favourable one, because a bound that only appears when it is welcome is not a bound at all. All numerical statements in this abstract are reproduced by the deposited script from the stated inputs. Status: DESIGN BLUEPRINT — SIMULATED, NOT MEASURED — NOT BUILT. All numbers are simulated values from the source paper's computation, deposited with this blueprint as compute.py and results.json. AI-use disclosure: Dave, an AI research system, assisted with the computation and drafting under the author's direction; author of record William Alexander Aldred.
In this paper we measure the efficiency of photon pair production from parametric downconversion at the photon counting level, generated in a laboratory prototype of PPLN waveguide at 1572 nm. The aim is to enhance the brightness of an heralded single photon source and to improve compactness for quantum communication applications, namely quantum key distribution. We observed a maximum of the conversion efficiency corresponding to 7.9 106 photon pairs/s for a pump power of 28 μW. This value is compared with the measured pairs production in a classical experiment of difference frequency generation, realized with the same sample by analogue classical detection. We highlight the advantages of using the waveguide versus a bulk PPLN for such applications and we foresee further possible improvements by adopting waveguide in a single photon level experimental setup.
For the single-pass and double-pass wavelength converters based on cascaded sum and difference frequency generation in quasi-phase matched lithium niobate waveguide, we present the criteria for a choice of the waveguide length and the assignment of pump powers to achieve the desired efficiency, ripple and bandwidth with a large pump wavelength difference of 75 nm. It is found that using the double-pass structure with a 95% reflectivity for the sum frequency wave, the mean efficiency is increased at least 5 dB compared to that of the single-pass one with the same length and input pump powers and the bandwidth is large enough to cover the entire conventional telecommunications C-band and long-wavelength L-band. To flatten the excessive response fluctuation of the double-pass scheme, we propose tuning one of the pumps to longer wavelengths incurring a small penalty of a slight reduction in the mean efficiency which can be compensated by increasing the input pump powers. We also show that to achieve the same amount of flatness, double-pass converters have a smaller efficiency penalty compared to the single-pass converters.