10.07.2015 Views

Performing high-quality multi-photon experiments with parametric ...

Performing high-quality multi-photon experiments with parametric ...

Performing high-quality multi-photon experiments with parametric ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

J. Phys. B: At. Mol. Opt. Phys. 42 (2009) 114008 T Jennewein et alcoming from different pulsed SPDC processes. Each effectwill be assigned <strong>with</strong> a visibility, and mostly defined as usualas V = (C max − C min )/(C max + C min ), where C are the counted<strong>photon</strong> signals fringing between a maximal and minimalvalue. There is the exception for the case of the two-<strong>photon</strong>interference (Hong–Ou–Mandel dip), where the visibility mustbe measured as V = (C inf − C min )/C inf , where C inf are thecounted two-<strong>photon</strong> events well outside the interference range,and C min are the minimal counted events at the point of optimalinterference (dip). In the first approximation, we estimate thetotal expected correlation visibility as the product of all theindividual visibilities, V tot ≃ ∏ i V i.3.1. Two-fold visibility of pulsed SPDCFigure 2. Measured two-<strong>photon</strong> interference visibilities compared<strong>with</strong> the calculation of our model, for three experimental scenarios.First, the Hong–Ou–Mandel interference (HOM) is considered,where the two <strong>photon</strong>s b and c interfere at the beam splitter, andpolarizers at 45 ◦ are placed in both of their paths. Second, anentanglement swapping experiment, where the correlation between<strong>photon</strong>s a and d is observed <strong>with</strong> polarizers at 45 ◦ setting. Third,again an entanglement swapping experiment, but both the polarizersare oriented at 0 ◦ . Some measurements show larger error bars thanthe final results presented in table 2, since they were taken just foralignment purposes <strong>with</strong>in a shorter experimental duration.Table 1. Summary of the estimated visibility for the entanglementswapping <strong>experiments</strong>. See the text for details.Filter bandwidth FWHM 1 nm 2 nmEffectTwo-fold correlation at 0 ◦ V I,0 ◦ 0.980 0.980Two-fold correlation at 45 ◦ V I,45 ◦ 0.935 0.904Splitting-ratio of the fibre BS V II,45:55 0.996 0.996Polarization alignment in the fibres V IV,3 ◦ 0.998 0.998Center wavelength of <strong>photon</strong>s b and c V V,0.25nm 0.977 0.977Distinguishability of <strong>photon</strong>s in HOM V VI 0.997 0.991Temporal jitter between <strong>photon</strong> pairs V VII 0.964 0.881Final swapping visiblity at 0 ◦ V 0 ◦ 0.896 0.815Final swapping visibility at 45 ◦ V 45 ◦ 0.816 0.693measured results are shown in figure 2 and are compared <strong>with</strong>the estimation from our model which includes various sourcesof errors. With these insights it is possible to understandthe limitation of the experimental <strong>quality</strong> of <strong>multi</strong>-<strong>photon</strong><strong>experiments</strong> based on <strong>parametric</strong> down-conversion. Moreimportantly, these estimations show that in our situation itis necessary to choose the filter bandwidth as narrow as 1 nmFWHM to finally achieve a level of entanglement well abovethe classical limits.3. Optimization of the entanglement swapping<strong>quality</strong>In the following, we give a detailed overview of the variouseffects limiting the <strong>quality</strong> for the interference of <strong>photon</strong>sFirst we consider the attainable <strong>quality</strong> of the two-foldcorrelations of the entangled <strong>photon</strong> pairs. An obvioussource of error is the alignment accuracy of the polarizingoptical elements in the down-conversion system (such as theSPDC crystals, the compensator crystals, wave plates [26]).In addition it has already been shown in several theoreticalcalculations and in <strong>experiments</strong> [30–34] that the in SPDCpumped <strong>with</strong> femtosecond laser pulses the achievable <strong>quality</strong>of two-fold entanglement is inherently reduced. Thereforethe correlation visibility for measurements along the 45 ◦ orthe circular polarization basis will have less <strong>quality</strong> thanmeasurements in the 0 ◦ basis. In this experiment we observedin the 0 ◦ /90 ◦ polarization basis a visibility of V I,1nm,0 ◦ =0.980, and in the 45 ◦ basis, corresponding to a two-foldcorrelation visibility of V I,1nm,45 ◦ = 0.935 <strong>with</strong> filters for<strong>photon</strong>s b and c having a 1 nm FWHM bandwidth, and avisibility of V I,2nm,45 ◦ = 0.904 <strong>with</strong> filters of 2 nm.3.2. The Bell-state analyserSince we are utilizing a fibre-optic beam splitter at the heart ofthe Bell-state analyser, the mode overlap is inherently perfect.As the only source of error we must take into account thatthe required splitting ratio of 50:50 is not perfectly achieved.The non-ideal splitting ratio leads to an achievable visibilityof the two-<strong>photon</strong> interference of [35] V II = 2RT , <strong>with</strong>R 2 +T 2reflectivity R and transmittivity T. Our actual device had asplitting ratio of about 45:55, which reduced the visibility ofthe HOM interference by V II,45:55 = 0.996.3.3. Higher order emission terms of SPDCThe emission characteristics of our <strong>photon</strong> pair productionimplies a similar probability for producing two <strong>photon</strong> pairsin separate modes (one <strong>photon</strong> each in modes a, b, c, d) or twopairs in the same mode (two <strong>photon</strong>s each in modes a and b orin modes c and d). The latter events lead to ‘false’ coincidencesin the detectors in the BSA hence creating a noise background.However, our optimal Bell-state analyser partially diminishesthis problem, since it reduces the unwanted <strong>high</strong>er order eventsby at least a factor of 2. We exclude part of these casesby only accepting events where one <strong>photon</strong> each in modes aand d is registered. Furthermore, the Bell-state analyser willonly detect a valid event when the two <strong>photon</strong>s (b and c) take3


J. Phys. B: At. Mol. Opt. Phys. 42 (2009) 114008 T Jennewein et al<strong>photon</strong> pairs and obtain the overall two-<strong>photon</strong> coincidenceprobability asP b,c ′ (τ) = 1 ∫ d/2 ∫ d/2Pd 2 b,c (t a,b − t c,d − τ)dt a,b dt c,d , (3)−d/2−d/2where τ is the offset of the two arms of the <strong>photon</strong>s to the beamsplitter. We further assume that the <strong>photon</strong> wave-packets area Gaussian shape <strong>with</strong> the width equal to the coherence timeτ cohr of the <strong>photon</strong>s, which interfere <strong>with</strong> time delay τ andthe offset t a,b − t c,d . Therefore the coincidence probability forthe interference of two <strong>photon</strong>s isP b,c (t a,b − t c,d − τ)= 1 2(1− 2V VI1+ V VIexp (t a,b − t c,d − τ) 2τ 2 cohr).In our experiment the crystal had a thickness of d = 2 mm,hence the <strong>photon</strong> pairs are undefined <strong>with</strong>in the range of|t a,b | 180 fs (large compared to the <strong>photon</strong> coherenceτ cohr = 775 fs for 1 nm bandwidth filters). The HOM visibilityis defined as V VII = (P ′ (inf) − P ′ (0))/P ′ (inf). The explicitnumerical calculation of the integral (3) for filter band widthsof λ FWHM of 1 nm, 2 nm, 3 nm leads to the HOM visibilitiesof V VII,1nm = 0.9641,V VII,2nm = 0.8817,V VII,3nm = 0.7778respectively. A further advanced calculation should alsoinclude the synchronization jitter of two pump lasers, andis given by Kaltenbaek [39].The standard approach for minimizing or even avoidingthe effects of this group velocity mismatch is to use narrowbandwidth filters and thereby extend their coherence timeof the SPDC <strong>photon</strong>s beyond the group velocity dispersion.Unfortunately this leads to a huge loss of the generated<strong>photon</strong>s. This timing effect is finally resolved by utilizingparticular crystals <strong>with</strong> well chosen or even specificallytailored group velocity properties, that do not require anyfilters for the SPDC and still achieve <strong>high</strong>-<strong>quality</strong> two-<strong>photon</strong>interference [40–43].4. Experimental resultsThe expected contributions of all the above discussed effectsare summarized in table 1. The total visibility is calculatedby taking the product of all contributing visibilities. Inour experiment we performed a range of test measurementsin order to compare the experimentally attainable HOMvisibilities <strong>with</strong> the estimated visibilities, these discussedabove. The measured results are summarized in figure 2and compared <strong>with</strong> the respective estimates. Note the verygood agreement of the experiment <strong>with</strong> the calculation.Now it is possible to understand the various effects thatreduce the experimental <strong>quality</strong> in our entanglement swappingexperiment. More importantly we are able the adequatelychoose the filter bandwidth as 1 nm FWHM, such thatall further measurements on the entanglement swapping areperformed at the required <strong>quality</strong> of the experiment.The correlations of the entangled <strong>photon</strong>s are quantifiedby measuring the correlations between <strong>photon</strong>s a and d <strong>with</strong>the expectation value for joint polarization measurementsE(φ a ,φ d ), where φ a and φ d are the polarizer setting for <strong>photon</strong>s(4)a and d, respectively [44]. The correlation coefficients aredefined asE(φ a ,φ d ) = (N ++ (φ a ,φ d ) − N +− (φ a ,φ d )∑− N −+ (φ a ,φ d ) + N −− (φ a ,φ d ))/Nij , (5)where as usual the N ij (φ a ,φ d ) are the coincidence countsbetween the i-channel of the polarizer of <strong>photon</strong> a set at angleφ a , and the j-channel of the polarizer of <strong>photon</strong> d set at angleφ d . Each measurement run of the correlation coefficientsE(φ a ,φ d ) for <strong>photon</strong>s a and d at settings φ a = φ d = 45 ◦involved a scan through the relative time delay betweenthe <strong>photon</strong>s b and c by moving a mirror in the path ofthe pump laser for one of the SPDC sources, see figure 3.Clearly the dependence of the <strong>photon</strong> interference on the delaybetween the two <strong>photon</strong>s can be observed and <strong>high</strong>lights theoperation of the interferometric Bell-state analysis used inthis entanglement swapping experiment. This measurementactually shows the very <strong>high</strong> correlation of the entangled<strong>photon</strong>s after the entanglement swapping of E = 0.91 forthe | − 〉 ad and E = 0.86 for the | + 〉 ad .Several such expectation values E were used to obtainmeasures for the state fidelity, defined as F =〈|ρ〉. Tosee how this was done, first take the general definition of atwo-qubit density matrix ρ( 3∑)ρ = 1 T mk σ m ⊗ σ k (6)40where T mk are the expectation values Tr{ρσ m σ k } for aparticular spin measurement, σ 0 is the identity and σ 1,2,3are the Pauli matrices. We also consider the projectorsof the two Bell states used in our experiment, which are| ∓ 〉〈 ∓ |=0.25(1 − σ 1 σ 1 ∓ σ 2 σ 2 ∓ σ 3 σ 3 ) and obtain thefidelity via the trace F | ∓ 〉 = Tr{ρ| ∓ 〉〈 ∓ |}, leading toF | ∓ 〉 = 1 4 [1 − T 11 ∓ T 22 ∓ T 33 ]. (7)This expression is experimentally easily accessible, since itonly requires three different measurements and not the setof the nine measurements (combinations of the three basissettings on either <strong>photon</strong>) needed for performing a tomographyof the state – which made a huge practical difference giventhe limited count rates of our experiment. In our case theaxis of the Pauli matrices are identified <strong>with</strong> the polarization,where σ 1,2,3 correspond to analyser setting in the H, 45 andL polarization basis. Therefore the coefficients in the aboveexpression for the fidelity are T 11 = E(H, H), and likewise forthe other polarization bases.We have performed the correlation measurements on theswapped entangled <strong>photon</strong>s which end up in the | − 〉 ad or| + 〉 ad depending on the outcome of the bell-state analyser, seetable 2. These results lead to overlap fidelities of the swappedentangled <strong>photon</strong>s <strong>with</strong> the ideal Bell states of F | − 〉 = 0.892and F | + 〉 = 0.879 for states | − 〉 ad and | + 〉 ad , respectively.As a sufficient indication of the presence of entanglementwe tested a Bell ine<strong>quality</strong>, where the suitable version for<strong>experiments</strong> is the Clauser–Horne–Shimony–Holt (CHSH)variant [29], which has the following form:S =|E(φ a ′ ,φ′ d ) − E(φ′ a ,φ′′ d )| + |E(φ′′ a ,φ′ d ) + E(φ′′a ,φ′′ d)| 2.(8)5


J. Phys. B: At. Mol. Opt. Phys. 42 (2009) 114008 T Jennewein et alTable 2. Measured correlation values for <strong>photon</strong>s a and d forpolarizer settings θ a ; θ d , after the entanglement swapping protocol.The outcome of the Bell-state measurement on <strong>photon</strong>s b and cdetermines the type of correlation to be a − or + . The first threemeasurements show the <strong>high</strong> <strong>quality</strong> of the entangled state in thethree bases. These measurements give us the very <strong>high</strong> state fidelityof F = 0.892 and F = 0.879. The following four measurementsallow us to test the CHSH ine<strong>quality</strong>. As the local realistic upperlimit for S is 2, above results for both Bell states clearly violate theCHSH ine<strong>quality</strong>. Note that the correlation coefficients for the twoBell states were measured <strong>with</strong>in the same measurement run for aparticular polarizer setting by sorting the data into correspondingsubsets. The given errors are the statistical uncertainties due to<strong>photon</strong> counting. The differences in the correlation coefficientscome from the <strong>high</strong>er correlation fidelity for analyser settings closerto 0 ◦ and 90 ◦ .E(θ a ; θ d ) | − 〉 ad | − 〉 bc | + 〉 ad | + 〉 bcE(0 ◦ ; 0 ◦ ) −0.972 ± 0.0 −0.957 ± 0.0E(45 ◦ ; 45 ◦ ) −0.832 ± 0.0 0.771 ± 0.0E(L; L) −0.765 ± 0.0 0.789 ± 0.0E(0 ◦ ; 22.5 ◦ ) −0.61 ± 0.04 −0.55 ± 0.04E(0 ◦ ; 67.5 ◦ ) 0.80 ± 0.04 0, 60 ± 0, 04E(45 ◦ ; 67.5 ◦ ) −0.64 ± 0.05 0.48 ± 0.05E(45 ◦ ; 22.5 ◦ ) −0.55 ± 0.05 0.67 ± 0.05Bell-parameter S 2.60 ± 0.09 2.30 ± 0.09It is straightforward to show that the prediction of quantummechanics leads to E QM (φ a ,φ d ) =−cos(2(φ a − φ d )). Withthe settings chosen as (φ a ′,φ′ d ,φ′′ a ,φ′′ d ) = (0◦ , 22.5 ◦ , 45 ◦ ,67.5 ◦ ) the maximal value of S QM = 2 √ 2, which clearlyviolates the limit of 2 imposed by ine<strong>quality</strong> (8). Thenecessary series of correlation coefficients of <strong>photon</strong>s a andd were measured, and are given in table 2. Note that thecorrelation coefficients for the two different Bell states weremeasured <strong>with</strong>in the same measurement run for a particularpolarizer setting by sorting the data into corresponding subsetsdepending upon the observed Bell state. Our results show clearviolations of the Bell ine<strong>quality</strong> <strong>with</strong> S = 2.60 ± 0.09 for the| − 〉 ad and S = 2.30 ± 0.09 for the | + 〉 ad Bell state. Thelower value of S for | + 〉 ad is due to the difficult polarizationalignment of the output fibres of the beam splitter.5. ConclusionWe demonstrate an entanglement swapping experimentusing <strong>photon</strong>s produced by spontaneous <strong>parametric</strong> downconversion(SPDC) and an optimal Bell-state analyser <strong>with</strong>50% efficiency, i.e. capable of detecting two of the four Bellstates perfectly. Via a thorough analysis and optimization ofthe experimental errors in our system, we achieved a very <strong>high</strong>experimental entanglement correlation and tested the violationof two Bell inequalities on the swapped <strong>photon</strong>s in the sameexperiment run. In addition, we showed that the final entangledBell states had an observed state fidelity F = 0.892 andF = 0.879 <strong>with</strong> the respective ideal states | − 〉 bc and | + 〉 bc .In particular, we studied the various deteriorating effects in thespontaneous <strong>parametric</strong> down conversion which are crucialfor performing <strong>high</strong>-<strong>quality</strong> <strong>multi</strong>-<strong>photon</strong> <strong>experiments</strong>, andgave a simple model that allowed us to estimate the finalstate fidelity. We consequently improved the experimentalperformance by carefully optimizing the several sources oferrors. As the main experimental obstacle for SPDC systemswe also estimated the group velocity mismatch between thepump and the down-conversion <strong>photon</strong>s in the SPDC crystal ina simple mathematical model, and compared the calculations<strong>with</strong> several test <strong>experiments</strong>. These calculations also allowedus to understand why in our case the required bandpass filtersfor the interfering <strong>photon</strong>s must be the narrow value of 1 nmFWHM. We note that finally the problem of the groupvelocity mismatch will be solved by cancelling the dispersioneffects <strong>with</strong> advanced phase-matching methods [40–43]. Thevarious effects of errors must be considered in most types of<strong>multi</strong>-<strong>photon</strong> <strong>experiments</strong> that rely on two-<strong>photon</strong> interferenceeffects and are based on <strong>parametric</strong> down-conversion.From the fundamental side it is interesting to note that theobserved violations of the Bell inequalities in our entanglementswapping experiment drastically <strong>high</strong>lights the non-classicalnature of entanglement, since the outcome of the Bellmeasurementof the two inner <strong>photon</strong>s b and c (| − 〉 bc and| + 〉 bc ) even determines the type of entanglement of the twoouter <strong>photon</strong>s a and d (| − 〉 ad and | + 〉 ad )[45].Since teleportation using a two-outcome Bell-statemeasurement is also one of the building blocks of linearopticalquantum gates [46], this experiment clearly representan important step for linear-optical quantum informationprocessing, the creation of <strong>multi</strong>-<strong>photon</strong> entangled states andquantum repeater protocols.AcknowledgmentsWe thank M Zukowski, D Greenberger, C Brukner, T Rudolph,A White, M Barbieri and R Kaltenbaek for fruitful discussionson experimental and theoretical questions. This work has beensupported by the Austrian Science Foundation (FWF) ProjectNo. F1506 and F1520, and by the European Commission underthe Integrated Project Qubit Applications (QAP) funded bythe IST directorate and the DTO-funded U.S. Army ResearchOffice.References[1] Bennett C H, Brassard G, Crépeau C, Jozsa R, Peres Aand Wootters W K 1993 Phys. Rev. Lett. 70 1895[2] Bouwmeester D, Pan J-W, Mattle K, Eibl M, Weinfurter Hand Zeilinger A 1997 Nature 390 575[3] Furusawa A, Sorensen J L, Braunstein S L, Fuchs C A,Kimble H J and Polzik E S 1998 Science 282 706[4] Riebe M et al 2004 Nature 429 734[5] Barrett M D et al 2004 Nature 429 737[6] Nielsen M A, Knill E and Laflamme R 1998 Nature 396 52(http://dx.doi.org/10.1038/23891)[7] Sherson J F, Krauter H, Olsson R K, Julsgaard B,Hammerer K, Cirac I and Polzik E S 2006 Nature 443 557(http://dx.doi.org/10.1038/nature05136)[8] Chen Y-A, Chen S, Yuan Z-S, Zhao B, Chuu C-S,Schmiedmayer3 J and Pan J-W 2008 Nature Phys. 4 103[9] Zukowski M, Zeilinger A and Weinfurter H 1995 Ann. NYAcad. Sci. 755 91[10] Jennewein T, Weihs G, Pan J-W and Zeilinger A 2002 Phys.Rev. Lett. 88 0179036


J. Phys. B: At. Mol. Opt. Phys. 42 (2009) 114008 T Jennewein et al[11] Kaltenbaek R, Prevedel R, Aspelmeyer M and Zeilinger A2008 High-fidelity entanglement swapping <strong>with</strong> fullyindependent sources http://www.citebase.org/abstract?id=oai:arXiv.org:0809.3991[12] Matsukevich D N, Maunz P, Moehring D L, Olmschenk Sand Monroe C 2008 Phys. Rev. Lett. 100 150404(http://link.aps.org/abstract/PRL/v100/e150404)[13] Yuan Z-S, Chen Y-A, Zhao B, Chen S, Schmiedmayer Jand Pan J-W 2008 Nature 454 1098[14] Bell J 1964 Physics 1 195[15] Kok P, Munro W J, Nemoto K, Ralph T C, Dowling J Pand Milburn G J 2007 Rev. Mod. Phys. 79 135(doi:10.1103/RevModPhys.79.135)[16] Browne D E and Rudolph T 2005 Phys. Rev. Lett. 95 010501(doi:10.1103/PhysRevLett.95.010501)[17] Briegel H-J, Dür W, Cirac J I and Zoller P 1998 Phys. Rev.Lett. 81 5932[18] Volz J, Weber M, Schlenk D, Rosenfeld W, Vrana J,Saucke K, Kurtsiefer C and Weinfurter H 2006 Phys. Rev.Lett. 96 030404[19] Genovese M 2005 Phys. Rep. 413 319 (http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0701071)[20] Calsamiglia J and Lütkenhaus N 2001 Appl. Phys. B 72 67[21] Mattle K, Weinfurter H, Kwiat P G and Zeilinger A 1996Phys. Rev. Lett. 76 4656[22] Michler M, Mattle K, Weinfurter H and Zeilinger A 1996Phys. Rev. A 53 R1209[23] Ursin R, Jennewein T, Aspelmeyer M, Kaltenbaek R,Lindenthal M, Walther P and Zeilinger A 2004 Nature430 849[24] Kwiat P G and Weinfurter H 1998 Phys. Rev. A58 R2623[25] Schuck C, Huber G, Kurtsiefer C and Weinfurter H 2006 Phys.Rev. Lett. 96 190501[26] Kwiat P G, Waks E, White A G, Appelbaum I andEberhard P H 1998 arXiv:quant-ph/9810003[27] Bose S, Vedral V and Knight P L 1999 Phys. Rev.A 60 194[28] Zhao Z, Yang T, Chen Y-A, Zhang A-N and Pan J-W 2003Phys. Rev. Lett. 90 207901[29] Clauser J F, Horne M A, Shimony A and Holt R A 1969 Phys.Rev. Lett. 23 880[30] Mattle K 1997 PhD Thesis Universität Innsbruck[31] Czeija G 2001 Master’s Thesis Universität Wien[32] Keller T and Rubin M 1997 Phys. Rev. A 56 1534[33] Pittman T B, Strekalov D V, Migdall A, Rubin M H,Sergienko A V and Shih Y H 1996 Phys.Rev.Lett.77 1917[34] Scarcelli G, Valencia A and Shih Y 2004 Phys. Rev. A70 051802 (http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0409210)[35] Weihs G 1994 Master’s Thesis Universität Innsbruck[36] Weinhold T J, Gilchrist A, Resch K J, Doherty A C, O’BrienJ L, Pryde G J and White A G 2008 arXiv:0808.0794[37] Kaltenbaek R, Blauensteiner B, Zukowski M, Aspelmeyer Mand Zeilinger A 2006 Phys.Rev.Lett. 96 240502(doi:10.1103/PhysRevLett.96.240502)[38] Sandia National Laboratories 2007, Snlo software, v.41,http://www.sandia.gov/imrl/XWEB1128/xxtal.htm[39] Kaltenbaek R 2008 PhD Thesis Universität Wien[40] Mosley P J, Lundeen J S, Smith B J, Wasylczyk P, U’Ren A B,Silberhorn C and Walmsley I A 2007 Heralded generation ofultrafast single <strong>photon</strong>s in pure quantum states (http://www.citebase.org/abstract?id=oai:arXiv.org:0711.1054)[41] U’Ren A B, Erdmann R K, de la Cruz-Gutierrez Mand Walmsley I A 2006 Phys. Rev. Lett. 97 223602(http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0611018)[42] Kim Y-H, Kulik S P and Shih Y 2001 Phys. Rev. A 63 060301(http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0007067)[43] U’Ren A B, Silberhorn C, Erdmann R, Banaszek K,Grice W P, Walmsley I A and Raymer M G 2005 LaserPhys. Lett. 15 146(http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0611019)[44] Aspect A, Grangier P and Roger G 1982 Phys.Rev.Lett.49 91[45] Greenberger D M, Horne M and Zeilinger A 2005 A belltheorem <strong>with</strong>out inequalities for two particles, usingefficient detectors (http://www.citebase.org/abstract?id=oai:arXiv.org:quant-ph/0510201)[46] Knill E, Laflamme R and Milburn G 2000 Nature 409 46[47] Kwiat P G, Mattle K, Weinfurter H, Zeilinger A, Sergienko Aand Shih Y 1995 Phys. Rev. Lett. 75 4337[48] Jennewein T 2002 PhD Thesis Universität Wien7

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!