deletions | additions       diff --git a/Intro.tex b/Intro.tex  index bee7edf..19201c6 100644  --- a/Intro.tex  +++ b/Intro.tex 
...
Both χ(2) and χ(3) implementation have to grapple with technical noise, usually generated by the strong pump driving the nonlinear process. In the case of the former implementation conversion efficiency close to 100% with limited residual noise is routinely obtained, however, for the latter, conversion at unitary efficiency was only obtained at the expenses of a signal polluted by large Raman noise [Clark2013a].   In this work, thanks to an optimal choice of medium and operating wavelength, we show Four Wave Mixing Bragg Scattering (FWM-BS) setup that performs frequency translation at very low noise regime, with almost unity efficiency on weak coherent states and single photons alike [Cleo2015?].  2. Four Wave Mixing Bragg Scattering  Fig. 1. Cartoon of FWM-BS. A) Energy level depiction. B) Dispersion picture.  FWM-BS interactions are driven by two strong pump fields, whose frequency separation Δω define the amount of frequency translation between signal and idler (see fig. 1) []. The efficiency of the translation process is , where is the coupling strength term, the scattering wavevector, and γ, L and P respectively the medium nonlinearity, the interaction length and the pump power. is the phasematching term: it is convenient to describe all fields in respect of the zero dispersion frequency of the nonlinear medium (i.e. and ), and to introduce the average pump frequency and the signal offset so that and (see figure 1b). The phasematching now reads    and simplifies to     In the straightforward case of = 0, the term (as well as all the other odd terms) cancels out, leaving only smaller contribution from higher order dispersion.  In the approximation we obtain the simple expression  (2)  in which we can identify the process acceptance-bandwidth and the frequency shift of the optimal conversion from the symmetry point due to higher-order dispersion [Machin10].  One prominent feature of FWM-BS, already noticed in [Inoue94,Marhic96] is highlighted by equation (2), that is translation for any given pair of signal and idler frequency can be exacly phasematched by choosing the appropriate pumps: this gives the flexibility of tuning the parameters of the interaction without the modifing the dispersion of the nonlinear medium.  3. Experimental Setup  An optimal choice of nonlinear medium and operating regime enables BS frequency translation in the quantum regime. There are two major design decisions that affect the choice of medium. First, one must fulfill the phase matching for a given set of pumps and signals. In fact, given the selectivity of the process, a very specific dispersion profile must be used for any given configuration. The second design decision is the balance between nonlinearity and dispersion properties: an optimal ratio between the amount of pump power required for full conversion and the acceptance bandwidth, as well as the minimum frequency separation for which cascaded BS becomes prominent. In order to maximize the former with respect to the latter, it is convenient to operate at large dispersion values . As a reference, the demonstrations reported above showed translation over Δ ω > 10 nm.  As mentioned, one obstacle for most quantum implementation of BS is technical noise, in the form of spontaneous FWM and Raman noise [lefrancois2015]. Modulation instability (purple in figure 2) is a competing FWM process, its gain profile given by where and is the linear phase mismatch for a frequency detuning from the pump. Modulation instability affects BS in multiple ways, either by depleting the pump and generating spurious sidebands, or amplifying the vacuum fluctuations and generating pairs of energy correlated photons. It can be managed placing the pump on the normal dispersion, where parametric gain is forbidden and the bandwidth for pair generation is minimized, as well as detuning the signal far from the pump frequencies.  Spontaneous and Stimulated Raman Scattering are processes that couple light with the thermal phonons of the medium. Scattering strength depends on the density of occupied states as well as the Raman spectrum . The probability of a spontaneous Stokes scattering (i.e. the photon losing energy) is given by while for anti-Stokes it is   The spectrum depends on the material: amorphous materials have a broad spectrum (i.e. for glass it extends to about 40 THz [Stolen1973]), while crystalline materials have strong, sharp features. Both processes become less probable for very large detuning, though the anti-Stoke probability depends exponentially on the temperature when (6 THz for room temperature), so that cooling the fiber can further reduce the noise by several order of magnitude (red 300 K and blue 90 K on figure 2) [Li2004, Takesue2005].     Fig. 2. Technical noise contributions to BS for large detuning. Spontaneus pairs generation (purple) . Spontaneus Raman scattering at room temperature (red) and at 76 K (blue)  Thanks to the BS phasematching flexibility, there are no limitations on the amount of detuning [Mechin2006a] between pumps and signal, the only fundamental parameter being ωZDW (rather then β(3) ), easily tunable via dispersion engineering.  We operate using a dispersion-shifted fiber Vistacor from Corning inc.. Although the fiber is not optimized for nonlinear interactions ( γ ~ 3 W/km), a sufficiently long spool makes up for the reduced nonlinear parameter. Measurement of the dispersion measures , with λZDW = 1420nm that corresponds to a ωZDW =1330 THz ,the O-Band (1260 nm - 1320 nm ) and the C-Band (1530 nm - 1565 nm) ΔΩ ~ 120 THz . This is an attractive configuration because it enables the large detuning needed for a low-noise operation, while still operating at wavelengths where off-the-shelf equipment is available. For = 5 THz (about 6.5 nm at 1560 nm and 4.3 nm at 1300 nm) the acceptance bandwidth is ( FWHM).     Fig. 3. Vistacor dispersion    Fig. 4. Experimental setup. LD Laser Diode. EDFA Erbium Doped Fiber Amplifier. WDM Wavelenght Division Multiplexer. DSF Dispersion Shifted Fiber. SpAPD. Single photon avalanche photodioide. PPLN Periodiacally Poled Lithium Niobate Crystal.  To generate pump fields, we use temperature stabilized laser diodes (QPhotonics) that are current modulated via a pulse generator, producing pulses of duration τ = 1-10 ns and peak power 5 mW. The pumps are amplified with cascaded c-band erbium-doped fiber amplifier (EDFA). The last EDFA (Keopsys) is optimized for high power pulsed amplification at low duty cycle (output P out = 30 dBm at 100/1 duty cycle). Both pulses are temporally separated when traversing the EDFA to avoid mutual nonlinear effects in the gain medium, and synchronized afterward using an unbalance combination of 1551.7 nm fiber wavelength division multiplexers (WDM).   A free space filtering setup removes higher frequency noise (low-pass filter, cut-off 1550 nm) and selects one polarization with a combination of λ/2- and λ/4- waveplates, and a polarizing beam splitter. Signal and pumps are coupled together using O-Band/C-band WDM, temporally synchronized and injected in the nonlinear fiber. At the end of the interaction, second WDM removes most of the pumps, and the signal is sent to the detection stage.  The 100 m of nonlinear fiber is spooled and placed in a Styrofoam box that operates as a cryostat. Signal losses through the setup are as low as 2.6 dB, due to splices, connectors and WDMs. Since the second WDM removes > 30dbB of pump power, we consider most of the Raman noise generated between the two WDMs, and we take care of placing as much amount of fiber as possible in the cryostat.  Before detection, the signal has to be carefully filtered of all residual pump photons and thermal noise: we used a combination of a fiber-based 1300 nm/1550 nm pass/reject filter and a free-space filtering setup (3 dB losses) composed of a lowpass sharp-edge filter (1300-Semrock). A fiber-based WDM (passband 1290 ± 6.5 nm nominal) can be used to separate the signal and idler fields at approximately 1283.3 nm and 1278.5 nm. For some of the measurements, we selected a narrow band (0.5 nm) using a grating filter (2.6 dB), or we used a Dispersion Compensating Module (DCM D =-1200nm/km at 1300 nm, 7.6 dB losses) for spectral characterization using time of arrival information to recover the wavelength.  For detection, we use homebrew InGaAs single photon detectors in gated configuration. The diodes (Princeton Lightwave) are cooled to -40 C with thermo-electric coolers to reduces dark counts. (q.e. = 15%, gate duration = 20-60 ns, rep. rate ≤ 100 KHz). The gate signal is generated from the same delay generator that provides the pump current pulses.  Temporal measurements are performed using a Time Tagging Module (TTM) with internal resolution 86 ps. The time delay between triggering and detection is about 1200 ns.  A tunable laser (1260 nm-1340 nm) is used both for testing purposes and, in combination with a tunable attenuator, to generate a weak coherent pulse.  To generate single photons we use a source based on spontaneus parametric downconversion in PPLN crystal. In a 10 mm crystal, phasematching is achieved via temperature tuning, a CW pump at 543 nm generates photon pairs: 940-nm photons are detected by a Si APD to herald the presence of 1283-nm photons and generate a synchronization signal that is used trigger the pulse generator. The marginal bandwidth of the signal photons is larger > 10 nm. The heralding photons are filtered with a grating to match signal photon to the acceptance bandwith of BS-FWM, (unless stated, at 1283.8 with 0.5 nm FWHM). An heralding signal of approx. 67 kcps is observed, with a probability of coincidence detection of the heralded 1283.5 nm photon.    Fig. 5. Characterization of the noise dependency on temperature. Total noise photon generated vs. temperature.  We monitor the noise at one of the outputs while the fiber is changing temperature. If we collect the full signal setup, where contributions come over 12 nm bandwidth of the WDM, we can observe the temperature dependency (fig 5). The reduction of noise is limited to 2 orders of magnitude, which is expected because of the amount of fiber not placed in the cryostat (about 1 meter over 100 m of fiber in the cooler).  Taking losses into account, we calculate the probability of generating a photon of noise is, while being already extremely low compared with other BS demonstrations ( P = 1 per gate in highest efficiency reported [Clark2013a]) can be additionally filtered both temporally and spectrally to match the acceptance bandwidth.    Fig. 6. Traces of signal (left) and idler (right) field for increasing pump power. (top) Weak coherent state. (bottom) Single photon.   4. Experimental results  For this pump configuration, we measure perfect phasematching for λs = 1283.74(5) nm and λi = 1279.28(5) nm. We tune our CW source at λs , attenuated to a level of less than one photon per gate, and measure the conversion efficiency η as the total pump power varies, while keeping the two pumps balanced and the polarization aligned.   We tune the single photon source to emit an heralded photon at λs into the BS setup. When the pump is temporally overlapped, we observe depletion and conversion. In figure 6 we observe the detected trace on the APD for increasing pump power. To obtain this curve, signal and idler are separated in the freespace setup and each one filtered with a 0.5 nm filter (signal 3 dB losses, idler 6 dB). We show the conversion curve in figure 7.    Fig. 7. Conversion efficiency   Without filtering the conversion efficiency is limited by the fact that single photon bandwidth ( δλ =0.57 nm FWHM) and the acceptance ( δλBS = 1.17(2) nm FWHM) are comparable: theoretical calculation of translation efficiency, show a maximum efficiency η of about 89%.   We measure the second order correlation of the output state to verify the preservation of the quantum statistic on the translated output. We inject the output field into a 50/50 fiber beamsplitter, and record detection rates A(t) and B(t) , and coincidence rate C(τ) from the two coupler outputs. Normalization of the g(2)(τ)=C(τ)/N(τ) is obtained from the discrete cross-correlation of the singles N(τ)=A(t) ∗ B(t) . Error bars are extracted from statistical error on the count events. To remove the dark counts contribution to the measurement, we observe that the contribution of the dark counts to the coincidence count is (with the assumption of uncorrelated noise where and   We compare the g(2) (τ) measurements of the single photon traversing the setup, with pump either off or on. Even if the fields are not selected, because of the large conversion efficiency, we are measuring respectively at λs and λi : we observe an input (i.e. converted) g(2)(0)= 0.19. Though this value is below 0.5, the non classical boundary for single photons, it is limited by the presence of generated photons at that, unheralded, create a background noise. Upon frequency translation, the correlation takes the value of g(2)(0)= 0.23, showing a slight increase but still well within the quantum regime.   Next, we select the idler field with the free space filter and repeat the measurement, shown in figure 8: fitting the raw data with a Gaussian factor, we measure g(2)(0)= 0.2(1) (the losses introduced by the filtering reduced the SNR). When we correct for dark counts, we obtain g(2)(0)= 0.02(2). Because g(2)(0) has a bounded value, we utilize Bayes statistics to estimate the uncertainty.    Fig. 8. g(2) (τ) measurement of filtered idler: g (2) (0)= 0.2(1) . When we correct for dark counts, we obtain g (2) (0)= 0.02(2).  5. Conclusions  We demonstrated a FWM-BS configuration capable of achieving, for the first time, high efficiency, low-loss and low-noise frequency translation. This high performance setup, operating within the standard telecommunication bands, can be easily replicated to provide a toolbox and test-bed for different manipulations of the temporal and spectral properties of single-photons and weak-coherent states