Weak-signal conversion from 1550nm to 532nm with 84% efficiency
Aiko Samblowski, Christina E. Vollmer, Christoph Baune, Jaromir Fiurasek, Roman Schnabel
WWeak-signal conversion from 1550 nm to 532 nm with 84% efficiency
Aiko Samblowski, Christina E. Vollmer, Christoph Baune, Jarom´ır Fiur´aˇsek, andRoman Schnabel , ∗ Institut f¨ur Gravitationsphysik, Leibniz Universit¨at Hannover and Max-Planck-Institut f¨ur Gravitationsphysik(Albert-Einstein-Institut), Callinstr. 38, 30167 Hannover, Germany Department of Optics, Palack´y University, 17. listopadu 12, 77146 Olomouc, Czech Republic ∗ Corresponding author: [email protected]
Compiled November 2, 2018We report on the experimental frequency conversion of a dim, coherent continuous-wave light field from 1550 nmto 532 nm with an external photon-number conversion efficiency of (84.4 ± (cid:13) OCIS codes:
The efficient frequency conversion of photons is an im-portant task in photonics to shift light fields into thevisible wavelength regime where high quantum efficiencysingle-photon detectors are available [1, 2]. In princi-ple an arbitrarily dim propagating light field can beup-converted to another frequency with an efficiency of100% preserving the photon number and its statistics [3].The required energy needs to be supplied in terms of anintense pump field. The dim signal field and the pumpfield need to be overlapped inside a phase-matched non-linear crystal, possibly supported by a cavity that is res-onant for one or several of the wavelengths involved.For intense continuous-wave (cw) light we realizedsecond-harmonic generation (SHG) with an externalconversion efficiency of up to 95% taking into accountall optical losses [4]. The SHG process, however, is notsuitable for the frequency up-conversion of dim fields,i.e. for the frequency up-conversion of arbitrarily weaksignals.In this Letter, we report on a highly efficient frequencyconversion of a dim coherent light field from 1550 nm to532 nm in the cw regime using (non-degenerate) sum-frequency generation. The overall conversion efficiencywas measured to be (84 . ± . spatial mode profile andto suppress technical noise. The MC cavity finesse was F = 560, corresponding to a linewidth of 1.3 MHz. Thecavity length was controlled using the Pound-Drever-Hall (PDH) locking scheme [5] with a phase modulationat a sideband frequency of 29.5 MHz.To provide the signal field at 1550 nm and the pumpfield at 810 nm, a non-degenerate, doubly resonant op- BS DBSPBSPD PD input @1550 nm output @532 nmPD PD PD SFGpump @ 810 nm λ /2 Fig. 1. (Color online) Schematic of the core setup. SFG:sum-frequency generator; (P)BS: (polarizing) beamsplitter; PD: photo diode; λ/
2: waveplate for power vari-ation. The generation of the intense pump field at 810 nmand the generation of the weak signal field at 1550 nmare not shown. Details are provided in the main text.tical parametric oscillator (OPO) was pumped with thefiltered 532 nm light. The OPO was built as a mono-lithic standing-wave non-linear cavity and was operatedabove threshold. The non-linear medium inside the OPOcavity was a periodically poled potassium titanyl phos-phate (PPKTP) crystal. The phase matching for 532,810 and 1550 nm was achieved at a temperature of 68 ◦ C,to which the crystal was stabilized actively. The lengthof the crystal was 8.9 mm and the coatings were chosento form a cavity with a finesse of F = 100 for the twinbeams at 810 nm and 1550 nm. Hence, the linewidth ofboth modes was 91 MHz. The radii of curvature of 8 mmled to a waist size of 24 µ m for the green pump beam,that simply double-passed the crystal. The thresholdpower was about 70 mW. The bright output fields wereco-propagating and spatially separated with a dichroicbeam splitter.The pump field at 810 nm with a tunable power of upto 190 mW was mode-matched into the sum-frequency1 a r X i v : . [ qu a n t - ph ] O c t enerator (SFG), see Fig. 1. The SFG was built asa standing-wave two-mirror non-linear cavity contain-ing another PPKTP crystal. The quasi phase matchingfor 810 and 1550 nm was achieved at a temperature of67 ◦ C, to which the crystal was also stabilized actively.The length of the crystal was 9.3 mm and the coatingswere chosen to be
R > .
9% on the right side and R = (96 . ± . µ m for the pump beam. The coatings forthe converted light at 532 nm are chosen to be R < . R > .
9% on the left side toensure that the converted light leaves the SFG to theright. The length of the cavity was actively controlledwith the PDH scheme, using a frequency modulation ofthe pump light at 24.5 MHz. The losses of the systemare mainly determined by the anti-reflective coatings ofthe crystal surface and by the absorption of the crystalitself. They can be combined in a total absorption of thecrystal, that was measured to be α = 0 . / cm and α = 0 . / cm.A 4 mW signal beam at 1550 nm (red) is sent througha 50/50 beam splitter (BS) to monitor the input powerwith the photo detector PD , in , see Fig. 1. The re-maining 2 mW incident power is overlapped with thepump beam at a dichroic beam splitter (DBS) and cou-pled into the SFG. Half of the reflected light is detectedat PD , refl . In transmission of the SFG, the opti-cal fields are separated by a set of DBSs and detectedwith the corresponding photo detectors PD , trans ,PD , trans and PD , trans . All signals from the photodetectors are recorded with a data acquisition system( PCI-6133 from
National Instruments ) and analyzedwith PC software (
Labview ).To characterize the conversion efficiency, the lightpowers of the optical fields were measured with the photodetectors mentioned above, which were calibrated withpower meters. The conversion efficiency was calculatedas the ratio of the converted photons at 532 nm and theinitial photons at 1550 nm η = QE QE · n n = γ · · P · P . (1)Here, γ denotes a correction factor for the quantum effi-ciency (QE) of the photo detectors and the power meters.To determine the correction factor γ , the depletion ofthe signal field at 1550 nm was measured in reflectionand in transmission of the cavity, respectively. When nopump light was coupled into the SFG, the light reflectedby the cavity far from its resonance, corresponding tothe total incident power, and the light transmitted bythe cavity on resonance were measured and normalized C on v e r s i on e ff i c i en cy η R e l a t i v e dep l e t i on δ a t n m Pump power in mW (810 nm)Relative depletion δ Conversion efficiency η Fig. 2. (Color online) Measurement results. The conver-sion efficiency (blue) and relative depletion (yellow) areshown over the pump power. The conversion efficiencyreaches its maximum of (84.4 ± δ = 1 − (cid:18) P refl + P trans P in (cid:19) = 1 − (cid:32) P refl P refl , max + P trans , max P refl , max (cid:124) (cid:123)(cid:122) (cid:125) κ · P trans P trans , max (cid:33) (2)depends on the normalized signals and on the ratio of themaximal transmitted and reflected power κ . In low losssystems, the relative depletion is a measure for the con-version efficiency [4] and Eq. (2) yields the same resultsas Eq. (1). Due to the finesse of F = 150 and to thetotal loss α = 0 . / cm, the two differ. However, therelative depletion provides additional data to determinethe parameters of our theoretical model more precisely.In particular the correction factor γ from Eq. (1) couldbe obtained accurately. Thus, both measurements arerequired to obtain an accurate value for the conversionefficiency.To measure the conversion efficiency the pump powerwas varied. For every pump power we took time seriesof each photo detector over a couple of seconds and an-alyzed them in a Labview script. The time series did notvary and the system was stable over hours.Figure 2 shows the results of our measurement. Theconversion efficiency (blue) and the relative depletion(yellow) are plotted against the pump power. The solidlines depict numerical simulations of our system, includ-ing mirror reflectivities, the non-linearity, losses and thephase mismatching as parameters. The simulations arein excellent agreement with our measurements and sup-port the experimentally obtained conversion efficiencyof (84.4 ± γ = 1 . ± · − , which waswell within the specified error range of the power me-ters. Furthermore, our simulation showed that it shouldbe possible to improve the conversion efficiency of oursetup to about 93% by reducing the mirror reflectivityat the signal wavelength from R = 96 .
5% to R = 90%.Since sum-frequency generation maintains the quan-tum properties of a state [3], our device will be suitableto reach a high fidelity if a non-classical state of lightis used as an input field. Squeezed vacuum stateswith a noise suppression of more than 12 dB havebeen demonstrated at a wavelength of 1550 nm in thecontinuous-wave regime [7]. In contrast, comparativelysmall squeezing factors were achieved at visible wave-lengths so far [8]. Assuming that the main optical lossis given by the imperfect conversion efficiency andassuming that the detection efficiency at 532 nm will bethe same as at 1550 nm, non-degenerate sum frequencygeneration might be a feasible approach for produc-ing strongly squeezed states of light in the visible regime.This work was supported by the Deutsche Forschungs-gemeinschaft (DFG), Project No. SCHN 757/4-1, by theCentre for Quantum Engineering and Space-Time Re-search (QUEST) and by the International Max PlanckResearch School for Gravitational Wave Astronomy(IMPRS-GW). J.F. acknowledges support from the Eu-ropean Social Fund and MSMT under project No.EE2.3.20.0060. References
1. M. A. Albota and F. N. Wong, Optics Letters , 1449(2004).2. A. P. VanDevender and P. G. Kwiat, Journal of the Op-tical Society of America B , 295 (2007).3. P. Kumar, Optics Letters , 1476 (1990).4. S. Ast, R. M. Nia, A. Sch¨onbeck, N. Lastzka, J. Stein-lechner, T. Eberle, M. Mehmet, S. Steinlechner, andR. Schnabel, Optics Letters , 3467 (2011).5. E. D. Black, American Journal of Physics , 79 (2001).6. N. Lastzka, “Numerical modelling of classical and quan-tum effects in non-linear optical systems”, Dissertation,Gottfried Wilhelm Leibniz Universit¨at Hannover (2010).7. M. Mehmet, S. Ast, T. Eberle, S. Steinlechner,H. Vahlbruch, and R. Schnabel, Optics Express , 25763 (2011).8. H. Tsuchida, Optics Letters , 2240 (1995)., 2240 (1995).