Deconvoluting the Photonic and Electronic Response of 2D Materials: The Case of MoS2

Evaluating and tuning the properties of two-dimensional (2D) materials is a major focus of advancing 2D science and technology. While many claim that the photonic properties of a 2D layer provide evidence that the material is “high quality”, this may not be true for electronic performance. In this work, we deconvolute the photonic and electronic response of synthetic monolayer molybdenum disulfide. We demonstrate that enhanced photoluminescence can be robustly engineered via the proper choice of substrate, where growth of MoS2 on r-plane sapphire can yield >100x enhancement in PL and carrier lifetime due to increased molybdenum-oxygen bonding compared to that of traditionally grown MoS2 on c-plane sapphire. These dramatic enhancements in optical properties are similar to those of super-acid treated MoS2, and suggest that the electronic properties of the MoS2 are also superior. However, a direct comparison of the charge transport properties indicates that the enhanced PL due to increased Mo-O bonding leads to p-type compensation doping, and is accompanied by a 2x degradation in transport properties compared to MoS2 grown on c-plane sapphire. This work provides a foundation for understanding the link between photonic and electronic performance of 2D semiconducting layers, and demonstrates that they are not always correlated.

-The HRSTEM image of the monolayer-bilayer MoS2. The low-magnification HRSTEM image of two aligned monolayer domains (a) and misaligned monolayer domains (b). The low magnification HRSTEM image of aligned bilayer domains (c), and the atomic-resolved micrograph in the grain boundary region (d).

Sapphire annealing study
The as received sapphire is cleaned by the following process: acetone sonication for 10 minutes, IPA sonication for 10 minutes, followed by the DI water rinse. After that, the substrate is subsequently soaked in Nanostrip™ at 80˚C for 20 minutes. After the cleaning, the substrate is annealed in the air with the conditions shown in table S1, followed by a second cleaning as described above. Thermal treatment of the sapphire substrate is required to achieve long-range commensurability and epitaxy of MoS2 on sapphire surfaces. 1,2 Such thermal treatments lead to the formation of atomic steps on sapphire, 3 and the particular annealing conditions can significantly alter the physical size and chemical functionality of the steps, which substantially modifies the quality and yield of monolayer MoS2. The atomic force microscope (AFM) images of annealed sapphire are shown in Figure S4 (a-l). The nano-steps start to form at 900˚C for 1 hour and the step height is characterized as ~0.2 nm, agreeing with the 1/6 c distance of sapphire. 3,4 The initial steps are zig-zag shape while the step edges become more linear with increased annealing temperature and time. When the annealing temperature/time hit the critical condition (ex: Figure S4 (c, d, h)), the sapphire exhibits a particle-free surface with smooth step edges. The particles and step bunching are observed on sapphire when we keep increasing the annealing temperature and annealing time (ex: Figure S4 (e, i)), indicating the over-etching by the annealing. 3,4 Figure S4-AFM micrographs of thermally treated sapphire substrates. The surface reconstruction is clearly observed when the annealing temperature is > 1000 C. The optimum surface (smooth terrace edge and particle-free) can be achieved in the arrow-indicated row (black arrow), indicating the medium strength of annealing (temperature and time) results in the optimum sapphire surface for MoS2 growth. Strengthening the annealing (increase annealing time and temperature) results in step bunching and particles, and weakening the annealing results in curve steps.
Discontinuous films of monolayer MoS2 domains are subsequently grown by powder vaporization technique. 2 mg MoO3 (99.8%, Sigma Aldrich) is put in an alumina crucible located in the center of a hot wall furnace. 200 mg of S powder (99.995%, Alfa Aesar) is put in a quartz crucible located ~12 inches upstream of the MoO3 crucible. After 5 min of flowing 100 sccm Ar at 710 Torr, the furnace is elevated to 550 ˚C for 2 min as a nucleation step, followed by dwelling at 800 ˚C for 15 min for the growth. The sulfur powder is heated by an individual heat tape wrapped outside the furnace, and the temperature is 130 ˚C. The as grown MoS2 on sapphire is characterized by AFM, scanning electron microscope (SEM) and Raman spectroscopy ( Figure S5 (a-c), respectively). The AFM image shows triangular MoS2 domain, and the thickness is ~6.7 Å, agreeing with previous reports for monolayers. [5][6][7] The low magnification SEM image shows there is no alignment. Three peaks: E2g (385.1 cm -1 ), A1g (402.9 cm -1 ) and a tiny sapphire peak (419.1 cm -1 ) are observed. Interestingly, the A1g is significantly quenched and red shifted compared to some other PV-grown MoS2 monolayers on SiO2, 6,8 strongly indicating electron doping and charge transfer to the MoS2. 9 The photoluminescence (PL) measurement is also carried out at 4.1 mW and in 488nm laser wavelength. The peak position is 1.83 eV, and the PL/A1g ratio is ~1.3. It is clear that PL is quenched and red shifted by 0.5 eV in our case compared to others. 10 The PL change also suggests the electron doping of the MoS2. 11 The AFM and SEM micrographs of the synthetic MoS2 monolayers on the sapphire with different thermal treatments are shown in Figure S6 and Figure S7. It is clear that the particle density and alignment varies based on the variation of annealing process.  For a better comparison, we pick some representative cases and summarized them in Figure S8. As-received sapphire substrates ( Fig. S8 (a)), sapphire annealed at 900˚C for 1hr ( Fig. S8 (b)), 1150˚C for 8hrs ( Fig. S8 (c)), and 1350˚C for 1hr ( Fig. S8 (d)) illustrate the evolution in surface topography as a function of annealing conditions. We find no significant morphology changes for sapphire annealed at 900˚C for 1hr compared to the as-received sapphire. In contrast, sapphire annealed at 1150˚C for 8 hr (Fig. S8 (c)) exhibits significantly smoother step-edges, indicating a surface reconstruction of the sapphire surface. 3 Finally, upon annealing at 1350˚C for 1hr ( Fig. S8 (d)), significant step-bunching -where multiple steps merge to form a single larger step -is observed, indicating an "over etching" of the sapphire. 4,12 Subsequently, MoS2 was grown on the as-received and annealed sapphire substrates ( Fig. S8 (e), (f), (g), (h)), where well aligned, particlefree MoS2 domains are achieved on sapphire annealed at 1150˚C for 8 hr. In contrast, MoS2 growth on the sapphire substrates annealed under the other two conditions yields lower fractional alignment, as well as "3D" particles. To better understand the role of the substrate annealing process on the MoS2/substrate interactions, the water contact angle (controlled by the substrate surface energy) before growth is measured, followed by a statistical survey of the MoS2 interdomain alignment after the growth reaction for the different annealing conditions. Upon comparison of contact angle on the annealing conditions ( Fig. S8 (i)), we clearly find that increased annealing duration reduces the contact angle for all evaluated temperatures up to an optimal time. A summary of the inter-domain alignment ( Fig. S8 (j)) reveals that the highest degree of crystalline alignment is observed on the substrate which was annealed at 1250˚C for 1hr, which also exhibits the lowest contact angle. There is an inversely proportional trend between contact angle (surface energy) and alignment percentage and domain size. As the surface energy (contact angle) increases (decreases), the sapphire surface becomes more reactive due to a higher density of ordered Al-and O-dangling bonds on the surface. This ultimately leads to enhanced nucleation on the sapphire surface and a reduced energy barrier to achieve long-range registry of the MoS2 atoms with the Al2O3. 2,13 Interestingly, at the lowest evaluated annealing temperature (900˚C), the alignment is more dominated by the annealing time rather than contact angle ( Fig. S8 (i, j)), suggesting that under certain conditions one is able to achieve similar surface chemistries required for alignment, while maintaining a relatively low surface energy (high contact angle). In order to understand the surface chemistry of the sapphire, angle resolved XPS (ARXPS) was used. ( Figure S9) A Phi Versa Probe II is used for this analysis, with a passing energy of 23.5 eV and a step size of 0.1 eV. All of the samples are integrated at the same dwell time (200 ms) at each step. There is no noticeable beam damage to the sample. The high-resolution spectra for Al at various temperatures did not change. There was no noticeable change as we went down to 30 0 which would result in observing the top 1 nm of the sample surface. It can be noted that the intensity went down as the total probe depth decreased to 1 nm. There is no change in the overall peak structure, indicating that the surface has the same composition similar to the bulk of the substrate. This suggests that the large, systematic changes in the surface energy with annealing ( Figure S8 (i, j)) may be the result of the removal of defective (or even amorphous) Al2O3 at the surface generated by the polishing process during wafer manufacturing. Step bunching is observed on the sapphire that was annealed at 1350˚C for 1 hr, indicating the over-etching of the sapphire. (e-h) the AFM micrographs of the monolayer MoS2 grown on the sapphire substrates that were annealed under the corresponding conditions (by row) specified in (a-d).
While particle-free, triangular MoS2 is grown on the as-received sapphire (e), it is misaligned. The well-aligned and particle free MoS2 is grown on the 1150C 8 hour annealed sapphire (g), and imperfect alignment and particles are found in the other two cases (f and h). (i-j) the dependence of the contact angle of the sapphire surface (i) and the alignment fraction of MoS2 (j) on annealing temperatures for various annealing times. The contact angle can be decreased by strengthening the annealing (i.e., increasing temperature and time) to a minimum value, which yields the maximum alignment fraction. Figure S9 -The ARXPS spectra of sapphire after thermal treatment. The ARXPS spectra in Al 2p range at 85 (a) and at 30 (b). It is clear that there is no significant peak shift and peak shape change after thermal treatment. Note that the peak intensity is higher when the detection angle is at 85 than it is at 30, which is due to the probe thickness is higher at 85. The surface termination of sapphire can be modeled and determined by visualizing the height profile of the sapphire surface. 14 To understand the surface termination of our sapphire, we conducted a careful AFM characterization on our as-received c-sapphire, annealed sapphire (1150 ˚C, 8 hours) and as received r-sapphire, respectively ( Figure S10 (a-c)). The height profile is extracted in Figure S10 (d-f) as black line, and the height profile of the reference 14 is drawn in red. It is clear that the height profiles of the as-received c-sapphire, annealed sapphire and rsapphire match the referenced height profile of Al-terminated sapphire, Al-O terminated sapphire and O-terminated sapphire, respectively, indicating the corresponding surface terminations of our sapphire substrates. Interestingly, the step width of our r-sapphire is 3 larger than that of the referenced O-terminated sapphire, which can be due to the larger O-O distance on the r-sapphire surface. 15  The PL of monolayer MoS2 grown on c-sapphire is significantly different from that of monolayer MoS2 grown on r-sapphire. The PL spectral of MoS2 on c-sapphire (a) and r-sapphire (b) are shown in Figure S11. After Lorentzian peak fitting, the trion peak (red) and exciton peak (green) is clearly identified at 1.83 eV and 1.87 eV and agrees with the previous reports. 11,16 It is clear that the trion-dominated PL changes to exciton-dominated PL when the growth is on r-sapphire instead of c-sapphire, indicating that less negative charge is transferred to the MoS2 grown on r-sapphire. From the peak fitting, we are able to estimate the electron concentration of the monolayer MoS2 on c-sapphire and r-sapphire, 17,18 and the estimated electron concentration is 3.14 × 10 12 −2 and 9.82 × 10 10 −2 on c-sapphire and r-sapphire respectively. This value also agrees well with the electron concentration vs. decay time in our measurements and previous reports from Amani et al. 19 Figure S11-The PL peak fitting of MoS2 grown on c-sapphire (a) and r-sapphire (b).

Rate equation analysis of power-law scaling and multi-exciton threshold
A simple rate equation can be used to model the multi-exciton effects on the power-law scaling curves reported in Figure 2 of the main text and here in the SI for edges, grain boundaries and interior regions 20 : where is the density of excitons, is the time after pulsed excitation, is the excited state lifetime due to radiative and non-radiative recombination in absence of multi-exciton effects, and is the exciton-exciton annihilation rate. This equation can be solved analytically for: where 0 is the initial exciton density created by an optical excitation pulse, and = 1/ is the excited state relaxation rate (in absence of multi-exciton effects). From this solution, the time-integrated emission intensity for different initial exciton densities (i.e., 0 ) is can be calculated by numerically integrating ( ): Figure S11 shows the dependence of the time-integrated emission on excitation density for two different excited state lifetimes. For these numerical calculations, we used an exciton-exciton annihilation rate of = 4.3 2 / that was determined in previous studies 21 and calculated 0 assuming 5% absorption at an excitation energy of 2.40 eV. The excited state lifetimes used were extracted from the PL decay transients in Figure 3 by performing a fitting a convolution of ( ) with the instrument response function of our experimental setup to enable a rough comparison to the data of Figure 2. A decrease in the excited state lifetime by ~100, which is comparable to the differences between the r-sapphire and c-sapphire substrates in the main text, results in a power-law scaling that is largely linear over these excitation densities. However, with a longer excited-state lifetime, the transition from a linear to sublinear power-dependence occurs at a much lower excitation density. These trends qualitatively reproduce what is observed in the main text for the c-sapphire and r-sapphire substrates and further support the hypothesis that nonradiative rate suppression plays a significant role in the enhancement of the photoluminescence of MoS2 on sapphire and at the edges and misaligned grain boundary regions. Figure S12 -Dependence of the excitation density threshold that marks the onset of multi-exciton dominated recombination.

Field effect transistors (FET) device fabrication
Following MoS2 growth on sapphire substrates, an array of Ti/Au (10/90 nm) alignment/fiduciary are deposited by standard optical lithography, e-beam evaporation, and lift-off processes. Large triangular monolayer MoS2 flakes (40-60 μm side length) are then etched into smaller rectangular channels of varying widths and lengths (dictated by original flake size) to be used as a TLM bar of uniform composition/thickness and defined width. Following electron beam lithography (EBL), the MoS2 etch is carried out in a Plasma Therm PT-720 Reactive Ion Etch (RIE) tool using a gas chemistry of SF6/Ar/O2 (30/10/10 sccm) at 100 W power and 10 mTorr pressure for a total etch time of 20 seconds. Minimal etching of the sapphire substrate is observed. Two subsequent metallization steps are then carried out to fabricate TLM (FET) structures suitable for electrolyte gating in the form of an initial thin source/drain metallization (actually contacting the MoS2 film) and a secondary large-area thick metallization for easier probing. First, 25 nm Ni source/drain extensions (grey metal fingers in Figure 3a) are formed by EBL, ebeam evaporation, and lift-off processes. Ni is deposited at 0.5 Å/sec at a deposition pressure of 3x10 -6 Torr. The width of these Ni source/drain fingers is 2 μm. Prior to Ni source/drain deposition and after ebeam resist development, samples are O2/He (150/50 sccm) plasma treated (50 W and 500 mTorr) for 45 seconds in an M4L etch RIE tool in order to chemically remove resist polymer residues and thereby improve the metal/MoS2 interface. Finally, 15/165 nm Ti/Au interconnects and pads are deposited to ensure ease of probing. In addition to thick pads and leads, a "side-gate" structure adjacent to the TLM structure is also written in this same EBL step; this side-gate is electrically isolated from the MoS2 channel and source/drain contacts, and is used to apply a voltage to the electrolyte with a third-terminal probe which is discussed in more detail in the electrolyte measurement section. This device geometry is chosen for several reasons. First, it provides a constant gate-MoS2 distance (~ 6 μm) and electric field which allows for consistent measurements (i.e. sweep rate) between devices. Secondly, it is much easier to make electrical contact with a probe-metal pad configuration than it is to consistently insert a probe into a thin electrolyte directly above a device at a constant height. 22

Field effect transistors (FET) device fabrication
Following MoS2 growth on sapphire substrates, an array of Ti/Au (10/90 nm) alignment/fiduciary are deposited by standard optical lithography, e-beam evaporation, and lift-off processes. Large triangular monolayer MoS2 flakes (40-60 μm side length) are then etched into smaller rectangular channels of varying widths and lengths (dictated by original flake size) to be used as a TLM bar of uniform composition/thickness and defined width. Following electron beam lithography (EBL), the MoS2 etch is carried out in a Plasma Therm PT-720 Reactive Ion Etch (RIE) tool using a gas chemistry of SF6/Ar/O2 (30/10/10 sccm) at 100 W power and 10 mTorr pressure for a total etch time of 20 seconds. Minimal etching of the sapphire substrate is observed. Two subsequent metallization steps are then carried out to fabricate TLM FET structures suitable for electrolyte gating in the form of an initial thin source/drain metallization (i.e., the contact metals that actually touch the MoS2) and a secondary, large-area thick metallization to provide robust contacts that can withstand repeated probing. First, 25 nm Ni source/drain extensions (grey metal fingers in Figure 3a) are formed by EBL, e-beam evaporation, and lift-off processes. Ni is deposited at 0.5 Å/sec at a deposition pressure of 3x10 -6 Torr. The width of these Ni source/drain fingers is 2 μm. Prior to Ni source/drain deposition and after e-beam resist development, samples are O2/He (150/50 sccm) plasma treated (50 W and 500 mTorr) for 45 seconds in an M4L etch RIE tool in order to chemically remove resist polymer residues and thereby improve the metal/MoS2 interface. Finally, 15/165 nm Ti/Au interconnects and pads are deposited to ensure ease of probing. In addition to thick pads and leads, a "side-gate" structure adjacent to the TLM structure is also written in this same EBL step; this side-gate is electrically isolated from the MoS2 channel and source/drain contacts, and is used to apply a voltage to the electrolyte with a third-terminal probe which is discussed in more detail in the electrolyte measurement section. This device geometry is chosen for several reasons.
First, it provides a constant gate-MoS2 distance (~ 6 μm) and electric field which allows for consistent measurements (i.e. sweep rate) between devices. Secondly, it is much easier to make electrical contact with a probe-metal pad configuration than it is to consistently insert a probe into a thin electrolyte directly above a device at a constant height. 22

Electrolyte gating application
A solid polymer electrolyte (PEO:CsClO4) is used for ionic gating. The preparation of the polymer electrolyte is similar to previously published procedures 23 with the exception that the electrolyte is prepared and deposited in an argon-filled glovebox where the concentrations of H2O and O2 are maintained to be < 0.1 part-per-million (ppm). Poly(ethylene-oxide) (PEO) (molecular weight 95,000 g/mol, Polymer Standards Service) and anhydrous CsClO4 (99.9%, Sigma-Aldrich) are dissolved in anhydrous acetonitrile (Sigma-Aldrich) with an ether oxygen to Cs molar ratio of 76:1 to make a 1 wt% solution. The solid polymer electrolyte is deposited on the sample by drop-casting 25 µL onto the ∼1 x 1 cm 2 sample. After a 15 minute wait time to allow the majority of the solvent to evaporate, the sample is annealed on a hotplate at 80 °C for 3 mins to drive off remaining solvent. The sample is then transferred from the glovebox to the probe station through an Ar-filled load lock. The entire process of electrolyte preparation, deposition, transfer to the probe station, and measurement are completed under an inert gas environment with no sample exposure to ambient. Electrical measurements are performed on a Lake Shore cryogenic vacuum probe station (CRX-VF) under ∼ 10 -6 Torr at room temperature using a Keysight B1500A semiconductor parameter analyzer.

Film tearing and delamination
Both MoS2/c-sapphire and MoS2/r-sapphire samples are processed with the identical procedure as follows: both samples are dip-coated in Surpass 4000 adhesion promoter prior to spin-coating of the ebeam resist ZEP 520A (1:1 ZEP:Anisole) to improve coating over the ~ 100 nm thick metal alignment mark features. This is needed to prevent etching of alignment marks during removal of the sacrificial Au layer (~15 nm) that is thermally evaporated on top of the e-beam resist stack to decrease charging effects during EBL exposure on insulating sapphire substrates. This Au wet etch (TFA Au Etchant) is performed after EBL exposure and prior to resist development. Following 45 sec Surpass dip, samples are then placed in DI water, and N2 dried. After the process, we find that MoS2/c-sapphire sample displays partially torn channels while all MoS2/r-sapphire channels remain intact. This discrepancy in 'adhesion' to the substrate could be due to the increased chemical bonding with the substrate in the case of rsapphire. Meanwhile, the MoS2/r-sapphire samples exhibit micro-cracking within the MoS2 monolayer film which is observed predominantly for channel lengths of 1 μm and smaller. These micro-cracks can be seen running parallel and directly adjacent to the source/drain fingers of the TLM structure. This micro-cracking is observed only after deposition and lift-off of Ni contacts. This is also observed for MoS2 on c-plane sapphire but with less frequency and for mainly 0.5 μm channels and smaller. This minor discrepancy could also have to do with the difference in degree of chemical bonding between MoS2 and the underlying substrate for r-plane vs c-plane, which could impact the ability of a monolayer film to respond to and accommodate external stresses. In qualitative terms, the MoS2 monolayers on r-sapphire devices are more constrained by the substrate, which could lead to critical mechanical failure possibly experienced during metal deposition from a combination of strain and thermal cycling. Nevertheless, this tendency to fracture for both cases could have important implications for channel length scaling on sapphire substrates and is worth further investigation on how to potentially mitigate this issue in the future. Cracking is normally observed for channel lengths < 1 μm. (b) An SEM image of a TLM structure on MoS2/csapphire showing the tearing and delamination of monolayer film from the substrate encountered during device fabrication, observed on most TLM devices on c-sapphire. In addition, cracking was also observed in MoS2/csapphire devices but to a lesser extent (not shown). No tearing or delamination was observed for MoS2/r-sapphire, except for 1 out of 16 TLM structures. Scale bars are 2 μm.

Transfer Length Measurement (TLM)
Prior to doing any gated measurements (transfer, output characteristics) on a particular TLM structure corresponding to an individual flake, two-terminal IV curves were measured on all devices of varying channel lengths with a voltage range of -100 -100mV (Vg = 0V) ( Figure S13) It is important to note that total resistance (y-axis) is plotted in log-scale for (a-b) in order to more clearly display the change in magnitude as a result of electrolyte deposition, which is consistent across every TLM structure and device on both samples. Resistance data from devices with channel lengths below 0.5 μm (1 μm) for MoS2/r-sapphire (MoS2/c-sapphire) were not included due to extremely high resistance values or negligible current (open-circuit), most likely from channel cracking and tearing discussed in Figure S12. The reduction in total resistance is due to a combination of both decreased sheet resistance (linear slope) and contact resistance (linear y-intercept) from ionic electrostatic doping. In fact, TLM data is unreliable prior to electrolyte deposition; there is more scatter in the data, and linear extrapolation of the data gives a negative yintercept and hence a negative contact resistance. (c) TLM plots for the same device sets from (a-b) from both MoS2/c-sapphire and MoS2/r-sapphire, plotted in linear scale, showing the lower total resistances for MoS2/csapphire compared with MoS2/r-sapphire. Contact resistance, sheet resistance, transfer length, and specific contact resistivity values are tabulated in Table S2 below. Table S2-Summary of TLM data for MoS2/r-sapphire and MoS2/c-sapphire devices. (a) Contact resistance (Rc), sheet resistance (Rsh), transfer length (Lt), and specific contact resistivity (pc) for each TLM structure/device is shown. Also included in this table are the exact channel lengths used for linear fitting as well as the linear fit coefficient (R 2 ). *Denotes linear trendline fit with a negative y-intercept, and hence negative contact resistance, which is not a realistic.

Electrolyte gating measurements
All transfer curves (Id-Vg) shown are taken by sweeping Vg from +3 to -4 V. Prior to collecting each transfer curve, a constant gate bias of Vg = 3 V is applied for 5 min to allow the ions in the electrolyte/MoS2 system to reach equilibrium. A 5 min hold time prior to initiating transfer curve measurement is determined by monitoring Id under constant gate bias (Vg = 3 V) and small drain bias over time; when Id is stabilized, this indicates equilibrium has been established. Because ion transport is slow compared to electrons/holes, sufficient time must be provided for the ions to respond to the changing gate bias (i.e., slow sweep rates are required to ensure repeatability and reduce hysteresis). All transfer curves reported in this manuscript, unless otherwise stated, are initially swept from +3 to -4 V with voltage steps of 150 mV and a dwell time between voltage steps of 30 sec under a drain bias of 500 mV, corresponding to a sweep rate of 5 mV/s. For each sample, devices were first measured using multiple sweep rates to determine the sweep rate that yields the most reproducible results with the smallest hysteresis. A sweep rate of 5 mV/s was determined as the optimal sweep rate because further decreasing the rate did not change the transfer measurement to an appreciate extent. To ensure a fair device-device and samplesample comparison, all transfer curves are taken with the same sweep direction and sweep conditions.

Field-effect mobility calculations
Field-effect mobility (μFE) was calculated using the equation = • ℎ ℎ • • . In this case, gm is the transconductance of the device which is the slope of the transfer curve (Id-Vg) in the linear region (when plotted in linear scale). Lch and Wch are the length and width of the device channel, respectively. Vds = 500 mV is the drain bias that was used for all transfer curve measurements and subsequent calculations of Vth and SS reported in the main text. CEDL in this case is the capacitance of the electrolyte double layer that is formed at the interface. For PEO:CsClO4 on monolayer MoS2, a CEDL = 1 μF/cm 2 was used.