---------------------------------------------------------------------- Report of Referee A -- NX10171/Lee ---------------------------------------------------------------------- This paper reports the synchronization of separated clocks using single photons generated via SPDC. The interference of single photons ensures security and the time offsets observed as g(2) peaks supply timing information. It contains experimental efforts and supporting theory. It is of interest to the quantum optics and timing communities. With some revision I believe it is possible to recommend publication in Physical Review Applied. Response to questions from the editor: Is the main achievement of this study obvious to you, and significant? The authors have simplified their work presented in Appl. Phys. Lett. 114, 101102 (2019) by removing the remote SPDC source. This can be viewed as a positive since it allows the system to function as a server/client network. Is the emphasis sufficiently on application? The immediate and long term utility of this work is not as clear to me as it could be. This method is secure against symmetric delay attacks but, it is not clear that it is compatible with a large scale network architecture that would be worth attacking. The significance of such attacks is also not clear, some references could help this. Star network (worth attacking?) MDI-QKD: Measurement-device independent quantum key distribution over untrustful metropolitan network classical: https://link-springer-com.libproxy1.nus.edu.sg/chapter/10.1007/978-3-642-35795-4_39, telecommunications frequency, references of symmetric delay attacks Questions and comments to the authors: What is the spatial separation limit to clock comparison? Sub-questions: (a) Since this method relies on single photons and count rates adequate to achieve good signal to noise, is the range of clock synchronization intrinsically limited by optical losses in the fiber? (Would any subsequent optical amplification destroy the correlations necessary for this protocol?) 3dB per 10km... compare to max time of temp drift. (b) Is there an optimum reflectivity (as opposed to the effort free 3.5%) for the distant fiber tip? Optimal R for balancing precision for foward and backward measurement. In-line partial reflectors https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=10784 What is the limit to timing accuracy imposed by this method? Is it adequate for the comparison/synchronization of Cs beam clocks, Masers, or optical clocks? Compare: precision in DT < clock fluctuation in DT < clock drift in DT (see frequency inaccuracy) < timing jitter It would be nice to quantify the statement “Photon pairs emerging from SPDC are tightly time correlated [9].” (DONE) What is the APD jitter in “The SNSPD has relatively low jitter (~40ps) compared to APDs, and allows Alice to measure the round-trip time more accurately regardless of the choice of detector by the remote party” InGaAs APD jitter ~300ps. (DONE) Figure 4 and discussion: The time variant of the Allan deviation “TDEV” is a more conventional measure of timing stability than the “Precision” defined in the text (the two are qualitatively similar). It should be relatively easy to reprocess the data; the program Stable 32 is freely available from the IEEE and is accompanied by substantial documentation http://www.stable32.com/ ---------------------------------------------------------------------- Report of Referee B -- NX10171/Lee ---------------------------------------------------------------------- In the manuscript, NX10171, authors, Jianwei Lee, et al., demonstrated absolute clock synchronization using a single time-correlated photon pair source. In the protocol, entangled photons are distributed to two communicating parties, and coincidence measurement between two communicating parties has been used to establish the clock synchronization. Also, a fraction of the transmitted photons is sent back to probe the symmetric channel delay attack in the quantum channel. Experimental tests have been conducted to demonstrate the distance-independence clock synchronization over 10 km with the same and different reference clocks. This experiment is very interesting, and the results sound good. However, the technical issues put serious limitations on the system performance. First, time jitter in the single-photon detector and coincidence electronics set the ultimate limitation of the timing synchronization precision, which is more than 40 ps. The jitter of the underlying devices contribute, but do not ultimately limit the precision. Second, the variation of the channel length due to the temperature drift will further increase time jitter. Unfortunately, the coincidence rate of entangled sources is too low to apply any feedback for improving the performance. We still managed to recover meaningful clock parameters despite the temperature drifts Third, in the present experiment, the precision of timing synchronization is almost 100 ps. But, in the conventional ultrafast laser techniques, timing jitter has reached a precision of sub-fs regime in a multi-km fiber network. So, the performance of the present experiment is about 5 orders of magnitude worse than the one of the conventional ultrafast technologies, illustrating significant drawback. So, unless some kind of quantum advantage has been demonstrated, I will not be able to recommend its publication in Physical Review Applied. We thank the reviewer for highlighting the difference in synchronization precision between our present work and those using ultrafast laser techniques. First, we would like to point out that the quoted sub-fs precision applies only to frequency transfer [1] and not to time and frequency transfer scenarios. The latter, to our knowledge, has only achieved a precision of 0.3ps in 5s [2]. Thus the difference in precision between the two techniques is not as extreme as asserted by the reviewer. Second, the use of time-correlated photon pair sources is pursued as they can be reconfigured to produce entangled photon pairs e.g. Ref. [4] which are useful for authenticating the synchronization signal (described in Section X in our work). This promise has thus far motivated similar work such as Ref. [3] and [5]. However, we chose not to demonstrate a Bell violation over multi-km using a source of comparable performance since a recent work in Ref. [6] has already reported an equivalent result where a QBER of 6.4% (QBER < 10% is equivalent to a Bell violation) was observed for a BBM92 quantum key distribution scheme deployed over 10km. Similarly, previous work focused not on demonstrating a Bell violation useful for authentication but rather on the practicality of using time-correlated photon pairs to synchronize widely used frequency standards, such as those based on Rubidium and Hydrogen-masers. For example, Ref. [5] focused on demonstrating that time-correlated photon pairs could be used for time and frequency transfer at a precision comparable to the frequency instability of independent clocks, and Ref [3] focused on using frequency entangled photon pairs to improve the precision of time transfer via quantum nonlocal dispersion cancellation. In the same spirit, the current work focuses on demonstrating a level of precision suitable for synchronizing widely used Rb clocks, but in a client/server configuration suitable for synchronizing multiple parties where previously such a scheme would require each party to possess a photon pair source [5]. [1] Xin, M., Şafak, K., Peng, M. Y., Callahan, P. T., Kalaydzhyan, A., Wang, W., ... & Hawthorne, T. (2018). Sub-femtosecond precision timing synchronization systems. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 907, 169-181. [2] Lessing, M., Margolis, H. S., Brown, C. T. A., & Marra, G. (2015, May). Simultaneous time and frequency transfer over a 158-km-long fiber network using a mode-locked laser. In 2015 Conference on Lasers and Electro-Optics (CLEO) (pp. 1-2). IEEE. [3] Hou, F., Quan, R., Dong, R., Xiang, X., Li, B., Liu, T., ... & Zhang, S. (2019). Fiber-optic two-way quantum time transfer with frequency-entangled pulses. Physical Review A, 100(2), 023849. [4] Kim, T., Fiorentino, M., & Wong, F. N. (2006). Phase-stable source of polarization-entangled photons using a polarization Sagnac interferometer. Physical Review A, 73(1), 012316. [5] Lee, J., Shen, L., Cerè, A., Troupe, J., Lamas-Linares, A., & Kurtsiefer, C. (2019). Symmetrical clock synchronization with time-correlated photon pairs. Applied Physics Letters, 114(10), 101102. [6] Shi, Y., Moe Thar, S., Poh, H. S., Grieve, J. A., Kurtsiefer, C., & Ling, A. (2020). Stable polarization entanglement based quantum key distribution over a deployed metropolitan fiber. Applied Physics Letters, 117(12), 124002. We agree with the reviewer that the present technique using time-correlated photons does not match the precision of conventional ultrafast laser techniques. Nevertheless, the use of time-correlated photon pairs in synchronization has been pursued thus far due to its promise of enhancing security through the use of a Bell test as a means to authenticate the origin of the photons. The approach of demonstrating synchronization as a proof-of-principle concept independent of demonstrating security has been adopted by several papers, including [1] [1] Fiber-optic two-way quantum time transfer with frequency-entangled pulses