Chemically identifying single adatoms with single-bond sensitivity during oxidation reactions of borophene

The chemical interrogation of individual atomic adsorbates on a surface significantly contributes to understanding the atomic-scale processes behind on-surface reactions. However, it remains highly challenging for current imaging or spectroscopic methods to achieve such a high chemical spatial resolution. Here we show that single oxygen adatoms on a boron monolayer (i.e., borophene) can be identified and mapped via ultrahigh vacuum tip-enhanced Raman spectroscopy (UHV-TERS) with ~4.8 Å spatial resolution and single bond (B–O) sensitivity. With this capability, we realize the atomically defined, chemically homogeneous, and thermally reversible oxidation of borophene via atomic oxygen in UHV. Furthermore, we reveal the propensity of borophene towards molecular oxygen activation at room temperature and phase-dependent chemical properties. In addition to offering atomic-level insights into the oxidation of borophene, this work demonstrates UHV-TERS as a powerful tool to probe the local chemistry of surface adsorbates in the atomic regime with widespread utilities in heterogeneous catalysis, on-surface molecular engineering, and low-dimensional materials.

I will summarize my impressions of the work first. The authors have demonstrated synthesis of high quality samples of both primary borophene polymorphs on Ag(111). For each of these polymorphs, they perform STM-based cryogenic TERS. They identify Raman modes related to the pure and oxidized borophene for both phases. Oxidation is accomplished by both atomic O and molecular O2 exposure. Differences in adsorption sites and oxidation behavior are highlighted between the v1/6 and v1/5 phases. Moreover, thermal reduction and tip-induced desorption are observed. The latter could easily be a platform for patterning! The methods employed are appropriate. Technical details such as the use of the iridium filament are sound (to avoid potential WOx deposition).
My questions (or perhaps, groupings of questions) are as follows: (1) Can the authors more explicitly relate the local oxidation sites with the expected electronic structure of the borophene atoms? That is, given that the boron atoms in borophene are in a number of inequivalent coordinations, how does the observed oxidation relate to the expected stability of these various coordinations? Does this imply anything about the nature of bonding in borophene? And finally, does this suggest a route to optimize the stability of borophene in new polymorphs? (this might lead the field of borophene synthesis in interesting directions).
(2) Is there a dose limit for recovery after oxidation via atomic O or molecular O2 exposure? Or restated differently, how much oxidation is fully destructive to the borophenes? The oxidation of borophene, and ostensibly the modification of the intrinsic properties, are a strong barrier to further applications and straightforward device fabrication. It would be great to explore whether the borophene can be recovered from intense oxidation via a similar protocol detailed in Fig. 6.

Reviewer #3 (Remarks to the Author):
In this manuscript, the authors used ultrahigh vacuum tip-enhanced Raman spectroscopy (UHV-TERS) to chemically interrogate individual oxygen adatoms on a boron monolayer, which is helpful for the deep understanding of the atomic-scale reactions. It is nice to see that the authors successfully manage to manipulate single atom and induce the single atom reaction with the substrate, and to chemically identify the formation of the bond via vibrational fingerprint. The article can be published after addressing the following questions: 1. While discussing Figure  , c (d comes before b and c, which is awkward). Thereafter, Figures e-g are described in the next paragraph. Such a logic is awkward. I highly suggest the authors to describe the phenomenon in Figure a first thoroughly, followed by simulation of the most stable structure and the possible vibrational modes. Then, the authors can try to correlate their result of simulation results. 2. In Figure 2i, from the description, it seems the authors assign both 205 and 189 cm-1 modes to the adsorbed single oxygen adatom. If this is the case, their trends of the TERS intensity profiles should not be inverse. I assume the authors would rather mean the 189 cm-1 mode comes from the mode of borophene under the influence of the neighboring O atom. If this is the case, the authors should clearly state this in the main text. Furthermore, the vibrational modes shown in Figure b-d can not be easily read. I would suggest the authors to increase the size or enlarge a portion of the model. 3. In the 2D TERS intensity mapping shown in Figure 2k and i, the authors mentioned that the oval-shaped rather a circle came from the thermal drift during TERS collection. To validate this assumption, the authors should provide the drift value of this TERS system, so that the result in the image distortion can be estimated. 4. In Figure 3d, the cluster shows no 205 cm-1 features in the 2D TERS mapping. It is better to also show the complete TERS spectra on the cluster site and the oxygen adatom site. 5. When comparing the activity of borophenes of different phases, why different amount of oxygen was introduced into the system (3600 L and 4800 L)? 6. Were Figures 5c and d obtained from the same position? I guess they are from different positions, as the substrate has undergone annealing. If this is the case, the authors should clearly state they are from different position, otherwise, the comparison of Figures 5a and b will mislead the understanding of c and d. 7. In figure 3b, why the two borophene with ν1/6 phase showed two different contrast? 8. It is not stated but can be observed that oxygen clustering tends to form near borophene edges, which need to be stated and explained. 9. The authors state that the simulations were performed based on a freestanding borophene monolayer without taking into account interactions with the Ag substrate, which led to the mismatch between theoretical values and experimental measurements. So, I am wondering are there any challenge in phonon simulations to include the effect of the Ag substrate. 10. In Figure 6c, when the bias is changed to 4.0 V, why the corresponding oxidized borophene height shows an immediate change at one line rather than changes from the beginning position? 11. The three isolated oxygen adatoms always exist in Figure 6a, e and f. It seems that the corresponding STM images show no features of these three adatoms when the bias is changed. Then, they will appear at the same position when the bias returned to 1.3 V. The authors should provide more explanations and the contrast bar of these STM images which is helpful for reading. 12. In Figure 6g, the author showed the scheme that there will be some clusters remaining on tip after the STM manipulation. We suggest the authors to check the contamination on the Ag(111) surface instead of clean borophene, which will be more convincing. The authors should also show the spectra of after tip retraction in all TERS results, which can convincingly demonstrate that the collected Raman signals are indeed from the sample instead of clusters on the tip.
In the point-by-point summary of the revisions below, the reviewers' original comments are italicized (in black), followed by our response (in blue) and corresponding changes to the manuscript (in green).

Reviewer #1 (Remarks to the Author):
The work has beautifully demonstrated several capabilities of TERS that are deeply desired by surface scientists. 4.8Å spatial resolution of TERS images in an STM set up matches the state of the art reported before. Most demanding measurements reported in the article such as imaging bonds of single adatom on different phases of borophene will push the TERS technique forward and encourage utilisation of TERS in answering intriguing questions likes of oxygen reactivity with borophene (or other materials).

This body of work addressed a number of aspects beyond utilisation of TERS as a technique. Examples include, reactivity of atomic and molecular oxygen, differences in reactivities between different phases, evolution of TERS spectra across a bonded oxygen, hopping of an adatom captured using STM imaging, thermal desorption and stability of oxidised borophene.
Adequate detail of the work has been given in the main manuscript and in the supplementary information. The article reads well and figures are presented clearly.
The work is significant in pushing the sensitivity and spatial resolution of STM TERS as well as interrogating behaviour of adatoms using chemical signature such as bond vibrations using Raman spectroscopy.

Reply:
We appreciate the reviewer's very positive comments and recommendation of publication.

Reviewer #2 (Remarks to the Author):
This manuscript by L.

Reply:
We are grateful for the positive feedback from the reviewer, and we also appreciate the valuable comments and questions. We have responded to each of them below and revised the manuscript accordingly.

My questions (or perhaps, groupings of questions) are as follows:
(1) Can the authors more explicitly relate the local oxidation sites with the expected electronic structure of the borophene atoms? That is, given that the

boron atoms in borophene are in a number of inequivalent coordinations, how does the observed oxidation relate to the expected stability of these various coordinations? Does this imply anything about the nature of bonding in borophene? And finally, does this suggest a route to optimize the stability of borophene in new polymorphs? (this might lead the field of borophene synthesis in interesting directions).
Reply: Boron has three valence electrons with an electronic configuration of [He]2s 2 2p 1 , but has four valence orbitals. That is, there are not enough electrons to fill all the electronic orbitals in the chemical bonding of boron atoms, in contrast to the case of carbon systems (e.g. graphene) that can adopt normal two-centered-two-electron bonds and form a hexagonal (honeycomb) carbon lattice with sp 2 bonding. Consequently, boron-related crystals are prone to multiple-centered-two-electron bonding due to electron deficiency. That is why the energetically favorable structure of borophene is a mixed hexagonal (electron deficiency)-triangular (electron surplus) lattice (e.g. V1/6 and V1/5 phase), which leads to inequivalent coordination numbers, such as 4, 5, and 6 in the V1/6 phase. According to the TERS and theoretical studies shown in Figure 2 and 4, bridge sites in either V1/6 or V1/5 structures, which link two boron atoms with a coordination number of 4, are the most vulnerable to oxygen attack. This implies that π bonds formed between the boron atoms with low coordination numbers are susceptible to chemical modifications, such as H, F, Cl functionalization (Phys. Chem. Chem. Phys. 2019, 21, 7630;Comput. Mater. Sci. 2019, 156, 56) or to interlayer bonding. Consequently, it suggests that chemical passivation or the formation of multi-layer structures via those active sites could enhance the chemical stability of borophene. We appreciate the reviewer's insight into this issue and admire his/her foresight, as we're happy to see that the improvement of ambient stability of borophene by the abovementioned approaches has been experimentally demonstrated in the recent past months. Li et al. reported the hydrogenation of borophene via atomic hydrogen in UHV, where the H atoms chemically adsorb to the B-B bridge sites with three-center-twoelectron B-H-B bonds (Science 371, 1143(Science 371, -1148. The resulting hydrogenated structure (i.e., borophane) showed enhanced chemical stability (negligible oxidation) even after 1 week of ambient exposure. Another work reported on the synthesis of bilayer borophene which features covalent interlayer B-B bonding via those boron atoms on the ends of bridge sites. The bilayer borophene is also inert to ambient oxidation compared to the monolayer counterparts, resulting from the significant charge transfer and redistribution due to the interlayer covalent bonding.

Revision:
We have added a brief comment in the Conclusion section to address the reviewer's concerns.
"We noticed that boron atoms with low coordination numbers (e.g., 4 in v1/6 or v1/5 phase) and the bridge sites linking them show remarkable activity to atomic adsorption (e.g., H, O, F) via multi-centered-two-electron bonding or to the formation of covalent interlayer bonding (in multilayered borophene). This insight has been demonstrated by the recent reports on the synthesis of hydrogenated borophene 39 and bilayer borophene 47 and their enhanced inertness to ambient oxidation due to the chemical passivation of active sites via B-H-B bonding or charge transfer and redistribution through covalent interlayer B-B bonds." (2) Is there a dose limit for recovery after oxidation via atomic O or molecular O2 exposure? Or restated differently, how much oxidation is fully destructive to the borophenes? The oxidation of borophene, and ostensibly the modification of the intrinsic properties, are a strong barrier to further applications and straightforward device fabrication. It would be great to explore whether the borophene can be recovered from intense oxidation via a similar protocol detailed in Fig. 6.
Reply: There is no dose limit for recovery after oxidation via atomic oxygen due to the predominance of chemically uniform surface species (O adatoms). In contrast, even a low dose of O2 (e.g., typically 1200 L) can lead to heterogeneous oxidized surface decorated with various boron oxide clusters, which cannot be thermally reduced and remain on the surface even after annealing up to 400 ˚C. The protocol described in Fig. 6 can reduce some complex boron oxides to simple species (e.g., oxygen adatoms), providing an opportunity to recover the pristine borophene surface for practical applications. Unfortunately, it is a daunting task to apply this approach to a variety of surface species on an intensely oxidized and chemically inhomogeneous sample, as (1) the manipulation conditions are highly dependent on the chemical structure and adsorption configuration of surface species; (2) tip-induced reactions take place very locally and are hardly performed on a large scale that is appropriate for device fabrication; (3) the case will be more complicated when the sample is exposed to ambient conditions due to the presence of water.

Reviewer #3 (Remarks to the Author):
In this manuscript, the authors used ultrahigh vacuum tip-enhanced Raman spectroscopy (UHV-TERS) to chemically interrogate individual oxygen adatoms on a boron monolayer, which is helpful for the deep understanding of the atomic-scale reactions. It is nice to see that the authors successfully manage to manipulate single atom and induce the single atom reaction with the substrate, and to chemically identify the formation of the bond via vibrational fingerprint. The article can be published after addressing the following questions: Reply: We appreciate the reviewer's very positive comments and recommendation of publication. We respond to the questions below and have revised the manuscript accordingly. Figure Figure 2i, from the description, it seems the authors assign both 205 and 189 cm -1 modes to the adsorbed single oxygen adatom. If this is the case, their trends of the TERS intensity profiles should not be inverse. I assume the authors would rather mean the 189 cm -1 mode comes from the mode of borophene under the influence of the neighboring O atom. If this is the case, the authors should clearly state this in the main text. Furthermore, the vibrational modes shown in Figure b-d can not be easily read. I would suggest the authors to increase the size or enlarge a portion of the model.

Reply:
We agree with the reviewer that if the 189 cm -1 mode solely originates from the adsorbed atomic oxygen (i.e., B-O bonding), it should show the same (instead of inverse) trend of intensity profile as that of the 205 cm -1 mode. However, as the reviewer speculated, the observed 189 cm -1 mode actually includes the mode of borophene lattice under the influence of the nearby O atom as demonstrated in Figure 2a, which shows a low intensity at the O atom site due to the higher tip position therein. In order to further clarify this issue, we would like to add a sentence in the main text.
In addition, following the reviewer's suggestion, we would like to increase the size of schematic models of Raman modes shown in Figure 2.

Revision:
We have added a sentence "Note that the evolution of the 189 cm -1 mode reflects the Raman intensity change of both bare and O-adsorbed v1/6 borophene" in the paragraph starting with "To see exactly how TERS spectra evolve across an oxygen adatom".
We have also increased the size of atomic models of Raman modes in the rearranged Figure 2. Figure 2k and i, the authors mentioned that the oval-shaped rather a circle came from the thermal drift during TERS collection. To validate this assumption, the authors should provide the drift value of this TERS system, so that the result in the image distortion can be estimated.

In the 2D TERS intensity mapping shown in
Reply: Typically, our TERS system has a thermal drift of 0.15-0.2 nm/min in X and Y directions under laser. However, the drift value is highly dependent on laser power (4-10 mW for the present research system), so sometimes it could be beyond the abovementioned drift range.

Revision:
We have added the typical drift value in the manuscript. Figure 3d, the cluster shows no 205 cm -1 features in the 2D TERS mapping. It is better to also show the complete TERS spectra on the cluster site and the oxygen adatom site.