Generation of nondegenerate narrow-band photon pairs for hybrid quantum network
Jian Wang, Peng-YinJie Lv, Jin-Ming Cui, Bi-Heng Liu, Jian-Shun Tang, Yun-Feng Huang, Chuan-Feng Li, Guang-Can Guo
GGeneration of nondegenerate narrow-band photon pairs for hybrid quantum network
Jian Wang, Peng-YinJie Lv, Jin-Ming Cui, Bi-Heng Liu, Jian-ShunTang, Yun-Feng Huang, ∗ Chuan-Feng Li, † and Guang-Can Guo Key Laboratory of Quantum Information, University of Science and Technology of China,CAS, Hefei, 230026, People’s Republic of ChinaSynergetic Innovation Centre in Quantum Information and Quantum Physics,University of Science and Technology of China, Hefei, Anhui 230026, China
In a hybrid quantum network, linking two kinds of quantum nodes through photonic channelsrequires excellent matching of central frequency and bandwidth between both nodes and their inter-facing photons. However, pre-existing photon sources can not fulfill this requirement. Using a novelconjoined double-cavity strategy, we report the generation of nondegenerate narrow-band photonpairs by cavity-enhanced spontaneous parametric down-conversion. The central frequencies andbandwidths of the signal and idler photons are independently set to match with trapped ions andsolid-state quantum memories. With this source we achieve the bandwidths and central frequenciesof 4 MHz at 935 nm and 5 MHz at 880 nm for the signal and idler photons respectively, with a nor-malized spectrum brightness of 4.9/s/MHz/mW. Due to the ability of being independently lockedto two different wavelenghts, the conjoined double-cavity is universally suitable for hybrid quantumnetwork consisting of various quantum nodes.
PACS numbers: 03.67.Hk, 42.50.Ar, 42.50.Dv, 42.50.Ex
The realization of quantum network composed withquantum nodes and quantum channels is of great im-portance to the distributed quantum computation andquantum communication. The quantum node works foroperating and storing quantum state and the quantumchannel works for distributing quantum information [1].Many experimental progresses have been achieved in thisaspect with different physical systems, such as atomicensembles, single atoms and trapped ions [2–4]. To becompared, hybrid quantum network consisting of variousphysical systems [5, 6] can combine the different advan-tages of diversities of physical systems. However, thecentral frequencies of photons for the interfaces linkingthe nodes built with different physical systems are usu-ally not identical. To solve this problem, the availableproposals are coherent quantum frequency conversion [7],tailoring the frequency of one kind of node to be samewith the other one [6] or interconnecting the nodes withnondegenerate photon pairs [8].As an advantage of the hybrid quantum network, dif-ferent physical systems can be chosen as computing orstoring nodes respectively. For the computing nodes,trapped ions can be a suitable candidate, because it isone of the most promising physical systems to realizequantum computation and simulation [9, 10]. The cre-ation of Greenberger-Horne-Zeilinger states in ion traphas been up to 14 qubits [11], and two-dimensional Isinginteractions simulated with hundreds of spins [12] hasbeen reported in experiment. To build a quantum net-work based on trap ions, it should have the ability tointerconvert stationary and flying qubits [13]. The emit-ted photon-ion entanglement and photon-mediated en-tanglement have been demonstrated in previous exper-iments [14, 15]. Recently, there have been experimen- tal results showing that single photons can be absorbedby the trapped single ion [6, 16, 17]. However, to thebest of our knowledge, the customized narrow-band pho-ton pair source suitable for trapped ions has not beenreported yet. For the storing nodes, a promising candi-date is the rare-earth ion-doped crystal system, becauseit has already shown some excellent merits as a quan-tum memory [18, 19]. There have been a lot of majoradvances in rare earth ion-doped crystal based quantummemory, including demonstration of stopped light andimage storage up to the regime of one minute [20], and upto 0.999 process fidelity for the storage and retrieval pro-cess of single-photon-level coherent pulse [21]. Moreover,recent progresses include the storage of photonic high-dimensional OAM Entanglement [22], storage of telecomwavelength time-bin entanglement [23], quantum telepor-tation of the polarization state of a telecom-wavelengthphoton onto the state of a solid-state quantum memory[24], and the preservation of quantum coherence on anhour-long timescale [25]. In view of these, the realiza-tion of photonic channels linking the trapped ions andthe rare earth ion-doped crystals should be useful for thehybrid quantum network.Compared to generated photon pairs from cold atomicsystem [17, 26], cavity-enhanced spontaneous paramet-ric down-conversion (SPDC) process is much more flexi-ble and less complicated for providing photonic channels.Passive filtering with optical etalon is also a direct way toget band-matched narrow-band photon pairs from single-pass SPDC source [16], but it will decrease the photoncounting rate extremely. During the process of cavity-enhanced SPDC, the down conversion photons with thefrequency and spatial mode matched to the cavity modeare greatly enhanced, otherwise greatly suppressed [27], a r X i v : . [ qu a n t - ph ] O c t FIG. 1: Experimental setup for narrow-band photon pair source. BS, beam splitter; HWP, half-wave plate; QWP, quarter-wave plate; PBS, polarization beam splitter; WG, waveguide; EOM, electro-optic modulator; IF, interference filter; FR, faradyrotator; FC, fiber coupler; PD1, PD2, photodiodes; SMF, single-mode fiber; SPCM, single photon counting module. The lockingbeams are frequency-modulated at 50 MHz by EOMs to generate the sidebands for the cavity locking. The combination ofHWP and FR is to get the reflected laser from the cavity, and the PDs detect the optical signal. The etalon is used to filterout the multimode components. resulting in generation of photon pairs with flexible band-width and perfect spatial mode. Since the first cavity-enhanced SPDC was demonstrated [27], lots of experi-ments have been performed for different purposes [28–33]. Recently, nondegenerate narrow-band photon pairsare generated with different methods [33–35]. However,linking two kinds of physical systems requires excellentmatch of the central frequencies and the bandwidths foreach node at the same time. In experiment [33], the fre-quency of idler photons can only be some certain valuesdetermined by the frequency of the signal photons be-cause of the locking system. In experiments [34, 35], thefrequencies of two photons are changing together withthe tuning of temperature. So the requirement can notbe fulfilled in above experiments.In this letter, we demonstrate a new way to gen-erate nondegenerate narrow-band photon pairs with aconjoined double-cavity (CDC) structure, and the cen-tral frequency and bandwidth are matched well to thetrapped ion Yb + [6] and the Nd -doped solid-statequantum memory [21]. The good match of central fre-quency is perfectly guaranteed by the fact that the twocavities in the CDC structure are independently lockedon the resonance absorption lines of two different kinds ofnodes, and the bandwidth match is ensured by modifieddesign of two individual cavities. The measured band-widths of the generated photon pairs is 4 MHz at 935 nmand 5 MHz at 880 nm. To further take advantage of longcoherence time of narrow-band photons, single-mode out-put of photons is realized by the temperature-controlledetalon filter.The experimental setup of our narrow-band photonpair source is shown in Fig. 1. Two semiconductor ECDL lasers with target wavelengths at 880 nm (Moglabs) and935 nm (Toptica DL pro) are tuned to the working wave-lengths of quantum memory and trapped ion. Two lasersare locked to the same ultrastable Fabry-P´erot (FP) cav-ity with 1 MHz bandwidth and 1500 finesses (StableLaser Systems) with Pound-Drever-Hall scheme, which isnot shown here [37]. The 453 nm pump laser is providedby the sum frequency generation (SFG) of the two 880nm and 935 nm lasers for the next step SPDC process.The SFG process mainly consists of a 1 cm long periodi-cally poled lithium niobate (PPLN) waveguide (HCPho-tonics Corp). In the cavity-enhanced SPDC process, a2 cm long periodically poled KTiOPO4 (PPKTP) crys-tal is used as the nonlinear crystal. The type-II quasi-phase matching is fabricated for the polarization of 453nm pump laser being paralleled with the polarization of935 nm laser. The calculated theoretical bandwidth ofthe phase-matching is 120 GHz. As shown in Fig. 1, acustomized dichroic mirror (DM) is inserted in the opticalcavity, which has a high transmissivity at the wavelengthof 880 nm ( T (cid:62) .
5% for s/p polarization) and a highreflectivity at the wavelength of 935 nm ( R (cid:62) .
99% fors/p polarization). The incident angle of the DM is about10 degree, resulted from the harsh demand for the goodperformance of two wavelengths at the same time. Thecavity mirror M1 is high-reflection-coated to work as aninput coupler ( R (cid:62) .
8% for 880 nm and 935 nm), andthe cavity mirrors M2, M3 work as the output couplersfor 880 nm and 935 nm respectively with a PZT attachedon each of them ( R = 97%). Thus the CDC structureis constructed with the help of the inserted dichroic mir-ror and the input coupler M1. The curvature of all thecavity mirrors is 10 cm, and the FSRs of the two cavities - 1 5 0 - 1 0 0 - 5 0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 002 0 0 04 0 0 06 0 0 08 0 0 01 0 0 0 01 2 0 0 01 4 0 0 01 6 0 0 01 8 0 0 0 Coincidence counts per time-bin
T i m e d e l a y ( n s )
FIG. 2: The correlation function G (2) S,I ( τ ) is measured at 0.8mW pump power. The time-bin size here is 4 ns and theintegration time is 1200 s. are both about 800 MHz. The nonlinear PPKTP crystalis also shared by the two cavities. This design providesa flexible method to overcome the critical difficulty tomatch the central frequencies and bandwidths of the sig-nal and idler photons to different quantum systems. Allof the optical elements except inserted DM is antireflec-tively coated at 453 nm such that the pump laser passesthe nonlinear PPKTP crystal only once.To realize the cavity-enhanced SPDC process, the twocavities in our CDC structure are locked to two referencelaser beams independently. Owing to the same travelingdirection of the locking beam and SPDC photons, the sin-gle photon counting module (SPCM) is under the threatof being excessively illuminated. A widely used methodis intermittently locking the cavities with the help of amechanical chopper [31, 33]. However, in nondegeneratecase [33], two mechanical choppers are usually needed toblock two different locking beams and the critical phaselock should be achieved between two choppers, due tothe inevitable difference frequency generation process inthe cavity. In our design, as shown in Fig. 1, two lockingbeams are combined together, so are the SPDC photonpairs, to pass through the chopper, then separated bythe dichroic mirror. These will ensure that the SPCMis definitely safe even if the chopper is broken down orthe electricity shuts down. Besides, thanks to the CDCdesign, no double resonances in one cavity is needed asbefore [31], which will decrease the demand for the accu-racy of the temperature control of PPKTP crystal. Allthe above improvements make our setup rather robustagainst the exotic environmental disturbances.It has been shown that the SPDC photon pairs emit-ted from the cavity are often multimode output [27],which will decrease the coherence and limit the use ofthe narrow band photon pairs in experiments. So we putan etalon filter after the output coupler of the 880 nmcavity (shown in Fig. 1) to eliminate the extra multi-mode components [31]. The etalon (central wavelength880 nm, bandwidth 120 MHz, FSR 8.4 GHz) is placedin a temperature-controlled oven with an accuracy of 10mK to achieve a good filtering effect. To learn more about the generated photon pairs, thetime-correlated measurements are conducted in our ex-periments. The specific method is to measure the second-order cross-correlation function, G (2) S,I ( τ ), between photonpairs [32, 33]. The signal and idler photons are separatedby a dichroic mirror and fiber-coupled into the SPCMsfor detection after necessary optical filtering. The timedistribution of photon pairs arriving at the detectorsis recorded by the time-to-digital-converter (Picoquant400). The measured results of G (2) S,I ( τ ) are plotted in Fig.2. The correlation function is fitted separately on the twosides of curve referred to the fitting function e − π ∆ υτ , re-sulting in the signal and idler photons bandwidths being4 MHz at 935 nm and 5 MHz at 880 nm. The smalldifference is mainly caused by the different losses of theinserted dichroic mirror at the two wavelengths.For the purpose of quantum information processing,the high value of normalized cross-correlation functionat zero delay, g (2) S,I (0), is very important and necessary.The detail form of normalized cross-correlation functioncan be expressed as: g (2) S,I ( τ ) = (cid:104) E † S ( t ) E † I ( t + τ ) E I ( t + τ ) E S ( t ) (cid:105)(cid:104) E † I ( t + τ )( t ) E I ( t + τ ) (cid:105)(cid:104) E † S ( t ) E S ( t ) (cid:105) (1)where E S,I ( τ ) indicates the operators of signal and idlerfields. It is calculated using the correlation function asthe peak value divided by the average count value in theaccidental region (in the range from 200 ns to 250 ns inFig. 2 [33]. The measured g (2) S,I (0) varying with differentinput pump power is plotted in Fig. 3, showing a well fit-ted inverse proportion relation [35]. The g (2) S,I (0) is about88 when the input pump power is 1 mW, much higherthan the classical threshold of 2 for two-mode squeezedstates. Decreasing the pump power to less than 100 µ W,the g (2) S,I (0) gets a high value of about 800. All these val-ues are calculated without subtracting any backgroundcounts. The quite high values of g (2) S,I (0) indicate the goodquality of the quantum source and reliable use in hybridquantum network. The coincidence rate varing with dif-ferent pump power is also shown in Fig. 3.The time-resolved measurement G (2) S,I ( τ ) of generatedcavity-enhanced photon pairs should be a time-domaincomblike structure of the curve, for the reason of interfer-ence between different frequency modes [28]. The periodof the interference fringe is t = 1 /F SR , be equal to thecavity round trip time. In our case, as the SPCMs havea resolution time of 350 ps, and the cavity round triptime of 1.25 ns, we get a unperfect comblike interferencefringes, as plotted in Fig. 4. The destructive interferencepart cannot reach zero, resulted from the small ratio be-tween cavity round trip time and resolution time. Theasymmetry of interference fringes can be blamed for themode difference between the two cavities. Then an etalonis inserted to filter the unwanted multimodes. Based on P u m p p o w e r ( m W ) g (2) S,I (0)
05 01 0 01 5 02 0 02 5 03 0 03 5 04 0 0
Coincidence (Hz)
FIG. 3: The measured g (2) S,I (0) varying with the pump power isplotted at left axis (blue square). The multimode coincidencerate versus the pump power is plotted on right axis (green tri-angle). The coincidence rate is measured by adding up all thebins of coincidence curve divided by the overall measurementtime with the subtraction of background counts [30].
Coincidence
T im e d e la y ( n s )
Coincidence
T i m e d e l a y ( n s )
FIG. 4: The measured time-resolved correlation function G (2) S,I ( τ ) with the time-bin size of 256 ps. The imperfect comb-like interference fringes is due to the low ratio between cavityround trip time and resolution time. The inset is measuredwith the etalon inserted after the 880 nm cavity, showing agood elimination of the multimode components. the same time-resolved measurement, the comblike struc-ture of the curve almost disappears, as in the inset of Fig.4. This relatively smooth curve indicates a quite largeelimination of multimode photon pairs.To further ensure the single mode output of generatedphoton pairs, we use a Michelson interferometer to mea-sure the coherence of the filtered 880 nm photons [33, 36].The results are shown in Fig. 5. To scan the interferencefringes, a reflective mirror in one arm is mounted on atranslation stage with a piezoelectric transducer attachedfor fine tuning. With a 880 nm laser beam fed into theinterferometer, a high average interference visibility ofabout 0.99 is observed, showing the good quality of in-terferometer. Without the etalon filter, the interferencefringes shows decreasing visibilities with a certain perioddue to the multimode components. While the average in-terference visibility of the filtered case is 92% in the test - 8 - 4 0 4 80 . 00 . 20 . 40 . 60 . 81 . 0 Visibility
L ( m m )
FIG. 5: Measured results of the visibility of the Michelson in-terferometer versus its optical path difference in three cases,illustrating the coherence of the input light. High interferencevisibilities of classical light shows the good quality of the inter-ferometer (red circle). Owing to the existence of multimodes,the unfiltered case shows overall decreasing interference visi-bilities (black triangle). For filtered case, its interference visi-bilities are almost as high as classical case, indicating a goodfiltering of multimodes (blue square). region, with a stable trend of high interference visibil-ity when increasing the optical path difference betweentwo arms. The combined measurement of second ordercross correlation and first order autocorrelation (coher-ence) provides a good evidence that our source can pro-vide single-mode output.After the filtering of multimodes, the single mode pho-ton pairs generation rate is detected at pump power of0 . . ∼ / ∗ Electronic address: [email protected] † Electronic address: cfl[email protected][1] H. J. Kimble, Nature (London) , 1023 (2008).[2] N. Sangouard, C. Simon, H. de Riedmatten, and N.Gisin, Rev. Mod. Phys. , 33 (2011).[3] S. Ritter, C. N¨olleke, C. Hahn, A. Reiserer, A. Neuzner,M. Uphoff, M. M¨ucke, E. Figueroa, J. Bochmann, andG. Rempe, Nature (London) , 195 (2012).[4] L.-M. Duan and C. Monroe, Rev. Mod. Phys. , 1209(2010).[5] M. Wallquist, K. Hammerer, P. Rabl, M. Lukin and P.Zoller, Phys. Scr. T137 , 014001 (2009).[6] H. M. Meyer, R. Stockill, M. Steiner, C. Le Gall, C.Matthiesen, E. Clarke, A. Ludwig, J. Reichel, M. Atatre,and M. K¨ohl, Phys. Rev. Lett. , 123001 (2015).[7] S. Ramelow, A. Fedrizzi, A. Poppe, N. K. Langford, andA. Zeilinger, Phys. Rev. A , 013845 (2012).[8] C. Simon, H. de Riedmatten, M. Afzelius, N. Sangouard,H. Zbinden, and N. Gisin, Phys. Rev. Lett. , 190503(2007).[9] R. Blatt, D. Wineland, Nature (London) , 1008(2008).[10] J.-M. Cui, Y.-F. Huang, Z. Wang, D.-Y. Cao, J. Wang,W.-M. Lv, Y. Lu, L. Luo, A. del Campo, Y.-J. Han, C.-F.Li, G.-C. Guo, arXiv: 1505. 05734.[11] T. Monz, P. Schindler, J. T. Barreiro, M. Chwalla, D.Nigg, W. A. Coish, M. Harlander, W. H¨ansel, M. Hen-nrich,and R. Blatt, Phys. Rev. Lett. , 130506 (2011).[12] J. W. Britton, B. C. Sawyer, A. C. Keith, C.-C. J. Wang,J. K. Freericks, H. Uys, M. J. Biercuk, and J. J. Bollinger,Nature (London) , 489 (2012).[13] H. H¨affner, C. F. Roos, R. Blatt, Phys. Rep. , 155(2008).[14] A. Stute, B. Casabone, P. Schindler, T. Monz, P. O.Schmidt, B. Brandst¨atter, T. E. Northup, and R. Blatt,Nature (London) , 484 (2012).[15] D. L. Moehring, P. Maunz, S. Olmschenk, K. C. Younge,D. N. Matsukevich, L.-M. Duan, and C. Monroe, Nature(London) , 68 (2007). [16] N. Piro, F. Rohde, C. Schuck, M. Almendros, J. Huwer,J. Ghosh, A. Haase, M. Hennrich, F. Dubin, and J. Es-chner, Nature Physics. , 17 (2011).[17] M. Schug, J. Huwer, C. Kurz, P. M¨uller, and J. Eschner,Phys. Rev. Lett. , 213603 (2013).[18] H. de Riedmatten, M. Afzelius, M. U. Staudt, C. Simonand N. Gisin, Nature (Londn) , 773 (2008).[19] D. Riel¨ander, K. Kutluer, P. Ledingham, M. G¨undoˇgan,J. Fekete, M. Mazzera and H. de Riedmatten, Phys. Rev.Lett. , 040504 (2014).[20] G. Heinze, C. Hubrich, and T. Halfmann, Phys. Rev.Lett. , 033601(2013).[21] Z.-Q. Zhou, W.-B. Lin, M. Yang, C.-F. Li, and G.-C.Guo, Phys. Rev. Lett. , 190505 (2012).[22] Z.-Q. Zhou, Y.-L. Hua, X. Liu, G. Chen, J.-S. Xu, Y.-J.Han, C.-F. Li and G.-C. Guo, arxiv: 1412. 5243v1.[23] E. Saglamyurek, J. Jin, V. B. Verma, M. D. Shaw, F.Marsili,S. Woo Nam, D. Oblak and W. Tittel, NaturePhotonics , 83 (2015).[24] F. Bussi`eres, C. Clausen, A. Tiranov, B. Korzh, V. B.Verma,S. Woo Nam, F. Marsili, A. Ferrier, P. Goldner,H. Herrmann, C. Silberhorn, W. Sohler, M. Afzelius andN. Gisin, Nature Photonics , 775 (2014).[25] M. Zhong, M. P. Hedges, R. L. Ahlefeldt, J.G.artholomew, S. E. Beavan, S. M. Wittig, J. J. Longdell,and M. J. Sellars, Nature (London) , 177 (2015).[26] J. K. Thompson, J. Simon, H. Loh, and V. Vuletic, Sci-ence , 74 (2006).[27] Z. Y. Ou and Y. J. Lu, Phys. Rev. Lett. , 2556 (1999).[28] H. Wang, T. Horikiri, and T. Kobayashi, Phys. Rev. A , 043804 (2004).[29] C. E. Kuklewicz, F. N. C. Wong, and J. H. Shapiro,Phys.Rev. Lett. , 223601 (2006).[30] Z.-Y. Zhou, D.-S. Ding, Y. Li, F.-Y. Wang, and B.-S.Shi, J. Opt. Soc. Am. B, , 1 (2014).[31] X.-H. Bao, Y. Qian, J. Yang, H. Zhang, Z.-B. Chen,T.Yang, and J.-W. Pan, Phys. Rev. Lett. , 190501(2008).[32] F. Wolfgramm, Y. A. de Icaza Astiz, F. A. Beduini,A.Cer`e, and M. W. Mitchell, Phys. Rev. Lett. ,053602 (2011).[33] J. Fekete, D. Riel¨ander, M. Cristiani, and H. de Ried-matten, Phys. Rev. Lett. , 220502 (2013).[34] M. F¨ortsch, J. U. F¨urst, C. Wittmann, D. Strekalov, A.Aiello, M. V. Chekhova, C. Silberhorn, G. Leuchs, andC. Marquard, Nature communications , 1818 (2013).[35] K.-H. Luo, H. Herrmann, S. Krapick, R. Ricken, V. Quir-ing, H. Suche, W. Sohler, and C. Silberhorn, arXiv: 1306.1756.[36] F.-Y. Wang, B.-S. Shi, C. Zhai, and G.-C. Guo, J. Mod.Opt. , 330 (2010).[37] E. D. Black, Am. J. Phys.69