1.Introduction

Lithium ion batteries(LIBs),the most important and widely used rechargeable power sources,are commonly used in portable consumer electronics such as smart phones,computers,and power tools because of their high energy density,long lifespan,light weight and environmental benignity[1–5].With rapid development of electric vehicles and hybrid electric vehicles,increasing attention has been paid to LIBs to pursue for higher energy density and better cycling stability[6–11].For anode materials,the commercial ones are usually the carbon-based graphite microspheres and mesophase carbon microbeads with high graphitization degree. Unfortunately, graphitized carbon materials cannot be satisfied for the increasing demand of high energy density due to their low specific capacity (only 372 mA h g-1) [12,13].Therefore,it is essential to develop alternative anode materials with high energy density and large power output.

Layer-structured tin disulfide(SnS2)is one of most promising anode candidates for LIBs[14–16].It has a layered CdI2-type structure with tin atoms sandwiched between two layers of hexagonal close-packed sulfur atoms,which can contribute high capacity and improved cycling stability of Li host[17–19].Serving as LIB anode,SnS2 has four electrons that can be reversibly transferred during charge/discharge processes based on the reaction SnS2+4Li++4e-↔2Li2S+Sn,leading to a theoretical capacity as high as 650 mA h g-1[20,21].In addition,it shows competitive advantages of low cost,environmental friendliness and easy fabrication as compared with other anode candidates.The large volume change of SnS2 during electrochemical cycles and low conductivity,however,are two main drawbacks that may result in fast capacity decay and poor rate capability[22,23].Recently,great effort has been made to improve the electrochemical performance of SnS2-based electrodes by using various preparation methods,such as hydrothermal technique[24,25],solution chemistry synthesis[26],and template route[27]etc.Nanosized SnS2-C nanocomposites are quite promising for practical applications,in which the introduced carbon networks such as grapheme[25,28–30]and carbon nanotubes[31,32],can not only alleviate the volume change of SnS2 during the charge/discharge processes,but also improve the conductivity[14,16,28].Furthermore,the nanosized SnS2 can provide stable electronic and ionic transfer channels to shorten the diffusion length of Li+ions[19].However,since the above-mentioned methods are too complex,costly and only limited to the level of milligram samples,there is still a great challenge to realize the large scale production of such sulfide carbon nanocomposites.Therefore,it is important to develop an effective method to economically fabricate the nanosized SnS2-based anode materials with improved electrochemical performance.

Fig.1.XRD patterns(a)and Raman spectra(b)of graphite,SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites.

In the present work,a simple but controllable ball-milling method has been developed for the synthesis of SnS2-C nanocomposite anodes.Up to now,only few works have been done on the SnS2-based anode materials using high graphite carbon content,particularly for up to 60 wt%,and the effect of carbon content on the electrochemical properties of SnS2.Therefore,it is the aim of this study to investigate the effect of nanostructure and carbon content,as represented in the SnS2-C nanocomposite materials,on the electrochemical performance of SnS2.

2.Experimental

2.1.Synthesis of SnS2-C nanocomposites

All of the chemicals were of analytical grade and used as received.Low-cost microsized tin disulfide(SnS2,particle size<10µm,purity 99%)and microsized graphite(C,particle size<10µm)were both purchased from Sinopharm Chemical Reagent Co.,Ltd.,China.The raw SnS2 was mechanically milled to reduce particle size in a vertical rotating ball mill at a rotation speed of 400 rpm on a planetary ball-mill machine(Fritsch P6,Germany)for 20 h at room temperature under argon atmosphere with a pressure of 100 kPa.Then the graphite microparticles were loaded into a stainless steel milling container and mixed with the as-prepared SnS2 NPs in different weight ratios(with 40%,50%and 60%graphite).The SnS2-C nanocomposites were produced in the same machine for another 20 h under same condition.The milled powders were removed from the ball milling container in an argon-filled glove box (MBraun, Labstar 1200/780, Germany) to achieve the final nanocomposites,named SnS2/C-40,SnS2/C-50 and SnS2/C-60,respectively.

2.2.Characterization

The phase of the synthesized samples was characterized by powder X-ray diffraction(XRD,BRUKER D8)using a Bruker D8 Advance instrument equipped with monochromatized Cu Kα radiation at 40 kV and 40 mA from 10°to 90°with a step size of 0.02°and step time of 5 s.The microstructures and sizes of the powders was observed by fieldemission scanning electron microscope(FE-SEM,FEI NOVA-450)and transmission electron microscope(TEM,PHILIPS TECNAI F30).The surface areas of the samples were estimated by the Brunauer–Emmett–Teller(BET)method using a Quanta Autosorb-iQ2-MP-ANG-VP.Raman scattering spectra were recorded at room temperature on a HORIBA JOBIN YVON S.A.S.system(model LabRAM HR800)with 532 nm incident wavelength radiation.All samples were taken from 50 to 2000 cm-1 spectra range and with a 40×objective lens.X-ray photoelectron spectroscopy(XPS)measurements were carried out with a VG ESCALAB 220iXL.

2.3.Electrochemical measurement

Fig.9 shows the rate performance of the pure SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites at different current densities.It is clear that the capacity of pure SnS2 NPs decreased significantly even after the initial 5 cycles.The charge capacities of SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites at 0.1 A g-1 after 5 cycles were 538,608 and 570 mA h g-1,respectively.The capacity retentions at 0.2,0.5,1.0 and 2.0 A g-1 for SnS2/C-50 wt%nanocomposite were 87.6%,77.4%,67.7%and 50.5%,they were higher than those of SnS2/C-40 wt%nanocomposite(89.0%,56.4%,20.2%and 5.4%)and SnS2/C-60 wt%nanocomposite(94.4%,74.9%,57.8%and 31.4%),especially at the current densities of 1.0 and 2.0 A g-1,indicating that SnS2/C-50 wt%nanocomposite exhibits the best rate performance.

Cyclic voltammetry(CV)was recorded using VMP3 electrochemical workstation(BIO-LOGIC SA France)in a voltage range of 0.01-3.0 V at a scan rate of 0.1 mV s-1.Electrochemical impendence spectroscopy(EIS)was performed on the cells by applying a sine wave with amplitude of 5 mV in the frequency range from 1000 kHz to 1 Hz.

2010～2015年，山东省基层卫生机构卫生人员数呈现先增加后降低的趋势，整体增幅11.75%，其中，社区卫生服务中心人员增幅最大，为62.35%，其次是社区卫生服务站，为22.11%，乡镇卫生院增幅为9.15%，而村卫生室人员到2015年呈现降低趋势，整体降幅5.63%。(详见表2)

3.Results and discussion

The cycling performance in Fig.8 shows that the charge capacities of pure SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites are 326,807,698 and 561 mA h g-1 at the first cycle,and 50,428,543 and 660 mA h g-1 after 100 cycles.Thus,their capacity retentions are 15.3%,53.0%,77.8%and 117.7%after 100 cycles,respectively,suggesting that SnS2/C-50 wt%nanocomposite exhibits the best cycling stability. SnS2/C-50 wt% nanocomposite has a higher specific capacity than SnS2/C-60 wt%one,which is due to the higher content of SnS2 in the sample.Increasing the content of graphite from 40 wt%to 60 wt%in the SnS2/C nanocomposite,the cycling stability is also significantly improved.More graphite in the SnS2/C-60 wt%nanocomposite can lead to better porous carbon network,higher conductivity but lower specific capacity.

In the present work,a high energy ball-milling method is adopted to synthesize SnS2-carbon nanocomposites as promising high-capacity anode materials for reversible lithium storage,and the following conclusions have been drawn:

The above plateaus agree exactly with the CV curves of SnS2/C-50 wt%nanocomposite in Fig.6(b).During the first discharge process,there were three reductive peaks at 1.62,1.89 and 1.97 V.These peaks disappeared in the following cycles,indicating that the initial electrode reaction was irreversible.After the first cycling,all the oxidative and reductive peaks were overlapped very well,which means that SnS2/C-50 wt%nanocomposite is endowed with an excellent cycling stability[29,32].

作为中国最大的钾肥工业基地，盐湖股份在致力于钾盐工业的探索和发展，碳酸钾、氢氧化钾、储能熔盐等钾盐项目已形成规模，盐湖生态镁锂钾园已具雏形的基础上，也在积极谋求“巩固钾、走出钾、提升钾”，为中国钾盐钾肥工业的发展壮大和实现石化产业“十三五”“走出去”目标做出应有贡献。未来十年，盐湖将构建现代盐湖产业体系，以“活下来、好起来、强起来”三部曲为途径，用大智慧、大手笔，利用大优势、大机遇，化解大困难、大挑战，释放大潜力、大活力，实现大转型、大发展。

几个月之后，她在香港有个演唱会，我的经纪人去看了。演唱会完了之后，她们一起出去喝东西，她跟我的经纪人讲，你知道吗？我恨死他了。经纪人回来讲给我听，我才知道，原来我这么坏，伤了人家的心，自己却不知道。

Fig.2.XPS spectra of the SnS2/C-50 wt%nanocomposite:(a)survey spectrum,(b)C1s,(c)Sn 3d,and(d)S 2p core levels.

Fig.3.SEM images:(a)graphite,(b)bulk SnS2 microparticles,(c)SnS2 NPs,(d)SnS2/C-40,(e)SnS2/C-50,(f)SnS2/C-60 wt%nanocomposites.

Fig.4.TEM image of SnS2/C-50 wt%nanonanocomposite.

Fig.5.N2 Adsorption-desorption isotherms for the SnS2/C-50 wt%nanonanocomposite.

Fig.6.Electrochemical performances of SnS2/C-50 wt%nanocomposite:(a)glavanostatic discharge-charge profiles,(b)CV curves.

Fig.6(a)shows the glavanostatic discharge-charge profiles of SnS2/C-50 wt%nanocomposite at cut-offvoltages from 0.01 to 3.0 V at a current density of 100 mA g-1 at room temperature.During the first discharge,two distinct defined voltage plateaus occurred at 1.32 and 0.32 V below 1.5 V,indicating that the main detectable reactions occurred in two steps.The former can be attributed to the insertion of Li+ions into the SnS2 layers,and the reaction was SnS2+4 Li++4 e-→Sn+2 Li2S and corresponded to the subsequent conversion to Li2S and Sn metal;the later at 0.32 V was a typical alloying reaction plateau of Li and Sn to form LixSn alloy:xLi+Sn →LixSn.However,three poorly defined voltage plateaus occurring at 1.53,1.85 and 1.97 V were found at high voltages, corresponding to reactions (1)–(3), respectively[15,19,21,25].

Fig.3 shows the SEM images of graphite,bulk SnS2 microparticles,SnS2 NPs,SnS2/C-40,SnS2/C-50,SnS2/C-60 wt%nanocomposites.The pure original graphite and bulk SnS2 microparticles with an irregular morphology in a size range of 500 nm are shown in Figs.3(a)and 3(b),respectively.Fig.3(c)shows SEM image of the bulk SnS2 microparticles after high energy ball-milling.It was seen that the obtained nanosized SnS2 was composed of porous cross-linked nanoparticles with average size of about 60 nm.The SEM images of SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites are shown in Figs.3(d),3(e)and 3(f),respectively.After the ball-milling treatment,nanosized graphite particles were uniformly dispersed in the porous SnS2 network which was composed of interconnected SnS2 NPs.

TEM image (Fig. 4) further elucidates that the nanocomposite manifests well-structured SnS2 coated by surrounding carbon.It was demonstrated that the ball-milling route gives rise to a highly uniform carbon coating,which favors to promote the electrochemical performance.BET analysis is also employed to study the surface state of the SnS2/C-50 wt%nanocomposite with the nitrogen adsorption–desorption isotherm shown in Fig.5.The BET surface area of the nanocomposite is 29.7 m2 g-1(Barrett-Joyner-Halenda(BJH))analysis.

However,SnS2/C-40 wt%nanocomposite(Fig.7(a))and SnS2/C-60 wt%nanocomposite(Fig.7(b))show different electrochemical performances.For SnS2/C-40 wt%nanocomposite,all the CV curves during different cycles cannot overlap.In addition,the intensity of the oxidative/reductive peaks derived from the reaction between lithium ions and SnS2 also decreased,indicating that the carbon content in the nanocomposites had a big influence on the electrochemical performance.Furthermore,compared with the CV profiles of bulk SnS2(Fig.7(c)),it is obvious that after the ball-milling treatment,the pure SnS2 NPs(Fig.7(d))show a significantly improvement of cycling stability,demonstrating that the introduction of nanostructure obtained from ballmilling is beneficial to the long cycling life.

Fig.8.The cycling performance of pure SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites.

Fig.1(a)shows the XRD patterns of graphite,pure SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites.For the pure SnS2 NPs,there were several distinct diffraction peaks at 2θ values of 14.8°,28.8°,30.0°,34.9°,38.1°,45.8°,50.2°,52.6°,and 62.6°,which is well assigned to a hexagonal SnS2 phase(JCPDS no.83-1705)[33].In addition,a strong peak around 26.5°in the pattern of graphite can be readily indexed to the(002)plane of graphite carbon(JCPDS no.65-6212)[34,35].Furthermore,both of the SnS2 and graphite peaks were found in other three different SnS2/C samples,indicating that the phase of SnS2 is well maintained even after high-energy ball milling.Fig.1(b)shows the Raman spectra for five samples.It was observed that there are two peaks for graphite and only one peak for SnS2.The weak peak at about 1350 cm-1(D-band)is associated with the disordered structure,while the strong peak at around 1590 cm-1(G-band)is attributed to the graphite carbon with high graphitization.According to the data,it is found that the intensity of D-band in SnS2/C nanocomposites increases significantly,while the G-band decreases,which means that the disordered structure derived from the ball-milling treatment is introduced inside the nanocomposites[35].On the other hand,the intensity and position of SnS2 peaks located at 318.2 cm-1 had no obvious change,indicating that the phase of the pristine SnS2 is protected well during the ball-milling process,which is in well agreement with the results obtained from XRD test[36].

Fig.7.The CV curves of(a)SnS2/C-40 and(b)SnS2/C-60 wt%nanocomposites,(c)bulk SnS2,(d)pure SnS2 NPs.

Fig.9.The rate performance of the pure SnS2 NPs,SnS2/C-40,SnS2/C-50 and SnS2/C-60 wt%nanocomposites at different current densities.

Fig.10.EIS curves for the pure ball-milled SnS2 and SnS2/C-50 wt%electrodes.

The electrodes were made from a mixture of active material,carbon black(AB)and polyvinylidene fluoride(PVDF)in weight ratio of 80:10:10.The mixture was dispersed in N-methylpyrrolidone(NMP)to form homogenous slurry.The slurry was spread onto a copper foil and dried overnight at 100°C in a vacuum oven.After being dried,the electrode foils were pressed and then punched into circular shape pieces.For electrochemical testing,CR2032-type coin cells were assembled in an argon-filled glove box(MBraun,Labstar 1200/780,Germany).Li metal foil was used as a counter electrode,1 mol l-1 LiPF6 in mixed solution of ethylene carbonate(EC)and dimethyl carbonate(DMC)(1:1,v/v)as the electrolyte,and Celgard 2400 as the separator.The galvanostatic charge-discharge test was conducted at cut-offvoltages from 0.01 to 3.0 V at different current densities at room temperature on an automatic Land battery measurement system(Land BT2001A,Wuhan LAND Electronics,Ltd.China).

EIS was employed to further understand the different electrochemical performances of the SnS2-C electrodes.As shown in Fig.10,the SnS2/C-50 wt%electrode shows the smallest radius of semicircle in the Nyquist plots,suggesting a lowest contact and charge-transfer resistance.The lowered resistance facilitates the electron and Li ion transfer in the electrode.Comparatively,the raw bulk SnS2 shows the highest contact and charge-transfer resistance due to its low conductivity and large particle size.Herein,the EIS analysis further confirms that the charge transfer resistance is significantly reduced due to the integration of carbon and SnS2 NPs,which offers more active sites for Li+insertion and extraction,resulting in high capacity and rate capability.

4.Conclusions

The XPS spectra recorded on the SnS2/C-50 wt%nanocomposite for C 2 s,Sn 5d,and S 2p core levels are shown in Fig.2.The survey spectrum showed the distinguished peaks corresponding to the Sn,S,and C.To understand the binding nature of the elements in SnS2/C-50 wt%nanocomposite,the C 1 s,Sn 3d,and S 2p binding energies were investigated in detail.The high resolution C 1 s peak centers at 284.4 eV(Fig.2(b))is ascribed to the presence of nonoxygenated sp2 carbon in the basal plane of graphite[34].The long tail observed at the higher binding energy indicates the presence of trace C-O(286.5 eV)bonding.However,there was no significant information on C-Sn direct bonding that is normally observed at lower binding energy(283.3 eV)[37].The sharp peaks at 487.1 and 495.5 eV are attributed to Sn 3d 5/2 and Sn 3d 3/2,respectively(Fig.2(c)).The absence of a peak at 486.3 eV indicates the absence of Sn-C direct bonding,which is in good agreement with the above C 1 s peak.The high-resolution S 2p core level analysis shows the presence of S2-species(Fig.2(d)),which shows two typical S2-peaks at 161.9 eV and 163.0 eV.

(1)The SnS2/C-50 wt%nanocomposite exhibits remarkably large capacity to 700 mA h g-1,stable cyclability with a remained capacity of 540 mA h g-1 after 100 cycles at 100 mA g-1,high coulombic efficiency(80.8%)and excellent rate capability.

灵山岛尖生态景观超级堤把单一城市防洪工程与城市道路建设、码头建设、景观建设等其他基础设施建设融合起来，将休闲公园、慢行系统等纳入其中，建设以现代公共设施为核心内容的滨海服务型经济带、景观带和文化带，成为人、水、城和谐的防洪（潮）工程典范。灵山岛尖生态景观超级堤典型剖面及俯瞰图分别如图1、图2所示。

③在开始学习阅读或识字的阶段，阅读很容易走神，可能需要手指帮忙才能顺利地开始阅读，并且在开始阅读很长时间以后还是不能摆脱这种习惯。

(2)The electrochemical performance is strongly related with the carbon content.After the ball-milling,SnS2 NPs are uniformly embedded inside the graphite nanoparticles network.The obtained structure provides abundant Li-ion storage sites,high electronic conductivity and fast ion diffusion,together with a decrease of the volume expansion of SnS2 during cycling,resulting a greatly improvement of the electrochemical performance of SnS2.

Acknowledgements

覆盖和连通度是无线传感器网络最基本的两个问题。如图2(a)所示,当满足全覆盖时,任意节点和相邻节点之间的最短距离一定小于等于2Rs,只要节点的通信半径Rc大于每个节点的最短通信距离,即满足Rc≥2Rs,就可以构成连通的网络,不需单独考虑连通度的问题[17]。图2(b)所示的百分比覆盖场景中,节点间的距离不再受限于2Rs,因此必须单独考虑网络的连通度问题。

Financial support from the national high technology research and development program (863 Program) (Grant no. 2015AA034601,2013AA032002),the National Key Scientific Instrument and Equipment Development Project(Grant no.2014YQ120351),the Beijing Natural Science Foundation(Grant no.2184134),China Iron&steel Research Institute Group Foundation(Grant no.SHI11AT0540A)is acknowledged.

References

[1] J.M.Tarascon,M.Armand,Nature 414(2001)359–367.

[2] M.Armand,J.M.Tarascon,Nature 451(2008)652–657.

[3] P.G.Bruce,B.Scrosati,J.M.Tarascon,Angew.Chem.Int.Ed.47(2008)2930–2946.

[4] B.Kang,G.Ceder,Nature 458(2009)190–193.

[5] S.Chu,A.Majumdar,Nature 488(2012)294–303.

[6] P.P.Sun,H.Y.Zhang,K.Shen,Q.Fan,Q.Y.Xu,J.Nanosci.Nanotechnol.15(2015)2667–2672.

[7] J.Choi,J.McDonough,S.Jeong,J.Yoo,C.Chan,Y.Cui,Nano Lett.10(2010)1409–1413.

[8] H.Zhao,H.Zeng,Y.Wu,S.Zhang,B.Li,Y.Huang,J.Mater.Chem.A 3(2015)10466–10470.

[9] K.M.Jeon,J.S.Cho,Y.C.Kang,J.Power Sources 295(2015)9–15.

[10] Q.Xiong,H.Chi,J.Zhang,J.Tu,J.Alloy.Compd.688(2016)729–735.

[11] D.T.Tran,S.S.Zhang,J.Mater.Chem.A 3(2015)12240–12246.

[12] X.Tang,Y.H.Wei,H.N.Zhang,F.L.Yan,M.Zhuo,C.M.Chen,P.Y.Xiao,J.J.Liang,M.Zhang,Electrochim.Acta 186(2015)223–230.

[13] S.H.Ng,J.Z.Wang,D.Wexler,K.Konstantinov,Z.P.Guo,H.K.Liu,Angew.Chem.Int.Ed.45(2006)6896–6899.

[14] T.Momma,N.Shiraishi,A.Yoshizawa,T.Osaka,A.Gedanken,J.Zhu,L.Sominski,J.Power Sources 97–98(2018)198–200.

[15] M.K.Jana,H.B.Rajendra,A.J.Bhattacharyya,K.Biswas,CrystEngComm 16(2014)3994–4000.

[16] M.S.Balogun,W.Qiu,J.Jian,Y.Huang,Y.Luo,H.Yang,C.Liang,X.Lu,Y.Tong,ACS Appl.Mater.Interfaces 7(2015)23205–23215.

[17] P.Kulkarni,S.K.Nataraj,R.Geetha Balakrishna,D.H.Nagaraju,M.V.Reddy,J.Mater.Chem.A 5(2017)22040–22094.

[18] C.Ma,J.Xu,J.Alvarado,B.Qu,J.Somerville,J.Y.Lee,Y.S.Meng,Chem.Mater.27(2015)5633–5640.

[19] B.Qu,C.Ma,G.Ji,C.Xu,J.Xu,Y.S.Meng,T.Wang,J.Y.Lee,Adv.Mater.26(2014)3854–3859.

[20] P.Norby,S.Johnsen,B.B.Iversen,Phys.Chem.Chem.Phys.17(2015)9282–9287.

[21] D.Kong,H.He,Qi Song,B.Wang,Q.H.Yang,L.Zhi,RSC Adv.4(2014)23372–23376.

[22] T.-J.Kim,C.Kirn,D.Son,M.Choi,B.Park,J.Power Sources 167(2007)529–535.

[23] Y.Idota,T.Kubota,A.Matsufuji,Y.Maekawa,T.Miyasaka,Science 276(1997)1395–1397.

[24] L.Yin,S.Chai,J.Ma,J.Huang,X.Kong,P.Bai,Y.Liu,J.Alloy.Compd.698(2017)828–834.

[25] M.S.A.S.Shah,A.R.Park,A.Rauf,S.H.Hong,Y.Choi,J.Park,J.Kim,W.J.Kimd,P.J.Yoo,RSC Adv.7(2017)3125–3135.

[26] J.T.Zai,K.X.Wang,Y.Z.Su,X.F.Qian,J.S.Chen,J.Power Sources 196(2011)3650–3654.

[27] J.Zai,X.Qian,K.Wang,C.Yu,L.Tao,Y.Xiao,J.Chen,Crystengcomm 14(2012)1364–1375.

[28] J.Jiang,Y.Feng,N.Mahmood,F.Liu,Y.Hou,Sci.Adv.Mater.5(2013)1667–1675.

[29] M.Sathish,S.Mitani,T.Tomai,I.Honma,J.Phys.Chem.C.116(2012)12475–12481.

[30] M.H.Modarres,J.H.W.Lim,C.George,M.D.Volder,J.Phys.Chem.C.121(2017)13018–13024.

[31] C.X.Zhai,N.Du,H.Zhang,J.X.Yu,D.R.Yang,ACS Appl.Mater.Interfaces 3(2011)4067–4074.

[32] H.Sun,M.Ahmad,J.Luo,Y.Shi,W.Shen,J.Zhu,Mater.Res.Bull.49(2014)319–324.

[33] S.Liu,X.Lu,J.Xie,G.Cao,T.Zhu,X.Zhao,ACS Appl.Mater.Interfaces 5(2013)1588–1595.

[34] L.Zhang,R.Rajagopalan,H.Guo,X.Hu,S.Dou,H.Liu,Adv.Funct.Mater.26(2016)440–446.

[35] L.Zhang,M.Zhang,Y.Wang,Z.Zhang,G.Kan,C.Wang,Z.Zhong,F.Su,J.Mater.Chem.A 2(2014)10161–10168.

[36] Z.X.Huang,Y.Wang,J.I.Wong,H.Y.Yang,2D Materials 2(2015)024010.

[37] M.He,L.-X.Yuan,Y.-H.Huang,RSC Adv.3(2013)3374–3383.