Enhanced C-O Functionality on Carbon Papers Ensures Lowering Nucleation Delay of ALD for Ru towards Unprecedented Alkaline HER Activity

2300974 (1 of 8) © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com Enhanced CO Functionality on Carbon Papers Ensures Lowering Nucleation Delay of ALD for Ru towards Unprecedented Alkaline HER Activity Sitaramanjaneya M. Thalluri, Jhonatan Rodriguez-Pereira, Raul Zazpe, Bilal Bawab, Eva Kolíbalová, Ludek Jelinek, and Jan M. Macak* DOI: 10.1002/smll.202300974 Our previous reports provided an understanding on fine tuning of ALD process parameters to successfully decorate high surface area supports, such as titania nanotube layers with noble metals.[3,4] The low surface energy of carbon substrates and the difficulty in surface modification make ALD nucleation and growth a great hurdle for most of the noble metals with high surface energies.[5] Thus, ALD of Ru on carbon supports with cycles less than 30 has been a significant issue in recent years.[6] Having hydrophobic surface for carbon substrates is one of the main reason for Ru nucleation delay.[6] Hence initial introduction of another metal center on carbon supports as an adsorption and nucleation center for Ru has been a common practice. For instance, Palladium (Pd) cubic centers were considered as nucleation centers for Ru, thus realizing selective deposition of Ru on Pd by ALD.[6] Pt single atoms were also considered as nucleation centers to achieve selective deposition of Ru single atoms by ALD on Pt single atoms realizing PtRu single atom dimers.[7] Ru has drawn significant attention for hydrogen evolution reaction (HER) due to its stability in acidic and alkaline conditions (unlike Pt, whose long-term stability in alkaline media is not optimal), less expensive than Pt, and presenting comparable hydrogen bond energies to Pt.[8–10] Previous works reported that the synergistic effect of nitrogen-doped graphene and Ru nanoclusters, provided an onset potential of 8mV and a Tafel slope of 30mV dec−1.[11] in 0.1m KOH. Similarly, an exceptional performance related to single atoms along with Ru NPs on carbon nanowires resulted in an onset potential of 12mV in 1M KOH and a Tafel slope of 14 mV dec−1 in 0.1 M KOH.[12] Recently, innovative synthesis of hierarchical 4H/face-centered cubic (fcc) Ru nanotubes (NTs) provided an onset potential of 23mV with a Tafel slope of 29.3mV dec−1 in 1M KOH solution. The outstanding performance was related to the abundance of atomic steps/kinks, crystal phase boundaries, and a better electron mobility due to the Ru nanorod morphology.[13] In this context, current work deals with an innovative strategy to modify carbon papers (CP) and deploy ALD tool to deposit Ru species as single atoms to nanoparticles (NP) with high density and good conformality. The overall distribution The success in lowering the nucleation delay for Atomic Layer Deposition (ALD) of Ru on carbon surfaces is mitigated by constructive pretreatments resulting enhancement of CO functionality. Treatment of the carbon papers (CP) allowed Ru species deposition for minimum number of ALD cycles (25 cycles) with good conformality. The development of electrocatalysts from single atoms to nanoparticles (NPs) on conductive supports with low metal loadings, thus improving performance, is essential in electrocatalysis. For alkaline hydrogen evolution reaction, ALD decorated CPs with Ru exhibit low onset potentials of ≈4.7mV versus reversable hydrogen electrode (RHE) (at 10mAcm−2) and a high turnover frequency of 1.92 H2 s−1 at 30mV versus RHE. The Ru decorated CPs show comparable to higher catalytic activity than of Platinum (Pt) decorated CP also developed by ALD. The current representation of unfamiliar catalytic activities of Ru active centers developed by ALD, pave a bright and sustainable path for energy conversion reactions. ReseaRch aRticle S. M. Thalluri, J. Rodriguez-Pereira, R. Zazpe, B. Bawab, E. Kolíbalová, J. M. Macak Central European Institute of Technology Brno University of Technology Purkynova 123, 61200 Brno, Czech Republic E-mail: [email protected] S. M. Thalluri, J. Rodriguez-Pereira, R. Zazpe, J. M. Macak Center of Materials and Nanotechnologies Faculty of Chemical Technology University of Pardubice Nam. Cs. Legii 565, 53002 Pardubice, Czech Republic L. Jelinek Department of Power Engineering University of Chemistry and Technology Technická 5, Prague, 166 28 Prague-6, Czech Republic The ORCID identification number(s) for the author(s) of this article can be found under https://doi.org/10.1002/smll.202300974. 1. Introduction Atomic Layer Deposition (ALD) holds promise in the criteria of low catalyst loading possibilities and scaled uniform and conformal deposition on supports regardless of the shape.[1–3] © 2023 The Authors. Small published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. Small 2023, 19, 2300974 2300974 (2 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com of Ru species on CPs leads to an exceptional performance for HER in alkaline medium having onset potential values as low as ≈4.7 mV, a high turnover frequency (TOF) of 1.92 H2 s−1 at 30 mV and good stability, highlighting the advantages CP treatments and Ru deposition by ALD technique to improve the performance with low metal loadings. 2. Results and Discussions Carbon substrates are known to have defects that can act as nucleation centers. Further acidic treatments (AT) in conjunction with formalin treatments (FT) and annealing (A) currently employed, improves defects along with CO functionality over CPs. These were the main reasons for improving the Ru footprint on CPs resulting from the ALD process employed. The enhanced CO functionality in carbon supports was shown to improve the loading of noble metal Pt, as observed by C. PradoBurguete etal.[14] Additionally, the CO functionality on carbon provides key anchor points that demonstrated good dispersion during the decomposition – reduction step.[14] The effect of CO functionality on the dispersion of uniform Ru nanoparticles within the 1.5 to 1.8 nm range on carbon supports received a great deal of attention, and further studies were underway to gain an in-depth understanding.[15] An effect of CO in the CVD process has also been observed to improve the nucleation of Ru species on Si/SiO2.[16] An increased ratio of CO/CO (1.22) was observed according to the X-ray photoelectron spectroscopy (XPS) analysis for treatment induced activated CPs, evaluating peaks at 531.1 and 532.6 for CO and CO respectively,[17] (Figure1a). Figure S1 and Tables S1 and S2 (Supporting Information) provides detailed evaluation of XPS for CPs untreated and after treatments along with deconvolution of C 1s and O1s high resolution spectra. The developed surface functionality with a higher CO content on CP, reinforce lowering of the Ru nucleation delay related to the ALD process employed. On substrates of CPs (with AT+FT+A treatments), ALD depositions were planned for number of ALD cycles (NALD) of 25, 50, 100, and 200 Small 2023, 19, 2300974 Figure 1. XPS analysis for normalized O 1s high resolution spectra a) CP having details treatments and corresponding CO/CO ratios, SEM top-view images obtained from b) The blank CP after treatment, and CPs with c) 25, d) 50, e) 100, and f) 200 NALD of Ru. 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (3 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com with 2x NALD increments. The corresponding scanning electron microscopy (SEM) top-view images in Figure 1b–f confirmed the presence of discrete Ru NPs as bright spots on the carbon paper homogeneously distributed all over the CP suggesting the success in the deposition. As one could expect, the bright spots gradually increase in density and size with increase in the NALD from 25 to 200. Noteworthy, Ru NPs coalescence was observed upon 200 Ru NALD. X-ray diffraction (XRD) patterns shown in Figure S2 (Supporting Information) (along with details in supporting information (SI)) exhibit characteristic Ru diffraction peaks indicated the growing of crystalline Ru NPs with hexagonal crystal system indexed from PDF 00-06-0663.[18] More detailed understanding was provided from the transmission electron microscopy (TEM) analysis of the samples. High-resolution TEM (HR-TEM) imaging revealed that the asprepared CP 100c Ru sample consists of Ru NPs (Figure2a,f), with an average size of 2.9 ± 0.6 nm, and single Ru atoms (Figure S2, Supporting Information). The elemental distribution was performed by combination of high-angle annular dark-field scanning transmission electron microscopy (HAADFSTEM) imaging STEM-energy-dispersive X-ray (STEM-EDX) spectroscopy (Figure 2a–e). The observed distribution of Ru NPs from the STEM-EDX mapping suggested an overall very uniform deposition by ALD. The lattice fringes in HR-TEM image (Figure2f) exhibit a lattice d-spacing of 0.205nm, which is characteristic of the (0 11 1) plane from hcp Ru.[12] The fast Fourier transformation (FFT) pattern taken from a single Ru NP (Figure2g), where the marked spots signified the diffractions resulting from Ru (0 11 1) , ( 101 2) , and (1 210 ) crystal planes with d-spacing values of 0.205, 0.158, and 0.135nm, respectively.[12] STEM-EDX line scan on CP 25c Ru shown in Figure 2h was conducted to take advantage of isolated Ru NP on CP. The distribution of C, O, and Ru indicated the metallic nature of asdeposited Ru with increased amount of oxygen at the edges of Ru NP surface, suggesting surface oxidation, further confirm by XPS. The presence of Ru single atoms (Figure S3a–c, Supporting Information) in the CP 25, 50, and 100c Ru samples, clearly suggested the co-existence of Ru single atoms along with Ru NPs, as it was reported in recent literature.[12] Our hypothesis suggests that up to certain NALD (in our case 100 NALD), ALD process mainly yields Ru single atoms in comparison to the Ru NPs while upon higher NALD (in our case 200 NALD) the growth of Ru NPs is dominant. The formation of Ru NPs can be explained by the classical theory of adatom aggregation suggesting that the growth of NP’s on low surface energy substrates does not rely only as such on ALD cycles, but is also supported by diffusion of adsorbed noble metal species leading to the transformation of single atoms to clusters and further growth to larger NP’s.[19] Last, but not least, the ALD process parameters did not affect the underlying CP morphology. Surface elemental analysis of CP 100c Ru sample was carried out by XPS measurements (Figure S4 and Tables S3 and S4, Supporting Information). No other chemical species were identified than C, O, and Ru. The Ru chemical state revealed that the Ru on CP deposited by ALD was mostly in the metallic form. The complete description of the survey and high resolution from individual elements is provided in the Supporting Information. Being highly surface sensitive technique, XPS provides information on surface oxide layer that develops upon atmosphere exposure of the sample. The observation from the activity profile of linear scan voltammetry (LSV) (Figure3a), suggests that all the samples have excellent performance for HER in alkaline medium. The electrocatalytic activities were comparable to those observed in CP 50c Pt. To provide a comparative analysis, CP 100c Ru displayed the lowest onset potential value of 4.7mV. The onset potentials of all the samples including Pt were below 15mV, signifying the benefits of ALD in designing the noble metal based electrocatalysts. The Tafel slope analysis was performed within a decade current density range of 10 – 100mA cm−1. The calculated Tafel slopes were 44, 30, 20, and 15mV dec−1 for CP 25, 50, 100, and 200 c Ru respectively (Figure3b) compared to 56mV dec−1 for CP 50c Pt. Having such low Tafel slopes from CP 100c Ru and Cp 200c Ru suggests the HER mechanism follows a different mechanism rather than conventional Volmer-Tafel which can Small 2023, 19, 2300974 Figure 2. TEM analysis: a–e) A HAADF-STEM image of CP 100c Ru sample corresponding STEM-EDX elemental mapping providing CK, OK, RuL, and overlapped maps. f) A HR-TEM image of deposited Ru NPs used for lattice d-spacing measurements. The top-right insert with lattice spacing measurement is performed from a profile indicated by a “green line” and it indicates presence of Ru (0 111) plane with d= 0.205 nm. g) FFT analysis corresponding to a blue rectangle in (f) is revealing a Ru NP with hcp lattice oriented to [ 1011 ] zonal axis. h) STEMEDX elemental mapping and line scan analysis from CP 25c Ru sample revealing single Ru NPs at the edge of carbon paper. The elemental distribution is in net intensities and corresponds to the profile indicated by a green arrow in the insert (zoomed area of the original map). 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (4 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com be explained by considering higher intrinsic activity of active sites or TOF as described by previous report.[20] Variation in Tafel slopes when performing the analysis at different potential windows, results in different Tafel slopes[21] (Figure3c), for which details are available in Supporting Information. Charge transfer kinetics for HER in alkaline medium was evaluated by the electrochemical impedance spectroscopy (EIS), as shown in Figure3d. The Nyquist EIS plots were recorded at an overpotential of 14mV. A small incomplete semicircle for all the samples at higher frequencies relates to the surface properties of CPand it is not linked to HER activity of the sample.[22] All the Ru samples along with Pt, exhibited one prominent semicircle, suggesting that no competing electrochemical process other than HER was involved. The diameter of the semicircle represents the equivalent charge transfer resistance (Rct). Interestingly, Rct values for the samples range from 1.037 to 2.797 ohms (Ω). In our consideration all the Ru samples exhibited comparable HER activities from the observed Rct values. Considering the difficulty in evaluating electrochemical surface areas (ECSA) for noble metal electrocatalysts by potential dependent method (Figure S5, Supporting Information), copper underpotential deposition (Cuupd) analyses was performed to determine EC SACuupd and estimate active surface area and number of active sites for the samples.[23] Figure4a-e represents Cuupd for CP 25c, 50c, 100c, 200c Ru, and CP 50c Pt, respectively. The calculated EC SACuupd are 12.6, 33, 162.8, and 136.9 cm2 for CP 25, 50, 100, and 200c Ru, respectively. The number of active sites for the samples of Ru were 0.1654, 0.4369, 2.134, and 1.794× 1017 for CP 25c, 50c, 100c, and 200c Ru respectively supporting observed trend of active surface areas. The trend of active area and active sites collaborates with onset potentials form LSV. In general, the activity of an electrocatalyst is dependent on the intrinsic activity of an active site and the number of active sites. The intrinsic activity of active sites was evaluated from TOF analyses. The calculated TOF for CP 100c Ru (Figure S6a, Supporting Information) at an overpotential of 30mV versus RHE was 1.92 H2 s−1, whereas for CP 50c Pt the TOF was 0.16 H2 s−1 (Figure S6b, Supporting Information). The higher adsorption energies of RuH and RuOH compared to PtH and PtOH are one of the main reasons for the poor catalytic activity of Ru toward HER. Strategies to weaken those adsorption energies are of the utmost significance to improve HER performance. Engineering electrolytes by adding cations,[24] heterostructuring Ru with metal hydroxides/oxides,[25] Ru crystal structure engineering,[26,27] and nano particles in conjunction with single atoms of Ru on catalyst supports.[28] are some of the strategies that can effectively tune the RuH and RuOH bond energies. It is observed as a fact that improved adsorption/desorption of RuH on Ru single atoms and conductivity of carbon supports by Ru NP, lead to improved electrochemical performance of the hybrid catalysts.[28] The current ALD process on pretreated CPs developed Ru species consisting of both single atoms and NPs. The electrochemical results suggests that the enhanced activities are majorly achieved due to the single atoms in conjunction with surface atoms over NPs of Ru on the CP support, which constitute to the electrochemical active area. Further, potential dependent ECSA was also calculated for Cp 50c Pt sample giving a value of 4.225 cm2. The calculated surface area from H2 desorption peak and Cuupd ( EC SACuupd) stripping was 410 and Small 2023, 19, 2300974 Figure 3. Electrochemical characterizations in 1m KOH electrolyte for the different Ru and Pt samples. Corresponding a) Linear sweep voltammetry curves SV, b) Tafel slope evaluation employing a decade of change in current density from 10 to 100mAcm−2, c) Tafel slopes in different defined overpotential windows and d) EIS analysis of the samples. 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (5 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com 393.6 cm2 per loaded mass of Pt and the closeness of values demonstrated the reliability of the Cu stripping to determine active surface areas (Figure4e). The normalized activities with active surface area derived from Cuupd provide the specific activity per unit area. The values were 1.69, 1.54, 0.589, and 0.435mA cm−2 at 25mV for CP 25c, 50c, 100c, and 200c Ru respectively (Figure4f) This suggests that the sample with low active surface area (CP 25c Ru) surpasses specific activity performance per unit active area. The specific activity per unit area for CP 50c Pt was 0.054mA cm−2 per EC SACuupd suggesting ≈21 and 10 times lower compared to CP 25c Ru and CP 100c Ru. Inductively coupled plasma-atomic emission spectroscopy (ICP-OES) was carried out for all samples to determine the metal loadings. The estimated weight percent (wt.%) of Ru was 0.9, 2.3, 5.05 and 9.07 wt.% for 25c, 50c, 100c, and 200c Ru, respectively. Thus, as per the ICP-OES analysis, 5.05wt.% seems to be the optimal loading for Ru for best HER performance. The mass activity for CP@Ru samples were evaluated considering geometric area and EC SACuupd at an applied potential of 25mV (Figure S6c,d, Supporting Information). In both the scenarios, mass activity was highest for CP 25c Ru as observed for area specific activity. However, the activities found for the CP@Ru samples (25c, 50c, 100c, and 200c Ru) have no strong correlation with either the loading mass or the potential dependent ECSA values of the electrodes as observed in previous reports.[29] and it is the active surface area that is responsible for the real activity. Catalytic stability is critically important aspect for an electrocatalyst to suit for industrial applications. The stability of CP 100c Ru and CP 50c Pt were evaluated in 1M KOH electrolyte at a constant potential. Morphological and electrochemical analyses for ALD decorated CP 50c Pt are provided in Figure S7a–f (Supporting Information) for further reference. The stability of CP Ru (Figure5a) and CP Pt (Figure S7e, Supporting Information) were evaluated for an estimated duration of 10 h. There was a very minimum decrease in the activity Small 2023, 19, 2300974 Figure 4. Cuupd analysis for a) CP 25c Ru, b) CP 50c Ru, c) 100c Ru, d) 200c Ru, and e) CP 50c Pt samples. f) Specific activity per unit active area calculated at 25mV (area obtained from Cuupd measurements). 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (6 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com Small 2023, 19, 2300974 profile of both Ru and Pt. Corresponding change of onset potential ≈14 mV and 15.65 mV for CP 100c Ru and CP 50c Pt, respectively (Figure 5b; Figure S7f, Supporting Information). A minor increase in Rct was observed from 2.11 to 2.38 ohms for CP 100c Ru (Figure5c). Accelerated degradation tests were performed for the CP 100c Ru sample (Figure S8, Supporting Information) in a potential range of ± 100mV with a scan rate of 20 mV s−1 for 2500 cycles and the electrode was very stable with a small change in onset potential of 13.3mV. The variations can be attributed to the fact of the hanging electrode lost some bits and pieces of CP due to constant H2 bubbling, which might lead to loss noble metal from the electrode, however, the loss amount is minimum. After stability analysis of CP Ru samples, SEM and TEM characterization proved no considerable variation in the morphology of sample (Figure S9, Supporting Information). The lattice fringes in HRTEM image (Figure S9b, Supporting Information) exhibit a lattice d-spacing of 0.214nm, which is characteristic of the (0002) plane from hcp Ru.[12] XPS analysis after stability measurements showed signal from K which arises from the electrolyte solution involved. Considering fresh sample deposited at 275 °C and sample after immersing in 1 M KOH for prolonged duration during chronoamperometry analyses, resulting in loss of CP along with attached Ru metal, explains the decreased Ru content and increased oxidized component (to a minimal percentage due to physical exposure) observed in XPS. (Figure S4, Supporting Information). This suggests that the sample is stable physically and chemically after long term operation in alkaline environment. Figure S10 and Table S4 (Supporting Information) shows comparative analysis of onset potentials and Tafel slopes from published reports will provide an essence of current work to reveal the importance of surface pretreatments and ALD in developing highly efficient noble metal electrocatalysts for alkaline electrocatalysis. 3. Conclusion In summary, this work demonstrates that there is an enormous potential in noble metals (Ru in the present case) in energy conversion reactions that needs to be unearthed. Innovative treatments of highly conductive substrates (such as CP in the present case), shows great potential to decrease the nucleation delay commonly observed for noble metals on low energy surfaces in an ALD process. The current strategic chemical treatment of CP has minimized the nucleation delay of Ru deposited by ALD. The high density of single atoms and NPs of Ru decorated CP by ALD provided an exceptional onset potential of 4.7mV versus RHE at 10mAcm−2 with sufficiently high TOF of 1.92 H2s−1 at 30mV versus RHE and better Tafel kinetics in comparison to CP@Pt. The current success in Ru ALD supports strong future research work on single atom catalysis over different catalytic supports of interest. 4. Experimental Section Materials: Carbon papers (CT GDS310) (Fuel cell store); KOH, H2SO4 and HNO3,(Lach:ner); Formalin and CuSO4. 5H2O (Penta); Bis(ethylcyclopentadienyl)ruthenium(II), 98% (99.9%-Ru) (Strem (Europe)). CP Treatment and Ru ALD Deposition: ALD of Ru ranging from single atoms to nanoparticles was performed on CP of commercial Figure 5. Stability analysis. a) Chronoamperometric analysis in 1 M KOH conducted for 10h, b) LSV analysis of before and after CA for CP 100c Ru and c) EIS analysis before and after CA. 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (7 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com Small 2023, 19, 2300974 origin. The CP were initially treated with 1:3 ratio of HNO3:H2SO4 from 1 hour under sonication at 60°C to develop hydrophilic nature to the CPs. The as treated CPs were later thoroughly rinsed in distilled water several times. Subsequently, the CPs were soaked in 2% v/v of formalin under stirring for 1 h at room temperature. Care must be taken to avoid damage from the stirrer bar. Formaldehyde is known to act as a linking agent.[12] The resultant CPs were rinsed and subjected to annealing in air at 300°C @ 5°C min−1 for 1h. These treated CPs (activated CPs) were used in ALD to deposit Ru. ALD was performed by means of BENEQ TFS 200 thermal reactor with the stop-flow configuration. Ru deposition on the activated CPs was performed by using bis(ethylcyclopentadienyl)ruthenium(II) as Ru precursor, and Oxygen (O2) and H2 as coreactants.[30] The choice of Ru precursor relies on the literature studies and employability of different co-reactants to tune the process conditions.[30] The inert carrier gas was ultra-high purity N2 (99.9999%). The container for bis (ethylcyclopentadienyl) ruthenium(II) was held at 90 °C and the chamber temperature was kept at 275 °C. The gas lines were held ≈90°C with proper insulation. During the Ru ALD process, a 1 s pulse of Ru precursor (split into two 500ms pulses) with 10 s of exposure and 20 s of purge was employed. Ru pulse was followed by O2 pulse of 800ms with 10 s of exposure and 20 s of purge. The O2 combusts the EtCp* ligands from the Ru surface and from the surface of carbon and forms Ru oxide. By a further sequential 1.25 s pulse of H2 with 10 s of exposure and 20 s of purge the Ru oxide was reduced to metallic Ru.[30] Thus, the ALD sequence Ru(EtCp)2−O2−H2 comprises one ALD cycle, which applied different NALD results in metallic Ru species ranging from single atoms to NPs. In this work, four different NALD of Ru were performed: NALD= 25, 50, 100, and 200 cycles. Additional samples of CP with Pt (NALD= 50) were deposited following our previous published report.[4] All the samples were named as CP Xc M, where “X” represents NALD and “M” represent the deposited metal (Ru or Pt). Characterization: SEM analyses were carried out by field emission SEM (FE-SEM JEOL JSM 7500F) and image-aberration corrected TEM Thermofisher Scientific TITAN Themis 60–300 operated at 300keV and equipped with a HAADF detector for STEM imaging and with Super-X detector for EDX spectroscopy. XRD analysis was carried out using Panalytical Empyrean with Cu tube and Pixcel3D detector. The surface chemical composition of CP@Ru was monitored by XPS using AXIS Supra instrument (Kratos Analytical, Japan) with monochromatic Al Kα (1486.7 eV) excitation source. The resulting spectra were fitted using CasaXPS software. The binding energy scale correction was performed using the sp2 carbon at 284.0 eV for CP samples and the fermi level cut-off for CP 100c Ru samples. The quantitative analysis was performed using the elemental sensitivity factors provided by the manufacturer. ICPanalyses were performed using (Thermo-Fischer iCAP 7400) axial mode, RF power 1150W, 267.876, 266.161, and 240.272nm after dissolving the Cp@Ru in a solution of 1:3 mixture of HCl and HNO3 in a microwave reactor (Multiwave 5000 – Anton-Parr). HER Activity Measurements: The electrochemical HER measurements were performed using a three-electrode system using AUTOLAB (PGSTAT 204; Metrohm Autolab B. V.; Nova 1.10 software). Ru decorated CPs with a geometrical area of 1 cm2 were used as working electrodes. Graphite rod and Ag/AgCl (3M KCl) were used as counter and reference electrodes respectively. For LSV a scan rate of 2mV s−1 was employed. A 100% iR-correction was performed to evaluate Tafel slopes for all the samples. EIS measurements were performed in the frequency range of 10 mHz – 100kHz with an AC voltage amplitude of 10mV. The active surface area of the catalyst was determined by Cuupd method. Cuupd measurements were performed in acidic solutions with a scan rate of 10 mV sec−1 having 0.5 M H2SO4 and 20 mM CuSO4.5H2O while the base line was recorded employing 0.5 M H2SO4 electrolyte. To evaluate Cuupd, the electrode was polarized at 0.3V for 100 s followed by an LSV from 0.3 to 0.8V in a solution of 0.5M H2SO4 and 20mM CuSO4.5H2O. During the polarization process a monolayer of Cu was formed on the active Ru sites on the surface of CPs. The as-formed Cu monolayer was stripped by the following LSV performed from 0.3 to 0.8 V. The area under the peak of Cu stripping represents the charge involved in stripping the deposited Cu monolayer. The baseline to evaluate the charge associated with the stripping of absorbed Cu, was obtained by performing a CV from the same solution with no Cu salt added. Stability analyses were performed by chronoamperometry and accelerated degradation methods in 1M KOH on a geometric area of 1 cm2 of the samples was employed. Unless otherwise stated, all the potentials reported here are versus reversible hydrogen electrode (RHE), by converting the potentials measured versus Ag/AgCl (3M KCl). pH for 1M KOH is considered as 14 for all the calculations. Supporting Information Supporting Information is available from the Wiley Online Library or from the author. Acknowledgements The authors gratefully acknowledge support from European Union Horizon 2020 program (project HERMES, nr. 952184) and the Ministry of Education, Youth and Sports of the Czech Republic (projects LM2018103 and LM 2018110). The authors acknowledge CzechNanoLab Research Infrastructure supported by MEYS CR (LM2018110). Conflict of Interest The authors declare no conflict of interest. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Keywords alkaline hydrogen evolution reaction, atomic layer deposition, low surface energy substrates, ruthenium, single atoms Received: February 3, 2023 Revised: March 17, 2023 Published online: April 17, 2023 [1] N.Cheng, Y.Shao, J.Liu, X.Sun, Nano Energy 2016, 29, 220. [2] R. Zazpe, M. Knaut, H. Sopha, L. Hromadko, M. Albert, J. Prikryl, V. Gärtnerová, J. W. Bartha, J. M. Macak, Langmuir 2016, 32, 10551. [3] F. Dvorak, R. Zazpe, M. Krbal, H. Sopha, J. Prikryl, S. Ng, L.Hromadko, F.Bures, J. M.Macak, Appl. Mater. Today 2019, 14, 1. [4] V. C. Anitha, R. Zazpe, M. Krbal, J. E. Yoo, H. Sopha, J. Prikryl, G.Cha, S.Slang, P.Schmuki, J. M.Macak, J. Catal. 2018, 365, 86. [5] L. Assaud, N. Brazeau, M. K. S. Barr, M. Hanbücken, S. Ntais, E. A. Baranova, L.Santinacci, ACS Appl. Mater. Interfaces 2015, 7, 24533. [6] L. Zhang, Z. J. Zhao, M. Norouzi Banis, L. Li, Y. Zhao, Z. Song, Z. Wang, T. K. Sham, R. Li, M. Zheng, J. Gong, X. Sun, J. Mater. Chem. A 2018, 6, 24397. [7] L. Zhang, R. Si, H. Liu, N. Chen, Q. Wang, K. Adair, Z. Wang, J. Chen, Z. Song, J. Li, M. N. Banis, R. Li, T. K. Sham, 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 2300974 (8 of 8) www.advancedsciencenews.com © 2023 The Authors. Small published by Wiley-VCH GmbH www.small-journal.com Small 2023, 19, 2300974 M. Gu, L. M. Liu, G. A. Botton, X. Sun, Nat. Commun. 2019, 10, 1. [8] G.Zhao, Y.Jiang, S. X.Dou, W.Sun, H.Pan, Sci. Bull 2021, 66, 85. [9] S. Anantharaj, S. Noda, V. R. Jothi, S. C. Yi, M. Driess, P. W.Menezes, Angew. Chemie – Int. Ed. 2021, 60, 18981. [10] Y.Yang, Y.Yu, J.Li, Q.Chen, Y.Du, P.Rao, R.Li, C.Jia, Z.Kang, P.Deng, Y.Shen, X.Tian, Nano-Micro Lett. 2021, 13, 160. [11] R.Ye, Y.Liu, Z.Peng, T.Wang, A. S.Jalilov, B. I.Yakobson, S. H.Wei, J. M.Tour, ACS Appl. Mater. Interfaces 2017, 9, 3785. [12] B. Lu, L. Guo, F. Wu, Y. Peng, J. E. Lu, T. J. Smart, N. Wang, Y. Z.Finfrock, D.Morris, P.Zhang, N.Li, P.Gao, Y.Ping, S.Chen, Nat. Commun. 2019, 10, 631. [13] Q. Lu, A. L. Wang, H. Cheng, Y. Gong, Q. Yun, N. Yang, B. Li, B. Chen, Q. Zhang, Y. Zong, L. Gu, H. Zhang, Small 2018, 14, 1801090. [14] C. Prado-Burguete, A. Linares-Solano, F. Rodríguez-Reinoso, C. S. M.deLecea, J. Catal. 1989, 115, 98. [15] Y. Li, G. Lan, G. Feng, W. Jiang, W. Han, H. Tang, H. Liu, ChemCatChem 2014, 6, 572. [16] W.Liao, J. G.Ekerdt, Chem. Mater. 2013, 25, 1793. [17] P. G.Rouxhet, M. J.Genet, Surf. Interface Anal. 2011, 43, 1453. [18] S. Y. Tee, C. J. J. Lee, S. S. Dinachali, S. C. Lai, E. L. Williams, H. K.Luo, D.Chi, T. S.Andy Hor, M. Y.Han, Nanotechnology 2015, 26, 415401. [19] N. E.Richey, C.De Paula, S. F.Bent, J. Chem. Phys. 2020, 152, 040902. [20] J.Dai, Y.Zhu, Y.Chen, X.Wen, M.Long, X.Wu, Z.Hu, D.Guan, X.Wang, C.Zhou, Q.Lin, Y.Sun, S. C.Weng, H.Wang, W.Zhou, Z.Shao, Nat. Commun. 2022, 13, 1189. [21] T.Shinagawa, A. T.Garcia-Esparza, K.Takanabe, Sci. Rep. 2015, 5, 23955. [22] A. F.Sardinha, D. A. L.Almeida, N. G.Ferreira, J Mater Res Technol 2020, 9, 10841. [23] C. L.Green, A.Kucernak, J. Phys. Chem. B 2002, 106, 1036. [24] B. Huang, R. R. Rao, S. You, K. Hpone Myint, Y. Song, Y. Wang, W. Ding, L. Giordano, Y. Zhang, T. Wang, S. Muy, Y. Katayama, J. C.Grossman, A. P.Willard, K.Xu, Y.Jiang, Y.Shao-Horn, JACS Au 2021, 1, 1674. [25] Y.Wang, J.Wang, T.Xie, Q.Zhu, D.Zeng, R.Li, X.Zhang, S.Liu, Appl. Surf. Sci. 2019, 485, 506. [26] M.Zhao, Y.Xia, Nat. Rev. Mater. 2020, 5, 440. [27] Q. Zhang, K. Kusada, D. Wu, T. Yamamoto, T. Toriyama, S.Matsumura, S.Kawaguchi, Y.Kubota, H.Kitagawa, J. Am. Chem. Soc. 2022, 144, 4224. [28] J. N. Tiwari, A. M. Harzandi, M. Ha, S. Sultan, C. W. Myung, H. J. Park, D. Y. Kim, P. Thangavel, A. N. Singh, P. Sharma, S. S. Chandrasekaran, F. Salehnia, J. W. Jang, H. S. Shin, Z. Lee, K. S.Kim, Adv. Energy Mater. 2019, 9, 1900931. [29] S.Anantharaj, H.Sugime, S.Noda, J. Electroanal. Chem. 2021, 903, 115842. [30] J.Lu, J. W.Elam, Chem. Mater. 2015, 27, 4950. 16136829, 2023, 32, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smll.202300974 by Technical University In Brno, Wiley Online Library on [23/01/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License