Electrocatalytic Water Oxidation at Quinone-on-Carbon: A Model

Oct 16, 2018 - Open Access ... of New South Wales, Sydney , Kensington, New South Wales 2052 , ... The turnover frequency per C═O (∼0.323 s–1 at...
0 downloads 0 Views 1MB Size
Subscriber access provided by UNIV OF LOUISIANA

Article

Electrocatalytic Water Oxidation at Quinoneon-Carbon: A Model System Study Yangming Lin, Kuang-Hsu Wu, Qing Lu, Qingqing Gu, Liyun Zhang, Bingsen Zhang, Dang Sheng Su, Milivoj Plodinec, Robert Schlögl, and Saskia Heumann J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b07627 • Publication Date (Web): 16 Oct 2018 Downloaded from http://pubs.acs.org on October 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Electrocatalytic Water Oxidation at Quinone-on-Carbon: A Model System Study Yangming Lin,[a] Kuang-Hsu Wu,[b] Qing Lu,[a] Qingqing Gu,[a,

c]

Liyun Zhang,[c] Bingsen Zhang,[c]

Dangsheng Su,[c] Milivoj Plodinec [d] Robert Schlögl [a, d] and Saskia Heumann, [a]* a

Max Planck Institute for Chemical Energy Conversion, Stiftstrasse 34–36, Mülheim an der Ruhr,

45470, Germany b

School of Chemical Engineering, The University of New South Wales, Sydney, Kensington, NSW 2052,

Australia c

Shenyang National Laboratory for Materials Science, Institute of Metal Research, Chinese Academy of

Sciences, Shenyang, 110016, P. R. China d

Department of Inorganic Chemistry, Fritz Haber Institute of the Max Planck Society, Faradayweg 4-6,

Berlin, 14195, Germany

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 20

Abstract Nanocarbon can promote robust and efficient electrocatalytic water oxidation through active surface oxygen moieties. The recent mechanistic understandings (e.g., active sites) of metal-free carbon catalysts in oxygen evolution reaction (OER), however, are still rife with controversies. In this work, we describe a facile protocol in which eight kinds of aromatic molecules with designated single oxygen species were used as model structures to investigate the explicit roles of each common oxygen group in OER at a molecular level. These model structures were decorated onto typical nanocarbon surfaces like onion-like carbons (OLC) or multi-walled carbon nanotubes (MWCNT) to build aromatic molecule-modified carbon systems. We show that edge (including zigzag and armchair) quinones in a conjugated π network are the true active centers, and the roles of ether and carboxyl groups are excluded in the OER process. The plausible rate-determining step could be singled out by H/D kinetic isotope effects. The turnover frequency per C=O (~0.323 s-1@=340 mV) in 0.1 M KOH and the optimized current density (10 mA/cm2 at 1.58 V vs. RHE) of quinone-modified carbon systems are comparable to those of promising metal-based catalysts.

ACS Paragon Plus Environment

Page 3 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Introduction Electrochemical water oxidation to produce hydrogen (HER) and oxygen (OER) is considered to be one of the most facile and green methods to generate hydrogen.1,2 However, the latter has been regarded as a rate-limiting step in the realization of a fully integrated water-splitting system because of its high energy barrier for O–H bond breaking and attendant O–O bond formation.3 It is therefore necessary to develop the promising OER catalysts which is able to decrease energy barrier and to improve O2 evolution rate, and finally accelerating the overall water splitting system. Recently, the primarily involved catalysts include Ir/Ru,4 Co,5 Ni,6 Fe,7 Cu8 or Mn based materials.9 Carbon materials as non-metal catalysts have been demonstrated recently to be alternatives to meet requirement of renewable energy technologies.10 They provide chemical inertness, tunable electronic structures, low-cost and sustainable advantages, and even exhibit comparable electrocatalytic activity and durability for the OER to those of metal-based catalysts. Although different oxidation methods have been attempted to elaborate the roles of each oxygen group for OER,11,12 difficulties still remain in understanding the intrinsic OER mechanism of carbon materials. The reasons are: (1) the inevitable presence of diverse functional groups at carbon surface (e.g., ketone, phenolic, carboxyl, ether, etc.) during the preparation process; (2) multifarious edge defect configurations. Therefore, understanding the OER mechanism on carbon-based materials at a molecular level still remains challenging. Nanocarbon materials can be considered as sp2 carbon atoms in ring configurations building planar or curved two dimensional structures. Oxygen heteroatoms can form different kinds of functional groups at the edges of the graphite layers as depicted in (Scheme 1a) .13 In comparison to nanocarbon materials, functionalized aromatic organic molecules provide oxygen species (Scheme 1b) with a similar carbon network structure. The conjugated π systems can be tuned in the electronic and reactive properties by extending the benzene unit. These molecules can be directly used as model active matrices to mimic the intrinsic structures of oxidized carbon catalysts and to identify specific oxygen active components, such as different C=O configurations.14 However, aromatic organic molecules have a low conductivity and poor electron transfer properties that are hampering their direct application in the electrocatalytic field. A simple, convenient and nondestructive solution with nanocarbons as supports

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 20

has been recently proposed in which an effective electron transfer was observed between active components and the support by the strong non-covalent π-π interactions. The obtained catalysts exhibited promising performance in various electrochemical reactions like OER,15,16 hydrogen evolution,17 oxygen reduction18,19 or CO2 conversion20,21.

Scheme 1. (a) Structure of nanocarbon materials with typical oxygen functional groups. (b) Structures of different aromatic organic molecules with isolated oxygen groups (e.g., C=O, C-OH, -COOH, C-O-C). (c) The synthetic process of aromatic molecules-modified OLC using solvothermal method. The detailed information is shown in supporting information (Scheme S2).

Inspired from the previous work, we used two typical carbon support materials, one dimensional onion-like carbon (OLC) with 460 m2/g high surface area and trace inherent oxygen species and multi-walled carbon nanotubes (MWCNTs) with 230 m2/g moderate surface area22,23 to engineer highly

ACS Paragon Plus Environment

Page 5 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

dispersed aromatic organic molecules with specific oxygen species via a facile solvothermal method (Scheme 1c). The interactions between organic molecules and carbon supports are confirmed by near-edge X-ray absorption fine structure (NEXAFS) spectroscopy. The obtained mono-modified carbon materials not only have a high concentration of desired active oxygen species, but also exhibit an impressive efficiency for OER based on the good synergy between support and active organic molecules. It should be noted that by extending the π-conjugated matrix of organic molecules having quinone groups, the catalytic performance (e.g., current density) of modified catalysts could be further improved, even being comparable to the promising metal-based catalysts. The activities of modified catalysts mainly originate from carbon catalysis rather than from carbon corrosion in 0.1M KOH. The C=O group is determined to be advantageous for OER. Results and discussion Synthesis, Characterization and Interaction Investigation Different kinds of oxygenated aromatic molecules-modified OLC samples were prepared via a simple solvothermal process. Organic molecules with representative oxygen-involved functional groups like

carboxyl-,

ether-,

9-phenanthrenecarboxylic

phenolic-, acid

quinione,

(PCA),

were

xanthene

used (X),

as

model

9-phenanthrenol

precursor, (PE)

namely and

9,

10-phenanthrenequinone (PQ). As reference, an only solvent (cyclohexane)-treated OLC (T-OLC) sample without organic molecule additives, was also prepared under the same preparation conditions. Detailed synthesis procedures are found in the supporting information and shown in Scheme S2. X-ray photoelectron spectroscopy (XPS) spectra are shown in Figure 1a. T-OLC and modified OLC samples involve three oxygen components: C=O related groups (ca. 531.8 eV), C-O-C or O=C-O groups (ca. 533±0.2 eV) and C-O(H) groups (ca. 534 eV).24 All modified-OLC samples show a similar surface oxygen content. The net contents of aromatic organic molecules on modified OLC range from 2.3 to 3.3 at% as calculated by subtracting the amount from the referencing T-OLC (Figure 1a, Table S1, Figure S1). The specific oxygen species itself exist in a high proportion (76%~95.6%, Table S1). By using thermogravimetric (TG) measurements, the net contents of the model molecules on modified

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 20

OLC were further quantified to be 1.6 wt%~2.1 wt% (Table S1, Figure S2). The molecular structures of PQ, PE, X and PCA remain unchanged during the preparation process as confirmed by attenuated total reflectance (ATR) spectra and high performance liquid chromatography (Figure S3-S7). Moreover, the rough coverage rates for PQ, PE, X and PCA molecules on OLC support were calculated to be 14.95%, 14.98%, 10.22% and 16.77%, respectively (Calculation Methods section 1 of Supporting information). The possible amounts of PQ, PE, X and PCA molecules on the surface of individual OLC nanoparticle were assessed to be 12, 12, 11 and 11, respectively.

Figure 1. Surface compositions and properties of T-OLC and various modified OLC. (a) XPS O1s and (e) XPS C1s spectra. NEXAFS spectra (incident angle θ=45o) of representative PQ-modified OLC and T-OLC samples at (b) C K-edge, (c) magnified spectra of C K-edge at 281 eV~287 eV and (d) O K-edge.

Figure 1b shows the normalized C K-edge NEXAFS spectra of representative PQ-modified OLC and the reference T-OLC. The spectra of the two samples have two main features, around 285 and 291 eV, which are assigned to π* antibonding and σ* antibonding orbitals at the sp2 (C=C band in the ring) site from the C1s level, respectively.25,26 It can be observed that the intensity of the π* feature at around 285 eV of modified OLC is higher than that of T-OLC. This points to an increased π-π stacking between PQ and OLC. The onset temperature of thermal decomposition of PQ-modified OLC is higher than that

ACS Paragon Plus Environment

Page 7 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

of pure unsupported PQ, which further supports the fact of π-π stacking between PQ and OLC (Figure S8). It should be pointed out that compared to T-OLC, an additional peak of modified OLC can be observed at 287.9 eV, which can be assigned to a π* bond of C=O. No other carbon-oxygen species can be observed in C K-edge NEXAFS spectra. Moreover, a weak peak emerged at about 286.5 eV over PQ-modified OLC is attributed to π* (ring) resonances of PQ molecule (Figure 1c). In general, π* (ring) resonances of PQ are split into two peaks by a variable of 1.6 eV and hence the other peak may be overlapped by π* bond of OLC at 285 eV.27 As illustrated in O K-edge NEXAFS spectra (Figure 1d), modified OLC exhibits a pronounced π* bond of C=O (at 531.3 eV) and a similar σ* bond of C-O or C-OH (at 535.6 eV),26 suggesting the main oxygen species over PQ-modified OLC is C=O group, confirming the presence of the functional groups from PQ on the composite. No charge correction was fulfilled, the minor shift of about 0.2 eV of the sp2 carbon position (~284.6 eV) in the XPS C1s spectra suggests that PQ is electronically closely coupled to the π electronic structure of OLC support (Figure 1e). The peak located at 288.2 eV is regarded as C=O bond. In addition, oxygen species have been demonstrated to be homogeneously distributed on the surface of OLC by using high angle annular dark field-scanning

transmission

electron

microscopy-energy

dispersive

x-ray

spectrometry

(HAADF-STEM-EDX) image (Figure S9 a-c and S9 e-g). The oxygen content is increasing from 0.7 at% for T-OLC to 2.9 at% for PQ-modified OLC. The introduction of PQ slightly changes the surface morphology of OLC (Figure S9d and S9h). All results from XPS, TG and NEXAFS have shown that well-controlled mono group-modified OLC composites could be obtained with similar oxygen concentrations building a proper basis for the study of different functional groups.

Figure 2. Comparative study of cyclic voltammetry curves of (a) pure aromatic organic molecules, (b) various modified

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

OLC and T-OLC samples in argon-saturated 0.1M KOH measured at a scan rate of 100 mV s-1. The loadings of all samples on electrode are 0.051 mg/cm2. (c) A plausible reversible proton-coupled electron transfer (PCET) between phenolic and quinone groups and the possible intermediate radicals during electrochemical process.

Cyclic voltammetry (CV) has been considered as an experimental straightforward method to examine the noncovalent π-π stacking interactions between organic molecules with redox-active sites and carbon support.28 In Figure 2a, characteristic redox peaks located at 0.4~0.6 V can be indistinctly observed in the CV for all organic molecules except for PQ. The current densities of all organic molecules are approximately between -0.14 mA/cm2 and 0.08 mA/cm2. Such low current densities may be ascribed to their low intrinsic electric conductivity. Apparently, the current density of each modified OLC electrode is much higher than that of the pure aromatic organic molecules and of T-OLC (Figure 2b). The recorded CV of modified OLC and T-OLC are very similar and overlap in wide regions. It should be noted that the intensities of characteristic redox peaks (at 0.4~0.6 V) of organic molecules are enhanced significantly to currents ranges between -0.62 mA/cm2 to 0.52 mA/cm2. It seems that during the electrochemical reaction a good interaction between the adsorbed organic molecules and OLC supports, most likely π-π stacking, is involved to achieve the better charge and electron transfers. Moreover, PQ-modified OLC and PE-modified OLC show similar features of the redox peaks since there is a reversible proton-coupled electron transfer (PCET) process between PQ and PE (Figure 2c and Figure S10a).29,30, PE can completely transfer into PQ at high oxidation potential. Furthermore, the good electron transfer can be experimentally supported by the face that PQ-modified OLC shows a shorter potential difference (22 mV), such as more positive (ΔE=+15 mV) and more negative (ΔE=-7 mV) values compared with pure PQ (Figure S10b). All these findings point out the direct existence of fast charge transfer between organic molecules and OLC support. This behavior indicated that there may have a good synergistic effect between OLC and active organic molecules, and thus could provide great potential for electrochemical applications.

ACS Paragon Plus Environment

Page 9 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Identification of Possible Active Components by Electrochemical Performance Evaluation

Figure 3. Electrochemical measurements of aromatic organic molecules-modified OLC catalysts for OER. The loadings of all samples on the glassy carbon electrode are 0.051 mg/cm2. (a) LSV curves of modified OLC and unmodified T-OLC catalysts measured in argon-saturated 0.1M KOH with a scan rate of 5 mV s-1 with iR correction. (b) Tafel slopes of various catalysts. (c) Theoretical TOFs of various organic molecules-modified OLC catalysts at two different overpotentials ().

The OER activities of various aromatic organic molecules and their corresponding modified OLC catalysts were assessed. In Figure 3a, PQ-modified OLC catalyst imparts the smallest onset potential (Eonset) of 1.53 V vs RHE and achieves a current density of 10 mA cm-2 (J10) at an overpotential () of 430 mV, whereas X- and PCA-modified OLC catalysts do not surpass the catalytic performance of T-OLC. Interestingly, PE-modified OLC affords an enhanced current density and demands a  = 510 mV to reach 10 mA cm-2. The Eonset is also close to 1.53 V vs RHE. Pure aromatic organic molecules were also tested under identical conditions as reference to the OLC supported ones (Figure S11). Compared with a blank experiment of glassy carbon, the pure aromatic organic molecules do not show any considerable performance. It seems that the significant current densities of modified OLC catalysts originate from water oxidation rather than from the corrosion of organic molecules themselves. The Tafel slope of PQ-modified OLC (Figure 3b, 69 mV dec-1) is much lower than that of X-modified OLC (189 mV dec-1), PCA-modified OLC (179 mV dec-1) and T-OLC catalyst (183 mV dec-1), suggesting its favorable reaction kinetic. The results of electrochemical impedance spectroscopy (EIS, Figure S12) imply a better charge transfer of the electrode/electrolyte interface over PQ-modified OLC because of its smaller semicircle diameter compared with that of T-OLC and pure PQ. To assess its intrinsic

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 20

catalytic ability, the theoretical turnover frequency (TOF) of various organic molecules-modified OLC catalysts at  = 340 and 430 mV were also calculated (Figure 3c, based on the results of ΔTG). The PQ-modified OLC is found to carry a dramatic TOF value of ~0.0494 s-1C=O-1, similar to that of PE-modified OLC catalyst with a TOF value of around 0.0489 s-1C-OH-1. These values can increase to ~0.476 s-1C=O-1 (PQ) and ~0.437 s-1C-OH-1 (PE) at  = 430 mV. Both X- and PCA-modified OLC catalysts perform very low TOF values, implying their non-promotion roles in OER. The calculated mass activity for PQ-modified OLC is 196.2 A g-1, which is about twofold to PE-modified OLC (87.3 A g-1, Figure S13). These discoveries on PE-modified OLC can be attributed to the reversible relationship between quinone and phenolic groups at high working potential as discussed for Figure 2c. Moreover, by employing the rotating ring-disk electrode (RRDE) technique (Figure S14), the OER process occurring on PQ-modified OLC is confirmed to be dominated by a desirable four-electron pathway with negligible peroxide intermediate formation (99.95%, 4 OH- → O2 + 2H2O + 4 e-) and a high Faradaic efficiency (FE(O2)) of 72.2% at a current density of 1.0 mA cm-2 @1.57 V or 85.9% at a J10. These impressive values are higher than those of T-OLC (8.5 %@1.57 V) and are comparable to reported carbon-supported metal catalysts.31,32 The loss of FE(O2) may be partly caused by interference due to wobbling oxygen bubbles and therefore inefficient oxygen collection by the Pt ring electrode or carbon corrosion in alkaline medium.33,34 The electrochemically active surface areas (ECSA) were determined to be 11.37 cm2 for PQ modified OLC and 5.89 cm2 for T-OLC, supporting its better activity (Figure S15). In addition, the results of decreasing contact angles reflect that the presence of the aromatic organic model molecules slightly improve the wettability of the catalysts which has a positive impact on the total performance (Figure S16).35 The contribution of metal impurities in the OER could be excluded by ICP-OES and elemental analysis measurements (Table S2). Exemplary, the content of Fe impurities is at a low level of a similar value for all modified and unmodified OLC. All these results show clearly that the quinone group has a positive effect on the OER. To study the stability of the supported and unsupported C=O functional group, CV curves before and after an applied constant current of 10 mA/cm2 of pure PQ, T-OLC and PQ-modified OLC were measured (Figure S17). After 9 h testing, the unsupported PQ still emerges apparent redox peaks,

ACS Paragon Plus Environment

Page 11 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

suggesting its good resistance towards oxidation. In comparison, fresh T-OLC does not show redox peaks in the beginning, but two new redox peaks (at 0.4~0.6 V) appear in used T-OLC, signifying that during the OER process, T-OLC would be oxidized/etched along with the in-situ formation of considerable oxygen species (Figure S17b). Additionally, the total current density decreases. These results provide reasonable explanation for the low Faradaic efficiency of T-OLC in OER. OLC supported PQ shows characteristic redox peaks in the beginning, which are still detectable after 9 h testing, similar to 2 h testing. The reduced current densities of the characteristic redox peaks of used modified OLC may be assigned to: 1) the desorption of PQ from OLC to some degree in the beginning of OER; 2) in-situ formed oxygen species of the neighboring benzene ring of C=O of PQ at applied potential will level the real CV redox peaks of original C=O. This can be supported by the higher current densities at 0.0~0.2 V and at 0.6~1.0 V regions of used modified OLC catalyst (Figure S17c). After a long time testing, the ΔE shifts from 22 mV of fresh catalyst to 52 mV of 9 h aging catalyst, suggesting the obvious diffusion mass transfer. Influences of Edge Configuration and π-Conjugated Structure of Active Components on OER

Figure 4. Electrochemical measurements of aromatic organic molecules-modified OLC catalysts for OER. The loadings of all samples on electrode are 0.051 mg/cm2. (a) LSV curves of aromatic organic molecules with different zigzag and armchair configurations and extended π conjugated structure-modified OLC and IrO2 supported OLC catalysts measured in

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 20

argon-saturated 0.1M KOH with a scan rate of 5 mV s-1 with iR correction. (b) The π conjugated structure as a function of Tafel slopes (black), onset potentials (Eonset, red) and potentials at a current density of 10 mA cm-2 (Ej=10, blue) over quinone-modified OLC catalysts. (c) LSV curves of AQ- and AT-modified OLC catalysts. (d) Theoretical TOFs of a series of C=O group-modified OLC catalysts at  of 340 mV. (e) Long-term stability tests of representative PQ-, 6, 13-PQ-modified OLC and IrO2 supported OLC catalysts at a constant current density of 10 mA/cm2 (J10). (f) Comparison of the electrocatalytic OER activity of 6,13-PQ-modified OLC, reported carbon-based and transition-metal catalysts recorded in 0.1 M and 1.0 M KOH. Detailed information are listed in Table S3.

Moreover, the effect of zigzag and armchair configuration of C=O on the catalytic activity was investigated as well as the conjugation of the π system of the organic model molecules. The composite materials were synthesized and characterized in analogy to the previous described modified OLC using anthraquinone (AQ), 5, 12-naphthacenequinone (5, 12-NQ) and 6, 13-pentacenequinone (6, 13-PQ) as model molecules. The net contents of AQ, 5, 12-NQ and 6, 13-PQ were determined to be 1.91 wt%, 1.72 wt% and 1.59 wt%, respectively by using TG-MS (Figure S 18a and Figure S19). The ATR spectra indicate that the molecular structures of AQ, 5, 12-NQ and 6, 13-PQ are not destroyed during the preparation process (Figure S20-S22). Based on the CV spectra (Figure S23), AQ-modified OLC (zigzag) show obvious redox peaks like PQ (armchair) modified OLC, indicating a good interaction between the AQ molecules and the OLC support. Similar Eonset, Ej=10 (1.66 V at J10) and Tafel slopes (69 mV dec-1 vs 65 mV dec-1, respectively, Figure 3a, Figure 3b, Figure 4a and Figure S24) of PQ- and AQ-modified OLC suggest that zigzag and armchair configurations linked with C=O groups do not obviously affect the catalytic performance. Furthermore, adjoining of the C=O has no impact indicating that each individual quinone group might play the same role in the OER. The π conjugation is also an important parameter of the carbon network structure that can influence the electronic and charge transfer, and thus has been investigated. The onset potential of quinone-modified OLC can be slightly optimized with increasing the number of benzene units (Figure 4a, 4b). The 6, 13-PQ-modified OLC catalyst exhibits the lowest Eonset of 1.50 V vs RHE and reaches a J10 at  of 390 mV (1.63 V), which is close to the IrO2/OLC catalyst (1.60 V). Details for the IrO2/OLC reference catalysts see supporting information. The  at J10 of 6, 13-PQ-modified OLC catalyst can be further decreased to =350 mV

ACS Paragon Plus Environment

Page 13 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

(1.58 V) with increasing catalyst loading and constant Eonset (Figure S25). Moreover, the calculated Tafel slope of 6,13-PQ-modified OLC (62 mV dec-1, Figure 4b and Figure S24) is lower than that of the IrO2/OLC catalyst, and similar to the PQ-, AQ- and 5, 12-NQ-modified OLC catalysts (69 mV dec-1 , 65 mV dec-1 and 63 mV dec-1, respectively). An inverse proportional relationships between the number of benzene unit and Eonset/Ej=10 in Figure 4b may be associated to the enhanced π conjugated electron effect of active molecules. It should be noted that the TOF values can be improved significantly depending on the π-conjugated size; the values range from 0.0371 s-1C=O-1 of AQ-modified OLC to 0.323 s-1C=O-1 over 6, 13-PQ-modified OLC at =340 mV (Figure 4d). Both values are significantly higher than that of reported highly active OER catalysts with TOF numbers of 0.03 s-1~0.0915 s-1 at  = 300 mV~400 mV in 0.1 M KOH or 1 M KOH.36-39 It is therefore reasonable to believe that the π-conjugated size of the carbon matrix is favorable to accelerate the OER process, although the reaction pathway will not be changed as clarified by the similar Tafel slopes. These phenomena can provide a new idea to develop highly efficient model OER catalysts by extending their π conjugated structures. As mentioned above, the introduction of PE with only one -OH group exhibits a similar catalytic performance related to the active sites (e.g. TOF) like PQ-modified OLC, which contributes two C=O functional groups per molecule as active component. This indicates the possibility that each C=O contributes independently to the OER. For further confirmation, anthrone (AT) with only one C=O group was studied in comparison to AQ. The net contents of AT is deliberate to be 2.01 wt% (Figure S18b). As shown in Figure 4c, 4d and Figure S24b, AT exhibits similar catalytic features compared to AQ, PQ and PE related to the TOF value and Tafel slope e.g., evidencing the independent contribution of each C=O to the OER. No significant potential or current changes could be observed for PQ- and 6, 13-PQ-modified OLC and the corresponding pure organic model structures after 9 h or 10 h testing, indicating their good catalytic and structure stabilities at the applied power (Figure 4e, Figure S26). The oscillating current increase during the chronopotentiometric measurement for PQ-modified OLC is caused by bubble formation and therefore blocking of the electrode surface. The stable behavior can not be observed for the referencing IrO2/OLC sample. The Tafel slope and the  of 6,13-PQ-modified OLC are lower than

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 20

these of distinguished carbon-based or metal-based catalysts at the same current density of 10 mA/cm2 in 1.0 M or 0.1 M KOH. A widespread comparison with literature known catalysts was conducted and was summarized in Figure 4f and Table S3. This occurrence well supports the significance role of the quinone group in OER. Mechanism Study

Figure 5. (a) Adsorption energy and structure of OH species on quinone molecule having C=O group obtained from DFT calculations. The negative adsorption energy means that the adsorption is exothermic. Color code: carbon is gray, hydrogen is white, and oxygen is red. Isotopic electrochemical studies of representative PQ-modified OLC catalyst. (b) CV spectra in 0.1 M KOH dissolved in H2O and D2O (99.9%). (c) LSV curves of PQ-modified OLC catalyst. The kinetic isotopic effect (KIE) values were roughly calculated from the current density ratio in H2O and D2O solutions, that is kH/kD=JH/JD.

In order to better understand the OER mechanism on carbon catalysts, it is significant to use effective measurements to interrogate the reaction kinetics of the OER processes. Density functional theory (DFT) calculations predicted that the binding energy of surface intermediate oxygen species such as *OH, *O, and *OOH governs the OER activity.40 In the present work, by using theoretical calculation (Figure 5a), an initial OH species would firstly adsorb at the meta-position (location 2) of C=O group to achieve the first step of OER. The negative adsorption energy (-2.23 eV) means that this step is easy to proceed. The other two positions (1, 3) are unable to provide a more optimized adsorption energy and structure to adsorb OH species. Actually, the use of computational methods to directly reveal the kinetics of heterogeneous catalytic systems is extremely challenging and the explicit role of

ACS Paragon Plus Environment

Page 15 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

protons in the OER remains poorly understood. The use of the kinetic isotope effect (KIE) is an established experimental technique to study chemical reactions involving protons. Specifically, the substitution of hydrogen with deuterium has been implemented extensively due to the large differences in reaction rates originating from the reduced mass differences between the isotopes.41 As mentioned above, a good PCET can be observed between PQ and PQH2 (Figure 2c). Here, the effect of protons on PCET process was firstly investigated. As displayed in Figure 5b, the PQ/PQH2 redox wave has a midpoint potential (E1/2) of 0.386 V. The PQ/PQD2 displays a redox wave with an E1/2 value of 0.465 V. The position of the redox wave in deuterated solution is ca. 79 mV more positive than that obtained in protonated solution, indicating that the oxidation of PQD2 to PQ is more difficult than the oxidation of PQH2 to PQ. This is caused by the stronger O-D bond in comparison to the O-H bond and thus breaking the O-D bond is energetically more costly and the oxidation peak shifts positively.42 Additionally, the cathodic wave also shifts positively which is likely caused by a tighter solvation shell of PQ than that of PQH2 and D2O forms a stronger deuterium-bonding network relative to the hydrogen-bonding network of H2O.41 Therefore, there is a larger increase in entropy when the deuterated solvent structure relaxes during the reduction of PQ to PQH2.43,44 AQ-modified OLC also shows a positive potential shift of 51 mV (Figure S27). Similar positive potential shifts on both the anodic and cathodic peaks upon deuteration have been reported for some metal-based catalysts.41 These results further suggest that the proton transfers are concerted with the electron transfers in the PCET system of PQ-modified OLC. When the potential is elevated to the OER working region, as shown in Figure 5c, the OER activity is decreasing when using deuterated water. The Tafel slopes increase from 69 mV dec-1 in protio solution to 93 mV dec-1 in deuterio solution (Figure S28), suggesting a difference in the Gibbs free energy of the intermediate that governs the reaction rate.45 Moreover, it has been demonstrated that the KIE values can be approximately calculated by the current density ratio in H2O and D2O solutions, that is kH/kD=JH/JD.41,46 In the present work, the kH/kD at 1.66 V carries a factor of 2.685 and the values are changed to 2.818 and 2.149 at lower potentials of 1.60 V and 1.57 V (the corresponding current density of latter is 1 mA/cm2), respectively, suggesting that there is a significant primary isotope effect due to a water oxidation reaction and C=O groups promote the reaction through a common OER PCET

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 20

mechanism. The primary isotope effect with a high value indicates that the rate-determining step (RDS) of water oxidation involves a cleavage of the O-H bond. It indicates that introduced deuterated water is most likely to influence the formation of active oxygen species *O from the deprotonation process of *OD or hinder the O-O bond formation from deprotonation process of active oxygen species *OOD. Moreover, based on a single-site mechanism assumption, Tafel slope close to 60 mV dec-1 has been considered to be indicator of a RDS involving the deprotonation process of active oxygen species *OOH into OO- species (that is, O-O bond formation for O2 production).47 As mentioned above, the Tafel slopes of quinone-modified OLC range from 62 to 69 mV dec-1 (Figure 4b) and simultaneously meet a single-site mechanism requirement (C=O as possible single active site clarified above). Therefore, it can be expected that the deprotonation process of active oxygen species *OOH into OOspecies is a possible RDS in OER.

Figure 6. Electrochemical activities of PQ-modified MWCNT sample for OER. (a) LSV curves of modified MWCNT, T-MWCNT and pure PQ at a scan rate of 5 mV-1 in argon-saturated 0.1M KOH with iR correction. (b) Long-term stability test of modified MWCNT catalyst at J10.

The influence of carbon support on OER was also investigated by using multi-walled carbon nanotubes (MWCNTs) as well-established support material in the field of electrocatalysis. When MWCNTs were used as support instead of OLC, the as-prepared PQ-modified MWCNTs sample with 1.21 wt% net content of PQ (Figure S29) outperforms pure PQ and T-MWCNTs samples in activity (Figure 6). The  of the modified MWCNTs catalyst at J10 is 460 mV (1.69 V). The Eonset = 1.52 V is similar to the quinone-modified OLC catalysts (e.g., AQ+OLC, PQ+OLC). It should be emphasized that

ACS Paragon Plus Environment

Page 17 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

the modified catalyst carries out an excellent stability, as the observed potential does not obviously change during 10 h operating time. All the above results support the fact that the aromatic organic molecules as model probes are excellent candidates to investigate the real active components of carbon materials at a molecular level. Conclusions In summary, we have demonstrated by aromatic organic molecules with designated oxygen species as model active components that the quinone functional groups play a crucial role for the electrochemical oxygen evolution reaction (OER) and each C=O is demonstrated experimentally to contribute independently to the OER. C=O groups located at edge zigzag and armchair configurations show the similar catalytic performance. The introduction of aromatic organic molecules with C=O configuration endows nanocarbon materials with an excellent performance in the OER with a low onset potential (1.50 V vs. RHE) and a low Tafel slope (62 mV dec-1) in 0.1 M KOH. Phenolic groups are also positive to OER derived from a proton-coupled electron transfer relationship between quinone and phenolic groups. We revealed that the OER activity can be further improved by extending the π-conjugated structure of active components. The deprotonation process of active oxygen species *OOH into OO- species is a possible rate-determining step in OER. The study in the multi-functionality in the surrounding of catalysts during OER process is ongoing in our lab. This work does not only offer a direct evidence to understand real catalytically active sites of carbon catalysts at a molecular level, but also introduces a highly efficient and durable electrocatalyst for OER. ASSOCIATED CONTENT Supporting Information. Full synthesis, additional characterization and activity date are provided in the supporting information. This material is available free of charge via the Internet at http://pubs.acs.org. Acknowledgment The authors would like to thank the Max Planck society for funding.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 20

AUTHOR INFORMATION Corresponding Author [email protected] Notes The authors declare no conflict of interest. References (1) Shui, J.-L.; Karan, N. K.; Balasubramanian, M.; Li, S.-Y.; Liu, D.-J. J. Am. Chem. Soc. 2012, 134, 16654. (2) Suen, N.-T.; Hung, S.-F.; Quan, Q.; Zhang, N.; Xu, Y.-J.; Chen, H. M. Chem. Soc. Rev. 2017, 46, 337. (3) Yu, X.-Y.; Feng, Y.; Guan, B.; Lou, X. W. D.; Paik, U. Energ. Environ. Sci. 2016, 9, 1246. (4) Reier, T.; Oezaslan, M.; Strasser, P. ACS Catal. 2012, 2, 1765. (5) Bergmann, A.; Martinez-Moreno, E.; Teschner, D.; Chernev, P.; Gliech, M.; De Araújo, J. F.; Reier, T.; Dau, H.; Strasser, P. Nat. Commun. 2015, 6, 8625. (6) Zhou, W.; Wu, X.-J.; Cao, X.; Huang, X.; Tan, C.; Tian, J.; Liu, H.; Wang, J.; Zhang, H. Energ. Environ. Sci. 2013, 6, 2921. (7) Zou, S.; Burke, M. S.; Kast, M. G.; Fan, J.; Danilovic, N.; Boettcher, S. W. Chem. Mater. 2015, 27, 8011. (8) Huan, T. N.; Rousse, G.; Zanna, S.; Lucas, I. T.; Xu, X.; Menguy, N.; Mougel, V.; Fontecave, M. Angew. Chem. 2017, 129, 4870. (9) Subbaraman, R.; Tripkovic, D.; Chang, K.-C.; Strmcnik, D.; Paulikas, A. P.; Hirunsit, P.; Chan, M.; Greeley, J.; Stamenkovic, V.; Markovic, N. M. Nat. Mater. 2012, 11, 550. (10) Zhang, J.; Zhao, Z.; Xia, Z.; Dai, L. Nat. Nanotechnol. 2015, 10, 444. (11) Li, L.; Yang, H.; Miao, J.; Zhang, L.; Wang, H.-Y.; Zeng, Z.; Huang, W.; Dong, X.; Liu, B. ACS Energy Lett. 2017, 2, 294. (12) Lu, X.; Yim, W.-L.; Suryanto, B. H.; Zhao, C. J. Am. Chem. Soc. 2015, 137, 2901. (13) Wu, S.; Wen, G.; Liu, X.; Zhong, B.; Su, D. S. ChemCatChem 2014, 6, 1558. (14) Lin, Y.; Li, B.; Feng, Z.; Kim, Y. A.; Endo, M.; Su, D. S. ACS Catal. 2015, 5, 5921. (15) Li, F.; Zhang, B.; Li, X.; Jiang, Y.; Chen, L.; Li, Y.; Sun, L. Angew. Chem. 2011, 123, 12484. (16) Garrido-Barros, P.; Gimbert-Suriñach, C.; Moonshiram, D.; Picón, A.; Monge, P.; Batista, V. S.; Llobet, A. J. Am. Chem. Soc. 2017, 139, 12907. (17) Tran, P. D.; Le Goff, A.; Heidkamp, J.; Jousselme, B.; Guillet, N.; Palacin, S.; Dau, H.; Fontecave, M.; Artero, V. Angew. Chem. 2011, 123, 1407. (18) Lei, H.; Liu, C.; Wang, Z.; Zhang, Z.; Zhang, M.; Chang, X.; Zhang, W.; Cao, R. ACS Catal. 2016, 6, 6429. (19) Morozan, A.; Campidelli, S.; Filoramo, A.; Jousselme, B.; Palacin, S. Carbon 2011, 49, 4839. (20) Blakemore, J. D.; Gupta, A.; Warren, J. J.; Brunschwig, B. S.; Gray, H. B. J. Am. Chem. Soc. 2013, 135, 18288. (21) Maurin, A.; Robert, M. J. Am. Chem. Soc. 2016, 138, 2492. (22) Lin, Y.; Feng, Z.; Yu, L.; Gu, Q.; Wu, S.; Su, D. S. Chem. Commun. 2017, 53, 4834. (23) Pech, D.; Brunet, M.; Durou, H.; Huang, P.; Mochalin, V.; Gogotsi, Y.; Taberna, P.-L.; Simon, P. Nat. Nanotechnol. 2010, 5, 651.

ACS Paragon Plus Environment

Page 19 of 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

(24) Lin, Y.; Su, D. ACS Nano 2014, 8, 7823. (25) Blöcker, J.; Werner, H.; Herein, D.; Schlögl, R.; Schedel-Niedrig, T.; Keil, M.; Bradshaw, A. In Materials Science Forum; Trans Tech Publ: 1992, 91, 337. (26) Schuster, M. E.; Hävecker, M.; Arrigo, R.; Blume, R.; Knauer, M.; Ivleva, N. P.; Su, D. S.; Niessner, R.; Schlögl, R. J. Phys. Chem. A 2011, 115, 2568. (27) Kuznetsova, A.; Popova, I.; Yates, J. T.; Bronikowski, M. J.; Huffman, C. B.; Liu, J.; Smalley, R. E.; Hwu, H. H.; Chen, J. G. J. Am. Chem. Soc. 2001, 123, 10699. (28) McQueen, E. W.; Goldsmith, J. I. J. Am. Chem. Soc. 2009, 131, 17554. (29) Boota, M.; Chen, C.; Bécuwe, M.; Miao, L.; Gogotsi, Y. Energ. Environ. Sci. 2016, 9, 2586. (30) Huynh, M. T.; Anson, C. W.; Cavell, A. C.; Stahl, S. S.; Hammes-Schiffer, S. J. Am. Chem. Soc. 2016, 138, 15903. (31) Swesi, A. T.; Masud, J.; Nath, M. Energ. Environ. Sci. 2016, 9, 1771. (32) Görlin, M.; Ferreira de Araújo, J.; Schmies, H.; Bernsmeier, D.; Dresp, S. r.; Gliech, M.; Jusys, Z.; Chernev, P.; Kraehnert, R.; Dau, H. J. Am. Chem. Soc. 2017, 139, 2070. (33) Qiu, Y.; Xin, L.; Li, W. Langmuir 2014, 30, 7893. (34) Yi, Y.; Weinberg, G.; Prenzel, M.; Greiner, M.; Heumann, S.; Becker, S.; Schlögl, R. Catalysis Today 2017, 295, 32. (35) Li, Y.; Zhao, C. ACS Catal. 2017, 7, 2535. (36) Lu, X.; Zhao, C. Nat. Commun. 2015, 6, 6616. (37) Lu, X. F.; Gu, L. F.; Wang, J. W.; Wu, J. X.; Liao, P. Q.; Li, G. R. Adv. Mater. 2017, 29, 1604437. (38) Song, F.; Hu, X. Nat. Commun. 2014, 5, 4477. (39) Wang, X-L.; Dong, L-Z.; Qiao, M.; Tang, Y-J.; Liu, J.; Li, Y.; Li, S-L.; Su, J-X.; Lan, Y-Q. Angew. Chem. Int. Ed. 2018, 57,9660. (40) Li, M.; Zhang, L.; Xu, Q.; Niu, J.; Xia, Z. J. Catal. 2014, 314, 66. (41) Tse, E. C. M.; Hoang, T. T. H.; Varnell, J. A.; Gewirth, A. A. ACS Catal. 2016, 6, 5706. (42) Boyarkin, O. V.; Koshelev, M. A.; Aseev, O.; Maksyutenko, P.; Rizzo, T. R.; Zobov, N. F.; Lodi, L.; Tennyson, J.; Polyansky, O. L. Chem. Phys. Lett. 2013, 568, 14. (43) Clough, S. A.; Beers, Y.; Klein, G. P.; Rothman, L. S. J. Chem. Phys. 1973, 59, 2254. (44) Edmund, C.; Barile, C. J.; Kirchschlager, N. A.; Li, Y.; Gewargis, J. P.; Zimmerman, S. C.; Hosseini, A.; Gewirth, A. A. Nat. Mater. 2016, 15, 754. (45) Yang, C.; Fontaine, O.; Tarascon, J.-M.; Grimaud, A. Angew. Chem. 2017, 129, 8778. (46) Zhang, B.; Chen, H.; Daniel, Q.; Philippe, B.; Yu, F.; Valvo, M.; Li, Y.; Ambre, R. B.; Zhang, P.; Li, F.; Rensmo, H.; Sun, L. ACS Catal. 2017, 7, 6311. (47) Shinagawa, T.; Garcia-Esparza, A. T.; Takanabe, K. Sci. Rep. 2015, 5, 13801.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

This is the new TOC for the Manuscript 456x279mm (150 x 150 DPI)

ACS Paragon Plus Environment

Page 20 of 20