Functionalization of graphene materials by

Progress in Natural Science: Materials International xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

HOSTED BY

Progress in Natural Science: Materials International journal homepage: www.elsevier.com/locate/pnsmi

Review

Functionalization of graphene materials by heteroatom-doping for energy conversion and storage Chuangang Hua, Dong Liub, Ying Xiaoa,b, Liming Daia,b,

⁎

a

Center of Advanced Science and Engineering for Carbon (Case4carbon), Department of Macromolecular Science and Engineering, Case Western Reserve University (CWRU), 10900 Euclid Avenue, Cleveland, OH 44106, USA b BUCT-CWRU International Joint Laboratories, College of Energy, Beijing University of Chemical Technology (BUCT), Beijing 100029, China

A R T I C LE I N FO

A B S T R A C T

Keywords: Heteroatom-doping Functionalization, Carbon-based catalysts Metal-free electrocatalysis, Energy conversion and storage

Recent development in nanoscience and nanotechnology has opened up new frontiers in materials science and engineering to create new materials for energy generation and storage. Owing to their earth abundance, lowcost, structural tunability, large-surface area, and unique physicochemical properties, graphitic carbon materials have attracted a great deal of attention for energy-related applications. However, the pristine graphene materials without functionalization is intractable (insoluble and infusible), which has hindered their practical applications. Therefore, considerable research effort has been devoted to the development of functionalized graphene materials with desirable properties for specific applications, including energy conversion and storage. It was demonstrated that functionalized graphene materials with tunable work functions were useful as charge-extraction materials to effectively improve solar cell performance while those with high electrocatalytic activities could be used as metal-free catalysts in fuel cells, metal-air batteries, water splitting and integrated energy systems. This article provides a timely focused review on the development of heteroatom-doped graphene materials for low-cost, but efficient, energy generation and storage.

1. Introduction The importance of developing new types of energy is evident from the fact that the global energy consumption has been accelerating at an alarming rate due to the rapid economic expansion worldwide, increase in world population, and ever-increasing human reliance on energybased appliances. Nanotechnology has opened up new frontiers in materials science and engineering to meet this challenge by offering unique enabling technology to create new materials for energy generation and storage. In particular, carbon nanomaterials have attracted a great deal of attention due to their earth abundance, low-cost, structural tunability, large-surface area, and unique physicochemical properties attractive for energy-related applications [1]. As a building block for all carbon materials (e.g., 0D buckyball, 1D nanotube, 3D graphite), graphene possesses the one-atom thick layer of sp2-bonded 2D honeycomb lattice of carbon with a fully conjugated structure of alternating C-C and C˭C bonds. The conjugated structure with alternating single and double bonds in graphitic carbon materials could provide a continuous p-orbital overlap for π electrons to be highly delocalized over the sp2 hybridized carbon atoms, leading to fascinating

electronic attractive for energy conversion and storage [2]. Like many other conjugated materials (e.g., conjugated conducting polymers or conjugated metal-organic frameworks), however, the pristine graphene sheets without functionalization is intractable (i.e., insoluble and infusible), which have hindered their large-scale practical application. Therefore, functionalization of graphene materials is essential, and considerable research effort has been devoted to the development of processable functionalized graphene materials with desirable properties for specific applications, including energy conversion and storage. The availability of solution-processable graphene oxides (GOs) [3–5] has facilitated functionalization of graphene materials. Consequently, various covalent and noncovalent functionalization methods, including basal-plane-functionalization via chemical grafting or noncovalent adsorption, asymmetrical functionalization of the basal plane with different moieties on the opposite graphene surfaces, and edge-functionalization (Fig. 1a) [6], have been developed, along with heteroatomdoping (Fig. 1b) [6]. The covalent and noncovalent functionalization of the graphene edge and/or basal plan shown in Fig. 1a can impart solubility, film forming capability, reactivity for further chemical functionalization,

Peer review under responsibility of Chinese Materials Research Society. ⁎ Corresponding author at: Center of Advanced Science and Engineering for Carbon (Case4carbon), Department of Macromolecular Science and Engineering, Case Western Reserve University (CWRU), 10900 Euclid Avenue, Cleveland, OH 44106, USA. E-mail address: [email protected] (L. Dai). https://doi.org/10.1016/j.pnsc.2018.02.001 Received 29 November 2017; Accepted 1 February 2018 1002-0071/ © 2018 Chinese Materials Research Society. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/).

Please cite this article as: Hu, C., Progress in Natural Science: Materials International (2018), https://doi.org/10.1016/j.pnsc.2018.02.001


C. Hu et al.

Fig. 1. (a) Functionalization possibilities for graphene: (1) edge-functionalization, (2) basal-plane-functionalization, (3) non-covalent adsorption on the basal plane, (4) asymmetrical functionalization of the basal plane, and (5) self-assembling of functionalized graphene sheets. (b) Doping of graphene with heteroatoms (e.g., N). Copyright: 2013, The American Chemical Society. (c) Electronegativity of elements increases along the Y axis. Adapted from http://users.rcn.com/jkimball.ma.ultranet/BiologyPages/E/Electronegativity.html. (a) Reproduced with permission from [6].

CNTs, make graphene a promising candidate for a large variety of applications [22–27], including energy conversion and storage. The pristine graphene without functionalization, however, is a zero bandgap material, which severely limited its application in electronic and many other devices. Heteroatom-doping is a process, in which some carbon atoms in the graphitic structure are replaced by the heteroatoms (Fig. 1b). As the size and electronegativity [28] of the heteroatoms are often different from those of carbon atom (Fig. 1c), the introduction of heteroatoms, regardless of whether the dopants have a higher (as N) or lower (as B, P, S) electronegativity than that of carbon, into graphitic carbon networks could cause electron modulation to change the charge distribution and electronic properties of carbon skeletons, which in turn affects their work function for electronic applications [29] and enhances interactions with reactants to impart catalytic activities [6,30–33]. By doping or functionalization of the edge of graphene sheets with heteroatoms without damage of the carbon basal plane can also change its work function and impart solubility and catalytic activity, while largely retaining the physicochemical properties of the pristine graphene [7]. As the dangling bonds at the edge of a graphene sheet have been demonstrated to be more reactive than the covalently bonded carbon atoms within the basal plane [34,35], edge-doping/functionalizing graphene materials can be taken as a promising approach to tune their properties for specific applications.

and even self-assemble graphene sheets into hierarchically structured materials with tunable properties for multifunctional applications. On the other hand, heteroatom-doping via the introduction of heteroatoms (e.g., nitrogen, boron, phosphorus) into the graphitic carbon network (Fig. 1b) could cause electron modulation to tune their optoelectronic properties and/or chemical activities useful particularly for energy conversion and storage (vide infra). Since both the edge and basal plane functionalization shown in Fig. 1a have been reviewed in our previously-published review articles [6,7], we present in this article a timely focused review on the use of heteroatom-doping for functionalization of graphene materials for energy conversion and storage by summarizing our own work, along with relevant research results as appropriate. Since focus is given to our own work with no intention for a comprehensive literature survey of the field, there will be no doubt that the examples to be discussed in this paper do not exhaust all significant work reported in the literature. 2. Heteroatom doping of graphene materials As is well known, there are three forms of carbon (i.e., amorphous carbon, graphite, and diamond) with different carbon atom arrangements and properties [1,8,9]. Carbon nanotubes (CNTs) can be view as a graphite sheet rolled into a nanoscale tubular form (single-wall carbon nanotubes, SWCNTs) or with additional concentric graphene tubes around the core of a SWCNT (multi-walled carbon nanotubes, MWCNTs) [10–12]. Depending on their diameter and the helicity of the arrangement of graphite rings along the tube length, carbon nanotubes can exhibit novel properties, making them potentially useful in diverse applications ranging from optoelectronic devices to energy-related systems [12]. The peculiar hollow geometry, coupled with the conjugated all-carbon structure, has enabled CNTs to exhibit many intriguing electrical, mechanical, and thermal properties with respect to other members of the carbon family. The high surface area is another important property of CNTs; the theoretical surface area of SWCNTs can reach up to 1315 m2 g−1 [13]. Owing to their unique structure and extraordinary electrical, mechanical, and optical properties, CNTs are promising for various applications [14–16]. To tune/optimize properties demanded for specific applications, various approaches for functionalization of carbon nanotubes have been developed [17,18]. As the building block for carbon nanotubes, graphene has also been demonstrated to exhibit an extremely large specific surface area (2630 m2 g−1) [19], good thermal conductivity (ca. 5000 W m−1 K−1 for single-layer graphene) [20], high Young's modulus (1.0 TPa) [21], high charge mobility (200 000 cm2 V−1 s−1) [22], excellent optical transparency, and flexibility. These properties, superior to those of

3. Functionalized graphene materials with tunable work function for solar cells 3.1. Polymer solar cells The photovoltaic effect involves the generation of electrons and holes in a semiconductor device under illumination, followed by charge separation, and charge collection at opposite electrodes. In inorganic semiconductors, the free charge carriers are directly produced on photon absorption [36]. Photon absorption of organic optoelectronic materials often creates bound electron-hole pairs (i.e., excitons). Charge collection, therefore, requires dissociation of the excitons, which occurs only at the heterojunction interface between semiconducting materials of different ionization potentials or electron affinities (Fig. 2a). The observation of photovoltaic effects arising from the photoinduced charge transfer at the interface between conjugated polymers as donors and C60 acceptor has provided interesting opportunities for improving energy conversion efficiencies of photovoltaic cells based on conjugated polymers [37]. Although the photoinduced charge transfer between the excited C60 acceptor and a conducting polymer donor can 2


C. Hu et al.

Fig. 2. (a) Schematic representation of the device structure of a polymer solar cell in the normal configuration. (b) Chemical structures and synthetic route of GO and GO-Cs. Device structures (c) and energy level diagrams (d) of the normal device and the inverted device with GO as hole-extraction layer and GO-Cs as the electron-extraction layer. Copyright: 2012, Wiley. (e) Schematic illustration of synthesizing GOR from oxidative unzipping of SWCNTs. (f) Device energy level alignment of the GOR-based polymer solar cell. (g) Current densityvoltage (J-V) curves under AM1.5 G illumination of the polymer solar cells without and with PEDOT:PSS, GO or GOR as the HEL. (h) PCE decay of a PEDOT:PSS-based device and a GORbased device unencapsulated and stored in a N2-filled glovebox. Copyright: 2014, Wiley. (a) Reproduced with permission from [38]. (b) Reproduced with permission from [46].

an electron extraction layer needs to have a low work function for electrons to efficiently transport to the cathode. GO has a work function of 4.7 eV, which matches the HOMO level of poly(3-hexylthiophene2,5-diyl) (P3HT) for hole extraction (see Fig. 2c). Moreover, the periphery –COOH groups in GO can dope P3HT of the active layer at the interface to facilitate an Ohmic contact for hole extraction. By replacing the periphery -COOH groups with the -COOCs groups through charge neutralization (Fig. 2b), the work function of GO-Cs modified Al can be reduced to 4.0 eV, which matches the LUMO level of PCBM for efficient electron-extraction (see Fig. 2d). Thus, the controlled functionalization renders GO derivatives to be the single electron-/hole- extraction material system, which is the first material to serve as both hole-extraction layer and electron-extraction layer in both the normal and inverted solar cells (Fig. 2c, d). Therefore, they should be versatile hole-/electron- extraction materials useful for various BHJ solar cells. Graphene thin films synthesized by chemical vapor deposition (CVD) have recently been proved to be efficient hole- and electron-collecting electrodes in BHJ solar cells [39,40], while BHJ solar cells with all-carbon active layer have also been demonstrated [41]. Therefore, the work on the GO/GO-Cs hole-/electron-extraction materials reported by Liu et al. [38], should open avenues to the design and development of novel electron-/hole- extraction materials and holds promise for developing all-carbon flexible solar cells. In a separate study, Liu et al. [42], have rationally designed and successfully developed sulfated graphene oxide (GO-OSO3H) with

occur very rapidly on a sub-picosecond time scale with a quantum efficiency of close to unity for charge separation from donor to acceptor, the conversion efficiency of a bilayer heterojunction device is still limited by several other factors. Firstly, since the efficient charge separation occurs only at the heterojunction interface, the overall conversion efficiency is diminished by the limited effective interfacial area available in the layer structure. Secondly, because the exciton diffusion range is typically at least a factor of 10 smaller than the optical absorption depth, the photoexcitations produced far from the interface recombine before diffusing to the heterojunction. Finally, the conversion efficiency is also limited by the carrier collection efficiency. In order to overcome these deficiencies, photovoltaic cells based on interpenetrating networks between the electron donor and acceptor phases and with charge extraction/transport materials at the interfaces between the active layer and electrodes have been developed, which showed significantly improved photovoltaic conversion efficiency [37]. In this context, GO derivatives have been demonstrated to be excellent hole- and electron-extraction layers in polymer solar cells. Of particular interest, Liu et al. [38], have demonstrated that simple charge neutralization of the -COOH groups in GO with Cs2CO3 (Fig. 2b) could tune the electronic structure of GO, making GO derivatives useful as both hole- and electron- extraction layers in bulk heterojunction (BHJ) solar cells. The work function of a hole-extraction material should be relatively high to allow for the built-in electrical field across the active layer and for holes to transport towards the anode. Similarly, 3


C. Hu et al.

Fig. 3. Device structure and energy levels of the polymer tandem solar cells. (a) Schematic of the tandem device structure, in which a GO-Cs/Al/GO/ MoO3 interconnecting layer was employed. The final device structure is: ITO/PEDOT:PSS (40 nm)/ PCDTBT:PCBM (1:4, 80 nm)/GO-Cs (2 nm)/Al (1 nm)/GO (2 nm)/MoO3 (20 nm)/PCDTBT:PCBM (1:4, 80 nm)/Ca (30 nm)/Al (100 nm). Top right: schematic representations of the chemical structures for the GO and GO-Cs used in the study. The white, red, and pink dots represent C, O, and Cs atoms, respectively. (b) Schematic energy-level diagram for the various layers of the optimized tandem device. (c) J-V characteristics of the subcell and tandem polymer solar cells with different interconnecting layers under illumination of a solar simulator (100 mW cm−2). Copyright: 2014, The American Chemical Society. Reproduced with permission from [47].

–OSO3H groups attached to the carbon basal plane of reduced graphene oxide surrounded with edge-functionalized –COOH groups. The resultant GO-OSO3H was demonstrated to be an excellent hole extraction layer (HEL) for polymer solar cells because of its proper work function for Ohmic contact with the donor polymer, its reduced basal plane for improving conductivity, and its -OSO3H/-COOH groups for enhancing solubility for solution processing. Compared with that of GO, the much improved conductivity of GO-OSO3H (1.3 S m−1 vs 0.004 S m−1) leads to greatly improved fill factor (0.71 vs 0.58) and power conversion efficiency (4.39% vs 3.34%) of the resulting polymer solar cells. Moreover, the device performance of GO-OSO3H is among the best reported for intensively studied P3HT:PCBM devices. These results imply that judiciously functionalized graphene materials can be used to replace existing HEL materials for specific device applications with outstanding performance. The pristine graphene is zero-bandgap material with metal-like conductivity, graphene nanoribbon (GNR) is semiconducting with an opened bandgap induced by the quasi-one-dimensional confinement of charge carriers [43,44]. As mentioned earlier, however, graphene and its nanoribbons without functionalization are insoluble and infusible. The poor processability has precluded the pristine graphene materials, including GNR, for various potential applications. This limitation has been circumvented by oxidizing graphene with acids (e.g., H2SO4/ KMnO4) to produce GO with oxygen-containing groups (e.g., -COOH, -OH) around and on the carbon basal plane [45], leading to low-cost mass production of soluble graphene derivatives for potential applications. By oxidative unzipping of SWCNTs with KMnO4 as oxidant in concentrated H2SO4 [44], the oxygen-rich groups were introduced around a graphene nanoribbon to produce graphene oxide nanoribbon (GOR, Fig. 2e), which should show a synergistic effect to have the bandgap of GNR and solution processability of GO. Therefore, GORs could be a new class of solution-processable semiconducting materials attractive for optoelectronic applications. To demonstrate the hole-extraction capability for GOR, Liu et al. [46], developed graphene oxide ribbons of controlled bandgaps with a good solubility and excellent film-forming capability, which were demonstrated to be an excellent hole extraction layer for polymer solar cells to outperform their counterparts based on conventional hole

extraction materials. Fig. 2g shows the current density-voltage (J-V) measured from the ITO/GOR/P3HT: PCBM/Ca/Al device, compared with other reference devices, under AM1.5 G illumination. As can be seen, the device with GO as the HEL exhibited an open-circuit voltage (VOC) of 0.62 V, shortcircuit current density (JSC) of 8.42 mA/cm2, fill factor (FF) of 0.60, and power conversion efficiency (PCE) of 3.08% [46]. This performance is much better than that of the corresponding device based on the bare ITO without HEL. When GOR was used as HEL, the device showed a VOC of 0.62 V, JSC of 9.96 mA/cm2, FF of 0.67, and PCE of 4.14%. The much better performance of the GOR-based device than its GO-counterpart indicates that GOR is an excellent hole extraction material for polymer solar cells. In view of the energy level alignment shown in Fig. 2f, the excellent performance for the GOR-based device was attributed to the improved hole extraction and electron blocking capabilities. In addition, the formation of a thin and uniform film of GOR on the ITO electrode offered an additional advantage for the polymer solar cell application. Both the proper energy level alignment and the excellent film forming property of GOR contributed to the observed excellent device performance. Furthermore, the long-term stability tested by recording the device performance as a function of storage time in a N2filled glovebox without any device encapsulation. As shown in Fig. 2h, the PCE of the PEDOT:PSS-based device dropped to 75% of its original value after storage for 90 days. In contrast, the GOR-based device remained 86% of the original value under the same condition, suggesting that GOR is a more stable hole extraction interfacial layer than PEDOT:PSS. Tandem polymer solar cells, consisting of more than one (normally two) subcells connected by a charge recombination layer (i.e., interconnecting layer), hold great promise for enhancing the performance of PSCs (Fig. 3a). For an ideal tandem solar cell, the Voc equals to the sum of those of the subcells while keeping the short circuit current the same as the lower one, leading to an increased overall power conversion efficiency. The interconnecting layer plays an important role in regulating the tandem device performance. By modifying a graphene oxide (GO)/GO-Cs (cesium neutralized GO) bilayer with ultrathin Al and MoO3, Chen et al. [47], prepared an efficient interconnecting layer for tandem polymer solar cells to achieve a significantly increased Voc, 4


C. Hu et al.

Fig. 4. (a) Scheme of DSSC with an N-GF counter electrode. (b) Iodine reduction on the N-doped graphene surface. (c) Current density-voltage characteristics and (d) IPCE of DSSCs with the Pt, N-GF, rGOF, N-graphene film, and rGO film counter electrode. Copyright: 2012, Wiley. Reproduced with permission from [49].

effect of nitrogen-doping and foam-like structure that makes the N-GF counter electrode to show the superb performance. Besides, Fig. 4d shows the incident photon to charge carrier efficiency (IPCE) for DSSCs with the N-GF, rGOF, rGO, N-Graphene, and Pt counter electrode, respectively. As expected, the foam-like electrodes (N-GF and rGOF) showed relatively higher IPCEs than their film-type counterparts (i.e., N-graphene film and rGO film) [49]. Defective CNT microballs have also been used to replace the Ptbased cathode in DSSCs to achieve an energy conversion efficiency of about 7% [51], and higher than 80% of the Pt-based DSSCs [52]. Vertically-aligned CNT arrays have also been used as the counterelectrode catalysts in DSCs to exhibit a high power conversion efficiency [53,54], while carbon catalyst comprised of CNTs coated onto a flexible graphene paper achieved a PEC of 83% of that for a Pt electrode [55]. Recently, DSSCs with Co2+/3+ redox pairs were reported to exhibit a higher open-circuit voltage and a higher PCE than the iodine/ triiodide counterpart with the same type of counter electrode. Kavan et al. [56], found that, for Co2+/3+ redox pairs, a graphene film on FTO exhibited higher electrocatalytic activity and higher device efficiency than a platinum–FTO electrode at high illumination intensity, indicating that graphene is an excellent counter electrode material for DSSCs. More recently, Baek and co-workers have used graphene materials doped with heteroatoms (e.g., N, Te) to replace the Pt cathode in DSSCs with either I-/I3- or Co2+/3+ redox pairs, showing PECs above 10% [57–60]. These results indicate that heteroatom-doped graphene materials can be used as effective metal-free electrocatalysts as the counter electrodes to replace Pt in high-performance DSSCs. More generally, heteroatom-doped and defect-rich graphene materials have been demonstrated to be useful metal-free electrocatalysts for various redox reactions important to renewable energy technologies, including oxygen reduction reaction (ORR), the oxygen evolution reaction (OER), and hydrogen evolution reaction (HER), as described below.

reaching almost 100% of the sum of the subcell Vocs under standard AM 1.5 conditions (Fig. 3b). The GO-based interconnecting layer modified by ultrathin Al and MoO3 could provide an efficient recombination region for electrons and holes generated from the front and rear cells duo to excellent energy level matches for efficient charge transport and recombination (Fig. 3c). The power conversion efficiency of the optimized tandem solar cells exhibited a maximum PCE of 3.90% and Voc of 1.69 V, reaching to ~100% of the sum of Vocs of the subcells. These results from the prototype devices indicate that graphene materials are very promising for application as interconnecting layer in tandem polymer solar cells. 3.2. Dye sensitized solar cells Dye sensitized solar cells (DSSCs) have attracted extensive attention as a class of low-cost, high-efficiency solar cells since 1991 [48]. A typical DSSC device consists of a dye-adsorbed TiO2 photoanode, counter electrode, and iodide electrolyte (Fig. 4a). The counter (cathode) electrode plays a key role in regulating the DSSC device performance by catalyzing the reduction of the iodide-triiodide redox species used as a mediator to regenerate the sensitizer after electron injection (Fig. 4b). The ideal counter electrode material should possess a low sheet resistance, high reduction catalytic activity, good chemical stability, and low cost. Because of its excellent electrocatalytic activity for the iodine reduction, high conductivity and good chemical stability, platinum has been widely used as a counter electrode in DSSCs. However, the high cost and scarcity of Pt have limited the commercialization of the DSSC technology. Recently, much effort has been made to reduce or replace Pt-based electrode in DSSCs. [49]. As alternative for Pt-based catalysts, metal-free heteroatom-doped carbon catalysts have been demonstrated to effectively catalyze the reduction of I3- to I- in DSSCs [30,31,49–52]. In particular, a DSSC with a 3D N-doped graphene foam (N-GF) cathode could display a power conversion efficiency of 7.07%, comparable to its counterpart based on a Pt cathode (7.44%) [49]. Fig. 4c indicate that the performances of graphene foams (i.e., NGF and rGOF) are much better than the corresponding spin-coated graphene films (i.e., N-graphene and rGO films). It is the combined

4. Functionalized graphene materials with high electrocatalytic activities for fuel cells Fuel cell is one of the most promising clean and sustainable energy 5


C. Hu et al.

Fig. 5. (a) Schematic representation of a fuel cell. Adapted from https://en.wikipedia.org/wiki/Solid_oxide_fuel_cell (b) Scheme illustration of nitrogen species in nitrogen-containing graphitic carbons. Reproduced with permission from [65]. Copyright: 2015, AAAS. (c) Left, The calculated charge density distribution on N-doped carbon nanotube (NCNT) and Right, schematic representations of possible adsorption modes of an oxygen molecule. Theoretical calculations by B3LYP hybrid density functional theory (DFT) indicated that the improved catalytic performance is contributed to the electron-accepting ability of the nitrogen atoms, which creates net positive charge on adjacent carbon atoms in the CNT plane of NCNT. More importantly, the nitrogen-induced charge delocalization could also change the chemisorption mode of O2 from the usual end-on adsorption (Pauling model) at the nitrogen-free CNT (CCNT) surface (top, Right) to a side-on adsorption (Yeager model) onto the NCNT electrodes (bottom, Right), which could effectively weaken the O−O bonding, facilitating oxygen reduction at the NCNT electrodes. Reproduced with permission from [64]. Copyright: 2009, American Association for the Advancement of Science. (d) Schematic representation of the ORR mechanism on metal-free N-doped carbon catalysts. Reproduced with permission from [66]. Copyright: 2016, American Association for the Advancement of Science. (a) Reproduced with permission from [65]. (b) Reproduced with permission from [64]. (c) Reproduced with permission from [66].

based electrode in fuel cells, Dai and co-workers [64] have discovered a new class of carbon nanomaterials, which, as alternative ORR catalysts, could dramatically reduce the cost and increase the efficiency of fuel cells. In particular, it was found that vertically-aligned nitrogen-doped carbon nanotube (VA-NCNT) arrays can act as a metal-free electrode to catalyze an ORR process free from CO “poisoning” with a 3-times higher electrocatalytic activity, much smaller crossover effect, and better long-term operational stability than that of the commercial platinum-based electrode (C2–20, 20% platinumonVulcanXC-72R; E-TEK) in alkaline fuel cells [64]. Based on the experimental observations and quantum mechanics calculations [64], the improved catalytic performance is attributed to the electron-accepting ability of the nitrogen atoms, which creates net positive charge and unpaired electron density on adjacent carbon atoms in the nanocarbon structure for facilitating the four-electron ORR (Fig. 5b, c) [64,65]. Uncovering this new ORR mechanism in the nitrogen-doped carbon nanotube electrode is significant as the same principle could be applied to the development of various other metal-free efficient ORR catalysts for fuel cell applications and even beyond fuel cells. Indeed, this seminal paper [64] has made a large impact on the materials and energy communities, which has led to a huge amount of literature on carbon electrocatalysts doped with

technologies that can directly convert fuel into electricity with a high efficiency [61]. As schematically shown in Fig. 5a, hydrogen is split into its constituent electrons and protons on electrode (the anode) in a fuel cell. While the electrons flow out of the anode to provide electrical power and end up at the cathode to reduce oxygen, the protons diffuse through an electrolyte to combine with the oxygen species at the cathode to produce the only by-product water. For fuel cells, cathodic oxygen reduction plays an essential role in producing electricity and is a major limit to the cell performance [62]. The ORR would naturally happen very slowly without catalysts on the cathode. Noble metals (e.g., platinum) are needed to accelerate the ORR. Although platinum nanoparticles have long been regarded as the best catalyst for the ORR, the Pt-based electrode suffers multiple drawbacks, including its timedependent drift, methanol crossover, and CO deactivation [63]. This, together with the high cost of platinum and its scarcity, has made these catalysts the primary barrier to mass market fuel cells for commercial applications. Therefore, the large-scale practical application of fuel cells is difficult to realize if the expensive platinum-based electrocatalysts for ORR cannot be replaced by other efficient, low cost, and stable electrodes. Along with the intensive research efforts in reducing or replacing Pt6


C. Hu et al.

Fig. 6. (a) A schematic representation for mechanochemically drive edge-heteroatoms reaction between the in-situ generated active carbon species (gold balls) and reactant halogens (twin green balls). Active carbon species were generated by homolytic bond cleavages of graphitic C-C bonds and reacted with halogen molecules to produce edge-halogenated graphene nanoplatelets (XGnPs) in a sealed ball-mill capsule and the remnant active carbon species are terminated upon subsequent exposure to air moisture. Red and gray balls stand for oxygen and hydrogen, respectively. (b) A schematic representation for the edge expansions of XGnPs caused by the edge-halogens. (c) Linear sweep voltammograms (LSV) show a gradual increase in current and a positive shift in the onset potential along the order of the pristine graphite < ClGnP < BrGnP < IGnP < Pt/C. Copyright: 2013, AAAS. (d) HOMO distribution of edge-sulfurized graphene nanoplatelets. Copyright: 2013, Wiley-VCH. Copyright: 2013, Wiley-VCH. And (f) 3D graphene foam. Copyright: 2013, Nature Group. (a) Reproduced with permission from [69]. (b) Reproduced with permission from [72]. (c) Schematic representation of (e) 3D CNT-graphene pillar, Reproduced with permission from [75]. (d) Reproduced with permission from [76].

are totally distinguished when the dopant locations of the heteroatoms are different even with the same doping elements and types. Edge functionalization by selectively doping heteroatoms at the edge of graphene sheets with minimal damage of the carbon basal plane [7,67,68], which has been considered to be of particular importance for imparting the electrocatalytic activity to graphene materials for specific applications. Recent studies indicate that various heteroatoms, such as F, Cl, Br, I, S, Sb, can be introduced at the edge of graphene nanoplatelets by ball milling of graphite powder in the presence of appropriate chemicals (Fig. 6a) [69–73]. During the ball milling of graphite, the strong shear forces generated between the high-speed rotating balls caused the mechanochemical cracking of the graphitic C-C bonds, leading to spontaneous incorporation of heteroatoms on the surface of the broken graphite particles. It was found that there was a beneficial charge transfer for favorable O2 adsorption and a weakened O-O bond for

various heteroatoms, such as N, B, O, P, S, Cl, Se, Br, and I, and the number of publications is still rapidly increasing every year. The dopant type of heteroatoms is of vital importance to make metal-free carbon catalysts to be efficient electrochemical catalysts. In particular, nitrogen can exist in many different types in N-doped carbon catalysts, including pyridine-like, pyrrole-like, graphitic nitrogen, and pyridine-N-oxide (Fig. 5b) [30,65]. Thus, it is important to understand effects of the chemical nature of the N dopant on the electrocatalytic performance. In this context, Guo et al. [66], have recently revealed pyridinic N as the active site for ORR in acidic electrolyte by using highly oriented pyrolitic graphite (HOPG) model catalysts with welldefined π conjugation and well-controlled doping of N species. Particularly, it was found that the ORR active sites in N-doped carbon materials are carbon atoms with Lewis basicity next to pyridinic N (Fig. 5d) [66]. The property and activity of the heteroatom-doped carbon catalysts 7


C. Hu et al.

Fig. 7. (a) Cross-section scanning electron microscope (SEM) image of the porous catalyst layer of N-G-CNT/CB/Nafion. Purple arrows in (a) indicate the parallelly separated N-G-CNT sheets with inter dispersed porous KB agglomerates. Schematic drawings of O2 efficiently diffused through the carbon black separated N-G-CNT sheets (b) but not the densely packed N-GCNT sheets (c). Copyright: 2015, AAAS. (d) Fabrication of the N, S-CNT following a two-step “graft-and-pyrolyze” route. The oxidized CNT (ox-CNT) was mixed with dopamine (DA) in PBS buffer to obtain CNT-PDA hybrid. After addition of R-SH, CNT-PDA modified with thiol (CPS) was synthesized, which produced N, S-CNT after pyrolysis. R-SH = 2-mercaptoethanol. (e) Transmission electron microscopy (TEM) image of the N, S-CNT. Copyright: 2017, Wiley-VCH. (a) Reproduced with permission from [79]. (b) Reproduced with permission from [80].

cells [79]. This is because O2 efficiently diffused through the CB separated N-G-CNT sheets (Fig. 7b), but not the densely packed N-G-CNT sheets (Fig. 7c). In an independent study, Qu et al. [80], prepared the N and S surface-doped CNTs (Fig. 7d, e) with improving the exposure of active sites, which display superb bifunctional catalytic activities for both HER and OER in alkaline solutions. Also, a N and O codoped graphene-CNT (NG-CNT) hydrogel film has been demonstrated to be an efficient electrocatalyst [81]. Combined experimental and theoretical data indicate that the excellent performance is attributable to synergistic effects arising from the multiple doping and the efficient mass and charge transfer characteristic of the 3D porous architecture, leading to high multifunctional electrocatalytic activities for both HER and OER attractive for metal-batteries and water splitting, as discussed in the succeeding sections.

edge-halogenation graphene, resulting in high-performance ORR catalysts. As atomic size decreased in the order of I < Br < Cl (Fig. 6b) [69], the d spacings were 0.35 nm, 0.36 nm, and 0.37 nm for ClGnP, BrGnP, and IGnP, respectively. Thus, the ORR activities of XGnPs along the order of ClGnP < BrGnP < IGnP (Fig. 6c). First-principles investigation based on density functional theory (DFT) methods has revealed the distributions of the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) are strongly polarized for edge-sulfurized graphene, which also facilitates electron transfer during the ORR process (Fig. 6d) [72]. For heterogeneous catalysis, the structural architecture to support the active sites of the catalysts are also expected to affect catalytic performance as those unexposed active sites induced by the bulk dopant atoms are not accessible to reactants, and consequently contribute little to the catalytic activity. For instance, the novel N-doped carbon coaxial nanocables (CNT@NCNT) with the pristine CNT as the core and the Ndoped carbon layers as the shell are ideal metal-free catalysts [74]. In this particular case, the outer CNT wall provides the reaction sites while the intact inner tubes act as the effective electronic conducting pathway for effective charge transfer through electron tunneling between the outer wall and inner tubes. More generally, porous 3D carbon architectures, such as the 3D CNT-graphene pillars (Fig. 6e) [75] and graphene foams (Fig. 6f) [76], have been demonstrated to be ideal electrocatalysts with an efficient charge transfer via the carbon framework, electrolyte/reactant transport through the connected porous channel, and good durability [30,77,78]. In this reagrd, Shui et al. [79], have recently demonstrated that both the VA-NCNT array and a rationally designed 3D micro structure consisting of carbon black (KB) separated N-doped carbon nanotubes and graphene composites (N-G-CNT/KB) (Fig. 7a) to exhibit significantly better long-term operational stabilities and comparable gravimetric power densities with respect to the best nonprecious metal catalysts in acidic polymer electrolyte membrane

5. Functionalized graphene materials with multifunctional electrocatalytic activities for metal-air batteries and water splitting Although N, B, S, O, and P-doped graphene materials have been demonstrated to be promising active catalysts for different types of reactions, their electrocatalytic performance can be further improved by codoping with two or more different heteroatoms (e.g., N and B, N and S, or, N and P), as the increased number of dopant heteroatoms and the electronic interactions between the different doped heteroatoms often generate synergistic effects in respect to single heteroatom-doped counterparts [82]. In this context, Dai's group was the first to report significantly enhanced ORR activity for N, B co-doped CNTs with respect to single N- or B-doped CNTs [83]. Since then, codoping and even multiple doping have been used as effective ways to impart multifunctional catalytic activities and to enhance the catalytic performance 8


C. Hu et al.

Fig. 8. (a) NBO population analysis of six different nonmetallic heteroatoms in a graphene matrix. pN and gN represent pyridinic and graphitic types of N, respectively. The inset shows the proposed doping sites for different elements: sites 1 and 2 are the edge and center in-plane sites, respectively, and site 3 is an out-of-plane center site in graphene. (b) The calculated free energy (DGH* ) diagram for the HER at the equilibrium potential (URHE= 0 V) for N and/or P-doped graphene models. Copyright: 2014, American Chemical Society. ORR (c) and OER (d) volcano plots of respectively, for N-doped, P-doped and N, P-doped graphene. (e) LSV curves of NPMC-1000, NPMC-1100, RuO2, and the commercial Pt/C catalyst on an RDE in 0.1 M KOH solution, showing the electrocatalytic activities for both the ORR and OER. Copyright: 2015, Nature Publishing Group). (f) Scheme of N, P, and F Tri-doped graphene as a multifunctional catalyst for ORR, OER, and OER. (g) Polarization and power density curves of Zn-air batteries using GOPANi31-FP, GO-PANi51-FP as ORR/OER catalyst. (h) O2 and H2 production volumes (powered by Zn-air batteries) as a function of water-splitting time. Copyright: 2016, Wiley-VCH. (a) Reproduced with permission from [85]. (b) Reproduced with permission from [86]. (c) Reproduced with permission from [87].

isolated P-doped counterparts, respectively. Overall, the minimum overpotentials of N, P co-doped graphene for ORR and OER are lower than those of the best catalysts identified theoretically (ORR on Pt and OER on RuO2), indicating that the bifunctional N, P co-doped graphene catalyst could outperform its metal/metal-oxide counterparts. Indeed, the resultant NPMCs showed superb bifunctional catalytic activities towards the ORR and OER in 0.1 M KOH (Fig. 8e). Based on the NPMC bifunctional catalyst, Zhang et al., further developed the first highly efficient Zn-air batteries (both primary and secondary) [86]. Primary batteries showed an open circuit potential of 1.48 V, a specific capacity of 735 mAh gZn−1 (corresponding to an energy density of 835 Wh kgZn−1), a peak power density of 55 mW cm−2, and could sustain stable operation for 240 h after mechanical recharging. Twoelectrode rechargeable batteries could be stably cycled for 180 cycles at 2 mA/cm2. Even at 2 mA/cm2, a three-electrode rechargeable battery showed a good stability (600 cycles for 100 h), which was comparable to that of Pt/C [86]. Furthermore, Zhang and Dai [87] prepared a N, P, and F tri-doped graphene foam for ORR, OER, and HER (Fig. 8f). The trifunctional metal-free catalyst was further used as an OER-HER bifunctional catalyst for oxygen and hydrogen gas production in an electrochemical water-splitting unit, which was powered by an integrated Zn-air battery based on an air electrode made from the same electrocatalyst for ORR

of carbon-based metal-free catalysts for metal-air batteries, water splitting, and many other applications. The potential applications of carbon-based electrocatalysts in metalair batteries and water splitting have led to intensive studies on the development of multifunctional catalysts for ORR, OER, and HER via heteroatom doping. Recently, graphene and other graphitic carbon materials codoped with nitrogen/phosphine and nitrogen/sulfur have been reported as efficient even multifunctional catalysts [84]. On the basis of the natural bond orbital (NBO) population analyses and theoretical prediction from DFT studies (Fig. 8a, b), Zheng et al. [85], explored graphene models doped or codoped with several heteroatoms (N, B, O, S, P, and/or F) and predicted that N and P codoping could afford the optimal HER activity. The incorporated N and P heteroatoms could coactivate the adjacent C atom in the graphene matrix by affecting the energy levels of its valence orbital to induce a synergistically enhanced reactivity toward the HER. In an independent study, Zhang et al. [86], fabricated N and P codoped mesoporous nanocarbon (NPMC) foams by pyrolysis at different temperatures (1000/1100 °C, denoted as NPMC-1000 or NPMC-1100, respectively). Fig. 8c, d present volcano plots of the overpotential versus descriptors for various reaction sites on N, P co-doped graphene structures in alkaline medium [86]. The minimum ORR and OER overpotentials of N, P-coupled structures were identified lower than those of the isolated N-doped and the 9


C. Hu et al.

Fig. 9. (a) Schematic representation of an ideal integrated energy system. (b) The multifunctional hybrid energy device of perovskite cells assembled with water splitting device and fuel cell. (c) J–V curves of the perovskite tandem cell under dark and simulated AM 1.5 G 100 mW cm−2 illumination, and the N, S-3DPG on carbon fiber woven electrodes in a two-electrode configuration. (d) Current density-time curve of the integrated water splitting device without external bias under chopped simulated AM 1.5 G 100 mW cm−2 illumination. (e) Gravimetric polarization and power density curves of the fuel cell using H2 and O2 collected from water splitting caused by the perovskite tandem cells. Reprinted with permission from Ref. [90], Copyright 2017, Elsevier.

through photocatalytic and/or photoelectrochemical water splitting in conjunction with a fuel cell. The electricity thus produced can even be stored in an integrated metal-air battery, as wish. To realize this ideal energy cycle, however, efficient catalysts are required for HER, OER, and ORR. ORR plays an important role in fuel cells while OER and HER are essential for metal (e.g., Zn)-air batteries and water splitting (Fig. 9a). As demonstrated above, low-cost graphitic carbon materials doped with multiple heteroatoms can act as efficient multifunctional catalysts for ORR, OER, and HER. By rationally designing N,S co-doped porous graphitic networks with a cross-linked three-dimensional (3D) structure (N,S-3DPG), Dai and co-workers have developed low-cost, highly efficient tri-functional catalysts for simultaneously catalyzing HER, OER, and ORR in alkaline electrolytes [90]. The unusual tri-functional electrocatalytic performance of N,S-3DPG was attributable to a synergistic effect arising from the N,S-co-doping and its unique 3D porous network structure. Based on the newly-developed ORR/OER/HER tri-functional catalyst from earth-abundant carbon materials, an integrated energy system was devised, in which assembled perovskite solar cells were used for photoelectrochemical water splitting (OER and HER) to produce hydrogen and oxygen gases for a fuel cell (ORR) to renewably generate clean electricity, which represents a new conceptually important approach to cost-effective generation of clean and renewable electricity from sunlight and water (Fig. 9b) [90]. Although the above work indicates that the construction of integrated energy systems consisting of (photo)

(Fig. 8g, h) [87]. Other examples for multifunctional carbon catalysts include the O,N,P tri-doped porous graphite nanocarbons directly grown on oxidized carbon cloth as an effective carbon nanomaterialbased full water splitting electrocatalysts [88], and N, O, and S-tridoped polypyrrole derived nanoporous carbons (NOSCs) that can serve as metal-free, selective electrocatalysts for ORR and alcohol oxidation reaction, respectively [89]. Thus, multiple doping is a powerful approach for the design and development of carbon-based metal-free catalysts for a variety of electrocatalytic reactions. Despite tremendous progress in the synthesis of dual-doped carbon nanomaterials, research on trip or multi-doped carbon-based multifunctional catalysts is a recent development and the associated mechanistic understanding is largely lacking. However, this is clearly an area in which future work would be of value. 6. Functionalized graphene materials with multifunctional electrocatalytic activities for integrated energy systems As can be seen from above discussions, chemical energy in the form of H2 can be generated by splitting water, which, in turn, can be used as fuel for fuel cells to produce clean electricity. Therefore, low-cost production of H2 by either photocatalytic water splitting or electricity generated from photovoltaic devices for photoelectrochemical water splitting is highly desirable. As schematically shown in Fig. 9a, clean electricity can be continuous generated from water and sunlight, 10


C. Hu et al.

1400274, DOD-AFOSR-MURI (FA9550-12-1–0037)) NASA (NNX16AD48A, NNC16CA42C), Key Program of National Natural Science Foundation of China (51732002), The National Key Research and Development Program of China (2017YFA0206500), Distinguished Scientist Program at BUCT (buctylkxj02), and National Basic Research Program of China (2011CB013000).

electrochemical water splitting coupled with a fuel cell to continuously generate clean electricity from water and sunlight is feasible, there is still much to do to realize the ideal integrate energy system (Fig. 9a), particularly with photocatalytic water splitting coupled with fuel cell and metal-air battery for practical applications. Despite significant progress has been made in the area of metal-free carbon-based catalysts for various electrochemical catalyses, the exploration of their applications for other than the ORR is a recent development and still facing multiple challenges. For instance, the longterm stability of carbon-based catalysts is a critical issue. There are some reported nanocarbon materials with special doped structures exhibited super stability, even better than that of commercial platinum catalysts. In most cases, however, the long-term durability of carbonbased catalysts, particularly in practical devices, has not been tested with a standard evaluation protocol [32]. It is important for the longterm stability of carbon-based metal-free catalysts in practical devices to be evaluated, though somewhat challenging and time-consuming.

References [1] L. Dai (Ed.), Carbon Nanotechnology: Recent Developments in Chemistry, Physics, Materials Science and Device Applications, Elsevier, Amsterdam, 2006. [2] L. Dai, Intelligent Macromolecules for Smart Devices: From Materials Synthesis to Device Applications, Springer-Verlag, Berlin, 2004. [3] S. Wang, D. Yu, L. Dai, D. Chang, J. Baek, ACS Nano 5 (2011) 6202–6209. [4] C. Hu, L. Song, Z. Zhang, N. Chen, Z. Feng, L. Qu, Energy Environ. Sci. 8 (2015) 31–54. [5] H. Bai, C. Li, G. Shi, Adv. Mater. 23 (2011) 1089–1115. [6] L. Dai, Acc. Chem. Res. 46 (2013) 31–42. [7] Z. Xiang, Q. Dai, J. Chen, L. Dai, Adv. Mater. 28 (2016) 6253–6261. [8] H.W. Kroto, J.R. Heath, S.C. Obrien, R.F. Curl, R.E. Smalley, C60: buckminsterfullerene, Nature 318 (1985) 162–163. [9] P.J.F. Harris, Carbon Nanotubes and Related Structures: New Materials for the Twenty-first Century, Cambridge university press, Cambridge, 1999. [10] Z. Zhu, N. Wei, H. Xie, R. Zhang, Y. Bai, Q. Wang, C. Zhang, S. Wang, L. Peng, L. Dai, F. Wei, Sci. Adv. 2 (2016) e1601572. [11] R.F. Zhang, Y.Y. Zhang, Q. Zhang, H.H. Xie, W.Z. Qian, F. Wei, ACS Nano 7 (2013) 6156–6161. [12] H. Chen, A. Roy, J.-B. Baek, L. Zhu, J. Qu, L. Dai, Mater. Sci. Eng. Rep. 70 (2010) 63–91. [13] A. Peigney, C. Laurent, E. Flahaut, R.R. Bacsa, A. Rousset, Carbon 39 (2001) 507–514. [14] X. Lu, Z.F. Chen, Chem. Rev. 105 (2005) 3643–3696. [15] D. Tasis, N. Tagmatarchis, A. Bianco, M. Prato, Chem. Rev. 106 (2006) 1105–1136. [16] N. Karousis, N. Tagmatarchis, D. Tasis, Chem. Rev. 110 (2010) 5366–5397. [17] L. Dai, T. Lin, T. Ji, V. Bajpai, Aust. J. Chem. 56 (2003) 635. [18] W. Lu, J. Baek, L. Dai, Carbon Nanomaterials for Advanced Energy Systems, John Wiley &Sons, New York, 2015. [19] A.K. Geim, K.S. Novoselov, Nat. Mater. 6 (2007) 183–191. [20] A.A. Balandin, S. Ghosh, W.Z. Bao, I. Calizo, D. Teweldebrhan, F. Miao, C.N. Lau, Nano Lett. 8 (2008) 902–907. [21] C. Lee, X.D. Wei, J.W. Kysar, J. Hone, Science 321 (2008) 385–388. [22] K.S. Novoselov, ECS Trans. 19 (2009) 3–17. [23] D. Yu, E. Nagelli, F. Du, L. Dai, J. Phys. Chem. Lett. 1 (2010) 2165–2173. [24] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V. Dubonos, I.V. Grigorieva, A.A. Firsov, Science 306 (2004) 666–669. [25] L. Qu, Y. Liu, J. Baek, L. Dai, ACS Nano 4 (2010) 1321–1326. [26] D. Yu, E. Nagelli, F. Du, W. Yang, K. Ratinac, S. Ringer, P. Thordarson, J. Gooding, F. Braet, Angew. Chem., Int. Ed. 49 (2010) 2114–2138. [27] X. Xie, L. Qu, C. Zhou, Y. Li, J. Zhu, H. Bai, G. Shi, L. Dai, ACS Nano 4 (2010) 6050–6054. [28] K.D. Sen, C.K. Jørgensen, J.A. Alonso, Electronegativity, Springer-Verlag, Berlin, 1987. [29] J. Liu, Y. Xue, M. Zhang, L. Dai, MRS Bull. 37 (2013) 1265–1272. [30] L. Dai, Y. Xue, L. Qu, H. Choi, J. Baek, Chem. Rev. 115 (2015) 4823–4892. [31] J. Zhang, L. Dai, ACS Catal. 5 (2015) 7244–7253. [32] X. Liu, L. Dai, Nat. Rev. Mater. 1 (2016) 16064. [33] D. Yu, Y. Xue, L. Dai, J. Phys. Chem. Lett. 3 (2012) 2863–2870. [34] A. Shen, Y. Zou, Q. Wang, R. Dryfe, X. Huang, S. Dou, L. Dai, S. Wang, Angew. Chem. Int. Ed. 53 (2014) 10804–10808. [35] W. Yuan, Y. Zhou, Y. Li, C. Li, H. Peng, J. Zhang, Z. Liu, L. Dai, G. Shi, Sci. Rep. 3 (2013) 2248. [36] R.H. Bube, Photoelectronic Properties of Semiconductors, Cambridge University Press, Cambridge, UK, 1992. [37] L. Dai, J. Macromol. Sci. Rev. Macromol. Chem. Phys. 39 (1999) 273–387. [38] J. Liu, Y. Xue, Y. Gao, D. Yu, M. Durstock, L. Dai, Adv. Mater. 24 (2012) 2227. [39] Y. Wang, S.W. Tong, X.F. Xu, B. Özyilmaz, K.-P. Loh, Adv. Mater. 23 (2011) 1514–1518. [40] S. Bae, H. Kim, Y. Lee, X.F. Xu, J.-S. Park, Y. Zheng, J. Balakrishnan, Y. Lei, H.R. Kim, Y.I. Song, Y.J. Kim, B. Özyilmaz, J.H. Ahn, B.H. Hong, S. Lijima, Nat. Nanotechnol. 5 (2010) 574–578. [41] V.C. Tung, J.-H. Huang, I. Tevis, F. Kim, J.Y. Kim, C.-W. Chu, S.I. Stupp, J.X. Huang, J. Am. Chem. Soc. 133 (2011) 4940–4947. [42] J. Liu, Y. Xue, L. Dai, J. Phys. Chem. Lett. 3 (2012) 1928–1933. [43] X. Li, X. Wang, L. Zhang, S. Lee, H. Dai, Science 319 (2008) 1229–1232. [44] D.V. Kosynkin, A.L. Higginbotham, A. Sinitskii, J.R. Lomeda, A. Dimiev, B.K. Price, J.M. Tour, Nature 458 (2009) 872–876. [45] M.J. Allen, V.C. Tung, R.B. Kaner, Chem. Rev. 110 (2010) 132–145. [46] J. Liu, G.-H. Kim, Y. Xue, J.Y. Kim, J.-B. Baek, M. Durstock, L. Dai, Adv. Mater. 26 (2014) 786–790. [47] Y. Chen, W.C. Lin, J. Liu, L. Dai, Nano Lett. 14 (2014) 1467–1471. [48] B. O'Regan, M. Grätzel, Nature 353 (1991) 737–740. [49] Y. Xue, J. Liu, H. Chen, R. Wang, D. Li, J. Qu, L. Dai, Angew. Chem. Int. Ed. 51 (2012) 12124–12127.

7. Concluding remarks Graphene materials show unique electronic structures and properties attractive for various potential applications. However, practical applications of graphene materials have been severely limited by their intractability and/or zero-bandgap. Heteroatom-doping via the introduction of heteroatoms (e.g., nitrogen, boron, phosphorus) into the graphitic carbon network could impart film forming capability and cause electron modulation to tune their optoelectronic properties and/ or chemical activities suitable for specific applications. Doping of graphene materials with single or multiple heteroatoms has opened up a rich field of charge-extraction materials with tunable work functions for optoelectronic applications and multifunctional metal-free catalysts for low-cost, but efficient, energy conversion and storage. In this article, we present a timely focused review by summerizing new concepts and strategies recently reported by us and some other groups for controlled doping/functionalization of graphene materials for the development of efficient energy conversion and storage devices, including solar cells, fuel cells, metal-air batteries, and their combinations. To prepare high-performance carbon electrocatalysts, a better understanding of catalytic mechanism is needed. A combine experimental and theoretical approach would be critical as computer simulation and calculation are powerful in searching the active centers and studying the basic science behind the electrocatalysis. The results from the computer simulation can guide the design and development of carbonbased catalysts with a desirable activity and stability for specific reactions in energy conversion and storage. Then, new synthetic and/or doping strategies must be developed to precisely control the atomic location, content, and the distribution of the catalytic active centers in carbon-based catalysts. It is also important to develop in-situ, real-time characterization techniques to probe the nature of the active sites and to understand the electrocatalytic activity and stability of metal-free carbon catalysts. Last but not least, macroscopic structure and threedimensional architecture design and evaluation are also important for fabricating the catalytic materials and electrodes with an optimized catalytic performance in practical devices. Recent studies have clearly revealed the versatility of heteroatomdoping for making functionalized graphene materials with desired structures and properties for integrated energy systems for low-cost, renewable generation of clean electricity from water and sunlight. Although tremendous progress has been achieved, there still are challenges to realize the ideal integrate energy systems of practical significance, continued research in this exciting field would be of value. Acknowledgements The authors thank our colleagues for their contributions to the work cited. We are also grateful for financial support from NSF (CMMI11


C. Hu et al.

[70] I. Jeon, H. Choi, S. Jung, J. Seo, M. Kim, L. Dai, J. Baek, J. Am. Chem. Soc. 135 (2013) 1386–1393. [71] I.-Y. Jeon, S.-Y. Bae, J.-M. Seo, J.-B. Baek, Adv. Funct. Mater. 25 (2015) 6961–6975. [72] I. Jeon, S. Zhang, L. Zhang, H. Choi, J. Seo, Z. Xia, L. Dai, J. Baek, Adv. Mater. 25 (2013) 6138–6145. [73] I. Jeon, M. Choi, H. Choi, S. Jung, M. Kim, J. Seo, S. Bae, S. Yoo, G. Kim, H. Jeong, N. Park, J. Baek, Nat. Commun. 6 (2015) 7123. [74] G.L. Tian, Q. Zhang, B.S. Zhang, Y.G. Jin, J.Q. Huang, D.S. Su, F. Wei, Adv. Funct. Mater. 24 (2014) 5956–5961. [75] G.K. Dimitrakakis, E. Tylianakis, G.E. Froudakis, Nano Lett. 8 (2008) 3166–3170. [76] X.B. Wang, Y.J. Zhang, C.Y. Zhi, X. Wang, D.M. Tang, Y.B. Xu, Q.H. Weng, X.F. Jiang, M. Mitome, D. Golberg, Y. Bando, Nat. Commun. 4 (2013) 2905. [77] Z.S. Wu, S.B. Yang, Y. Sun, K. Parvez, X.L. Feng, K. Mulle, J. Am. Chem. Soc. 134 (2012) 9082–9085. [78] Y. Zhao, C. Hu, L. Song, L. Wang, G. Shi, L. Dai, L. Qu, Energy Environ. Sci. 7 (2014) 1913–1918. [79] J. Shui, M. Wang, F. Du, L. Dai, Sci. Adv. 1 (2015) e1400129. [80] K.G. Qu, Y. Zheng, Y. Jiao, X.X. Zhang, S. Dai, S.Z. Qiao, Adv. Energy Mater. 7 (2017) 1602068. [81] S. Chen, J.J. Duan, M. Jaroniec, S.Z. Qiao, Adv. Mater. 26 (2014) 2925–2930. [82] J. Zhang, L. Qu, G. Shi, J. Liu, J. Chen, L. Dai, Angew. Chem. 128 (2016) 2270–2274. [83] S. Wang, E. Iyyamperumal, A. Roy, Y. Xue, D. Yu, L. Dai, Angew. Chem. Int. Ed. 50 (2011) 11756–11760. [84] C. Hu, L. Dai, Adv. Mater. 29 (2017) 1604942–1604950. [85] Y. Zheng, Y. Jiao, L.H. Li, T. Xing, Y. Chen, M. Jaroniec, S.Z. Qiao, ACS Nano 8 (2014) 5290–5296. [86] J. Zhang, Z. Zhao, Z. Xia, L. Dai, Nat. Nanotechnol. 10 (2015) 444–452. [87] J. Zhang, L. Dai, Chem. Int. Ed. 55 (2016) 13296–13300. [88] J.P. Lai, S.P. Li, F.X. Wu, M. Saqib, R. Luque, G.B. Xu, Energy Environ. Sci. 9 (2016) 1210–1214. [89] Y.Y. Meng, D. Voiry, A. Goswami, X.X. Zou, X.X. Huang, M. Chhowalla, Z.W. Liu, T. Asefa, J. Am. Chem. Soc. 136 (2014) 13554–13557. [90] C. Hu, X. Chen, Q. Dai, M. Wang, L. Qu, L. Dai, Nano Energy 41 (2017) 367–376.

[50] J.D. Roy-Mayhew, D.J. Bozym, C. Punckt, I.A. Aksay, ACS Nano 4 (2010) 6203–6211. [51] S.H. Seo, S.Y. Kim, B.K. Koo, S.I. Cha, D.Y. Lee, Langmuir 26 (2010) 10341–10346. [52] S.I. Cha, B.K. Koo, S.H. Seo, D.Y. Lees, J. Mater. Chem. 20 (2010) 659–662. [53] K.S. Lee, W.J. Lee, N.-G. Park, S.O. Kim, J.H. Park, Chem. Comm. 47 (2011) 4264–4266. [54] P. Dong, C.L. Pint, M. Hainey, F. Mirri, Y.J. Zhan, J. Zhang, M. Pasquali, R.H. Hauge, R. Verduzco, M. Jiang, H. Lin, J. Lou, ACS Appl. Mater. Interfaces 3 (2011) 3157–3161. [55] S. Li, Y. Luo, W. Lv, W. Yu, S. Wu, P. Hou, Q. Yang, Q. Meng, C. Liu, H.-M. Cheng, Adv. Energy Mater. 1 (2011) 486–490. [56] L. Kavan, J.-H. Yum, M.K. Nazeeruddin, M. Grätzel, ACS Nano 5 (2011) 9171–9178. [57] M.J. Ju, J.C. Kim, H.-J. Choi, I.T. Choi, S.G. Kim, K. Lim, J. Ko, J.-J. Lee, I.-Y. Jeon, J.-B. Baek, H.K. Kim, ACS Nano 7 (2013) 5243–5250. [58] I.-Y. Jeon, H.M. Kim, D.H. Kwon, S.-M. Jung, J.-M. Seo, S.-H. Shin, I.T. Choi, Y.K. Eom, H.K. Kim, M.J. Ju, J.-B. Baek, Nano Energy 30 (2016) 867–876. [59] M.-J. Ju, I.Y. Jeon, H.M. Kim, J.I. Choi, S.-M. Jung, J.-M. Seo, I.T. Choi, S.H. Kang, H.S. Kim, M.J. Noh, J.-J. Lee, H.Y. Jeong, H.K. Kim, Y.-H. Kim, J.-B. Baek, Sci. Adv. 2 (2016) e1501459. [60] S.-Y. Bae, D.H. Kweon, J. Mahmood, M.-J. Kim, S.-Y. Yu, S.-M. Jung, S.-H. Shin, M.J. Ju, J.-B. Baek, Nano Energy 34 (2017) 533–540. [61] K. Kordesch, G. Simader (Eds.), Fuel Cells and Their Applications, Wiley-VCH, Berlin, 2006. [62] B.C.H. Steele, A. Heinzel, Nature 414 (2001) 345–352. [63] C. Sealy, Mater. Today 11 (2008) 65–68. [64] K. Gong, F. Du, Z. Xia, M. Dustock, L. Dai, Science 323 (2009) 760–764. [65] J. Zhang, Z. Xia, L. Dai, Sci. Adv. 1 (2015) e1500564. [66] D.H. Guo, R. Shibuya, C. Akiba, S. Saji, T. Kondo, J. Nakamura, Science 351 (2016) 361–365. [67] L. Tao, Q. Wang, S. Dou, Z. Ma, J. Huo, S. Wang, L. Dai, Chem. Commun. 52 (2016) 2764–2767. [68] A. Shen, Y. Zou, Q. Wang, R. Dryfe, X. Huang, S. Dou, L. Dai, S. Wang, Angew. Chem. Int. Ed. 53 (2014) 10804–10808. [69] I. Jeon, H. Choi, M. Choi, J. Seo, S. Jung, M. Kim, S. Zhang, L. Zhang, Z. Xia, L. Dai, N. Park, J. Baek, Sci. Rep. 3 (2013) 1810.

12