Nanoscale redox mapping at the MoS2-liquid interface

Layered MoS2 is considered as one of the most promising two-dimensional photocatalytic materials for hydrogen evolution and water splitting; however, the electronic structure at the MoS2-liquid interface is so far insufficiently resolved. Measuring and understanding the band offset at the surfaces of MoS2 are crucial for understanding catalytic reactions and to achieve further improvements in performance. Herein, the heterogeneous charge transfer behavior of MoS2 flakes of various layer numbers and sizes is addressed with high spatial resolution in organic solutions using the ferrocene/ferrocenium (Fc/Fc+) redox pair as a probe in near-field scanning electrochemical microscopy, i.e. in close nm probe-sample proximity. Redox mapping reveals an area and layer dependent reactivity for MoS2 with a detailed insight into the local processes as band offset and confinement of the faradaic current obtained. In combination with additional characterization methods, we deduce a band alignment occurring at the liquid-solid interface.

Remarks to the Author: Summary: The authors studied the relationship between the work function and observed electrochemical reactivity of MoS2 using several scanning probe techniques (viz, SECM, AFM, STS, STM, and KPFM) with a ferrocene (Fc) redox mediator. Their main results are that multilayers exhibit a higher reactivity than monolayers due to the lower conduction band potential of the multilayers and that bipolar electrochemical activity is proportional to flake size. The data demonstrate clearly a difference in electrochemical reactivity related to the number of MoS2 layers. However, the analysis and interpretation of the SECM data as well as ambiguous experimental conditions are not convincing. This paper has the potential to be of interest to researchers in a variety of fields, but the paper's contribution to the body of knowledge on TMDCs is not apparent, as the main results of the paper are not discussed in comparison to the literature on TMDCs. The novelty of AFM-SECM (Eifert and Kranz, "Hyphenating Atomic Force Microscopy," Anal Chem 2014, 86, 5190-5200) and observation of layer number-dependent electrochemical activity and band gap (Velicky and Toth, "From two-dimensional materials to their heterostructures: An electrochemist's perspective," Appl Mater Today 2017, 8, 68-103) of MoS2 is unclear.
Specific comments and questions: -Abstract is misleading. Photocatalytic and water splitting properties of MoS2 should be deemphasized, as the authors do not present any photocatalytic results or relate their results to the photoactivity of MoS2. If water splitting is of interest, why was organic solution used? What are "local processes" (line 24)?
-What does "near field" (lines 22 and 54) mean? Do the authors mean small SECM probe-sample distances? Similarly, the description "large distance above the sample" in line 47 is vague. What is a large distance? SECM measurements are done with probe-sample distances less than the probe radius size.
-Instead of the focus on AFM-SECM (see, for example, the paragraph starting on line 53) and achieving high resolution, what may be more interesting in this paper is the ability to compare directly local work function and electrochemical activity at the same location on the same MoS2 sample. However, this capability is not apparent in the paper as written.
-Several experimental conditions need to be clarified. (a) Supporting electrolyte and solvent need to be verified. Did the authors mean tetrabutylammonium perchlorate (TBAP), which is a more common electrolyte, instead of tetrabutylammonium perruthenate (TBAP) (line 300)? If perruthenate, an oxidizing agent, was used, would it oxidize the redox mediator or MoS2? Polypropylene carbonate (line 301) is a thermoplastic material and not a liquid solvent. (b) SECM probe dimensions (lines 303-304) are the same as in Ref 20 (Michael et al), and I assume that the authors used similar, commercially available probes. However, cyclic voltammograms (CVs) of the probe in the redox mediator would be helpful as well as an approach curve (i.e., plot of probe current vs. probe-sample distance) of the probe over the Si/SiO2 surface to confirm the probe dimensions used in this paper. Additional characterization such as SEM would also be helpful. (c) Is the insulating coating on the probes stable in the electrolyte solution? (d) What was probe step size and scan rate in AFM and SECM images? How long did each image take? Can the solvent evaporate within the duration of the measurements? (e) Lift mode in lines 47 and 295-296 as well as how AFM and SECM measurements are performed in this paper need to be better explained. Is SECM probe current measured in the main mode, where the probe is in direct contact with the sample, as shown in Figure S3a (line 297)? In my understanding of PeakForce SECM, two steps are involved: (i) main mode (tapping) is AFM to determine topography and (ii) lift mode is SECM to determine electrochemical reactivity with a constant probe-sample distance based on AFM profile. How was SECM probe-sample distance of 100 nm achieved and chosen for this paper? (f) Why was SECM current normalized with respect to negative feedback at Si/SiO2 instead of current at semi-infinite conditions, as is done typically in SECM measurements, and how was this normalization done? "cell geometry related background" in line 79 needs to be explained. (g) For Figure S4, experimental details (e.g., sweep rate, electrode area/size, concentration of mediator) for the CVs are missing. I assume that the CVs are for the SECM probe but don't understand the decaying, diffusion-limited current. I would expect steady-state current (i.e., sigmoid shape and independent of potential) for ultramicroelectrodes such as SECM probes. (h) Experimental details, for example bias current and scan rate for STM data and frequency used for Raman imaging in Figure S6, need to be given. Figure captions in the SI need more detail.
-Fc passivated the probe within a few hours, rendering quantitative comparison between samples difficult (SI section 1). Could impurities in the Fc or solvent absorb on the probe or sample? Did other mediators (e.g., DmFc) exhibit similar, passivating behavior? -How was the Ag QRE potential determined with respect to the SHE ( Figure S4)? Is the potential of the Ag QRE expected to be 0.00 V vs. SHE? Is the QRE potential stable during the measurements?
-The interpretation of the SECM needs to be reevaluated. The statement "the distance limits the maximum achievable resolution and, as resistance for the ion flow increases, the current sensitivity" in lines 49-50 is confusing. SECM probe electrode size determines spatial resolution. I agree with the authors that SECM probe current is affected by probe-sample distance, but this current arises from Faradaic processes (i.e., charge transfer, electron transfer in this case, across the electrode-electrolyte interface) between the probe and redox mediator, not ion flow. For electron transfer to occur, the mediator needs to diffuse to the probe. In negative feedback (such as over an insulator like SiO2), current decreases at small probe-sample distances because mediator diffusion to the probe is hindered by the sample. In positive feedback (such as over a conductor), current increases due to increased flux of mediator generated at the sample. Commonly, feedback current is analyzed using models for negative and positive feedback(e.g., LeFrou and Cornut, "Analytical Expressions for Quantitative Scanning Electrochemical Microscopy (SECM)," ChemPhysChem 2010, 11, 547-556). The authors mention positive/negative feedback in lines 89-92 and 308-316 (although I don't understand why Ref 26 is cited here) but do not apply this analysis to their results. If ion flow does play a role in the SECM current, would current/resistance vary with ionic strength? -I agree with the authors that resistance in the sample can affect the observed SECM current (e.g., by altering the apparent HET kinetics), but how sample resistance relates to heterogeneous electron transfer (HET) of Fc at MoS2 needs to be better explained as well as how well the author's model compares with the SECM literature on unbiased samples of finite size (e.g., Ref. 27: Wipf andBard, J Electrochem Soc 1991, Ref 28: Amemiya, Anal. Chem. 2007). The discussion of resistance throughout the main text and SI (section S2) is confusing. Is resistance due to the sample or due to electron transfer between the MoS2 and the electrolyte? What gives rise to this resistance? In line 19 of the SI, where does the contact resistance originate? MoS2 is electrically isolated from the circuitry (line 304 of the main text). Units need to be added to figure S2. Numbers need to be checked in the SI. On line 17, the area of the reference resistance is (15 mm)2 and in the equation on line 43, it is 15 um2. Line 18 has 3 um2 (is it supposed to be 3 um on a side?).
-In the discussion, results need to be compared to those in literature. How do band gaps, fermi level, and work functions compare with reported values? Can you determine the HET rate constant of Fc and dmFc, and how does it compare with other mediators and reported values? -In Figure 2e, 3 um flakes are not observed with SECM, as mentioned in the main text (line 108) and SI (line 7). Could the flakes detach from the Si/SiO2? Was AFM/optical imaging done after SECM to confirm that the flakes were still attached? -On line 115, what area is used to normalize the feedback current (8fA/nm2)? -Why does the resolution change based on the number of monolayers (see lines 140, 145-146)? I would expect the same resolution for similar SECM parameters and probe size regardless of sample features, as the resolution is determined by the SECM probe radius.
-I disagree with assuming that the SECM probe is a point-like source (line 147). Theoretical models are available for analyzing a conical SECM probe of finite size.
-The following needs to be clarified. (a) Lines 109-111: The statement beginning "AFM resolution is of the order of 100 nm ( Fig. 2(d)), which is image size but not technically limited, in comparison to AFM-SECM resolution of ~2.5 μm.
i.e. on the length scale beyond 100 nm." (b) Lines 119-120: "strong correlation between flake area and feedback is dominantly controlled by the ion flow, i.e. the product of ion conductivity and oxidation rate towards the recharge area (Region II)." (c) Line 148-149: The statement beginning "All dimensions of the used ultra-microelectrode probe are on the 0.1" (d) Discussion on lines 154-157, particularly which conductivity is being determined and how d-3 is determined.
-The statements "When the probe is located over an insulating surface like SiO2, the feedback is driven by the charge flow only towards the more distant, continuously recharged monolayer" (lines 158-159) and "Moreover, a detectable feedback current is still present at a distance of μm from the flow towards a distant MoS2 island "(lines 272-273) are misleading. The probe reaction (oxidation of Fc to Fc+) is driven by potential applied by the potentiostat and not a bipolar reaction with a distant (how distant?) MoS2 flake. The distant flake is on an insulator (SiO2) and is unable to deliver charge (electrons) through the sample to the area under the probe. For similar reasons, I disagree with interpreting the paper's results based on Ref. 24 (Amatore et al.). In their experiment, the distant electrode is electrically connected to the sample area under the probe and thus can act like a bipolar electrode.
-Could the enhanced positive feedback current for the center of the larger area flakes in figure 3g (line 167) be due to the large density of bilayer flakes in this region? -In Figure 3f and line 170, was the unexpected area of decreased current reproducible (i.e., was it observed on subsequent SECM images) and were similar features observed in other samples? -Could the more negative current on the left side of the flake in figure 3d and lines 172-173 be due to sample tilt or differences in the apparent HET of the sample? Were any adjustments made (experimental or software) to correct sample tilt relative to the probe in all of the SECM images? -Would it be possible to perform AFM-SECM on MoS2 on graphite to compare the electrochemical reactivity with STS results? I agree with the authors that the underlying substrate could affect the electronic properties of the MoS2.
-In figure 5, numbers to vertical axis would be helpful.
-Why are intervals given for some numbers (e.g., the work function) and not others (saturation ratio of 30%)? Are intervals standard deviation or confidence intervals. The intervals for the monolayer and bilayer overlap at 5.17 eV. Are the work functions of the monolayer and bilayer different?
Response to Referee Report 1: We deeply thank Referee 1 for his/her comments and suggestions. We agree that our manuscript has many, more technically oriented aspects, which would certainly qualify it for a respective journal. Explicitly considering Referee Report 2, we see it is mandatory to give sufficient background details so that the reader can clearly identify the differences in comparison to conventional SECM. However, our modification is only a tool -with the previously hidden physics at the interface revealed and discussed. As also pointed out by the other referees, our approach gives an insight into the present system, which was not previously explored and is benefitting from the chosen mode of SECM operation and the combination with additional techniques. In so far, our approach suggests an application not limited to MoS 2 but to all systems with local electro-chemical processes, as already presented in the discussion section. Nevertheless, to demonstrate a more general value, we performed additional light irradiation experiments (see Fig. 5c), which further supports our interpretation and -as suggested by Referee 3 -we extended the introduction to widen the focus.
Review 1: (Q1.1) In the present work, He-Yun et al. study the heterogeneous charge transfer behavior between MoS 2 flakes of various layer numbers and sizes and the liquid interface, using the ferrocene/ferrocenium (Fc/Fc+) as redox probe in near-field scanning electrochemical microscopy (SECM). Spatially resolved redox maps reveal a reactivity dependence on the area of the flakes and the number of layers of MoS 2 . In combination with scanning tunneling spectroscopy, they finally discuss a band alignment occurring at the liquid-solid interface. Fundamental knowledge of the electronic structure at the solid-liquid interface in MoS 2 layer is relevant for the understanding and control of the photo-and electrocatalytic behavior of layered MoS 2 regarding, for example, hydrogen evolution and oxygen evolution reactions. The idea of addressing this question using SECM is quite remarkable as it provides parallel topological and electrochemical information. However, it is not clear to me how the results from the work of He-Yun et al. can be translated to the former electrochemical reactions.
In my opinion, the paper is more technically oriented to validate the SECM technique in 2D TMDs, but lacks of new insights in the physics at the MoS 2 -liquid interface. Therefore, I do not consider this work, as presented, to be published in Nature Communications.
Answer: See general discussion above. Our research, which was initially indeed aimed for electrochemical reactions, did reveal that first a detailed understanding of the band alignment is mandatory. Extremely highly cited, the review article of M. Grätzel (Nat. 414, 338 (2001)) identifies band alignment and its understanding as crucial for improvements on photoelectrochemical cells. "The flat band potential is a very useful quantity in photoelectrochemistry as it facilitates location of the energetic position of the valence and conduction band edge of a given semiconductor material. "; "This depends on the applied bias voltage according to the Mott-Schottky equation, where ∆Φ SC is the voltage drop in the space-charge layer." Now we compare the same issue (∆Φ SC ) by the observation of feedback current difference between ML and BL MoS 2 single crystal, which owing to their band offset difference. Our articles delivers unprecedented access with high spatial resolution into the local processes as band offset and confinement of the faradic current obtained, and unravel the charge transfer behavior at the solid-liquid interface of MoS 2 flakes of various layers and sizes.
(Q2.1) The authors provide a fair description of the observed experimental facts and intuitively argue the space charge in the MoS 2 -electrolyte junction -which is carried out by work function difference-to induce a band offset. However, they miss a deeper discussion about the meaning of the observed area and layer dependent reactivity in the monolayer regime, in contrast to the fixed 30% feedback enhancement in bilayer islands. A feedback enhancement vs. flake size comparative plot would help to visualize the correlation between mono-and bilayer flakes, depending on the size range, which is expected to face smaller bilayer flakes.
Answer: This is indeed a very interesting question, which we would love to be able to give an answer -but we are not able with the reasons stated in the manuscript. The feedback enhancement is driven by the recycling area (see Fig. 1), with its size determined to be of the order of (200 nm) 2 . The detection threshold is given by the current resolution in dependence of the size (8 fA/μm 2 ) and would require measurements in the range of Atto-A (10 -15 A), which is not feasible. This is demonstrated in Fig. 2a/e with the current vanishing for an island of 100 times the size, when feedback enhancement is expected to be affected. Fig. 5c gives now even more additional experimental evidence for our intuitive physical interpretation.
(Q3.1) Hydrogen evolution reaction has been largely related based on the emergence of deep in gap states associated to the presence of S vacancies [Li et al., Nat. Mater. 15, 48-53 (2016)]. More recent studies have identified oxygen substituents as the most abundant point defects in monolayers of TMD semiconductors, which remove the vacancy in gap state [Barja et al., Nat. Comm., 10, 3382 (2019)]. Relevantly, oxide samples show an enhanced catalytic activity for HER [Petô et al., Nat. Chem. 10, 1246-1251(2018] compared to fresh ones; and formation of lateral heterointerfaces between defective and pristine regions have also been demonstrated [Kastl et al., ACS Nano 13, 1284, (2019, with spatial extent similar to the feedback dependence inhomogeneity observed by the authors. The authors need to consider and justify the effect of the former point defects in the observed local variations of the reactivity. Answer: Vacancies and defects are very interesting aspects additionally contributing to catalytic activity. In the present manuscript, selected data were explicitly chosen from clean samples with a low and over the surface constant abundance of defects and vacancies. In so far, for the study of the principal effects, selected samples offer a constant background variation, unlike the system studied in [Kastl et al., ACS Nano 13, 1284, (2019]. In Kastl et al. a pronounced rim variation in KPFM can be found (Supplement Figure 3 in above article with add. details) which originates from the specific growth. We also find such rims under altered growth conditions but for the purpose of this manuscript, we excluded such growth studies. S10 clarifies the mentioned issue for the studied films.
Some more comments are: (Q4.1) 1. Units must be included in all graphs.
Answer: As suggested we added: (Q5.1) 3. In page 2, line56, it is the authors claim "Electrochemical and topological observations on the scale of nm in x-y direction and atomic resolution in z direction are achieved". Such resolution seems unlikely under the experimental conditions presented in this work. This sentence needs to be justified or removed.
Answer: The chosen phrasing is indeed (unintentionally) misleading. The stated precision is technically given but observed structures are obviously not on the same scale. We respectively clarified:
Answer: Supplement S1 (as referenced in the manuscript) is the text under S1. For clarity, we added a table S1.

Reviewer 2:
We thank the referee for the intense and knowledgeable questioning, which is often directly or indirectly linked to our interpretational and strongly criticized approach -see Q14.2 "I disagree with assuming that the SECM probe is a point-like source". We see it therefore mandatory to clarify this issue before discussing detailed aspects, which we thoroughlyand we are optimistic that the referee agrees -handled to his/her fullest satisfaction and certainly improved the readability of the manuscript.
We do fully agree, that a numerical approach for the handling of the (underlying) Poisson equation with all the side conditions (recycling flow etc.) is in general required to properly and reliably interpret any effect related to vertical displacements and structural features under the tip apex.
Our experimental approach, i.e. where we treat/discuss the probe as a point-like source, with the probe operated in closer proximity to the surface than in conventional SECM, relies dominantly on information obtained during horizontal probe displacement with AFM (which is in conventional SECM mode only used after or before but not in parallel) recording the respective vertical displacement. Only because of this available additional information, we are able to identify the crosstalk between proximity related SECM variations and to know the apparent proximity to any surface topography related structure at any time of the SECM recording. Therefore, the SECM contrast observed for the bi-layers at a resolution of ~200 nm defines the lower limit of structural analysis to be performed by numerical methods.
However, we do fully disagree, that in the context of the manuscript, discussion under usage of a point-like source anything more than a minor (if any) variation can be expected from a more complex treatment -and can be traced back to the problem of the E-Field calculation (the gradient drives the mass flow in SECM) between two charged objects with the separating distance significantly larger than object size.
First: The length scale is larger than the range within we can experimentally trace any proximity effects.
Second: The case of SECM is complicated by the mass transport, explicitly of the recycling flow, which is rather inhomogeneous in close sample-probe proximity and in variation with distance. An influence of the vertical separation on our interpretation can be excluded based on the experimental evidence given. An inhomogeneity (which is not described by our point-like probe approach) can be expected -however, critical is only any change of homogeneity with the experimental conditions, which is here the position. When approaching an edge in a distance much larger than the actual probe-sample distance, where proximity effects do not contribute anymore, the change of inhomogeneity is essentially negligible and respectively the relative recycling rate variation over the entire probe (which again is much smaller than the distance to the edge). The treatment of the given case is in so far equivalent to the classical case of two differently charged objects at a separation much larger than objects dimensions. The limitation is, that we cannot give an interpretation of the absolute values -which we did not do and would require a more complex approach. However, the interpretation of relative changes as we did throughout the entire manuscript -as we chose the experiments and samples -remains unaffected.
(Q1.2) Summary: The authors studied the relationship between the work function and observed electrochemical reactivity of MoS 2 using several scanning probe techniques (viz, SECM, AFM, STS, STM, and KPFM) with a ferrocene (Fc) redox mediator. Their main results are that multilayers exhibit a higher reactivity than monolayers due to the lower conduction band potential of the multilayers and that bipolar electrochemical activity is proportional to flake size. The data demonstrate clearly a difference in electrochemical reactivity related to the number of MoS 2 layers. However, the analysis and interpretation of the SECM data as well as ambiguous experimental conditions are not convincing.
Answer: See the above discussion. In the specific experimental realization with the interpretation basing on the horizontal displacement only (in difference to essential all previous publications) -and vertical effects resolved by the AFM-SECM signal cross-talk -we cannot agree with the reviewer's statement, that the conditions are not convincing and we do not see any ambiguity, explicitly, as not further specified.
More specifically, although a seemingly rough approximation is used, strong evidence is given by the experimental observations itself, gathered in Fig. 2, that no additional high order terms outside the approximation contribute. Numerically treating the system in a fully 3D approach under the known geometrical parameters (as suggested by the Referee), the feedback value is described by a double integral (sample and probe surface -considering actual ion acceleration along the gradient, it can be simplified to a single integral) of a path integral (ion flow) under consideration of an inhomogeneous distribution of charged and uncharged mediator (cell geometry / boundary conditions). It is then instructive to realize, that in the case of a horizontal movement alone, the mathematical complexity vanishes for the interpretation exactly in the case, that the feedback reacts on changes in the near field (i.e. approaching step) but does not capture the geometrical properties in the far field. Applied to the results in Fig. 2, we find that the feedback varies in dependence of the distance towards the nearest edge -but is independent of the distance to the corners of the triangularly shaped islands. Mathematically, this implies that only the highest order term contributes, which is given by a point-like probe and all other terms, which come from a geometrically more realistic descriptions of the probe-sample system are negligible. This approximation holds till we experimentally observe deviations, which are given below ~200 nm -from this point onwards, an approach as suggested by the Referee is mandatory and this range is explicitly not included in our discussion -and therefore not considered in the discussion of the lateral extension of BL mapping. Instead, the discussion on the BL is restricted to the case of saturation. i.e. where higher order far-field terms do not contribute to an experimentally observable signal.
This paper has the potential to be of interest to researchers in a variety of fields, but the paper's contribution to the body of knowledge on TMDCs is not apparent, as the main results of the paper are not discussed in comparison to the literature on TMDCs. The novelty of AFM-SECM (Eifert and Kranz, "Hyphenating Atomic Force Microscopy," Anal Chem 2014, 86, 5190-5200) and observation of layer number-dependent electrochemical activity and band gap (Velicky and Toth, "From two-dimensional materials to their heterostructures: An electrochemist's perspective," Appl Mater Today 2017, 8, 68-103) of MoS 2 is unclear.
Answer: We need to emphasize, the value of our work is not to give specific answers to the physics of TMDCs nor to establish an entirely new technique -However, new insight is given, as such, we would not dare to submit to this prestigious journal. Our experiments demonstrate that the evolution of existing techniques and the combination with a broad range of other methods give highly valuable access to a deeper understanding of the fundamental aspects of catalytically active liquid-solid interfaces not previously obtained. TMDCs serves here as an example with new insight obtained -but our approach is not limited to the physics of TMDCs.
Specific comments and questions: (Q2.2) -Abstract is misleading. Photocatalytic and water splitting properties of MoS 2 should be de-emphasized, as the authors do not present any photocatalytic results or relate their results to the photoactivity of MoS 2 . If water splitting is of interest, why was organic solution used? What are "local processes" (line 24)?
Answer: Although we cannot agree with the reviewer's comment of de-emphasis, Referee 3 even suggests to enlarge scope of the introduction -as it is a central motivation for the current research, we did tune the localization of potential relevance to less pronounced positions within the abstract and introduction for the current revision. We specified: "….the local processes as band offset and confinement of the faradic current obtained." (Q3.2) -What does "near field" (lines 22 and 54) mean? Do the authors mean small SECM probe-sample distances? Similarly, the description "large distance above the sample" in line 47 is vague. What is a large distance? SECM measurements are done with probe-sample distances less than the probe radius size.
Answer: we specified in the modified version "large distance" with "of typically more than 100 nm" and "near field" with ", with the probe operated in the AFM mode in close nm proximity" -and in the abstract, we added "i.e. in close nm probe-sample proximity" See above discussion, our point-like probe approach requires the identification where validity can be expected -and respectively, we more precisely defined near and far field / respectively large distance.
(Q4.2) -Instead of the focus on AFM-SECM (see, for example, the paragraph starting on line 53) and achieving high resolution, what may be more interesting in this paper is the ability to compare directly local work function and electrochemical activity at the same location on the same MoS 2 sample. However, this capability is not apparent in the paper as written.
Answer: A very interesting suggestion, we are currently upgrading our system for such a purpose but is currently (and for this paper) not available -and the limitation of the solvent and the mediator on the KPFM measurements needs first to be accurately determined with unknown outcome.
Answer: Changed as suggested.
-Several experimental conditions need to be clarified.
(Q6.2) (a) Supporting electrolyte and solvent need to be verified. Did the authors mean tetrabutylammonium perchlorate (TBAP), which is a more common electrolyte, instead of tetrabutylammonium perruthenate (TBAP) (line 300)? If perruthenate, an oxidizing agent, was used, would it oxidize the redox mediator or MoS 2 ? Polypropylene carbonate (line 301) is a thermoplastic material and not a liquid solvent.
Answer: Yes, we agree with the referee that this chemical is miswritten. Therefore, the "tetrabutylammonium perruthenate" is changed to "tetrabutylammonium perchlorate". Also, the "polypropylene" is changed to "propylene". The solvent is propylene carbonate indeed. We do apologize for these mistakes. and I assume that the authors used similar, commercially available probes. However, cyclic voltamograms (CVs) of the probe in the redox mediator would be helpful as well as an approach curve (i.e., plot of probe current vs. probe-sample distance) of the probe over the Si/ SiO 2 surface to confirm the probe dimensions used in this paper. Additional characterization such as SEM would also be helpful.
Answer: Within the framework of relative changes in the chosen experimental approach and with the confinement area -which reflects directly the actual tip geometry -experimentally determined, the knowledge on the actual tip geometry is redundant. We agree that for the future development, the requested information are interesting for the technically interested readers. We therefore added to the supplement CVs and the approaching curve of the probes as Figure S5  Answer: The stability of the tip is indeed highly relevant. The tip quality decays with time in terms of signal sensitivity. However, as relevant for the present manuscript, the signal enhancement scales proportionally -and the relative signal remains constant (till current saturation terminates running experiments). We added S6 for clarification, see also S1. See above discussion: we focus on relative changes. (Q11.2) Is SECM probe current measured in the main mode, where the probe is in direct contact with the sample, as shown in Figure S3a (line 297)?
Answer: AFM works on the peak force mapping module, which is modified tapping mode instead of contact mode.
(Q12.2) In my understanding of PeakForce SECM, two steps are involved: (i) main mode (tapping) is AFM to determine topography and (ii) lift mode is SECM to determine electrochemical reactivity with a constant probe-sample distance based on AFM profile. How was SECM probe-sample distance of 100 nm achieved and chosen for this paper?
Answer: We found that the feedback current (normalized with background) is reduced with the probe-sample distance within 700 nm (see Fig.S5). We choose the probe-sample distance around half of the maximum feedback current which is roughly 100 nm as the lift mode scan distance to record. For SECM measurements in this paper, here explicitly the recording of the feedback current was performed in the main scan mode.
(Q13.2) (f) Why was SECM current normalized with respect to negative feedback at Si/ SiO 2 instead of current at semi-infinite conditions, as is done typically in SECM measurements, and how was this normalization done? "cell geometry related background" in line 79 needs to be explained.
Answer: The background current (cell geometry related background) means the current of probe on top of SiO 2 in the same AFM scan line. The probe current on SiO 2 is kept similar in the same AFM scan line. However, the probe current on SiO 2 (background current) will slightly increase with time, as shown in Fig. S6(a). Figure  S6 (b) compares the variation of the probe current on SiO 2 surface. The normalization of SECM image is done with respect to the background current since the SECM probe current will reduce the background current (SiO 2 position) in the same scan line. Figure S6(c) shows the SECM image after normalization of probe background current along Y direction.
Answer: We added to the figures as suggested.
(Q15.2) I assume that the CVs are for the SECM probe but don't understand the decaying, diffusion-limited current. I would expect steady-state current (i.e., sigmoid shape and independent of potential) for ultramicroelectrodes such as SECM probes.
Answer: Fig S4 is the CVs of the Pt wire with a diameter around 0.25 mm. The tip CV is attached in Fig. S5a. (Q16.2) (h) Experimental details, for example bias current and scan rate for STM data and frequency used for Raman imaging in Figure S6, need to be given. Figure captions in the SI need more detail.
Answer: Added to figures as suggested.
(Q17.2) -Fc passivated the probe within a few hours, rendering quantitative comparison between samples difficult (SI section 1). Could impurities in the Fc or solvent absorb on the probe or sample? Did other mediators (e.g., DmFc) exhibit similar, passivating behavior?
Answer: As discussed above in our reply to question (Q13.2), the probe current is continuously increasing with time both on top of MoS 2 and SiO 2 . We observed this phenomenon for both Fc and DmFc in organic solvent. We also observed a current drift in the acid solution but on a significantly larger time scale (days). Although we cannot give a definite answer on potential impact of impurities, we have no indication that additional impurities are present for Fc and DmFc.
We need to stress, that within the framework of the manuscript all analyses are based on relative measurements, as within the same scanline, and different sample systems are probed in SECM and AFM. Therefore, the evolution of the increasing current was monitored during the entire experimental course in parallel for all systems and we verified, that the current drift has no effect on the statements of this manuscript.
(Q18.2) -How was the Ag QRE potential determined with respect to the SHE ( Figure S4)? Is the potential of the Ag QRE expected to be 0.00 V vs. SHE? Is the QRE potential stable during the measurements?
Answer: The CV of the Pt wire is measured both with the references of Ag wire and non-aqueous silver/silver chloride reference electrode at the same time in Figure  S4(a,b). The Ag QRE potential is determined by comparing with non-aqueous silver/silver chloride reference electrode first, then normalized to SHE as shown in Figure S4(c).
(Q19.2) -The interpretation of the SECM needs to be reevaluated. The statement "the distance limits the maximum achievable resolution and, as resistance for the ion flow increases, the current sensitivity" in lines 49-50 is confusing. (Q19.2a) SECM probe electrode size determines spatial resolution. I agree with the authors that SECM probe current is affected by probe-sample distance, but this current arises from Faradaic processes (i.e., charge transfer, electron transfer in this case, across the electrode-electrolyte interface) between the probe and redox mediator, not ion flow (Q19.2b). For electron transfer to occur, the mediator needs to diffuse to the probe. In negative feedback (such as over an insulator like SiO 2 ), current decreases at small probe-sample distances because mediator diffusion to the probe is hindered by the sample.
In positive feedback (such as over a conductor), current increases due to increased flux of mediator generated at the sample. Commonly, feedback current is analyzed using models for negative and positive feedback(e.g., LeFrou and Cornut, "Analytical Expressions for Quantitative Scanning Electrochemical Microscopy (SECM)," ChemPhysChem 2010, 11, 547-556). The authors mention positive/negative feedback in lines 89-92 and 308-316 (although I don't understand why Ref 26 is cited here) but do not apply this analysis to their results (Q19.2c). (Q19.2d) If ion flow does play a role in the SECM current, would current/resistance vary with ionic strength?
Answer: (to Q19.2a) The dimensions of the probe are smaller than the probe-sample distance in conventional SECM. The nano-electrode probe is coated with dielectric materials and has an exposed conical Pt tip apex of ∼200 nm in height and of ∼25 nm in end-tip radius. (to Q19.2b) Undoubtedly, the current at the liquid-solid interface is Faradaic. We need to admit, that a potential misunderstanding arises from an indeed imprecise wording with no distinction of the dominant contribution (Q19.2d)-as we realized when we tried to identify the source of the obvious misunderstanding. We respectively changed: Instead of "The observed size dependence of the feedback of approximately 8 fA/μm 2 is near-linear with the flake area S1" we changed to: "The feedback scales with approximately 8 fA/μm 2 (see S1)." And: "We conclude that here, the observed strong correlation between flake area and feedback is dominantly controlled by the charge flow (Instead of "Ion Flow") towards the recharge area (Region II) (moved from the end of the sentence), i.e. the product of ion conductivity and oxidation rate." We thank the referee for identifying this potentially misleading wording.

(to Q19.2c) Negative Feedback is certainly present. The experiments are performed in horizontal displacement only (the impact of vertical displacement is minor and discussed in the manuscript within the framework of confinement effects)
and different surface areas present in parallel. Therefore, the relative changeswhich are the base of our analysis -are entirely unaffected. Respectively, we can separate information originating from the geometrical structure and surface properties, not equally accessible in conventional SECM, which is covered by the given references.
(Q20.2) -I agree with the authors that resistance in the sample can affect the observed SECM current (e.g., by altering the apparent HET kinetics), but how sample resistance relates to heterogeneous electron transfer (HET) of Fc at MoS 2 needs to be better explained as well as how well the author's model compares with the SECM literature on unbiased samples of finite size (e.g., Ref. 27: Wipf andBard, J Electrochem Soc 1991, Ref 28: Amemiya, Anal. Chem. 2007). The discussion of resistance throughout the main text and SI (section S2) is confusing. (Q20.2a) Is resistance due to the sample or due to electron transfer between the MoS 2 and the electrolyte? (Q20.2b) What gives rise to this resistance? (Q20.2c) In line 19 of the SI, where does the contact resistance originate? MoS 2 is electrically isolated from the circuitry (line 304 of the main text). (Q20.2d) Units need to be added to figure S2. (Q20.2e) Numbers need to be checked in the SI. On line 17, the area of the reference resistance is (15 mm)2 and in the equation on line43, it is 15 um2. Line 18 has 3 um2 (is it supposed to be 3 um on a side?).

Heterogeneous electron-transfer rate constants for ferrocene and ferrocene carboxylic acid at boron-doped diamond electrodes in a room temperature ionic liquid. Electrochimica
(Note, we changed the order of the following two questions as answers build upon each other) (Q24.2) -I disagree with assuming that the SECM probe is a point-like source (line 147). Theoretical models are available for analyzing a conical SECM probe of finite size.
Answer: Here, we fully disagree with the referee -see discussion above -the point-like source approach is only applied when we experimentally cannot detect any higher-order contributions, which then would indeed require a more detailed 3D numerical modelling. This is here not given.
(Q25.2) Why does the resolution change based on the number of monolayers (see lines 140, 145-146)? I would expect the same resolution for similar SECM parameters and probe size regardless of sample features, as the resolution is determined by the SECM probe radius.
Answer: It is a seemingly surprising experimental finding that the resolution depends on layer number and did initially puzzle us as well (for the very same reasons mentioned by the Referee). More advanced numerical treatment of the geometry did not reveal any respective contribution. Our experimental approach (see discussion above), allows for an unambiguous conclusion that crosstalk and confinement effects can be disregarded -in difference to conventional SECM. We think, that the manuscript already gives a clear answer, which is in short the following: Whereas handling the charging and recharging -as described in the manuscriptas two independent resistance networks immediately reveals the origin of different distance dependences. When approaching the monolayer edge from SiO 2 , the hampered recharging of the insulating SiO 2 causes a redirection of the current flow to the more apart and efficiently recharged monolayer whereas when approaching the 2 nd layer, efficient recharging of the monolayer underneath is maintained by the conductive layer. Therefore, the enhancement current related to the 2 nd layer appears only at closer vicinity.
(Q26.2) -The following needs to be clarified. (a) Lines 109-111: The statement beginning "AFM resolution is of the order of 100 nm (Fig.  2(d)), which is image size but not technically limited, in comparison to AFM-SECM resolution of ~2.5 μm. i.e. on the length scale beyond 100 nm." Answer: We modified the manuscript for clarity: "The apparent structural resolution obtained in AFM is of the order of 100 nm (Fig. 2(d)) whereas in SECM of ~2.5 μm. i.e. on the length scale beyond 100 nm, therefore, variations in the SECM originate entirely from the local electro-chemical activity and are not affected by topographical cross-talk." It is worth noting that within the context of the referee's criticism on different points, that the observation of a cross-talk decoupling beyond 100 nm, a strong confinement can be deduced and respectively specifies the range within the validity of a point-like approach can be falsified and beyond which a point-like probe approach is reasonable, i.e. in our discussion when the distance (and variations) are on a length scale much beyond 100 nm. Otherwise, as demonstrated in Fig. 3f of the main manuscript, 2 nd ML islands would be traceable at larger horizontal distances.
(Q27.2) (b) Lines 119-120: "strong correlation between flake area and feedback is dominantly controlled by the ion flow, i.e. the product of ion conductivity and oxidation rate towards the recharge area (Region II)." Answer: See Q19.2b, the sentence is unintentionally misleading and respectively corrected.
(Q28.2) (c) Line 148-149: The statement beginning "All dimensions of the used ultra-microelectrode probe are on the 0.1" Answer: The unit appeared at a later position. This is indeed potentially confusing. We changed: 0.1 μm range -the recorded SECM resolution for the monolayer is on the μm (Q29.2) (d) Discussion on lines 154-157, particularly which conductivity is being determined and how d-3 is determined.
Answer: As we refer to the "redox concentration gradient", the charge conductivity due to ion flow can be unambiguously identified. In the case of large separations of a (respectively small -considering the dimensions of the one used) probe from a surface, the potential gradient (and of the concentration gradient respectively) scales with d-2. Considering the additionally extended diffusion path (scales with d) from probe to sample results in the above statement.
(Q30.2) -The statements "When the probe is located over an insulating surface like SiO 2 , the feedback is driven by the charge flow only towards the more distant, continuously recharged monolayer" (lines 158-159) and "Moreover, a detectable feedback current is still present at a distance of μm from the flow towards a distant MoS 2 island "(lines 272-273) are misleading. The probe reaction (oxidation of Fc to Fc+) is driven by potential applied by the potentiostat and not a bipolar reaction with a distant (how distant?) MoS 2 flake. The distant flake is on an insulator (SiO 2 ) and is unable to deliver charge (electrons) through the sample to the area under the probe. For similar reasons, I disagree with interpreting the paper's results based on Ref. 24 (Amatore et al.). In their experiment, the distant electrode is electrically connected to the sample area under the probe and thus can act like a bipolar electrode.
Answer: Here is an obvious misunderstanding. We clearly speak of the feedback current from the probe to the continuously recharged monolayer (btw (how distant? -is answered by the experimental data presented in the figure 3 and the respective manuscript statement: "MoS 2 monolayer flakes at a resolution of ~2 μm").
(Q32.2) -Could the enhanced positive feedback current for the center of the larger area flakes in figure 3g (line 167) be due to the large density of bilayer flakes in this region?
Answer: We considered this possibility for the original manuscript submission and we think we can now disregard this possibility. Such an effect should have a similar artefact emerging in 3d/e (and other data), which is not observed.
(Q33.2) -In Figure 3f and line 170, was the unexpected area of decreased current reproducible (i.e., was it observed on subsequent SECM images) and were similar features observed in other samples?
Answer: We find it important to include alternative effects, which are beyond our explanation if present, instead of selecting those images we like. Of course we verified that it is not an artefact -already the feature in 3f is a reproduction of the same effect in the enlarged view scanned We remain with our original statement: "In Fig. 3(f), SECM data reveal a local activity depression (reflected by the green color) which is not correlated to any topographical feature as recorded in AFM. Although we cannot comment on the physical nature, the experimental observation demonstrates the value of SECM to give an important additional access to surface activity beyond simple topographical features." (Q34.2) -Could the more negative current on the left side of the flake in figure 3d and lines 172-173 be due to sample tilt or differences in the apparent HET of the sample? Were any adjustments made (experimental or software) to correct sample tilt relative to the probe in all of the SECM images?
Answer: We refer to the statement on our original manuscript: "In Fig. 3(d), an asymmetry in SECM data of the monolayer is apparent. We verified that misalignment effects in the imaging and normalization processing could be disregarded, though the actual origin is still unclear." Note: As AFM and SECM are conducted in parallel an artificial wrong tilting for constant height measurements as in conventional SECM is not possible here.
(Q35.2) -Would it be possible to perform AFM-SECM on MoS 2 on graphite to compare the electrochemical reactivity with STS results? I agree with the authors that the underlying substrate could affect the electronic properties of the MoS 2 . (Q37.2) -Why are intervals given for some numbers (e.g., the work function) and not others (saturation ratio of 30%)? Are intervals standard deviation or confidence intervals. The intervals for the monolayer and bilayer overlap at 5.17 eV. Are the work functions of the monolayer and bilayer different?

Answer: The combination of all experiments within one setup and conducted in
Answer: We answer in reverse order. The work functions for ML and BL are different. The stated error values reflect statistical error (which is minor as seen from Fig. 4d) and systematic error from repetitive realization of the measurement (different tips, sample, solvents). Thereby, the relative change is not affected by systematic errors -but as demonstrated in the same figure, the relative change is beyond the statistical error. The seemingly precisive saturation ratio was not intended, but a meaningful error can also not be stated (as the exact exp conditions do play a role which we can not individually separate). We respectively changed at different positions within the manuscript to avoid the impression of an unintended accuracy (for example "….enhancement of the order of 30 % on ….").
(Q38.2) I disagree with using the term electronegative to describe redox-active species, as electronegativity is an atomic property. When saying more/strongly electronegative, do the authors mean that the mediator is more reducing or has a more negative reduction potential?
Answer: We agree, the wording is not appropriate. We mean more negative reduction potential using normal hydrogen electrode (NHE) as a reference.

Reviewer 3:
With thank the Referee for his/her valuable comments and we follow his/her suggestion to widen the scope of the introduction and respectively, widen the potential application of our finding by introducing with Fg. 5c new data on light irradiation experiments.
The authors report on the characterization of heterogeneous charge transfer behavior of MoS 2 flakes varying the layer numbers and sizes. They applied a combined AFM-SECM approach to probe the reactivity of different number of layers, and mapping the band alignment as well. This manuscript reports original and novel results, after reviewing the manuscript I think this work can be published in Nature Communications after major revisions.
My comments and questions: (Q1.3) The Introduction and Experimental sections are well written and every step is showed.
However, the Introduction is quite straightforward, probably the Authors could have a think to emphasize some points more extending this part a bit. (Q5.

Answer: In order to emphasize the importance of semiconductor properties of 2D
3) The authors mention that "Due to the respective changes of the environments, it can be assumed that absolute values will be different for MoS 2 / SiO 2 , but that relative changes remain qualitatively valid." in the "Layer-dependent work function and band offset" part. The effect of substrate (insulator vs. conductor) for electrochemical behavior of TMDs is also known, so authors should consider this as well, proving with control measurements on both substrates.
Answer: We agree, each subsystem of the Substrate/MoS 2 /Mediator system will impact the band structure and the electrochemical behavior and goal for the future should be to explore the specific aspects. In this manuscript and beyond the coherence of results of different methods with 5c added, we show the validity of our interpretation for a change of the mediator. For the benefit of a clear result, the change from an insulating to a conducting substrate will also affect the recharge flow through the substrate in a non-trivial way. Although we fully agree that the Referee suggestion is great and in line with our thoughts, it also covers significant experimental barriers. (Q6. 3) The comparison of different number of layers seems to be a bit random, only the ML and BL, or ML and BL and thicker samples are compared in some cases, therefore some summarizing explanations/graphs of the systematic study are necessary. More explanation and discussion on the HET dependence of the number of layers are also needed, to underline one of the selling points of the manuscript, the electrochemical reactivity dependence of different number of layers ("heterogeneous charge transfer behavior of MoS 2 flakes of various layer numbers").
Answer: Sample preparation was by purpose optimized for mono-and bi-layer growth, with both systems systematically addressed. Additional data for higher layer numbers are used only if available.  (2017)) demonstrating nanoscale redox mapping of 2D materials with the same commercial system (Bruker PeakForce AFM-SECM) used in the manuscript, and the sentence "The nanoelectrode probe is coated with dielectric materials and has an exposed conical Pt tip apex of ∼200 nm in height and of ∼25 nm in end-tip radius." (line 62) is word-for-word the same as that in the abstract of that reference. In their rebuttal, the authors mention "the evolution of existing techniques and the combination with a broad range of other methods" and that "TMDCs serves here as an example with new insight obtained -but our approach is not limited to the physics of TMDCs," but the manuscript, as written, does not clearly describe in more detail how these insights can be expanded to other applications or materials. Given the body of existing work on MoS2 electrochemistry and band structure (e.g., Choi, S et al. 2014;Trainer, D. J. et al. 2017;Velický, M. et al. 2016; Velicky and Toth 2017 cited in the manuscript) and the focus on demonstrating the use of scanning probe systems to characterize MoS2, this manuscript may be better suited for a more technically oriented journal.
These are the specific comments and concerns I have after reading the authors' rebuttal: (1) I appreciate the authors' diligence in reading and summarizing literature on heterogeneous electron transfer (HET) rates of Fc and agree with their caution in comparing values obtained under different experimental conditions (Q21.2). However, I think that putting their results (how apparent HET and band gap relate to the number of layers and flake size) in context with reported values is important.
(2) Although I appreciate the addition of the photoelectrochemical measurements (Fig. 5), I don't see a clear correlation between the photoelectrochemical water-splitting at TMDCs mentioned throughout the manuscript and the measurements performed in organic solvents. I would have liked to see an explanation as to why organic supporting electrolyte was used (e.g., organic electrolytes were used instead of aqueous ones due to delamination of the MoS2 flakes). I also would have liked to know how the band alignment in organic solvent would compare with that in aqueous solution (see for example, LF Schneerneyer and MS Wrighton "Flat-Band Potential of n-Type Semiconducting Molybdenum Disulfide by Cyclic Voltammetry of Two-Electron Reductants: Interface Energetics and the Sustained Photooxidation of Chloride" J. Am. Chem. Soc. 1979, 101, 6496) (3) I appreciate the authors clarifying the electrochemical measurements by adding more details to the figure captions as well as correcting units and electrochemical details (e.g., supporting electrolyte). However, I stand by my original assessment that the interpretation and discussion of the electrochemical results are ambiguous: (a) Regarding Q18.2, I still don't understand why the potential of Fc versus the SHE (standard hydrogen electrode) is the same that versus the Ag quasi-reference electrodes in Fig. S4c and how the authors arrived at their values. I would expect these values to be different, although I agree that 0 V vs. SHE is comparable to -4.44 eV for vacuum. The authors mention using both the (SHE) and Ag quasi-reference electrode (lines 100 and 107 in supporting information) as well as a "nonaqueous" silver-silver chloride reference in their response. The SHE is difficult to use in practice, and I wonder if they used a silver-silver ion electrode, which is more commonly used in organic electrolytes, instead of silver-silver chloride.
(b) In figure 5, I'm not sure why the current for the "off" photoelectrochemical measurements (280 to 300 pA) is much higher than that in the SECM images (ca. 1-15 pA). I would expect these values to be similar if the probe-MoS2 distance is similar for all of the measurements. (c) The authors answered my question about using the word electronegativity to describe the reducing potential of redox-active species (e.g., lines 276 and 295) in their written response (Q38.2) but did not update the corresponding manuscript text.
(4) I appreciate the authors including CVs and approach curves of the SECM probe in Figure S5 (Q7.2). These additions help me better understand the experimental conditions. However, the probe was operated at 0.6 V, where the current is still changing (in other words, it is potentialdependent), and the current value is obviously not stable or reproducible at 0.6 V, as the probe current at 0.6 V was different for all three probes. I would have expected the probe to be poised at a potential (e.g., 0.7 V) where the current is at steady state (independent of potential) as well as more consistent between probes so that any change in current can be attributed to a change in redox activity of the substrate and not due to factors such as a drift in potential of the silver quasireference electrode.
(5) For Q30.2, I am still unsure what a distant MoS2 island is. If the island is another, isolated MoS2 flake on Si/SiO2 (an insulator), then I don't understand how charge (electrons) are being passed through an insulator between the two flakes. If the island refers to two different areas on the same flake, then I can understand how the MoS2 (a semiconductor) is able to act as a bipolar electrode.
(6) For Q35.2, the authors may have misunderstood my question. I am not asking if they could combine all the measurement probes into one instrument but rather if they could do different measurements on the same type of sample, namely, measure MoS2 on graphite (a conductive substrate) using the AFM-SECM and comparing those results to AFM-SECM imaging on MoS2 on silicon oxide (an insulator) to see if the substrate (e.g., conductor vs. insulator) affects the observed electrochemical reactivity of the MoS2.
Reviewer #3: Remarks to the Author: Comments to the Authors regarding to the resubmitted version of their revised version of "Nanoscale redox mapping at the MoS2-liquid interface" The authors revised their original work, and resubmitted an improved version, which attempts to solve the three Reviewers' comments.
In my case, I see an improvement, and think that the manuscript had revised well, although it would improve more anytime. Regarding to my comments and question for the experimental and results parts are addressed, I am glad to see this. While one of my comment corning the fundamental aspects of the current work is remained untouched, namely the substrate effect. I can understand that in the current COVID-19 situation, we should focus current results and utilise those as much as possible without trying to extend the existing content applying further experiments. But the Introduction part is still confusing and contains missing parts, pieces of works for me. Please, consider to revise more on the Introduction part to try to emphasize more

Introduction
Two-dimensional (2D) crystalline materials are consisting of a single layer of atoms and expected to have a significant impact on a large variety of applications 1,2 . For transition metal dichalcogenide (TMDC) monolayers are direct semiconductors, of which a common representative is molybdenum disulfide (MoS 2 ). Monolayer MoS 2 has a device relevant direct bandgap of ~2 eV 3 and its electron concentration at the surface than in bulk is enhanced by more than three orders of magnitude 4 . Their stacking variability and the dependence of physical properties with thickness are attractive for property fine-tuning 5 . TMDCs are often combined with other 2D layered materials like graphene or hexagonal boron nitride for device application such as transistors, solar cells, water splitting 6 , and sensors. The preparation of respective, high quality material on various substrates is established by means of chemical vapor deposition (CVD). The selection of growth parameters and precursors gives a precise control of its morphology, i.e. flake sizes and layer thicknesses 7 .
The well-defined band structure of MoS 2 flakes with fixed layer number is ideal to explore the localized electrochemistry of 2D materials as relevant for heterogeneous electron transfer (HET) 8 , photoelectrochemistry 9 , batteries/capacitors etc. Scanning electrochemical microscopy (SECM) is an established method 10 to locally probe catalytic properties 11 as demonstrated for graphene oxide 12 and MoS 2 flakes in feedback mode 13 and biased mode 14 .
Whereas the device relevant absolute catalytic activity not only reflects the local catalytic property but its performances, individually identified, although highly intertwined 8,9,15,16,40,41 . The route towards highly efficient materials is the study and discussion of application relevant properties in conjunction with the full bandwidth of physical properties in different environments and their dependencies among each other 17,18 . This is adressed here for MoS 2 and can be equivalently expected to be relevant for 2D materials in general, whenever dimensionality effects modulate electronic and chemical properties.
In SECM, when an AFM (atomic force microscopy) 42 controlled probe approaches the un-biased semiconductor surface (feedback mode), a charge-flow through the local probe is established and gives access to the local electro-