Surface Defects, Ni3+ Species, Charge Transfer Resistance, and Surface Area Dictate the Oxygen Evolution Reaction Activity of Mesoporous NiCo2O4 Thin Films Erik Dubai,[a] Qingyang Wu,[a] Stefan Lauterbach,[b] Jan P. Hofmann,[a] and Marcus Einert*[a] For catalyzing the oxygen evolution reaction, earth-abundant materials with high activity and stability need to be developed. NiCo2O4 has been proven to show high OER activity, however facile and inexpensive techniques for preparation of this compound as mesostructured thin film, possessing a high surface area, is lacking. In this study, the sol-gel synthesis of nanocrystalline, mesoporous spinel NiCo2O4 thin films by dip- coating and soft-templating using the evaporation-induced self- assembly approach and utilizing the tri-block-copolymer Plur- onic® F-127 as structure-directing agent is reported. The morphology and crystallographic structure were thoroughly probed by various physicochemical characterization techniques collectively validating the development of uniform mesoporous NiCo2O4 architectures crystallizing exclusively in the cubic spinel phase after calcination in air at ether 300 °C, 400 °C, or 500 °C. The surface area of thin films increased from 300 °C to 400 °C owing to degradation of the organic template, while the growth of the mesopores from 400 °C to 500 °C resulted in significant decline of the overall (electrochemical) surface area. XPS investigations showed that the amount of octahedrally coordi- nated Ni3+ and defective (low-coordinated) oxygen species increased for decreasing calcination temperatures. The nano- morphology and presence of catalytically active surface sites of the mesoporous NiCo2O4 electrodes were correlated with the electrochemical properties, presenting that the overall surface area, Ni3+ content, charge transfer resistance, and amount of defective oxygen sites collectively control the OER performance. After an optimized annealing procedure at 300 °C and chrono- potentiometric analysis at 10 mA/cm2 for 1.5 h, a low over- potential of 330 mV vs. RHE at 10 mA/cm2 in alkaline solution was achieved. The results highlight the necessity of precise selection of the appropriate calcination temperature and tailoring of the nanostructure and electrochemical pre-treat- ment conditions of NiCo2O4 sol-gel thin films for adjusting the concentration of electrocatalytically active reaction sites. Introduction Fossil fuels are made accountable for the increased CO2 levels in the atmosphere, which is causing climate change leading to a multitude of environmental, economic and social challenges. In this context, electrolysis of water is an important technology to produce hydrogen as both a storage medium of energy and as energy source. The electrolysis process of water in an electrochemical cell can be divided into the anode and cathode reaction where the oxygen evolution reaction (OER) and hydro- gen evolution reaction (HER) taking place, respectively.[1] Since the OER reaction requires four electrons to be transferred and multiple intermediates are produced, the OER process is a kinetically sluggish process. As a consequence, the energy input that is required for driving the OER has to be lowered in order to enhance the overall water splitting efficiency. Currently, iridium- and ruthenium-based materials are used as OER catalysts in PEM electrolyzers and are the best performing catalysts (under acidic conditions) so far, but are undesirable due to their high cost and elemental scarcity.[2] Hence, low-cost alternatives such as transition metal oxides (TMOs) are investigated.[3] Among the TMOs, cobaltates (MCo2O4), and more specifically NiCo2O4, was found to be a promising candidate as electrocatalyst for driving the OER.[1] NiCo2O4 crystallizes in the spinel structure and is known to be a p-type semiconductor with direct bandgap properties.[4] In principle, a spinel-type crystal structure consists of an octahedra and a tetrahedra substructure formed by O2� anions forming a cubic packed lattice, this leads to 8 tetrahedra and 16 octahedra sites which are formed. Since NiCo2O4 forms an inverse spinel, eight octahedral sites are occupied by Co3+ and the remaining 8 are filled with Ni ions with mixed valence state (Ni2+/Ni3+), while 8 tetrahedral sites are filled up with Co ions with mixed valence state (Co2+/Co3+) cations.[5,6] Experimental findings suggest that the formation of an intermediate structure deviates from the ideal inverse structure by an altered cation distribution.[5] An investigation by Zhou et al.[7] elucidate the octahedral sites to be the active centres in promoting the OER [a] E. Dubai, Q. Wu, J. P. Hofmann, M. Einert Surface Science Laboratory, Department of Materials- and Geosciences, Technical University of Darmstadt, Peter-Grünberg-Straße 4, 64287 Darm- stadt, Germany E-mail: meinert@surface.tu-darmstadt.de [b] S. Lauterbach Institute for Applied Geosciences, Geomaterial Science, Technical University of Darmstadt, Schnittspahnstraße 9, 64287 Darmstadt, Germany Supporting information for this article is available on the WWW under https://doi.org/10.1002/cnma.202400242 © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH. This is an open access article under the terms of the Creative Commons Attribution Non-Commercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 113/124] 1 ChemNanoMat 2024, 10, e202400242 (1 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Articledoi.org/10.1002/cnma.202400242 www.chemnanomat.org http://orcid.org/0000-0001-6717-656X https://doi.org/10.1002/cnma.202400242 www.chemnanomat.org owing to localisation of the octahedrally coordinated cations predominantly at surface sites, thus making these cations preferentially accessible and reactive. In contrast, TMs located in tetrahedral positions were found to be rather inhibitive to OER, which can partially be attributed to the unfavourable geometric position. In addition, the orbitals of tetrahedrally located cations are inhibited to overlap with oxygen orbitals, which are considered as critical in the OER process.[7] Both Ni3+ and Co3+ cations are regarded as the most active sites, especially when located at octahedral sites.[8] This can be attributed to the stronger metal oxygen bonds, facilitating the OER. Since the Ni valence is dominating the OER activity in inverse spinel NiCo2O4, they are considered to be catalytically more active than its cobalt counterpart.[8] An even stronger correlation was found with oxygen vacancies as active sites, also promoting the OER activity.[9] This correlation can be most plausibly explained by the promotion of the energetically more favourable lattice oxygen mechanism (LOM) being described as prevalent OER route for cobaltite materials in alkaline media.[10] Zhang et al. demonstrated that NiCo2O4 exhibits a strong correlation with the amount of Ni ions in the +3 oxidation state and that the inclusion of Ni3+ creates unoccupied hole states in the electronic structure, which reduces the energy barrier for electron transfer, and thus, facilitates the OER kinetics.[11] With respect to fabrication of mesostructured NiCo2O4 in literature, only few reports are available for various energy applications: Mesoporous NiCo2O4 powders with well-defined mesoporous architectures were prepared by soft-templating utilizing Pluronic® P-123 as structure-directing-agent (soft- template) for application as supercapacitor.[12] Nanoporous NiCo2O4 on indium-doped tin oxide (ITO) glass was prepared by dip-coating using hexadecyltrimethyl-ammonium bromide (CTAB) as organic template and metal acetates as precursors.[13] Sol-gel-derived mesoporous NiCo2O4 nanofibers were fabricated by electrospinning and applied as counter electrode in a solar cell.[14] Dag et al. fabricated a series of mesoporous metal cobaltite (MCo2O4, M=Mn, Ni, and Zn) thin lyotropic liquid crystalline films (about 400 nm in thickness on fluorine doped tin oxide (FTO) glasses) by the molten salt-assisted self- assembly process for electrocatalysis of the OER.[15] The best- performing mesoporous MnCo2O4 electrode achieved an over- potential of 300 mV at 10 mA/cm2 in alkaline solution.[15] Wang et al. reported about obtaining uniformly grown arrays of NiCo2O4 nanoneedles via a hydrothermal reaction route. After deposition of hierarchical structured NiFe2O4 nanoflakes and annealing, the formed heterointerface lead to an enhanced interfacial polarization effect yielding in high electrocatalytic OER activity of 265 mV@50 mAcm� 2.[16] Hierarchical sheet-like W-NiCo2O4 structures were prepared by exchange-hydrolysis- reconstruction-annealing. The W6+ doping shifted the electron density from Co to W, thereby energetically lowering the adsorption energy for oxygen intermediates involved in OER catalysis on W sites.[17] To the best of our knowledge, there are no studies available on the synthesis of Pluronic® F-127 templated NiCo2O4 thin films deposited directly on conductive substrates by dip-coat- ing. However, especially nanostructured thin films prepared by soft-templating and evaporation-induced self-assembly (EISA) are of great interest, since the synthetic route is easy-to-handle and up-scalable.[18] Furthermore, the thin film electrocatalysts can be derived by deposition of the corresponding metal and organic precursors on conductive (polar and non-polar) sub- strates with controlled thicknesses. Subsequent calcination in air allows for transformation of the amorphous (composite) phase into a nanocrystalline mesoporous metal oxide framework.[19] This mesostructure can be considered as benefi- cial for driving the OER due to increased availability of catalyti- cally active reaction sites provided by the nanoporous network.[20–22] This work aims to study the fundamental impact of a nanoscale framework on the electrocatalytic performance of mesoporous NiCo2O4 thin film electrodes. The mesostructure was modified by variation of the annealing temperature with the result that the lower the calcination temperature was chosen, the better was the observed OER performance. The highest OER activity was achieved by annealing at 300 °C, leading to the highest concentrations of Ni3+ species and low- coordinated oxygen groups as well as the largest surface area. Results and Discussion Structural Characterization The SEM images in Figure 1 show the surface morphology of the NiCo2O4-300 (Figure 1 A), NiCo2O4–400 (Figure 1B), and NiCo2O4–500 (Figure 1C), in which the NiCo2O4-grains/particles appear as brighter features, while the darker spots can be ascribed to the presence of mesopores (defined to be 2 � 50 nm in size). The NiCo2O4-300 exhibit a surface structure with grain sizes of 7�3 nm and rather small mesopores of 4�2 nm in diameter. Due to insufficient thermal treatment for complete degradation of the organic (copolymer) template, the meso- pores are partially blocked and still closed at the surface. Note that Pluronic® F-127 is known to be not fully decomposed above 300 °C according to thermogravimetric analysis from literature.[23] For NiCo2O4-400, the mesoporous architecture appears to be more developed with grain/particle size of 12� 5 nm and open pores estimated to be 10�5 nm in size, both indicating a progressed, thermally-induced particle growth and polymer-template decomposition. These observations are in accordance with other studies on Pluronic® F-127 templated mesoporous metal oxides prepared by dip-coating and EISA process.[21,24,25] The NiCo2O4–500 sample possessed the largest grain/particle sizes (24�7 nm) and mesopores, which were found to be 13�6 nm in size. The mesopores were not of cylindrical shape (as compared to NiCo2O4-300 and NiCo2O4- 400) owing to coalescence of the nanocrystals, which built the pore walls and led to growth of larger nanocrystalline domains upon heating (Figure 1C). To evaluate the nanomorphology of both the surface and bulk structure of the mesoporous NiCo2O4, bright field mode transmission electron microscopy (TEM, Figure 2A) and high-resolution TEM (HRTEM, Figure 2 B) were carried out. From these pictures the porous nature of the bulk Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 114/124] 1 ChemNanoMat 2024, 10, e202400242 (2 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article film can be observed and verifies a homogenously developed through-pore structure of the entire NiCo2O4-400 sample. The selected area electron diffraction (SAED) patterns in Figure 2 A can be consistently indexed to the cubic spinel phase and the by inverted Fourier transformation (iFFT) derived images show a lattice spacing of 2.45 Å and 2.91 Å for the (311)- and (220)- plane (Figure S1 A), respectively, additionally supporting the formation of a pure spinel phase. The TEM data correspond well with the observations made by SEM and illustrate nanocrystal- line (spinel oxide) particles in the size range of 6–10 nm. To determine the crystallographic structure of the meso- porous NiCo2O4 thin films, grazing-incidence X-ray diffraction (GIXRD) was conducted. Figure 3 A exhibits the diffraction patterns of the mesoporous NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500 thin film presenting peaks at 31.0°, 36.6°, 38.3°, and 44.5° belonging to the reflexes of the (220)-, (311)-, (222)-, and (400)-plane, respectively. These Bragg peaks can be exclusively assigned to the cubic NiCo2O4 spinel phase (refer- ence JCPDS no. 20–0781)[26] with space group Fd3m symmetry, overall representing a crystallographic spinel structure with inverse characteristics. No secondary phases, such as NiO, were observed. With increasing calcination temperature, the peak intensity in the patterns rises, since the crystallization process within the NiCo2O4 bulk structure proceeds. Additionally, the full-width at half-maximum value decreases for higher anneal- ing temperatures, indicating the formation of larger particles the mesoporous NiCo2O4 network is composed of. As a consequence, the NiCo2O4-300 sample has to be considered as (partially) amorphous with respect to the sensitivity of XRD analysis (since no diffraction peaks are present). To exclude the presence of any side-phases, which are not sensitive to XRD analysis, Raman spectroscopy was performed. The Raman spectra (Figure S3 A) show phonon modes located around 186 cm� 1, 437 cm� 1, 522 cm� 1, and 661 cm� 1, which correspond to the F2g, E2g, F2g, and the A1g peak, respectively, and confirms in accordance with the XRD data, the formation of the cubic spinel crystal structure.[27] Raman bands of NiCo2O4- 400 and NiCo2O4-500 occur with the highest intensity, while for NiCo2O4-300 only the A1g and F2g peaks are detectable (Fig- ure S3 A). This observation can be explained by a more pronounced response of vibrational phonon modes due to more developed and larger crystalline domains within the mesoporous framework in NiCo2O4-400 and NiCo2O4-500. How- ever, the F2g mode appearing at 522 cm� 1 coincides with the strong Si signal located around 521 cm� 1, originating from the silicon substrate (see Figure S2). The A1g is the vibration mode of Co� O, more specifically oxygen atoms located at the octahedral sites associated with Co3+ ions in the tetrahedral sites.[28] The peaks at 437 cm� 1 and 522 cm� 1 (E2g and F2g) correspond to the combined vibration of oxygen present at octahedral and tetrahedral sites in the lattice. In addition, the F2g peak can be attributed to the Ni� O vibration mode.[28] Comparing NiCo2O4-500 and NiCo2O4-400, a shift of the E2g peak from 438 cm� 1 to 469 cm� 1 was observed, which can most likely be attributed to the influence of the differences in grain size known to affect the vibrational modes in energy.[29] To study the surface composition and oxidation states of the near-surface elements in detail, X-ray photoelectron spec- troscopy (XPS) was accomplished and the survey spectra of NiCo2O4-300, NiCo2O4-400, and NiCo2O-500 are demonstrated in Figure S3. The identified peaks in the survey spectra were assigned to the photoemission lines of nickel, cobalt, and oxygen in accordance with the elemental composition of the prepared samples. Furthermore, carbon and traces of tin were identified to be present, the latter one most likely originating from the calcination process in the furnace. For Nickel, the photoemission lines were located at 66.8 eV, 103.6 eV, and 855.6 eV and could be assigned to the 3p, 3 s, and 2p orbitals, respectively.[30] The peaks detected at 647.6 eV, Figure 1. SEM images of mesostructured NiCo2O4 thin films stabilized for 1 h at 200 °C and calcined at A) 300 °C, B) 400 °C, and C) 500 °C for 30 min in air. Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 115/124] 1 ChemNanoMat 2024, 10, e202400242 (3 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article 714.0 eV, and 780.5 eV were assigned to the Auger lines LMM, LMM1, and LMM2. Co 2p emission was found to be present at a binding energy of 780.5 eV and the O 1s peak was identified at 530.0 eV accompanied with the corresponding KLL Auger line observed at 975.0 eV.[30] Furthermore, carbon was found to be present on the sample surface illustrated by the C1s peak at 285.2 eV[31] and the corresponding Auger line KLL at 1226.4 eV (Figure S3). An overview of all detected elements with their respective photoelectron and Auger-electron spectral lines at their corresponding binding energies can be found in Table S1. Since the Auger contribution of nickel with L3M23M23 at 778 eV[32] is in vicinity of the position of the Co 2p2/3 peak,[25] superposition of both peaks are likely and thus difficult to distinguish. As a consequence, calculation of the surface composition stoichiometry of NiCo2O4 thin films was not considered, since the integral peak areas would be significantly affected by the above-mentioned interfering emission lines. The carbon signal in form of the C 1s peak can be predominantly related to surface adsorbed hydrocarbons and residual Plur- onic® F-127, utilized as organic template and known to thermally decompose in a temperature range between 200 °C and 400 °C (main mass loss derivative determined by thermog- ravimetric analysis in air was reported at 292 °C).[23,33] Accord- ingly, the slight decrease in carbon content for NiCo2O4-300 (19.1%) compared to NiCo2O4–500 (17.4%) can be explained by not yet fully degraded Pluronic® F-127 at calcination temper- atures of 300–400 °C. The major part of the C 1s peak (over 17% contribution to the total elemental surface composition) can be attributed to surface adsorbed hydrocarbons, typically observed in such concentrations for other structurally related, but chemi- cally different mesoporous metal oxide thin films.[21,25,34] The high-resolution spectra of the deconvoluted Ni 2p peak for all NiCo2O4 samples were measured between binding energies of 890 eV and 850 eV (Figure 4). Comparison of the deconvoluted Ni 2p peaks reveals various contributions from distinct chemical states changing significantly by applying different calcination temperatures. The 2p3/2 peak was fitted into a Ni2+-component located at 854.2 eV and a Ni3+ -component identified at 855.6 eV (in terms of BE).[35] As for the 2p1/2 peak, the Ni2+ contribution can be found at 871.7 eV and for Ni3+ at 873.4 eV.[35] Additionally, a shake-up satellite peak was detected and its maximum was found at 880.3 eV. Furthermore, two satellite peaks were located at 866.2 eV and 861.7 eV and assigned to the +3 and +2 oxidation state contributions, respectively, in accordance with literature.[36] Based on the integral peak area of the Ni2+ and Ni3+ contributions in the respective 2p1/2 and 2p3/2 peaks, the average contribution (in percentage) to the overall correspond- ing integral peak area was calculated (see Table 1). NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500 exhibit an average Ni2+ content of 13.5%, 14.3%, and 17.3%, and for the average Ni3+ contribution to the Ni 2p peaks 86.5%, 85.7%, and 82.7% was found, respectively. This means that the concen- tration of Ni2+ increases with rising annealing temperatures, Figure 2. A) STEM bright field picture of mesoporous nanocrystalline NiCo2O4 films visualizing the porous nature. The inset exhibits SAED patterns from the bulk area indicating the formation of the cubic spinel phase. B) High-resolution TEM image of a nanocrystalline NiCo2O4 thin film including an iFFT filtered image of the HRTEM image illustrating the lattice plane spacing of d(311) for the cubic spinel structure. Table 1. Percentual contribution of Ni2+ and Ni3+ components to the high-resolution Ni 2p peak. Sample Ni2+ average Ni3+ average NiCo2O4-300 13.5% 86.5% NiCo2O4-400 14.3% 85.7% NiCo2O4-500 17.3% 82.7% Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 116/124] 1 ChemNanoMat 2024, 10, e202400242 (4 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article while it decreases for Ni3+-species. As demonstrated in literature, Ni3+ is considered as the electrocatalytically more active species among the octahedrally coordinated metal sites[8] and facilitates the formation of highly catalytically active holes states within the electronic band structure.[11] Hence, the decrease in Ni3+ concentration for samples produced at higher calcination temperature should be regarded as an important trend negatively impacting the OER and will be discussed in detail in the electrochemistry section. Figure 5 shows the high-resolution spectra of the Co 2p photoemission lines of NiCo2O4-300, NiCo2O4-400, and NiCo2O4- 500 including the deconvolution of the respective components and recorded for binding energies between 775 eV and 810 eV. The Co 2p spectrum consists of a 2p1/2 and 2p3/2 peak (due to spin-orbital splitting), which can be deconvoluted into the Co2+ contribution located at 780.9 eV and 796.1 eV, and the Co3+ contribution located at 779.6 eV and 794.6 eV, respectively.[37] In addition, satellite peaks were detected at 784.6 eV belonging to Co2+, while the satellite maximum at 789.1 eV was assigned to Co3+ species. According to literature, the third satellite (for BE of 800 � 807 eV) can be generally attributed to cobalt being present in mixed valence states 2+ /3+ .[38] By calculation of the integral area of the specific contributions of the Co oxidation states in NiCo2O4 thin films (see Table 2), two important trends are obvious: with increasing calcination temperature the average content of Co3+ species increases from 32.2% (NiCo2O4-300) to 43.3% (NiCo2O4-500), while for Co2+ the Figure 3. A) GIXRD patterns and B) Raman spectra of the mesoporous NiCo2O4 thin films annealed at 300 °C (red), 400 °C (blue), and 500 °C (green) including the presentation of the corresponding lattice planes and phonon modes, respectively, both indicative for the cubic spinel phase. Figure 4. High-resolution XP spectra of Ni 2p comprising the corresponding deconvolution of peaks for A) NiCo2O4-300, B) NiCo2O4-400, and C) NiCo2O4- 500. Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 117/124] 1 ChemNanoMat 2024, 10, e202400242 (5 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article contribution (to the total Co 2p peak area) decreases from 67.8% (NiCo2O4-300) to 56.8% (NiCo2O4-500). In other words, more Co2+ gets oxidized to catalytically (more) active Co3+ at higher annealing temperatures, while for NiCo2O4-300 a 1 :2 ratio of Co3+/Co2+ was detected. A tabular comparison of all Co2+ and Co3+ contributions to the respective 2p1/2 and 2p3/2 peaks, including the calculated average values of the concen- trations, are summarized in Table S3. The O 1s signal can be fitted and deconvoluted into three peaks, which were declared as O1, O2, and O3 and identified at binding energies of 529.8 eV, 531.5 eV, and 532.8 eV (Figure 6), respectively. The O1 peak represents metal oxide bonds inside Figure 5. High-resolution XP spectra of Co 2p showing the corresponding deconvolution of peaks for A) NiCo2O4–300, B) NiCo2O4-400, and C) NiCo2O4- 500. Table 2. Percentual contribution of Co2+ and Co3+ components to the high-resolution Co 2p peak. Sample Co2+ average Co3+ average NiCo2O4-300 32.2% 67.8% NiCo2O4-400 32.9% 67.1% NiCo2O4-500 43.3% 56.8% Figure 6. High-resolution XP spectra of O 1s showing the corresponding deconvolution of peaks for A) NiCo2O4-300, B) NiCo2O4-400, and C) NiCo2O4- 500. Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 118/124] 1 ChemNanoMat 2024, 10, e202400242 (6 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article the lattice, while the O2 peak denotes to the contribution from low-coordinated, defective oxygen species, typically stemming from oxygen vacancies, surface adsorbed oxygen, and hydroxyl groups.[20,39] Especially oxygen vacancies, significantly contribu- ting to defective oxygen sites, have been reported to favor the oxygen evolution reaction substantially.[20,39] Furthermore, the O3 peak can be attributed to the multiple states with physical or chemisorbed water.[40] For NiCo2O4–500, the formation of an additional peak (O4) at 535.3 eV was observed, which did not match with any oxygen species occurring in NiCo2O4. Due to the evolution of the mesopore network (larger mesopores with increasing temperature) and the high binding energy, it might be possible that molecular oxygen accumulated and phys- isorbed in the voids/cavities, as it was also found for other NiCo2O4 materials.[41] Most interestingly, the concentration of the O2 components increases substantially with decreasing annealing temperature from 45.9% (NiCo2O4-500) to 48.0% (NiCo2O4-400) to 56.7% (NiCo2O4-300), indicating that also the content of low-coordinated oxygen species increases signifi- cantly (see Table 3). This allows the assumption that with higher calcination temperatures the amount of oxygen vacancies increases, since the defective oxygen species (O2 contribution) indirectly represents this kind of surface defects. Since oxygen deficiency is proven to be advantageous for electrocatalytic processes,[42] a correlation between the electrochemical activity and the concentration of defective oxygen will be regarded as impor- tant control parameter for the discussion of the temperature- dependent OER performance. We highlight that the determi- nation of the elemental concentrations by the presented fitting method have to be considered as an estimation approach, indicating only relative trends among the control samples. We note that such simple peak fitting models for the features of overlapping 2+ and 3+ species provide estimated values and an unambiguous determination and quantification would require synchrotron-based analysis, which will be scope of a perspective study. For further characterization of the optoelectronic properties of the NiCo2O4 thin films, UV-vis spectroscopy was accom- plished (Figure S4). All NiCo2O4 films possess strong absorption for wavelengths lower than 800 nm and show an increase of absorbance of photons in the order NiCo2O4-500 > NiCo2O4- 400 > NiCo2O4-300 (Figure S4 A). This is most likely due to enhanced crystallinity and enlarged particle sizes resulting in enhanced optical absorptivity. The influence of the film thick- ness on the absorption behavior can be excluded, since the film thickness decreases with increasing annealing temperature and were found to be 197�13 nm, 157�3 nm, and 138�21 nm for NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500 (determined as averaged values by profilometry, Figure S5), respectively. In principle, the valence band of NiCo2O4 is composed of over- lapping O 2p orbitals, while the Ni 3 d and Co 3 d orbitals form the conduction band.[43] The absorbance spectra were trans- formed into Tauc plots (Figure S4 B) assuming direct optical transition (n = 2). The extrapolation of the graphical linear fitting to (αhv)2 = 0, leads to two direct optical band gap energies. The two intercept values can be attributed to excitation of electrons from Co 3d-t2g orbitals to the partially filled Co 3d-eg orbital states. Therefore, two band gap energies can be assigned to the co-existence of high-spin and low-spin states of Co3+ in the mesoporous NiCo2O4 thin films.[43] For all NiCo2O4 samples the first low-energy intercept is located around 1.6 eV. The second intercept was determined to be at 2.5 eV, 2.4 eV, and 2.3 eV for NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500, respectively. The experimental data agree with literature values showing optical transitions located at 1.55 eV and 2.85 eV for comparable NiCo2O4 structures (deposited at 450 °C).[44] A trend towards increasing optical band gap energies for decreasing calcination temperatures emerges. Since lower annealing temperatures for the synthesis of NiCo2O4 thin films lead to much smaller grain/particle sizes within the mesoporous framework, as confirmed by SEM investigations (see Figure 1), the increase in band gap energy can be related to the nanoconfinement effect, also observed in other TMOs com- posed of crystallites in the sub � 10 nm range.[45] Decreasing particle sizes should lead to an increase of the ground energy state of confined electrons in order to satisfy Heisenberg’s uncertainty principle according to the relationship: Δx ·Δp��h/2. As a consequence, the (band gap) energy increases for smaller, and thus more confined NiCo2O4 nanocrystals.[46] Electrochemical Characterization To elucidate the electrochemical performance of the nano- crystalline mesoporous NiCo2O4 thin films as OER electro- catalysts, cyclic voltammetry (CV, Figure 7A), linear sweep voltammetry (LSV, Figure 7B), Tafel plot analysis (Figure 7C), electrochemical impedance spectroscopy (EIS, Figure 7D), scan- rate dependent voltammetry (Figure S6), and chronopotentiom- etry (Figure S7) were conducted. Figure 7A shows the CVs of NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500 measured within a potential window of 0.6 V to 1.5 V vs. RHE in 1 M KOH (pH �13.6) and a scan rate of 50 mV/s. According to Zhu et al.[47] the redox peaks for sol-gel-derived spinel NiCo2O4 originate from the electron transfer of the cations to OH � species, resulting in change of oxidation states for Ni from +2 to +3 (Equation (1)) and for Co from +3 to +4 (Equation (2)), corresponding to the following reversible reac- tion balances: NiCo2O4 þ OH� þ H2OÐ NiOOHþ 2CoOOHþ e� (1) CoOOHþ OH� Ð CoO2 þ H2Oþ e� (2) Table 3. Percentual contributions of distinct oxygen species (O1, O2, O3, and O4) to the deconvoluted O 1s peak. Sample O1 O2 O3 O4 NiCo2O4-300 35.0% 56.7% 8.3% – NiCo2O4-400 43.2% 48.0% 8.8% – NiCo2O4-500 40.4% 45.9% 8.3% 5.4% Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 119/124] 1 ChemNanoMat 2024, 10, e202400242 (7 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article In agreement with literature,[48] the Ni2+/Ni3+ and Co3+/Co4+ transitions are represented as one oxidation peak at 1.35 V and one reduction peak at 1.15 V vs. RHE due to superposition of both (Ni and Co) redox processes, being reported to be too close to each other for differentiation by CV analysis.[47] Hence, the NiOOH, Ni(OH)2 and CoO2 species can be expected to be the catalytically active species for the OER in accordance with literature.[49] Oxyhydroxide formation and the presence of CoO2 has also been suggested as reactive compounds for structurally comparable mesoporous Co3O4 thin films, also prepared by the EISA-based dip-coating method.[21] For the NiCo2O4-500 sample a significant decrease in redox peak intensity was observed, indicating a decreasing tendency for oxidation and reduction reactions of the electrochemically active species (cations), which might also be interpreted as a decline in surface reactivity. To evaluate the OER activity of the various NiCo2O4 thin films, LSV curves were recorded between 1.0 V to 1.9 V vs. RHE with a 5 mV/s scan rate (Figure 7 B). At the characteristic geometrical current density of 10 mA/cm2, the overpotential for NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500 was determined to be 0.43 V, 0.47 V, and 0.49 V, respectively, presenting a pivotal trend toward the OER performance of mesoporous NiCo2O4 thin films: the OER activity increases for decreasing calcination temperatures. The underlying mesoporous NiCo2O4 thin films possess higher overpotentials compared to chemically and structurally related samples from literature.[15] In this reference, mesoporous NiCo2O4 thin films annealed at 300 °C were found to possess an overpotential of η10 =253 mV in 1 M KOH,[15] a difference which is (at least) partially attributed to the difference in film thickness (the OER activity is known to be function of the film thickness for mesoporous metal oxide thin films).[50] The LSV data (Figure 7 B) are supported by Tafel plot evaluation presented in Figure 7 C. Comparison of the Tafel slopes indicates the same trend as LSV does, revealing values of 332 mV/dec, 289 mV/dec, and 168 mV/dec for NiCo2O4-300, NiCo2O4-400, and NiCo2O4- 500, respectively. Consequently, the determined Tafel slopes suggest that the required activation energy for driving the OER increases for higher annealing temperatures of NiCo2O4 films, which is consistent with the LSV data. As the electrochemical stability is a profound material characteristic for (industrial) long-term applications, chronopotentiometry was conducted at a constant applied current density of 10 mA/cm2 for all mesoporous NiCo2O4-thin films for 1 hour (Figure S8) and for Figure 7. A) CV curves recorded with a scan rate of 50 mV/s, B) LSV plots collected at 10 mV/s, C) Tafel plots, and D) electrochemical impedance spectroscopy as Nyquist plot of A) NiCo2O4-300 (red), B) NiCo2O4-400 (blue), and C) NiCo2O4-500 (green). Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 120/124] 1 ChemNanoMat 2024, 10, e202400242 (8 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article the best-performing NiCo2O4-300 sample for 8 hours (Figure S9). As illustrated in Figure S8, electrocatalytic activity loss was found to be 1.5% and 2.0% for NiCo2O4-500 and NiCo2O4-400, respectively, while for NiCo2O4-300 an improvement of 0.7% was observed over one hour of CP analysis, showing the same trend as the LSV data. Hence, the calcination temperature does not only affect the OER activity, but also the OER stability of NiCo2O4 thin film catalysts. Therefore, a prolonged stability test was conducted for NiCo2O4-300 over 8 hours (Figure S9), presenting a decrease in OER activity of only 0.6%. Interestingly, the overpotential of NiCo2O4-300 decreased to only 330 mV vs. RHE after 1.5 h of chronopotentiometric experiments at 10 mA/ cm2. The enhanced stability and activity of NiCo2O4-300 is most likely assigned to surface reconstructing (e.g. by formation of NiOOH and CoO2 species)[51] and self-healing processes, which are facilitated at amorphous/less-crystalline surface structures,[52] resulting in enhanced electrochemical stability. For comparison, pristine FTO substrates were investigated under the same experimental conditions. The data (Figure S7) confirm that the FTO substrates have neglectable impact on the electrochemical performance of the mesoporous NiCo2O4 thin film electrocatalyst, since the overpotential was found to be about 1800 mV vs. RHE at 10 mA/cm2. To interpret and understand the temperature-dependent trends in OER performance, two further control parameters, the electrochemical active surface area (ECSA) and the charge transfer resistance � both known to affect the OER activity substantially[11,20] � were investigated. The ECSA was deter- mined, according to the method of McCrory et al.,[53] by scan- rate dependent voltammograms collected between 0.725 V and 0.925 V vs. RHE for scan rates ranging from 10 mV/s to 70 mV/s. The double layer capacitance was determined by linear fitting of the ~j-values, resulting in 4.5 mF, 6.9 mF, and 1.0 mF for NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500, respectively. The ECSA values were calculated by assuming 40 μF/cm2 as specific capacitance,[53] which lead to 113 cm2, 172 cm2, and 25 cm2 for NiCo2O4-300, NiCo2O4-400, and NiCo2O4-500, respectively. How- ever, we note that no standardization for the determination and application neither of the Helmholtz double layer capaci- tance nor the specific capacitance CS exist,[54] making exact comparison of the ECSA of distinct spinel transition metal oxides difficult. The empirically attained trend for ECSA of NiCo2O4 thin films can be generally explained based on the thermal decomposition behavior of Pluronic® F-127, which is known not to be fully degraded at 300 °C, and thus blocking the pores.[23] These data agree with the observed surface concentration of carbon as analyzed by XPS (see Figure S3). The highest carbon content was found for NiCo2O4-300 with 19%. The highest ECSA, however, was evaluated at 400 °C (172 cm2), a temperature at which the copolymer is known to be fully decomposed (carbon concentration reached a constant value of 17%) and where the mesopores are smaller compared to NiCo2O4-500, but larger compared to NiCo2O4-300. With increas- ing temperature (500 °C), a thermally induced growth of nano- crystals can be observed, resulting in 3-dimensional extension of the mesopore size (see SEM analysis Figure 1) and, in turn, a significantly lower ECSA of 25 cm2 for NiCo2O4-500. Since the charge transfer resistance (Rct) at the electrode- electrolyte interface is known to be an important descriptor for interpretation of the OER activity, the value was analysed by EIS and illustrated in form of Nyquist plots in Figure 7D. Rct demonstrates a clear trend towards decreasing interfacial charge transfer resistances for lower calcination temperatures and were identified to be 36.7 Ω, 15.5 Ω, and 6.0 Ω for NiCo2O4- 500, NiCo2O4-400, and NiCo2O4-300, respectively. This trend (NiCo2O4-300 < NiCo2O4-400 < NiCo2O4-500) was confirmed by additional EIS analysis performed at 1.4 V vs. RHE (Figure S10) and is in accordance with literature, reporting that shorter migration paths owing to smaller nanoparticle sizes (within the pore walls) result in reduced charge transfer resistances.[25] The RΩ value, (first intercept of semicircle with abscissa) represent- ing the total cell resistance, was amounted to 10.8 Ω (NiCo2O4- 300), 10.2 Ω (NiCo2O4-400), and 11.9 Ω (NiCo2O4-500). Hence, RΩ is comparable for all samples within an error range of �1 Ω with respect to the average value (11.0 Ω). Taken the ECSAs, the charge transfer resistances, concen- trations of Ni2+, Ni3+, Co3+, Co4+, and defective oxide groups into account (see Table 1, 2, and 3), the OER performance of mesoporous NiCo2O4 in dependence of the annealing temper- ature can be explained: the higher the (electrochemical) surface area, the more catalytically active surface sites are available, resulting in higher geometrical current densities.[20] Importantly, although the NiCo2O4-400 possess significantly higher ECSA (compared to NiCo2O4-300 and NiCo2O4-500), it shows only the second highest OER activity (after NiCo2O4-300). This deviation can be explained by having a closer look at the defective oxygen and Ni3+ concentrations determined by XPS (see Table 2 and 3). It can be stated that NiCo2O4-300 has the highest concen- tration of defective oxygen groups, which also compromise oxygen vacancies known to be highly electrocatalytically active reaction centers owing to appropriate adsorption energetics of OH � (according to Sabatier‘s law oxygen intermediates are bonded at the catalyst’s surface ideally to a moderate extent - not too strong and not too weak, i. e., close to thermoneutral).[1,2,20,39] Additionally, NiCo2O4-300 contains the largest amount of Ni3+ cations, also reported to dictate the OER performance of spinel NiCo2O4 electrocatalysts.[8] Therefore, the lower overpotential of NiCo2O4-300, compared to NiCo2O4-400, has to be ascribed to the absolute amount of electrocatalytically active and available OER-sites, which is the product of the ECSA and the combined concentration of oxygen vacancies and Ni3+ centres per area (ECSA � ([Ox] + [Ni3+])/cm2=#OER-active sites), predetermining the overall OER activity. This explanation approach is further supported by two studies: Liu et al.[8] describes that Ni in octahedral site is critical for the OER catalysis and that the Ni valence dominates the OER perform- ance (compared to Co3+). Additionally and in agreement, Cui et al.[11] shows that the OER activity of NixCo3-xO4 strongly correlate with the presence of Ni3+ species. It is reported that Ni3+ ions shift the occupied valence band maximum upwards (by 0.27 eV) and increases the hybridization of O 2p with Ni 3 d and Co 3 d orbitals. This modification of the electronic band structure enhances the adsorption of OH (rate-determining Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 121/124] 1 ChemNanoMat 2024, 10, e202400242 (9 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article step) on NiCo2O4 resulting in faster OER kinetics. In essence, the OER performance of mesoporous sol-gel-derived NiCo2O4 thin films is controlled by 1) the ECSA, 2) concentration of defective oxygen species, 3) amount of (octahedrally coordinated) Ni3+ sites present within the first few surface-layers, and 4) the charge transfer resistance at the electrode-electrolyte interface. Conclusions In this study, mesoporous NiCo2O4 thin films were produced, for the first time, by a soft-templating-based dip-coating approach in humidity-controlled atmosphere using a commercially avail- able copolymer (Pluronic® F-127) as structure-directing agent and metal nitrates as precursors. Temperature-dependent analyses were conducted to study the impact of calcination temperature on the OER activity of the nanocrystalline and mesostructured thin films. The NiCo2O4 samples fabricated at annealing temperature between 300 °C and 500 °C were all found to crystallize in a cubic spinel phase and composed of a mesoporous network well accessible for aqueous electrolytes. Thermally-induced growth of both the nanocrystals/grains (building unit of the pore walls) and the mesopores occur for increasing treatment temperatures. The electrochemical surface area increased from 300 °C to 400 °C due to ongoing decom- position of the copolymer (blocking accessibility to pores) and decreased from 400 °C to 500 °C owing to continuous growth of the mesopores. XPS analysis revealed a trend in the develop- ment of redox species on the surface of the thin films for decreasing calcination temperatures, which was an increase in both Ni3+ and under-coordinated oxygen species. Both param- eters showed substantially impact on the OER activity. Together with the surface area and the charge transfer resistance, these four control parameters were evaluated to determine the OER performance of sol-gel-derived mesoporous NiCo2O4, while the concentration of Co3+ was found to be of secondary impact. The underlying results showcase the importance of detailed and thorough surface analysis and provide a facile synthesis strategy for the preparation of high-surface-area NiCo2O4 thin films as stable OER electrocatalyst. Experimental Section Sample Preparation. Materials. Cobalt (II) nitrate hexahydrate (99.999%, Sigma-Aldrich), nickel (II) nitrate hexahydrate (99.999%, Sigma-Aldrich), citric acid (�99.5%, Sigma-Aldrich), ethanol (�99.8%, Carl Roth), 1 M KOH (Carl Roth), H-(C2H4O)x-(C3H6O)y-(C2H4O)z-OH (Pluronic® F-127, Sig- ma-Aldrich) and water (obtained from Milli Q water purification system, Merck KGaA). Synthesis. In principal, the synthesis of mesoporous NiCo2O4 thin films on polar and non-polar substrates by dip-coating was adapted from literature,[43,55] In detail, the dip-coating solution was obtained by mixing 1.2 ml EtOH with 1.1 ml H2O and dissolving 45 mg Pluronic® F-127 and 91.3 mg of citric acid (serving as complexing agent). Afterwards, 96.9 mg of Ni(NO3)2 · 6H2O and 194.0 mg Co- (NO3)2 · 6H2O were added. In the next step, the solution was stirred for an hour by a magnetic mixer under vigorous stirring for homogenization purposes until a transparent and clear solution was obtained. As for the substrates, silicon (100) wafer (from Siltronic) and fluorine-doped tin oxide (FTO) coated glass slides (from XOP glass) were utilized. All substrates were cleaned first with acetone, then with ethanol. Afterwards, remaining (carbonaceous) impurities were removed by a UV-Ozone-cleaner (from Ossila). The installed synthetic quartz UV lamp was operated at wavelengths of 185 nm and 254 nm for 10 minutes. Finally, dip-coating was performed with the cleaned substrates in a self-made chamber under controlled humidity conditions (15�3% relative humidity). Since sol-gel reactions highly depend on the humidity (by undergoing hydrolysis and condensation reactions) the control of this parameter is of great importance. Upon dip- coating, the substrate was immersed into the prepared solution and extracted at a constant withdrawal speed between 4 mm/s and 8 mm/s depending on the desired film thickness to be produced. During the withdrawal step, the evaporation-induced self-assembly (EISA) process takes place and leads to the formation of micelles within the formed precursor sol film. The micelles act as organic template for the latter mesopore formation. The prepared samples were transferred into a preheated muffle furnace (from Nabertherm, Model LT3/11) heated to a temperature of 125 °C for desorption of water. Thermal stabilization of the amorphous composite thin film was initiated by heating from 125 °C to 200 °C with a heating ramp of 10 °C/min and held for an hour. Afterwards, the temperature was increased using a heating ramp of 10 °C/min until the final calcination temperature, which was ether 300 °C, 400 °C, or 500 °C was reached. These samples are assigned to NiCo2O4-300, NiCo2O4- 400, and, NiCo2O4-500, respectively. At these temperatures, the thin films were annealed for 30 minutes in air. The heating process for temperatures above 400 °C should theoretically lead to the complete degradation of Pluronic® F-127 according to thermogra- vimetric analysis from literature.[23] At the same time, the thermal decomposition of the organic template (micelles) occurs, the amorphous inorganic-organic composite is transformed into a nanocrystalline NiCo2O4 framework. Structural characterization. Images of the topography from the various samples were obtained using high-resolution scanning electron microscopy (HR-SEM). The machine used for this purpose was a Philips XL30 FEG with a resolution of up to 2 nm. The measurement was conducted at UHV at an acceleration voltage of 30 keV. Prior SEM measurements the samples were sputter-coated with platinum, creating a few nm thick platinum layer on the surface, in order to increase electric conductivity. For transmission electron microscopy (TEM) measurements, flakes of the thin film were scraped off the substrate and collected in a small sample glass. After adding 2 mL of ethanol, the suspension was dispersed with the help of an ultrasonic bath. The obtained dispersion was allowed to settle for 20 s to separate large particles. Two to three droplets of the upper part of the dispersion were applied on a carbon coated copper grid (holey type, Plano GmbH, Wetzlar, Germany). To prevent charging from the incident electron beam in the TEM, the samples were coated with a thin carbon film (carbon coater MED 010, Bal-Tec AG, Balzers, Liechtenstein). Examination of the samples was carried out on a FEG TEM (JEM2100F, JEOL Ltd., Tokyo, Japan). XRD measurements were conducted on a diffrac- tometer D8-advance from Bruker-AXS. The accuracy of the meas- urement is given as � 0.01° (in 2Ɵ) from the manufacturer. As X- ray source Cu Kα radiation with a wavelength of 1.5406 Å was used. Diffraction patterns were recorded for 2-Theta angles between 10° to 60° and with a scan rate of 0.02° min� 1 by using an energy dispersive detector. For profilometry a Dektak XT Advanced System machine was used, the measurement was conducted with a Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 122/124] 1 ChemNanoMat 2024, 10, e202400242 (10 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article diamond tip (stylus) attached to a needle to determine small height differences. The measurement was conducted at two distinct spots of the sample. The average value was taken as the final film thickness. An instrumental error of several nanometers has to be considered. Spectroscopic characterization. XPS measurements were con- ducted with an ULVAC-PHI VersaProbe II machine at the DArm- stadt’s Integrated SYstem for BATtery Research“ (DAISY-BAT) located at the Institute of Materials Science at TU Darmstadt. The XPS spectra were measured at UHV with a pressure of p <5×10� 9 mbar, a pass energy of 23.5 eV and a step size of 0.1 eV. Monochromatic Al Kα radiation (hv =1486.6 eV) was used and directed to the sample at 45° with respect to the surface normal. Binding energy calibration was performed by setting the Au 4f7/2 emission of Au foil to 84.0 eV. Determining the corresponding peaks in the survey spectra and peak fitting was accomplished with the MultiPak software and CasaXPS. The background was sub- tracted by the Shirley method. Raman spectra were acquired with a confocal Raman microscope using 10×long working distance (LWD) objective. As microscope a SENTERRA II from Bruker was operated with a 531 nm laser. The measurement range was set between 50 cm� 1 to 2000 cm� 1. Optical characterization. Optical properties of the prepared samples were evaluated using a Lambda 950UV/Vis/NIR spectrom- eter from PerkinElmer. The absorption spectra were collected in transmission mode between 300 nm and 900 nm at a data interval of 2 nm and scan speed of 500 nm/min. The transmission spectra were converted into absorption spectra and the corresponding Tauc plots were calculated for direct optical transitions. The optical band gap energies were estimated by determining the intercept of the graphical fitting of the linear part of the data points with the x- axis. Electrochemical Characterization. For electrochemical investiga- tions, the prepared NiCo2O4 thin films were assembled in an electrochemical cell (Zahner cell, model PECC- 2) using 1 M KOH solution as electrolyte (pH �13.6). The electrochemical analyses were carried out by using a GAMRY Interface 1000 E potentiostat measuring the assembled electrochemical cell in a three-electrode setup. Hg/HgO was used as reference electrode, the prepared thin films were applied as working electrodes, and a Pt wire was utilized as counter electrode. Linear sweep voltammetry (LSV) measurement was conducted in the range of 0.9 V to 1.9 V vs. RHE with a scan rate of 10 mV/s and step size of 1 mV. The iR-correction of the LSV curves was executed by subtraction of the electrolyte resistance derived experimentally from impedance spectroscopy. The cyclic voltammetry (CV) scans were measured between 0.9 V to 1.9 V vs. RHE for 50 cycles with a scan rate of 50 mV/s and a step size of 1 mV. For evaluation of the Helmholtz double layer capacitance, scan rate-dependent CV curves were carried out between 0.72 V to 0.92 V vs. RHE with a step size of 1 mV. The applied scan rates were 20 mV/s, 30 mV/s, 40 mV/s, 50 mV/s, 60 mV/s, and 70 mV/s. Based on the differential current densities evaluated for distinct scan rates, the double-layer capacitance (DLC) was determined after the method/protocol of McCrory et al..[53] The obtained DLC value can be converted into the electrochemical surface area (ECSA), by dividing the value by 40 μF/cm2, assumed as average CS specific capacitance value used in alkaline media.[56] Electrochemical impedance spectroscopy (EIS) was recorded at 1.4 V and 1.6 V vs. RHE between 300 mHz and 100 kHz at AC modulation of 10 mV using a resolution of 25 points per decade. Chronopotentiometry measurements were conducted at a current density of 10 mA/cm2 vs. RHE for 1 h or 8 h. Acknowledgements The authors thank Jean-Christophe Jaud for assistance with XRD experiments, Julia Gallenberger for help with Raman spectroscopy and Chuanmu Tian for fruitful discussions. Marcus Einert acknowledges both the Deutsche Forschungsgemein- schaft (DFG, German Research Foundation, Walter Benjamin Program to M. Einert) with project no. 469377211 and the Federal Ministry of Education and Research (BMBF) within the project TWOB under Award Number 033RC036 for financial support, and the support from China Scholarship Council (CSC) of No. 202208320036 (Qingyang Wu). Open Access funding enabled and organized by Projekt DEAL. Conflict of Interests There are no conflicts to declare. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Keywords: Mesoporous · Oxygen evolution reaction · Sol-gel · Thin film · Electrocatalysis [1] N.-T. Suen, S.-F. Hung, Q. Quan, N. Zhang, Y.-J. Xu, H. M. Chen, Chem. Soc. Rev. 2017, 46, 337–365. [2] M. Tahir, L. Pan, F. Idrees, X. Zhang, L. Wang, J.-J. Zou, Z. L. Wang, Nano Energy 2017, 37, 136–157. [3] H. Wang, K. H. L. Zhang, J. P. Hofmann, F. E. Oropeza, J. Mater. Chem. A 2021, 9, 19465–19488. [4] I. T. Papadas, A. Ioakeimidis, G. S. Armatas, S. A. Choulis, Adv. Sci. 2018, 5, 1701029. [5] T.-C. Chang, Y.-T. Lu, C.-H. Lee, J. K. Gupta, L. J. Hardwick, C.-C. Hu, H.- Y. T. Chen, ACS Omega 2021, 6, 9692–9699. [6] X. Xu, C. Mellinger, Z. G. Cheng, X. Chen, X. Hong, J. Appl. Phys. 2022, 132, 020901 . [7] Y. Zhou, S. Sun, C. Wei, Y. Sun, P. Xi, Z. Feng, Z. J. Xu, Adv. Mater. 2019, 31, 1902509. [8] Y. Liu, P. Liu, W. Qin, X. Wu, G. Yang, Electrochim. Acta 2019, 297, 623– 632. [9] H. Park, B. H. Park, J. Choi, S. Kim, T. Kim, Y.-S. Youn, N. Son, J. H. Kim, M. Kang, Nanomaterials 2020, 10, 1727. [10] Ö. N. Avcı, L. Sementa, A. Fortunelli, ACS Catal. 2022, 12, 9058–9073. [11] M. Cui, X. Ding, X. Huang, Z. Shen, T.-L. Lee, F. E. Oropeza, J. P. Hofmann, E. J. M. Hensen, K. H. L. Zhang, Chem. Mater. 2019, 31, 7618–7625. [12] J. Wang, Y. Xiong, X. Zhang, J. Mater. Sci. 2017, 52, 3678–3686. [13] Y. Liu, N. Wang, C. Yang, W. Hu, Ceram. Int. 2016, 42, 11411–11416. [14] C. Zhang, L. Deng, P. Zhang, X. Ren, Y. Li, T. He, Dalton Trans. 2017, 46, 4403–4411. [15] A. Amirzhanova, N. Akmansen, I. Karakaya, O. Dag, ACS Appl. Energ. Mater. 2021, 4, 2769–2785. [16] X. Wang, J. Luo, Y. Tuo, Y. Gu, W. Liu, S. Wang, Y. Zhou, J. Zhang, Chem. Eng. J. 2023, 457, 141169. [17] J. Luo, X. Wang, Y. Gu, S. Wang, Y. Li, T. Wang, Y. Liu, Y. Zhou, J. Zhang, Chem. Eng. J. 2023, 472, 144839. [18] a) S. Dilger, M. Trottmann, S. Pokrant, ChemSusChem 2019, 12, 1931– 1938; b) D. R. Ceratti, B. Louis, X. Paquez, M. Faustini, D. Grosso, Adv. Mater. 2015, 27, 4958–4962. [19] a) B. Smarsly, M. Antonietti, Euro. J. Inorg. Chem. 2006, 6, 1111–1119; b) B. Smarsly, D. Fattakhova-Rohlfing in Solution Processing of Inorganic Materials, John Wiley & Sons, Inc Hoboken 2009, 283–306; c) E. Celik, Y. Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 123/124] 1 ChemNanoMat 2024, 10, e202400242 (11 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article https://doi.org/10.1039/C6CS00328A https://doi.org/10.1039/C6CS00328A https://doi.org/10.1016/j.nanoen.2017.05.022 https://doi.org/10.1016/j.nanoen.2017.05.022 https://doi.org/10.1039/D1TA03732C https://doi.org/10.1039/D1TA03732C https://doi.org/10.1021/acsomega.1c00295 https://doi.org/10.1016/j.electacta.2018.11.111 https://doi.org/10.1016/j.electacta.2018.11.111 https://doi.org/10.3390/nano10091727 https://doi.org/10.1021/acs.chemmater.9b02453 https://doi.org/10.1007/s10853-016-0658-1 https://doi.org/10.1016/j.ceramint.2016.04.071 https://doi.org/10.1039/C7DT00267J https://doi.org/10.1039/C7DT00267J https://doi.org/10.1021/acsaem.1c00064 https://doi.org/10.1021/acsaem.1c00064 https://doi.org/10.1016/j.cej.2022.141169 https://doi.org/10.1016/j.cej.2022.141169 https://doi.org/10.1016/j.cej.2023.144839 https://doi.org/10.1002/cssc.201802645 https://doi.org/10.1002/cssc.201802645 https://doi.org/10.1002/adma.201502518 https://doi.org/10.1002/adma.201502518 Ma, T. Brezesinski, M. T. Elm, Phys. Chem. Chem. Phys. 2021, 23, 10706– 10735; d) T. Brezesinski, M. Groenewolt, N. Pinna, H. Amenitsch, M. Antonietti, B. Smarsly, Adv. Mater. 2006, 18, 1827–1831; e) M. Einert, A. Waheed, D. C. Moritz, S. Lauterbach, A. Kundmann, S. Daemi, H. Schlaad, F. E. Osterloh, J. P. Hofmann, Chem. Eur. J. 2023, 29, e202300277. [20] M. Einert, M. Mellin, N. Bahadorani, C. Dietz, S. Lauterbach, J. P. Hofmann, ACS Appl. Energ. Mater. 2022, 5, 717–730. [21] Q. Wu, M. Mellin, S. Lauterbach, C. Tian, C. Dietz, J. P. Hofmann, M. Einert, Mater Adv 2024, 5, 2098–2109. [22] Q. Wu, A. Alkemper, S. Lauterbach, J. P. Hofmann, M. Einert, Energy Adv. 2024, 3, 765–773. [23] K. Kirchberg, R. Marschall, Sustain. Energy Fuels 2019, 3, 1150–1153. [24] a) L. Q. Wagner, E. Da Prates Costa, C. Glatthaar, F. Breckwoldt, M. Zecca, P. Centomo, X. Huang, C. Kübel, H. Schlaad, M. Kriechbaum, Chem. Mater. 2023, 35, 9879–9899; b) K. Kirchberg, S. Wang, L. Wang, R. Marschall, ChemPhysChem 2018, 19, 2313–2320. [25] M. Einert, A. Waheed, S. Lauterbach, M. Mellin, M. Rohnke, L. Q. Wagner, J. Gallenberger, C. Tian, B. M. Smarsly, W. Jaegermann, Small 2023, 19, 2205412. [26] Y. Li, X. Wu, S. Wang, W. Wang, Y. Xiang, C. Dai, Z. Liu, Z. He, X. Wu, RSC Adv. 2017, 7, 36909–36916. [27] Y. Zhang, Y. Ru, H.-L. Gao, S.-W. Wang, J. Yan, K.-Z. Gao, X.-D. Jia, H.-W. Luo, H. Fang, A.-Q. Zhang, J. Electrochem. Sci. Eng. 2019, 9, 243–253. [28] M. Pathak, P. Mutadak, P. Mane, M. A. More, B. Chakraborty, D. J. Late, C. S. Rout, Mater Adv 2021, 2, 2658–2666. [29] a) X. Yang, X. Yu, Q. Yang, D. Zhao, K. Zhang, J. Yao, G. Li, H. Zhou, X. Zuo, Ceram. Int. 2017, 43, 8585–8589; b) C. Suchomski, et al., J. Mater. Chem. A. 2017, 5, 16296–16306; c) M. Schröder, S. Sallard, M. Böhm, M. Einert, C. Suchomski, B. M. Smarsly, S. Mutisya, M. F. Bertino, Small 2014, 10, 1566–1574. [30] J.-G. Kim, D. L. Pugmire, D. Battaglia, M. A. Langell, Appl. Surf. Sci. 2000, 165, 70–84. [31] M. Einert, C. Wessel, F. Badaczewski, T. Leichtweiß, C. Eufinger, J. Janek, J. Yuan, M. Antonietti, B. M. Smarsly, Macromol. Chem. Phys. 2015, 216, 1930–1944. [32] J. Chastain, R. C. King Jr, Perkin-Elmer Corporation 1992, 40, 221. [33] N. U. Khaliq, J. Lee, S. Kim, D. Sung, H. Kim, Pharmaceutica 2023, 15, 2102. [34] K. Brezesinski, J. Haetge, J. Wang, S. Mascotto, C. Reitz, A. Rein, S. H. Tolbert, J. Perlich, B. Dunn, T. Brezesinski, Small 2011, 7, 407–414. [35] M. Ahmad, B. Xi, Y. Gu, S. Xiong, Inorg. Chem. Front. 2021, 8, 3740–3747. [36] T.-H. Ko, K. Devarayan, M.-K. Seo, H.-Y. Kim, B.-S. Kim, Sci. Rep. 2016, 6, 20313. [37] W. Huang, Y. Cao, Y. Chen, J. Peng, X. Lai, J. Tu, Appl. Surf. Sci. 2017, 396, 804–811. [38] J. F. Marco, J. R. Gancedo, M. Gracia, J. L. Gautier, E. I. Ríos, H. M. Palmer, C. Greaves, F. J. Berry, J. Mater. Chem. 2001, 11, 3087–3093. [39] S. Peng, et al., J. Am. Chem. Soc. 2018, 140, 13644–13653. [40] D. Ouyang, J. Xiao, F. Ye, Z. Huang, H. Zhang, L. Zhu, J. Cheng, W. C. H. Choy, Adv. Energy Mater. 2018, 8, 1702722. [41] A. I. Stadnichenko, L. S. Kibis, D. A. Svintsitskiy, S. V. Koshcheev, A. I. Boronin, Surf. Eng. 2018, 34, 1–5. [42] a) D. Kubba, I. Ahmed, P. Kour, R. Biswas, H. Kaur, K. Yadav, K. K. Haldar, ACS Appl. Nano Mater. 2022, 5, 16344–16353; b) J. Kang, D. Tang, Y. Liu, Y. Huang, W. He, Y. Liu, X. Ji, W. Li, J. Li, Ind. Eng. Chem. Res. 2023, 62, 3882–3888. [43] L. Hu, L. Wu, M. Liao, X. Hu, X. Fang, Adv. Funct. Mater. 2012, 22, 998– 1004. [44] I. Saafi, R. Dridi, A. Mami, J. Ben Naceur, A. Amlouk, R. Chtourou, M. Amlouk, Appl. Phys. A 2021, 127, 651. [45] a) L. Korell, S. Lauterbach, J. Timm, L. Wang, M. Mellin, A. Kundmann, Q. Wu, C. Tian, R. Marschall, J. P. Hofmann, Nanoscale Adv. 2024, 6, 2875– 2891; b) M. Einert, R. Ostermann, T. Weller, S. Zellmer, G. Garnweitner, B. M. Smarsly, R. Marschall, J. Mater. Chem. A. 2016, 4, 18444–18456. [46] M. Einert, P. Hartmann, B. Smarsly, T. Brezesinski, Sci. Rep. 2021, 11, 17687. [47] Y. Zhu, X. Ji, Z. Wu, W. Song, H. Hou, Z. Wu, X. He, Q. Chen, C. E. Banks, J. Power Sources 2014, 267, 888–900. [48] a) S. K. Meher, P. Justin, G. Ranga Rao, ACS Appl. Mater. Interfaces. 2011, 3, 2063–2073; b) R. N. Singh, J. -F. Koenig, G. Poillerat, P. Chartier, J. Electrochem. Soc. 1990, 137, 1408. [49] J. Gallenberger, H. M. Fernández, A. Alkemper, M. Li, C. Tian, B. Kaiser, J. P. Hofmann, Catal. Sci. Technol. 2023, 13, 4693–4700. [50] M. Bernicke, E. Ortel, T. Reier, A. Bergmann, J. Ferreira de Araujo, P. Strasser, R. Kraehnert, ChemSusChem 2015, 8, 1908–1915. [51] N. Weidler, J. Schuch, F. Knaus, P. Stenner, S. Hoch, A. Maljusch, R. Schäfer, B. Kaiser, W. Jaegermann, J. Phys. Chem. C 2017, 121, 6455– 6463. [52] T. Guo, L. Li, Z. Wang, Adv. Energy Mater. 2022, 12, 2200827. [53] C. C. L. McCrory, S. Jung, J. C. Peters, T. F. Jaramillo, J. Am. Chem. Soc. 2013, 135, 16977–16987. [54] P. Connor, J. Schuch, B. Kaiser, W. Jaegermann, Zeitschrift für Physikali- sche Chemie 2020, 234(5), 979–994. [55] B. Eckhardt, E. Ortel, D. Bernsmeier, J. Polte, P. Strasser, U. Vainio, F. Emmerling, R. Kraehnert, Chem. Mater. 2013, 25, 2749–2758. [56] J. Yan, L. Kong, Y. Ji, J. White, Y. Li, J. Zhang, P. An, S. Liu, S.-T. Lee, T. Ma, Nat. Commun. 2019, 10, 2149. Manuscript received: April 10, 2024 Revised manuscript received: July 12, 2024 Accepted manuscript online: August 7, 2024 Version of record online: October 2, 2024 Wiley VCH Samstag, 02.11.2024 2411 / 371535 [S. 124/124] 1 ChemNanoMat 2024, 10, e202400242 (12 of 12) © 2024 The Authors. ChemNanoMat published by Wiley-VCH GmbH Research Article https://doi.org/10.1039/D1CP00834J https://doi.org/10.1039/D1CP00834J https://doi.org/10.1002/adma.200600154 https://doi.org/10.1021/acsaem.1c03190 https://doi.org/10.1039/D3MA01054F https://doi.org/10.1039/D4YA00026A https://doi.org/10.1039/D4YA00026A https://doi.org/10.1039/C8SE00627J https://doi.org/10.1021/acs.chemmater.3c01255 https://doi.org/10.1021/acs.chemmater.3c01255 https://doi.org/10.1002/cphc.201800506 https://doi.org/10.1039/C7RA06172B https://doi.org/10.1039/C7RA06172B https://doi.org/10.5599/jese.690 https://doi.org/10.1039/D1MA00032B https://doi.org/10.1016/j.ceramint.2017.03.121 https://doi.org/10.1039/C7TA02316B https://doi.org/10.1039/C7TA02316B https://doi.org/10.1002/smll.201300970 https://doi.org/10.1002/smll.201300970 https://doi.org/10.1016/S0169-4332(00)00378-0 https://doi.org/10.1016/S0169-4332(00)00378-0 https://doi.org/10.1002/macp.201500169 https://doi.org/10.1002/macp.201500169 https://doi.org/10.3390/pharmaceutics15082102 https://doi.org/10.3390/pharmaceutics15082102 https://doi.org/10.1002/smll.201001333 https://doi.org/10.1039/D1QI00500F https://doi.org/10.1016/j.apsusc.2016.11.034 https://doi.org/10.1016/j.apsusc.2016.11.034 https://doi.org/10.1039/b103135j https://doi.org/10.1021/jacs.8b05134 https://doi.org/10.1179/1743294415Y.0000000010 https://doi.org/10.1021/acsanm.2c03395 https://doi.org/10.1021/acs.iecr.2c04294 https://doi.org/10.1021/acs.iecr.2c04294 https://doi.org/10.1002/adfm.201102155 https://doi.org/10.1002/adfm.201102155 https://doi.org/10.1039/D4NA00199K https://doi.org/10.1039/D4NA00199K https://doi.org/10.1039/C6TA06979G https://doi.org/10.1016/j.jpowsour.2014.05.134 https://doi.org/10.1021/am200294k https://doi.org/10.1021/am200294k https://doi.org/10.1149/1.2086682 https://doi.org/10.1149/1.2086682 https://doi.org/10.1039/D3CY00674C https://doi.org/10.1002/cssc.201402988 https://doi.org/10.1021/acs.jpcc.6b12652 https://doi.org/10.1021/acs.jpcc.6b12652 https://doi.org/10.1021/ja407115p https://doi.org/10.1021/ja407115p https://doi.org/10.1021/cm400535d Surface Defects, Ni³⁺ Species, Charge Transfer Resistance, and Surface Area Dictate the Oxygen Evolution Reaction Activity of Mesoporous NiCo₂O₄ Thin Films Introduction Results and Discussion Structural Characterization Electrochemical Characterization Conclusions Experimental Section Sample Preparation. Acknowledgements Conflict of Interests Data Availability Statement