Spontaneous doping of the basal plane of MoS2 single layers through oxygen substitution under ambient conditions


The chemical inertness of the defect-free basal plane confers environmental stability to MoS2 single layers, but it also limits their chemical versatility and catalytic activity. The stability of pristine MoS2 basal plane against oxidation under ambient conditions is a widely accepted assumption however, here we report single-atom-level structural investigations that reveal that oxygen atoms spontaneously incorporate into the basal plane of MoS2 single layers during ambient exposure. The use of scanning tunnelling microscopy reveals a slow oxygen-substitution reaction, during which individual sulfur atoms are replaced one by one by oxygen, giving rise to solid-solution-type 2D MoS2−xOx crystals. Oxygen substitution sites present all over the basal plane act as single-atom reaction centres, substantially increasing the catalytic activity of the entire MoS2 basal plane for the electrochemical H2 evolution reaction.


Layered materials display thickness-dependent properties when approaching the single layer limit. Their chemical properties are no exception, as evidenced by the oxidation and hydrogenation of graphene1,2. Chemically modifying the basal plane of 2D materials, given their fully exposed atomic structure, provides a particularly promising approach for engineering their properties. However, the chemistry of two-dimensional (2D) transition metal dichalcogenide (TMDC) crystals is mainly defined by their edges3, where coordinatively unsaturated sites prevail. These reactive edge sites are also proposed to be responsible for the catalytic activity of MoS2 (refs 4,5), one of the most widely studied TMDC materials. However, chemical modification restricted to edges is only efficient for nanoscale islands6, because the edge-to-surface ratio drastically decreases with increasing lateral dimensions of the sheets. Consequently, for large-area 2D crystals it is of particular importance to increase the chemical and catalytic activity of the entire basal plane.

The oxidation of graphene has been studied widely as a promising approach towards its efficient exfoliation7. For 2D TMDC crystals the oxidation reaction is also of particular importance, because for some crystals it can spontaneously proceed under ambient conditions. Therefore, the study of this reaction is essential for understanding the long-term environmental stability of such crystals, as well for the possibility of creating new routes towards chemically engineering their properties.

Some TMDC crystals, such as HfSe2, MoTe2 or WTe2, are known to be air-sensitive8,9,10 because they rapidly degrade under ambient conditions, but the most widely investigated members of the TMDC family (MoS2, MoSe2, WS2 and WSe2) have generally been considered air-stable11,12, based also on decades-long experience with their bulk crystals. Nevertheless, it has been demonstrated that in single-layer form the oxidation of MoS2 and WS2 also occurs under ambient conditions13. This detailed investigation revealed that the oxidation-induced etching observed originates from edges and grain boundaries and proceeds towards the interior of the flakes. Due to the higher strength of the Mo–O bonds compared to Mo–S, the substitutional oxidation of the 2D MoS2 basal plane is in principle also thermodynamically favourable14,15. However, although such oxidation is a fast, low-barrier process14 at under-coordinated atomic sites on edges and grain boundaries15, oxidation of the defect-free basal plane has been predicted to face relatively high kinetic barriers of ~1.6 eV (ref. 16), rendering the basal plane environmentally stable. Although experiments often reveal significantly lower activation energy values for MoS2 oxidation, such as 0.54 eV (ref. 17) or 0.98 eV (ref. 18), no direct evidence for the ambient oxidation of the pristine MoS2 basal plane has been reported so far. We propose that this is mainly due to the inability of previously employed structural characterization methods to resolve single-atom-level structural changes.

Harsh oxidation processes, such as oxygen plasma treatment, UV–ozone exposure, electrochemical exfoliation or high-temperature (>300 °C) annealing are able to oxidize the basal plane of MoS2 crystals. Based primarily on X-ray photoelectron spectroscopy (XPS) investigations it has been shown that such processes can lead either to covalent oxygen bonding to the top sulfur atoms14,19 or the formation of completely oxidized MoO3 areas20,21,22 that can subsequently volatilize, leading to etching. However, neither process is ideal for chemically tuning the properties of MoS2 sheets. The formation of MoO3 destroys the original MoS2 crystal lattice, yielding an overall disordered and fragmented structure, while the chemisorption of oxygen onto chalcogen atoms is predicted to have a relatively weak influence on band structure and properties14,16. It has also been shown that oxidation can both enhance and degrade the catalytic activity of MoS2, depending on the structural details23,24, further emphasizing the importance of investigating and controlling the resulting oxidized structure. The possibility of a substitutional oxidation reaction for MoS2 has also been raised as being thermodynamically preferred to oxygen chemisorption15,25; however, no clear experimental evidence has been reported so far. The controlled substitutional oxidation of the MoS2 basal plane would be a highly desirable reaction that preserves the original MoS2 crystal structure, and in contrast to O chemisorption, it is also expected to substantially influence the electronic band structure, enabling a more efficient engineering of its electronic14,16 and optical26 properties.

Here we show that the basal plane of the MoS2 monolayers, when subjected to long-term ambient exposure, spontaneously undergoes such O substitution reactions, giving rise to a highly crystalline 2D molybdenum oxy-sulfide phase.

Results and discussion

Ambient oxidation of the basal plane revealed at the single-atom level

We prepared mechanically exfoliated MoS2 single layers on atomically flat Au(111) substrates using a slightly modified version of a recently developed exfoliation technique27, yielding single layers with lateral dimension of hundreds of micrometres (see Supplementary Section 1 for details). Such large-area samples are characterized by an extremely low edge to surface ratio and grain boundary concentration, as well as a much smaller concentration and variety of intrinsic point defects28, establishing them as an excellent model system for studying the intrinsic chemical properties of the pristine basal plane. The exfoliated MoS2 samples were stored under ambient conditions (air, room temperature and ambient light) for periods of up to 1.5 years. We used scanning tunnelling microscopy (STM) measurements to follow the atomic-level structural changes in the basal plane structure of 2D MoS2 crystals during long-term ambient exposure. So far, optical, scanning electron and atomic force microscopy measurements have mainly been used to monitor the structural changes induced by oxidation in MoS2 layers; however, the spatial resolution of these methods does not allow the detection of single-atom-level modifications. Imaging the atomic-scale structure of the MoS2 basal plane is possible by high-resolution scanning transmission electron microscopy (TEM). However, detecting light atoms such as oxygen with TEM is challenging due to their low contrast and easy knockout29,30. By contrast, STM measurements can detect the unaltered structure of such oxygen defects due to its atomic resolution capability and the low energy of the tunnelling electrons31. Due to the slow oxidation reaction, and the noninvasive nature of the measurements, the structure and number of oxidation-induced defect sites did not change during the STM imaging, enabling us to acquire atomic-resolution snapshots of the oxidation process.

Atomic-resolution STM images of the basal plane of a mechanically exfoliated MoS2 single layer after 1 month and 1 year of ambient exposure are shown in Fig. 1b,c, respectively. The STM images were acquired on the same sample, but at different locations. In contrast to general expectations, STM measurements reveal clear modifications in the atomic structure of the MoS2 basal plane during ambient exposure. Freshly prepared 2D MoS2 crystals contain a native point defect density in the range of 1 × 1011 to 1 × 1012 cm−2 (Supplementary Fig. 2). After a month of ambient exposure our STM measurements revealed the formation of new point defects all over the basal plane, increasing their concentration into the 3 × 1012 to 2 × 1013 cm−2 range (Fig. 1b). The increase in the atomic-scale defect concentration (up to 5 × 1013 to 1 × 1014 cm−2) appears more strikingly after a year of exposure (Fig. 1c), when a substantial area of the sample surface is already covered by such defects, clearly providing evidence of the ability of the ambient exposure to create new defect sites in the basal plane structure of 2D MoS2 crystals. Although the defect density shows some spatial variation, as indicated by the concentration ranges given above, the effect of ambient exposure time is more significant. Higher-resolution STM images (Fig. 1d) shed light on the atomic structure of the defects induced by ambient exposure. Dark triangles in the STM images of 2D MoS2 crystals are characteristic of sulfur atom vacancies31. Within the dark triangles, bright spots can be clearly detected that have not been observed before. These central bright spots are a pronounced feature of the experimental STM images, and their contrast is even stronger than that of S atoms in the MoS2 lattice. The most straightforward explanation attributes the bright spots inside the S vacancies to oxygen atoms or molecules saturating the vacancies under ambient conditions. Among the species present under ambient conditions, O is predicted to be the energetically most favourable to saturate an S vacancy14,16. To confirm this, we performed detailed simulations of the STM images of various O-related defects in 2D MoS2 crystals based on density functional theory (DFT) calculations of their electronic structure (Supplementary Fig. 3). Our theoretical results reveal that the best agreement with the experimental STM data is provided by single O atoms substituting individual S atoms (saturating the S vacancy), giving a dark triangle with a bright spot inside. Such a simulated STM image is shown in the inset of Fig. 1d. Note that, according to both calculations and measurements, the contrast of the O substitutions in STM images is dependent on the tip–sample distance; they only become fully apparent at small tip–sample distances (see Supplementary Section 4 for details). While the agreement of the calculated STM image with experiments is far better for O substitution than for any other calculated defect structure, it is not perfect. This can be attributed partly to the STM image distortions that are unavoidable at room temperature, and partly to the dark ‘halo’ often emerging near single-atomic substitution sites in STM images32,33. The latter can be a signature of an electron depletion zone around a negatively charged defect site34. STM investigations could not directly confirm the chemical identity of the incorporated atoms, and XPS measurements are usually used for such purposes. However, due to the low spatial resolution, it was not possible to selectively measure the single-layer areas, as the exfoliated samples did not homogeneously cover the substrate. Furthermore, the relatively low concentration of the incorporated O atoms (<3 at%) makes such investigations challenging13. To gain more information on the nature of the defects progressively forming under ambient conditions, we also performed Raman spectroscopy (Supplementary Fig. 5) and photoluminescence measurements. Photoluminescence spectra on aged samples revealed a new peak at 1.75 eV that is not present in freshly exfoliated samples (Supplementary Fig. 6). This peak can be clearly associated with vacancy-type defect formation35,36. Moreover, it has been shown that such peak does not emerge under ultrahigh-vacuum conditions; the charge transfer induced by the saturation of vacancies with environmental species (O, N) is necessary36. These findings further support our STM and simulation data for the formation of O-saturated S vacancies.

Fig. 1: 2D MoS2−xOx solid solution crystals by ambient oxidation of MoS2.

a, Schematic atomic structure of the MoS2−xOx solid solution monolayer. b,c, Atomic-resolution STM images (5 mV, 2 nA) of an exfoliated MoS2 single layer after 1 month (b) and 1 year (c) of ambient exposure, revealing a progressive defect formation. d, Higher-resolution STM image displaying the incorporation of O atoms (bright spots) into S vacancies (dark triangles). Inset, simulated STM image based on DFT calculations of an O-saturated S vacancy site in the 2D MoS2 crystal.

The above findings clearly challenge the generally accepted view of an environmentally inert MoS2 basal plane, and evidence that the ambient oxidation of the basal plane spontaneously proceeds through formation of S vacancies and their saturation by O. This oxidation process preserves the original crystal lattice, with Mo sites in a trigonal prismatic configuration, but coordinated by five S atoms and an O atom. The oxidation speed of the MoS2 basal plane under ambient conditions was found to be of the order of 1 atom per minute per μm2. This ultraslow oxidation reaction in principle enables extremely precise control of O concentration in the MoS2 lattice. The slow reaction speed is most likely responsible for stabilizing the oxy-sulfide phase against the otherwise favourable full conversion to MoO3 (ref. 20).

Because the oxidation process is expected to be kinetically limited14,16, by increasing the temperature, the oxidation speed should also increase. We subjected a freshly exfoliated MoS2/Au(111) sample to air inside a furnace heated to 400 K for 1 week. Subsequent STM investigation revealed a similar defect formation as at room temperature, but the O site concentration reached values of 8 × 1012 to 3 × 1013 cm−2 even after a week of exposure (Supplementary Fig. 7). This clearly indicates that the O substitution process has been accelerated by increasing the temperature. Furthermore, this observation also excludes the possibility that the ambient oxidation occurs due to the accidental presence of some more reactive O species (for example, ozone, superoxide or singlet oxygen). Increasing the temperature from 21 to 127 °C is not expected to significantly increase the formation of such reactive species. The accelerated O substitution process at higher temperatures also provides a more feasible route for the synthesis of oxy-sulfide crystals.

The substitutional oxidation of the MoS2 basal plane yields a novel 2D MoS2−xOx solid solution crystal phase (Fig. 1a).The formation of this MoS2−xOx solid solution phase has already been proposed for sputter-deposited MoS2 films37,38,39. However, previously, the structure of the synthesized molybdenum oxy-sulfide films was found to be amorphous or highly disordered, with poor long-range crystalline order40,41. Here we provide clear evidence for the formation of a highly crystalline 2D molybdenum oxy-sulfide phase and thus a strategy for synthesizing new materials through chemical transformation of 2D crystals that are difficult to synthesize by other methods.

The STM investigations of 2D MoSe2 crystals (prepared and aged under the same condition as the 2D MoS2 crystals) revealed strikingly different behaviour following long-term ambient exposure. Figure 2 presents atomic-resolution STM images of freshly exfoliated and 1-year-old MoSe2 single layers. The basal plane of the 2D MoSe2 crystal does not show significant changes, even after a year-long ambient exposure. Native defects (Se vacancies and chemisorbed adatoms—see inset of Fig. 2a) are present with a relatively low (1 × 1011 to 1 × 1012 cm−2) density, but their concentration does not increase during ambient exposure.

Fig. 2: Stability of 2D MoSe2 basal plane under ambient conditions.

a,b, Atomic-resolution STM images (5 mV, 1 nA) of 2D MoSe2 crystals on Au(111) substrate, freshly prepared (a) and after 1 year of ambient exposure (b), revealing a remarkable stability of the basal plane under ambient conditions. Insets in a higher-resolution STM images and corresponding line cuts of bright (chemisorbed adatom) and dark (Se vacancy) defect sites. The number of defects does not increase with ambient exposure time.

Energetics and kinetics of O substitution

Based on the existing literature, it is difficult to interpret the striking experimentally observed differences in the ambient oxidation behaviour of 2D MoS2 and MoSe2 basal planes. However, most theoretical works so far have focused on the chemisorption mechanism. Indeed, our experimental results confirm that the pristine basal planes of both MoS2 and MoSe2 single layers are stable against chemisorption or MoO3 transformation. We found that O substitution proceeds in the case of MoS2, but not for MoSe2. In contrast to chemisorption, the O substitution mechanism implies the removal of S (Se) atoms during the oxidation process. To investigate the thermodynamics of this process, we performed DFT calculations (see Methods for details) regarding the removal of S (Se) atoms from the basal plane of the 2D MoS2 (MoSe2) crystals through oxidation.

First, we investigated the reaction of the surface S atoms of 2D MoS2 crystals with oxygen, yielding volatile SO2 species and creating a sulfur vacancy in the MoS2 basal plane (Fig. 3a). The enthalpy of the reaction was calculated using the formula ∆E = E (reactants) − E (products) = −0.49 eV. Because the enthalpy of the final structures is lower than for the initial structures, this implies that S atom removal by oxidation is thermodynamically favourable in the case of MoS2. By contrast, our DFT calculations show that removal of a Se atom through oxidation of the MoSe2 basal plane has a positive oxidation enthalpy of ∆E = +0.75 eV (Fig. 3b). This result indicates the thermodynamically unfavourable nature of selenium oxide formation, which hinders the formation of Se vacancies through oxidation and hence the substitutional oxidation of the MoSe2 basal plane, in agreement with the experimental observations.

Fig. 3: Energetics and kinetics of O substitution in the MoS2 and MoSe2 basal plane.

a,b, The process of chalcogenide atom vacancy formation through oxidation is characterized by a negative oxidation enthalpy (ΔE) for the defect-free 2D MoS2 basal plane (a) and a positive enthalpy for MoSe2, protected from oxidation by the endothermic nature of SeO2 formation (b). c,d, Kinetic energy barriers calculated by the NEB model for 2D MoS2 (c) and 2D MoSe2 (d) crystals reveal substantially lower barriers for MoS2 of ~1 eV height that can be overcome even at room temperature on a months-long timescale.

Although thermodynamically favourable, substitutional oxidation of the MoS2 basal plane is still expected to face kinetic barriers. To investigate this, we performed detailed nudged elastic band (NEB) model calculations42 to determine the energy of the transitional states and to find the potential barriers for the substitutional oxidation process. The results displayed in Fig. 3c show that, in the case of MoS2, the typical kinetic barrier height is ~1 eV for the proposed reaction pathway. Note that, with O saturation of the S vacancies the energy of the final state becomes ~4 eV lower (Supplementary Fig. 3), indicating the highly favourable nature of the O substitution process as a next step, in accordance with experimental findings. By contrast, the barriers for the substitutional oxidation of MoSe2 basal plane are about 1.5 eV, with a thermodynamically unfavourable final state (Fig. 3d). Consequently, both energetics and kinetics support our experimental findings regarding the differences between the oxidation of the 2D MoS2 and MoSe2 basal planes, reflecting the different oxidation mechanisms of S- and Se-based compounds. Furthermore, the kinetic barriers of ~1 eV for the substitutional oxidation of MoS2 are predicted to be surmounted at room temperature over a timescale of a month, according to transition state theory43.

Although the basal plane of the MoSe2 appears more stable, the overall environmental stability of the crystals is also dependent on their edge oxidation speed. We found that the oxidation proceeds much faster at the edges of MoSe2 single layers than on the MoS2 edges (Supplementary Figs. 8 and 9). Adsorbed contaminations are naturally present on both MoS2 and MoSe2 surfaces subjected to ambient conditions. However, their role in the ambient oxidation process is expected to be more pronounced at edges and grain boundaries13,15, as they are known to preferentially attach to these high-energy sites instead of the pristine basal plane.

Although structural disorder induced by invasive oxidation is unlikely to be fully reversible, the substitutional oxidation of the MoS2 basal plane does not damage the crystal lattice, so it can in principle be suitable for a fully reversible reduction. Indeed, we found that simple annealing of MoS2−xOx crystals in a H2S atmosphere at 200 °C for 30 min is able to fully restore the atomic structure of the pure 2D MoS2 phase. Two representative atomic-resolution STM images of the MoS2 basal plane before and after reduction (Fig. 4) clearly show that 2D MoS2−xOx solid solution crystals can be reduced to the pure, almost defect-free MoS2 phase. The feasibility of such a reduction process is also supported by our DFT calculations (Supplementary Fig. 10). The atomically perfect reduction of the oxidized 2D MoS2−xOx crystals enables a highly efficient engineering of their surface chemistry.

Fig. 4: Reduction of 2D MoS2−xOx to pristine MoS2.

a,b, Representative atomic-resolution STM images (5 mV, 2 nA) of 2D MoS2−xOx before (a) and after (b) 30 min annealing at 200 °C in H2S atmosphere, showing the atomically perfect reduction of the oxy-sulfide solid solution to the pure MoS2 phase through resubstitution of the single atomic O sites by S atoms.

Catalytic activity of 2D MoS2−xOx crystals towards hydrogen evolution

To investigate how the O substitution sites change the properties of MoS2 single layers, we investigated the catalytic activity of 2D MoS2−xOx crystals for the electrochemical hydrogen evolution reaction (HER) (see Methods for experimental details). Polarization curve (IE) measurements were performed on both MoS2−xOx and MoS2 single layers. To make the comparison more direct and relevant, we measured a 1-year-old MoS2−xOx flake before and after its reduction to the pure 2D MoS2 phase. This way, we could measure the catalytic activity of the same flake both with and without O substitution sites in the basal plane, as revealed by our STM characterization. The measured polarization curves and corresponding Tafel plots shown in Fig. 5 show a highly increased catalytic HER activity for the 2D MoS2−xOx solid solution crystals compared to the reduced pure 2D MoS2 phase. This enhanced catalytic activity can be clearly related to the presence of substitutional O sites, the only structural feature that displays a correlation with ambient exposure time, according to atomic-level structural STM data. We also investigated the stability of the 2D MoS2−xOx crystals during 1,000 catalytic cycles. After an initial slight decrease by ~10%, the current density stabilizes, indicating the good long-term stability of the catalytic process (Supplementary Fig. 11). Confocal Raman microscopy maps revealed that the 2D MoS2−xOx layer is still continuous after 1,000 cycles, and the measured Raman spectra remained practically unaltered (Supplementary Fig. 12), while atomic-resolution STM measurements confirmed that the O substitution sites are present after the catalytic process (Supplementary Fig. 11).

Fig. 5: Catalytic activity of 2D MoS2−xOx for hydrogen evolution.

a,b, Linear sweep voltammogram curves (a) and corresponding Tafel plots (b) for Au substrate, MoS2 single layer, MoS2−xOx single layer (1 year old) and Pt substrate, revealing a significantly higher catalytic activity of the 2D oxy-sulfide phase as compared to the pure MoS2 phase. The increased catalytic activity can be attributed to single atomic O sites progressively incorporating into the MoS2 basal plane during ambient exposure. Dashed lines in b correspond to linear fits of the low-current segments of Tafel plots, used for calculating their slopes.

A widely used parameter for predicting the catalytic activity of various sites is the H adsorption Gibbs free energy (ΔGH). We calculated ΔGH for O substitution sites and compared it to that of pristine MoS2 surface (see Supplementary Section 10 for calculation details). We found that ΔGH is lowered to about half on the O sites (+1.2 eV) compared to the S sites (+2.2 eV) of the basal plane, which means that hydrogen is much more probably absorbed on the O sites. However, even the substantially reduced ΔGH value of ~1 eV is still well above zero, indicating a not very favourable H adsorption. Nevertheless, other effects—not captured by ΔGH—can also play an important role in the catalytic activity. Because O atoms have a similar electron configuration to S atoms, it is often assumed that the O substitutions are fully passivating the S vacancies without significantly altering the electronic structure. While O substitutions indeed remove the midgap states characteristic of S vacancies44, they still substantially change the orbital composition of the valence and conduction bands (Supplementary Section 11), as well as electron affinity at the substitution sites. The electronegativity and electron affinity of the dopants (including O substitutions) have been shown to play an important role in defining the catalytic activity of graphene45,46. To evaluate this effect for the O substitution sites of 2D MoS2 crystals, we performed Bader charge analysis (see Supplementary Section 12 for details), which has proven useful for understanding the catalytic activity of sites where charge transfer plays an important role47. The Bader analysis indicated a strong acceptor-type behaviour of the O substitution sites, characterized by almost two times higher electron affinity (−0.88 e) as compared to S atoms (−0.47 e). The locally increased electron affinity combined with the experimentally measured overall n-doping of the MoS2 crystals (Supplementary Fig. 15) can give rise to localized negative charges on the O substitution sites. These partially screened negative charges can substantially facilitate the adsorption of positively charged H species from the acidic electrolyte. The effect of such charged dopants is not included in the ΔGH calculations, as it is challenging to treat charged impurities at the DFT level48. We therefore propose that the combined effects of the substantially decreased ΔGH and increased electronegativity at the O sites—both facilitating the absorption of H species—can give rise to the increased catalytic activity observed in 2D MoS2−xOx crystals. Understanding the catalytic process is a highly challenging task, largely due to the complexity of the experimentally investigated systems, in strong contrast with the idealized theoretical models49. 2D MoS2−xOx provides an ideal model system for understanding the atomic-level relations between active sites and catalytic HER activity, as it is characterized by a single type of active site, with experimentally known atomic structure, while edges and previously investigated more disordered MoS2 structures can host a variety of active sites with complex atomic configurations and often little experimental insight into their precise atomic nature. These findings clearly show that the substitutional oxidation process of the MoS2 basal plane reported here can open new routes for engineering 2D electrocatalysts with single O-atom active sites of a much higher site density than previously achieved for individual hetero-atom catalysts50.


Sample preparation

The investigated MoS2 (and MoSe2) single layers were prepared by mechanical exfoliation of bulk MoS2 (MoSe2) synthetic crystals of high structural quality (2DSemiconductors). We employed a slightly modified version (Supplementary Section 1) of the mechanical exfoliation technique developed by us and discussed in details in ref. 27, providing single-layer TMDC flakes with hundreds of micrometres lateral dimensions on atomically flat Au(111) surfaces. The single-layer nature of the investigated MoS2 and MoSe2 crystals was also confirmed by Raman spectroscopy. The samples were stored under ambient laboratory conditions, in air, at room temperature and under ambient light conditions.


Atomic-resolution STM measurements were performed on a Nanoscope E STM operating under ambient conditions. Tunnelling spectroscopy data were measured in an RHK PanScan STM under UHV conditions at room temperature. Large-area flakes could be easily identified under an optical microscope, with guided landing of the STM tip on MoS2 single layers. Atomic-resolution imaging was carried out under ambient conditions with typical parameter ranges of |Ubias| = 5–50 mV and Itunnel = 1–3 nA. Although the resolution of the hexagonal lattice of the top layer of S atoms can be routinely achieved, the atomic resolution of individual point defects is challenging. Defects were clearly resolved when imaging at bias voltages (energies) within the bandgap. This is possible due to the influence of the Au(111) substrate, which induce a small but finite density of states within the bandgap (confirmed by our tunnelling spectroscopy measurements), conferring a weak metallic character to the supported MoS2 single layers. Raman measurements were conducted on a confocal Raman microscope (Witec 300RSA) with λ = 532 nm and W = 1 mW. During Raman measurements the spectra of the sample did not change detectably.

Electrochemical measurements

Electrochemical measurements were carried out on selected sample areas (0.4–0.8 mm diameter) of a single 2D MoS2 flake supported by a 100-nm-thick Au(111) film on a glass substrate. Control measurements on the same Au(111) substrate as well as a Pt plate were conducted in the same experimental configuration. To compare the performance of different samples in HER, linear sweep voltammetry was performed in a three-electrode configuration using a 0.5 M sulfuric acid electrolyte at room temperature. Ag/AgCl and Pt wire were used as counter and reference electrodes, respectively. Potential sweeps were acquired at a scan rate of 2 mV s−1 using a Bio-Logic SP-150 potentiostat.

Computational details

All calculations were performed on TMDC single layers in the framework of spin-polarized DFT theory implemented in the VASP software package, using the plane-wave basis set and projector augmented wave method. Exchange-correlation effects were taken into account in the framework of the local density approximation (LDA) and generalized gradient approximation (GGA) by the Perdew–Burke–Ernzerhof (PBE) functional. Defects were modelled in periodically repeated 8 × 8 supercells. The Brillouin zone of the supercells was sampled with a (2 × 2 × 1) Monkhorst–Pack mesh of k-points for geometry optimization and (4 × 4 × 1) for STM image calculations. The cutoff energy for the plane-wave basis set was set to 400 eV. To avoid artificial interactions between periodic replicas of low-dimensional nanoclusters, a vacuum interval of 15 Å was introduced in all supercell geometry. The NEB method was applied to model transitional states and find the potential barriers of MoS2 and MoSe2 oxidation.

Data availability

The data supporting the findings of this study are available within the Article and its Supplementary Information files. All other relevant source data are available from the corresponding author upon request.


  1. 1.

    Liu, L. et al. Graphene oxidation: thickness-dependent etching and strong chemical doping. Nano Lett. 8, 1965–1970 (2008).

    CAS  Article  Google Scholar 

  2. 2.

    Luo, Z. et al. Thickness-dependent reversible hydrogenation of graphene layers. ACS Nano 3, 1781–1788 (2009).

    CAS  Article  Google Scholar 

  3. 3.

    Chhowalla, M. et al. The chemistry of two-dimensional layered transition metal dichalcogenide nanosheets. Nat. Chem. 5, 263–275 (2013).

    Article  Google Scholar 

  4. 4.

    Jaramillo, T. F. et al. Identification of active edge sites for electrochemical H2 evolution from MoS2 nanocatalysts. Science 317, 100–102 (2007).

    CAS  Article  Google Scholar 

  5. 5.

    Lauritsen, J. V. et al. Hydrodesulfurization reaction pathways on MoS2 nanoclusters revealed by scanning tunneling microscopy. J. Catal. 224, 94–160 (2004).

    CAS  Article  Google Scholar 

  6. 6.

    Lauritsen, J. V. et al. Size-dependent structure of MoS2 nanocrystals. Nat. Nanotech. 2, 53–58 (2007).

    CAS  Article  Google Scholar 

  7. 7.

    Dreyer, D. R., Park, S., Bielawski, C. W. & Ruoff, R. S. The chemistry of graphene oxide. Chem. Soc. Rev. 39, 228–240 (2010).

    CAS  Article  Google Scholar 

  8. 8.

    Mirabelli, G. et al. Air sensitivity of MoS2, MoSe2, MoTe2, HfS2, HfSe2. J. Appl. Phys. 120, 125102 (2016).

    Article  Google Scholar 

  9. 9.

    Yue, R. et al. HfSe2 films: 2D transition metal dichalcogenides grown by molecular beam epitaxy. ACS Nano 9, 474–480 (2015).

    CAS  Article  Google Scholar 

  10. 10.

    Lee, C. H. et al. Tungsten ditelluride: a layered semimetal. Sci. Rep. 5, 10013 (2015).

    CAS  Article  Google Scholar 

  11. 11.

    Li, Y., Zhou, Z., Zhang, S. & Chen, Z. MoS2 nanoribbons: high stability and unusual electronic and magnetic properties. J. Am. Chem. Soc. 130, 16739–16744 (2008).

    CAS  Article  Google Scholar 

  12. 12.

    Grønborg, S. S. et al. Synthesis of epitaxial single-layer MoS2 on Au(111). Langmuir 31, 9700–9706 (2015).

    Article  Google Scholar 

  13. 13.

    Gao, J. et al. Aging of transition metal dichalcogenide monolayers. ACS Nano 10, 2628–2635 (2016).

    CAS  Article  Google Scholar 

  14. 14.

    Longo, R. C. et al. Intrinsic air stability mechanisms of two-dimensional transition metal dichalcogenide surfaces: basal plane versus edge oxidation. 2D Mater. 4, 025050 (2017).

    Article  Google Scholar 

  15. 15.

    Martincova, J., Otyepka, M. & Lazar, P. Is single layer MoS2 stable in the air? Chem. Eur. J. 23, 13233–13239 (2017).

    CAS  Article  Google Scholar 

  16. 16.

    Santosh, K. C., Longo, R. C., Wallace, R. M. & Cho, K. Surface oxidation energetics and kinetics on MoS2 monolayer. J. Appl. Phys. 117, 135301 (2015).

    Article  Google Scholar 

  17. 17.

    Rao, R., Islam, A. E., Campbell, P. E., Vogel, E. M. & Marujama, B. In situ thermal oxidation kinetics in few layer MoS2. 2D Mater. 4, 025058 (2017).

    Article  Google Scholar 

  18. 18.

    Bonde, J., Moses, P. G., Jaramillo, T. F., Norskov, J. K. & Chorkendorff, I. Hydrogen evolution on nano-particulate transition metal sulfides. Faraday Discuss. 140, 219–223 (2009).

    Article  Google Scholar 

  19. 19.

    Angelica, A. et al. HfO2 on UV–O3 exposed transition metal dichalcogenides: interfacial reactions study. 2D Mater. 2, 014004 (2015).

    Article  Google Scholar 

  20. 20.

    Walter, T. N., Kwok, F., Simchi, H., Aldosari, H. M. & Mohney, S. E. Oxidation and oxidative vapor-phase etching of few-layer MoS2. J. Vac. Sci. Technol. 35, 021203 (2017).

    Article  Google Scholar 

  21. 21.

    Jaehyun, J. et al. Improved growth behavior of atomic-layer-deposited high-k dielectrics on multilayer MoS2 by oxygen plasma pretreatment. ACS Appl. Mater. Inter. 5, 4739–4744 (2013).

    Article  Google Scholar 

  22. 22.

    Pingli, Q. et al. In situ growth of double-layer MoO3/MoS2 film from MoS2 for hole-transport layers in organic solar cell. J. Mater. Chem. A 2, 2742–2756 (2014).

    Article  Google Scholar 

  23. 23.

    Voiry, D. et al. Conducting MoS2 nanosheets as catalysts for hydrogen evolution reaction. Nano. Lett. 13, 6222 (2013).

    CAS  Article  Google Scholar 

  24. 24.

    Juanfeng, X. et al. Controllable disorder engineering in oxygen-incorporated MoS2 ultrathin nanosheets for efficient hydrogen evolution. J. Am. Chem. Soc. 135, 17881–17888 (2013).

    Article  Google Scholar 

  25. 25.

    Liu, X. et al. Insight into the structure and energy of Mo27SxOy clusters. RSC Adv. 7, 9513–9520 (2017).

    CAS  Article  Google Scholar 

  26. 26.

    Shu, H., Li, Y., Niu, X. & Wang, J. Greatly enhanced optical absorption of a defective MoS2 monolayer through oxygen passivation. ACS Appl. Mater. Interfaces 8, 13150–13156 (2016).

    CAS  Article  Google Scholar 

  27. 27.

    Magda, G. Z. et al. Exfoliation of large-area transition metal chalcogenide single layers. Sci. Rep. 5, 14714 (2015).

    CAS  Article  Google Scholar 

  28. 28.

    inhua, H. et al. Exploring atomic defects in molybdenum disulphide monolayers. Nat. Commun. 6, 6293 (2015).

    Article  Google Scholar 

  29. 29.

    Krivanek, O. L. et al. Atom-by-atom structural and chemical analysis by annular dark field electron microscopy. Nature 464, 571–574 (2010).

    CAS  Article  Google Scholar 

  30. 30.

    Komsa, H. P. et al. Two-dimensional transition metal dichalcogenides under electron irradiation: defect production and doping. Phys. Rev. Lett. 109, 035503 (2012).

    Article  Google Scholar 

  31. 31.

    Vancsó, P. et al. The intrinsic defect structure of exfoliated MoS2 single layers revealed by scanning tunneling microscopy. Sci. Rep. 6, 29726 (2016).

    Article  Google Scholar 

  32. 32.

    Nagl, C., Haller, O., Platzgummer, E., Scmid, M. & Varga, P. Submonolayer growth of Pb on Cu (111): surface alloying and de-alloying. Surf. Sci. 321, 237–248 (1994).

    CAS  Article  Google Scholar 

  33. 33.

    Li, Z. et al. Spontaneous doping of two-dimensional NaCl films with Cr atoms: aggregation and electronic structure. Nanocale 7, 2366 (2015).

    CAS  Google Scholar 

  34. 34.

    Bampoulis, P. et al. Defect dominated charge transport and Fermi level pinning in MoS2/metal contacts. ACS Appl. Mater. Interfaces 9, 19278–19286 (2017).

    CAS  Article  Google Scholar 

  35. 35.

    Chow, P. K. et al. Defect induced photoluminescence in monolayer semiconducting transition metal dichalcogenides. ACS Nano 9, 1520–1527 (2015).

    CAS  Article  Google Scholar 

  36. 36.

    Tongay, S. et al. Defects activated photoluminescence in two-dimensional semiconductors: interplay between bound, charged and free excitons. Sci. Rep. 3, 2657 (2013).

    Article  Google Scholar 

  37. 37.

    Lince, J. R. Mo2−xOx solid solutions in thin films produced by rf-sputter-deposition. J. Mater. Res. 5, 218–222 (1990).

    CAS  Article  Google Scholar 

  38. 38.

    Lince, J. R., Hilton, M. R. & Bommannavar, A. S. Oxygen substitution in sputter deposited MoS2 films studied by extended X-ray absorption fine-structure, X-ray photoelectron spectroscopy and X-ray diffraction. Surf. Coat. Technol. 43–44, 640–651 (1990).

    Article  Google Scholar 

  39. 39.

    Fleischauer, P. D. & Lince, J. R. A comparison of oxidation and oxygen substitution in MoS2 solid film lubricants. Tribol. Int. 32, 627–636 (1999).

    CAS  Article  Google Scholar 

  40. 40.

    Benoist, L. et al. X-ray photoelectron spectroscopy characterization of amorphous molybdenum oxysulfide thin films. Thin Solid Films 258, 110–114 (1995).

    CAS  Article  Google Scholar 

  41. 41.

    Sung, H. S. et al. Bandgap widening of phase quilted, 2D MoS2 by oxidative intercalation. Adv. Mater. 27, 3152–3158 (2015).

    Article  Google Scholar 

  42. 42.

    Henkelman, G., Uberuaga, B. P. & Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 113, 9901–9904 (2000).

    CAS  Article  Google Scholar 

  43. 43.

    Nan, H. et al. Strong photoluminescence enhancement of MoS2 through defect engineering and oxygen bonding. ACS Nano 8, 5738–5745 (2014).

    CAS  Article  Google Scholar 

  44. 44.

    Hong, L. et al. Activating and optimizing MoS2 basal planes for hydrogen evolution through the formation of strained sulphur vacancies. Nat. Mater. 15, 48–53 (2016).

    Article  Google Scholar 

  45. 45.

    Xia, Z. Hydrogen evolution guiding principles. Nat. Energy 1, 16155 (2016).

    Article  Google Scholar 

  46. 46.

    Jiao, Y., Zheng, Y., Davey, K. & Qiao, S. Z. Activity origin and catalyst design principles for electrocatalytic hydrogen evolution on hetero-atom-doped graphene. Nat. Energy 1, 16130 (2016).

    CAS  Article  Google Scholar 

  47. 47.

    Yang, N., Zheng, X., Li, L., Li, J. & Wei, Z. Influence of phosphorus configuration on electronic structure and oxygen reduction reactions of phosphorus-doped graphene. J. Phys. Chem. C 121, 19321 (2017).

    CAS  Article  Google Scholar 

  48. 48.

    Komsa, H. P., Berseneva, N., Krashenninikov, A. & Nieminen, R. M. Charged point defects in the flatland: accurate formation energy calculations in two-dimensional materials. Phys. Rev. X 4, 031044 (2014).

    Google Scholar 

  49. 49.

    Su, Y., Gao, S., Lei, F. & Xie, Y. Atomically-thin two-dimensional sheets for understanding active sites in catalysis. Chem. Soc. Rev. 44, 623–636 (2015).

    Article  Google Scholar 

  50. 50.

    Qiao, B. et al. Single-atom catalysis of CO oxidation using Pt1/FeOx. Nat. Chem. 3, 634–641 (2011).

    CAS  Article  Google Scholar 

Download references


This work was performed in the framework of a NanoFab2D ERC starting grant, H2020 Graphene Core2 project no. 785219 and the Korea Hungary Joint Laboratory for Nanosciences. L.T. acknowledges OTKA grant K108753 and the ‘Lendület’ programme. The work was also supported by a VEKOP-2.3.2-16-2016-00011 grant, supported by the European Structural and Investment Funds. Z.I.P. and P.B.S. acknowledge financial support from the Ministry of Education and Science of the Russian Federation in the framework of the Increase Competitiveness Program of NUST ‘MISIS’ (no. K2-2017-001). Z.I.P. and P.B.S. acknowledge the supercomputer cluster provided by the Materials Modelling and Development Laboratory at NUST “MISIS” (supported via a grant from the Ministry of Education and Science of the Russian Federation no. 14.Y26.31.0005), and the Information Technology Centre of Novosibirsk State University for providing access to the cluster computational resources. Z.I.P. acknowledges the financial support of the Russian Scientific Foundation according to research project no. 18-73-10135 for stability calculations. P.V. acknowledges the Plateforme Technologique de Calcul Intensif (PTCI), which was supported by the FRS-FNRS under convention no. 2.5020.11. P.B.S. acknowledges financial support from the RFBR, via research project no. 16-32-60138 mol_а_dk. The authors thank J. S. Pap for useful discussions on electrochemistry.

Author information




L.T. conceived and designed the experiments. J.P. and G.Z.M. prepared the samples and performed the STM measurements. T.O. performed the chemical reduction and electrocatalytic experiments. Z.I.P., P.V. and P.B.S. performed theoretical calculations. G.D. and G.Z.M. conducted Raman and photoluminescence investigations. L.T., P.B.S. and C.H. supervised the project. L.T. wrote the paper. All authors discussed the results and commented on the manuscript.

Corresponding author

Correspondence to Levente Tapasztó.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information

Supplementary Figures 1–15, Supplementary Methods, Supplementary Characterization

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Pető, J., Ollár, T., Vancsó, P. et al. Spontaneous doping of the basal plane of MoS2 single layers through oxygen substitution under ambient conditions. Nature Chem 10, 1246–1251 (2018). https://doi.org/10.1038/s41557-018-0136-2

Download citation

Further reading


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing