# Fileset

[Kii_2025_Nano_Ex._6_025008.pdf](https://mdr.nims.go.jp/filesets/d0bb450a-22b2-4b9c-8379-fb43d4ae49d2/download)

## Creator

Hiroki Kii, [Naoka Nagamura](https://orcid.org/0000-0002-7697-8983), [Ryo Nouchi](https://orcid.org/0000-0002-7232-4827)

## Rights

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

## Other metadata

[Plasma treatment of MoS<sub>2</sub> field-effect transistors using solid-state fluorine source](https://mdr.nims.go.jp/datasets/a1d18194-5004-4b8a-b31d-ece635841009)

## Fulltext

Plasma treatment of MoS2 field-effect transistors using solid-state fluorine sourceNano Express            PAPER • OPEN ACCESSPlasma treatment of MoS2 field-effect transistorsusing solid-state fluorine sourceTo cite this article: Hiroki Kii et al 2025 Nano Ex. 6 025008 View the article online for updates and enhancements.You may also likeNonlinear deformation of end-supportednanorods based on consistent couplestress and surface elasticity theoriesSitti Prasittikulwat, Jianjun Zhang, TawichPulngern et al.-Ordered arrays of metal nanostructures oninsulator/metal film: dependence ofplasmonic properties on lattice orientationEugene Bortchagovsky, Fang Dai, JiaTang et al.-Bioinspired nanocarriers for advanceddrug deliveryKalyani Pathak, Mohammad Zaki Ahmad,Jon Jyoti Sahariah et al.-This content was downloaded from IP address 144.213.253.16 on 04/01/2026 at 19:48https://doi.org/10.1088/2632-959X/addadb/article/10.1088/2632-959X/adea8b/article/10.1088/2632-959X/adea8b/article/10.1088/2632-959X/adea8b/article/10.1088/2632-959X/adde79/article/10.1088/2632-959X/adde79/article/10.1088/2632-959X/adde79/article/10.1088/2632-959X/adff9f/article/10.1088/2632-959X/adff9fNano Express 6 (2025) 025008 https://doi.org/10.1088/2632-959X/addadbPAPERPlasma treatment of MoS2 field-effect transistors using solid-statefluorine sourceHirokiKii1, NaokaNagamura2,3,4 andRyoNouchi1,51 Department of Physics and Electronics, Osaka PrefectureUniversity, Sakai 599-8570, Japan2 Research Center for Advanced Measurement and Characterization, National Institute for Materials Science, 3-13, Sakura, Tsukuba,Ibaraki 305-0003, Japan3 Faculty of Advanced Engineering, TokyoUniversity of Science, 6-3-1, Niijuku, Katsushika, Tokyo, 125-8585, Japan4 Research Institute of Electrical Communication, TohokuUniversity, 2-1-1Katahira, Aobaku, Sendai 980-8577, Japan5 Department of Physics and Electronics, OsakaMetropolitanUniversity, Sakai 599-8570, JapanE-mail: r-nouchi@omu.ac.jpKeywords: transition-metal dichalcogenide, surface passivation, hysteresis, water repellency, substitutional dopingSupplementarymaterial for this article is available onlineAbstractPlasma treatmentusing a solid-statefluorine (F) source canmitigate the emissionof F-containing gasesinto the environment. In this study,we investigated theprocessability ofAr-plasma-mediatedF treatmentusing apolytetrafluoroethylene (PTFE) sheet as the F source. Surface treatment of two-dimensional (2D)semiconductordevices using thismethod resulted in an improvement infield-effectmobility andareduction inhysteretic behavior. Theprolonged treatment led toheavyp-doping, possibly owing tosubstitutional Fdoping.The treatment strengthwas controllable by the treatment timeand samplepositionduring the treatment. Placing the samplesupstreamresulted in amilder treatment compared tothat positioneddownstream.The controllability of the proposedmethod enables us tofine-tune theproperties of devices basedon2Dmaterials. The treatment elements couldbe controlledusing sheetsmadeofmaterials other thanPTFE, indicating thebroad applicability of thismethod.1. IntroductionTwo-dimensional (2D) semiconductors are promising candidates for post-silicon electronics because of theiratomically thin nature, which enables the realization of ultrascaled field-effect transistors (FETs) [1]. The entirebodies of such ultrathinmaterials can be altered bymodifying their surfaces. Plasma exposure is widely used tocontrol the electronic properties of 2D semiconductors [2]. Oxygen-containing plasma improves photodetectorperformance [3, 4] andfield-effectmobilities in FETs [5–7] via surface oxide formation [8] and defectpassivation [9]. Argon- and nitrogen-containing plasmas have been used to improve the electrocatalytic activity[10–13] and supercapacitor performance [14] through defect formation and nitrogen doping [15].Another important class of plasma gases in the semiconductor industry is those containing fluorine (F).F-containing plasmas have also been used to etch 2Dmaterials [16–18], similar to their application in etchingsilicon-basedmaterials. In addition to its use in etching processes, F-containing plasma has been confirmed toapply to surfacemodifications such asfluorination [18–20] and F-atomdoping [21]. However, F-containinggases generally exhibit significantly higher global warming potential (GWP) values than that of CO2. Forexample, GWPvalues for a 100-year time horizonwere reported to be 7380, 9290, 10200, 12400, 14600, 17400,and 24300 for CF4, C3F8, cyclic C4F8, C2F6, CHF3,NF3, and SF6, respectively (1 for CO2) [22]. Therefore, theemission of F-containing gases into the environment should beminimized.However, the degree of ionization inlow-pressure plasma, commonly employed for surfacemodification and etching processes, is typically less thana fewpercent [23], andmost of the inlet gas is exhausted from the process chamber as it is unionized. Althoughdownstream abatement equipment has been developed in the semiconductor industry, the net destructionefficiency of these systems cannot achieve 100%owing to the byproducts generated during the treatment ofOPEN ACCESSRECEIVED31December 2024REVISED8May 2025ACCEPTED FOR PUBLICATION20May 2025PUBLISHED28May 2025Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 4.0licence.Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.© 2025TheAuthor(s). Published by IOPPublishing Ltdhttps://doi.org/10.1088/2632-959X/addadbhttps://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7232-4827https://orcid.org/0000-0002-7232-4827mailto:r-nouchi@omu.ac.jphttps://doi.org/10.1088/2632-959X/addadbhttps://crossmark.crossref.org/dialog/?doi=10.1088/2632-959X/addadb&domain=pdf&date_stamp=2025-05-28https://crossmark.crossref.org/dialog/?doi=10.1088/2632-959X/addadb&domain=pdf&date_stamp=2025-05-28https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/F-containing gases [24]. In addition, destroying stable F-containing species such as CF4 requires high-temperature processing, resulting in high energy consumption [25, 26].In this study, a solid-state F sourcewas employed for the surfacemodification ofmolybdenumdisulfide(MoS2), a representative 2D semiconductor. For a solid source, the unionized species remain inside the solid,which is expected tomitigate exhaust gas problems. An x-ray photoelectron spectroscopy study on argon plasmatreatment of polytetrafluoroethylene (PTFE) [27] showed that the F/C ratio of the PTFE surface reduced from2.0 to 1.4 by the treatment. This defluorination indicates the generation of reactive F species by breakingC—Fbonds [28, 29]. In this study, we performed the F treatment ofMoS2 FETs using F species produced by the argonplasma treatment of a PTFE sheet. The strength of the F treatment can be controlled by the treatment time andposition of the sample relative to that of the PTFE sheet. Although harsh treatment by placing the sampledownstream resulted in heavy p-doping, as expected from the high electronegativity of F,milder treatment at theupstreamposition significantly improved thefield-effectmobility. This study provides a facilemethod fortunable surfacemodification using F species by changing their position relative to the PTFE sheet.2. ExperimentalHighly doped Si wafers with a 285 nm thick thermal oxide layer were used as substrates for device fabrication.The substrate surfacewas cleaned by ultrasonicationwith acetone and 2-propanol and treatedwith an oxygenplasma cleaner (PDC-32G,Harrick Plasma). TheMoS2 flakeswere transferred to a cleaned substrate bymechanical exfoliation of a natural crystal (SPI Supplies). Specifically, thickflakes were exfoliated from thesourceMoS2 crystal using an adhesive tape. Theflakes were thinned by repeated exfoliation using another tape.The thinnedflakes on the tapewere transferred to a dimethylpolysiloxane (PDMS) sheet (Gel-Pak PF-20/17-X4) by attaching the tape surface to the sheet and pressing it with the fingers. The PDMS sheet with thin flakeswas thenfixed on a glass slide. Theflakes on the PDMS sheet were finally transferred to the SiO2/Si substrate byattaching the surface of the PDMS sheet to the substrate and pressing it with the fingers. Au electrodes with Cradhesion layers were fabricated on the transferred flakes by electron-beam lithography. Figures 1(a) and (b)show a schematic and opticalmicrograph of the fabricated FETs, respectively.Figure 1. (a) Schematic of the FETs fabricated in this study. ID,VD, andVG denote the drain current, drain voltage, and gate voltage,respectively. (b)Opticalmicrograph of a typical sample. (c) Setup for Ar-plasma-mediated F treatment.2Nano Express 6 (2025) 025008 HKii et alThe electrical characteristics of the FETsweremeasured using a semiconductor device analyzer (B1500A,Keysight) under ambient air in the dark. The FETs underwent Ar-plasma-mediated F treatment using aPTFE sheet. As shown infigure 1(c), the sample was placed downstreamor upstream at a distance of 1.0 cmfroma 1-mm-thick PTFE sheet (TOMBO™No. 9000,Nichias, used after being cut into the dimensions of4.5 cm× 2.0 cm) placed in the quartz tube of a plasma apparatus (PDC-32G,Harrick Plasma, operatedwith the rf power of 18W). The typical base pressure of the systemwas 15 Pa. Arwas introducedwhen thepressure reached approximately 40 Pa.The surfacemorphology of theMoS2 FETswas acquired using an atomic forcemicroscope (AFM;AFM5200S,HitachiHigh-Tech) in the dynamic forcemode. Chemical bonding states ofMoS2flakes wereinvestigated by x-ray photoemission spectroscopy (XPS) using a scanning photoelectronmicroscopy (SPEM)equipped at BL07LSUof SPring-8 [30, 31]. The excitation photon energywas set to 1000 eV. The incident x-rayswere focused down to∼100 nm. The energy resolutionwas set at∼200meV. Themeasurements wereperformed at room temperature. TheAu 4f7/2 core-level peak at 84.0 eV from the gold filmdeposited close totheMoS2flakeswas used for the binding energy scale calibrations.3. Results and discussionFigure 2(a) shows the transfer characteristics of theMoS2 FETs before and after Ar-plasma-mediated Ftreatment. The treatment was repeated thrice for 5, 5, and 10 min (20 min in total). The transfer characteristicsweremeasured before and after each treatment period. Field-effectmobilities (μ) and threshold voltages (Vth)were extracted using the following equation,( )m= -IWLC V V VD ox G th DwhereW and L are the channel width and length, respectively;Cox is the gate capacitance per unit area and equals12.1 nF cm−2 for the 285-nm-thick SiO2 layer. Least-squares fitting of the transfer curves within the 20-V rangearound the gate voltage corresponding to themaximum transconductance (dID/dVG) yielded the values showninfigure 2(b). Theμ values are limited by the electrode contacts as explained below. The thicknesses of the flakesused in this studywere greater than 10 nm (see the supplementarymaterial). The source and drain electrodeswere fabricated as the top contacts. Thus, charge carriers were injectedmainly from the top surface of the thickMoS2 flakes. On the other hand, the gate stackwas fabricated as a global back gate. Thus, transistor channelswere formed at the bottom surface of the flakes. Therefore, the devices in the present study suffered from a largeFigure 2.Effect of repeated treatments on the transfer characteristics ofMoS2 FETs placed downstream. The drain voltage,VD,was setto 0.7V. (a)Transfer characteristicsmeasured before the series of treatments and after each treatment period. The arrows indicate thesweep direction of the gate voltage. Changes in (b)field-effectmobility, threshold voltage, and (c) hysteresis window. The thresholdvoltage difference between the forward and backward sweeps of the gate voltage defines the hysteresis window. The absolute humidityof themeasurement environment is also shown in (c). See Figures S1(a) and S1(b) in the supplementarymaterial for semi-logarithmictransfer characteristics and output characteristics, respectively.3Nano Express 6 (2025) 025008 HKii et alaccess resistance from the top to the bottom surface of theflakes. In addition, the contactmetal used in thisstudy, Au (with an adhesion layer of Cr), is not a good electron injector, considering its work function.As the total treatment timewas prolonged,μ improved, as evidenced by the increased current. In addition,hysteretic behavior was significantly suppressed. As shown infigure 2(c), the hysteresis window, defined as thedifference in the threshold voltage between the forward and backward sweeps, was reduced from71 to 25V aftera series of treatment processes. The hysteretic behavior ofMoS2 FETs is attributed to the presence of water[32–34]. A smaller hysteresis with a longer treatment time indicates fewerwatermolecules near the FET channel.The hysteresis window shows a stronger correlationwith the treatment time thanwith the absolute humidity,which suggests that the plasma treatment has a stronger impact on the device characteristics. The possibleacquisition of thewater-repellent property is likely attributable to F passivation of theMoS2flakes by thetreatment.Watermolecules possess an electric dipole that induces potential fluctuations and leads to charge-carrier scattering. Therefore, the improvedmobility can be attributed to the possible acquisition of water-repellent properties of F treatment.The effect of the F treatment on the surfacemorphology was investigated byAFM. Figure 3(a) shows a seriesof AFM images acquired before starting the treatment and after each treatment. Before starting the treatment,the surfaces of theMoS2flakes were covered by particles, which could have originated from the exfoliation andlithography processes. The particles were removedmainly after the initial treatment. As shown infigure 3(b),this was evidenced by an initial decrease in the root-mean-square roughness of theMoS2flake surface. It is wellknown that surface scattering of carriers is significant for carrier transport in thin films [35, 36]. The presence ofsurface adsorbates transforms the specular reflectance of carriers at the film surface into diffusive scattering,increasing the electrical resistance. Such adsorbates include gasmolecules [37], chemisorbed species [38, 39],and largemolecules [40]. Process residues on theMoS2 surface can also act as scattering centers. Therefore, theimprovedmobility due to F treatment is partially attributable to removing process residues.However, plasma treatments for 5 minwithout the PTFE sheet were found to degrade the electricalproperties (see figure S2 in the supplementarymaterial). The devicewas placed in a position surrounded by an rfcoil. Therefore, the device was subjected to direct exposure to high-energy plasma species, which damaged thedevice. On the other hand, figure 2(a) shows that plasma treatmentwith a PTFE sheet for 5 min enhancedelectrical properties. Resist residues should also be removed by treatmentwithout PTFE sheets. Therefore,mechanisms other than resist-residue removal should be the leading cause of the improvement by treatmentwith PTFE sheets. One possiblemechanism is the passivation effect of thefluorocarbon species originating fromthe PTFE sheet. Fluorocarbonfilms are known to be deposited byCF4-gas plasma treatments [41]. A similarphenomenonwas expected in the case of the solid-state F source. As shown infigure 2(c), an increase inflakethickness after repeated treatments was confirmed byAFMobservations. After forming a thinfilm of CF-relatedspecies, the CFx layer covering theMoS2 surfacemay have prevented the direct impingement of high-energyplasma species. In addition, CFx deposition can change thewater repellency and dielectric environment ofMoS2flakes. The formermitigates Coulomb scattering of charge carriers as explained above. For the latter, an increasein dielectric constant is known to reduce charged-impurity scattering of charge carriers in the channel, leading toan increase inmobility [42, 43]. The dielectric constant of PTFE has been reported to be around 2.0 [44]. Thealteration of the dielectric on the top surface of theMoS2 layer from air to a CFxfilm increases the dielectricconstant, thus increasing themobility.To confirm the introduction of F species by plasma treatment ofMoS2flakes, pinpointmeasurements wereperformed using synchrotron radiationmicro-XPSwith high spatial resolution, which allows selective analysisofmicroscaleMoS2flakes. OneMoS2flakewas processed by plasma treatment for a short period (30 s at theFigure 3.Effect of repeated treatments on the surfacemorphology ofMoS2 FETs placed downstream. (a)AFM images before the seriesof treatments and after each treatment. (b)Rootmean square roughness of area indicated by thewhite rectangle in (a) andflakethickness.4Nano Express 6 (2025) 025008 HKii et aldownstreamposition) to prevent the deposition of a thick fluorocarbon film. Figure 4 shows theC 1s core-levelspectra of plasma-treated and untreatedMoS2flakes. TheCpeaks observed in the untreated samplewereattributable to hydrocarbon contamination from air or residues from the adhesive tape used in the exfoliationprocess. As shown infigure 3, these C contaminationswere largely removed by plasma treatment. The plasmatreatment should have introduced fluorocarbon species originating from the PTFE source in addition tounremoved contamination. The plasma-treated sample showed an increase in the intensity of the high-energyside (∼286 eV) of theC 1smain peak, which is attributable to the introduction of C-CF. In addition, a peakcentered at∼293 eVwas discernible only in the plasma-treated sample, originating fromC-F2. The observeddifferences in theC 1s peak shape in the pinpoint XPS spectra confirm the introduction of F-related species intothe plasma-treated sample.Ar-plasma-mediated F treatmentwas conducted under amedium vacuum (∼40 Pa). Under theseconditions, the gas in the process chamber exhibits a viscous flow. The F species generated from the PTFE sheetwere transported downstreamby the residual gasflow. Therefore, the sample position during treatment isexpected to affect the strength of the treatment. Figures 5(a)–(d) and (e) show the changes in the transfercharacteristics with treatment at the downstream and upstreampositions, respectively. It should be noted thatthe current level of the device shown infigure 5(c) is higher than that of the other devices. This can be explainedby the thinner flake used in the device (seefigure S4 in the supplementarymaterial), which results in the lowestaccess resistance between the top electrode contact and the bottom channel. As shown infigures 5(a)–(d), thecharacteristics after treatment at the downstreamposition gradually changed fromn- to p-type conduction asthe treatment time increased. The appearance of a p-type branch and positive shift in the threshold voltage of then-type branch (see figures 5(b) and (c)) are clear signs of substitutional p-doping [21]. The strong electron-withdrawing ability offluorine, due to its high electronegativity, leads to hole doping. Because theelectronegativity of carbon is nearly the same as that of sulfur, substitutional p-doping by carbon ismuchweakerthan that by fluorine [45]. The carbon-to-fluorine ratio of the PTFE sheet was 1:2. Therefore, the carrier-dopingeffect ismost likely governed by the incorporation of F atoms. The changes in carrier concentration resultingfrom the plasma treatment were calculated based on the difference in the threshold voltage before and aftertreatment. The results of these calculations are presented in table 1. The carrier concentration resulting from theaddition of F as a dopantwas higher than 1013 cm−2 in the case with the highest dopant concentration. Althoughthe total treatment time infigure 2was the same as that shown infigure 5(c), the difference in the resultanttransfer characteristics suggests the presence of an incubation time for substitutional doping, similar towhatwasobserved in the etching ofMoS2 byCF4 plasma [16]. If the treatment period is insufficient for doping, plasma-induced damagemay be recovered during air exposure during electrical characterization. For the treatment atthe upstreamposition, the change in the transfer characteristics was similar to that shown infigure 2. This resultFigure 4.Micro-XPS analysis of plasma-treated and untreatedMoS2flakes. (a)Opticalmicroscopy and (b) spectral intensitymappingofMo3d obtained by SPEM for a plasma-treatedMoS2flake. (c)Opticalmicroscopy and (d) spectral intensitymapping ofMo3dobtained by SPEM for an untreatedMoS2flake. (e)Pinpoint C 1s core-level spectra at positions indicated in (b) and (d). (f)Results ofpeak deconvolution of theC 1s spectrum shown in the red curve of (e)measured on a plasma-treatedMoS2flake. (g)Results of peakdeconvolution of theC 1s spectrum shown in the blue curve of (e)measured on an untreatedMoS2flake. See figure S3 in thesupplementarymaterial forMo 3d and S 2s spectra.5Nano Express 6 (2025) 025008 HKii et alindicates that the strength of the F treatment was reduced by positioning the sample upstream. Therefore, Ar-plasma-mediated F treatment can be controlled by controlling the position of the sample during treatment.Thinner flakes should be affectedmore severely than thicker ones; for example, the ratio of a portion dopedwith F atoms to that remaining undoped is expected to be higher in thinner flakes. The sample positionedupstream (thickness: 23.2 nm; see figure S4 in the supplementarymaterial) showed an increase in drain currentafter 20-min of treatment. The drain current increasedwith the gate voltage, indicating n-channel conduction.At the downstreamposition, another sample with a similar thickness (23.3 nm)was treated for 15 min. Evenwith a shorter treatment time, the treatment resulted in a reduction in n-channel conduction and theappearance of p-channel conduction. This is a clear indication of significant F doping. These results indicate thatthe sample position during treatment is amore dominant factor than flake thickness.PTFE sheets are aggregates of TFE-basedmolecular chains. Therefore, plasma-induced damage to the PTFEsurface should be considered at themolecular scale. A single chain can be easily removed from the surface byexposure to the plasma. After removing the chain, another fresh chainwill soon appear on the underlyingsurface. Looking at a single site, the surface condition changes repeatedly. The stage of chain removal differsfromposition to position on the PTFE surface. Therefore, the surface condition can be regarded as constant onaverage during plasma exposure. As shown infigures 5(a)–(d), the present process possesses a certaincontrollability of the doping level by changing the treatment time. Themonotonic dependence on the treatmenttimemight support the constancy of the PTFE surface conditions.The transfer characteristics shown in Figures S1 and S3 in the supplementarymaterial show that the gateleakage current tends to increase after exposure to the fluorine plasma. The slight increase in the gate leakagemay be attributed to the carbon components that are deposited in theCFx layers during the plasma treatment.These CFx films are deposited onto the entire substrate surface because the plasma treatment was conductedwithout a shadowmask. Thefinite conductivity of the amorphous carbon (a-C) parts embedded in theCFxfilmsmay provide a gate leakage pathwhen the a-C reaches the substrate edges and bypasses the gate insulator. Thegate leakage problem is expected to bemitigated by the co-introduction of oxygen gas during the plasmatreatment, whichwould serve to burn out the residual carbon by oxidation. In addition, a certain number ofcarbon atoms are expected to be incorporated into theMoS2 lattice. The incorporation of carbon atomsmightFigure 5.Effect of treatment time and sample position during treatment on the transfer characteristics ofMoS2 FETs. The sampleswere positioned (a)–(d) downstream and (e) upstream. The treatment timewas indicated in eachfigure. The drain voltageVDwas setto 0.5V. The data points before and after treatment are displayed as open and solid circles, respectively. The arrows indicate the sweepdirection of the gate voltage. See figure S4 in the supplementarymaterial for semi-logarithmic transfer characteristics, outputcharacteristics, andAFM images.Table 1.Changes in the carrier concentration,Δn, calculated from the differencein the n-channel threshold voltage before and after the treatment,ΔVth, usingthe relationship /∆ ∆=n C V eox th , where e is the elementary charge.FigureNo. 5 (a) 5 (b) 5 (c) 5 (d) 5 (e)Vth (before) [V] −49.2 −53.6 −59.8 −79.8 −54.0Vth (after) [V] −27.3 −34.3 61.3 >80 −51.5ΔVth [V] 21.9 19.3 121.1 >159.8 2.5Δn [1012 cm−2] 1.7 1.5 9.2 >12.1 0.26Nano Express 6 (2025) 025008 HKii et alcause deteriorative effects, such as degradation of the carriermobility, which are also expected to bemitigated bythe co-introduction of oxygen gas during the plasma treatment.4. ConclusionTheAr-plasma-mediated F treatment ofMoS2 FETswas conducted using a PTFE sheet as the solid-state Fsource. In this treatment, the unionized F species remained inside the solid,mitigating the emission of high-GWPF-based gases into the environment. Repeating the short-term treatment gradually improved the field-effectmobilities and reduced the hysteretic behavior of the transfer characteristics. The latter effect is attributedto the possible acquisition of water-repellent properties through F passivation, which reduces the potentialfluctuation caused by the electric dipoles of thewatermolecules. Additionally, the treatment was confirmed toremove the fabrication process residues from theMoS2 surface. The reduced number of watermolecules andsurface contaminants results in a reduction of the concentration of charge-carrier scattering centers, which leadsto improvedmobility. The strength of the treatment was controlled by the treatment time and position of thesample during the treatment.While the placement of samples at upstreampositions led to a similar result with ashorter treatment time at downstreampositions, the samples placed downstreamwere found to be heavilyp-doped, possibly owing to substitutional F doping. This study elucidates the processability of F-based surfacetreatments using a solid-state F source. The device properties imparted by this treatment can be tuned byexploiting the controllability of thismethod. In principle, plasma-mediated treatment with a solid-state sourcecan be applied to elements other than F by simply changing the solid-sourcematerial, which allows for alteringsurface properties withoutmodifying the plasma gas species.AcknowledgmentsThis studywas supported by JSPSKAKENHIGrantNumbers JP19H02561, JP22H01912, JP23K23180,JP21H01638, JP21H04696, and JP24K01243; JST PRESTOGrantNumbers JPMJPR20T8, JPMJPR20T7, andJPMJPR17NB; and theMurata Science Foundation. The spectral datasets were obtainedwith the support of theUniversity of TokyoOutstation Beamline (BL07LSU) at SPring-8 (Proposal Numbers: 2021A7421, 2021A7422,2021B7433, 2021B7435, 2022A7444).We thank ShunKonno (TokyoUniversity of Science) andWenxiongZhang (TheUniversity of Tokyo) formeasurements in SPring-8.Data availability statementAll data that support thefindings of this study are includedwithin the article (and any supplementary files).ORCID iDsNaokaNagamura https://orcid.org/0000-0002-7697-8983RyoNouchi https://orcid.org/0000-0002-7232-4827References[1] Zeng S, LiuC andZhouP 2024Transistor engineering based on 2Dmaterials in the post-silicon eraNat. Rev. Electr. Eng. 1 335–48[2] Jadwiszczak J, KellyD J, Guo J, Zhou Y andZhangH2021 Plasma treatment of ultrathin layered semiconductors for electronic deviceapplicationsACSAppl. Electron.Mater. 3 1505–29[3] Jadwiszczak J, Li G, CullenCP,Wang J J,Maguire P, DuesbergG S, Lunney JG andZhangH2019 Photoresponsivity enhancement inmonolayerMoS2 by rapidO2:Ar plasma treatmentAppl. Phys. Lett. 114 091103[4] Wu J, Li S,WangX,HuangY,HuangY, ChenH,Chen J, She J andDeng S 2023 Plasma treatment for achieving oxygen substitution inlayeredMoS2 and the room-temperaturemid-infrared (10 μm) photoresponseACSAppl.Mater. Interfaces 15 58556–65[5] Dhall R, Li Z, Kosmowska E andCronin S B 2016Charge neutralMoS2field effect transistors through oxygen plasma treatment J. Appl.Phys. 120 195702[6] Jadwiszczak J et al 2018Oxide-mediated recovery offield-effectmobility in plasma-treatedMoS2 Sci. Adv. 4 eaao5031[7] Li Z, Liu L andXu J-P 2021 Largely enhancedmobility ofMoS2field-effect transistors by optimizingO2-plasma treatment onMoS2IEEETrans. ElectronDevices 68 4614–7[8] KangN, PaudelHP, LeuenbergerMN,Tetard L andKhondaker S I 2014 Photoluminescence quenching in single-layerMoS2 viaoxygen plasma treatment J. Phys. Chem. C 118 21258–63[9] NanH,WuZ, Jiang J, Zafar A, YouY andNi Z 2017 Improving the electrical performance ofMoS2 bymild oxygen plasma treatmentJ. Phys. D: Appl. Phys. 50 154001[10] NguyenAD et al 2019Nitrogen-plasma-treated continuousmonolayerMoS2 for improving hydrogen evolution reactionACSOmega4 21509–157Nano Express 6 (2025) 025008 HKii et alhttps://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7697-8983https://orcid.org/0000-0002-7232-4827https://orcid.org/0000-0002-7232-4827https://orcid.org/0000-0002-7232-4827https://orcid.org/0000-0002-7232-4827https://doi.org/10.1038/s44287-024-00045-6https://doi.org/10.1038/s44287-024-00045-6https://doi.org/10.1038/s44287-024-00045-6https://doi.org/10.1021/acsaelm.0c00901https://doi.org/10.1021/acsaelm.0c00901https://doi.org/10.1021/acsaelm.0c00901https://doi.org/10.1063/1.5086726https://doi.org/10.1021/acsami.3c11962https://doi.org/10.1021/acsami.3c11962https://doi.org/10.1021/acsami.3c11962https://doi.org/10.1063/1.4967398https://doi.org/10.1126/sciadv.aao5031https://doi.org/10.1109/TED.2021.3089562https://doi.org/10.1109/TED.2021.3089562https://doi.org/10.1109/TED.2021.3089562https://doi.org/10.1021/jp506964mhttps://doi.org/10.1021/jp506964mhttps://doi.org/10.1021/jp506964mhttps://doi.org/10.1088/1361-6463/aa5c6ahttps://doi.org/10.1021/acsomega.9b03205https://doi.org/10.1021/acsomega.9b03205https://doi.org/10.1021/acsomega.9b03205[11] Wang J, FengX,HedmanD,WuX, PanH andZhangQ 2022 Enhancing the hydrogen evolution reaction onMoS2flakes by coldplasma treatment Electrochem. Commun. 137 107250[12] HuangY-S, Liu Y-T, Perng T-P, LuM-Y, Chueh Y-L andChen L-J 2023 Enhancing photocatalytic properties of continuous few-layerMoS2 thinfilms for hydrogen production bywater splitting through defect engineering withAr plasma treatmentNano Energy 109108295[13] Li Y,WanY, Yao J, ZhengH,WangX, LiuX,Ouyang B,HuangC,DengK andKanE 2023Enhanced electrocatalytic hydrogenevolution fromnitrogen plasma-tailoredMoS2 nanostructures Phys. Chem.Chem. Phys. 25 31628–35[14] Rui S, Li Z,Meng L,WangQ, Xu J, Zhao Y, JiaQ, LiH, Lu S andZhang Y 2025Citric acid and plasma treatedMoS2 for high-performance supercapacitors J.Mater. Chem. C 13 849–57[15] Azcatl A et al 2016Covalent nitrogen doping and compressive strain inMoS2 by remoteN2 plasma exposureNano Lett. 16 5437–43[16] JeonMH,AhnC,KimH,KimKN, LiNTZ,QinH,KimY, Lee S, KimT andYeomGY2015ControlledMoS2 layer etching usingCF4plasmaNanotechnology 26 355706[17] Ghasemi F, Abdollahi A andMohajerzadeh S 2019Controlled plasma thinning of bulkMoS2flakes for photodetector fabricationACS.Omega 4 19693–704[18] Shin Y et al 2024 Facilitated fluorination and etching of 2DmaterialsAppl. Surf. Sci. 645 158857[19] Bon S B, Valentini L, Verdejo R,Garcia Fierro J L, Peponi L, Lopez-ManchadoMAandKenny JM2009Plasma fluorination ofchemically derived graphene sheets and subsequentmodificationwith butylamineChem.Mater. 21 3433–8[20] ZhangH, Fan L,DongH, Zhang P,NieK, Zhong J, Li Y, Guo J and SunX2016 Spectroscopic investigation of plasma-fluorinatedmonolayer graphene and application for gas sensingACSAppl.Mater. Interfaces 8 8652–61[21] Chee S-S, JangH, Lee K andHamM-H2020 Substitutional fluorine doping of large-areamolybdenumdisulfidemonolayerfilms forflexible inverter device arraysACSAppl.Mater. Interfaces 12 31804–9[22] Climate Change 2021: The Physical Science Basis 2021Table 7.SM.7 SixthAssessment Report of the Intergovernmental Panel onClimate Change (https://www.ipcc.ch/report/ar6/wg1/)[23] Suzuki K,Okudaira S, SakudoNandKanomata I 1977Microwave plasma etching Jpn. J. Appl. Phys. 16 1979[24] KoDG,Ko S J, Choi EK,Min SG,Oh SH, Jung J, KimBMand Im I-T 2014 Perfluorocarbon destruction and removal efficiency:considering the byproducts and energy consumption of an abatement system formicroelectronicsmanufacturing IEEE Trans.Semicond.Manuf. 27 456–61[25] Iacopi F andMcIntoshM2019Opportunities and perspectives for green chemistry in semiconductor technologiesGreenChem. 213250–5[26] ChangMB andChang J-S 2006Abatement of PFCs from semiconductormanufacturing processes by nonthermal plasmatechnologies: a critical review Ind. Eng. Chem. Res. 45 4101–9[27] KimSR 2000 Surfacemodification of poly(tetrafluoroethylene)film by chemical etching, plasma, and ion beam treatments J. Appl.Polym. Sci. 77 1913–20[28] Carbone EAD,VerhoevenMWGM,KeuningW,Mullen J J AMand van der 2016 PTFE treatment by remote atmospheric Ar/O2plasmas: a simple reaction schememodel proposal J. Phys. Conf. Ser. 715 012011[29] PerryCC,Wagner A J andHoward FairbrotherD 2002 Electron stimulatedC–F bond breaking kinetics influorine-containing organicthinfilmsChem. Phys. 280 111–8[30] HoribaK,Nakamura Y,NagamuraN, Toyoda S, KumigashiraH,OshimaM,AmemiyaK, Senba Y andOhashiH 2011 Scanningphotoelectronmicroscope for nanoscale three-dimensional spatial-resolved electron spectroscopy for chemical analysisRev. Sci.Instrum. 82 113701[31] NagamuraN,Konno S,MatsumotoM, ZhangW,KotsugiM,OshimaMandNouchi R 2023 Photoelectron spectromicroscopyanalysis of graphene during gate-controlled photo-oxidation processNano Express 3 044003[32] KimW, Javey A,VermeshO,WangQ, Li Y andDaiH 2003Hysteresis caused bywatermolecules in carbon nanotube field-effecttransistorsNano Lett. 3 193–8[33] WangH,WuY, CongC, Shang J andYuT2010Hysteresis of electronic transport in graphene transistorsACS.Nano 4 7221–8[34] LateD J, Liu B,MatteH S SR,Dravid VP andRaoCNR2012Hysteresis in single-layerMoS2field effect transistorsACS.Nano 65635–41[35] FuchsK 1938The conductivity of thinmetallic films according to the electron theory ofmetalsMath. Proc. Cambridge Philos. Soc. 34100–8[36] Sondheimer EH1952Themean free path of electrons inmetalsAdv. Phys. 1 1–42[37] WissmannP 1972The effect of gas adsorption on the conductivity of thinmetal filmsThin Solid Films 13 189–93[38] Yang F,DonavanKC, Kung S-C and Penner RM2012The surface scattering-based detection of hydrogen in air using a platinumnanowireNano Lett. 12 2924–30[39] DingM,HeQ,WangG, ChengH-C,HuangY andDuanX2015An on-chip electrical transport spectroscopy approach for in situmonitoring electrochemical interfacesNat. Commun. 6 7867[40] Nouchi R andKanno I 2005Charge transfer and formation of conducting C60monolayers at C60/noble-metal interfaces J. Appl. Phys.97 103716[41] JinH, Fan F, YuanZ and Li Y 2020 Investigation of the formationmechanism of thefluorocarbon film inCF4 plasma processing offused silicaOptik 202 163693[42] JangC, AdamS, Chen J-H,Williams ED,Das Sarma S and FuhrerMS 2008Tuning the effective fine structure constant in graphene:opposing effects of dielectric screening on short- and long-range potential scattering Phys. Rev. Lett. 101 146805[43] Radisavljevic B andKis A 2013Mobility engineering and ametal–insulator transition inmonolayerMoS2Nat.Mater. 12 815–20[44] Wang L, LiuC, Shen S, XuMand LiuX 2020 Lowdielectric constant polymers for high speed communication networkAdv. Ind. Eng.Polym. Res. 3 138–48[45] YueQ,Chang S,Qin S and Li J 2013 Functionalization ofmonolayerMoS2 by substitutional doping: afirst-principles studyPhys. Lett. A377 1362–78Nano Express 6 (2025) 025008 HKii et alhttps://doi.org/10.1016/j.elecom.2022.107250https://doi.org/10.1016/j.nanoen.2023.108295https://doi.org/10.1016/j.nanoen.2023.108295https://doi.org/10.1039/D3CP04951Ehttps://doi.org/10.1039/D3CP04951Ehttps://doi.org/10.1039/D3CP04951Ehttps://doi.org/10.1039/D4TC04432Khttps://doi.org/10.1039/D4TC04432Khttps://doi.org/10.1039/D4TC04432Khttps://doi.org/10.1021/acs.nanolett.6b01853https://doi.org/10.1021/acs.nanolett.6b01853https://doi.org/10.1021/acs.nanolett.6b01853https://doi.org/10.1088/0957-4484/26/35/355706https://doi.org/10.1021/acsomega.9b02367https://doi.org/10.1021/acsomega.9b02367https://doi.org/10.1021/acsomega.9b02367https://doi.org/10.1016/j.apsusc.2023.158857https://doi.org/10.1021/cm901039jhttps://doi.org/10.1021/cm901039jhttps://doi.org/10.1021/cm901039jhttps://doi.org/10.1021/acsami.5b11872https://doi.org/10.1021/acsami.5b11872https://doi.org/10.1021/acsami.5b11872https://doi.org/10.1021/acsami.0c07824https://doi.org/10.1021/acsami.0c07824https://doi.org/10.1021/acsami.0c07824https://www.ipcc.ch/report/ar6/wg1/https://www.ipcc.ch/report/ar6/wg1/https://www.ipcc.ch/report/ar6/wg1/https://doi.org/10.1143/JJAP.16.1979https://doi.org/10.1109/TSM.2014.2362942https://doi.org/10.1109/TSM.2014.2362942https://doi.org/10.1109/TSM.2014.2362942https://doi.org/10.1039/C9GC01058Khttps://doi.org/10.1039/C9GC01058Khttps://doi.org/10.1039/C9GC01058Khttps://doi.org/10.1039/C9GC01058Khttps://doi.org/10.1021/ie051227bhttps://doi.org/10.1021/ie051227bhttps://doi.org/10.1021/ie051227bhttps://doi.org/10.1002/1097-4628(20000829)77:9<1913::AID-APP7>3.0.CO;2-#https://doi.org/10.1002/1097-4628(20000829)77:9<1913::AID-APP7>3.0.CO;2-#https://doi.org/10.1002/1097-4628(20000829)77:9<1913::AID-APP7>3.0.CO;2-#https://doi.org/10.1088/1742-6596/715/1/012011https://doi.org/10.1016/S0301-0104(02)00561-Xhttps://doi.org/10.1016/S0301-0104(02)00561-Xhttps://doi.org/10.1016/S0301-0104(02)00561-Xhttps://doi.org/10.1063/1.3657156https://doi.org/10.1088/2632-959X/aca0afhttps://doi.org/10.1021/nl0259232https://doi.org/10.1021/nl0259232https://doi.org/10.1021/nl0259232https://doi.org/10.1021/nn101950nhttps://doi.org/10.1021/nn101950nhttps://doi.org/10.1021/nn101950nhttps://doi.org/10.1021/nn301572chttps://doi.org/10.1021/nn301572chttps://doi.org/10.1021/nn301572chttps://doi.org/10.1021/nn301572chttps://doi.org/10.1017/S0305004100019952https://doi.org/10.1017/S0305004100019952https://doi.org/10.1017/S0305004100019952https://doi.org/10.1017/S0305004100019952https://doi.org/10.1080/00018735200101151https://doi.org/10.1080/00018735200101151https://doi.org/10.1080/00018735200101151https://doi.org/10.1016/0040-6090(72)90173-3https://doi.org/10.1016/0040-6090(72)90173-3https://doi.org/10.1016/0040-6090(72)90173-3https://doi.org/10.1021/nl300602mhttps://doi.org/10.1021/nl300602mhttps://doi.org/10.1021/nl300602mhttps://doi.org/10.1038/ncomms8867https://doi.org/10.1063/1.1897840https://doi.org/10.1016/j.ijleo.2019.163693https://doi.org/10.1103/PhysRevLett.101.146805https://doi.org/10.1038/nmat3687https://doi.org/10.1038/nmat3687https://doi.org/10.1038/nmat3687https://doi.org/10.1016/j.aiepr.2020.10.001https://doi.org/10.1016/j.aiepr.2020.10.001https://doi.org/10.1016/j.aiepr.2020.10.001https://doi.org/10.1016/j.physleta.2013.03.034https://doi.org/10.1016/j.physleta.2013.03.034https://doi.org/10.1016/j.physleta.2013.03.034 1. Introduction 2. Experimental 3. Results and discussion 4. Conclusion Acknowledgments Data availability statement References