# Fileset

[chen-et-al-2024-locally-doped-transferred-contacts-for-wse2-transistors.pdf](https://mdr.nims.go.jp/filesets/37253ed9-7aab-4628-9d87-1a45717b9d07/download)

## Creator

[He-Yu Chen](https://orcid.org/0009-0004-7101-9179), [Jheng-Jie Lin](https://orcid.org/0009-0002-1263-9453), Sheng-Shong Wong, Zhen-You Lin, [Yu-Chiang Hsieh](https://orcid.org/0009-0001-3346-9548), Kuo-En Chang, Chung-Lin Wu, [Kenji Watanabe](https://orcid.org/0000-0003-3701-8119), [Takashi Taniguchi](https://orcid.org/0000-0002-1467-3105), Tse-Ming Chen, [Luke W. Smith](https://orcid.org/0000-0003-2194-4708)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Locally Doped Transferred Contacts for WSe<sub>2</sub> Transistors](https://mdr.nims.go.jp/datasets/c300fad0-96db-442a-8980-b2a1427af710)

## Fulltext

Locally Doped Transferred Contacts for WSe2 TransistorsLocally Doped Transferred Contacts for WSe2 TransistorsHe-Yu Chen, Jheng-Jie Lin, Sheng-Shong Wong, Zhen-You Lin, Yu-Chiang Hsieh, Kuo-En Chang,Chung-Lin Wu, Kenji Watanabe, Takashi Taniguchi, Tse-Ming Chen, and Luke W. Smith*Cite This: ACS Appl. Electron. Mater. 2024, 6, 8319−8327 Read OnlineACCESS Metrics & More Article Recommendations *sı Supporting InformationABSTRACT: While two-dimensional (2D) materials have showngreat promise for scaling technology nodes beyond the limits ofsilicon devices, key challenges remain for realizing high-quality andpractical 2D field-effect transistors (FETs), including loweringcontact resistance, demonstrating device structures with highelectrical stability, reducing interface charge trapping, andintegrating n- and p-FETs for beyond-complementary metaloxide semiconductor devices. High contact resistance often stemsfrom Schottky contacts and Fermi level pinning and can bereduced by local doping or transferred contacts, respectively.However, these approaches to date have been mutuallyincompatible. Here, we combine both into a single structure anddemonstrate a locally doped, transfer-contact stack containing access regions adjacent to the metal via contacts embedded inhexagonal boron nitride. Doping is applied by oxygen plasma treatment of access regions, while the fully encapsulated WSe2 channelremains pristine, creating a lateral p+−i−p+ junction. We demonstrate a reduction in contact resistance by up to >30,000 times withthe contact strategy, with a lowest individual contact resistance of ∼3.6 kΩ · μm, limited by the doping density at the contacts. Ourresults highlight increasing doping in the contact region as being crucial for achieving improved contact resistance in p-type WSe2devices. For our FET devices, the geometry of gates, doped access regions, and the channel are all defined by an electron beamlithography giving full and precise control over size and position. The p-FET behavior is strongly enhanced with a high on/off ratioup to 107, but ambipolar characteristics from the intrinsic channel are still retained. Negligible, temperature-independent hysteresis isachieved from T = 10 to 300 K, with only back gate carrier control. High electrical stability is evident in the excellent reproducibilityof transfer characteristics between multiple contact sets on a single device and different devices. The doping reduces contactresistance by reducing the Schottky barrier height and width, achieving Ohmic IV characteristics. The doping appears very stable,with negligible degradation of performance, keeping the device for 50 days in atmosphere. This reasonably simple device structureincorporates two important strategies to enhance contact quality, improving p-FET performance and retaining intrinsic channelquality.KEYWORDS: Two-dimensional, transition metal dichalcogenides, tungsten diselenide, oxygen plasma, p-type doping, hole injection,field effect transistor, tungsten oxide■ INTRODUCTIONTwo-dimensional (2D) materials have been revolutionizingfundamental research and technology by the presence ofremarkable physical and electric properties in a naturallyultrathin body down to a single monolayer. Graphene, perhapsthe most widely researched of the 2D material family, possessesdistinctive electrodynamic and transport properties;1−3 how-ever, the zero bandgap makes it unsuitable for logic operations.In response, semiconducting 2D transition-metal dichalcoge-nides (TMDs) have emerged with significant potential intransistor applications and as a possible solution for the scalingof such technologies.4 Challenges for silicon devices stem frominterface effects including surface roughness and danglingbonds that degrade carrier mobility, or short channel effects,5whereas 2D TMDs are free from dangling bonds, can beintegrated with ultrasmooth 2D dielectrics such as hexagonal-boron nitride (hBN),6 and retain high mobility in the ultrathinlimit, thereby overcoming the performance degradation thatoccurs for silicon devices as body thickness and channel lengthare reduced in the few nanometer limit.7However, despite their significant potential, challengesremain for high-performance TMD field-effect transistors(FETs),8 including contact resistance, reducing chargetrapping from interface stages, and integrating both n- and p-Received: September 3, 2024Revised: October 27, 2024Accepted: November 5, 2024Published: November 8, 2024Articlepubs.acs.org/acsaelm© 2024 The Authors. Published byAmerican Chemical Society8319https://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−8327This article is licensed under CC-BY 4.0Downloaded via NATL INST FOR MATLS SCIENCE (NIMS) on November 27, 2024 at 05:37:47 (UTC).See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.https://pubs.acs.org/action/doSearch?field1=Contrib&text1="He-Yu+Chen"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Jheng-Jie+Lin"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Sheng-Shong+Wong"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Zhen-You+Lin"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Yu-Chiang+Hsieh"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kuo-En+Chang"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Chung-Lin+Wu"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Chung-Lin+Wu"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kenji+Watanabe"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takashi+Taniguchi"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Tse-Ming+Chen"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Luke+W.+Smith"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/showCitFormats?doi=10.1021/acsaelm.4c01574&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?goto=articleMetrics&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?goto=recommendations&?ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?goto=supporting-info&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=tgr1&ref=pdfhttps://pubs.acs.org/toc/aaembp/6/11?ref=pdfhttps://pubs.acs.org/toc/aaembp/6/11?ref=pdfhttps://pubs.acs.org/toc/aaembp/6/11?ref=pdfhttps://pubs.acs.org/toc/aaembp/6/11?ref=pdfpubs.acs.org/acsaelm?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://pubs.acs.org/acsaelm?ref=pdfhttps://pubs.acs.org/acsaelm?ref=pdfhttps://acsopenscience.org/researchers/open-access/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/behavior for complementary metal oxide semiconductor(CMOS) applications.9 For p-type FETs in particular, contactquality and the metal−semiconductor interface are cruciallyimportant, where interface effects can significantly impactcontact resistance, semiconductor polarity, and subthresholdswing (S.S.). Issues may include interfacial roughness,contaminants/defects, and Fermi-level pinning (FLP) withthe formation of metal-induced gap states (MIGS), interfacedipoles, and disorder-induced gap states (DIGS).10 Reducingthe height and/or width of the subsequent Schottky barrier atthe metal−semiconductor interface is crucial to loweringcontact resistance, and strategies to mitigate these issues can bebroadly grouped into interface/contact engineering,11−15surface treatment,16 and the use of suitable contact metalswith optimized work functions.12,17,18 In particular, there iscurrently intense interest in two specific methods that improvecontact quality and enhance p-FET performance, namely, (1)transferred contact integration11,19−21 and (2) surface chargetransfer doping by oxygen plasma treatment.16,22−27 So far,these techniques have only been used separately; here, wecombine both in a single structure. Relying on either methodalone has disadvantages; for example, doping techniques havepreviously been combined with direct metal evaporation ofcontacts, which has potential to damage the 2D material.20,28On the other hand, transferred contacts have not yet beenintegrated with highly doped contact areas. Our hybridapproach applies oxygen plasma doping locally to the contactregions within a transferred contact structure, improvingcontact resistance while maintaining the high quality of theWSe2 channel. The channel remains encapsulated in hBN,thereby retaining pristine characteristics, while gate-adjacentregions are oxidized to tungsten oxide (WOx),24,29,30 creating ap-doped access region between the metal and the semi-conducting channel. We achieve strongly enhanced p-FETbehavior in ambipolar, ultrathin, three-layer (3L) WSe2 FETsand demonstrate negligible temperature-independent hyste-resis, indicating a low trap density with only back-gate control.The device shows excellent electrical stability, with highlyreproducible transfer curves between different devices as wellas for multiple contacts on the same device. We also showexcellent stability over time for at least 50 days in atmosphericconditions.■ RESULTS AND DISCUSSIONThe device concept is shown in Figures 1(a) and (b), andoptical micrographs of two samples are given in Figures 1(c)and (d). Holes are etched in a hBN flake, and metal contactsdefined by electron beam (e-beam) lithography create the viacontacts. After evaporation and lift-off, the metal partiallycovers the holes leaving a gap for local doping. The structure isassembled by a dry transfer process which avoids contactingthe 2D material with processing contaminants, preserving itshigh pristine quality, and WSe2 exfoliation and device assemblyis carried out inside a glovebox. Exposing the device to mildoxygen plasma oxidizes the WSe2 surface layer in the accessregion to form the WOx surface charge transfer doping layer.Parameters are chosen to ensure that only the top WSe2 layeris oxidized, as described in Supporting Information. Asimplified process flow is outlined in Figures 1(e) to (k),and a detailed description is given in Methods. In particular,the etching approach is able to create high-quality smootharbitrary structures in hBN.31,32 The WSe2 thickness isidentified as trilayer for both devices by atomic forcemicroscopy (AFM), Raman spectroscopy, and photolumines-cence (PL) spectroscopy, as shown in Figures S1 and S2.Distinctive advantages of our device structure includeprecise and full control over the size and location of thedoping area and the intrinsic channel length/geometry, sinceall etching and metal deposition steps and their alignment aredefined by electron beam lithography. Alternative techniquesof separately transferring hBN masks27,33,34 or hBN/top-gatestacks35 onto the 2D TMD rely on user skill thus limitingcontrol over the position and size of the doping area and thechannel. Additionally, there is potential for contamination ofthe FET channel when the TMD and hBN are transferred inseparate steps, or when spin coating a photoresist masklayer.36,37 Ours is a relatively simple method for 2D FETdesign incorporating transferred contacts and selective areadoping, with potential for systematic and precise investigationof the impact of doped region and intrinsic channel geometryon electrical performance. The channel regions are fullyencapsulated (above and below) with hBN, which is vital forensuring pristine and atomically smooth interfaces withminimized surface defects.Figure 2(a) shows transfer characteristics for sample A,before and after 60 s oxygen plasma treatment and red andblue traces, respectively. The oxidation process is performedusing a reactive ion etching (RIE) system at room temperature,15 sccm O2 flow rate, and low power of 10 W to ensureoxidation is limited to the topmost WSe2 layer for all exposuretimes used in this study. Post oxidation transfer characteristicsshow a significantly enhanced p-branch and clear off-state nearback gate voltage VBG = 0 V, with the p-FET enhancementarising from hole injection via surface charge transfer doping inthe access regions. An n-branch is still observed at positive gatevoltages, indicating that the intrinsic ambipolar nature of themechanically exfoliated WSe2 in the channel region ispreserved, even after the p-doping process, and highlightingthe efficacy of the encapsulation process. The ability tomaintain both n- and p-type behavior within a single materialcould be advantageous for beyond-CMOS technologies,9,38where both behaviors are required, and where integrating theFigure 1. (a, b) Schematic device structure illustrating WOxformation after oxidation and 3D cartoon image, respectively. (c, d)Optical images of samples A and B, respectively. Sets of two-terminalcontact groups are labeled 1 to 5. In (d), the WSe2 location is outlinedfor clarity. (e−k) Basic step-by-step fabrication process. An hBN flakearound 30 nm is selected for via contact preparation and patternedwith etched holes. Metal (PdAu) is deposited partially covering theholes leaving an open region for selective area doping. A PC/PDMSstack is used to transfer the hBN/via contact layer onto trilayer WSe2,and then both layers are transferred onto a bottom hBN flake toencapsulate the device. Finally Cr/PdAu is deposited connecting viacontacts to bond pads.ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278320https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig1&ref=pdfpubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-astwo may help reduce device footprint and enable moreversatile applications.For comparison, Figure 2(b) shows transfer characteristicsfor a control device of few-layer WSe2 on a SiO2/Si substratewithout hBN encapsulation. Prior to oxidation, the pristineWSe2 displays ambipolar transport (red trace). In contrast tothe locally doped transferred contact stacks, the entire surfaceis exposed to oxygen plasma, and the current does not turn offwithin the measurement range. This example illustrates theessential requirement for local area doping for functional p-type WSe2 FETs with the threshold voltage (Vth) within anacceptable measurement range. The control device also showsrapid degradation over time; the gray trace shows transfercharacteristics after the device is placed in atmosphericconditions for 1 day, with significantly lower current and Vthdrift indicating surface deterioration and changes in dopinglevels.In contrast, exceptionally high electrical stability isdemonstrated for our locally doped transferred contact devices,in terms of high reproducibility both between multiple contactsets on a single device and for different devices, as well as innegligible degradation over time. Figure 2(c) shows two-terminal transfer characteristics for different contact groups onsample A. The turn-on behavior, on state-current, and Vth areexceptionally similar.Sample A was oxidized for 60 s initially, as this is sufficient toconvert the top WSe2 layer to WOx (Supporting InformationFigures S1, S2, and S3). However, after this step, the IV sweepsremain slightly asymmetric, as will be discussed with referenceto Figure 4. Therefore, additional oxidation steps of 60 s eachwere performed, with a total oxidation time of 180 s resultingin more uniform doping and quasi-ohmic contacts. Based onthis finding, sample B was initially oxidized for 180 s, followedby a further 60 s to increase the uniformity of IVcharacteristics, although transfer characteristics followingboth oxidation steps are very similar.The transfer characteristics for sample B are listed in Figure2(d). Similar to sample A, the p-branch is strongly enhancedafter oxidation with Vth close to VBG = 0 V. Transfercharacteristics are also highly reproducible between samplesA and B as shown in Figure 2(e), and data are shown aftertheir initial oxidation steps of 60 and 180 s, respectively. Forcomparison, data are shown from a second control device ofWSe2 on a SiO2/Si substrate, also oxidized for 60 s, whichshows shallower turn on, lower on current, and the absence ofan n-branch at positive VBG. Since electrical stability is a keyissue for doped 2D materials,39,40 sample B is measured for upto 50 days after the second oxidation step. Figure 2(f) showstransfer characteristics as a function of day number, whilekeeping the device in the atmosphere in a desiccant cabinet.Remarkable air stability is achieved with negligible change inthe off current, on current, or Vth position.After oxidation, the channel resistance (RCH) and thecombined resistance of the metal−semiconductor contacts anddoped access regions (2RC + 2Racc) are estimated from four-point measurements, with data presented in Figure S4. It is notFigure 2. (a) Transfer characteristics of IDS as a function of VBG, sample A, before and after 60 s oxidation. Insets in (a) and (d) show schematicillustrations of contacts with the source-drain pair measured shown in red. Data are representative of device behavior since contact pairs show highreproducibility. (b) Transfer characteristics for a control device of unencapsulated WSe2 on SiO2/Si before and after 60 s oxidation, red and bluetraces, respectively. The gray trace shows device behavior after exposure to atmosphere for 1 day. (c) Transfer characteristics measured for variouscontact arrangements in sample A, defined in the legend. The inset shows a schematic diagram of the contacts where colors correspond to differenttraces in the figure. (d) Representative transfer characteristics for sample B. Data are presented in the unoxidized condition (gray), after 180 soxidation (blue), and after further 60 s oxidation (red). (e) Comparison of transfer characteristics for samples A and B and a second control deviceof unencapsulated WSe2 on an SiO2/Si substrate. Samples A and B are oxidized for 60 and 180 s, respectively. (f) Transfer characteristics forsample B after the second oxidation (total cumulative oxidation time 240 s) and after being left at atmosphere in a desiccant cabinet for 10, 20, and50 days. Data in (e) and (f) are measured for the same contact arrangement as (d).ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278321https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig2&ref=pdfpubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-aspossible to separate 2RC + 2Racc to independently estimatecontact resistance; therefore, separate devices are fabricatedwithout an hBN capping layer so that there are no transitionsbetween doped and undoped regions after oxidation, therebyallowing the contact resistance to be extracted.27,34 Weestimate a minimum individual contact resistance RC ∼ 3.6kΩ · μm and a maximum resistance reduction from the pristinestate of up to >30,000 times, shown in Figures S5 and S6.Plotting contact resistance as a function of doping density,Figure S6(d) shows a clear trend of reduced resistance withhigher doping density, indicating that the contact resistance ofour devices is limited by the doping level achieved through ouroxidation process, p2D ∼ 1.2 × 1013 cm−2. This is consistentwith the current lowest reported contact resistance fromsurface charge transfer doping by surface oxidation of 2RC ∼0.642 kΩ · μm being obtained at a higher doping density of3.94 × 1013 cm−2. The oxidation strategy used ozone plasmatreatment for 30 min at an oxygen flow rate of 3 L/min, forPd/Au contacts in exfoliated three-layer WSe2.27 Thisrelationship highlights increasing doping at the contacts asessential for achieving low contact resistance. For context,other reported values of individual contact resistance RC rangefrom ∼0.9−6.1 kΩ · μm;41 ∼1.4 kΩ · μm;30 ∼1.45 kΩ · μm;35∼3−4 kΩ · μm;24 ∼3.6 kΩ · μm34 for exfoliated WSe2 flakes,and ∼40 kΩ · μm for metal−organic chemical vapordeposition (MOCVD) grown bilayer WSe2,15 and on/offratio up to 105.For a broader contextual comparison, various otherstrategies have been employed with the goal of achievinglow-resistance contacts for p-type WSe2 FETs. An ultralowcontact resistance of 0.5 kΩ · μm was achieved using anepitaxially grown VSe2 contact layer on bilayer CVD-grownWSe2. The transistors were located between naturallyoccurring cracks in the VSe2 layer, which formed van derWaals (vdW) contacts,42 and devices demonstrated very highon-currents up to 1.72 mA· μm−1 and on/off ratios up to 106.An alternative charge transfer doping strategy, integrating few-layer α-RuCl3 at the contact regions in monolayer WSe2transistors, achieved a contact resistance of ∼4 kΩ · μm,with an on-current of 35 μA· μm−1 and on/off ratio exceeding109.43 Near-ideal vdW interfaces between TMDs and contactelectrodes have also been demonstrated using metal evapo-ration techniques, where carefully minimizing radiative heatingled to a low contact resistance of ∼3.3 kΩ · μm for Pt contactsto multilayer CVD WSe2, with saturation currents >10 μA ·μm−1 and an on/off ratio of 107. This resistance increased to∼229 kΩ · μm for monolayer WSe2, attributed to a higherSchottky barrier due to the elevated position of the valenceband edge.44 For the transferred contacts approach to contactintegration, Pt contacts embedded in hBN achieved a lowcontact resistance of ∼5 kΩ · μm for monolayer WSe2, with anon-state current of 7.4 μA · μm−1 and an on/off ratio of 108.45In another study, bilayer WSe2 p-FETs using transfer contactsdemonstrated a contact resistance of ∼3.5 kΩ · μm with Pt asthe contact metal, achieving an on/off ratio of 106 and an on-current of 5 μA · μm−1.20Returning to the electrical characteristics of our locallydoped via contact stacks. Devices display negligible hysteresisindependent of temperature; Figure 3(a) shows typical forwardand backward transfer characteristics for sample B at T = 10and 300 K. Similar characteristics are observed for sample A asshown in Figure S7(a). For context, noticeable hysteresis haspreviously been observed for plasma-doped samples with back-gate control, which modulates carriers in the contact regions,doped areas, and the channel. In these studies, devices werefabricated on SiO2 substrates, i.e., without hBN encapsulation.For example, in ref 46, significant hysteresis was present for agate voltage sweep range of ±70 V, attributed to interface trapcharges, becoming almost negligible for a ±20 V sweep range.In a separate study, noticeable room-temperature hysteresisbecame negligible below T ∼ 50 K for back-gated samples,whereas top-gated samples showed negligible hysteresis acrossall temperatures. This led to the conclusion that the hysteresiswas primarily due to water trapped at the WSe2/SiO2interface.33 Significant hysteresis has also been observed for aWSe2 device with the entire surface exposed to oxygen plasma,but it became negligible for devices with hBN covering thechannel region.34 This is consistent with other studies showingthat capping layers can significantly reduce hysteresis, asdemonstrated in ref 18 using Al2O3 grown by atomic layerdeposition (ALD). For transfer contact devices, negligiblehysteresis can be achieved when devices are fully encapsulatedwith hBN and fabricated in a glovebox;20 however, when thesame devices are fabricated in air or when metal is directlyevaporated through hBN vias, hysteresis becomes morepronounced.20 Additionally, ref 45 reported low hysteresis ofFigure 3. Hysteresis characterization. (a) Transfer characteristics for sample B at T = 10 and 300 K, for sweep rate ∼0.24 V s−1. The arrowsindicate the sweep directions. Insets: Contacts measured (red) and mobility as a function of temperature at VBG = −50 V. (b) Transfercharacteristics of a control device of WSe2 on a SiO2/Si substrate at T = 10 and 300 K. Inset: linear relation between ln(μ) as a function of T−1/3 atVBG = −50 V. (c) Carrier trap density for samples A and B and the control device as a function of temperature. Labeling, e.g., “A-C1” refers tosample (A or B) and contact pair as defined in Figures 1(c) and (d).ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278322https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig3&ref=pdfpubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as<0.5% for transfer contacts with a gate voltage sweep range of±100 V. Our device achieves negligible hysteresis consistentwith high-quality transfer contact devices,20 even with ourinclusion of an etched access region in which the WSe2 (WOx)is exposed to the atmosphere prior to (and after) oxidation.This suggests good stability of the doped contact areas, andtrap states within the WOx do not negatively impact thechannel characteristics of our devices.39 Incidentally, a highon/off ratio is achieved for sample B, up to 1 × 107, Figure3(a).We compare samples A and B with a control device, whereWSe2 is directly transferred onto a SiO2/Si substrate withouthBN encapsulation. In contrast, the control device showssignificant hysteresis, Figure 3(b), which we attribute to trapsat the SiO2/Si interface. Here the channel is fully encapsulatedin hBN in comparison to previous reports where temperature-dependent hysteresis is observed for doped access regions inWSe2 with back-gate control,33 illustrating the significance forfull encapsulation to improve device performance. To quantifythe hysteresis we estimate a corresponding change in carrierdensity due to trapped charges47 as Nit = CoxΔVth/e, whereΔVth is the difference between forward and backward sweeps atIDS around 10 pA, and Cox is capacitance per unit area.48 Figure3(c) shows data for samples A and B and the control device.The control device shows the largest carrier trap densitycompared to samples A and B, which we attribute to theabsence of an hBN layer between the WSe2 and SiO2 sinceinserting hBN can lower surface roughness and reduce disordereffects from trap sites in the SiO2. Additionally theencapsulation fabrication method where WSe2 is directlypicked up by a top hBN layer likely results in cleaner interfacesand a lower trap density. We also measure the electricalhysteresis after sample B is aged for 50 days in atmosphere, andwe find it continues to remain very low, as shown in Figure S8,Supporting Information. The IV characteristics over time arealso given for completeness in Figure S9.We also estimate the field effect mobility μ2P by a simpletwo-point measurement. The mobility is extracted as μ2P =Lgm/w Cox VDS, where the L is channel length, and w is channelwidth. The transconductance gm is estimated from linearrelation of transfer characteristics as gm = dIDS/dVBG.48 Highestroom-temperature mobilities ∼80 cm2 V−1 s−1 are obtained forsample B with a trend of increasing mobility with decreasingtemperature. A typical example is shown in Figure 3(a) inset.Data for other contacts on sample B and for sample A areshown in Figure S7. For comparison, we estimate a room-temperature mobility for sample B using a four-point (4P)measurement of around 145 cm2 V−1 s−1. The higher 4Pestimate is consistent with inclusion of channel and contactresistance in the two-point measurement.16,48 As shown inFigure 3(b) inset, the mobility for the control device shows alinear trend in ln(μ) as a function of T−1/3, consistent with 2Dvariable range hopping from disorder-induced localizedstates.49,50 This can be attributed this to greater disorderfrom surface roughness, dangling bonds, and impuritiescreating traps at the WSe2/SiO2 interface.51Reducing the Schottky barrier (SB) is essential to reducingcontact resistance.9 We find a reduction of both barrier heightFigure 4. IV sweeps and Schottky barrier height/width extraction. (a) Pristine IV characteristics of sample A at VBG = −50 V as a function oftemperature, for the contact arrangement shown in the inset. (b, c) Corresponding IV characteristics after 60 s oxidation at fixed gate voltages fromVBG = −10 to −50 V in steps of −10 V and after different oxidation times at VBG = −50 V, respectively, at T = 300 K. (d) IV characteristics ofsample B after 240 s oxidation at fixed gate voltages from VBG = −10 to −50 V in steps of −10 V at T = 300 K. The inset shows the contactarrangement. (e) Fowler-Nordheim (FN) plot of ( )ln IVDSDS2 as a function ofV1DSat T = 10 to 300 K in steps of 10 K. The linear fit to T = 10 K data(black dashed line) is used to estimate barrier width. (f) Corresponding transfer characteristics, the blue dash-dotted line shows a fit to theexponential subthreshold behavior. The deviation from this trend identifies the flat band current and voltage and thereby Schottky barrier height.This is subsequently used in (e) to extract barrier width by fitting data at 10 K using the FN-tunneling relation. (d, e) Data for sample B for thecontact arrangement shown in the inset of (f). Data for other contacts are shown in Supporting Information.ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278323https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?fig=fig4&ref=pdfpubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asand width for the locally doped transferred contact stacks,evidenced in analysis of tunneling mechanisms and also ingreater hole injection, i.e., higher hole currents after oxidation.We note that the utilization of transferred contacts11,20 andsurface charge transfer doping from WOx formation16,29 canboth lead to Fermi-level depinning and reducing the SB,although estimating the pinning factor requires comparison ofdifferent metal work functions.11,29 Prior to oxidation, i.e., inthe pristine condition, low-temperature data typically showsflat regions around VDS = 0 V characteristic of tunnelingbehavior, as shown in Figure 4(a) for sample A. CorrespondingIV characteristics as a function of VBG after 60 s oxidation areshown in Figure 4(b). The behavior is slightly asymmetricaround VDS = 0 V but is neither strictly ohmic nor Schottkyand may suggest the doping concentration in the accessregions may be nonuniform or potential barriers at the sourceand drain contacts may be asymmetric. We therefore performrepeat O2 plasma oxidation steps for 60 s each time, as shownin Figure 4(c), and the total oxidation time is given in thelegend. A linear trend appears after 180 s oxidation, suggestingmore uniform doping concentration and nearly ohmiccontacts. Similar data are obtained for other contacts tosample A, as shown in Figure S11. Sample B is thereforeexposed to an initial 180 s oxygen plasma treatment. However,we find that a further 60 s of oxidation is required to producemore symmetric and quasi-ohmic characteristics, as shown inFigure 4(d). We speculate that if there is any directionality toexposure during the oxygen plasma treatment process, thedepth of the etched holes may cause shadowing of particularregions and therefore nonuniform exposure, necessitating theneed for a second oxidation step to produce more uniformdoping despite the 180 s initial exposure time. This couldpotentially be investigated by mounting the sample on anangled stage and continuously rotating it during the oxidationprocess to ensure uniform plasma exposure across all regions.Additionally, oxidation parameters such as plasma power, flowrate, and exposure time could be further optimized to improveuniformity. Reducing the thickness of the top hBN cappinglayer may also help minimize any shadowing effects.To further investigate the transport mechanisms, data arerepresented in Fowler-Nordheim (FN) plots, i.e., ( )ln IVDSDS2 as afunction ofV1DS. Representative data for sample B are shown inFigure 4(e). Briefly, transport through a tunnel barrier at lowT, i.e., where thermionic emission is negligible, can becharacterized either by direct tunnelling (DT) or FNtunneling.52,53 This analysis was initially motivated by theobservation of tunneling-like characteristics in Figure 4(a).Direct tunnelling, where current tunnels through the entirewidth of the barrier at low source-drain bias, is described byIV Vd mln ln1 2 2DSDS2DSikjjjjjy{zzzzzikjjjjjy{zzzzz (1)For sufficiently high source-drain bias, FN tunneling occursand current tunnels through a triangular region at the top ofthe barrier expressed byIVd me Vln4 231DSDS23DSikjjjjjy{zzzzz (2)where d is barrier width, m is effective mass, ϕ is barrier height,ℏ is the reduced Planck constant, and e is the unit of electroniccharge. Our devices show a transition from DT to FNtunneling as bias increases, Figure 4(e). Data for sample Acorresponding to Figure 4(a) are plotted in an FN manner inFigure S10, and data for other contact groups for sample B areshown in Figure S12. We estimate the slope of the lowesttemperature data, T = 10 K in the FN regime, indicated by thedashed black line in Figure 4(e). The lowest available T data isused to remove the contribution of thermionic emission, andthe slope can be related to barrier height ϕ and barrier width daccording to eq 2. We first independently estimate barrierheight ϕ is from transfer characteristics at flat band current andvoltage as shown in Figure 4(f).27,54,55 This value is then usedin eq 2 to estimate d. Values between ϕ ∼ 260 to 315 meV andd ∼ 0.3 to 0.7 nm are estimated for different contact groups onsample B, Figure S12. For comparison, the van der Waals gapthat exists for transferred contacts ∼0.3 nm.29 Prior tooxidation, we estimate ϕ ∼ 395 meV and d ∼ 2.3 nm forsample A, Figure S10. It has been suggested that a scenariowhere the WOx region does not extend under contacts resultsin a reduction of the Schottky barrier width instead of height,33whereas extending far under contacts would change the barrierheight rather than width. Our estimates appear to be a mix ofthese scenarios but with a larger relative change of d comparedto ϕ, suggesting the barrier width change is the majormechanism for greater charge injection into the channel, andthis device architecture and fabrication process mostly controlsthe barrier width. In our work, we employ high work functionPdAu contacts, for which p-type characteristics are expected todominate and the relatively high work functions of these metalsfacilitate hole injection. In light of previous studies,11,12,16metals with even higher work functions such as Pt may resultin stronger p-type behavior for TMDs, along with higher on-currents due to improved hole injection efficiency. In theabsence of Fermi-level pinning, using Pt could potentiallyfurther reduce the Schottky barrier height and, thus, enhancethe p-FET performance. This presents a promising avenue forfuture work aimed at optimizing contact resistance andmaximizing the device performance.■ CONCLUSIONSWe have demonstrated a contact-doping approach compatiblewith transferred contacts, achieving dramatic contact resistancereduction and enabling negligible, temperature-independenthysteresis in transport measurements of WSe2. In particular,the hole injection is greatly increased through the reduction ofthe Schottky barrier in WSe2. High electrical stability isachieved both in terms of high reproducibility in transfercharacteristics for different contact arrangements and differentsamples, as well as excellent air stability for up to 50 days atatmosphere. The results highlight the importance ofheterostructure design for WSe2 FETs, particularly the needfor full encapsulation with a 2D dielectric when doped accessregions are incorporated by selective oxidation. By definingaccess regions and contacts a priori in the hBN dielectric byelectron beam lithography, we achieve precise control of thedoped region location and size, as well as channel geometry.Incorporating contacts embedded in hBN directly transferredonto the WSe2 material minimizes contamination andpreserves the pristine nature of the TMD channel. This is apromising approach for the development of doped TMDstructures with high-quality channel and contacts. Our resultsACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278324https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfpubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asalso highlight achieving higher-density doping at contacts asbeing highly important for achieving low-resistance p-typecontacts for WSe2.■ METHODSDevice Fabrication and Characterization. WSe2 from HQgraphene is mechanically exfoliated on a 285 nm SiO2 substrate, andthe trilayer WSe2 is identified by optical contrast. The thickness isfurther confirmed by atomic force microscopy (AFM), Raman shift,and photoluminescence spectroscopy (PL) measurements, asdescribed in Supporting Information. Hexagonal boron nitride(hBN) is also mechanically exfoliated on a SiO2 substrate andannealed by ramping the temperature to 300 °C in 1 h, maintaining at300 °C for 3 h, then naturally cooling to room temperature. Around30 nm thick hBN is used for via contact fabrication. The via holepatterns are defined in the hBN using e-beam lithography after spincoating the chip with poly(methyl methacrylate) (PMMA) A4photoresist. After developing, the hBN is etched using an InductiveCoupled Plasma (ICP) system with parameters Ar (20 sccm), O2 (5sccm), and SF6 (20 sccm), for a total time of 120 s. The sample isthen cleaned using Propylene Glycol (PG), acetone (ACE), thenisopropanol (IPA) to remove the photoresist. The via contact metalpattern is then defined by e-beam lithography after spin coating thechip with PMMA (A4). Around 35−40 nm PdAu is deposited,partially covering the etched areas in the hBN such that a holeremains adjacent to the contacts for selective area doping. A lift-offprocess is used to remove unwanted metal by rinsing the chip in ACEthen IPA. For the dry transfer process, Polydimethylsiloxane (PDMS)covered with a Polycarbonate (PC) thin film is used as a stamp andused to pick up the via contact hBN. We then pick up the trilayerWSe2 and transfer to the bottom hBN flake, thus fully encapsulatingthe device. The entire stack is then annealed by heating to 300 °C in 1h, maintaining at 300 °C for 1 h, and then allowed to cool naturally toroom temperature. We used e-beam lithography to define contactmetal tracks and bond pads, followed by e-beam evaporation of Cr/PdAu 10/80 nm. A final lift-off process is required to remove theunwanted metal. The gate electrode and etched hole layout andgeometry are identical for samples A and B. The only difference is thatsample A is etched into a rectangular shape to ensure the via contactmetal crosses the WSe2 edge. For sample B, the via contacts design isadjusted so that contacts cross the edge without an extra etchingprocess.Electrical Measurements. Electrical transport measurements areperformed using a Keysight B1500A semiconductor parameteranalyzer and LakeShore cryogenic probe station between 10−2 and10−3 mBar.■ ASSOCIATED CONTENT*sı Supporting InformationThe Supporting Information is available free of charge athttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574.WSe2 thickness calibration; Conductive properties of theWOx oxide layer; Samples A and B channel andcombined contact/access region resistance; Contactresistance estimate; Hysteresis and mobility estimates;Aging characteristics; Fowler-Nordheim analysis, sampleA; Oxidation time dependence of IV characteristics;Schottky barrier height and width estimates (PDF)■ AUTHOR INFORMATIONCorresponding AuthorLuke W. Smith − Department of Physics, Department ofPhysics and Center for Quantum Frontiers of Research &Technology (QFort), and Academy of InnovativeSemiconductor and Sustainable Manufacturing, NationalCheng Kung University, Tainan 701, Taiwan; orcid.org/0000-0003-2194-4708; Email: lukesmith@phys.ncku.edu.twAuthorsHe-Yu Chen − Department of Physics, National Cheng KungUniversity, Tainan 701, Taiwan; orcid.org/0009-0004-7101-9179Jheng-Jie Lin − Department of Physics, National Cheng KungUniversity, Tainan 701, Taiwan; orcid.org/0009-0002-1263-9453Sheng-Shong Wong − Department of Physics, National ChengKung University, Tainan 701, TaiwanZhen-You Lin − Department of Physics, National Cheng KungUniversity, Tainan 701, TaiwanYu-Chiang Hsieh − Department of Physics, National ChengKung University, Tainan 701, Taiwan; orcid.org/0009-0001-3346-9548Kuo-En Chang − Department of Physics, National ChengKung University, Tainan 701, TaiwanChung-Lin Wu − Department of Physics and Department ofPhysics and Center for Quantum Frontiers of Research &Technology (QFort), National Cheng Kung University,Tainan 701, TaiwanKenji Watanabe − Research Center for Electronic and OpticalMaterials, National Institute for Materials Science, Tsukuba305-0044, Japan; orcid.org/0000-0003-3701-8119Takashi Taniguchi − Research Center for MaterialsNanoarchitectonics, National Institute for Materials Science,Tsukuba 305-0044, Japan; orcid.org/0000-0002-1467-3105Tse-Ming Chen − Department of Physics and Department ofPhysics and Center for Quantum Frontiers of Research &Technology (QFort), National Cheng Kung University,Tainan 701, TaiwanComplete contact information is available at:https://pubs.acs.org/10.1021/acsaelm.4c01574NotesThe authors declare no competing financial interest.■ ACKNOWLEDGMENTSThis work is supported by the National Science andTechnology Council in Taiwan (grant number 111-2112-M-006-036-MY3) and the Higher Education Sprout Project,Ministry of Education to the Headquarters of UniversityAdvancement at the National Cheng Kung University(NCKU). L.W.S., T.-M.C., and C.-L.W. acknowledge supportfrom TSMC Advanced Research Project (ARP). K.W. andT.T. acknowledge support from the JSPS KAKENHI (GrantNumbers 21H05233 and 23H02052) and World PremierInternational Research Center Initiative (WPI), MEXT, Japan.■ REFERENCES(1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang,Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effectin Atomically Thin Carbon Films. Science 2004, 306, 666−669.(2) Zhang, Y.; Tan, Y.-W.; Stormer, H. L.; Kim, P. Experimentalobservation of the quantum Hall effect and Berry’s phase in graphene.Nature 2005, 438, 201−204.(3) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.;Katsnelson, M. I.; Grigorieva, I. V.; Dubonos, S. V.; Firsov, A. A. Two-dimensional gas of massless Dirac fermions in graphene. Nature 2005,438, 197−200.ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278325https://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?goto=supporting-infohttps://pubs.acs.org/doi/suppl/10.1021/acsaelm.4c01574/suppl_file/el4c01574_si_001.pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Luke+W.+Smith"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0000-0003-2194-4708https://orcid.org/0000-0003-2194-4708mailto:lukesmith@phys.ncku.edu.twmailto:lukesmith@phys.ncku.edu.twhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="He-Yu+Chen"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0009-0004-7101-9179https://orcid.org/0009-0004-7101-9179https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Jheng-Jie+Lin"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0009-0002-1263-9453https://orcid.org/0009-0002-1263-9453https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Sheng-Shong+Wong"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Zhen-You+Lin"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Yu-Chiang+Hsieh"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0009-0001-3346-9548https://orcid.org/0009-0001-3346-9548https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kuo-En+Chang"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Chung-Lin+Wu"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kenji+Watanabe"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0000-0003-3701-8119https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takashi+Taniguchi"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0000-0002-1467-3105https://orcid.org/0000-0002-1467-3105https://pubs.acs.org/action/doSearch?field1=Contrib&text1="Tse-Ming+Chen"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/doi/10.1021/acsaelm.4c01574?ref=pdfhttps://doi.org/10.1126/science.1102896https://doi.org/10.1126/science.1102896https://doi.org/10.1038/nature04235https://doi.org/10.1038/nature04235https://doi.org/10.1038/nature04233https://doi.org/10.1038/nature04233pubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as(4) Wang, S.; Liu, X.; Zhou, P. The Road for 2D Semiconductors inthe Silicon Age. Adv. Mater. 2022, 34, 2106886.(5) Wang, S.; Liu, X.; Xu, M.; Liu, L.; Yang, D.; Zhou, P. Two-dimensional devices and integration towards the silicon lines. Nat.Mater. 2022, 21, 1225−1239.(6) Huang, X.; Liu, C.; Zhou, P. 2D semiconductors for specificelectronic applications: from device to system. npj 2D Mater. Appl.2022, 6, 51.(7) Patoary, N. H.; Xie, J.; Zhou, G.; Al Mamun, F.; Sayyad, M.;Tongay, S.; Esqueda, I. S. Improvements in 2D p-type WSe2transistors towards ultimate CMOS scaling. Sci. Rep. 2023, 13, 3304.(8) Schram, T.; Sutar, S.; Radu, I.; Asselberghs, I. Challenges ofWafer-Scale Integration of 2D Semiconductors for High-PerformanceTransistor Circuits. Adv. Mater. 2022, 34, 2109796.(9) Knobloch, T.; Selberherr, S.; Grasser, T. Challenges fornanoscale CMOS logic based on two-dimensional materials. Nano-materials (Basel) 2022, 12, 3548.(10) Liu, X.; Choi, M. S.; Hwang, E.; Yoo, W. J.; Sun, J. Fermi LevelPinning Dependent 2D Semiconductor Devices: Challenges andProspects. Adv. Mater. 2022, 34, 2108425.(11) Liu, Y.; Guo, J.; Zhu, E.; Liao, L.; Lee, S.-J.; Ding, M.; Shakir, I.;Gambin, V.; Huang, Y.; Duan, X. Approaching the Schottky−Mottlimit in van der Waals metal−semiconductor junctions. Nature 2018,557, 696−700.(12) Ngo, T. D.; Yang, Z.; Lee, M.; Ali, F.; Moon, I.; Kim, D. G.;Taniguchi, T.; Watanabe, K.; Lee, K.-Y.; Yoo, W. J. Fermi-LevelPinning Free High-Performance 2D CMOS Inverter Fabricated withVan Der Waals Bottom Contacts. Advanced Electronic Materials 2021,7, 2001212.(13) Jang, J.; Ra, H.-S.; Ahn, J.; Kim, T. W.; Song, S. H.; Park, S.;Taniguch, T.; Watanabe, K.; Lee, K.; Hwang, D. K. Fermi-LevelPinning-Free WSe2 Transistors via 2D Van der Waals Metal Contactsand Their Circuits. Adv. Mater. 2022, 34, 2109899.(14) Zheng, Y.; Gao, J.; Han, C.; Chen, W. Ohmic ContactEngineering for Two-Dimensional Materials. Cell Reports PhysicalScience 2021, 2, 100298.(15) Oberoi, A.; Han, Y.; Stepanoff, S. P.; Pannone, A.; Sun, Y.; Lin,Y.-C.; Chen, C.; Shallenberger, J. R.; Zhou, D.; Terrones, M.;Redwing, J. M.; Robinson, J. A.; Wolfe, D. E.; Yang, Y.; Das, S.Toward High-Performance p-Type Two-Dimensional Field EffectTransistors: Contact Engineering, Scaling, and Doping. ACS Nano2023, 17, 19709−19723.(16) Lee, K.; Ngo, T. D.; Lee, S.; Shin, H.; Choi, M. S.; Hone, J.;Yoo, W. J. Effects of Oxygen Plasma Treatment on Fermi-LevelPinning and Tunneling at the Metal−Semiconductor Interface ofWSe2 FETs. Advanced Electronic Materials 2023, 9, 2200955.(17) Nakajima, R.; Nishimura, T.; Ueno, K.; Nagashio, K. WorkFunction Modulation of Bi/Au Bilayer System toward p-Type WSe2FET. ACS Applied Electronic Materials 2024, 6, 144−149.(18) Pendurthi, R.; et al. Monolithic three-dimensional integrationof complementary two-dimensional field-effect transistors. Nat.Nanotechnol. 2024, 19, 970−977.(19) Telford, E. J.; Benyamini, A.; Rhodes, D.; Wang, D.; Jung, Y.;Zangiabadi, A.; Watanabe, K.; Taniguchi, T.; Jia, S.; Barmak, K.;Pasupathy, A. N.; Dean, C. R.; Hone, J. Via Method for LithographyFree Contact and Preservation of 2D Materials. Nano Lett. 2018, 18,1416−1420.(20) Jung, Y.; Choi, M. S.; Nipane, A.; Borah, A.; Kim, B.;Zangiabadi, A.; Taniguchi, T.; Watanabe, K.; Yoo, W. J.; Hone, J.;Teherani, J. T. Transferred via contacts as a platform for ideal two-dimensional transistors. Nature Electronics 2019, 2, 187−194.(21) Satterthwaite, P. F.; et al. Van der Waals device integrationbeyond the limits of van der Waals forces using adhesive matrixtransfer. Nature Electronics 2024, 7, 17−28.(22) Yamamoto, M.; Dutta, S.; Aikawa, S.; Nakaharai, S.;Wakabayashi, K.; Fuhrer, M. S.; Ueno, K.; Tsukagoshi, K. Self-Limiting Layer-by-Layer Oxidation of Atomically Thin WSe2. NanoLett. 2015, 15, 2067−2073.(23) Yamamoto, M.; Nakaharai, S.; Ueno, K.; Tsukagoshi, K. Self-Limiting Oxides on WSe2 as Controlled Surface Acceptors and Low-Resistance Hole Contacts. Nano Lett. 2016, 16, 2720−2727.(24) Wang, S.; Zhao, W.; Giustiniano, F.; Eda, G. Effect of oxygenand ozone on p-type doping of ultra-thin WSe2 and MoSe2 field effecttransistors. Phys. Chem. Chem. Phys. 2016, 18, 4304−4309.(25) Kang, W.-M.; Lee, S.; Cho, I.-T.; Park, T. H.; Shin, H.; Hwang,C. S.; Lee, C.; Park, B.-G.; Lee, J.-H. Multi-layer WSe2 field effecttransistor with improved carrier-injection contact by using oxygenplasma treatment. Solid-State Electron. 2018, 140, 2−7.(26) Sivan, M.; Li, Y.; Veluri, H.; Zhao, Y.; Tang, B.; Wang, X.;Zamburg, E.; Leong, J. F.; Niu, J. X.; Chand, U.; Thean, A. V.-Y. AllWSe2 1T1R resistive RAM cell for future monolithic 3D embeddedmemory integration. Nat. Commun. 2019, 10, 5201.(27) Borah, A.; Nipane, A.; Choi, M. S.; Hone, J.; Teherani, J. T.Low-Resistance p-Type Ohmic Contacts to Ultrathin WSe2 by Usinga Monolayer Dopant. ACS Applied Electronic Materials 2021, 3, 2941−2947.(28) Kong, L.; Zhang, X.; Tao, Q.; Zhang, M.; Dang, W.; Li, Z.;Feng, L.; Liao, L.; Duan, X.; Liu, Y. Doping-free complementaryWSe2 circuit via van der Waals metal integration. Nat. Commun.2020, 11, 1866.(29) Ngo, T. D.; Choi, M. S.; Lee, M.; Ali, F.; Yoo, W. J.Anomalously persistent p-type behavior of WSe2 field-effecttransistors by oxidized edge-induced Fermi-level pinning. J. Mater.Chem. C 2022, 10, 846−853.(30) Moon, I.; Lee, S.; Lee, M.; Kim, C.; Seol, D.; Kim, Y.; Kim, K.H.; Yeom, G. Y.; Teherani, J. T.; Hone, J.; Yoo, W. J. The device levelmodulation of carrier transport in a 2D WSe2 field effect transistor viaa plasma treatment. Nanoscale 2019, 11, 17368−17375.(31) Hsieh, Y.-C.; Lin, Z.-Y.; Fung, S.-J.; Lu, W.-S.; Ho, S.-C.; Hong,S.-P.; Ho, S.-Z.; Huang, C.-H.; Watanabe, K.; Taniguchi, T.; Chan, Y.-H.; Chen, Y.-C.; Wu, C.-L.; Chen, T.-M. Engineering the Strain andInterlayer Excitons of 2D Materials via Lithographically EngravedHexagonal Boron Nitride. Nano Lett. 2023, 23, 7244−7251.(32) Ho, S.-C.; Chang, C.-H.; Hsieh, Y.-C.; Lo, S.-T.; Huang, B.; Vu,T.-H.-Y.; Ortix, C.; Chen, T.-M. Hall effects in artificially corrugatedbilayer graphene without breaking time-reversal symmetry. NatureElectronics 2021, 4, 116−125.(33) Kato, R.; Uchiyama, H.; Nishimura, T.; Ueno, K.; Taniguchi,T.; Watanabe, K.; Chen, E.; Nagashio, K. p-Type Conversion of WS2and WSe2 by Position-Selective Oxidation Doping and Its Applicationin Top Gate Transistors. ACS Appl. Mater. Interfaces 2023, 15,26977−26984.(34) Liu, X.; Pan, Y.; Yang, J.; Qu, D.; Li, H.; Yoo, W. J.; Sun, J.High performance WSe2 p-MOSFET with intrinsic n-channel basedon back-to-back p−n junctions. Appl. Phys. Lett. 2021, 118, 233101.(35) Ngo, T. D.; Huynh, T.; Moon, I.; Taniguchi, T.; Watanabe, K.;Choi, M. S.; Yoo, W. J. Self-Aligned Top-Gate Structure in High-Performance 2D p-FETs via van der Waals Integration and ContactSpacer Doping. Nano Lett. 2023, 23, 11345−11352.(36) Arnold, A. J.; Schulman, D. S.; Das, S. Thickness Trends ofElectron and Hole Conduction and Contact Carrier Injection inSurface Charge Transfer Doped 2D Field Effect Transistors. ACSNano 2020, 14, 13557−13568.(37) Im, H. H.; Lee, D.; Lee, G.; Kim, H.-Y.; Kim, J. Self-AlignedContact Doping of WSe2 Metal−Insulator−Semiconductor Field-Effect Transistors Using Hydrogen Silsesquioxane. ACS AppliedElectronic Materials 2023, 5, 2394−2400.(38) Ren, Y.; Yang, X.; Zhou, L.; Mao, J.-Y.; Han, S.-T.; Zhou, Y.Recent Advances in Ambipolar Transistors for Functional Applica-tions. Adv. Funct. Mater. 2019, 29, 1902105.(39) Ho, P.-H.; Chang, J.-R.; Chen, C.-H.; Hou, C.-H.; Chiang, C.-H.; Shih, M.-C.; Hsu, H.-C.; Chang, W.-H.; Shyue, J.-J.; Chiu, Y.-P.;Chen, C.-W. Hysteresis-Free Contact Doping for High-PerformanceTwo-Dimensional Electronics. ACS Nano 2023, 17, 2653−2660.(40) Kwon, S.-J.; Han, T.-H.; Ko, T. Y.; Li, N.; Kim, Y.; Kim, D. J.;Bae, S.-H.; Yang, Y.; Hong, B. H.; Kim, K. S.; Ryu, S.; Lee, T.-W.ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278326https://doi.org/10.1002/adma.202106886https://doi.org/10.1002/adma.202106886https://doi.org/10.1038/s41563-022-01383-2https://doi.org/10.1038/s41563-022-01383-2https://doi.org/10.1038/s41699-022-00327-3https://doi.org/10.1038/s41699-022-00327-3https://doi.org/10.1038/s41598-023-30317-4https://doi.org/10.1038/s41598-023-30317-4https://doi.org/10.1002/adma.202109796https://doi.org/10.1002/adma.202109796https://doi.org/10.1002/adma.202109796https://doi.org/10.3390/nano12203548https://doi.org/10.3390/nano12203548https://doi.org/10.1002/adma.202108425https://doi.org/10.1002/adma.202108425https://doi.org/10.1002/adma.202108425https://doi.org/10.1038/s41586-018-0129-8https://doi.org/10.1038/s41586-018-0129-8https://doi.org/10.1002/aelm.202001212https://doi.org/10.1002/aelm.202001212https://doi.org/10.1002/aelm.202001212https://doi.org/10.1002/adma.202109899https://doi.org/10.1002/adma.202109899https://doi.org/10.1002/adma.202109899https://doi.org/10.1016/j.xcrp.2020.100298https://doi.org/10.1016/j.xcrp.2020.100298https://doi.org/10.1021/acsnano.3c03060?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.3c03060?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1002/aelm.202200955https://doi.org/10.1002/aelm.202200955https://doi.org/10.1002/aelm.202200955https://doi.org/10.1021/acsaelm.3c01091?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.3c01091?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.3c01091?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41565-024-01705-2https://doi.org/10.1038/s41565-024-01705-2https://doi.org/10.1021/acs.nanolett.7b05161?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.7b05161?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41928-019-0245-yhttps://doi.org/10.1038/s41928-019-0245-yhttps://doi.org/10.1038/s41928-023-01079-8https://doi.org/10.1038/s41928-023-01079-8https://doi.org/10.1038/s41928-023-01079-8https://doi.org/10.1021/nl5049753?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/nl5049753?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.6b00390?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.6b00390?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.6b00390?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1039/C5CP07194Ahttps://doi.org/10.1039/C5CP07194Ahttps://doi.org/10.1039/C5CP07194Ahttps://doi.org/10.1016/j.sse.2017.10.008https://doi.org/10.1016/j.sse.2017.10.008https://doi.org/10.1016/j.sse.2017.10.008https://doi.org/10.1038/s41467-019-13176-4https://doi.org/10.1038/s41467-019-13176-4https://doi.org/10.1038/s41467-019-13176-4https://doi.org/10.1021/acsaelm.1c00225?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.1c00225?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41467-020-15776-xhttps://doi.org/10.1038/s41467-020-15776-xhttps://doi.org/10.1039/D1TC04148Ghttps://doi.org/10.1039/D1TC04148Ghttps://doi.org/10.1039/C9NR05881Hhttps://doi.org/10.1039/C9NR05881Hhttps://doi.org/10.1039/C9NR05881Hhttps://doi.org/10.1021/acs.nanolett.3c01208?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c01208?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c01208?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41928-021-00537-5https://doi.org/10.1038/s41928-021-00537-5https://doi.org/10.1021/acsami.3c04052?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsami.3c04052?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsami.3c04052?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1063/5.0036343https://doi.org/10.1063/5.0036343https://doi.org/10.1021/acs.nanolett.3c04009?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c04009?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c04009?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.0c05572?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.0c05572?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.0c05572?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.3c00211?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.3c00211?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaelm.3c00211?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1002/adfm.201902105https://doi.org/10.1002/adfm.201902105https://doi.org/10.1021/acsnano.2c10631?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.2c10631?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-aspubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asExtremely stable graphene electrodes doped with macromolecularacid. Nat. Commun. 2018, 9, 2037.(41) Lee, D.; Choi, Y.; Kim, J.; Kim, J. Recessed-Channel WSe2Field-Effect Transistor via Self-Terminated Doping and Layer-by-Layer Etching. ACS Nano 2022, 16, 8484−8492.(42) Wu, R.; et al. Bilayer tungsten diselenide transistors with on-state currents exceeding 1.5 milliamperes per micrometre. Nature.Electronics 2022, 5, 497−504.(43) Xie, J.; et al. Low Resistance Contact to P-Type MonolayerWSe2. Nano Lett. 2024, 24, 5937−5943.(44) Wang, Y.; Kim, J. C.; Li, Y.; Ma, K. Y.; Hong, S.; Kim, M.; Shin,H. S.; Jeong, H. Y.; Chhowalla, M. P-type electrical contacts for 2Dtransition-metal dichalcogenides. Nature 2022, 610, 61−66.(45) Liu, Y.; Liu, S.; Wang, Z.; Li, B.; Watanabe, K.; Taniguchi, T.;Yoo, W. J.; Hone, J. Low-resistance metal contacts to encapsulatedsemiconductor monolayers with long transfer length. NatureElectronics 2022, 5, 579−585.(46) Ngo, T. D.; Choi, M. S.; Lee, M.; Ali, F.; Hassan, Y.; Ali, N.;Liu, S.; Lee, C.; Hone, J.; Yoo, W. J. Selective Electron BeamPatterning of Oxygen-Doped WSe2 for Seamless Lateral JunctionTransistors. Advanced Science 2022, 9, 2202465.(47) Alexander-Webber, J. A; Sagade, A. A; Aria, A. I; VanVeldhoven, Z. A; Braeuninger-Weimer, P.; Wang, R.; Cabrero-Vilatela, A.; Martin, M.-B.; Sui, J.; Connolly, M. R; Hofmann, S.Encapsulation of graphene transistors and vertical device integrationby interface engineering with atomic layer deposited oxide. 2DMaterials 2017, 4, No. 011008.(48) Mitta, S. B.; Choi, M. S.; Nipane, A.; Ali, F.; Kim, C.; Teherani,J. T.; Hone, J.; Yoo, W. J. Electrical characterization of 2D materials-based field-effect transistors. 2D Materials 2021, 8, No. 012002.(49) Pradhan, N. R.; Rhodes, D.; Memaran, S.; Poumirol, J. M.;Smirnov, D.; Talapatra, S.; Feng, S.; Perea-Lopez, N.; Elias, A. L.;Terrones, M.; Ajayan, P. M.; Balicas, L. Hall and field-effect mobilitiesin few layered p-WSe2 field-effect transistors. Sci. Rep. 2015, 5, 8979.(50) Yu, Z.; Ong, Z.-Y.; Li, S.; Xu, J.-B.; Zhang, G.; Zhang, Y.-W.;Shi, Y.; Wang, X. Analyzing the Carrier Mobility in Transition-MetalDichalcogenide MoS2 Field-Effect Transistors. Adv. Funct. Mater.2017, 27, 1604093.(51) Chan, M. Y.; Komatsu, K.; Li, S.-L.; Xu, Y.; Darmawan, P.;Kuramochi, H.; Nakaharai, S.; Aparecido-Ferreira, A.; Watanabe, K.;Taniguchi, T.; Tsukagoshi, K. Suppression of thermally activatedcarrier transport in atomically thin MoS2 on crystalline hexagonalboron nitride substrates. Nanoscale 2013, 5, 9572−9576.(52) Beebe, J. M.; Kim, B.; Gadzuk, J. W.; Daniel Frisbie, C.;Kushmerick, J. G. Transition from Direct Tunneling to Field Emissionin Metal-Molecule-Metal Junctions. Phys. Rev. Lett. 2006, 97,No. 026801.(53) Smith, L. W.; Batey, J. O.; Alexander-Webber, J. A.; Hsieh, Y.-C.; Fung, S.-J.; Albrow-Owen, T.; Beere, H. E.; Burton, O. J.;Hofmann, S.; Ritchie, D. A.; Kelly, M.; Chen, T.-M.; Joyce, H. J.;Smith, C. G. Giant Magnetoresistance in a Chemical VaporDeposition Graphene Constriction. ACS Nano 2022, 16, 2833−2842.(54) Das, S.; Chen, H.-Y.; Penumatcha, A. V.; Appenzeller, J. HighPerformance Multilayer MoS2 Transistors with Scandium Contacts.Nano Lett. 2013, 13, 100−105.(55) Alharbi, A.; Shahrjerdi, D. Analyzing the Effect of High-kDielectric-Mediated Doping on Contact Resistance in Top-GatedMonolayer MoS2 Transistors. IEEE Trans. Electron Devices 2018, 65,4084−4092.ACS Applied Electronic Materials pubs.acs.org/acsaelm Articlehttps://doi.org/10.1021/acsaelm.4c01574ACS Appl. Electron. Mater. 2024, 6, 8319−83278327https://doi.org/10.1038/s41467-018-04385-4https://doi.org/10.1038/s41467-018-04385-4https://doi.org/10.1021/acsnano.2c03402?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.2c03402?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.2c03402?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41928-022-00800-3https://doi.org/10.1038/s41928-022-00800-3https://doi.org/10.1021/acs.nanolett.3c04195?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.nanolett.3c04195?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1038/s41586-022-05134-whttps://doi.org/10.1038/s41586-022-05134-whttps://doi.org/10.1038/s41928-022-00808-9https://doi.org/10.1038/s41928-022-00808-9https://doi.org/10.1002/advs.202202465https://doi.org/10.1002/advs.202202465https://doi.org/10.1002/advs.202202465https://doi.org/10.1088/2053-1583/4/1/011008https://doi.org/10.1088/2053-1583/4/1/011008https://doi.org/10.1088/2053-1583/abc187https://doi.org/10.1088/2053-1583/abc187https://doi.org/10.1038/srep08979https://doi.org/10.1038/srep08979https://doi.org/10.1002/adfm.201604093https://doi.org/10.1002/adfm.201604093https://doi.org/10.1039/c3nr03220ehttps://doi.org/10.1039/c3nr03220ehttps://doi.org/10.1039/c3nr03220ehttps://doi.org/10.1103/PhysRevLett.97.026801https://doi.org/10.1103/PhysRevLett.97.026801https://doi.org/10.1021/acsnano.1c09815?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsnano.1c09815?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/nl303583v?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/nl303583v?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1109/TED.2018.2866772https://doi.org/10.1109/TED.2018.2866772https://doi.org/10.1109/TED.2018.2866772pubs.acs.org/acsaelm?ref=pdfhttps://doi.org/10.1021/acsaelm.4c01574?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as