Significant enhancement of proton conductivity in solid acid at the monolayer limit

Proton transport in nanofluidic channels is not only fundamentally important but also essential for energy applications. Although various strategies have been developed to improve the concentration of active protons in the nanochannels, it remains challenging to achieve a proton conductivity higher than that of Nafion, the benchmark for proton conductors. Here, taking H3Sb3P2O14 and HSbP2O8 as examples, we show that the interactions between protons and the layer frameworks in layered solid acid HnMnZ2O3n+5 are substantially reduced at the monolayer limit, which significantly increases the number of active protons and consequently improves the proton conductivities by ∼8 ‒ 66 times depending on the humidity. The membranes assembled by monolayer H3Sb3P2O14 and HSbP2O8 nanosheets exhibit in-plane proton conductivities of ~ 1.02 and 1.18 S cm−1 at 100% relative humidity and 90 °C, respectively, which are over 5 times higher than the conductivity of Nafion. This work provides a general strategy for facilitating proton transport, which will have broad implications in advancing both nanofluidic research and device applications from energy storage and conversion to neuromorphic computing.

This paper reports solid acids that have high proton conductivity.The proton conductivity of H3Sb3P2O14 is reported > 1 S cm-1 at 90 C and the authors claim that this may be a new strategy for fuel cells and other devices.The high proton conductivity of solid acids and metal phosphates is well-known from numerous previous studies.(e.g., Sossina Haile et al. Nature, 410, 910, 2001).However, despite the high conductivity of inorganic conductors, realizing the high performance of fuel cells using such materials seems to be much more challenging.After more than 20 years of research, the fuel cell performance using inorganic materials is still much inferior to the Nafion-based system.This is because there are more requirements than just conductivity for fuel cell membranes.The requirements include thin-film forming capability, stability in the presence of water, hydrogen impermeability, etc.
In this manuscript, the authors presented high conductivity, good stability under hydrated conditions, and film-forming properties.But they do not provide any single-cell performance.If they do not provide single-cell performance, the impact of this paper is greatly reduced.It is unfair to ask for very high fuel cell performance using these materials since the author's expertise is not the device testing.However, the authors should provide high-frequency resistance data with reasonable performance data (500-800 mW/cm2) to convince the readers that the proposed materials are promising.If the authors cannot provide such data, this material can be reported in a more material-specific journal.
Reviewer #3 (Remarks to the Author): The work has not been logically carried out, presented and compared with other materials.It will be necessary to compare the materials first before any clear conclusion can be made, because the membrane ones are more complicate which involves some engineering issues.An excellent work will need to examine the single crystal's directional dependence of the proton conductivity to figure out the mechanism.The work is quite far away from the quality of NC publication.

Reviewer #1:
The authors report ultrahigh proton conductivity in monolayer solid acid assembled membranes.With monolayer H3Sb3P2O14 and HSbP2O8 membranes as examples, they found extreme enhancement of proton conductivity at monolayer limit due to the significant reduction of interactions between protons and the layer framework.The proton conductivity of monolayer solid acid membranes is very impressive, exceeding those of the benchmark Nafion and other materials in a broad range of temperature and humidity.This work demonstrates a new concept for developing high-performance proton conductors, which have potential applications in many fields such as fuel cells and batteries.Furthermore, it provides new insight into the unique properties of 2D materials at the monolayer limit in addition to the known electronic, optical, thermal and mechanical properties.Thus, I am happy to recommend the publication of this manuscript in Nature Communications after a minor revision.
1.The authors show that the centrifugal force affects the thickness of H3Sb3P2O14 nanosheets.Does it affect the size?Can the size be tuned?Does the size of nanosheets affect the proton conductivity of the m-H3Sb3P2O14 membranes?
Response: We thank the reviewer very much for the valuable suggestions.
The average lateral sizes of H3Sb3P2O14 nanosheets are 0.92 μm, 0.93 μm, 1.20 μm, and 1.27 μm for a relative centrifugal force of 13000 g (g = 9.80 m s -2 ), 7312 -13000 g, 3250 -7312 g, and 1106 -3250 g, respectively (Figure R1).This indicates a very little influence of centrifugal force on the lateral size of H3Sb3P2O14 nanosheets.To reveal the relationship between the proton conductivity of H3Sb3P2O14 membranes and the lateral size of H3Sb3P2O14 nanosheets, m-H3Sb3P2O14 nanosheets with the same thickness (~1.0 nm, monolayer) and different average lateral size were synthesized by controlling the ultrasound process.As shown in Figure R1a and Figure R2, the average lateral sizes of m-H3Sb3P2O14 nanosheets are decreased from 0.92 μm, 0.38 μm, 0.22 μm to 0.15 μm by increasing the ultrasonic power and time.We characterized the proton transport properties of the m-H3Sb3P2O14 membranes assembled by these four kinds of nanosheets.The proton conductivities are 0.49  0.021 S cm -1 , 0.52  0.035 S cm -1 , 0.52  0.025 S cm -1 and 0.53  0.020 S cm -1 at 30 C and 100% relative humidity (RH) for the m-H3Sb3P2O14 membranes assembled by the nanosheets with lateral size of 0.92 μm, 0.38 μm, 0.22 μm and 0.15 μm, respectively (Figure R3).Moreover, the values are also almost the same at 60 C and 90 C.These results suggest that the influence of lateral size of m-H3Sb3P2O14 nanosheets on the proton conductivity is negligible for m-H3Sb3P2O14 membranes.
We have added these data and related discussions in the revised manuscript.2. The EIS test temperature for the m-H3Sb3P2O14 membranes ranges from 30 to 90 C.
What will happen for the proton conductivity at lower temperature?Is it still superior to that of Nafion at lower temperature?
Response: We thank the reviewer very much for the valuable suggestions.
The proton conductivities of m-H3Sb3P2O14 membranes and Nafion117 membranes (DuPont company) were measured by EIS at low temperature from 20 to -30 C.As shown in Figure R4, the obtained proton conductivity of m-H3Sb3P2O14 membranes ranges from 0.41  0.020 S cm -1 to 0.01  0.001 S cm -1 , which are much higher than those of Nafion 117 membranes at the same temperature (from 0.063  0.0038 S cm -1 to 0.007  0.0010 S cm -1 ).This result indicates that m-H3Sb3P2O14 membranes is superior to Nafion in proton conductivity at lower temperature.
We have added these data and related discussions in the revised manuscript.Response: We thank the reviewer very much for the insightful comment.
We have tested the proton conductivities of m-H3Sb3P2O14 and m-HSbP2O8 membranes with different thicknesses.As shown in Figure R5a, the proton conductivity of m-H3Sb3P2O14 membranes is 0.52  0.048 S cm -1 , 0.53  0.024 S cm -1 and 0.56  0.04 S cm -1 for a membrane thickness of 7 μm, 11 μm and 15 μm, respectively, at 30 C and 100% RH.Moreover, such membranes also show a very little change in proton conductivity at 60 C and 90 C.These results suggest that the thickness of m-H3Sb3P2O14 membranes has very little influence on their proton conductivity.The m-HSbP2O8 membranes show similar trend (Figure R5b).For instance, the proton conductivity is 0.68  0.005 S cm -1 , 0.67  0.010 S cm -1 and 0.60  0.016 S cm -1 for a membrane thickness of 5 μm, 8 μm and 11 μm, respectively.
We have added these data and related discussions in the revised manuscript.

Response:
We thank the reviewer very much for the valuable suggestion.
The stability of m-H3Sb3P2O14 membranes in strong acid were characterized by EIS, Raman and FT-IR spectroscopy.As shown in Figure R6a, m-H3Sb3P2O14 membranes shows no obvious change in proton conductivity after immersing in 10 M H2SO4 for 12 days.Moreover, both the Raman and FT-IR spectra remain almost the same before and after H2SO4 treatment (Figure R6b,c).These results demonstrate that m-H3Sb3P2O14 membranes are very stable in strong acid environment.
We have added these data and related discussions in the revised manuscript. 5.The unit for the y-axis is missing in the inset of Fig. 4g.

Response:
We have added the unit for the y-axis in the inset of Fig. 4g in the revised manuscript.

Reviewer #2:
This paper reports solid acids that have high proton conductivity.The proton conductivity of H3Sb3P2O14 is reported > 1 S cm -1 at 90 C and the authors claim that this may be a new strategy for fuel cells and other devices.The high proton conductivity of solid acids and metal phosphates is well-known from numerous previous studies.
(e.g., Sossina Haile et al.Nature, 410, 910, 2001).However, despite the high conductivity of inorganic conductors, realizing the high performance of fuel cells using such materials seems to be much more challenging.After more than 20 years of research, the fuel cell performance using inorganic materials is still much inferior to the Nafion-based system.This is because there are more requirements than just conductivity for fuel cell membranes.The requirements include thin-film forming capability, stability in the presence of water, hydrogen impermeability, etc.
In this manuscript, the authors presented high conductivity, good stability under hydrated conditions, and film-forming properties.But they do not provide any singlecell performance.If they do not provide single-cell performance, the impact of this paper is greatly reduced.It is unfair to ask for very high fuel cell performance using these materials since the author's expertise is not the device testing.However, the authors should provide high-frequency resistance data with reasonable performance data (500-800 mW/cm 2 ) to convince the readers that the proposed materials are promising.If the authors cannot provide such data, this material can be reported in a more material-specific journal.

Response:
We thank the reviewer very much for the insightful comments and valuable suggestions.
Due to the highly anisotropic structure, m-H3Sb3P2O14 membranes exhibit much higher proton conductivity along in-plane direction than through-plane direction.For instance, as shown in the manuscript, the in-plane proton conductivity of the membrane is ~0.38 S cm -1 at 95% RH and 60 C.However, the through-plane conductivity is ~7.7  10 -4 S cm -1 at the same test conditions (Figure R7).In the typical structure of a fuel cell such as PEMFC, large-area (~cm 2 ) proton exchange membrane is usually used to achieve a large electricity (e.g., Jiao, K., et al, Nature 595, 361, 2021;Kraysberg, A., et al, Energy Fuels 28, 7303, 2014;Steele, B. C. H., et al, Nature 414, 345, 2001).As for the m-H3Sb3P2O14 membranes, the area of the membrane for proton transfer is only ~10 -3 cm 2 when we try to use its superhigh in-plane proton conductivity since the thickness of the membranes synthesized by the vacuum method is on the order of 10 -4 cm, which is difficult to meet the requirement of a typical fuel cell.Moreover, the low through-plane proton conductivity limits the direct use of such membranes in fuel cell as Nafion does.The main finding of our work is that the interactions between protons and the layer frameworks in layered solid acid HnMnZ2O3n+5 are substantially reduced at the monolayer limit, which significantly increases the concentration of active protons and consequently improves the proton conductivity of membranes dramatically.Thus, one possible way to use this material for fuel cell is to make composites using monolayer H3Sb3P2O14 nanosheets and Nafion, which is beyond the scope of this work and deserves thorough studies in the future.
Alternatively, to improve the impact of this work, we have used m-HSbP2O8 membrane to construct an all-2D solid-state micro-supercapacitor (MSCs) to demonstrate its practical use, which fully utilized the highly anisotropic structure and proton transport characteristic of the membranes.The superhigh proton conductivity and electronic insulating nature enable monolayer solid acid membrane an excellent H + solid-state electrolyte and electrode separator in high-safety energy storage devices.
Figure R9d shows that the m-HSbP2O8-MXene MSCs exhibit good capacitive behaviors, where the cyclic voltammetry (CV) curves maintain a rectangular-like shape with a capacitance over 1.9 mF cm -2 at a superhigh scan rate of 300 mV s -1 , which is similar to the devices using H2SO4 and H2SO4/PVA as electrolytes (Figure R10a,b).In contrast, the Nafion-MXene and GO-MXene MSCs show poor electrochemical performances (Figure R10c,d).Despite the similar interconnected proton transport nanochannels in GO-MXene MSCs, they show a much lower capacitance of 0.37 mF cm -2 at 300 mV s -1 .Importantly, the m-HSbP2O8-MXene MSCs output a high operating voltage (0.9 V) (Figure R9e), which is over 1.5 times larger than those of the MSCs using other electrolytes, ranging from 0.4 V for GO membrane to 0.6 V for H2SO4/PVA gels (Figure R9f, Figure R11).As a result, the m-HSbP2O8-MXene MSCs show similarly high volumetric energy densities of 18.5 -18.3 mWh cm -3 with the corresponding power densities in the range of 1.7 -6.8 W cm -3 (Figure R9g).These performances are significantly better than those of H2SO4-, H2SO4/PVA-, Nafion-, and GO-MXene MSCs with the same electrodes.Furthermore, such all-2D MSCs can also be used to power the electronic devices even under repeated bending without packaging (Figure R9h,i, Supplementary movie 1), demonstrating the great potential of m-HSbP2O8 membranes for practical use in flexible solid-state micro energy storage devices.
We have added these data and related discussions in the revised manuscript.

Reviewer #3:
The work has not been logically carried out, presented and compared with other materials.It will be necessary to compare the materials first before any clear conclusion can be made, because the membrane ones are more complicate which involves some engineering issues.An excellent work will need to examine the single crystal's directional dependence of the proton conductivity to figure out the mechanism.The work is quite far away from the quality of NC publication.

Response:
We thank the reviewer very much for the kind comments and suggestions.
The main finding of our manuscript is that the interactions between protons and the layer frameworks in layered solid acid HnMnZ2O3n+5 are substantially reduced at the monolayer limit, which increases the concentration of active protons and consequently improves the proton conductivities of monolayer HnMnZ2O3n+5 assembled membranes significantly.We agree with the reviewer that the directional dependence of the proton conductivity might be a possible reason for the improved proton conductivity of m-H3Sb3P2O14 membranes compared to the pellets made by H3Sb3P2O14 particles considering that the latter ones have engineering issue and may not have a highly oriented nanochannels as m-H3Sb3P2O14 membranes do.Thus, as the reviewer pointed out, comparing the proton transport properties of H3Sb3P2O14 single crystals and m-H3Sb3P2O14 membranes is important to clarify the mechanism for the improved proton conductivity.Unfortunately, the synthesis of large single crystals of HnMnZ2O3n+5 is very challenging (Deniard-courant, S., et al, Solid State Ion. 27, 189, 1988) and no single crystals large enough for proton conductivity testing are available so far.Notably, all the membranes made by H3Sb3P2O14 nanosheets with four different average thicknesses (from 1.0 nm to 8.6 nm, including monolayers) have no engineering issue and they show significantly improved proton conductivity as the thickness of nanosheets reduces.Thus, to confirm the mechanism of significantly improved proton conductivity in m-H3Sb3P2O14 membranes, we thoroughly investigated the structure, in particular the orientation degree, of H3Sb3P2O14 membranes made by H3Sb3P2O14 nanosheets with four different average thicknesses.
As shown in Figure R12, R13, the four kinds of membranes show almost the same XPS, Raman and FT-IR spectra, indicating almost the same chemical composition, bonding and crystal structure of the nanosheets with different average thicknesses.
Moreover, these nanosheets have similar average lateral size from 0.92 to 1.27 m (Figure R1).
We then thoroughly characterized the orientation of H3Sb3P2O14 nanosheets in the membranes.In our experiments, all the membranes were fabricated by vacuum filtration, where the atmospheric pressure was used to squeeze the liquid towards the other side of the filter membranes and the nanosheets will be aligned by the directional water flow due to high aspect ratio (lateral size/thickness >10 3 ).SEM and XRD measurements clearly show that all the membranes have well-aligned layered structure (Figure R14 and R15).Wide-angle X-ray scattering (WAXS) pattern derives from the diffraction of an incident X-ray beam parallel to the surface of a membrane and has been widely used to quantitatively characterize the orientation degree of 2D nanosheets in the membranes (e.g., Wan, S., et al, Science 374, 96, 2021& Nat. Mater. 20, 624, 2021).We quantitatively characterized the orientation degree of nanosheets in the four kinds membranes made of H3Sb3P2O14 nanosheets with different average thicknesses by using WAXS (Figure R16).The orientation degree of H3Sb3P2O14 nanosheets was quantified by using the Herman's orientations factor (f), which was defined as Where <cos 2 > is the average value of the square of the cosine of the azimuthal angle for the (003) peak of the membranes, which was calculated as Where I() is the peak intensity at an azimuthal angle of .
As shown in Figure R15 and R16, all the membranes show a set of high-order diffraction peak along (00l) crystal plane and very similar f values from 0.97 to 0.99 for the (003) peak, confirming that these membranes have highly oriented structure along (00l) crystal plane with almost the same orientation degree.However, as shown in our manuscript, these four kinds of membranes show significantly different proton conductivities.Among them, the m-H3Sb3P2O14 membranes assembled by monolayers (~1.0 nm thick) show the highest proton conductivity over the investigated temperature range at 100% RH, which is about 2, 3 and 4 times larger than that of the membranes assembled from H3Sb3P2O14 nanosheets with an average thickness of ~1.4,3.1 and 8.6 nm, respectively.Moreover, the lateral size of nanosheets has negligible influence on the proton conductivity, as shown in Figure R3.Therefore, these results give strong evidence that the significantly improved proton conductivity in m-H3Sb3P2O14 membranes is attributed to the reduced thickness of the nanosheets rather than the directional dependence.Using solid state NMR, we revealed that the thickness dependence of proton transport in H3Sb3P2O14 is due to the greatly reduced interactions between protons and the layer frameworks at the monolayer limit, which results in substantial increase in the number of active protons and consequently the superhigh proton conductivities of m-H3Sb3P2O14 membranes.
We have added these data and related discussions in the revised manuscript.Figure R16 The azimuthal scan profiles for the (003) peak of 1.0 nm-H3Sb3P2O14 membranes (a), 1.4 nm-H3Sb3P2O14 membranes (b), 3.1 nm-H3Sb3P2O14 membranes (c), and 8.6 nm-H3Sb3P2O14 membranes (d).The insets are the corresponding WAXS patterns for an incident Cu-Kα X-ray beam parallel to the sheet plane.The derived Herman's orientation factors (f) for the (003) crystal plane of 1.0 nm-, 1.4 nm-, 3.1 nm-, and 8.6 nm-H3Sb3P2O14 membranes are ~0.987,0.966, 0.984, and 0.978, respectively, demonstrating that all the membranes have almost the same orientation degree along the (00l) crystal plane.

REVIEWERS' COMMENTS
Reviewer #1 (Remarks to the Author): I am happy to recommend the acceptance of this version since the authors have well addressed the comments raised by the reviewers.
Reviewer #2 (Remarks to the Author): Thank you for answering the Reviewers' questions.The authors answered clearly that the proposed material is inadequate for fuel cell applications due to the anisotropy.Instead, the authors added data for 2D solid-state micro-supercapacitors using the proposed materials.If the materials can be used with high performance in micro-supercapacitor applications, they may be publishable in Nature Communications.Unfortunately, this reviewer does not have enough knowledge to judge the micro-supercapacitor performance.This manuscript should be reviewed by experts in this area.In order to change direction to micro-supercapacitors, the authors should resubmit the manuscript after changing the title, introduction, and discussion.It should be clear that the advancement of these materials in the device performance compared with the state-ofthe-art.
Reviewer #3 (Remarks to the Author): The manuscript has been improved.However, this reviewer still cannot be fully convinced for the publication in NC.If other two reviewers are very positive, this reviewer is alright for the editor to make the decision of the acceptance though the work is too much engineering instead of science oriented.

Figure
Figure R1 The lateral size distributions of the H3Sb3P2O14 nanosheets obtained at a relative centrifugal force of 13000 g (g = 9.80 m s -2 ) (a), 7312 -13000 g (b), 3250 -7312 g (c), and 1106 -3250 g (d).The average lateral sizes are shown in the upper corner of each figure.

Figure
Figure R2 The lateral size distributions of the m-H3Sb3P2O14 nanosheets obtained at ultrasonic power and time of 160 W and 5 mins (a), 160 W and 20 mins (b), and 320 W and 20 mins (c).The average lateral sizes are shown in the upper corner of each figure.

Figure
Figure R3 The relationship between the proton conductivities of m-H3Sb3P2O14 membranes and the lateral sizes of m-H3Sb3P2O14 nanosheets at 100% RH and different temperatures.Error bars represent standard deviations.

Figure
Figure R4The proton conductivities of m-H3Sb3P2O14 membranes and Nafion117

Figure
Figure R5 The relationships between the proton conductivities and the thicknesses of m-H3Sb3P2O14 (a) and m-HSbP2O8 (b) membranes at 100% RH and different temperatures.Error bars represent standard deviations.

Figure
Figure R6 The proton conductivities (a), Raman spectra (b), and FT-IR spectra (c) of m-H3Sb3P2O14 membranes before and after immersing in 10 M H2SO4 for 12 days.Error bars represent standard deviations.

Figure
Figure R7The through-plane conductivities of m-H3Sb3P2O14 membranes at 95% RH

Figure R8 .
Figure R8.The LSV curves of fully hydrated m-HSbP2O8 membrane and 3.0 M H2SO4 solution at the polarization scanning of 2 mV s −1 .

Figure
Figure R9 Demonstration of all-2D flexible solid-state m-HSbP2O8-MXene MSCs.a, Photograph of a m-HSbP2O8-MXene MSC, showing good flexibility.b, Cross-sectional SEM images and the corresponding EDS mappings of the m-HSbP2O8 membrane (top)

Figure
Figure R12 XPS spectra of the membranes assembled by H3Sb3P2O14 nanosheets with different average thicknesses.a-d Survey XPS spectrum (a) and Sb 3d (b), P 2p (c) and

Figure
Figure R15 The XRD patterns of membranes assembled by H3Sb3P2O14 nanosheets with different average thicknesses, showing highly oriented structure along (00l) crystal planes.