{"id":"39d0fe74-cf2b-4b8b-8a3f-ba6eaec1c39d","arxiv_id":"2504.17273","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Split-source electrodes allow independent tuning of parallel silicon single-electron pumps, yielding 4ef plateaus at 200 MHz and combined currents over 2 nanoamperes at 2.1 GHz.","lead":"A silicon chip with eight single-electron pumps, each with its own source electrode, was used to stack current from multiple pumps and reach over 2 nanoamperes. The method lets each pump be tuned independently, a step toward accurate nanoampere current standards for metrology.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 2 nA plateau is not quantitatively validated: no accuracy or uncertainty is reported for the 2.1 GHz three-pump stack, so the metrological reading of the central claim is unsupported.","rationale":"The paper convincingly demonstrates independent source-voltage tuning: the two-dimensional current maps in Figs. 2 and 3 show clear plateaus and boundaries that move as expected, and the four-pump 4ef plateau at 200 MHz in Fig. 3d is a real device-level achievement. The central issue is not the existence of the plateaus but what they mean for the abstract's metrology-oriented framing. The authors themselves state that the two-pump accuracy at 200 MHz is only about 10^-3, which is several orders of magnitude above the 0.1 ppm target mentioned in the introduction. At 2.1 GHz and with three pumps, no accuracy or uncertainty is reported, so the 2 nA plateau is not quantitatively tied to 6ef. Because plateau flatness in Fig. 4d does not by itself establish quantization, the strongest reading of the central claim (nanoampere-level current with the accuracy needed for a current standard) is not supported by the presented data. Secondary, the claim that parallelization is limited only by the number of integrated TSEPs is an extrapolation from four working pumps, and the note that one source line leaked and had to be floated is a reminder that the eight-pump device was not operated as an eight-pump array; this is less central than the missing accuracy at 2 nA. The reader's CONDITIONAL verdict correctly captures the situation: the core parallelization result appears credible, but the metrological claim requires the missing accuracy measurement. I therefore keep the verdict unchanged and agree with the reader's identification of the weakest assumption.","tokens_in":9350,"tokens_out":12843,"duration_ms":126592,"concrete_test":"Re-measure the Fig. 4d plateau with the same three TSEPs at 2.1 GHz using a calibrated ultrastable current amplifier or a substitution measurement against a reference current source. For at least five points across the plateau, integrate for 100 s, subtract the current measured with the RF drive off, and compute (I_pump - 6ef)/(6ef) with combined uncertainty. If the magnitude of the deviation exceeds the combined uncertainty or exceeds 10^-3 (the accuracy level already reported at 200 MHz), the plateau is not quantitatively established as a 6ef quantized current; if it is within both, the central claim survives this check.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing condition for the central claim is that the Fig. 4d current plateau exceeding 2 nA is a quantized pump-current plateau equal to 6ef (three pumps at 2.1 GHz), not merely a flat region of the measured current-voltage curve. The only accuracy quantification in the paper is at 200 MHz and for two pumps: in the text around Figs. 2d and 2e the authors report that the two-pump accuracy is approximately 10^-3, limited by the first TSEP, while the second reaches about 10^-5. No equivalent measurement is given for the 2.1 GHz three-pump configuration in Fig. 4d. The current readout is not calibrated against a reference, and no uncertainty bars, integration times, or offset-subtraction procedures are reported. The 2 nA stack includes a single TSEP operating on a 3ef plateau, so any error in multi-electron capture enters directly; the paper provides no evidence about the accuracy of that 3ef contribution. If the deviation of the plateau from 6ef is large, the literal claim of a plateau exceeding 2 nA could still hold while the metrology-oriented conclusion (a path to high-accuracy quantum current standards) would not follow. This is an unsupported extrapolation, not an internal inconsistency.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes and demonstrates a parallelization scheme for tunable-barrier single-electron pumps (TSEPs) in silicon, in which each pump has its own source electrode so that individual source voltages can be used to tune the dynamic-capture condition of each device. The authors show, at 200 MHz, that they can sequentially activate and optimize up to four parallel TSEPs and stack four ef plateaus into a 4ef current. They then operate three TSEPs at 2.1 GHz and report a stacked current plateau exceeding 2 nA (ef + 2ef + 3ef). The central claim is that split-source control provides a scalable and practical route to nanoampere-level current generation toward quantum current standard applications.","tokens_in":9520,"tokens_out":3745,"duration_ms":35721,"significance":"If the central claim holds, the split-source control scheme is a genuine practical advance: it allows independent tuning of parallel TSEPs with only one additional electrode per pump and no change to the shared entrance/exit gates. The experimental data in Figs. 2-4 clearly demonstrate controllable sequential activation, and the observed stacked plateaus up to 4ef at 200 MHz and exceeding 2 nA at 2.1 GHz are notable relative to prior parallel-pump demonstrations. The paper also benefits from a clean device concept and a straightforward tuning protocol. However, the metrology-oriented significance, which is emphasized in the abstract, hinges on the accuracy of the nanoampere-level plateaus, and the paper does not yet provide a quantitative accuracy or uncertainty assessment for the 2 nA plateau.","major_comments":[{"comment":"The central claim of a current plateau exceeding 2 nA is not quantitatively validated. The only accuracy quantification in the paper is the two-pump, 200 MHz measurement shown in Figs. 2d and 2e, where the combined accuracy is reported as approximately 10^-3. For the 2.1 GHz three-pump stack in Fig. 4d, no deviation from the ideal 6ef value, no uncertainty bars, and no integration-time or offset-subtraction information are provided. Since the abstract and conclusions explicitly connect this result to a path toward high-accuracy quantum current standards, the authors should either (i) provide a calibrated measurement of Ipump at the center of the 2 nA plateau, with a stated integration time and an uncertainty budget, or (ii) explicitly state that the accuracy at 2 nA has not yet been assessed and adjust the metrological claim accordingly.","section":"Fig. 4d and surrounding text"},{"comment":"The 2 nA plateau stacks ef + 2ef + 3ef contributions, so the accuracy of the individual multi-electron plateau from a single TSEP at 2.1 GHz is directly load-bearing for the claim. The paper provides no error-rate data for the 3ef plateau of any TSEP at 2.1 GHz. The statement in Sec. 2 that an ef + ef combination is expected to be more accurate than a single 2ef plateau does not apply to a 3ef plateau from one device, where multi-electron capture errors enter directly. Please quantify the deviation from the ideal 3ef value for the constituent TSEP, or temper the claim that the stacked 2 nA plateau is a path to high-precision current standards.","section":"Fig. 4d and Sec. 2"}],"minor_comments":[{"comment":"The analytical expression for the dynamic-capture boundary contains an unspecified constant and parameters (αe, αeB, αs, αsB, g, EC) that are not extracted from the present data. The paper states that the boundary slope 'is determined by the ratio of the prefactors' but does not quantitatively compare the predicted slope with the dashed line in Fig. 2a. Please either provide this comparison or explicitly state that the model is used only qualitatively.","section":"Supporting Information, 'Analytical Expression of the Pump Current'"},{"comment":"The sentence 'the limitation of the parallelization is only the number of TSEPs that can be integrated at once' is too strong given that the demonstration uses four TSEPs. Effects such as high-frequency line loading, gate cross-coupling, and device inhomogeneity could limit scaling; suggest rewording to indicate that the design is in principle extendable.","section":"Sec. 3, near Fig. 3d"},{"comment":"The definition of 'accuracy' (approximately 10^-3 and about 10^-5) is not stated. Please specify whether this is |1 - Ipump/(ef)| or |2 - Ipump/(ef)|, and describe how it was measured, including integration time and number of repetitions.","section":"Sec. 2, Figs. 2d and 2e"},{"comment":"The floating of the S5 terminal due to a current leak is described only in the Supporting Information. Since Figs. 3 and 4 use a subset of TSEPs, the main text should state which terminals were active and that the floating S5 did not contribute to the measured currents, so that readers can interpret the four-pump and three-pump data without reading the supplement.","section":"Supporting Information, 'Measurement Setup'"}],"recommendation":"major_revision","confidential_remarks":"The experimental demonstration is plausible and the split-source concept is attractive, but the gap between the abstract's metrological promise and the missing accuracy assessment for the 2 nA plateau is the key issue to resolve. This is a fixable problem (additional measurement or a more cautious claim), so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a solid experimental demonstration that splitting the source electrode gives independent control of parallel silicon TSEPs, and with that control the authors stack ef, 2ef, and 3ef plateaus from three pumps at 2.1 GHz to exceed 2 nA. That's a real first for parallel single-electron pumps, and the data in Figs. 2–4 look clean. The tuning procedure is practically useful and easy to reproduce: find the optimal V-exit for the first TSEP, then use the source voltage to bring up each additional pump sequentially.\n\nWhat's genuinely new is the split-source control itself. Previous parallelization used separate wiring, extra gates, or addressable exit gates; here just splitting the source and using its DC bias as an independent knob is simpler, and the paper shows it works for up to four pumps at 200 MHz. The current maps and line cuts are consistent and support the claim of independent activation and stacking.\n\nThe soft spot is exactly where the stress-test note lands: no accuracy is reported for the 2 nA plateau. The only quantitative accuracy statement is for two pumps at 200 MHz, about 1e-3, limited by the least accurate TSEP. The 2.1 GHz three-pump stack—which is the centerpiece—has no error bars, no integration time, no calibration against a reference amplifier, and no estimate of how far the plateau deviates from 6ef. Since the stack includes a single pump running on a 3ef plateau, any multi-electron capture errors enter directly. So the paper can legitimately claim 'current exceeding 2 nA,' but not that this is an accurate current at the level needed for quantum metrology. The phrase 'path toward advancing high-accuracy quantum current standards' is fair as a future direction, not as a demonstrated outcome.\n\nAlso note the scalability claim says the limitation is only the number of pumps that can be integrated, but only four of eight devices are used and one source line had a leak. That's an extrapolation, though a modest one.\n\nIn sum: the experimental method is credible, the central demonstration holds up, and the missing accuracy data are a clear limitation rather than a fatal flaw. The paper deserves a serious referee. I'd ask the authors to add uncertainty analysis for the high-current plateaus or soften the metrology claims.","headline":"Split-source control is a genuinely useful new knob for parallelizing silicon TSEPs, and the >2 nA stacked plateau is a first, but the accuracy at that current is unmeasured, so the metrology claims outrun the data.","tokens_in":10125,"tokens_out":2327,"would_cite":false,"duration_ms":22141,"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":"Four parallel silicon single-electron pumps, each with its own source electrode, can be tuned independently and their stacked currents exceed 2 nA.","keywords":["single-electron pump","tunable barrier","silicon quantum dot","parallelization","split-source electrode","nanoampere current","quantum current standard","dynamic capture"],"falsifier":"Compare the plateau current of the three-pump 2.1 GHz stack against an independent current reference and check the deviation from the ideal $6ef$ (about 2.018 nA); a deviation far above the sub-0.1 ppm target, or a plateau that is not flat when each source voltage is individually swept, would show that plateau flatness does not certify accuracy.","tokens_in":9077,"feed_emoji":"⚡","tokens_out":9819,"duration_ms":84026,"temperature":0.7,"pith_summary":"This paper claims that the obstacle to parallelizing single-electron pumps—each pump wants slightly different voltage settings—can be removed by giving every pump its own source electrode and tuning those source voltages independently. Using silicon tunable-barrier single-electron pumps, devices that transfer one electron per clock cycle, the authors operate four pumps in parallel at 200 MHz and stack four $ef$ plateaus into a single $4ef$ current. At 2.1 GHz they stack the $ef$, $2ef$, and $3ef$ plateaus of three pumps and reach currents above 2 nA, an output that would require a single pump running at roughly 12.6 GHz. Reaching the nanoampere range matters because current measurement accuracy degrades as current falls, and a quantum current standard needs accurate current in exactly this range.","feed_headline":"Four silicon pumps stack currents past 2 nA","feed_subtitle":"Each pump gets its own source voltage, so parallel single-electron pumps can finally reach the nanoampere regime.","key_machinery":"The load-bearing object is the split-source electrode: an array of eight silicon pumps shares the entrance gate, exit gate, and drain, but each pump has its own source terminal $S_n$ with an independent DC voltage $V_{Sn}$. The theory used here is dynamic capture: as the entrance barrier rises, the quantum-dot level crosses the source Fermi level, and the electron either tunnels back to the source or is captured; the source voltage sets where that crossing happens, hence the escape rate $\\Gamma_1$ and the capture probability $P_1$. The paper takes the analytical form $P_1 = \\exp[-\\exp(\\cdots)]$ for the capture probability from Ref. [33]; its level lines in the $(V_{Sn}, V_{\\mathrm{exit}})$ plane give the slanted dynamic-capture boundary observed in the current maps. That boundary is the tuning handle: fixing the first pump's optimum from the log-deviation curve, then sweeping the next pump's $V_{Sn}$ while reading $\\log_{10}|m - I_{\\mathrm{pump}}/ef|$ for the stacked level $m$, selects each pump's operating point in order.","core_discovery":"The central discovery is that a single split-source electrode per pump is enough to give each parallel tunable-barrier single-electron pump independent, high-precision tuning. In these silicon devices the source voltage shifts the Fermi level of the source relative to the rising quantum-dot level, which changes the escape rate during dynamic capture; a more positive source voltage lowers the Fermi level, makes the capture decision happen earlier, increases back-tunneling, and lowers the probability of capturing one electron. That makes the source voltage both a switch, since a large positive bias disables a pump, and a trimmer, since small adjustments set each pump exactly on its optimal $ef$ plateau. With this control the authors sequentially tune four pumps at 200 MHz to produce a $4ef$ plateau, and three pumps at 2.1 GHz to produce $ef+2ef+3ef$ current exceeding 2 nA.","pith_inferences":["The authors leave implicit that source and exit-gate controls move different boundaries, so the two-dimensional current maps can serve as a lookup table for programming arbitrary integer stacks $mef$ per pump, not only $1ef$ each.","A testable extension is to close the loop: read the combined current and feed small corrections back to individual source voltages, turning the manual sequential tuning into a self-calibrating array that tracks drift.","Because the analytical capture-probability expression ties the boundary slope to capacitive-coupling ratios, measuring that slope across many devices would show whether the required source-voltage trim is predictable from layout, a useful input for designing much larger arrays."],"forward_implications":["Current scales by adding pumps rather than raising frequency: four pumps at 200 MHz already form a clean $4ef$ plateau.","Nanoampere output is reached at a modest clock frequency: three pumps at 2.1 GHz deliver what a single pump would need about 12.6 GHz to supply.","The tuning protocol is sequential and scalable: each new pump's source voltage is optimised against the already-stacked current, so the number of tuning steps grows with the number of pumps.","If the individual pumps reach sub-0.1 ppm accuracy, the same control scheme yields a metrology-ready current source in the nanoampere range."],"supporting_citations":[{"why":"Defines the tunable-barrier single-electron pump and its $I=mef$ output, the object being parallelized.","marker":"[1]"},{"why":"Supplies the log-deviation method used to locate the optimal operating point of the first pump.","marker":"[17]"},{"why":"Establishes the roughly 2 GHz upper limit for electrically defined quantum-dot pumps, the frequency constraint that motivates parallelization.","marker":"[18]"},{"why":"Reports nanoampere current from a trap-level pump at 7.4 GHz, the prior current-level benchmark.","marker":"[24]"},{"why":"States that optimal conditions differ among individual pumps, the problem the split-source tuning solves.","marker":"[25]"},{"why":"Represents the previous strategy of wiring each pump separately, the scalability comparison for the proposed shared-gate split-source layout.","marker":"[27]"},{"why":"Represents an earlier individually addressable exit-gate parallelization scheme that this work extends.","marker":"[31]"},{"why":"Provides the dynamic-capture model that underlies the source-voltage control of capture probability.","marker":"[32]"},{"why":"Gives the analytical capture-probability expression whose level lines set the dynamic-capture boundary in the current maps.","marker":"[33]"}],"fun_headline_variants":["Split-source electrodes let silicon pumps scale to 2 nA","Parallel silicon pumps hit 2 nA with single-electrode control","One tuning electrode per pump: silicon pumps reach 2 nA","Silicon pumps go parallel, split-source control unlocks 2 nA","Scalable silicon pumps: split gates enable nanoampere currents"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the flat stacked plateaus—especially the >2 nA plateau at 2.1 GHz—represent accurate current at the metrology level; the paper reports an error estimate near one part in a thousand only for two pumps at 200 MHz and does not quantify how far the 2 nA plateau deviates from its ideal value.","fun_headline_variants_meta":{"raw":{"variants":["Split-source electrodes let silicon pumps scale to 2 nA","Parallel silicon pumps hit 2 nA with single-electrode control","One tuning electrode per pump: silicon pumps reach 2 nA","Silicon pumps go parallel, split-source control unlocks 2 nA","Scalable silicon pumps: split gates enable nanoampere currents"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000434,"raw_usage":{"total_tokens":2179,"prompt_tokens":880,"completion_tokens":1299,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":496,"completion_tokens_details":{"reasoning_tokens":1210}},"tokens_in":496,"tokens_out":1299,"duration_ms":9614,"temperature":1.0,"reasoning_tokens":1210,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T10:43:47.919923+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compare the plateau current of the three-pump 2.1 GHz stack against an independent current reference and check the deviation from the ideal $6ef$ (about 2.018 nA); a deviation far above the sub-0.1 ppm target, or a plateau that is not flat when each source voltage is individually swept, would show that plateau flatness does not certify accuracy.","supporting_citations":[{"cited_title":"Kaestner and V","cited_arxiv_id":null,"evidence_quote":"Defines the tunable-barrier single-electron pump and its $I=mef$ output, the object being parallelized."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the log-deviation method used to locate the optimal operating point of the first pump."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the roughly 2 GHz upper limit for electrically defined quantum-dot pumps, the frequency constraint that motivates parallelization."},{"cited_title":"Yamahata, S","cited_arxiv_id":null,"evidence_quote":"Reports nanoampere current from a trap-level pump at 7.4 GHz, the prior current-level benchmark."},{"cited_title":"Norimoto, P","cited_arxiv_id":null,"evidence_quote":"States that optimal conditions differ among individual pumps, the problem the split-source tuning solves."},{"cited_title":"Nakamura, D","cited_arxiv_id":null,"evidence_quote":"Represents the previous strategy of wiring each pump separately, the scalability comparison for the proposed shared-gate split-source layout."},{"cited_title":"Mirovsky, B","cited_arxiv_id":null,"evidence_quote":"Represents an earlier individually addressable exit-gate parallelization scheme that this work extends."},{"cited_title":"Kashcheyevs and B","cited_arxiv_id":null,"evidence_quote":"Provides the dynamic-capture model that underlies the source-voltage control of capture probability."},{"cited_title":"Yamahata, N","cited_arxiv_id":null,"evidence_quote":"Gives the analytical capture-probability expression whose level lines set the dynamic-capture boundary in the current maps."}],"review_version":1}