{"id":"b27b30ef-a5d5-48b6-8f19-809c4608493f","arxiv_id":"2504.15258","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A birefringent interferometer with superconducting single-photon detectors performs broadband, time-resolved, spin-selective photoluminescence spectroscopy of quantum emitters, demonstrated on silicon carbide divacancies and diamond nitrogen-vacancy centers.","lead":"This paper shows that a compact interferometer paired with fast single-photon detectors can record the spectra of single quantum emitters across a wide wavelength range, roughly 600 to 1200 nanometers. The same setup can track how spectra change in time, letting the authors observe spin-dependent light emission from diamond defects using time bins as short as half a nanosecond.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uncalibrated FT wavelength axis shows a 2.4 nm offset from the grating reference, which directly undermines the flagship claim of identifying emitters from their zero-phonon line.","rationale":"The reader's weakest assumption is the interferometer calibration, and I agree that this is the most load-bearing issue. The 2.4 nm offset is concrete, visible in the paper's own data, and directly contradicts the text's 'comparable' characterization. The other main concern, that the sub-nanosecond timing claim is not shown in the displayed 10 ns bin data, is a presentation mismatch; the hardware timing resolution is plausibly adequate and the 500 ps bin data are mentioned as taken. The calibration error, by contrast, bears on the central application of emitter identification and could lead to concrete misidentification of defects. It is also readily testable: calibrating the wavelength axis against a known line is a standard step in FT spectroscopy. Since the issue is addressable and does not invalidate the demonstrated signals, the verdict should remain CONDITIONAL.","tokens_in":21204,"tokens_out":5788,"duration_ms":51081,"concrete_test":"Reanalyze the Section III interferogram data after calibrating the delay axis with a known line: e.g., set the FT ZPL peak to the grating value 1077.96 nm (or use a stabilized laser at a known wavelength) and apply the resulting linear correction to the wavelength scale. Then check the corrected FT ZPL center for the single divacancy and, independently, the NV singlet line in Section IV (expected ~1042 nm). If the corrected centers agree with the reference values to within ~0.3 nm, the offset is a constant calibration factor and the identification claim holds. If a wavelength-dependent residual remains, the birefringent dispersion model underlying the mapping is wrong and the wavelength axis must be rederived before quantitative emitter identification can be claimed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III reports the same single divacancy with a zero-phonon line at 1077.96 +/- 0.01 nm on the grating spectrometer and at 1075.6 +/- 0.1 nm on the FT spectrometer, a 2.36 nm shift that is roughly an order of magnitude larger than the FT fit uncertainty and comparable to the FT resolution of 2.54 nm. The text calls the center wavelengths 'comparable' and states this 'confirms the accuracy and resolution of the FT spectrometer in identifying quantum emitters in the IR region.' A resolution-broadened Lorentzian does not shift its center, so the offset indicates a systematic error in the delay-to-wavelength mapping, not a simple resolution artifact. The mapping is taken from the manufacturer's formula (Fig. 2(b)) with no in-situ calibration against a known spectral line. Because the paper's flagship application is identifying unknown defects from their ZPL wavelength, and common 4H-SiC divacancy ZPLs are separated by only about 1 nm (PL1/PL2 at 1133/1132 nm) to about 25 nm, a 2.4 nm systematic error is large enough to cause misidentification. The claim of accurate emitter identification is therefore not yet established, although the error may be correctable by calibrating the wavelength axis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the use of a compact common-path birefringent interferometer (TWINS) combined with superconducting nanowire single-photon detectors (SNSPDs) for Fourier-transform photoluminescence spectroscopy of quantum emitters in the near-infrared and telecom range. Simulations compare a grating spectrometer with an InGaAs camera against the FT approach for a single divacancy-like emitter. Experiments identify a single (kh) divacancy in 4H-SiC from its zero-phonon line in about 37 seconds, compare the FT spectrum with a grating reference, demonstrate parallel broadband detection of NV-center emission in the visible/near-infrared and infrared channels, and perform spin-selective time-resolved FT measurements using microwave initialization, extracting spin-dependent difference spectra with nanosecond time bins. The authors argue that FT spectroscopy with SNSPDs offers sensitivity, speed, and time-resolution advantages over InGaAs-camera grating spectrometers in the infrared.","tokens_in":21354,"tokens_out":5376,"duration_ms":49801,"significance":"If the claims hold, the work provides a practical route to broadband photoluminescence spectroscopy of single emitters at wavelengths where silicon detectors fail, using commercial components, and it adds time resolution to FT spectroscopy via single-photon timing. The demonstrations are benchmarked against known NV dynamics and a grating reference, and the paper is explicit about instrument parameters and measurement times. The main experimental results—single-divacancy zero-phonon-line identification in tens of seconds, parallel VIS/NIR and IR spectra, and spin-dependent difference dynamics—are internally consistent and would be of immediate use to the quantum emitter community. The accuracy of the wavelength axis and the scope of the sub-nanosecond claim need attention before the stated capabilities are fully supported.","major_comments":[{"comment":"The FT spectrum yields a zero-phonon line at 1075.6±0.1 nm while the grating measurement on the same divacancy gives 1077.96±0.01 nm; the 2.36 nm shift is much larger than the fit uncertainties and comparable to the FT resolution of 2.54 nm. A resolution-broadened Lorentzian does not shift its center, so this offset indicates a systematic error in the delay-to-wavelength mapping, which is taken from the manufacturer formula without in-situ calibration. Since the flagship application is emitter identification from the ZPL wavelength, and common divacancy lines are separated by about 1-25 nm, the current data do not establish the claimed wavelength accuracy. Please calibrate the axis against a known spectral line and re-report the identification claim with corrected values.","section":"Section III, Fig. 2(c)-(d)"},{"comment":"The abstract claims 'monitoring of spin-dependent spectral changes on sub-nanosecond timescales,' but the spin-selective experiments use time bins δt=10 ns and δt=500 ps, and the reported dynamics decay on roughly 1 μs and 5-7 μs scales. No sub-nanosecond spectral feature is actually resolved in the data. Please either demonstrate a sub-nanosecond spin-dependent change or soften the claim to state that the detector jitter permits sub-nanosecond binning in principle, while the demonstrated dynamics are nanosecond to microsecond.","section":"Abstract and Section V"},{"comment":"The claimed difference in decay timescales between the VIS/NIR channel (5-7 μs) and the IR channel (~1 μs) is stated without fits or confidence intervals. Because this is a central demonstration of spin-resolved FT spectroscopy, please provide quantitative fits (e.g., exponential or model-based) with reported uncertainties, or explicitly label the comparison as qualitative.","section":"Section V, Fig. 5(e)"}],"minor_comments":[{"comment":"The text states a microwave π pulse at 2995 GHz; throughout the paper the resonance is 2995 MHz. Please correct the unit.","section":"Section V, Rabi measurement"},{"comment":"The sentence 'a brightness of 1 Mcps, corresponding to 10 cps in a 10 µs pulse' is confusing; 1 Mcps over a 10 μs pulse yields 10 photons per pulse, not 10 cps. Please rephrase.","section":"Section V, temporal resolution paragraph"},{"comment":"Equation (1) adds the readout noise R to the mean photo-electron count before treating the result as a Poisson mean; this is not the standard noise model, in which readout noise is an additive Gaussian variance after the Poisson draw. The impact is minor for the order-of-magnitude comparison, but the model should be stated more carefully.","section":"Section II, Eq. (1)"},{"comment":"The factor 0.605 in the spectral-resolution formula is not derived or cited; please define all symbols (Δn, xmax, α) and give the origin of the numerical constant.","section":"Section III, Fig. 2(b)"},{"comment":"The VIS/NIR spectrum in Fig. 4(c) appears not to show the 637 nm zero-phonon line expected for NV−; please comment on whether it is suppressed by the ensemble measurement, the excitation conditions, or the spectral window shown.","section":"Section IV, Fig. 4(c)"}],"recommendation":"major_revision","confidential_remarks":"This is a solid experimental demonstration with commercial components, and the main technical claims are plausible. The wavelength-calibration offset is correctable and should be addressed before publication, as it bears directly on the emitter-identification application. The sub-nanosecond temporal-resolution claim should also be aligned with the data actually presented. I do not see grounds for rejection, but the revisions above are needed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing to know: this is a methods demonstration, and a reasonably convincing one. The genuinely new combination is common-path TWINS birefringent interferometry with SNSPDs for Fourier-transform photoluminescence spectroscopy of quantum emitters. The paper demonstrates single-divacancy ZPL identification in 4H-SiC in about 37 seconds, broadband parallel two-channel detection of NV ensembles (VIS/NIR and IR singlet emission), and spin-selective time-resolved FT spectra that match known NV dynamics. The simulations comparing grating+InGaAs versus FT+SNSPD for weak IR emitters are useful, and the sensitivity argument (Fellgett advantage plus low SNSPD dark counts) is honestly presented. I believe the core capability is real and available with commercial parts.\n\nThe soft spots, in proportion. The most serious is the uncalibrated wavelength axis. The same divacancy ZPL appears at 1075.6 ± 0.1 nm on the FT instrument and 1077.96 ± 0.01 nm on the grating spectrometer — a 2.4 nm offset, larger than the FT fit uncertainty and comparable to the FT resolution. The text calls them “comparable” and claims this confirms accuracy in identifying emitters. A resolution-broadened line does not shift its center; this looks like a systematic error in the manufacturer delay-to-wavelength formula. Since the flagship application is identifying unknown defects from ZPL position, and 4H-SiC divacancy lines can be about 1 nm apart, the claimed identification capability is not established until the axis is calibrated against a known line. This is correctable, but it is a real gap.\n\nThe sub-nanosecond headline is also oversold. Detector jitter is tens of picoseconds, but the displayed spin-resolved data use 10 ns bins and show microsecond-scale dynamics; 500 ps bins are mentioned but not shown. The claim that the system “enables monitoring on sub-nanosecond timescales” is plausible but not demonstrated in the displayed data. Third, the 12.6-day measurement is presented with no drift analysis and no data availability statement; less important, but worth requesting in revision.\n\nThe simulation section has some free parameters (brightness, camera noise, detection efficiency), but they are stated and reasonable. No circularity problem: the comparison is benchmarked on known spectra and a published NV model.\n\nWho this is for: experimental groups working on IR and telecom quantum emitters who need a practical alternative to InGaAs grating spectrometers. It deserves a serious referee. My recommendation: send it to review, with a request for in-situ wavelength calibration, a demonstration or softening of the sub-nanosecond claim, and a short drift/stability note.","headline":"Solid experimental methods paper: TWINS + SNSPD Fourier-transform PL spectroscopy is a real, useful combination, but the uncalibrated wavelength axis (2.4 nm ZPL offset) undercuts the flagship emitter-identification claim until fixed.","tokens_in":22071,"tokens_out":2134,"would_cite":true,"duration_ms":20006,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper shows that a compact birefringent interferometer plus superconducting detectors can measure single-emitter spectra in the infrared in seconds and add nanosecond-scale spin-resolved dynamics.","keywords":["Fourier transform spectroscopy","TWINS interferometer","superconducting nanowire single-photon detectors","quantum emitters","nitrogen-vacancy centers","divacancy in silicon carbide","spin-resolved photoluminescence","time-resolved spectroscopy"],"falsifier":"Measure the zero-phonon line of the same divacancy with the FT spectrometer and with a wavelength-calibrated grating spectrometer, then compare the centers; if the FT axis is offset by more than the resolution-limited uncertainty, the absolute-wavelength claim fails. A direct check uses a known narrow atomic or molecular line as a calibration source and asks whether the formula $\\Delta$-$\\lambda$ = 0.605 $lambda^{2}$/($\\Delta$-n times x times sin-$\\alpha$) reproduces it.","tokens_in":20907,"feed_emoji":"🔬","tokens_out":6935,"duration_ms":59614,"temperature":0.7,"pith_summary":"The paper tries to establish that time-domain Fourier transform spectroscopy, built from a compact birefringent interferometer and superconducting nanowire single-photon detectors, is a practical and advantageous way to measure the photoluminescence of single quantum emitters in the near-infrared and telecom range, where grating spectrometers with InGaAs cameras are noise-limited. It shows that a single divacancy center in silicon carbide can be identified from its zero-phonon line in tens of seconds to a few minutes, without the background subtraction an InGaAs setup needs. It also shows that the same instrument can record spectra and photon arrival times simultaneously, resolving spin-dependent emission dynamics of nitrogen-vacancy centers on nanosecond time bins. If the paper is right, laboratories studying telecom-range quantum emitters can obtain broadband, time-resolved spectra with commercial components and without liquid-nitrogen-cooled cameras.","feed_headline":"FT spectrometer identifies single emitters in under a minute","feed_subtitle":"A birefringent interferometer plus superconducting detectors records emitter spectra in tens of seconds and spin dynamics in nanoseconds.","key_machinery":"The device that carries the argument is the TWINS interferometer: a common-path birefringent interferometer in which two wedges of alpha-BBO create a controllable delay between two orthogonally polarized replicas of the input waveform, and a polarizer recombines them so that stepping the wedge delay produces an interferogram whose Fourier transform is the spectrum, following the Wiener-Khinchin theorem. Because both replicas travel the same path, the delay is stable and insensitive to vibration. The second half of the mechanism is the single-pixel SNSPD: its high efficiency and low dark count provide the sensitivity that InGaAs cameras lack, and its photon-arrival time tagging provides the temporal resolution, so each delay step yields a time-resolved count histogram that becomes a time-resolved spectrum after the Fourier transform.","core_discovery":"The central demonstration is that a common-path birefringent interferometer, the TWINS design, combined with superconducting nanowire single-photon detectors, reconstructs photoluminescence spectra by Fourier transforming an interferogram, and that this approach outperforms a grating spectrometer with an InGaAs camera for single emitters in the infrared. For a single divacancy in 4H-SiC, the FT instrument recovered the zero-phonon line in 37 seconds using 10 ms per delay step, while the grating instrument needed a five-minute background-subtracted exposure; simulations with a divacancy spectrum show that the grating camera cannot identify such a dim emitter even in five minutes. Broadband operation is shown by splitting the interferometer output with a dichroic and detecting the visible phonon sideband and the 1042 nm singlet emission of an NV ensemble in parallel. Finally, using photon arrival times binned at 10 ns and 500 ps, the paper reports spin-selective spectra: the difference between ms=0 and ms=1 emission decays on roughly 5-7 microseconds in the visible but about 1 microsecond for the 1042 nm peak, in line with the known NV dynamics.","pith_inferences":["A direct extension not demonstrated in the paper: the alpha-BBO interferometer is transparent to about 3 micrometers, so the same instrument should cover C-band telecom emitters such as erbium, T-centers, or G-centers by exchanging detectors and optics; this is an extrapolation, since the paper's measurements stop at 1200 nm.","The roughly 2.4 nm discrepancy between the FT and grating zero-phonon-line centers, if systematic, would weaken absolute emitter identification but would not affect the dynamical contrast results; an in-situ wavelength calibration would separate those cases.","The 12.6-day duration of the spin-resolved measurement implies that routine use on dim single emitters will need higher spin contrast, faster repetition, or compressed sensing; the paper mentions compressed sensing conceptually but does not quantify the speedup."],"forward_implications":["A single divacancy center in 4H-SiC can be identified from its zero-phonon line in about 37 seconds with the FT spectrometer, while the grating/InGaAs comparison requires a five-minute background-subtracted acquisition and still cannot recover a dim single emitter in simulation.","Because the interferometer output can be split onto multiple detectors, one scan yields spectra in separate bands simultaneously, demonstrated for the NV visible phonon sideband and the 1042 nm singlet emission.","Photon arrival times make every spectral point time-resolved; the paper demonstrates 10 ns and 500 ps bins and argues that the detector jitter, about 50 ps here and a few picoseconds in the best devices, sets the ultimate limit.","Spin-selective difference spectra resolve the known NV dynamics, with visible emission contrast decaying over roughly 5-7 microseconds and the 1042 nm contrast over about 1 microsecond, showing that the technique can separate emission pathways tied to different spin sub-levels."],"supporting_citations":[{"why":"Supplies the divacancy photoluminescence spectrum parameters, including zero-phonon-line wavelengths and Debye-Waller factor, used as simulation input and for emitter identification.","marker":"[46]"},{"why":"Supplies the NV center singlet-transition dynamics and the 1042 nm emission used to benchmark the spin-resolved time-resolved spectra.","marker":"[73]"},{"why":"Supplies the TWINS common-path birefringent interferometer design that the FT spectrometer is built around.","marker":"[64]"},{"why":"Gives the manufacturer's spectral-resolution formula that sets the FT instrument's wavelength axis and resolution.","marker":"[60]"},{"why":"Supplies the detector noise model, including dark noise, readout noise, and well capacity, used to simulate the InGaAs camera baseline.","marker":"[43]"},{"why":"Reports near-unity SNSPD efficiency, underpinning the sensitivity advantage over InGaAs cameras.","marker":"[52]"},{"why":"Documents few-picosecond SNSPD jitter, the stated limit for the temporal resolution claimed in the paper.","marker":"[58]"}],"fun_headline_variants":["Sub-ns FT spectra of quantum emitters via compact interferometer","Superconducting detectors boost FT spectroscopy to ns timescales","Birefringent FT captures emitter spectra in seconds with ns spin info","Fast, broadband FT spectroscopy of single emitters with ns resolution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The delay-to-wavelength calibration comes from the manufacturer's formula and is not checked in situ against a known spectral line, so the absolute wavelength of every FT spectrum inherits that unverified mapping; the paper's own data show a roughly 2.4 nm offset from the grating value for the same zero-phonon line.","fun_headline_variants_meta":{"raw":{"variants":["Sub-ns FT spectra of quantum emitters via compact interferometer","Superconducting detectors boost FT spectroscopy to ns timescales","Birefringent FT captures emitter spectra in seconds with ns spin info","Fast, broadband FT spectroscopy of single emitters with ns resolution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000818,"raw_usage":{"total_tokens":3571,"prompt_tokens":924,"completion_tokens":2647,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":540,"completion_tokens_details":{"reasoning_tokens":2572}},"tokens_in":540,"tokens_out":2647,"duration_ms":19673,"temperature":1.0,"reasoning_tokens":2572,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:30:45.924748+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the zero-phonon line of the same divacancy with the FT spectrometer and with a wavelength-calibrated grating spectrometer, then compare the centers; if the FT axis is offset by more than the resolution-limited uncertainty, the absolute-wavelength claim fails. A direct check uses a known narrow atomic or molecular line as a calibration source and asks whether the formula $\\Delta$-$\\lambda$ = 0.605 $lambda^{2}$/($\\Delta$-n times x times sin-$\\alpha$) reproduces it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the divacancy photoluminescence spectrum parameters, including zero-phonon-line wavelengths and Debye-Waller factor, used as simulation input and for emitter identification."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the NV center singlet-transition dynamics and the 1042 nm emission used to benchmark the spin-resolved time-resolved spectra."},{"cited_title":"Oriana, J","cited_arxiv_id":null,"evidence_quote":"Supplies the TWINS common-path birefringent interferometer design that the FT spectrometer is built around."},{"cited_title":"Perri, Journal of Physics B: Atomic, Molecular and Optical Physics 54, 113001 (2021)","cited_arxiv_id":null,"evidence_quote":"Gives the manufacturer's spectral-resolution formula that sets the FT instrument's wavelength axis and resolution."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports near-unity SNSPD efficiency, underpinning the sensitivity advantage over InGaAs cameras."},{"cited_title":"Korzh, Q.-Y","cited_arxiv_id":null,"evidence_quote":"Documents few-picosecond SNSPD jitter, the stated limit for the temporal resolution claimed in the paper."}],"review_version":1}