Accepted Manuscript Two-dimensional siligraphenes as cathode catalysts for nonaqueous lithium-oxygen batteries Huilong Dong, Yujin Ji, Tingjun Hou, Youyong Li PII: S0008-6223(17)31075-8 DOI: 10.1016/j.carbon.2017.10.077 Reference: CARBON 12504 To appear in: Carbon Received Date: 9 August 2017 Revised Date: 3 October 2017 Accepted Date: 23 October 2017 Please cite this article as: H. Dong, Y. Ji, T. Hou, Y. Li, Two-dimensional siligraphenes as cathode catalysts for nonaqueous lithium-oxygen batteries, Carbon (2017), doi: 10.1016/j.carbon.2017.10.077. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT Two-dimensional siligraphenes as cathode catalysts for nonaqueous lithium-oxygen batteries Huilong Dong,* Yujin Ji, Tingjun Hou and Youyong Li* RI PT Institute of Functional Nano & Soft Materials (FUNSOM), Soochow University, Suzhou, Jiangsu 215123, China. SC Abstract The nonaqueous lithium-oxygen (Li-O2) battery is promising as a superior energy M AN U storage medium, whose practical applications are limited by the catalytic performance of cathode catalysts. Using density functional theory calculations, we systematically investigate the catalytic mechanism and catalytic performance of siligraphenes as cathode catalysts in nonaqueous Li-O2 batteries. For the siligraphenes studied here, Li2O is the final discharge product. Among them, the single-layered SiC (SL-SiC) TE D exhibits potential as a suitable cathode catalyst due to the considerably low overpotential values during discharge and charge process. That is, 0.73 V for oxygen reduction reaction (ORR) and 1.87 V for oxygen evolution reaction (OER). The quantitative analysis indicates that the ORR overpotential on siligraphenes is linearly EP correlated with the ∆Eads of the last added Li in Li4O2* (* denotes the adsorbed surface), while the OER overpotential is linearly correlated with the Eads of LiO2 on AC C surface. Our investigations elucidate quantitative correlations between the catalytic performance of siligraphenes and the adsorption performance of Li-contained intermediates on them, which provides promising approach to develop metal-free cathode catalysts with high catalytic activity. * Corresponding authors. E-mail: firstname.lastname@example.org, Tel: (86)-512-65882037 (H. Dong); E-mail: email@example.com, Tel: (86)-512-65882037 (Y. Li). ACCEPTED MANUSCRIPT 1. Introduction Since its first proposal by Abraham and Jiang, the lithium-oxygen (Li-O2) battery has attracted tremendous scientific interest. Due to the extremely high theoretical specific energy density [5–10 times higher than the traditional lithium ion RI PT batteries (LIBs)],[2, 3] nonaquous Li-O2 battery is viewed as competitive energy storage medium for electric vehicles (EVs) or hybrid electric vehicles (HEVs).[4, 5] In a Li-O2 battery, the pure lithium metal is employed as anode material, while atmospheric oxygen is used as a reactant in cathode, which is stored externally and SC can be readily obtained from the surrounding air.[6, 7] Though it seems promising, the practical application of Li-O2 battery still faces lots of challenge, such as poor stability, short cycle life, low discharge capacity and high M AN U electrolyte overpotential.[8-10] Most of the problems can be attributed to the slow kinetics of the oxygen reduction reaction (ORR) during discharging process and the oxygen evolution reaction (OER) during charging process.[6, 11] To effectively facilitate the ORR and OER, the development of suitable cathode catalysts is of the great TE D importance. Among the various cathode catalysts, carbon-based nanomaterials take advantages in low cost, large specific surface area, high conductivity, and good EP stability,[13, 14] making them attractive in applications of Li-O2 battery. As the most representative two-dimensional carbon nanomaterial, graphene was deeply researched as the cathode catalyst for Li-O2 battery.[15-17] The experimental work by Qiao and AC C coworkers had proven that the flat surface of two-dimensional carbon nanomaterial could exhibit high catalytic activity. Sun and coworkers first reported the application of graphene nanosheets (GNSs) as cathode active materials in a nonaqueous lithium-oxygen battery, and found that the GNSs electrode exhibit extremely high discharge capacity in comparison to carbon powders. Furthermore, Jiang et al. theoretically demonstrated that the existence of some specific point defects (like single vacancy, DV5555-6-7777 defect) on graphene surface is the origin of graphene’s excellent performance. However, the surface defects are difficult to control, and are easy to migrate or be healed. Doping is the effective way to ACCEPTED MANUSCRIPT overcome the drawbacks of point defects by introducing more stable and more active sites into graphene. The hetero-atom doped graphenes (including N-doped,[21-24] B-doped, S-doped, or N, B-codoped) have been widely investigated as the cathode catalysts in Li-O2 battery, both experimentally and theoretically. RI PT Particularly, Ren et al. studied a series of hetero-atom doped graphene materials (B, N, Al, Si and P) as OER catalysts to correlate catalytic activity with adsorption structure and charge transfer properties. From their results it is found that Si-doped graphene provides relatively favorable charging voltage as well as low barrier for SC oxygen evolution. Actually, silicon always plays important role in energy storage materials. For nanoparticles for M AN U instance, Son et al. reported the SiC-free graphene growth over silicon high-performance lithium-ion battery application, which demonstrates the synergistic effect of silicon and carbon in energy storage. Lots of previous studies validated the great effect of Si-doped graphene or siligraphenes in oxygen dissociation or ORR catalysis[31, 32]. Comparing with the Si-doped TE D graphene, the siligraphene that formed by incorporating silicon into graphene honey-comb structure benefits from more stable composite and richer reaction sites for adsorption of oxygen molecules. Here we made comparative investigations on EP three kinds of siligraphenes (SL-SiC, g-SiC2,[31, 33] and g-SiC3) that show high affinity to oxygen adsorption. By performing density functional theory (DFT) calculations, we uncover the potential of siligraphenes (especially SL-SiC) as AC C candidates of cathode catalysts for Li-O2 battery. The high Si : C ratio provides plenty active sites for adsorption and dissociation of O2 molecules, and the strong binding between the dissociated oxygen and Li promotes the four-electron pathway to form Li2O as the final product. Our calculations also provide new insight into the relationship between adsorption performance of LixO2 intermediates and the ORR/OER overpotential, which is helpful to the material design on cathode of Li-O2 battery. 2. Computational details and modeling ACCEPTED MANUSCRIPT All the geometry optimization and energy calculations are carried out by DMol3[35-37] module available in Materials Studio 7.0. The Perdew-Burke-Ernzerhof (GGA-PBE) functional is employed to describe the electronic exchange and correlation effects. The double numerical atomic orbital RI PT including polarized p-function (DNP) is used as the basis set, and orbital cutoff is set as 4.6 Å. To correctly deal with the long-range weak interaction like van der Waals forces, Grimme scheme is adopted for dispersion correction. The geometry optimization converges with energy, gradient and displacement reach the threshold SC values of 10−5 Ha, 0.001 Ha/Å, and 0.005 Å, respectively. Electron thermal smearing effect is imposed by employing a smearing value (0.005 Ha) for all the calculations, M AN U which could effectively enhance the SCF convergence efficiency. The crystal structures of SL-SiC, g-SiC2 and g-SiC3 share graphene-like hexagonal lattice, and their unit cells are derived from the recently reported work.[31, 32, 34] We expanded SL-SiC into 5×5×1 supercell (a = 15.42 Å), and expanded g-SiC2 and g-SiC3 into 3×3×1 supercell (a = 15.00 Å and 16.84 Å for g-SiC2 and TE D g-SiC3). The Brillouin zone is sampled by a 3×3×1 k-point Monkhorst-Pack grid, which is sufficient to reach convergence as tested [see Table S1 in Supplementary material]. To determine the activation barriers (Eb) during the dissociation of EP adsorbed O2 molecule on siligraphene, transition state searches are conducted by performing the complete linear synchronous transit/quadratic synchronous transit (LST/QST) method. All the obtained transition state (TS) structures show only AC C one imaginary frequency, and are confirmed to directly connect corresponding reactants and products by nudged-elastic band (NEB) algorithm. The adsorption energy (Eads) for Li atom or O2 molecule on siligraphene is defined as () = + ∗ − ∗ (1) ( ) = + ∗ − ∗ (2) where the ∗ represents the energy of the siligraphene substrate (* denotes the adsorbent), ∗ or ∗ represents the energy of optimized Li or O2 adsorbed ACCEPTED MANUSCRIPT siligraphene, and is the energy of O2 molecule (triplet as the ground state). It should be pointed out that the used here is the energy per atom in bulk Li metal (bcc structure), so Eads of Li indicates the energy difference between adsorbed Li atom against Li atom in its bulk structure. For LixO2 adsorbed surface with several Li atoms, as ∆ = ∗ + − ∗ RI PT we defined the consecutive adsorption energy (∆Eads) of a newly added Li in LixO2* (8) SC where the ∗ represents the energy of LixO2* and ∗ represents the energy of Lix-1O2*. The ∆Eads of Li indicates the energy released when a new Li is M AN U added into Lix-1O2*, which can also be viewed as the insertion energy of Li atom. The equilibrium potential can be obtained by the Nernst equation = − ∆ (3) In this equation n is the number of transferred electron and the standard formation energy (∆ ) could be derived from where () , () − !() − "(#) (4) TE D ∆ = () , and (#) are the calculated free energies of LixOy (denotes the final product in crystal), Li (from bulk bcc structure) and O2 molecule (triplet as EP the ground state), respectively. Through above method we determined the equilibrium potential of Li2O at 298.15 K and 1atm as 2.83 V, which agrees well with the AC C experimental result of 2.91 V. The free energy change (∆G) is calculated following ∆G = ∆E + ∆ZPE − T∆S (5) where ∆E is the change in total energy, ∆ZPE is the difference in zero point vibrational energy, T = 298.15 K and ∆S is the change in entropy. The zero point vibrational energy and entropy are obtained from frequency calculations on the optimized structures. The Bader charge analysis developed by Henkleman group is applied for quantitative charge analysis by using the Vienna Ab initio Simulation Package ACCEPTED MANUSCRIPT (VASP),[45, 46] due to its more reasonable results on positive charge assigned on Li. More details on VASP calculations and the comparison on charge population can be found in Supplementary material. RI PT 3. Results and discussion The geometric structures of the siligraphenes investigated here are shown in Figure 1. Due to the larger electronegativity of C than Si, the electron tends to accumulate around the C-rich zone. As displayed in Figure 1, the deformation SC electron density contours clearly depict that the electron density increases around the C atoms in SL-SiC, the 4-C domain in g-SiC2 and the 6-C domain in g-SiC3. The M AN U accumulation of electron leads to the positively charged Si sites. According to our charge analysis, the Si in SL-SiC, g-SiC2 and g-SiC3 carries positive charge of +1.40 |e|, +1.25 |e|, and +1.06 |e|, respectivey. Compared with the pristine graphene with highly delocalized electronic distribution, the positively charged Si atoms could AC C EP TE D provide suitable adsorption sites for the nucleophilic O2 molecule. Figure 1. The geometric structures and deformation electron density contours of (a) SL-SiC, (b) g-SiC2 and (c) g-SiC3. The grey spheres denote C atoms and the yellow spheres denote Si atoms. The “H” or “C” indicates different kinds of adsorption sites ACCEPTED MANUSCRIPT for Li atoms. For cathode catalysis in nonaqueous Li-O2 battery, the ORR process have two proposed mechanisms depending on the final product, that is, two-electron pathway or reduced to Li2O2 following, 2( + + , - ) + → (6) RI PT four-electron pathway. For two-electron pathway, the O2 molecule will be directly While for four-electron pathway, the O2 molecule binds with four Li atoms and is SC reduced to Li2O as following, 4( + + , - ) + → 2 (7) M AN U During the four-electron pathway, the Li2O2 is formed as intermediate. For both pathways the dissociation of O2 and the formation of LiO2 is the beginning step. Before we consider the O2 dissociation, it is necessary to compare the adsorption performance of Li and O2, which determines the initial growing process of LiO2 on siligraphene. For the substrate, if the adsorption of O2 is more preferred than Li, the TE D initial pathway should be O2* → LiO2* (* denotes the adsorbed substrate), and if the adsorption of Li is more preferred than O2, the initial pathway will be Li* → O2Li*. The adsorption of O2 on siligraphene has been deeply investigated in previous EP work.[31, 32, 47] It has been validated that the positively charged Si site possesses high adsorption affinity for O2. The moderate distance between the neighboring Si atoms in our investigated g-SiC2 and g-SiC3 lead to stable bridged configuration of O2 AC C with the two neighboring Si atoms, while for SL-SiC the Si-C bridged site is more energetically stable for adsorption of O2 (as shown in Figure S1). For adsorption of a single Li atom on siligraphene, we systematically investigated all the possible adsorption sites. Due to the symmetry restriction, for SL-SiC there is only one C top site, one Si top site and one hollow (H) site for Li adsorption. For g-SiC2, all the six-member rings are equal for Li adsorption (the hollow site), but there are two different types of carbon atoms, the one directly bonds with Si is marked as C1, and the one surrounded by three C1 atoms is marked as C2. For g-SiC3, all the C top sites are equal to each other, so as the Si top site. But there are two different types of ACCEPTED MANUSCRIPT hollow sites, the one composed with 2 Si atoms and 4 C atoms is marked as H1, while the one composed with 6 C atoms is marked as H2. The detailed location of (C1, C2) or (H1, H2) sites could be found in Figure 1. As listed in Table 1, the O2 show very stable adsorption with bridged configuration. However, for Li, only some of the sites RI PT can be adsorbed (the Li will automatically move from the initial site to another after optimization). More importantly, only g-SiC2 could provide small positive adsorption energy for Li, and the SL-SiC and g-SiC3 cannot prevent the clustering of individual Li atoms. The comparison results demonstrate that O2 take dominant advantages in SC adsorption on the siligraphenes, so in the following discussion we will take O2* → M AN U LiO2* as the initial pathway. Table 1. The adsorption site and corresponding adsorption energy (Eads) of Li and O2 on different siligraphenes. The H indicates hollow site. The detailed location of (C1, C2) or (H1, H2) sites are marked in Figure 1. Adsorption site Adsorbate Eads (eV) Before optimization After optimization Li Li AC C g-SiC2 EP O2 Si top Si top -0.22 C top C top -0.72 H Si top -0.22 TE D SL-SiC Li O2 0.67 Si-Si bridge 0.66 Si top H 0.22 C1 top C2 top 0.15 C2 top C2 top 0.15 H H 0.22 O2 g-SiC3 Si-C bridge Si-Si bridge 1.10 Si-C bridge 0.83 Si top H1 -0.24 C top H2 -0.16 H1 H1 -0.18 H2 H2 -0.16 Si-Si bridge 2.88 Si-C bridge 1.25 ACCEPTED MANUSCRIPT To stimulate the formation of LiO2 on siligraphene, the dissociation of adsorbed O2 molecule is the essential step. Here we performed transition state (TS) search to simulate the reaction process, as displayed in Figure S1. Our calculation results RI PT indicate that the activation barriers (Eb) during the dissociation of adsorbed O2 are all considerably low (0.30 eV on SL-SiC, 0.34 eV on g-SiC2 and 0.04 eV on g-SiC3) so that the O2* is easy to dissociate on the siligraphenes under mild condition. To exclude the influence from adsorption density, we also performed additional SC calculations on the same O2 dissociation reaction for different supercells. The calculation results in Table S3 show that the enlarging of siligraphene supercell has M AN U merely influence on both Eads of O2 and the Eb (less than 0.1 eV). Based on the fact that the adsorption of O2 is superior to Li atom and O2 is easily dissociated on the siligraphene surface, we propose that the ORR processed on the siligraphenes follows the four-electron pathway as silicene does. The detailed reaction steps should be (a) (Li++e-)+2O* Li3O2*, and (d) (Li++e-)+Li3O2* Li2O2*, (c) Li4O2* (the * denotes the TE D (Li++e-)+Li2O2* LiO2*, (b) (Li++e-)+LiO2* siligraphene surface). To make sure if a second O2 molecule will adsorb onto the Li2O2* intermediate, we constructed the Li2O4* structures and performed calculations, EP the optimized structures of Li2O4* are shown in Figure S2. As results, the second O2 molecule shows repulsive interaction against the Li2O2 on SL-SiC and g-SiC2, and weak physisorption on Li2O2-SiC3 (0.30 eV). The O-O bond of the second O2 AC C molecule merely elongates (1.282 Å in SL-SiC, 1.289 Å in g-SiC2, and 1.307 Å in g-SiC3), very close to the O-O bond length obtained from the isolated O2 molecule (1.232 Å), indicating weak electronic interaction with the Li2O2*. Moreover, previous experimental results also indicate that Li2O is preferred to form on catalysts with high oxygen catalyst bond strength. Based on these, we consider that the O2 molecule is hardly reduced by the the Li2O2* intermediate and Li4O2* is more likely to be the final product on siligraphenes. In Figure 2, we depicted the most stable structures of the optimized intermediates LiO2*, Li2O2*, Li3O2* and Li4O2*. It is worth noting that the possible adsorption site ACCEPTED MANUSCRIPT of the third Li on the top of Li2O2* is also considered (see Figure S3). The adsorption of Li on the top site of Li2O2* exhibits significantly weaker consecutive adsorption energy (∆Eads = 1.30, 1.20 and -0.10 eV for SL-SiC, g-SiC2 and g-SiC3, respectively) than that on the corresponding side site (the ∆Eads values for the third Li in Li3O2* is RI PT listed in Table 1). That’s why we chose the structures of Li3O2* in Figure 2 as the AC C EP TE D M AN U SC most stable ones. Figure 2. Schematics of the growing pathways of Li4O2 on (a) SL-SiC, (b) g-SiC2 and (c) g-SiC3. The grey, yellow, red and purple spheres represent C, Si, O and Li atoms, respectively. It should be noted that the final product Li4O2 is taken as a representative of the bulk Li2O. As we can see in Figure 3, the structures of Li4O2 unit formed on different siligraphene are highly similar with the Li4O2 units in the (110) facet of Li2O. Considering that there are plenty of Si sites, it is expected the Li2O (110) facet could form with the siligraphene surface is fully covered by Li4O2 units, as shown in Figure ACCEPTED MANUSCRIPT RI PT 3(b-d). Figure 3. (a) The (110) facet cutting from bulk Li2O, the Li4O2 units in the surface are SC displayed as ball and stick model for clarity. (b-d) The simulated siligraphene surfaces AC C EP TE D M AN U (SL-SiC, g-SiC2 and g-SiC3) that are fully covered by Li4O2 units. Figure 4. The calculated free energy diagrams for ORR/OER catalyzed by (a) SL-SiC, (b) g-SiC2 and (c) g-SiC3 at zero potential (U = 0 V), equilibrium potential (Ueq), discharge potential (UDC) and charge potential (UC). (d) The correlation between ORR overpotential and ∆Eads of last adsorbed Li in Li4O2* (the black line), as well as the correlation between OER overpotential and Eads of LiO2 on different siligraphenes ACCEPTED MANUSCRIPT (the blue line). Based on the optimized structures, we further calculated their free energy changes to elucidate the ORR and OER performance of the siligraphenes in nonaqueous Li-O2 RI PT batteries. In Figure 4 (a-c), the calculated free energy diagram for ORR/OER catalyzed by the three siligraphenes at zero potential (U = 0 V), equilibrium potential (Ueq), discharge potential (UDC) and charge potential (UC) are depicted. Here, the UDC is the highest voltage needed to make the free energies of ORR steps downhill, while SC the UC is the lowest voltage needed to make the free energies of OER steps downhill. Due to the difference between practical potential and equilibrium potential, the M AN U overpotential arises, which is useful to evaluate the catalytic performance by calculating 011 = − 23 , and 041 = 5 − . According to the definition, 011 is the discharge overpotential and 041 is the charge overpotential. Obviously, the low overpotential value means high catalytic activity for the catalyst. From the free energy diagram we can see that the discharge potential of TE D siligraphene catalyzed ORR follows the sequence of SL-SiC (2.10 V) < g-SiC2 (1.98 V) < g-SiC3 (0.97 V). As comparison, SL-SiC and g-SiC2 possess higher discharge potential than that of graphene (calculated as 1.21 - 1.35 V[14, 19]), showing the EP potential as candidates for high density energy storage. The charge potential of siligraphene catalyzed ORR also follows the sequence of SL-SiC (4.70 V) < g-SiC2 (5.03 V) < g-SiC3 (7.44 V). The results indicate that among the three siligraphenes the AC C SL-SiC has the best performance due to its high discharge voltage as well as low charge voltage, while g-SiC3 is unsuitable to be used as cathode catalytic material. Concluded from Figure 4 (a-c), another finding is that for siligraphenes, the rate determining step (RDS) during ORR process is the formation of Li4O2*, and the RDS during OER process is the formation of LiO2*. To reveal the possible factor that affects ORR overpotential, we calculated all the ∆Eads values of the newly inserted Li in LixO2*. As listed in Table 2, all of the ∆Eads values are positive, indicating that the insertion of Li atom is exothermic process. Considering that the formation of Li4O2* is the RDS for ORR process, we found that ACCEPTED MANUSCRIPT the ∆Eads of Li in Li4O2* shows strong correlation with the ORR overpotential, that is, the higher ∆Eads of Li in Li4O2* (the last adsorbed Li) is, the lower ORR overpotential is needed. This correlation is linear and is depicted in Figure 4(d) (the solid black triangle), with a fitting coefficient (R2) value as high as 0.995, which can be expressed 011 = −0.97∆ + 2.73 (9) RI PT as Similarly, to find out the factor that directly determines the OER overpotential, we calculated the adsorption energies (Eads) of LixO2 on siligraphenes following (10) SC = + ∗ − ∗ M AN U where the represents the energy of isolated LixO2 cluster. As listed in Table 2, we also disclose that the Eads of LiO2 is strongly related with the OER overpotential since the formation of LiO2 is the RDS of OER process. The linear correlation depicted in Figure 4(d) (the hollow blue triangle) shows a fitting coefficient (R2) value almost equal to 1. This correlation can be described as (11) TE D 041 = 1.03 − 3.32 It is not accidental that there is such quantitative correlation between the adsorption performance of LixO2 intermediates and the catalytic performance of cathode catalysts. Similar correlation is also reported when the N or B doped EP graphene is used as the cathode catalysts. From the microscopic view, it is found that the ∆Eads of Li in LixO2 gradually decreases as the number of Li increases, which AC C makes the binding of the last Li atom the most difficult. The energy released when the last Li is inserted could directly determine the discharge voltage. Here the ORR overpotential actually means the potential needed to modulate the difference between surface of catalyst and bulk Li2O. That’s why during ORR process, the overpotential highly depends on the ∆Eads of last Li in Li4O2. For the OER process, the strongest adsorption affinity of LiO2 among the intermediates also makes its removal the most difficult. Combining with the fact that the charge process aims at oxidize all the formed reduced products into mobile Li atoms and O2 molecules again, the OER overpotential could help to provide the extra energy to desorb the LiO2 from the ACCEPTED MANUSCRIPT surface of catalyst. Table 2. The calculated consecutive adsorption energies (∆Eads, eV) of Li in LixO2 adsorbed siligraphenes, and the corresponding adsorption energies (Eads, eV) of LixO2 LiO2* Li2O2* ∆Eads 2.70 2.46 3.71 ∆Eads 2.29 2.57 2.29 Eads 6.00 6.45 8.59 ∆Eads 2.18 2.21 1.02 Eads 6.44 6.92 7.87 Li4O2* ∆Eads 2.03 1.95 0.89 Eads 6.51 6.92 6.81 SC SL-SiC g-SiC2 g-SiC3 Eads 5.11 5.28 7.70 Li3O2* RI PT on siligraphenes. To find out how the difference on adsorption performance origins, the Bader M AN U charge analysis is performed. For Li3O2 adsorbed siligraphenes, we revealed that the charge distributed on Li3O2 is negative and the charge value decreases following the sequence of SL-SiC > g-SiC2 > g-SiC3 for different surface, as listed in Table 3. This is consistent with the insertion energy of Li into Li3O2* [the ∆Eads (Li) in Table 3], due to the fact that larger negative charge has stronger attraction to the inserted Li TE D atom. For adsorption of LiO2 on surface of siligraphenes, we used electrostatic energy term as a descriptor. The electrostatic energy term indicates the weak electrostatic interaction between the adsorbate and substrate, which can be expressed as < < 1 . In this expression, the Q1 and Q2 stands for the charge on interacting EP ES = − particle 1 and 2, respectively, and R is the distance between the interacting particles. AC C In Table 3, we listed the charge distributed on LiO2 and siligraphene surface for LiO2* and the distance between LiO2 and surface (R). The calculated electrostatic energy terms (ES) in Table 3 show positive correlation with the corresponding Eads of LiO2. This finding reveals that weaker electrostatic energy between the adsorbed LiO2 and the surface is favorable to weaken the binding of LiO2, thus helps to its desorption during charge process and leads to a lower OER overpotential. Another important conclusion on the charge population is that to obtain lower ORR and OER overpotential, smaller negative charge on LiO2 and larger negative charge on Li3O2 is recommended, which is directly related with the catalytic surface. ACCEPTED MANUSCRIPT Table 3. The charge population Q (LiO2) and Q (surface) on LiO2 and siligraphene surface in LiO2*, the distance between LiO2 and surface (R), as well as the calculated values for electrostatic energy terms (ES), the charge population Q (Li3O2) on the also listed for comparison. Q (LiO2) Q (surface) R (Å) ES Eads (LiO2) Q (Li3O2) ∆Eads (Li) SL-SiC -1.80 1.80 1.60 2.04 5.11 g-SiC2 -1.83 1.83 1.57 2.15 5.28 g-SiC3 -2.09 2.09 1.60 2.75 7.70 -0.53 2.03 -0.51 1.98 -0.50 0.89 SC Surface RI PT adsorbed Li3O2 for different siligraphenes. The Eads (LiO2) and ∆Eads (Li) in Li4O2* is M AN U Additionally, we can summarize from Eq (9) and Eq (11) that when the final discharging product is Li2O, the ideal cathode catalyst should have proper Eads of LiO2 as well as ∆Eads of the last Li atom in Li4O2*. The obtained slope and intercept values indicate that the ideal ∆Eads of Li inserted into Li3O2* should be close to 2.81 eV (2.73/0.97) and ideal Eads value of LiO2 should be close to 3.22 eV (3.32/1.03). 2D TE D materials like g-SiC3 is unsuitable as cathode catalysts because that the extremely strong adsorption of O2 will result in exceedingly high Eads of the initially formed LiO2, and the strong binding between dissociated O and Si atom leads to relatively EP low ∆Eads for the following adsorbed Li atoms. As an example, the previously reported silicene cathode also yields very high OER overpotential (4.09 V), due to the AC C extremely strong binding of LiO2 on silicene (6.17 eV), which is consistent with our findings. Thus our investigations elucidate that the siligraphene can be used as good model system to quantitatively correlate the adsorption performance of Li-O intermediates with the catalytic performance of cathode catalyst in Li-O2 battery. It is worth noting that there has been theoretical investigation reported that the binding of Li and LiO2 are the key descriptors of catalytic activity of Pt-based bimetallic catalysts in nonaqueous Li-O2 batteries, and low ORR and OER overpotentials are closely related to strong binding of Li and weak adsorption of LiO2 on catalytic surface, respectively. Our results demonstrate that the metal-free ACCEPTED MANUSCRIPT two-dimensional siligraphene also shows similar rules and we further provide quantitative formulas to correlate the adsorption of Li and LiO2 on catalytic surface with ORR and OER overpotentials for the first time. Since the exploit on better cathode catalysts for nonaqueous Li-O2 batteries is still RI PT in its preliminary stage, the SL-SiC investigated here is promising candidate as the cathode catalyst. Comparing with many other graphene-based cathode catalysts, the SL-SiC exhibits fairly high discharge potential and acceptable charge potential. As comparison, the well-known N doped graphene exhibits discharge overpotential SC ranges from 0.89-1.66 V with the different doping configurations (Li4O4 as the discharge product). Moreover, the high concentration of Si and great structural M AN U stability will also be helpful to suppress the formation of side product Li2CO3, which could significantly limits the performance of carbon-based nanomaterial. It has been demonstrated that the C-O radical arisen in defective graphene or N-doped graphene may account for the formation of Li2CO3.[19, 24] However, in siligraphene the C-O radical won’t generate because O2 molecule is tightly captured by Si sites and results TE D in the formation of Si-O bonds. Besides, the experimental realization of SiC nanosheets has been reported by solution exfoliation or reaction between graphene EP and silicon source,[50, 51] making it practical in applications. 4. Conclusion In this article, we investigated the potential application of siligraphenes AC C (including SL-SiC, g-SiC2 and g-SiC3) as cathode catalysts in nonaqueous Li-O2 batteries by DFT calculations. The O2 molecule is easily dissociated on the surface of siligraphene, which provides adsorption sites for the binding of Li atoms to form the final product of Li4O2. Through DFT calculations, we established the reaction pathway for the discharge process and calculated the free energy diagrams for ORR/OER catalyzed by siligraphenes. Our results reveal that the SL-SiC exhibits considerably low ORR overpotential (0.73 V) and OER overpotential (1.87 V), showing the potential as suitable cathode catalyst for nonaqueous Li-O2 battery. The quantitative analysis indicates that the ORR overpotential on siligraphenes is linearly ACCEPTED MANUSCRIPT correlated with the ∆Eads of the last added Li in Li4O2*, while the OER overpotential is linearly correlated with the Eads of LiO2 in LiO2*. The charge population is used to find out the influencing factor on the adsorption performance. Moreover, our research summarized quantitative correlations between the catalytic performance of RI PT siligraphenes and the adsorption performance of LixO2 intermediates on them, which provides promising approach to develop metal-free cathode catalysts with high catalytic activity. SC Acknowledgments This work is supported by the Ministry of Science and Technology of China M AN U (Grants No. 2017YFA0204802 and 2017YFB0701601), the National Natural Science Foundation of China (Grants No. 51761145013, 21673149 and 21703145), China Postdoctoral Science Foundation (Grant No. 2017M611892), and a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). This is also a project supported by the Fund for Innovative Research Teams TE D of Jiangsu Higher Education Institutions, Jiangsu Key Laboratory for Carbon Based Functional Materials and Devices, Collaborative Innovation Center of Suzhou Nano AC C EP Science and Technology. ACCEPTED MANUSCRIPT References AC C EP TE D M AN U SC RI PT  Abraham K, Jiang Z. A polymer electrolyte based rechargeable lithium/oxygen battery. J Electrochem Soc 1996;143(1):1-5.  Armand M, Tarascon JM. Building better batteries. Nature 2008;451(7179):652-7.  Lu Y-C, Gasteiger HA, Parent MC, Chiloyan V, Shao-Horn Y. The Influence of Catalysts on Discharge and Charge Voltages of Rechargeable Li–Oxygen Batteries. Electrochem Solid State Lett 2010;13(6):A69-A72.  Li F, Kitaura H, Zhou H. The pursuit of rechargeable solid-state Li-air batteries. Energy Environ Sci 2013;6(8):2302-11.  Li Y, Wang J, Li X, Geng D, Li R, Sun X. Superior energy capacity of graphene nanosheets for a nonaqueous lithium-oxygen battery. Chem Commun 2011;47(33):9438-40.  Padbury R, Zhang X. Lithium–oxygen batteries—limiting factors that affect performance. J Power Sources 2011;196(10):4436-44.  Lu J, Li L, Park J-B, Sun Y-K, Wu F, Amine K. Aprotic and aqueous Li–O2 batteries. Chem Rev 2014;114(11):5611-40.  Xu J-J, Wang Z-L, Xu D, Zhang L-L, Zhang X-B. Tailoring deposition and morphology of discharge products towards high-rate and long-life lithium-oxygen batteries. Nat Commun 2013;4:2438.  Lu Y-C, Gallant BM, Kwabi DG, Harding JR, Mitchell RR, Whittingham MS, et al. Lithium– oxygen batteries: bridging mechanistic understanding and battery performance. Energy Environ Sci 2013;6(3):750-68.  Girishkumar G, McCloskey B, Luntz A, Swanson S, Wilcke W. Lithium− air battery: promise and challenges. J Phys Chem Lett 2010;1(14):2193-203.  Kim H-J, Jung SC, Han Y-K, Oh SH. An atomic-level strategy for the design of a low overpotential catalyst for Li− O 2 batteries. Nano Energy 2015;13:679-86.  Shao Y, Park S, Xiao J, Zhang J-G, Wang Y, Liu J. Electrocatalysts for nonaqueous lithium– air batteries: status, challenges, and perspective. ACS Catal 2012;2(5):844-57.  Xu Y, Shelton WA. Oxygen Reduction by Lithium on Model Carbon and Oxidized Carbon Structures. J Electrochem Soc 2011;158(10):A1177-A84.  Jiang HR, Zhao TS, Shi L, Tan P, An L. First-Principles Study of Nitrogen-, Boron-Doped Graphene and Co-Doped Graphene as the Potential Catalysts in Nonaqueous Li–O2 Batteries. J Phys Chem C 2016;120(12):6612-8.  Xia B, Yan Y, Wang X, Lou XWD. Recent progress on graphene-based hybrid electrocatalysts. Mater Horiz 2014;1(4):379-99.  Kim H, Lim H-D, Kim J, Kang K. Graphene for advanced Li/S and Li/air batteries. J Mater Chem A 2014;2(1):33-47.  Lee JH, Kang SG, Moon HS, Park H, Kim IT, Lee SG. Adsorption mechanisms of lithium oxides (LixO2) on a graphene-based electrode: A density functional theory approach. Appl Surf Sci 2015;351:193-202.  Zheng Y, Jiao Y, Ge L, Jaroniec M, Qiao SZ. Two-Step Boron and Nitrogen Doping in Graphene for Enhanced Synergistic Catalysis. Angew Chem 2013;125(11):3192-8.  Jiang H, Tan P, Liu M, Zeng Y, Zhao T. Unraveling the Positive Roles of Point Defects on Carbon Surfaces in Nonaqueous Lithium–Oxygen Batteries. J Phys Chem C 2016;120(33):18394-402. ACCEPTED MANUSCRIPT AC C EP TE D M AN U SC RI PT  Wu L, Hou T, Li Y, Chan KS, Lee S-T. First-principles study on migration and coalescence of point defects in monolayer graphene. J Phys Chem C 2013;117(33):17066-72.  Li Y, Wang J, Li X, Geng D, Banis MN, Li R, et al. Nitrogen-doped graphene nanosheets as cathode materials with excellent electrocatalytic activity for high capacity lithium-oxygen batteries. Electrochem Commun 2012;18:12-5.  Jing Y, Zhou Z. Computational Insights into Oxygen Reduction Reaction and Initial Li2O2 Nucleation on Pristine and N-Doped Graphene in Li–O2 Batteries. ACS Catal 2015;5(7):4309-17.  Wang H, Maiyalagan T, Wang X. Review on Recent Progress in Nitrogen-Doped Graphene: Synthesis, Characterization, and Its Potential Applications. ACS Catal 2012;2(5):781-94.  Yun K-H, Hwang Y, Chung Y-C. Effective catalytic media using graphitic nitrogen-doped site in graphene for a non-aqueous Li–O2 battery: A density functional theory study. J Power Sources 2015;277:222-7.  Ren X, Zhu J, Du F, Liu J, Zhang W. B-Doped Graphene as Catalyst To Improve Charge Rate of Lithium–Air Battery. J Phys Chem C 2014;118(39):22412-8.  Li Y, Wang J, Li X, Geng D, Banis MN, Tang Y, et al. Discharge product morphology and increased charge performance of lithium–oxygen batteries with graphene nanosheet electrodes: the effect of sulphur doping. J Mater Chem 2012;22(38):20170-4.  Ren X, Wang B, Zhu J, Liu J, Zhang W, Wen Z. The doping effect on the catalytic activity of graphene for oxygen evolution reaction in a lithium-air battery: a first-principles study. Phys Chem Chem Phys 2015;17(22):14605-12.  Son IH, Park JH, Kwon S, Park S, Rümmeli MH, Bachmatiuk A, et al. Silicon carbide-free graphene growth on silicon for lithium-ion battery with high volumetric energy density. Nat Commun 2015;6.  Li Y, Li F, Zhou Z, Chen Z. SiC2 silagraphene and its one-dimensional derivatives: where planar tetracoordinate silicon happens. J Am Chem Soc 2010;133(4):900-8.  Chen Y, Yang X-c, Liu Y-j, Zhao J-x, Cai Q-h, Wang X-z. Can Si-doped graphene activate or dissociate O2 molecule? J Mol Graph Model 2013;39(0):126-32.  Dong H, Lin B, Gilmore K, Hou T, Lee S-T, Li Y. Theoretical investigations on SiC2 siligraphene as promising metal-free catalyst for oxygen reduction reaction. J Power Sources 2015;299:371-9.  Zhang P, Xiao B, Hou X, Zhu Y, Jiang Q. Layered SiC Sheets: A Potential Catalyst for Oxygen Reduction Reaction. Sci Rep 2014;4:3821.  Zhou L-J, Zhang Y-F, Wu L-M. SiC2 Siligraphene and Nanotubes: Novel Donor Materials in Excitonic Solar Cells. Nano Lett 2013;13(11):5431-6.  Zhao M, Zhang R. Two-dimensional topological insulators with binary honeycomb lattices: SiC3 siligraphene and its analogs. Phys Rev B 2014;89(19):195427.  Delley B. An all-electron numerical method for solving the local density functional for polyatomic molecules. J Chem Phys 1990;92(1):508-17.  Delley B. From molecules to solids with the DMol3 approach. J Chem Phys 2000;113(18):7756-64.  Delley B. Fast Calculation of Electrostatics in Crystals and Large Molecules. J Phys Chem 1996;100(15):6107-10.  Perdew JP, Burke K, Ernzerhof M. Generalized Gradient Approximation Made Simple. Phys Rev Lett 1996;77(18):3865-8. ACCEPTED MANUSCRIPT AC C EP TE D M AN U SC RI PT  Grimme S. Semiempirical GGA-type density functional constructed with a long-range dispersion correction. J Comput Chem 2006;27(15):1787-99.  Monkhorst HJ, Pack JD. Special points for Brillouin-zone integrations. Phys Rev B 1976;13(12):5188-92.  Govind N, Petersen M, Fitzgerald G, King-Smith D, Andzelm J. A generalized synchronous transit method for transition state location. Comp Mater Sci 2003;28(2):250-8.  Henkelman G, Jónsson H. Improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points. J Chem Phys 2000;113(22):9978-85.  Radin MD, Rodriguez JF, Tian F, Siegel DJ. Lithium peroxide surfaces are metallic, while lithium oxide surfaces are not. J Am Chem Soc 2011;134(2):1093-103.  Henkelman G, Arnaldsson A, Jónsson H. A fast and robust algorithm for Bader decomposition of charge density. Comput Mater Sci 2006;36(3):354-60.  Kresse G, Furthmüller J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys Rev B 1996;54(16):11169-86.  Kresse G, Hafner J. Ab initio molecular dynamics for open-shell transition metals. Phys Rev B 1993;48(17):13115-8.  Dong H, Wang L, Zhou L, Hou T, Li Y. Theoretical investigations on novel SiC5 siligraphene as gas sensor for air pollutants. Carbon 2017;113:114-21.  Hwang Y, Yun K-H, Chung Y-C. Carbon-free and two-dimensional cathode structure based on silicene for lithium–oxygen batteries: A first-principles calculation. J Power Sources 2015;275:32-7.  Lu Y-C, Gasteiger HA, Crumlin E, McGuire R, Shao-Horn Y. Electrocatalytic Activity Studies of Select Metal Surfaces and Implications in Li-Air Batteries. J Electrochem Soc 2010;157(9):A1016-A25.  Lin S. Light-emitting two-dimensional ultrathin silicon carbide. J Phys Chem C 2012;116(6):3951-5.  Lin S, Zhang S, Li X, Xu W, Pi X, Liu X, et al. Quasi-two-dimensional SiC and SiC2: Interaction of Silicon and Carbon at Atomic Thin Lattice Plane. J Phys Chem C 2015;119(34):19772-9.