# Fileset

[chen-et-al-detecting-dna-translocation-through-a-nanopore-using-a-van-der-waals-heterojunction-diode.pdf](https://mdr.nims.go.jp/filesets/8eef21c6-172d-48de-ba20-7c4befda5e2f/download)

## Creator

Sihan Chen, Siyuan Huang, Jangyup Son, Edmund Han, [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), Pinshane Y. Huang, William P. King, Arend M. van der Zande, Rashid Bashir

## Rights

[Creative Commons BY-NC-ND Attribution-NonCommercial-NoDerivs 4.0 International](https://creativecommons.org/licenses/by-nc-nd/4.0/)

## Other metadata

[Detecting DNA translocation through a nanopore using a van der Waals heterojunction diode](https://mdr.nims.go.jp/datasets/62f6416f-d665-4eb3-ae50-09d24a8c604f)

## Fulltext

Detecting DNA translocation through a nanopore using a van der Waals heterojunction diodePNAS  2025  Vol. 122  No. 18 e2422135122� https://doi.org/10.1073/pnas.2422135122 1 of 8RESEARCH ARTICLE | Significance Nanopores are nanometer-sized openings that enable detailed analysis of molecular structures by analyzing individual biomolecules as they pass through. Solid-state nanopores integrated with local electronic sensors hold promise for high-throughput, label-free, single-molecule sensing. However, integrating solid-state nanopores with local electrical sensing modalities remains challenging, with successful implementations limited to in-plane architectures such as tunneling junctions and transistors. This work demonstrates the integration of a nanopore with a vertical two-dimensional heterojunction diode, enabling electrical sensing of DNA translocation through the nanopore via interlayer tunneling current, opening the possibility for out-of-plane electrical sensing and control of single biomolecules.The authors declare no competing interest.This article is a PNAS Direct Submission. M.T. is a guest editor invited by the Editorial Board.Copyright © 2025 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).1Present address: Functional Composite Materials Research Center, Korea Institute of Science and Technology, Wanju-gun 55324, Jeonbuk, Republic of Korea.2Present address: Division of Nanoscience and Technology, Korea Institute of Science and Technology School, University of Science and Technology, Seoul 02792, Republic of Korea.3Present address: Department of Jeonbuk National University-Korea Institute of Science and Technology Industry-Academia Convergence Research, Jeonbuk National University, Jeonju 54896, Jeonbuk, Republic of Korea.4To whom correspondence may be addressed. Email: arendv@illinois.edu, or rbashir@illinois.edu.This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.​2422135122/-/DCSupplemental.Published May 1, 2025.APPLIED PHYSICAL SCIENCESDetecting DNA translocation through a nanopore using a van der Waals heterojunction diodeSihan Chena , Siyuan Huangb, Jangyup Sonb,1,2,3, Edmund Hanc , Kenji Watanabed, Takashi Taniguchie, Pinshane Y. Huangc,f , William P. Kinga,b,c,f,g ,  Arend M. van der Zandea,b,c,f,4 , and Rashid Bashira,b,c,f,g,h,i,j,4Affiliations are included on p. 7.Edited by Mehmet Toner, Massachusetts General Hospital, Charlestown, MA; received October 25, 2024; accepted April 4, 2025 by  Editorial Board Member John A. RogersA long-unrealized goal in solid-state nanopore sensing is to achieve out-of-plane electrical sensing and control of DNA during translocation, which is a prerequisite for base-by-base ratcheting that enables DNA sequencing in biological nanopores. Two-dimensional (2D) heterostructures, with their capability to construct out-of-plane electronics with atomic layer precision, are ideal yet unexplored candidates for use as electrical sensing mem-branes. Here, we demonstrate a nanopore architecture using a vertical 2D heterojunction diode consisting of p-type WSe2 on n-type MoS2. This diode exhibits rectified interlayer tunneling currents modulated by ionic potential, while the heterojunction potential reciprocally rectifies ionic transport through the nanopore. We achieve concurrent detection of DNA translocation using both ionic and diode currents and demonstrate a 2.3-fold electrostatic slowing of average translocation speed. Encapsulation layers enhance chemical and mechanical stability and durability while preserving the spatial resolution of atomically sharp 2D heterointerface for sensing. These results establish a paradigm for out-of-plane electrical sensing of single biomolecules.nanopore | van der Waals heterojunction | DNA | single molecule | ion transport Nanopore sensors, with a sensing volume comparable to the analyte size, are a powerful tool for single-biomolecule analysis ( 1 ,  2 ). Eighteen years ago, IBM introduced the idea of out-of-plane electrical sensing and control using a “DNA transistor” ( 3 ). This design employs electrodes separated by a thin dielectric to control and sense DNA translocation through a nanopore. However, limitations in thin-film processes hindered the scaling of the transistor to the molecular thicknesses needed for sensing ( 4 ). Vertically stacked two-dimensional (2D) heterostructures provide layer-by-layer control to construct out-of-plane electronics ( 5 ,  6 ), opening possibilities beyond single 2D materials for sensing with out-of-plane electric fields and interlayer currents. Unlike three-dimensional (3D) diodes with a depletion region, vertical 2D junctions exhibit an atomically sharp energy band discontinuity at their heterointerface ( 6 ), which represents the ultimate limit for DNA transistors, with the conducting layers separated by only an angstrom-sized van der Waals gap. These attributes make vertical 2D heterojunction diodes an ideal candidate for use as electrical sensing membranes. In this work, we present a nanopore architecture integrated with a vertical 2D hetero-junction diode from p-type WSe2  on n-type MoS2 , referred to as the HJD-NP sensor. We first electrically characterized the heterojunction to demonstrate the p-n diode behavior, then characterized the transport characteristics of both the ionic and diode channels in the absence of analytes, and finally demonstrated DNA sensing using an HJD-NP sensor. Three key advances from these proof-of-principle experiments include: i) concurrent detection of DNA translocation through a vertical diode nanopore using both ionic and diode currents, ii) slowing of DNA translocation via electrostatic interactions between the DNA and diode, and iii) rectified ionic transport controlled by heterojunction poten-tial. These results highlight the potential of HJD-NP sensors for sensitive single-biomolecule analysis with high spatial resolution and extended interrogation time. Results and DiscussionNanopore Integrated with a Vertical 2D Diode. Fig. 1 presents the basic concept of the HJD-NP sensors for DNA sensing. As illustrated in Fig. 1A, a nanopore is formed in an electrically contacted 2D heterojunction, which consists of a p-type WSe2 flake stacked on an n-type MoS2 flake. This heterojunction forms an out-of-plane p-n diode. When a double-stranded DNA (dsDNA) molecule translocates through the nanopore under an Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.https://creativecommons.org/licenses/by-nc-nd/4.0/https://creativecommons.org/licenses/by-nc-nd/4.0/mailto:arendv@illinois.edumailto:rbashir@illinois.eduhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2422135122/-/DCSupplementalhttps://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2422135122/-/DCSupplementalhttps://orcid.org/0000-0003-1216-6422https://orcid.org/0000-0002-1910-9052https://orcid.org/0000-0002-1095-1833https://orcid.org/0000-0001-8606-1290mailto:https://orcid.org/0000-0001-5104-9646mailto:https://orcid.org/0000-0002-7225-9180http://crossmark.crossref.org/dialog/?doi=10.1073/pnas.2422135122&domain=pdf&date_stamp=2025-4-282 of 8   https://doi.org/10.1073/pnas.2422135122� pnas.orgexternal ionic bias Vionic, the sensor simultaneously interrogates the conventional ionic current Iionic through the pore and the electrical current Id across the diode under a drain-source bias Vds. The detailed device fabrication procedures are described in Materials and Methods  and SI Appendix, Fig. S1 . Briefly, we started the fabrication with HfOx  membranes formed by atomic layer deposition (ALD) (SI Appendix, Fig. S2 ), which is hydrophilic ( 7 ) and durable in salt solutions ( 8 ). Then we transferred 2 to 5 nm (2 to 7 layers) thick MoS2  and WSe2  flakes onto the membrane to form the heterojunction stack, with a 1 to 10 µm2  overlap region. While thinner diodes are preferred for better spatial reso-lution, it remains challenging to fabricate working diodes from monolayer 2D semiconductors, primarily limited by p-type con-tacts ( 6 ). Therefore, we used few-layer structures instead of mon-olayers. Next, we fabricated n-type and p-type ohmic contacts to MoS2  and WSe2  using evaporated Ni/Au contacts and transferred Au contacts, respectively (SI Appendix, Fig. S3 ). Subsequently, we transferred a multilayer hBN flake to fully encapsulate the WSe2 /MoS2  heterostructure ( 9 ). This hBN buffer layer preserves the electrical conductivity of p-type WSe2  during ALD ( 10 ). Afterward, we deposited a layer of ALD HfOx  for electrical insu-lation and stable operation. Finally, we drilled a nanopore in the overlap region through the membrane stack using a focused elec-tron beam in STEM mode, which minimizes electron-beam induced damage to electronic materials compared to TEM mode ( 11 ).  Fig. 1B   shows the STEM image of an example nanopore. Further details on nanopore drilling are provided in SI Appendix, Supplementary Text 1  and Fig. S4 . We confirmed the structure of the heterojunction and cleanli-ness of the interface between layers with cross-section STEM.  Fig. 1C   shows a representative membrane stack, which consists of HfOx , hBN, five-layer WSe2 , five-layer MoS2 , and HfOx  from top to bottom. The heterointerface between WSe2  and MoS2  is atom-ically clean, which is critical for efficient interlayer charge carrier transport ( 12 ). Raman microscopy in SI Appendix, Fig. S5  further suggests interlayer coupling between WSe2  and MoS2 . To use the heterojunction current for sensing, we must first understand the electrical transport properties and sensitivity of the van der Waals heterojunction. Due to defects incorporated during synthesis and environmental interactions ( 13 ,  14 ), few-layer WSe2  is lightly p-doped while few-layer MoS2  is heavily n-doped, as confirmed by FET transport measurements in SI Appendix, Fig. S3 . Vertical stacking of p-WSe2  on n+-MoS2  forms a p−n diode with a type-II energy band alignment ( Fig. 2A  ). Under forward bias, Id  is dominated by interlayer recombination between the majority carriers of WSe2  and MoS2  over diffusion current, and Id  is maximized when the hole density of WSe2  and the electron density of MoS2  are balanced ( 5 ,  15 ).  Fig. 2B   shows the Id −Vds  curve of an example HJD-NP sensor measured in dry air in dark, exhibiting forward rectification characteristic of a p-n diode.         However, unlike an ideal vertical p-n diode, the total van der Waals heterojunction resistance Rds  also includes series resistances ( Fig. 2C  ), which consist of the contact resistances to WSe2  and MoS2 , as well as the channel resistances of WSe2  and MoS2  in both the non-overlapping and overlapping regions. We establish that the p-n junction is the dominant resistance in the system using photocurrent microscopy (SI Appendix, Fig. S6A ). Since photocurrent is sensitive to the electric field gradient within the junction, its magnitude is greatest at the junction with the highest resistance.  Fig. 2D   shows the photocurrent map for an example device at Vds  = 0 mV. The strongest photo response came from the MoS2 /WSe2  overlap region, confirming the formation of a p-n junction ( 5 ). SI Appendix, Fig. S6B  shows additional photocurrent maps at Vds  ≠ 0 mV for this device, which also supports this find-ing. Other heterojunction devices exhibited a strong photo response from the Au–WSe2  junction or the WSe2  channel region (SI Appendix, Fig. S6 C  and D ) and/or non-rectifying Id −Vds  char-acteristics (SI Appendix, Fig. S7 ). We screened all diodes using photocurrent microscopy and selected those with both strong rectification and a dominant photocurrent in the overlap region for sensing measurements. Next, we studied the device behavior in an ionic environment to identify the optimal measurement parameters for electrical sensing. There are two control knobs for device operation, namely Vionic  and Vds .  Fig. 2E   illustrates the coupling between the ionic and diode channels, which is comparable to that of FET-nanopore sensors ( 16       – 20 ), except here the sensing element is a vertical van der Waals heterojunction diode rather than a planar FET. The ionic voltage controls the heterojunction current transport via electrostatic gating. As shown in  Fig. 1A  , the backside HfOx  serves as the gate dielectric, while the front side is grounded to minimize electrochemistry.  Fig. 2 F  and G   shows the output and transfer characteristics of the heterojunction in an HJD-NP device, respec-tively. The current rectification improved with a more negative Vionic , indicating that the p-WSe2  partly limits the charge transport in the heterojunction and that the contact and channel resistances of the n+-MoS2  are comparably small. Liquid gating is insensitive to salt concentrations, as the gate capacitance of the electric double layers far exceeds that of the backside HfOx . The FET transport characteristics of a backgated p-WSe2 /n+-MoS2  heterojunction on a SiO2 /Si substrate also supports these findings (SI Appendix, Fig. S8 ). Therefore, p-WSe2 /n+-MoS2  heterojunction under for-ward bias can be modeled as a vertical p-n diode in series with an additional p-FET. We assume a uniform electric potential of αVd  for WSe2  and 0 mV for MoS2  in the overlap region, where α  is A B CFig. 1.   Device architecture and structural characterization. (A) Schematic of device architecture and measurement setup of an HJD-NP, with n-type MoS2 and p-type WSe2. When a dsDNA molecule translocates through the nanopore, it induces changes in the interlayer current. (B) High-angle annular dark-field (HAADF) scanning transmission electron microscopy (STEM) image of an example nanopore drilled through a membrane stack. (C) Cross-sectional HAADF-STEM image of a representative device showing the membrane stack.Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialsPNAS  2025  Vol. 122  No. 18 e2422135122� https://doi.org/10.1073/pnas.2422135122 3 of 8the ratio of the vertical p-n diode resistance to the total hetero-junction resistance. Consequently, applying a more negative Vionic  could further reduce the series resistances of the heterojunction in an HJD-NP, thereby enhancing the sensitivity of electri-cal sensing.  Electrical Rectification of Ionic Current. Fig.  3 examines the opposite situation of how the diode potential affects ionic transport. Fig. 3 A and B show the output and transfer curves of ionic transport of a liquid-gated HJD-NP with a pore size dpore of 4.0 ± 0.3 nm in 10 mM KCl. The ionic I–V characteristics were modulated by Vds. Specifically, i) |Iionic| increased (decreased) when Vds and Vionic had opposite (the same) polarities. ii) Iionic exhibited greater variation with Vds under forward bias compared to reverse bias. iii) |Iionic| was most rectified at the most negative Vionic. To understand the mechanism behind the ionic current control by Vds , we measured four HJD-NP sensors with dpore  from 1.1 ± 0.3 nm to 7.7 ± 0.3 nm and KCl concentrations cKCl  from 1 mM to 1 M (SI Appendix, Fig. S9 ).  Fig. 3C   plots the ionic current on/off ratio Iionic,ON /Ii  onic,OFF  and the maximum change in ionic cur-rent Iionic,ON  − Ii  onic,OFF  versus dpore  in a 10 mM KCl buffer, where Iionic,ON  and Iionic,OFF  were measured at Vds  = ±200 mV.  Fig. 3D   plots Iionic,ON /Ii  onic,OFF  versus cKCl  for two nanopore devices, where Iionic,ON  and Iionic,OFF  were measured at Vds  = ±100 mV. The data show that Iionic,ON /Ii  onic,OFF  decreased significantly with dpore  to the power of −1.25 ± 0.10 and slightly with cKCl  to the power of −0.03 to −0.06. In addition, Iionic,ON  - Ii  onic,OFF  increased sublinearly with dpore .  Fig. 3E   shows the cyclic changes in Iionic  as Vds  switched between ±200 mV at a fixed Vionic  of −400 mV, demonstrating stability and repeatability. There are two common mechanisms for modulating the ionic currents with interlayer potentials: i) edge electrochemistry at the nanopore, and ii) electrostatic interactions. Both applied biases, |Vd  − Vionic | and |Vs  − Vionic |, were capped at 600 mV, below the 800-mV onset for electrochemical reactions in monolayer MoS2  and WSe2  in 1 M KCl ( 21 ,  22 ). If edge chemistry dominated, then Iionic,ON  − Iionic,OFF  would increase linearly with the surface area of the pore edges Apore , as the electrochemical current increases linearly with Apore  ( 23 ). However, this predicted linear relationship between Iionic,ON  − Ii  onic,OFF  and Apore  contradicts the observed trend (SI Appendix, Fig. S10A ), suggesting that edge chemistry did not play a dominant role. Additionally, Iionic,ON  − Ii  onic,OFF  for one HJD-NP sensor with a pore size of 2.2 ± 0.3 nm remained relatively constant as cKCl  increased from 1 mM to 100 mM (SI Appendix, Fig. S10B ), confirming negligible electrochemistry. After ruling out edge electrochemistry, we attribute the control of ionic transport by Vds  to electrostatic interactions between the biased heterojunction and the ions within the nanopore channel. There are two possible origins of electrostatic modulation of the ionic transport: i) interlayer voltage drop across the vertical het-erojunction ( 24 ,  25 ), and ii) field-effect gating like an embedded electrode ( 26 ,  27 ). We ruled out the effect of the vertical potential drop, as reversing the heterojunction stacking sequence, and hence the direction of the electric field from Vds , did not reverse the induced ionic transport (SI Appendix, Fig. S9 C  and D ). Therefore, we attribute the control of ionic transport by Vds  to field-effect gating, which effectively explains the observed modulation of ionic transport by Vds  (SI Appendix, Supplementary Text 2 ) and can be leveraged for dynamic control of DNA translocation speed ( 28     – 31 ). We acknowledge that other mechanisms could also account for the modulation of the ionic currents by interlayer potentials, such as chemical modifications of the pore edges, changes in the hydra-tion structure of water, or changes in ion mobility in the vicinity of the diode junction. Such phenomena could reduce the effective A B CDF GEFig. 2.   Equivalent circuit diagrams and electrical characterization. (A) The ideal band diagram of a vertical p-WSe2/n+-MoS2 diode under forward Vds bias. The current due to interlayer recombination between the majority carriers of WSe2 and MoS2 dominates over the diffusion current (5, 15). (B) Id−Vds curve of an example HJD-NP sensor measured in dry air in dark. (C) Equivalent circuit of the van der Waals heterojunction. RHJ is the total resistance of the heterojunction. Rc,WSe2 and Rc,MoS2 are the contact resistances to WSe2 and MoS2, respectively. Rch,WSe2 and Rch,MoS2 are the channel resistances of WSe2 and MoS2 in the non-overlapping region, respectively. Rsho,WSe2 and Rsho,MoS2 are the sheet resistances of WSe2 and MoS2 in the overlapping region, respectively. Rdiode is the interface resistance between WSe2 and MoS2 per unit area. (D) Optical image (Left) and corresponding photocurrent map (Right) at Vds = 0 mV of an example HJD-NP. The red dot indicates the location of the nanopore. (E) Equivalent circuit of an HJD-NP. Rcis and Rtrans are the resistances of the salt solution in the cis and trans chambers, respectively, Cchip is the chip capacitance, Rpore is the nanopore resistance, Clead is the capacitance between the electrolyte and the contact leads on the drain or source side, and Cg is the capacitance between the electrolyte and the heterojunction. The ionic channel is shown in blue and the diode channel in red. (F) Output and (G) Transfer curves of the heterojunction in an HJD-NP under liquid gating in a 1 M LiCl buffer.Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materials4 of 8   https://doi.org/10.1073/pnas.2422135122� pnas.orgdiameter of the pore, with a more pronounced impact on small pores compared to larger ones. Further simulations are required to evaluate and identify specific mechanisms. Additionally, we modeled ion transport in HJD-NP with a floating diode (SI Appendix, Supplementary Text 3 , Table S1, and  Fig. S11 ). This ionic model was used to estimate dpore  in liquid measurements based on the measured ionic conductance. We also examined the effect of grounding the diode on ionic transport (SI Appendix, Fig. S12 ). The ionic I–V curves before and after grounding the diode remained linear and nearly identical, indi-cating negligible leakage and that connecting the diode with Vds  = 0 mV does not affect ion transport.  Electrical Sensing of DNA. In Fig. 4, we demonstrate the electrical sensing of dsDNA using an HJD-NP, with the measurement setup depicted in Fig. 1A. Both the cis and trans chambers were filled with 10 mM KCl. 50 nM of 2,686 bp circular dsDNA was added into the trans chamber. Vionic was varied from −300 to −400 mV, while Vds was varied from 0 to 200 mV. As expected, DNA translocation was detected only in the ionic channel at Vds = 0 mV (SI Appendix, Fig. S13 A and B). In contrast, DNA translocation was simultaneously detected in both ionic and diode channels at Vds = 200 mV (SI Appendix, Fig. S13 C and D). Fig. 4A shows example traces of these concurrently detected events. Both the ionic and diode currents increased upon DNA translocation. The increase of the ionic current is a signature of DNA translocation in dilute salt solutions like 10 mM KCl, where the transport of the counterions of DNA dominates over bulk ion transport (32, 33). The simultaneous increase of the diode current suggests the presence of DNA inside the nanopore channel reduced the diode resistance. When a dsDNA molecule translocates through an HJD-NP, two non-capacitive electrical sensing mechanisms contribute to changes in Id  ( Fig. 4B  ): i) charge sensing and ii) volumetric blockade-based sensing. The negative charges from the translocat-ing DNA strand reduce the charge imbalance at the heterointer-face of p-WSe2 /n+-MoS2 , leading to an increase in Id  (SI Appendix, Supplementary Text 4 ). Volumetric blockade–induced local poten-tial change would otherwise lead to a decrease in Id  and therefore can be ruled out. The transconductance of this heterojunction was only 0.8 pA/mV (SI Appendix, Fig. S14 ). Based on analytical cal-culations and simulations ( 16 ,  18 ), we estimate a maximum poten-tial change of 1 mV at the nanopore entrance, corresponding to a change in Id  of <1 pA. No capacitive signals were observed in the ionic or diode channel. Since Rlead  exceeded 10 GΩ (SI Appendix, Fig. S15 ) and Clead  was only 9.2 pF (Materials and Methods ), both capacitive crosstalk and direct leaking between the ionic and diode channels should be negligible ( 17 ,  18 ). Above all, we ascribe the observed increase in diode current upon DNA translocation to charge sensing. A total of 409 concurrent events were detected at Vionic  = −300 mV, Vds  = 200 mV over 9.6 min, and 238 events at Vionic  = −400 mV, Vds  = 200 mV over 7.8 min, as shown in the scatterplots in  Fig. 4 C  and D  . The current changes of the concurrent events detected in the ionic channel ∆Iionic  and the diode channel ∆Idiode  are weakly correlated, with a Pearson’s correlation coefficient of r2  = 0.41 ( Fig. 4E  ), suggesting different physical origins.  Fig. 4F   shows the scatterplots of the events detected in the ionic channel at Vionic  = −300, −400 mV, Vds  = 0 mV. The median and SD of the dwell times and current changes were extracted from these scat-terplots and presented in  Fig. 4 G  and H  . The details of data processing and analysis are in Materials and Methods . Notably, the median dwell time at Vionic  = −300 mV increased from 1.2 to 2.8 ms as Vds  increased from 0 to 200 mV. As Vionic  changed from −300 to −400 mV, ∆Iionic  increased by 12%, from 97 ± 18 to 109 ± 18 pA, while ∆Idiode  increased by only 4%, from 129 ± 24 to 134 ± 24 pA. The ionic signal-to-noise ratio (SNR) was 3.5 to 3.6, while the diode SNR was 3.9 to 4.2. This diode SNR is comparable to the FET SNR of 2D FET-nanopore sensors (SI Appendix, Table S2 ). Both the ionic and diode channels exhib-ited similar noise levels (SI Appendix, Fig. S16 ). Interestingly, the median dwell time increased with increasing Vds  regardless of Vionic , suggesting that applying Vds  slowed the DNA translocation. Additionally, dwell time increased slightly with |Vionic |, differing from conventional nanopore sensors at high salt concentrations with ionic bias alone, where dwell time decreases with |Vionic | ( 34 ). We ascribe the reduced DNA translo-cation speed to field-effect gating of the nanopore channel induced by the biased heterojunction. Specifically, as Vionic  becomes more negative and Vds  more positive, the effective gate potentials at the nanopore channel relative to WSe2  and MoS2 , i.e., |Vionic  − αVd | and |Vionic |, both increase ( 29 ,  30 ). This gating effect on DNA translocation may remain significant even when the pore diameter substantially exceeds the Debye length ( 29 ,  30 ). We also note the transport behavior of charged macromolecules at low salt concentrations may differ from that at high salt con-centrations ( 35 ,  36 ), which could also contribute to the uncon-ventional dependence of dwell time on |Vionic |. In addition, acquiring additional DNA translocation events with increased A BDCEFig. 3.   Ionic current modulation. (A) Output and (B) Transfer curves of the ionic transport of an HJD-NP with a dpore of 4.0 ± 0.3 nm under varying Vds. The heterojunction stack consists of 3.4 nm thick WSe2 on top of 3.8 nm thick MoS2. (C) Ionic current on/off ratio Iionic,ON/Iionic,OFF and the maximum difference in ionic current Iionic,ON − Iionic,OFF versus dpore at Vionic = −400 mV in a 10 mM KCl buffer. ON state: Vds = 200 mV; OFF state: Vds = −200 mV. (D) Iionic,ON/Iionic,OFF versus cKCl at Vionic = −400 mV for two nanopore devices. ON state: Vds = 100 mV; OFF state: Vds = −100 mV. (E) Response of Id and Iionic of an HJD-NP to cyclic switching of Vds between ±200 mV at a fixed Vionic of −400 mV. The pore size was 7.7 ± 0.3 nm, and the buffer was 10 mM KCl.Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialsPNAS  2025  Vol. 122  No. 18 e2422135122� https://doi.org/10.1073/pnas.2422135122 5 of 8bandwidth and higher low-pass filter frequencies is necessary to obtain accurate event statistics that include fast events ( 37 ). The slight increase in ∆Iionic  with |Vionic | is consistent with previous studies in dilute salt solutions ( 38 ,  39 ). The relatively small increase in ∆Idiode  with |Vionic | further supports the charge sensing mechanism, as the effective charge of DNA does not change with Vionic  ( 40 ).  Potential Improvements. HJD-NP sensors are hypothesized to offer high spatial resolution due to their atomically sharp 2D heterointerfaces. However, this hypothesis has yet to be experimentally validated, which will require a 1 to 2 orders of magnitude improvement in both SNR and dwell time. Strategies to improve the diode's SNR and slow DNA translocation include minimizing series resistances arising from contact and channel resistances through degenerate doping (41–43), reducing background noise by shrinking the overlap region between WSe2 and MoS2 from 1 to 10 µm2 to ≲0.01 µm2 using nanolithography (41, 44), and enhancing the 2D diode’s electrostatic control over DNA by using a smaller nanopore below the Debye screening length (28–31). Additional discussion on slowing DNA translocation and the potential for base-by-base sensing is provided in SI Appendix, Supplementary Text 5. Detailed discussion on improving the diode’s SNR is provided in SI Appendix, Supplementary Text 6.Conclusion We demonstrated electrical sensing of DNA translocation through a nanopore using the interlayer current of a vertical 2D diode, concurrently observing slowed transport due to electrostatic ABG HC D EFFig. 4.   Concurrent diode and ionic current sensing of DNA translocation. (A) Concatenated signal traces (Left) and zoom-in view of a single concurrent event (Right). Vionic = −300 mV, Vds = 200 mV. (B) Two potential electrical sensing mechanisms for detecting translocating dsDNA through a p-WSe2/n+-MoS2 heterojunction nanopore: charge sensing and volumetric blockade-based sensing. In charge sensing, as negatively charged dsDNA enters the nanopore, it improves the balance between hole density p and electron density n at the heterointerface, leading to an increase in Id. In volumetric blockade-based sensing, when a negative Vionic is applied, the amplitude of the voltage at the nanopore entrance Ven decreases upon DNA translocation, resulting in a decrease in Id. The sensitive region is shown in the golden zone. Scatterplots of the concurrent events detected in the (C) ionic channel and (D) diode channel, showing changes in ionic current ∆Iionic and diode current ∆Idiode, as well as dwell times for both ionic τionic and diode τdiode signals. Vionic = −300, −400 mV, Vds = 200 mV. (E) Signal amplitude correlation between concurrent events in the ionic and diode channels. Vionic = −300, −400 mV, Vds = 200 mV. (F) Scatterplots of the events detected in the ionic channel. Vionic = −300, −400 mV, Vds = 0 mV. (G) τionic (median ± SD) versus Vionic. Vds = 0, 200 mV. (H) ∆Iionic and ∆Idiode (median ± SD) versus Vionic. Vds = 200 mV. The heterojunction stack consists of 3.5 nm thick WSe2 on top of 2.0 nm thick MoS2. The pore size was 15 ± 0.3 nm. The DNA molecule was 2,686 bp pUC19 plasmid. The electrolyte was 10 mM KCl in both cis and trans chambers.Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materials6 of 8   https://doi.org/10.1073/pnas.2422135122� pnas.orginteractions between the DNA and diode. This heterojunction nanopore platform enables sensing capabilities beyond single 2D materials with local out-of-plane electric fields and interlayer currents. The local voltage control Vds  is preferred over the global voltage control Vionic  to regulate DNA translocation through the nanopore, because i) Vds  can be adjusted much more rapidly than Vionic , as the diode channel can potentially operate at GHz fre-quencies ( 45 ) compared to the kHz frequencies of the ionic channel ( 46 ), and ii) Vds  allows for individual addressing of nan-opore sensors in a multiplexed sensor array, which is challenging with Vionic . Exciting future directions include exploring band-to-band tun-neling with type-III band alignment and measuring correlated intralayer currents within each semiconductor layer to comple-ment interlayer measurement. Future work could also add a volt-age source between the reference electrode in the electrolyte solution and the diode ( 19 ). This modification would enable independent control of the source and drain sides of the diode relative to the solution. By allowing the diode potential to vary with respect to the electrolyte without affecting the potential drop across the diode, this setup could provide enhanced control of DNA translocation. Collectively, the vertical 2D diode nanopore presents promising pathways toward solid-state nanopore sequencing.  Materials and MethodsHJD-NP Device Fabrication. SI Appendix, Fig. S1 illustrates the fabrication pro-cedure. The process started with a 20 ± 2 nm thick commercial SiNx membrane (NBPX4001Y-LR, Norcada) with a 10 × 10 µm window size on a 4 mm large, 10 to 20 Ωcm low-resistivity silicon substrate. A lightly doped silicon substrate was chosen over an intrinsic silicon substrate to prevent electrostatic discharge, which could further reduce the already low yield (47). Next, i) a 10 to 15 nm thick HfOx layer was deposited onto the front side of the SiNx membrane at 200 °C using ALD (Savannah S100 ALD system). The thickness of the HfOx layer was measured using ellipsometry (J.A. Woollam VASE). Following this, ii) the SiNx was etched from the HfOx/SiNx membrane stack using a 60 s XeF2 gas etch (XACTIX XeF2 etching system) at 3 Torr (SI Appendix, Fig. S2), with the substrate placed upside down on a clean glass slide during etching. ALD HfOx served as the etch stop. Consequently, the supporting membrane was converted from 20 nm SiNx to 10 to 15 nm HfOx, which offers improved wettability (7) and durability (8).Subsequently, iii) 2 to 5 nm thick MoS2 and WSe2 flakes were exfoliated from synthetic crystals (HQgraphene), stacked together, and transferred onto the membrane using the pick-up technique (9). The thickness of exfoliated 2D flakes (MoS2, WSe2, and hBN) was optically identified after exfoliation and verified with atomic force microscopy (AFM) after transfer. The size of the WSe2/MoS2 overlap region was 1 to 10 µm2. AFM tip-based cleaning was used to remove surface residues from the transfer process (48). Next, iv) a 100 × 100 µm large, 50 nm thick gold pad was transferred onto the WSe2 flake as the p-type contact electrode in a N2-filled glovebox (49). Subsequently, v) 5 nm Ni/30 nm Au was deposited onto the MoS2 flake as the n-type contact electrode using optical or ebeam lithography followed by ebeam evaporation. Then, vi) a 5 to 20 nm thick hBN flake was exfoliated and transferred onto the membrane to fully encapsu-late the WSe2/MoS2 heterostructure using the pick-up technique (9). This mul-tilayer hBN flake reduces electron doping and damage to WSe2 from ALD HfOx. Next, a gentle remote O2 plasma process (1 min, 10 W, Tergeo plasma cleaner) was used to nucleate the hBN surface for complete and uniform ALD coverage (50). Afterward, vii) a 10 to 15 nm thick HfOx layer was deposited onto the front side of the device at 200 °C using ALD for electrical insulation. Leakage tests in SI Appendix, Fig. S15 A and B found a minimum of 10 nm ALD HfOx was sufficient for electrical insulation. The total thickness of the final membrane stack (HfOx/hBN/WSe2/MoS2/HfOx) was 30 to 60 nm.Finally, viii) a 1 to 15 nm nanopore was drilled through the membrane stack in the WSe2/MoS2 overlap region using a focused electron beam at 300 kV using STEM (Themis Z, Thermo Fisher Scientific). A beam current of 13 nA (4 nA) was required to drill through the membrane stack within a few minutes in nanoprobe (microprobe) mode with spot size 1, corresponding to a beam size of about 1 nm (a few nm). The electron irradiation dose for initial imaging and locating the drilling spot was 107 e−/nm2 (106 e−/nm2). After drilling, the nanopore was imaged in STEM with a 30 pA beam current and an atomic beam size (nanoprobe mode with spot size 9).Cross-Sectional STEM Sample Fabrication and Imaging. A protective layer of 5 to 30 nm thick amorphous carbon was thermally evaporated onto the nan-opore device membrane. The cross-sectional STEM sample was then fabricated using standard focused ion beam (FIB) lift-out procedures with a FIB-SEM system (Helios 600i DualBeam, Thermo Fisher Scientific). A cryo-can was used during FIB thinning to minimize redeposition. The cross-sectional sample was imaged using an aberration-corrected STEM (Themis Z, Thermo Fisher Scientific), oper-ated at 300 kV with a convergence angle of 25.2 mrad. Elemental mapping was performed using the Super-X EDS detector.Photocurrent Mapping. Scanning photocurrent measurements were performed by rastering a focused laser spot across the device surface, using a source measure unit (2450 SourceMeter, Keithley) for voltage sourcing and a current preamplifier (SR570, Stanford Research System) for current measurement. The laser spot had a diameter of 1 µm, a wavelength of 488 nm, and a power of 70 µW.Raman Measurements. Raman measurements were performed with a confocal Raman microscope (Nanophoton Raman 11) using a 532 nm laser with a 100× objective. Raman spectra were obtained using a grating of 2,400 lines/mm at a laser power of 0.5 mW/cm2 and an acquisition time of 30 s. Raman maps were obtained using a grating of 2,400 lines/mm at a laser power of 0.5 mW/cm2, an acquisition time of 3 s/pixel and a pixel size of 300 nm.Nanopore Measurements and Data Analysis. Before the experiment, the fabricated chip was attached onto a custom-made printed circuit board (PCB) by applying silicone elastomer (Kwik-cast, World Precision Instruments) around the edges of the chip. The silicone elastomer was then further applied around the membrane to reduce the chip capacitance and provide additional insulation between the metal leads and the electrolyte, leaving an exposed area of less than 0.01 mm2. Clead was 5.2 nF without the silicone paint and 9.2 pF with the silicone paint, measured using triangular Vionic waves (51). Afterward, the chip was electrically connected to the PCB by wire bonding. The PCB was cleaned with acetone, IPA, and deionized (DI) water (18 MΩ cm, Milli-Q, Millipore), and then sandwiched between two chambers of mechanically clamped custom-made PMMA flow cells.Next, both chambers were sequentially flushed with DI water, IPA, DI water, and a salt solution of 10 mM KCl, 10 mM Tris, 1 mM EDTA at a pH of 7.4 ± 0.2. Both the DI water and salt solution were degassed overnight using a dry scroll vacuum pump (SVF-E3M-20PC, Scroll Labs), which is critical for successful nano-pore wetting. In most cases, the nanopore could not be fully wetted immediately. To promote wetting, three methods were employed: i) flushing both chambers with IPA, ii) leaving both chambers with degassed 10 mM KCl solution overnight, and iii) ramping the ionic voltage between ±500 mV in 10 mM KCl.All current measurements were performed using an integrated patch-clamp amplifier (MultiClamp 700B, Axon Instruments), with channel 1 dedicated to the ionic channel and channel 2 to the diode channel. Channel 1 was connected to a pair of Ag/AgCl electrodes to apply Vionic and measure Iionic. The ionic voltage was applied to the trans chamber, while the cis chamber was grounded. Channel 2 was connected to the PCB to apply Vds and measure Id. The electrical voltage was applied to the drain side with p-WSe2, while the source side with n-MoS2 was grounded. Current offsets in both channels were adjusted to zero at zero biases. The entire setup was housed in a Faraday cage with a dedicated low-noise ground connection on a vibration isolation table.After complete nanopore wetting, both Vionic and Vds were varied while measur-ing Iionic and Id to study the interactions between the ionic channel and the diode channel. Additional leakage tests in SI Appendix, Fig. S15 C and D established the operation limits of HJD-NPs as |Vd − Vionic| ≤ 600 mV and |Vs − Vionic| ≤ 600 mV. The sampling rate was 1 kHz, and the data were low-pass-filtered at 20 Hz using the built-in 8-pole Bessel filter. The output signal was digitized by a Digidata 1440A (Axon Instruments) and recorded using pClamp 10.7 software. Finally, for DNA sensing experiments, 50 nM of 2,686 bp circular dsDNA (pUC19 plasmid, New England Biolabs) was added to the trans chamber. The sampling Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialsPNAS  2025  Vol. 122  No. 18 e2422135122� https://doi.org/10.1073/pnas.2422135122 7 of 8rate for DNA sensing was 100 kHz. A 2 kHz low-pass 8-pole Bessel filter was applied to both channels to obtain adequate SNRs for event detection and statis-tical analysis. Blank experiments were conducted at the applied voltages before DNA insertion. Event detection was performed using the open-source Matlab code package Transalyzer (52). Events shorter than 0.25 ms were excluded due to low-pass filtering, which however skews the complete dwell time distribution (37) and constitutes a limitation of this study. Concurrent events were identified when the event start time in both channels differed by 0.5 ms or less. τionic and τdiode were determined using the full-width, half-maximum (FWHM) values of the event (52). ∆Iionic and ∆Idiode were determined by dividing the event charge deficit by the FWHM time (52). τionic and τdiode were analyzed statistically by fitting the event dwell time histogram with a Gaussian peak for Vds = 200 mV and an exponential decay for Vds = 0 mV. ∆Iionic and ∆Idiode were analyzed statistically by fitting the event amplitude histogram with a Gaussian peak.Yield. After establishing the fabrication protocol, we started device fabrication with 82 SiNx membranes, out of which 57 reached the final fabrication step. Out of the fabricated devices, 36 (63%) were excluded from liquid measurements due to an insensitive WSe2/MoS2 overlap region, as determined by both I−V character-istics and photocurrent mapping, and 13 (23%) failed to fully wet. We managed to perform DNA translocation measurements on nine devices and successfully detected concurrent events in both ionic and diode channels in two devices. SI Appendix, Fig. S17 reports concurrent DNA sensing from the other working device, which exhibits a diode current decrease upon DNA translocation, in con-trast to the diode behavior shown in Fig. 4, where the diode current increased. Since this device features a modified structure with a strong p-dopant on WSe2, we attribute its different diode sensing behavior to enhanced p-doping in WSe2 dominating over n-doping in MoS2.This low yield aligns with previously reported challenges in fabricating FET-nanopore devices (16, 17, 53–55), particularly those using 2D materials (SI Appendix, Table S2). SI Appendix, Table S3 compares the diode current den-sity, noise, and measured or estimated signals and SNRs across all nine devices with DNA sensing measurements, including seven devices with unresolved diode signals and two devices with resolved diode signals. Devices with unresolved diode signals exhibit similar noise levels to those with resolved diode signals but lower current densities, resulting in reduced and ultimately inadequate SNRs. Low diode current density arises from insufficient p-doping in WSe2 in devices where the electron density in MoS2 exceeds the hole density in WSe2 (SI Appendix, Supplementary Text 4).Taken together, the analysis of the devices with both unresolved and resolved diode signals, combined with the DNA sensing data from the device with enhanced p-doping in WSe2, suggests that high device variability and low yield primarily arise from insufficient doping control in the 2D materials. To make 2D heterojunction diode nanopore a viable technique with sufficient yield and SNR, doping control in each of the 2D layer is essential, which remains a critical challenge not only for 2D nanopores integrated with local electronic sensors but for 2D electronics in general (56). Further discussions on yield and reproducibil-ity limiting processes are provided in SI Appendix, Supplementary Text 7. While further measures can be taken to optimize the yield and SNR, we focused on proof-of-principle of electronic out-of-plane DNA sensing and device behavior characterization, which was accomplished in this study.Data, Materials, and Software Availability. The dataset supporting these findings is openly available in the Illinois Data Bank at https://doi.org/10.13012/B2IDB-7678688_V1 (57).ACKNOWLEDGMENTS. This work was partially funded by Taiwan Semiconductor Manufacturing Company under Grant No. 089401, the NIH under Grant No. R21HG010701, the NSF through the University of Illinois Urbana-Champaign Materials Research Science and Engineering Center (Illinois MRSEC) under Grant No. DMR-2309037, and funding from University of Illinois Urbana-Champaign to R.B. This work was carried out in part in the Holonyak Micro and Nanotechnology Laboratory and Materials Research Laboratory Central Facilities at University of Illinois Urbana-Champaign. We acknowledge the use of facilities and instrumentation supported by NSF through the Illinois MRSEC under Grant No. DMR-2309037. K.W. and T.T. acknowledge support from the Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research program (Grant Nos. 21H05233 and 23H02052) and World Premier International Research Center Initiative, Ministry of Education, Culture, Sports, Science and Technology, Japan.Author affiliations: aHolonyak Micro and Nanotechnology Laboratory, The Grainger College of Engineering, University of Illinois Urbana-Champaign, Urbana, IL 61801; bDepartment of Mechanical Science and Engineering, The Grainger College of Engineering, University of Illinois Urbana-Champaign, Urbana, IL 61801; cDepartment of Materials Science and Engineering, The Grainger College of Engineering, University of Illinois Urbana-Champaign, Urbana, IL 61801; dResearch Center for Electronic and Optical Materials, National Institute for Materials Science, Tsukuba 305-0044, Ibaraki, Japan; eResearch Center for Materials Nanoarchitectonics, National Institute for Materials Science, Tsukuba 305-0044, Ibaraki, Japan; fMaterials Research Laboratory, The Grainger College of Engineering, University of Illinois Urbana-Champaign, Urbana, IL 61801; gDepartment of Biomedical and Translational Science, The Carle Illinois College of Medicine, University of Illinois Urbana-Champaign, Urbana, IL 61801; hDepartment of Bioengineering, The Grainger College of Engineering, University of Illinois Urbana-Champaign, Urbana, IL 61801; iChan Zuckerberg Biohub Chicago, Chicago, IL 60642; and jCarl R. Woese Institute for Genomic Biology, University of Illinois Urbana-Champaign, Urbana, IL 61801Author contributions: S.C., A.M.v.d.Z., and R.B. designed research; S.C., S.H., E.H., and P.Y.H. performed research; J.S., K.W., T.T., and W.P.K. contributed new reagents/analytic tools; S.C., S.H., E.H., and P.Y.H. analyzed data; A.M.v.d.Z. and R.B. supervised research; and S.C., A.M.v.d.Z., and R.B. wrote the paper.1.  B. M. Venkatesan, R. Bashir, Nanopore sensors for nucleic acid analysis. Nat. Nanotechnol. 6, 615–624 (2011).2.  S. Chen, R. Bashir, Advances in field-effect biosensors towards point-of-use. Nanotechnology 34, 492002 (2023).3.  S. Polonsky, S. Rossnagel, G. Stolovitzky, Nanopore in metal-dielectric sandwich for DNA position control. Appl. Phys. Lett. 91, 153103 (2007).4.  S. Harrer et al., Electrochemical characterization of thin film electrodes toward developing a DNA transistor. Langmuir 26, 19191–19198 (2010).5.  C. H. Lee et al., Atomically thin p–n junctions with van der Waals heterointerfaces. Nat. Nanotechnol. 9, 676–681 (2014).6.  P. V. Pham et al., 2D heterostructures for ubiquitous electronics and optoelectronics: Principles, opportunities, and challenges. Chem. Rev. 122, 6514–6613 (2022).7.  J. Shim, J. A. Rivera, R. Bashir, Electron beam induced local crystallization of HfO2 nanopores for biosensing applications. Nanoscale 5, 10887 (2013).8.  Y. C. Chou, P. Masih Das, D. S. Monos, D. S. Monos, M. Drndić, Lifetime and stability of silicon nitride nanopores and nanopore arrays for ionic measurements. ACS Nano 14, 6715–6728 (2020).9.  K. Kim et al., Van der Waals heterostructures with high accuracy rotational alignment. Nano. Lett. 16, 1989–1995 (2016).10.  S. H. H. Shokouh et al., High-performance, air-stable, top-gate, p-channel WSe2 field-effect transistor with fluoropolymer buffer layer. Adv. Funct. Mater. 25, 7208–7214 (2015).11.  M. Puster, J. A. Rodríguez-Manzo, A. Balan, M. Drndić, Toward sensitive graphene nanoribbon-nanopore devices by preventing electron beam-induced damage. ACS Nano 7, 11283–11289 (2013).12.  X. Zhang et al., Van der Waals-interface-dominated all-2D electronics. Adv. Mater. 35, 2207966 (2023).13.  Y. Z. Zhang, G. J. Zhu, J. H. Yang, Origin of p-type conductivity in a WSe2 monolayer. Nanoscale 15, 12116–12122 (2023).14.  M. D. Siao et al., Two-dimensional electronic transport and surface electron accumulation in MoS2. Nat. Commun. 9, 1442 (2018).15.  T. Roy et al., Dual-gated MoS2/WSe2 van der Waals tunnel diodes and transistors. ACS Nano 9, 2071–2079 (2015).16.  P. Xie, Q. Xiong, Y. Fang, Q. Qing, C. M. Lieber, Local electrical potential detection of DNA by nanowire-nanopore sensors. Nat. Nanotechnol. 7, 119–125 (2012).17.  F. Traversi et al., Detecting the translocation of DNA through a nanopore using graphene nanoribbons. Nat. Nanotechnol. 8, 939–945 (2013).18.  M. Puster et al., Cross-talk between ionic and nanoribbon current signals in  graphene nanoribbon-nanopore sensors for single-molecule detection. Small 11, 6309–6316 (2015).19.  S. J. Heerema et al., Probing DNA Translocations with Inplane Current Signals in a Graphene Nanoribbon with a Nanopore. ACS Nano 12, 2623–2633 (2018).20.  M. Graf, M. Lihter, D. Altus, S. Marion, A. Radenovic, Transverse detection of DNA using a MoS2 nanopore. Nano. Lett. 19, 9075–9083 (2019).21.  J. Feng et al., Electrochemical reaction in single layer MoS2: Nanopores opened atom by atom. Nano Lett. 15, 3431–3438 (2015).22.  M. Brunet Cabré, A. E. Paiva, M. Velický, P. E. Colavita, K. McKelvey, Electrochemical kinetics as a function of transition metal dichalcogenide thickness. Electrochim. Acta 393, 139027 (2021).23.  S. Banerjee et al., Electrochemistry at the edge of a single graphene layer in a nanopore. ACS Nano 7, 834–843 (2013).24.  M. Niu et al., Light-driven ion transport through single-heterojunction nanopores. Nano Lett. 23, 1010–1016 (2023).25.  Z. Yuan et al., Light-driven ionic and molecular transport through atomically thin single nanopores in MoS2/WS2 heterobilayers. ACS Nano 18, 24581–24590 (2024).26.  S. W. Nam, M. J. Rooks, K. B. Kim, S. M. Rossnagel, Ionic field effect transistors with sub-10 nm multiple nanopores. Nano Lett. 9, 2044–2048 (2009).27.  Z. Jiang, D. Stein, Charge regulation in nanopore ionic field-effect transistors. Phys. Rev. E 83, 31203 (2011).Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.http://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttp://www.pnas.org/lookup/doi/10.1073/pnas.2422135122#supplementary-materialshttps://doi.org/10.13012/B2IDB-7678688_V1https://doi.org/10.13012/B2IDB-7678688_V18 of 8   https://doi.org/10.1073/pnas.2422135122� pnas.org28.  Y. Ai, J. Liu, B. Zhang, S. Qian, Field effect regulation of DNA translocation through a nanopore. Anal. Chem. 82, 8217–8225 (2010).29.  Y. Liu, L. Yobas, Slowing DNA translocation in a nanofluidic field-effect transistor. ACS Nano 10, 3985–3994 (2016).30.  P. C. Yen, C. H. Wang, G. J. Hwang, Y. C. Chou, Gate effects on DNA translocation through silicon dioxide nanopore. Rev. Sci. Instrum. 83, 034301 (2012).31.  Y. He, M. Tsutsui, C. Fan, M. Taniguchi, T. Kawai, Controlling DNA translocation through gate modulation of nanopore wall surface charges. ACS Nano 5, 5509–5518 (2011).32.  R. M. M. Smeets et al., Salt dependence of ion transport and DNA translocation through solid-state nanopores. Nano Lett. 6, 89–95 (2006).33.  H. Chang et al., DNA counterion current and saturation examined by a MEMS-based solid state nanopore sensor. Biomed. Microdevices 8, 263–269 (2006).34.  M. Wanunu, J. Sutin, B. McNally, A. Chow, A. Meller, DNA translocation governed by interactions with solid-state nanopores. Biophys. J. 95, 4716–4725 (2008).35.  Y. M. N. D. Y. Bandara, N. Farajpour, K. J. Freedman, Nanopore current enhancements lack protein charge dependence and elucidate maximum unfolding at protein’s isoelectric point. J. Am. Chem. Soc. 144, 3063–3073 (2022).36.  L. S. Lastra, Y. M. N. D. Y. Bandara, M. Nguyen, N. Farajpour, K. J. Freedman, On the origins of conductive pulse sensing inside a nanopore. Nat. Commun. 13, 2186 (2022).37.  C.-Y. Lin et al., Ultrafast polymer dynamics through a nanopore. Nano Lett. 22, 8719–8727 (2022).38.  K. J. Freedman et al., Nanopore sensing at ultra-low concentrations using single-molecule dielectrophoretic trapping. Nat. Commun. 7, 10217 (2016).39.  K. Chen, N. A. W. Bell, J. Kong, Y. Tian, U. F. Keyser, Direction- and salt-dependent ionic current signatures for DNA sensing with asymmetric nanopores. Biophys. J. 112, 674–682 (2017).40.  U. F. Keyser et al., Direct force measurements on DNA in a solid-state nanopore. Nat. Phys. 2, 473–477 (2006).41.  S. Chen, Y. Zhang, W. P. King, R. Bashir, A. M. Van Der Zande, Edge-passivated monolayer WSe2 nanoribbon transistors. Adv. Mater. 36, 2313694 (2024).42.  R. Younas, G. Zhou, C. L. Hinkle, A perspective on the doping of transition metal  dichalcogenides for ultra-scaled transistors: Challenges and opportunities. Appl. Phys. Lett. 122, 160504 (2023).43.  S. Chen, Y. Zhang, W. P. King, R. Bashir, A. M. van der Zande, Extension doping with low-resistance contacts for p-type monolayer WSe2 field-effect transistors. Adv. Electron. Mater. 2400843 (2024), 10.1002/aelm.202400843.44.  S. Chen et al., Monolayer MoS2 nanoribbon transistors fabricated by scanning probe lithography. Nano Lett. 19, 2092–2098 (2019).45.  S. Hong et al., Ultrafast van der Waals diode using graphene quantum capacitance and Fermi-level depinning. Sci. Adv. 9, eadh9770 (2023).46.  Z. Mi et al., Ten thousand recaptures of a single DNA molecule in a nanopore and variance reduction of translocation characteristics. ACS Nano 18, 23243–23252 (2024).47.  K. Healy et al., Fabrication and characterization of nanopores with insulated transverse nanoelectrodes for DNA sensing in salt solution. Electrophoresis 33, 3488–3496 (2012).48.  S. Chen et al., Tip-based cleaning and smoothing improves performance in monolayer MoS2 devices. ACS Omega 6, 4013–4021 (2021).49.  Y. Liu et al., Approaching the Schottky–Mott limit in van der Waals metal–semiconductor junctions. Nature 557, 696–700 (2018).50.  M. Snure et al., Two-dimensional BN buffer for plasma enhanced atomic layer deposition of Al2O3 gate dielectrics on graphene field effect transistors. Sci. Rep. 10, 14699 (2020).51.  A. Balan et al., Improving signal-to-noise performance for DNA translocation in solid-state nanopores at MHz bandwidths. Nano Lett. 14, 7215–7220 (2014).52.  B. Doekemeijer et al., Data analysis methods for solid-state nanopores. Nanotechnology 26, 084003 (2015).53.  S. J. Heerema et al., Probing DNA translocations with inplane current signals in a graphene nanoribbon with a nanopore. ACS Nano 12, 2623–2633 (2018).54.  M. Graf, M. Lihter, D. Altus, S. Marion, A. Radenovic, Transverse detection of DNA using a MoS2 nanopore. Nano Lett. 19, 9075–9083 (2019).55.  X. Zhu et al., Monolithic integration of vertical thin-film transistors in nanopores for charge sensing of single biomolecules. ACS Nano 15, 9882–9889 (2021).56.  Y. Wang, S. Sarkar, H. Yan, M. Chhowalla, Critical challenges in the development of electronics based on two-dimensional transition metal dichalcogenides. Nat. Electron 7, 638–645 (2024).57.  S. Chen et al., Data from “Detecting DNA translocation through a nanopore using a van der Waals heterojunction diode.” Illinois Data Bank. https://doi.org/10.13012/B2IDB-7678688_V1. Deposited 15 April 2025.Downloaded from https://www.pnas.org by "BUSSHITSU ZAIRYO KENKYUKIKO, LIBRARY" on May 1, 2025 from IP address 144.213.253.16.https://doi.org/10.1002/aelm.202400843https://doi.org/10.13012/B2IDB-7678688_V1 Detecting DNA translocation through a nanopore using a van der Waals heterojunction diode Significance Results and Discussion Nanopore Integrated with a Vertical 2D Diode. Electrical Rectification of Ionic Current. Electrical Sensing of DNA. Potential Improvements. Conclusion Materials and Methods HJD-NP Device Fabrication. Cross-Sectional STEM Sample Fabrication and Imaging. Photocurrent Mapping. Raman Measurements. Nanopore Measurements and Data Analysis. Yield. Data, Materials, and Software Availability ACKNOWLEDGMENTS Supporting Information Anchor 26