High carrier mobility in graphene doped using a monolayer of tungsten oxyselenide

Doped graphene could be of use in next-generation electronic and photonic devices. However, chemical doping cannot be precisely controlled in the material and leads to external disorder that diminishes carrier mobility and conductivity. Here we show that graphene can be efficiently doped using a monolayer of tungsten oxyselenide (TOS) that is created by oxidizing a monolayer of tungsten diselenide. When the TOS monolayer is in direct contact with graphene, a room-temperature mobility of 2,000 cm2 V−1 s−1 at a hole density of 3 × 1013 cm−2 is achieved. Hole density and mobility can also be controlled by inserting tungsten diselenide interlayers between TOS and graphene, where increasing the layers reduces the disorder. With four layers, a mobility value of around 24,000 cm2 V−1 s−1 is observed, approaching the limit set by acoustic phonon scattering, resulting in a sheet resistance below 50 Ω sq−1. To illustrate the potential of our approach, we show that TOS-doped graphene can be used as a transparent conductor in a near-infrared (1,550 nm) silicon nitride photonic waveguide and ring resonator. A monolayer of tungsten oxyselenide, created by oxidizing a layer of tungsten diselenide, can be used to efficiently dope graphene, leading to a room-temperature mobility of 2,000 cm2 V–1 s–1 at a hole density of 3 × 1013 cm–2.

D oping techniques that control the carrier density of a material over a wide range are crucial for fabricating advanced electronic and photonic devices [1][2][3] . For graphene, doping is used to achieve high conductivity and tunable work functions, which are necessary for use in transparent electrodes and near-infrared (IR) and mid-IR photonics 4,5 . However, conventional doping techniques, such as ion implantation, are ineffective for graphene and other two-dimensional (2D) materials 6 . Instead, surface charge transfer doping is typically used, and hole doping with nitrogen dioxide can, for example, achieve a hole density of <2 × 10 12 cm −2 and sheet resistance of around 300 Ω sq −1 (ref. 7 ). Wet chemical doping using a variety of inorganic and organic dopants has also been explored, delivering carrier densities of >5 × 10 12 cm −2 and sheet resistances between 150 and 200 Ω sq −1 (refs. [8][9][10] ). However, chemical doping can suffer from poor long-term stability and large device-to-device variability 10,11 , and it introduces considerable charge disorder that limits the carrier mobility to below 1,000 cm 2 V −1 s −1 . Solid-state dopants have the advantages of high repeatability, compatibility with complementary metal-oxide-semiconductor processes and long-term stability; non-stoichiometric insulators, such as SiN x and AlO x (refs. 12,13 ), have been used to optimize charge injection in 2D semiconductors. However, the doping approach has not been widely explored with graphene.
In this Article, we report low-disorder, tunable high-density doping of graphene using a monolayer of the solid-state oxide of tungsten oxyselenide (TOS), which is created via the room-temperature (RT), ultraviolet (UV)-ozone oxidation of tungsten diselenide (WSe 2 ). The high-work-function TOS induces a mismatch-induced charge transfer with the nearby graphene layer. The oxidation process, which has been previously shown to be self-limiting 14 , allows layer-by-layer control and low disorder compared with other oxidation techniques; further, by starting from a crystalline layered source, charge disorder is reduced 15 . We show that the TOS layer induces a hole density of above 3 × 10 13 cm −2 in a neighbouring graphene layer, which has a high mobility of 2,000 cm 2 V −1 s −1 and sheet resistance of 118 Ω sq −1 . When the TOS is separated from graphene by WSe 2 , the mobility rises further, similar to modulation doping in III-V semiconductor heterostructures, where the separation reduces impurity scattering 16 , showing phonon-limited mobility and a reduced sheet resistance of 48 Ω sq −1 . Doped graphene is also highly transparent in the IR, and we show that near-ideal transmittance (99.2%) and low insertion loss (0.012 dB μm −1 ) at IR frequencies is possible using the TOS-doped graphene in a silicon nitride (SiN) photonic waveguide and ring resonator. Figure 1a shows the process flow of fabricating TOS-doped monolayer graphene. The device is fabricated by first stacking monolayer WSe 2 (1L-WSe 2 ), graphene and bottom hexagonal boron nitride (hBN) using a polycaprolactone (PCL) polymer-based dry-transfer process 17 . The heterostructure is then etched into Hall bar structures for the accurate extraction of carrier densities and conductivities of graphene, with edge-contacted metal electrodes (Cr/Au) Doped graphene could be of use in next-generation electronic and photonic devices. However, chemical doping cannot be precisely controlled in the material and leads to external disorder that diminishes carrier mobility and conductivity. Here we show that graphene can be efficiently doped using a monolayer of tungsten oxyselenide (TOS) that is created by oxidizing a monolayer of tungsten diselenide. When the TOS monolayer is in direct contact with graphene, a room-temperature mobility of 2,000 cm 2 V −1 s −1 at a hole density of 3 × 10 13 cm −2 is achieved. Hole density and mobility can also be controlled by inserting tungsten diselenide interlayers between TOS and graphene, where increasing the layers reduces the disorder. With four layers, a mobility value of around 24,000 cm 2 V −1 s −1 is observed, approaching the limit set by acoustic phonon scattering, resulting in a sheet resistance below 50 Ω sq −1 . To illustrate the potential of our approach, we show that TOS-doped graphene can be used as a transparent conductor in a near-infrared (1,550 nm) silicon nitride photonic waveguide and ring resonator.

Electrical and Raman characterization of TOS-doped graphene
formed using standard lithography processes (Fig. 1b) 18 . Finally, the monolayer WSe 2 is oxidized into monolayer TOS at RT by exposing the sample to ozone under UV illumination for 30 min (Methods provides further device fabrication and UV-ozone oxidation details). In particular, the combination of UV and ozone is important as UV exposure creates local surface defects that facilitate the oxidation of the topmost WSe 2 layer on subsequent ozone exposure at RT. We note that this is in contrast to earlier studies utilizing only ozone without any UV exposure, which required an elevated temperature together with longer exposure time to completely oxidize the topmost WSe 2 layer 19 . We also note that another added benefit of our TOS-doping method is the self-cleaning nature of UV-ozone oxidation, resulting in the reduction in surface roughness and suppression of hysteresis arising from carrier (de-)trapping at the surface states formed by the polymer residue after oxidation (Supplementary Figs. 1-3) 20 .
We first determine the type of carrier induced in our TOS-doped monolayer graphene using Raman spectroscopy, as shown in Fig. 1c. We find a clear blueshift in both G and 2D peaks (18.1 and 4.3 cm −1 , respectively) to their original positions in pristine graphene 21 , which indicates that graphene becomes hole doped. This is corroborated by the I 2D /I G ratio getting reduced by threefold and full-width at half-maximum (FWHM) of the G peak reduced from 7.3 to 5.9 cm −1 after oxidation (Fig. 1c, inset). Four-probe resistance (R 4p ) of graphene as a function of the back-gate bias (V GS ) at RT also shows a large shift in the Dirac peak from V GS = 30 V to beyond the measurement range after the formation of the TOS layer (Fig. 1d). This, together with R 4p decreasing with the negative bias voltage, clearly indicates that an ultrahigh density of holes is induced in graphene. As a result, the RT sheet resistance (R sh ) of our TOS-doped graphene shows a low value of ~118 Ω sq −1 at zero gate-bias voltage. Note that the weak secondary peak shown in R sh can be attributed to small spatial inhomogeneity in the sample or formation of the moiré potential from an unintentional atomic alignment of graphene with the bottom hBN (ref. 22 ). We further note that any contribution of the TOS layer to the measured conductivity can be ruled out by independent electrical measurements, which confirm that it is insulating ( Supplementary Fig. 4).

Structural characterization of monolayer TOS
To gain further insight into the nature of the TOS layer, we investigate its structural properties using selected-area electron diffraction (SAED), as shown in Fig. 2. Figure 2b shows the SAED patterns of 1L-WSe 2 flake after oxidation, indicating the complete disappearance of hexagonal symmetry along the [0001] zone axis. This suggests that the resultant TOS layer is amorphous. However, few-layer WSe 2 shows hexagonal single-crystal diffraction patterns even after oxidation, as shown in Fig. 2c,d, confirming that the UV-ozone oxidation process presented in our work is self-limiting in nature and thus the underlying WSe 2 layers remain in the pristine form ( Supplementary Fig. 5 shows the transmission electron microscopy images of monolayer and few-layer WSe 2 ). The self-limited nature of our oxidation process is further corroborated by the energy-dispersive X-ray spectroscopy measurements of 1L-WSe 2 that show the presence of selenium atoms even after oxidation (Supplementary Table 1). We note that this self-limited nature allows for the repeated oxidization and etching of multilayer WSe 2 flake in a monolayer-by-monolayer fashion (Supplementary Fig.  1) 14 . On the atomistic level, this indicates that the TOS layer acts as a high diffusion barrier, preventing further penetration of ozone molecules into the underlying layers 23 . Figure 2e,f shows the X-ray photoelectron spectroscopy (XPS) spectra of the core levels of W 4f and Se 3d before and after the oxidation of chemical vapour deposition (CVD)-grown monolayer WSe 2 . We find a dominant formation of multivalent oxidation states of W (W 5+ and W 6+ ) after oxidation, which verifies that 1L-TOS is sub-stoichiometric. Note that weak W-O and Se-O signals present before oxidation are presumably attributed to intrinsic defects present in the CVD-grown samples. We further note that defects in the amorphous nature of the TOS layer can cause surface adsorption of water and oxygen molecules, leading to time-dependent degradation 24 . Supplementary Fig. 6 depicts the poly(methyl methacrylate) (PMMA) encapsulation layer as a potential solution to prevent this degradation. A slight initial decrease in the zero-gate bias hole density (p) of our TOS-doped graphene to 7 × 10 12 cm −2 immediately after PMMA encapsulation can be attributed to the chemical reaction with a solvent at a high baking temperature of 180 °C. However, p remains nominally unchanged thereafter (~14% decrease in p over one month), showing the enhanced stability of our TOS-doping method. Figure 3a shows the measured p of TOS-doped graphene extracted from Hall effect measurements ( Supplementary Fig. 7). We first focus on the case where the TOS is directly in contact with graphene. We find that the hole density is 3.2 × 10 13 cm −2 at V GS = 0 V, consistent with the estimate from the Raman measurements. Applying V GS can further tune p up to 3.7 × 10 13 cm −2 , demonstrating that the back-gate capacitance is electrostatically decoupled from the top TOS layer. This doping level is equivalent to ~1% of the graphene atomic density (3.82 × 10 15 cm −2 ), and it is similar to the maximum achievable value in silicon using substitutional doping 25 . The doping level is also beyond what can be achieved by electrostatic gating through solid dielectrics. For example, a graphite back-gated structure with an hBN dielectric can only achieve carrier densities of the order of ~6 × 10 12 cm −2 in graphene 26 , and a perfect dielectric with a high electrical-breakdown dielectric strength of 1 V nm −1 can only accumulate ~2 × 10 13 cm −2 .

Tunable carrier density and mobilities in TOS-doped graphene with WSe 2 interlayers
The self-limiting nature of the oxidation process provides a straightforward method to tune the induced hole density. When multilayer WSe 2 is utilized, the top layer is converted to TOS and the remaining layers remain pristine, increasing the separation between the dopant layer and the channel. As we vary the interlayer WSe 2 thickness from one to four layers, the induced p decreases monotonically to 0.4 × 10 13 cm −2 ( Fig. 3a and Supplementary Fig. 8 (1.5 K)), and remains additionally tunable by the back gate. The corresponding Fermi energy (E F ) can be extracted using the relation where E CNP is the energy of the charge neutrality point (Dirac point), ħ is the reduced Planck constant and v F is the Fermi velocity in graphene (10 6 m s −1 ). We find that E F can be tuned from −0.1 to −0.7 eV, by simply changing the WSe 2 layer number together with the back gate (Fig. 3b). We show the measured sheet resistance of graphene in Fig. 3c. At zero back-gate voltage, the TOS-graphene sample shows a R sh of 118 Ω sq −1 . For comparison, undoped graphene has R sh of the order of a few kΩ sq −1 and state-of-the-art chemically doped graphene has R sh of ~140 Ω sq −1 (ref. 4 ), illustrating the superiority of our doping method. Remarkably, for the three-interlayer and four-interlayer c,d, SAED patterns of few-layer WSe 2 before (c) and after (d) the UV-ozone oxidation process. Scale bars in a-d, 5 nm −1 . Single-crystal diffraction patterns with the [0001] zone axis are visible before UV-ozone oxidation and are completely removed after the UV-ozone oxidation process, indicating the amorphous nature of the TOS layer. In contrast, few-layer WSe 2 still shows crystalline patterns even after oxidation, indicating a self-limiting process. e,f, XPS characterization of 1L-WSe 2 (e) and 1L-TOS (f) indicating the W 4f and Se 3d peaks, respectively. A comparison of W 4f peaks shows the creation of multivalent oxidation states of W (W 5+ and W 6+ ) in the TOS layer, confirming the sub-stoichiometric nature of the TOS layer. The coloured lines show deconvolved Gaussian peaks fit to the data. For W 4f, the doublet peak of W 4+ is indicated by light blue (4f 7/2 ) and dark blue (4f 5/2 ). For Se 3d, the doublet peak of Se 2− is indicated by light blue (3d 5/2 ) and dark blue (3d 3/2 ). Multivalent oxidation states are shown in light/dark red and green lines. samples, R sh is even smaller, falling below 50 Ω sq −1 ( Supplementary  Fig. 9). This indicates an increase in carrier mobility that more than offsets the reduced carrier density.
To explore the electrical performance of TOS-doped graphene in more detail, we plot the carrier mobility (at RT) derived from the measured sheet resistance and carrier density, as shown in Fig. 3d,e. With no interlayer WSe 2 , the mobility is ~2,000 cm 2 V −1 s −1 , more than an order of magnitude higher than that achieved in graphene with a similar high carrier density induced by either chemical doping 4,8,9 or electrolyte gating [27][28][29][30] . In spite of this dramatic improvement, the mobility still falls below the limit predicted from acoustic phonon scattering, indicating dominant scattering from charged impurities. The density of such impurities can be estimated from the measured low-temperature mobility µ ≈ 20(e/hp imp ), where p imp is the impurity density, h is the Planck constant and e is the elementary charge ( Supplementary Fig. 8b,c) 31 . From this relation, we estimate p imp ≈ 4.6 × 10 11 cm −2 , well below that of electrolyte-gated graphene (ranging from 6 × 10 12 to 1 × 10 13 cm −2 ) (ref. 32 ). This indicates that charge disorder in the TOS layer is much lower than other dopants.
With the insertion of interlayer WSe 2 , the mobility improves dramatically, reaching values of ~17,000 cm 2 V −1 s −1 for the threeinterlayer sample and ~24,000 cm 2 V −1 s −1 for the four-interlayer sample at zero gate-bias voltage. Remarkably, these values are at the limit set by longitudinal acoustic (LA) phonon scattering 18 . This indicates that three layers of WSe 2 (~2 nm) can screen the charged impurities in the TOS layer such that impurity scattering is reduced by more than an order of magnitude. This is consistent with previous studies showing that the charged-impurity scattering rate rapidly decreases as a function of the distance between the impurities and graphene layer 33 . Mobility improvement achieved in this way is analogous to the modulation-doping technique in conventional semiconductor heterostructures, where the physical separation of dopant from the active channel drastically increases carrier mobility by minimizing impurity scattering by dopant atoms 16 .
To understand the origin of charge transfer between TOS and graphene, we focus on two important electrostatic boundary conditions imposed in our devices in equilibrium: the constant Fermi level E F across the entire system and the continuous vacuum level without any discontinuities. We first note that the charge neutrality point of graphene lies deep in the bandgap of WSe 2 ; therefore, WSe 2 can be effectively treated as a dielectric 34 . Therefore, the resultant electrostatic boundary condition in our TOS-doped graphene with interlayer WSe 2 can be expressed in terms of the work function Φ of the individual layers as where Φ TOS (Φ Gr ) is the work function of TOS (graphene) with respect to the vacuum level, t is the distance between graphene and TOS and ε is the dielectric constant of WSe 2 . Here we use ε = 3.8 (ref. 35 ). The final term in equation (2) is the potential drop developed across TOS and graphene as a result of charge transfer, which can be determined from Poisson's equation 36 . From equation (2), we find that our model effectively fits our data for p in TOS-doped graphene as a function of t (Fig. 4a), indicating that charge transfer is dictated by work-function mismatch between the two layers. Note that the extracted Φ TOS (~5.6 eV) is in good agreement with previous studies on non-stoichiometric tungsten oxides 37 . To validate our model of work-function-mediated charge transfer, we study the TOS-doped graphene devices with different stacking orders of graphene (Gr) and TOS layer (Gr/TOS and TOS/ Gr) as well as with different insulating interlayers (TOS/2L-hBN/ Gr; Supplementary Fig. 10). The Raman spectra clearly show the hole doping of graphene irrespective of the stacking order or type of insulating interlayer (Fig. 4b) 38 . This not only verifies the work-function-mediated charge transfer but also rules out the possibility of doping due to fixed dipoles, as seen for self-assembled monolayers and ferroelectric insulators 39,40 . The densities extracted  RT Doping Open symbols measured at 1.5 K c R sh (Ω sq -1 ) from the Raman shifts, namely, >2.5 × 10 13 cm −2 for the TOS/graphene sample, ~1 × 10 13 cm −2 for the graphene/TOS sample and ~2 × 10 13 cm −2 for the TOS/2L-hBN/graphene sample, are in good agreement with our Hall effect measurements 21 . A slight decrease in doping density of the graphene/TOS sample (compared with TOS/ graphene) can be attributed to the stacking and cleaning processes at elevated temperatures. The proposed doping mechanism indicates that TOS can be an effective hole dopant for many other materials with suitable work-function mismatch. We test this by interfacing TOS with semiconductors spanning various dimensions-one-dimensional semiconductor (single-walled carbon nanotube (SWCNT)), 2D semiconductor (4L-WSe 2 ) and three-dimensional organic semiconductor (dinaphtho[2,3-b:2′,3′-f]thieno[3,2-b]thiophene (DNTT)) (fabrication details are provided in Methods). Supplementary Figs. 11-13 clearly show that hole densities increase in these systems after the placement of TOS on their surface, indicating that high-work-function TOS can universally p-dope a wide range of materials.
We next perform self-consistent electrostatic simulations using the extracted Φ TOS value of ~5.6 eV to gain a further insight into the role of the WSe 2 interlayer in the charge transfer process (simulation details can be found in our previous work 41 ). The additional material parameters are provided in Supplementary Table 2. Specifically, we study the effect of defects in WSe 2 on the resultant p in graphene at RT, as shown in Fig. 4c. Here we assume that the defects are either acceptors (with density N A , solid line) or donors (with density N D , dashed line). For defect densities lower than 5 × 10 11 cm −2 , our simulations show that p can be well understood in terms of equation (2) for any number of WSe 2 interlayers. However, p departs from equation (2) at higher defect densities due to additional charge transfer from acceptors (donors) in WSe 2 to graphene, which further decreases (increases) the resultant hole density in graphene. In our studies, we use high-purity flux-grown WSe 2 with low defect densities (<10 11 cm −2 ) (ref. 42 ). Therefore, the simulations support the model of charge transfer dictated by work-function mismatch and validate the assumption of treating interlayer WSe 2 as a simple dielectric. In this low-defect-density limit, our simulations further show that the induced charge densities in the WSe 2 interlayer are orders of magnitude lower than those in graphene ( Fig. 4d and Supplementary Fig. 14). This indicates that the electrical characteristics of our devices are dominated by the bottom graphene layer with negligible WSe 2 contribution.

Optical characterization of TOS-doped graphene
We next illustrate the potential of TOS-doped graphene in optoelectronic applications. One immediate advantage of our technique is the ability to strongly suppress interband absorption for  21 . c, Self-consistent electrostatic simulation of induced hole density in TOS-doped graphene as a function of defect density for different thicknesses of WSe 2 interlayer. d, The same simulation for 4L-WSe 2 interlayer device shows negligible charges in the WSe 2 interlayers compared with the bottom graphene. photon energies up to 2E F due to Pauli blocking 5 . Figure 5a shows the measured transmittance spectra of CVD-grown 1L-WSe 2 /graphene films on quartz before and after UV-ozone oxidation (see Supplementary Fig. 15 and Methods for the detailed measurement setup). Before oxidation, the transmittance is near graphene's intrinsic value (97.7%) for photon energies less than 1.4 eV since the top WSe 2 is transparent in the near-IR region, and shows a dip at 1.67 eV that corresponds to the excitonic bandgap of WSe 2 . In contrast, the near-IR transmittance notably improves after oxidation, increasing to 99.2% at telecommunication wavelengths (λ ≈ 1,550 nm). From the transmittance data, we can infer that E F ≈ 0.6 eV for our TOS-doped graphene, in reasonable agreement with that from electrically measurements for the exfoliated sample discussed above (0.65 eV). Furthermore, the TOS-doped graphene is highly transparent even in the visible regime (Fig. 5a, insets) indicated by the reduction in the WSe 2 absorption peak. The weak presence of the excitonic peak is due to thickness inhomogeneity in the top CVD-grown WSe 2 layer within the area of illumination (~6% of the area is covered by 2L-WSe 2 , as shown in Supplementary  Fig. 15c). Figure 5b shows that our TOS-doped graphene displays superior performance at telecommunication wavelength (1,550 nm) compared with other conventional transparent conducting films, including indium tin oxide (ITO), zinc-doped indium oxide (IZO), zirconium-doped indium oxide (IO:Zr), hydrogen-doped indium oxide (IO:H), zinc oxide (ZnO) and aluminium-doped zinc oxide (AZO) [43][44][45][46][47] . In general, traditional transparent conductors suffer from a trade-off between transmittance and sheet resistance. In contrast, our CVD-grown TOS/Gr and exfoliated TOS/3L-WSe 2 /Gr samples provide exceptionally high transmittance while maintaining a remarkably low sheet resistance, namely, >99% transmittance at 197 Ω sq −1 and 97.7% at 48 Ω sq −1 , respectively. This highlights both intrinsic suppression of interband transitions in highly doped graphene and the ability of the TOS-doping technique to achieve high carrier density while maintaining high mobility. Before transfer (blue), the ring resonator is in the undercoupled regime, where the ring resonator and bus waveguide exhibit weak coupling due to the ultralow-loss waveguides (2.326 dB cm −1 ). After transfer (grey), the insertion loss in the composite SiN waveguides increases to 0.077 ± 0.014 dB µm −1 , which overcouples the ring resonator to the waveguide, causing considerable broadening of the resonator linewidth. The high loss in the ring can be attributed to undoped graphene. After UV-ozone oxidation, the transmission spectra exhibit a critical coupling condition, where insertion loss due to the stack is reduced by 85% to 0.012 ± 0.0022 dB µm −1 . The error range in the insertion loss measurement before and after oxidation arises due to uncertainty in the length of the stack transferred on the ring resonator (as shown in the inset). Finally, we demonstrate the ability to utilize TOS-doped graphene as a transparent gate electrode and high-speed phase modulator in near-IR photonic circuits 48 . We probe the optical response of TOS-doped graphene embedded on planarized low-loss SiN waveguides, in a microring resonator cavity ( Fig. 5c; Methods provides the detailed fabrication processes). Notably, our planar photonic structure comprises a TOS/Gr/hBN/SiN composite waveguide with a strong optical-mode overlap compared with out-of-plane measurements. The normalized ring transmission spectra show that the bare low-loss cavity is weakly coupled to the straight waveguide (undercoupled regime), thereby yielding a low extinction of ~3 dB on resonance, with a narrow linewidth (Fig. 5d). After the transfer of WSe 2 /Gr/hBN on the planarized SiN substrate, we extract an insertion loss of 0.077 ± 0.014 dB µm −1 in our composite waveguide from the optical response, as shown by the grey colour in Fig. 5d (Methods and Supporting Information provide the insertion loss extraction) 49 . We attribute the high insertion loss to undoped graphene in the WSe 2 /Gr/hBN stack, which causes the resonator linewidth to broaden considerably, increasing the cavity loss and overcoupling the waveguide to the cavity. The insertion loss is lowered by about 85% to 0.012 ± 0.0022 dB µm −1 after UV-ozone oxidation. The lowering of the insertion loss leads to the condition in which the coupling rate between the waveguide and ring resonator equals the optical decay rate in the cavity, thereby exhibiting a critically coupled resonance transmission response (Fig. 5d, red), where the extinction is ~60 dB with the spectral sharpening of the resonance. The magnified change in the in-plane transmission compared with the out-of-plane transmission (Fig. 5a) is attributed to the enhanced optical-mode overlap in the integrated photonic circuits. This low insertion loss of 0.012 dB µm −1 uniquely places TOS-doped graphene as an ideal alternative to traditional transparent conducting materials such as ITO that has at least two orders of magnitude higher insertion loss (1.6 dB µm −1 ) with similar device geometries 50 .

Conclusions
We have reported high-density p-type doping of graphene using work-function-mediated charge transfer from monolayer TOS, prepared by RT UV-ozone oxidation of monolayer WSe 2 . TOS-doped graphene without any WSe 2 interlayers exhibits hole mobilities of 2,000 cm 2 V −1 s −1 , even at a hole density above 3 × 10 13 cm −2 , which resulted in an RT sheet resistance of 118 Ω sq −1 . The carrier density can also be tuned and the mobility can be increased-up to the phonon-limited scattering rate-by inserting WSe 2 interlayers, with a minimum sheet resistance below 50 Ω sq −1 . A self-consistent electrostatic model based on work-function mismatch can be used to describe the charge transfer mechanism, as well as the carrier density dependence on WSe 2 interlayer number. Our TOS-doped graphene also exhibits high optical transmittance (>99%) and low insertion loss (0.012 dB μm −1 ) at telecommunication wavelength. Our approach offers a promising doping method that could be used to incorporate van der Waals heterostructures into photonic circuits as a transparent gate electrode and a high-speed phase modulator for near-IR applications.

Methods
Fabrication and characterization of graphene device. WSe 2 , graphene and hBN flakes were prepared on an SiO 2 /Si substrate by mechanical exfoliation. The thickness of each flake was determined by the contrast difference in the optical microscopic images and Raman spectra. Only monolayer graphene was used, whereas the WSe 2 thickness varied from one layer (1L) to five layers (5L) to verify the layer dependence. The stacking of flakes was conducted by the dry-transfer method using PCL polymer at 50-58 °C to pick up flakes that are then transferred onto 285 nm SiO 2 /Si at 80 °C to melt the polymer. To remove the polymer, the sample was annealed at 340 °C under a vacuum condition. Edge contacts were formed to the graphene layer by etching through the layers and subsequently depositing electron-beam (e-beam)-evaporated metal. Cr/Au (2/80 nm) contacts were deposited by e-beam evaporation after reactive-ion etching (RIE) of the WSe 2 , graphene and hBN layers with CHF 3 to form the edge contact. UV-ozone oxidation (Samco UV-2 system located in a class 100 cleanroom controlling humidity at <40% and temperature at <20 °C) was conducted at RT for 30 min at an oxygen flow rate of 3 l min −1 . The system has a separate discharge-type ozone generator and the generated ozone flows into the sample chuck underneath the UV-light source. The UV lamp is a mercury grid lamp with primary wavelengths of 254 and 185 nm and output power of ~10 mW cm −2 . The chuck temperature was maintained at <50 °C during the oxidation process. Etching to remove the TOS was conducted using 1 M KOH diluted in deionized water followed by rinsing with deionized water and annealing in a vacuum at 300 °C. Electrical measurements were performed with both semiconductor parameter analyser (Keysight B1500A) and lock-in amplifiers (Stanford Research Systems SR830) connected to a cryostat containing a tunable perpendicular magnetic field under vacuum conditions. Fabrication of SWCNT, WSe 2 and DNTT devices. SWCNT device. CNTs were grown on an SiO 2 /Si substrate at 890 °C. The locations of SWCNTs were identified using scanning electron microscopy and atomic force microscopy scans. Cr/ Pd (2/20 nm) contacts were then fabricated on select SWCNTs using the lift-off method. Here 1L-WSe 2 was transferred and subsequently oxidized after the initial electrical measurements of the SWCNT (without TOS) were completed. WSe 2 device. In this case, 4L-WSe 2 was stacked on hBN using the same dry-transfer process used with graphene. E-beam evaporation and lift-off method were used to create the top surface contacts of Pd/Au (20/50 nm) to WSe 2 . Electrical measurements of the same device before and after UV-ozone oxidation are presented in the manuscript.
DNTT device. Two devices were made on the same chip, one with TOS and one without TOS. To create the device with TOS, we first exfoliated 1L-WSe 2 on an SiO 2 /Si substrate followed by lift-off metallization with Ti/Pd/Au (2/20/20 nm). Further, 1L-WSe 2 was converted into monolayer TOS through UV-ozone oxidation. Contacts were also concurrently patterned in areas without TOS to form the second device. Then, 40 nm DNTT was deposited via sublimation. The channel areas were defined by coating the sample with PMMA, patterning with e-beam lithography and etching away the semiconductor with SF 6 plasma, leaving the active channel area (Supplementary Fig. 13).
Transmittance measurements. The transmittance of 1L-WSe 2 /Gr that was directly grown on a quartz substrate (purchased from 2D semiconductors) via CVD was measured under the ambient condition using a custom setup built around a Nikon TE-300 inverted microscope, as shown in Supplementary Fig. 15. The output of a tungsten halogen lamp was focused onto a 50 μm pinhole and an aspheric condenser lens to obtain collimated white light, which was focused at the sample plane through the quartz substrate using a long-working-distance ×50 objective (numerical aperture, 0.55). The transmitted light was collected with ×40 (numerical aperture, 0.60) objective focused at the sample plane from below and was sent to an f/4 (where f is the focal length) spectrograph (Princeton Instruments SpectraPro HRS-300) equipped with a cooled InGaAs array detector (Princeton Instruments PyLoN IR). Order-sorting filters were used to avoid higher-order diffraction signals. Transmittance was calculated by dividing the transmitted intensity measured on the sample by the transmitted intensity through a nearby blank area of quartz, with dark counts subtracted from both measurements. The system was optimized to ensure that the instrument error was <0.5% over the whole spectral range for each measurement, as indicated by the shaded area of the spectrum.

Fabrication of SiN photonic platform with TOS-doped graphene.
To fabricate low-loss SiN waveguides, we use e-beam lithography to define 1,300-nm-wide waveguides on a stack of 360 nm plasma-enhanced-CVD-grown SiO 2 hard mask and 330-nm-high SiN, deposited using low-pressure CVD at 800 °C and annealed at 1,200 °C for 3 h on 4,170 nm thermally oxidized SiO 2 . We use CHF 3 / O 2 (52/2 s.c.c.m.) chemistry in an Oxford Plasmalab 100 ICP RIE system to etch the hard mask of 360 nm SiO 2 and subsequently use an optimized etch recipe of CHF 3 /O 2 /N 2 (47/23/7 s.c.c.m.) that uses a higher oxygen content reducing both polymerization and side-wall roughness. We use 360 nm SiO 2 as a sacrificial hard mask for etching SiN, since increasing the O 2 flow lowers the selectivity of etching. Following the etch, we deposit about 600 nm plasma-enhanced-CVD-grown SiO 2 on the waveguides. We planarize the SiO 2 to about 70-100 nm above the SiN waveguides using standard chemical planarization techniques to create a uniform surface for the transfer of the WSe 2 /Gr/hBN stack on the ring resonator. Subsequently, a stack of 1L-WSe 2 /1L-Gr/~20-25 nm hBN was transferred onto the planarized SiN waveguides using a PCL dry-transfer method. To remove the polymer, the sample was annealed at 340 °C under vacuum conditions. Finally, UV-ozone oxidation is performed to oxidize WSe 2 into TOS and to dope graphene. Notably, microring resonators allow the high-fidelity measurement of the optical propagation loss induced by TOS-doped graphene due to the enhanced sensitivity of a high-quality-factor (Q) cavity (~300,000 in our case) to small changes in phase and absorption.
Optical loss measurement and insertion loss estimation. We couple transverse-electric-polarized light from a tunable near-IR laser (~1,550-1,650 nm) to the input of the SiN microring resonator (schematic shown in Fig. 5c) using a tapered single-mode fibre, which is then collected from the SiN ring output, using a similar tapered fibre. Transmission measurements were taken from the same device (1) before transfer to account for the insertion loss of planarized SiN substrates, (2) once after transfer to account for the insertion loss due to undoped graphene and (3) finally after UV-ozone oxidation to measure the propagation loss due to the composite waveguide with doped graphene. The original normalized transmission responses are shown in Supplementary Fig. 16. We normalize each ring transmission spectrum by the maximum transmission power at the output at wavelength detuning ∆λ, where the cavity does not interact with the incoming light. We estimate the insertion loss by fitting the experimental normalized ring transmission to the theoretical ring transmission (T) obtained from the following resonator equation: where 1 τe is the decay rate (coupling rate) from the waveguide (bus) to the ring resonator, 1 τ0 is the decay rate due to loss in the ring, and ωω 0 is the frequency detuning (ω 0 , the resonant frequency). The Q value of the ring resonator is related to τ 0 and α (insertion loss in the linear scale) through the equation Q = ω0τ0 ≈ 2πng αλ0 , where n g is the group index, ω0 = 2πc λ0 , c is the speed of light in vacuum and λ 0 is the resonant wavelength.

Data availability
The data that support the findings within this paper are available from the corresponding authors upon reasonable request.