RSC Advances

PAPER

Key Laboratory of Automobile Materials,

Materials Science and Engineering, Jilin

E-mail: jiangq@jlu.edu.cn; Fax: +86-431-85

Cite this: RSC Adv., 2014, 4, 28400

Received 14th April 2014Accepted 13th June 2014

DOI: 10.1039/c4ra03387f

www.rsc.org/advances

28400 | RSC Adv., 2014, 4, 28400–284

Pt monatomic wire supported on graphenenanoribbon for oxygen reduction reaction

B. B. Xiao, X. Y. Lang and Q. Jiang*

Among the issues inherent to fuel cells, the high cost of the Pt electrocatalyst restricts its widespread

application. Controlling nano- or subnano-material structures could improve the utilization of Pt. Here,

the catalytic activity of the oxygen reduction reaction (ORR) on Pt monatomic wire that is supported on

the zigzag edges of graphene nanoribbon (Pt–GNR) is studied using density functional theory. It is found

that Pt–GNR is inert for ORR, due to the strong binding of OH and H2O. However, when Pt–GNR is

covered by the chain of OH and H2O (cPt–GNR), it becomes catalytically active for ORR. Through the

free energy diagrams on cPt–GNR, we demonstrate that the highest potential U for ORR as an

exothermic process is 0.82 eV. When U is larger than 0.82 V, the rate-determined step (RDS) of ORR is

located at the reduction of O2* to OOH* (* denotes adsorbed species) where the energy barrier DG is

less than 0.41 eV. These results support cPt–GNR as a candidate for ORR.

1. Introduction

Oxygen reduction reaction (ORR) is one of the important elec-trochemical reactions in energy-conversion devices, such as fuelcells and metal–air batteries,1 where the four-electron electro-chemical reaction is usually catalyzed by Pt to accelerate theprocess of conversion of chemical energy to electrical energy.However, the scarce resources and high cost of Pt give strongmomentum to reduce or replace Pt.2,3 Nanostructure engi-neering is one of the feasible methods to increase the surface/volume ratio and decrease the amount of Pt-based nano-materials.4–17 However, to our knowledge, a few studies arefocused on Pt subnanomaterials, such as Pt clusters (n ¼4–60),18–21 single (double) wall nanotubes,22 and single atom.23 Itis well known that the free monatomic wires cannot exist exceptAu.24 Thus no attention has been focused on monatomic wires.

Graphene has attracted widespread scientic and techno-logical interest due to its unique properties, and graphene/metal contacts are inevitable and omnipresent components ingraphene-based devices, such as memory devices, plasmonicdevices and energy-conversion devices.25–28 It has been shownthat defects in graphene play a key role in stabilizing metalclusters by providing strong anchoring sites.29 Thus, it isexpected that the edge of graphene can provide active sites forchemical bonding, due to the unsaturated bonds, which hasserved as a good support for metal monatomic wires.30–32 To ourknowledge, Pt nanowires can be exclusively formed on the edgeof highly oriented pyrolytic graphite by open circuit deposition33

Ministry of Education, and School of

University, Changchun 130022, China.

095371

08

or atomic layer deposition technique,34 due to the strongchemical binding between Pt atoms and the edge C atoms.Thus, if the substrate changes from graphite to graphene, thebinding face location of graphite for the above Pt nanowiresbecomes a line one for forming a Pt monatomic wire, if otherexperimental conditions remain. In fact, this has been carriedout on graphene when the element is Cr, where a monatomicchain of Cr is built on the edge of graphene experimentally.31

Since the binding energy between Cr and C (4.9 eV) is smallerthan that between Pt and C (6.4 eV), Pt atoms can be trapped bythe edge of graphene.31,35 Recently, the atomic structure of theboundary between a graphene zigzag edge and a Pt(111) stephas been unveiled by STM experiments and DFT calculation,which shows the strong bonding between the edge C and Ptatoms.36 Thus, the synthesis of Pt monatomic wire that is sup-ported on the zigzag edges of graphene nanoribbon (Pt–GNR) isfeasible.

Here, we consider Pt–GNR as ORR catalyst by using densityfunctional theory (DFT) calculations. It is found that Pt–GNR isinert as ORR catalyst due to strong adsorption ability. However,when Pt–GNR is covered by a chain consisting of OH and H2O,dened as cPt–GNR, it is activated. Based on the free energydiagrams, we demonstrate that the highest potential for ORRunderway is 0.82 eV, being comparable with 0.75 or 0.78 eV ofPt(111).1,37

2. Computational method

All calculations are performed within the DFT framework asimplemented in DMol3 code.38,39 The generalized gradientapproximation with the Perdew–Burke–Ernzerhof (PBE) func-tional is employed to describe exchange and correlation

This journal is © The Royal Society of Chemistry 2014

Paper RSC Advances

effects.40 The DFT semi-core pseudopots (DSPP) core treatmethod is implemented for relativistic effects, which replacescore electrons by a single effective potential and introducessome degree of relativistic correction into the core.41 The doublenumerical atomic orbital augmented by a polarization function(DNP) is chosen as the basis set.38 A smearing of 0.005 Ha (1 Ha¼ 27.21 eV) to the orbital occupation is applied to achieveaccurate electronic convergence. The spin-unrestricted methodis used for all optimization calculations. A. vacuum larger than20 A thick is added to avoid the articial interactions betweenPt–GNR and its images. In all of the structure optimizationcalculations, all atoms are fully relaxed.

Combining high resolution STM experiments and DFTcalculations, Merino et al. have unambiguously unveiled theatomic structure of the boundary between a graphene zigzagedge and a Pt(111) step, and the corrugation of Pt atoms at thegraphene–Pt junction has been identied.36 Thus, the corru-gated Pt–GNR is considered in our work. The structure of Pt–GNR that is proposed in ref. 30 is shown in Fig. 1(a). The averagebond lengths of Pt–C bonds (lPt–C) and Pt–Pt bonds (lPt–Pt) are1.92 and 2.57 A, respectively. The monatomic wire is highcorrugated with the corrugation angle of q ¼ 148�. However,another structure is more stable with the energy 0.77 eV per celllower, as shown in Fig. 1(b) and (d). The average lPt–C and lPt–Ptare 1.93 A and 2.51 A, respectively, and q is 158�. Thus, in ourstudy, the latter one is adopted. On the other hand, in order todemonstrate the coadsorption phenomenon, Pt(111) ismodeled by a three-layer (2O3 � 2O3)R30� periodic slab wherethe two bottom layers are xed (36 atoms per cell). It is worthnoting that the convergence testing is performed on four layersof atoms with the bottom two xed, and the same adsorptionstructures are given. Thus, we expect the three layer periodicmodel is enough.

The adsorption energy (Eads) of adsorbates on thesesubstrates are calculated by,

Eads(M) ¼ Esys � EM � Esubstrate (1)

Fig. 1 The schematic structures of Pt–GNR and cPt–GNR. Gray, blue,white and red colors denote C, Pt, H andO atoms. (a) is the side view ofPt–GNR structure from ref. 30, (b) and (c) are the side view of Pt–GNRand cPt–GNR structures in this work while (d) and (e) are the corre-sponding top view, respectively.

This journal is © The Royal Society of Chemistry 2014

where Esys, EM and Esubstrate are the total energies of theadsorption systems, an isolated adsorbate species and thesubstrate, respectively. M denotes O2, OOH, O, OH and H2O.Eads < 0 corresponds to an exothermic adsorption process.

Due to the structural deformation during the adsorptionprocess, besides adsorption energy Eads, the interaction energyEint should give accurate information of the bonding strengthbetween the adsorbates and substrate,

Eint(M) ¼ Esys � EM-def � Esubstrate-def (2)

where EM-def and Esubstrate-def are the energies of M and thesubstrate with the deformed geometry, respectively. Eint <0 denotes an exothermic adsorption process.

Free energies of the ORR intermediates in electrochemicalreaction pathways are calculated based on a computationalhydrogen electrode (CHE) model.1,20 The CHE model denesthat the chemical potential of proton–electron (H+ + e�) insolution is equal to the half of the chemical potential of a gas-phase H2. Free energy change (DG) is dened as,

DG ¼ DE + DZPE � TDS (3)

where DE is the total energy change directly obtained from DFTcalculations, DZPE is the change in zero-point energies, Tdenotes temperature, and DS is the change in entropy. Theeffect of a bias can be applied shiing DG by�eU, where e is theelementary positive change and U is the applied bias. Freeenergies of adsorbates are calculated by treating 3N degrees offreedom of adsorbates as vibrational frequencies in theharmonic oscillator approximation and xing the substrate.The zero-point energies and entropies of the ORR intermediatesare calculated from the vibrational frequencies according tostandard methods. DG < 0 corresponds to an exothermicadsorption process. The free energy G of O2 is derived asG(O2)¼4.92 + 2G(H2O) � 2G(H2) by utilizing OER equilibrium at thestandard conditions.42,43

3. Results and discussionThe ORR process on Pt–GNR

In order to characterize the nature of the interaction betweenthe Pt atoms and the edge C atoms of Pt–GNR, the partialdensity of states (PDOS) is shown in Fig. 2(a). It is obvious thatPt-5d is highly hybridized with C-2p at the range from �6.8 to�2.7 eV while a little contribution comes from s–d hybridiza-tion. From the deformation electron density shown in the insertof Fig. 2(a), we can see that electrons accumulate between the Ptatoms and the edge C atoms, indicating the formation ofcovalent Pt–C bonds. In addition, since the low-temperaturepolymer electrolyte fuel cells are operated at 253 to 373 K, rst-principle molecular dynamics (MD) simulation at a constanttemperature of T ¼ 400 K in the NVT ensemble (i.e., constantparticle number, volume and temperature conditions) has beencarried out for 2 ps with the time step of 1 fs.44 This selection ofthe temperature, being higher than the working temperature,could conrm the stability of the studied Pt–GNR system at

RSC Adv., 2014, 4, 28400–28408 | 28401

Fig. 2 (a) Partial density of states (PDOS) between the d band of Ptatoms and the sp band of the edge C atoms of Pt–GNR and cPt–GNR.Insert is the deformation electron density. Blue/red denotes electronaccumulation/depletion. (b) The four structures from MD simulationare shown.

RSC Advances Paper

working temperature. In Fig. 2(b), four structures from MDcalculation are present. As shown in the gure, it is found thatPt atoms are xed at the edge of GNR and the Pt clustering isabsent at 400 K.

Before investigating the ORR mechanism, we rst study theadsorption of O2 and ORR intermediates on Pt–GNR. Theadsorption energies Eads are listed in Table 1 and the corre-sponding structures are shown in Fig. 3(a–e). The adsorption andactivation of O2 are the rst step for ORR. Fig. 3(a) presents the

Table 1 The adsorption energy Eads on Pt–GNR, cPt–GNR and Pt(111), aand cPt–GNR. The values in parentheses denote the interaction energy

Pt–GNRa cPt–GNRa Pt(1

O2 �1.39 (�2.01) 0.09 (�0.48) �0.OOH �2.33 �0.99 �1.O �3.72 �3.40 �4.OH �3.14 �2.08 �2.H2O �0.77

a Eads (Eint) of O2 and ORR intermediates are listed. b me of O2 and ORR i

28402 | RSC Adv., 2014, 4, 28400–28408

optimized geometry of O2 adsorption on Pt–GNR. O2 is located atthe bridge site on Pt–GNR with the Eads(O2) ¼ �1.39 eV and thebond length of O–O (lO–O) is elongated to 1.36 A compared with1.23 A of the gas O2. The Eads(O2) on Pt(111) has been determinedby experiments and simulations [experimentally Eads(O2) ¼0.3–0.5 eV on Pt(111) with low-coverage45,46 and theoreticallyEads(O2) ¼ 0.41–0.72 eV when O2 is adsorbed on bridge site ofPt(111)47–49]. Eads(O2) on Pt–GNR is much stronger than the aboveresults on Pt(111) due to the lower coordination number of Pt in amonatomic wire.50 In order to gain further insight into the originof the interaction between O2 and Pt–GNR, the correspondingelectronic structures are shown in Fig. 4 where the spin-polarizedPDOS projected onto the O–O bond has considerable change.Firstly, spin polarization of the adsorbed O2 orbitals is absent,denoting strong interaction between O2 and Pt–GNR. Further-more, 5s, 1p and 2p* orbitals of O2 are broadening, whichdominate the adsorption of O2 while 2p* orbitals are partiallylled. This lling is responsible for the catalytic activation of theadsorbed O2 and the stretching lO–O.51–53 The charge transfer fromPt–GNR to O2 is 0.29e according to Hirshfeld population chargedistribution listed in Table 1, which is agreement with theincrease of lO–O.

Fig. 3(b) depicts the favorable adsorption structure of OOHon Pt–GNR. OOH is located at the bridge site and binds stronglyto Pt–GNR with Eads(OOH) ¼ �2.33 eV. Therein, the lO–O of theadsorbed OOH increases to 1.54 A compared with the unad-sorbed one (1.34 A). The weakening of the O–O bond shouldpromote the dissociation of OOH into O and OH fragments. Inaccordance with the previous work,49 Eads(OOH) is �1.06 eV onPt(111), being weaker than that of Pt–GNR. For O adsorption onPt–GNR, as shown in Fig. 3(c), O prefers to locate at the bridgesite and Eads(O) is �3.72 eV, which is weaker than �4.21 eV ofPt(111).54 In Fig. 3(d), the OH adsorption is more stable at thebridge site with Eads(OH) ¼ �3.14 eV, which is stronger thanthat of Pt(111) (�2.24, �2.31 eV)49,55. H2O is the nal product ofthe ORR. As shown in Fig. 3(e), H2O is located at atop site withEads(H2O) ¼ �0.77 eV. This value is stronger than the solvationstabilization energy (about �0.40 eV) of bulk H2O.56 Thus, H2Ocould not be readily released into the bulk solution, hamperingthe next ORR cycle.

It has been demonstrated that the best Eads(O) and Eads(OH)of the ORR catalysts should be 0.0–0.4 eV and 0.0–0.2 eV weakerthan that of Pt(111),1,37,57,58 respectively, where there exist ascaling relationship between Eads(O) and Eads(OH).1,59 However,this scaling relationship is invalid for Pt–GNR where there are a

nd the electron number me of O2 and ORR intermediates on Pt–GNREint for O2 adsorption

11)a Pt–GNRb cPt–GNRb

41, �0.65, �0.72 (ref. 47–49) 0.29 0.1806 (ref. 49) 0.20 0.0821 (ref. 54) 0.31 0.2424, �2.31 (ref. 49 and 55) 0.25 0.12

ntermediates are shown.

This journal is © The Royal Society of Chemistry 2014

Fig. 3 The favorable adsorption structures of O2 and ORR intermediates. (a–e) are on Pt–GNR while (f–j) are on cPt–GNR. The O2 adsorptionstructures on covered Pt(111) is shown in (k). OH–H2O-I/II means OH–H2O line covered and OH–H2O-III is OH–H2O network covered. Theupper are the initial structures and the lower are the optimization structures.

Paper RSC Advances

weaker Eads(O) but a stronger Eads(OH), compared with Pt(111).Thus, the ORR reaction under a potential U should be studied indetails.

A standard ORR mechanism involves four proton and elec-tron transfers at the cathode, which mainly include threemechanisms: O2 dissociation, OOH dissociation and H2O2

dissociation.60 To study the activity of Pt–GNR, the OOH asso-ciative mechanism of ORR is taken as an example and consid-ered by the following steps (* denotes adsorbed species andg l�1 shows gas/liquid state),

O2(g) + * / O2*, (R1)

O2* + (H+ + e�) / OOH*, (R2)

OOH* + (H+ + e�) / 2OH*, (R3)

This journal is © The Royal Society of Chemistry 2014

2OH* + (H+ + e�) / H2O* + OH*, (R4)

H2O* + OH* + (H+ + e�) / 2H2O*, (R5)

2H2O* / 2H2O(l) + *. (R6)

According to this mechanism, (R1) is a process of themolecular adsorption and activation of O2. (R2) shows thereduction of O2* to OOH*, which occurs through coupledproton–electron transfer (CPET), as suggested by Damjanovicand Brusic.61 (R3) should be the reduction of OOH* to O* andH2O* through a CPET. However, since O* and H2O* are spon-taneously transferred to 2OH* aer the relaxation, OOH* isreduced to 2OH*. In order to further conrm the stability of2OH*, we perform geometry optimization with ve differentinitial congurations of O*/H2O* shown in Fig. 5. For the

RSC Adv., 2014, 4, 28400–28408 | 28403

Fig. 4 Partial density of states (PDOS) for O2 adsorbed on Pt–GNR andcPt–GNR. Fig. 6 Potential-dependent surface diagram for Pt–GNR (a) and cPt–

GNR (b). Zero on the free energy axis corresponds to the Pt–GNR orcPt–GNR and one O2(g) molecule at T ¼ 298 K, denoted as the greenline. The free energy of Pt–GNR or cPt–GNR and two H2O(l) mole-cules is shown in the orange line.

RSC Advances Paper

congurations I and II, the O* and H2O* locate at the adjacentatop sites with different H2O* orientations. Aer relaxation, O*and H2O* are transferred to 2OH* with the same structure. Onthe other hand, another three congurations are consideredwhere the location of H2O* is kept and O* is moved away inorder to increase the distance between O* and H2O*. For thecongurations III and IV, O* absorption is located on bridge sitewhile for the conguration V, O* is at atop site. The above threecongurations converge to the same coadsorption structure ofO* and H2O* aer relaxation where the total energy of thesestates are 0.45 eV higher than the case of 2OH* pair. Thus,compared with O* and H2O* on Pt–GNR, 2OH* is more stableand we take it as the product of (R3). (R4) and (R5) indicate therst and the second OH* reduction to H2O* through contin-uous CPET. The nal step (R6) is the desorption of H2O* andrecovery of Pt–GNR.

Firstly, in order to analyze the stability of ORR intermediatesthermodynamically, the ab initio phase diagram is established onPt–GNR, as shown in Fig. 6(a). Here, the stability of the ORRintermediates is characterized as a function of the electrode biasU. Obviously, as the potential U changes from 0 to 1.23 V, thecoadsorption system of H2O* and OH* is stable and identied asthe thermodynamic ground state, where its free energy G is evenlower than the system of H2O(l) and Pt–GNR. This suggests thatPt–GNR should be covered by H2O* and OH* and lose its activity.

To further understand the ORR activity of Pt–GNR, the Gunder two limiting potentials U of 0 V and 1.23 V are shown inFig. 7(a) and the free energy change DG of the endothermic

Fig. 5 The different O*/H2O* configurations adsorption on Pt–GNR. Tstructures. The values show the stability of different systems. More nega

28404 | RSC Adv., 2014, 4, 28400–28408

steps as a function of U are shown in Fig. 7(b). At U ¼ 0 V, (R5)and (R6) are endothermic with DGR5

¼ 0.12 eV and DGR6¼ 0.53

eV while other steps are strong exothermic. At the equilibriumbias of U¼ 1.23 V, corresponding to the situation where the fuelcell has the maximum potential allowed by thermodynamics,(R2) changes from exothermic to endothermic process withDGR2

¼ 0.69 eV. (R5) becomes more endothermic and DGR5

reaches 1.35 eV. DGR6is 0.53 eV, which is independent on U. As

shown in Fig. 7(b), when U is less than 0.42 V, the rate-deter-mined step (RDS) is located at (R6). On the other hand, when Ulocates at the range from 0.42 to 1.23 V, RDS changes to (R5),which is consistent with that of the graphene-supported Pt13cluster.20 Thus, the endothermic steps are always present for Pt–GNR. Due to the strong endothermic process of (R5) and thethermodynamic stability of H2O* and OH* coadsorption systemshown in Fig. 6(a), it is conrmed that the Pt–GNR would becovered by a chain combined with H2O* and OH* (cPt–GNR)whose structure is shown in Fig. 1(c) and (e). Thus, in thefollowing text, we test the ORR activity on cPt–GNR.

The ORR process on cPt–GNR

Compared with Pt–GNR, the average lPt–Pt, lPt–C and q of cPt–GNR are nearly unchanged with the values of 2.51 A, 1.97 A and159�, respectively. The bonding energy between Pt monatomic

he upper are the initial structures and the lower are the optimizationtive means more stable.

This journal is © The Royal Society of Chemistry 2014

Fig. 7 The ORR processes on Pt–GNR [(a) and (b)] and cPt–GNR [(c) and (d)]. Zero on the free energy axis corresponds to Pt–GNR or cPt–GNRand two H2O molecules.

Paper RSC Advances

wire and GNR in cPt–GNR is enhanced compared with Pt–GNR(4.08 vs. 3.52 eV per atom). From the PDOS shown in Fig. 2(a),Pt-5d/C-2p hybridization is strong at the range from �6.8 eV toFermi level. A little contribution comes from s–d hybridization.Compared with Pt–GNR, the p–d hybridization of cPt–GNRlocated around �2.7 eV is much stronger, which is consistentwith the corresponding stronger bonding energy. In addition,from the insert in Fig. 2(a), the covalent nature of Pt–C bonds isconrmed by the deformation electron density.

Eads(O2) is 0.09 eV as O2 adsorbs at bridge site of cPt–GNR orthe adsorption is absent. The corresponding adsorption structureis shown in Fig. 3(f). However, the lO–O of the adsorbed O2 iselongated to 1.35 A and O2 indeed gets activated. From the PDOSof the adsorbed O2 shown in Fig. 4, it is obvious that the 5s, 1pand 2p* orbitals of O2 are broadened, which dominates theadsorption of O2. The spin polarization disappears and 2p*orbital is partial lled. In light of Hirshfeld charge analysis listedin Table 1, 0.18e charge transfers from cPt–GNR to the 2p*orbital, further supporting the adsorption and activation of O2.Therein, the electronic structure of the adsorbed O2 is incontradictionwith the positive Eads. Since Eads can be divided intothe deformation energy and the interaction energy Eint,62 thepositive Eads could be originated from the deformation energy[Eint(O2) ¼ �0.48 eV on cPt–GNR, which is in accord with theelectronic structure shown in Fig. 4]. Compared with Pt–GNR[Eads(O2) ¼ �1.39 eV and Eint(O2) ¼ �2.01 eV], the adsorptionability of O2 on cPt–GNR is substantially weakened due to thepreferential adsorption of OH* andH2O*. It should be noted thatif Pt atoms in Pt monatomic wires are covered by the OH* andH2O*, which could exist in the ORR reaction, the O2 adsorption

This journal is © The Royal Society of Chemistry 2014

and activation remain due to the low-coordination number andmore open structure, whichmakes the ORR in proceed. However,this cannot be realized at the case of Pt(111). As shown inFig. 3(k), O2 cannot be bonded with Pt atoms covered by the OH*

and H2O* on Pt(111). For the end-on conguration, O2 locates atthe atop site of the adjacent Pt atom to form hydrogen bondingwith OH* (OH–H2O-I) while for side-on conguration, O2 is onthe bridge site to formhydrogen bonding with anotherOH* (OH–

H2O-II). Further increasing the coverage of the OH* andH2O*, O2

still locates at the uncovered Pt atom (OH–H2O-III). Thus, thecovered Pt atom on Pt(111) is inert due to the presence of OH*

and H2O*, which prevent the adsorption of the reactant O2 andthe ORR reaction is terminated.

The stable adsorption structures of ORR intermediates areshown in Fig. 3(g–j) and the Eads are summarized in Table 1.Eads(OOH) is �0.99 eV at atop site and lO–O is 1.40 A. Comparedwith 1.34 A of the gas one, the weakening of the O–O bondappears and should promote the dissociation of OOH into Oand OH fragments. Eads(O) is �3.40 eV at bridge site. For OHadsorption, atop site is slightly preferred compared with bridgesite [Eads(OH) is�2.08 eV at atop site and Eads(OH) is�1.92 eV atbridge site]. Note that H2O cannot locate at Pt monatomic wireand promote forming two hydrogen bonds with the chain ofH2O* and OH*, which benets for recovery of the catalyst forthe next ORR cycle. Obviously, the existence of chain of H2O*and OH* on Pt–GNR weakens the adsorption ability of O2 andORR intermediates where the electron number transferred fromcPt–GNR is smaller than that from Pt–GNR as listed in Table 1.

To provide a physical picture for the potential catalyticactivity of cPt–GNR, the deformation density is shown in Fig. 8.

RSC Adv., 2014, 4, 28400–28408 | 28405

Fig. 8 The deformation electron density for cPt–GNR and the related adsorption systems are plotted where the electron accumulation is shown.The value of isovalue is 0.1 e A�3.

RSC Advances Paper

Fig. 8(a) shows the electron accumulation of original cPt–GNR islocated on O atoms of the OH*/H2O* chain. In Fig. 8(b)–(e),even the adsorbates are different aer the adsorption, electronis still accumulated on O atoms of both adsorbates and OH*/H2O* chain. Thus, the weakening of absorption ability of cPt–GNR should originate from the electrostatic repulsion betweenthe OH*/H2O* chain and the adsorbates (O2 and ORR inter-mediates). This conclusion has been found in other systems,such as the adsorption of O or OH on Ru(0001), where theelectrostatic repulsion occurs due to the presence of Se.63

Therefore, we expect that the adsorbed species on cPt–GNRcould be easily removed and cPt–GNR possesses ORR activity.

Now, the energetics of the ORR on cPt–GNR is analyzed bythe simple potential-dependent model,

O2(g) + * / O2*, (R1)

O2* + (H+ + e�) / OOH*, (R2)

OOH* + (H+ + e�) / O* + H2O(l), (R7)

O* + (H+ + e�) / OH*, (R8)

OH* + (H+ + e�) / H2O(l). (R9)

Since H2O cannot be adsorbed on cPt–GNR, reduction ofOOH* to O* and H2O(l) through continuous CPET, dened as(R7). (R8) and (R9) indicate the reduction processes from O toOH* and from OH* to H2O(l).

In Fig. 6(b), the phase diagram of ORR intermediates on cPt–GNR is shown. Differing from Pt–GNR, when 0# U# 1.11 V, thesystem of H2O(l) and cPt–GNR is thermodynamic ground state.When U > 1.11 V, the adsorption system of O* on cPt–GNR is thestable phase. The G function at the limiting condition (U ¼ 0 Vand U ¼ 1.23 V) are shown in Fig. 7(c). At U ¼ 0 V, all steps areexothermic. On the other hand, when U is 1.23 V, (R2) and (R8)become endothermic with DGR2

¼ 0.41 eV and DGR8¼ 0.36 eV,

respectively. From the DG shown in Fig. 7(d), when U is less than0.82 V, all ORR steps are exothermic, which means that thereactions are thermodynamically favorable. (R2) and (R8) becomeendothermic at the values of U ¼ 0.82 and 0.87 V, respectively.Furthermore, DGR2

is always larger than DGR8. Thus, the RDS is

located at (R2) with the maximum of DGR2¼ 0.41 eV, which is

much smaller than the case of Pt–GNR (1.35 eV). The maximal Uvalue corresponding to the exothermic steps of ORR is 0.82 V,which is comparable with Pt(111) (0.75 or 0.78 V).1,37

28406 | RSC Adv., 2014, 4, 28400–28408

Since the presence of the overpotential would reduce thethermal efficiency well below the thermodynamic limit, thethermal efficiency 3 of this system is considered here. It is wellknown 3 is calculated by,

3 ¼ Dg/DhHHV (4)

where Dg ¼�nFU is the useful work, n is the number of transferelectron (n ¼ 2 for 1 mol hydrogen), F is the Faraday constant(F ¼ 96485C mol�1) and U denotes the maximal U for theexothermic ORR (in our case, U ¼ 0.82 V), DhHHV is the idealenergy released by 1 mol hydrogen combustion (DhHHV ¼ �286kJ mol�1). In light of eqn (4), 3 value of cPt–GNR reaches 55%[the theoretical limit 3 value is 83% at 1.23 V (ref. 64)]. When wecalculate the case of Pt(111),1,37 this value is 51–53%. Tosimplify, the effect of the fuel utilization is neglected.

From the aspect of the free energy barriers, the activity ofcPt–GNR could be compared with that of Pt–GNR. Usually, theactivity of the ORR catalysts is generally characterized underU ¼ 0.9 V.7,11,13,14 Thus, the DG values of cPt–GNR under thispotential are given in Fig. 7(c), which are 0.08 eV and 0.03 eV for(R2) and (R8), respectively. Since the activation barrier for therate-limiting proton-transfer step is equal to the maximal DGvalue1, the thermodynamic barrier associated with over ORRprocess is 0.08 eV. Compared with the barrier of Pt–GNR(1.02 eV) in Fig. 7(a), the cPt–GNR is more active than Pt–GNR(this value for Pt(111) is 0.07 eV,65 being comparable with cPt–GNR). Thus, it is expected that cPt–GNR would be viewed as agood candidate of ORR electrocatalyst.

4. Conclusions

In summary, we have reported cPt–GNR as a suitable ORRelectrocatalyst. It is found that the rate-determined step islocated at the reduction of O2* to OOH* with maximal barrier of0.41 eV as U is 1.23 V. Thus, cPt–GNR could be a candidate ofORR electrocatalyst.

Acknowledgements

We acknowledge the supports fromNational Key Basic Researchand Development Program (Grant no. 2010CB631001), theNational Natural Science Foundation of China (no. 51201069),the Keygrant Project of Chinese Ministry of Education (no.313026), the Program for New Century Excellent Talents in

This journal is © The Royal Society of Chemistry 2014

Paper RSC Advances

University (no. NCET-10-0437), and the Research Fund for theDoctoral Program of Higher Education of China (no.20120061120042).

References

1 J. K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist,J. R. Kitchin, T. Bligaard and H. Jonsson, J. Phys. Chem. B,2004, 108, 17886–17892.

2 A. Morozan, B. Jousselme and S. Palacin, Energy Environ. Sci.,2011, 4, 1238–1254.

3 C. Sealy, Mater. Today, 2008, 11, 65–68.4 C. Wang, H. Daimon, T. Onodera, T. Koda and S. Sun, Angew.Chem., Int. Ed., 2008, 47, 3588–3591.

5 M. K. Carpenter, T. E. Moylan, R. S. Kukreja, M. H. Atwan andM. M. Tessema, J. Am. Chem. Soc., 2012, 134, 8535–8542.

6 C. M. Sanchez-Sanchez, J. Solla-Gullon, F. J. Vidal-Iglesias,A. Aldaz, V. Montiel and E. Herrero, J. Am. Chem. Soc.,2010, 132, 5622–5624.

7 J. Wu, L. Qi, H. You, A. Gross, J. Li and H. Yang, J. Am. Chem.Soc., 2012, 134, 11880–11883.

8 C. Cui, L. Gan, H. H. Li, S. H. Yu, M. Heggen and P. Strasser,Nano Lett., 2012, 12, 5885–5889.

9 S. M. Alia, K. O. Jensen, B. S. Pivovar and Y. Yan, ACS Catal.,2012, 2, 858–863.

10 Z. Chen, M. Waje, W. Li and Y. Yan, Angew. Chem., Int. Ed.,2007, 46, 4060–4063.

11 L. Ruan, E. Zhu, Y. Chen, Z. Lin, X. Huang, X. Duan andY. Huang, Angew. Chem., Int. Ed., 2013, 52, 1–6.

12 C. Koenigsmann, W. P. Zhou, R. R. Adzic, E. Sutter andS. S. Wong, Nano Lett., 2010, 10, 2806–2811.

13 H. I. Karan, K. Sasaki, K. Kuttiyiel, C. A. Farberow,M. Mavrikakis and R. R. Adzic, ACS Catal., 2012, 2, 817–824.

14 C. Wang, N. M. Markovic and V. R. Stamenkovic, ACS Catal.,2012, 2, 891–898.

15 J. Kibsgaard, Y. Gorlin, Z. Chen and T. F. Jaramillo, J. Am.Chem. Soc., 2012, 134, 7758–7765.

16 M. Shao, A. Peles and K. Shoemaker, Nano Lett., 2011, 11,3714–3719.

17 C. Chen, Y. Kang, Z. Huo, Z. Zhu, W. Huang, H. L. Xin,J. D. Snyder, D. Li, J. A. Herron, M. Mavrikakis, M. Chi,K. L. More, Y. Li, N. M. Markovic, G. A. Somorjai, P. Yangand V. R. Stamenkovic, Science, 2014, 343, 1339–1343.

18 T. Imaoka, H. Kitazawa, W. J. Chun, S. Omura, K. Albrechtand K. Yamamoto, J. Am. Chem. Soc., 2013, 135, 13089–13095.

19 K. Yamamoto, T. Imaoka, W. J. Chun, O. Enoki, H. Katoh,M. Takenaga and A. Sonoi, Nat. Chem., 2009, 1, 397–402.

20 D. H. Lim and J. Wilcox, J. Phys. Chem. C, 2012, 116, 3653–3660.

21 A. Roudgar, M. Eikerling and R. van Santen, Phys. Chem.Chem. Phys., 2010, 12, 614–620.

22 I. Matanovic, P. R. C. Kent, F. H. Garzon and N. J. Henson, J.Phys. Chem. C, 2012, 116, 16499–16510.

23 M. A. Gabriel, L. Genovese, G. Krosnicki, O. Lemaire,T. Deutsch and A. A. Franco, Phys. Chem. Chem. Phys.,2010, 12, 9406–9412.

This journal is © The Royal Society of Chemistry 2014

24 D. Liu, W. T. Zheng and Q. Jiang, Appl. Surf. Sci., 2005, 252,1780–1784.

25 J. T. Smith, A. D. Franklin, D. B. Farmer andC. D. Dimitrakopoulos, ACS Nano, 2013, 7, 3661–3667.

26 X. Wang, W. Xie, J. Du, C. Wang, N. Zhao and J. B. Xu, Adv.Mater., 2012, 24, 2614–2619.

27 A. Politano and G. Chiarello, Nanoscale, 2013, 5, 8215–8220.28 S. Guo and S. Sun, J. Am. Chem. Soc., 2012, 134, 2492–

2495.29 I. Fampiou and A. Ramasubramaniam, J. Phys. Chem. C,

2012, 116, 6543–6555.30 G. J. Soldano, P. Quaino, E. Santos and W. Schmickler, J.

Phys. Chem. C, 2013, 117, 19239–19244.31 H. Wang, Q. Feng, Y. Cheng, Y. Yao, Q. Wang, K. Li,

U. Schwingenschlogl, X. X. Zhang and W. Yang, J. Phys.Chem. C, 2013, 117, 4632–4638.

32 Y. Wang, C. Cao and H. P. Cheng, Phys. Rev. B: Condens.Matter Mater. Phys., 2010, 82, 205429.

33 P. M. Quaino, M. R. Gennero de Chialvo, M. E. Vela andR. C. Salvarezza, J. Argent. Chem. Soc., 2005, 93, 215–224.

34 H. B. R. Lee, S. H. Baeck, T. F. Jaramillo and S. F. Bent, NanoLett., 2013, 13, 457–463.

35 H. Wang, K. Li, Y. Cheng, Q. Wang, Y. Yao,U. Schwingenschlogl, X. Zhang and W. Yang, Nanoscale,2012, 4, 2920–2925.

36 P. Merino, L. Rodrigo, A. L. Pinardi, J. Mendez, M. F. Lopez,P. Pou, R. Perez and J. A. Martın-Gago, ACS Nano, 2014, 8,3590–3596.

37 J. Greeley, I. E. L. Stephens, A. S. Bondarenko,T. P. Johansson, H. A. Hansen, T. F. Jaramillo,J. Rossmeisl, I. Chorkendorff and J. K. Nørskov, Nat.Chem., 2009, 1, 552–556.

38 B. Delley, J. Chem. Phys., 1990, 92, 508–517.39 B. Delley, J. Chem. Phys., 2000, 113, 7756–7764.40 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,

1996, 77, 3865–3868.41 B. Delley, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 66,

155125.42 J. Sun, Y. H. Fang and Z. P. Liu, Phys. Chem. Chem. Phys.,

2014, 16, 13733–13740.43 G. F. Wei and Z. P. Liu, Phys. Chem. Chem. Phys., 2013, 15,

18555–18561.44 A. Rabis, P. Rodriguez and T. J. Schmidt, ACS Catal., 2012, 2,

864–890.45 J. L. Gland, B. A. Sexton and G. B. Fisher, Surf. Sci., 1980, 95,

587–602.46 C. T. Campbell, G. Ertl, H. Kuipers and J. Segner, Surf. Sci.,

1981, 107, 220–236.47 L. Qi, J. Yu and J. Li, J. Chem. Phys., 2006, 125, 054701.48 A. Eichler and J. Hafner, Phys. Rev. Lett., 1997, 79, 4481–4484.49 Y. Sha, T. H. Yu, Y. Liu, B. V. Merinov and W. A. Goddard,

J. Phys. Chem. Lett., 2010, 1, 856–861.50 B. Hammer and J. K. Nørskov, Adv. Catal., 2000, 45, 71–129.51 B. Yoon, H. Hakkinen and U. Landman, J. Phys. Chem. A,

2003, 107, 4066–4071.52 X. Ding, Z. Li, J. Yang, J. G. Hou and Q. Zhu, J. Chem. Phys.,

2004, 120, 9594–9600.

RSC Adv., 2014, 4, 28400–28408 | 28407

RSC Advances Paper

53 R. Coquet, K. L. Howard and D. J. Willock, Chem. Soc. Rev.,2008, 37, 2046–2076.

54 X. Lin, N. J. Ramer, A. M. Rappe, K. C. Hass, W. F. Schneiderand B. L. Trout, J. Phys. Chem. B, 2001, 105, 7739–7747.

55 M. T. M. Koper, T. E. Shubina and R. A. van Santen, J. Phys.Chem. B, 2002, 106, 686–692.

56 Y. Sha, T. H. Yu, B. V. Merinov, P. Shirvanian andW. A. Goddard, J. Phys. Chem. C, 2012, 116, 21334–21342.

57 I. E. L. Stephens, A. S. Bondarenko, F. J. Perez-Alonso,F. Calle-Vallejo, L. Bech, T. P. Johansson, A. K. Jepsen,R. Frydendal, B. P. Knudsen, J. Rossmeisl andI. Chorkendorff, J. Am. Chem. Soc., 2011, 133, 5485–5491.

58 V. Viswanathan, H. A. Hansen, J. Rossmeisl andJ. K. Nørskov, ACS Catal., 2012, 2, 1654–1660.

28408 | RSC Adv., 2014, 4, 28400–28408

59 G. S. Karlberg, J. Rossmeisl and J. K. Nørskov, Phys. Chem.Chem. Phys., 2007, 9, 5158–5161.

60 J. A. Keith, G. Jerkiewicz and T. Jacob, ChemPhysChem, 2010,11, 2779–2794.

61 A. Damjanovic and V. Brusic, Electrochim. Acta, 1967, 12,615–628.

62 W. Liu, J. S. Lian and Q. Jiang, J. Phys. Chem. C, 2007, 111,18189–18194.

63 S. Stolbov, J. Phys. Chem. C, 2012, 116, 7173–7179.64 V. R. Stamenkovic, B. Fowler, B. S. Mun, G. Wang, P. N. Ross,

C. A. Lucas and N. M. Markovic, Science, 2007, 315, 493–497.65 H. S. Casalongue, S. Kaya, V. Viswanathan, D. J. Miller,

D. Friebel, H. A. Hansen, J. K. Nørskov, A. Nilsson andH. Ogasawara, Nat. Commun., 2013, 4, 2817.

This journal is © The Royal Society of Chemistry 2014