A demonstration of half-metallicity in graphene using Mn3O4 nanosheet

A demonstration of half-metallicity in graphene using Mn3O4 nanosheet

Accepted Manuscript A demonstration of half-metallicity in graphene using Mn3O4 nanosheet Sumit Mandal, Moni Baskey, Shyamal K Saha PII: DOI: Referenc...

2MB Sizes 0 Downloads 20 Views

Accepted Manuscript A demonstration of half-metallicity in graphene using Mn3O4 nanosheet Sumit Mandal, Moni Baskey, Shyamal K Saha PII: DOI: Reference:

S0008-6223(13)00411-9 http://dx.doi.org/10.1016/j.carbon.2013.05.002 CARBON 8024

To appear in:

Carbon

Received Date: Accepted Date:

31 October 2012 2 May 2013

Please cite this article as: Mandal, S., Baskey, M., Saha, S.K., A demonstration of half-metallicity in graphene using Mn3O4 nanosheet, Carbon (2013), doi: http://dx.doi.org/10.1016/j.carbon.2013.05.002

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.

A demonstration of half-metallicity in graphene using Mn3O4 nanosheet

Sumit Mandal, Moni Baskey and Shyamal K Saha

*

Department of Materials Science Indian Association for the Cultivation of Science Jadavpur, Kolkata, 700032, India

Abstract: The recent prediction of the edge modification by a magnetic impurity spin to create half-metallicity in a graphene sheet is demonstrated. Ultrafine Mn3O4 sheets of size 10 nm are grown on graphene to observe the remarkable effect of half-metallicity with a 90% change in magnetoresistance

at

room

temperature.

The

demonstration

of

a

huge

negative

magnetoresistance using a magnetic impurity spin will make graphene a potential candidate for spintronic devices.

*Corresponding author. Fax: + 91 33 24732805. E-mail address: [email protected]

1. Introduction Spintronics, in which the use of carrier spin as a new degree of freedom in electronic devices represents the most promising topic of research now a days [1,2]. Electronic transport can be completely spin polarized in a class of materials known as half-metals first observed in manganese perovskite in which the behavior is metallic for one spin whereas, it is insulating or semiconducting for the other [3]. In view of potential applications to use this property in spin based electronics, tremendous efforts are being paid to search new materials with half metallic property [4,5]. It is now well established that the peculiar localized states at each zigzag edge in graphene nanoribbon are ferromagnetic in nature and the two edges are coupled through antiferromagnetic interaction [6]. Based on theoretical investigation, it was predicted that zigzag graphene nanoribbon (ZGNR) would be half-metallic under the application of electric field [7]. Very recently, spin polarized charge transport in graphene nanoribbon has been predicted due to edge modification by magnetic atoms; however, experimental results have not yet been reported [8-10]. In the previous work [11], we have reported the observation of ferromagnetism and large magnetoresistance (MR) behavior due to edge states in ultra fine graphene quantum sheets. Although, room temperature ferromagnetism and MR due to edge states in ZGNR have already been reported [12,13], the effect of MR is still very controversial and not observed so far in large graphene sheets, which have potential applications in graphene based spintronic devices. It is believed that graphene has two sub lattices and the spins created due to defects in the same sub lattice are ferromagnetically coupled while the spins in different sub lattices are coupled through antiferromagnetic interaction [11]. Extrapolating the concept of edge modified spin density in ZGNR due to magnetic impurity atoms reported in reference 8 and 9, in the present work, we

have considered that the spins in different sub lattices of large graphene sheets will be modified differently due to presence of magnetic atoms. The advantage of invoking this concept is the emergence of half-metallicity (spin polarization) in large graphene sheet for spintronic applications. To create remarkable spin polarization effect in large graphene sheet, here we have used manganese ions to modify graphene magnetism to achieve 90% change in MR at room temperature.

2. Experimental 2.1. Materials Synthesis At first, graphene oxide (GO) was prepared using modified Hummers method [14]. 0.14 g GO was dispersed in 200 ml of 10:1 Dimethylformamide (DMF) and water mixture and sonicated for 6 hrs after 5 ml of 0.01(M) Manganese Acetate [Mn(OAc)2] was added to it. At the completion of reaction, the product was filtered, washed and dried at 70°C.

2.2. Characterizations Transmission Electron Microscopy (TEM) of the as-synthesized samples are carried out by Jeol-2011 high resolution transmission electron microscope and subsequently for X-Ray Diffraction (XRD) and Fourier Transform Infrared Spectroscopy (FTIR) profiles, we have used X-RAY Diffractometer (Rich Seifert) and NICOLET MAGNA IR 750 System respectively. Raman spectroscopy was studied in Agiltron Inc (R-3000) Raman Spectrometer. The magnetic measurements have been carried out by Physical Property Measurement System (PPMS) (Cryogenic). MR of the sample was measured by impedance spectroscopy with varying magnetic

field using Agilent E4980A Precision LCR Meter and an electromagnet supplied by M/S control systems and devices (Mumbai, India).

2.3. Structural analysis and growth mechanism To create half-metallicity in large graphene sheets using magnetic impurity spin, we have grown ferrimagnetic ultra fine Mn3O4 sheets of size ~10 nm on large graphene sheets (Fig. 1a). From the high resolution lattice image, Mn3O4 phase has been identified as shown in Fig. 1b.

Fig. 1 - (a) TEM images of Mn3O4 sheets of size ~10 nm grown on graphene layers. (b) High resolution lattice image showing d-value 2.47Å corresponding to (211) plane of Mn3O4. (c) TEM image showing an isolated graphene sheet of dimension ~ 0.25µ. (d) EDS spectrum of Fig. 1(c) and the atomic % of carbon (C) and manganese (Mn) are 83.86 and 16.14 respectively shown in the inset of Fig. 1(d).

For this sample, we have investigated many graphene flakes in TEM and all of them were found to be of size 0.1-0.4μ and one such flake shown in Fig. 1c is ~ 0.25μ. To estimate the atomic ratio of C:Mn in the present sample , we have also performed the EDS measurement on the TEM image shown in Fig. 1c. The obtained spectra are shown in Fig. 1d where atomic ratio is found to be 83.86 and 16.14 for carbon (C) and manganese (Mn) respectively. Fig. 2a shows the x-ray diffraction pattern from which the same Mn3O4 phase has also been verified with JCPDS card no. 24-0734. It is also noted that the diffraction peaks are much broader due to smaller size of Mn3O4 sheets grown on graphene. To investigate the interaction with graphene during formation of Mn3O4, we have carried out FTIR for GO as well as graphene/Mn3O4 samples. Comparing the two spectra as shown in Fig. 2b it is seen that in graphene/Mn3O4, intensities of all the usual GO peaks corresponding to functional groups/moieties –OH (3425 cm-1), –COOH (1729 cm-1), –C=C (1622 cm-1), –C-O-C (1034 cm-1) decrease significantly and two additional peaks at 620 and 510 cm-1 due to Mn-O stretching respectively for tetrahedral and octahedral sites are observed [15,16]. It is also noticed that these two peaks are blue shifted than the corresponding pure counterparts which occurs at 639 and 532 cm-1 indicating the attachment of Mn3O4 phase with the graphene surface [17]. The formation of Mn3O4 layer on graphene surface in the present case as shown in Fig. 2c, is explained on the basis of XRD, FTIR and TEM results. GO is anchored by negatively several reactive sites for chemical modification of the carbon network by grafting atoms or molecules. After the addition of Mn(OAc)2 solution, Mn+2 ions are coordinated with them by electrostatic interaction due to numerous number of oxygenated groups like hydroxyl, epoxy and carboxylic acid on GO sheets [18]. It is reported that in alkaline atmosphere, GO gets reduced to graphene [19] and also Mn+2 ion oxidized to Mn+3 ion which leads to the formation of Mn3O4. Here

Fig. 2 - (a) Shows the XRD pattern which gives both the graphene and Mn3O4 phase. (b) FTIR spectra for GO and Graphene/Mn3O4. In Graphene/Mn3O4, intensities get diminished significantly compared to the usual GO peaks and two additional peaks are appeared at 620 and 510 cm-1 due to Mn-O stretching for tetrahedral and octahedral sites respectively. (c) Schematic diagram showing the formation of Mn3O4 on graphene sheet. In the first stage, Mn2+ ions are coordinated with negatively charged functional groups like hydroxyl, epoxy, and carboxylic acid. Finally on sonication, ~10 nm Mn3O4 sheets are formed on graphene. DMF:H2O mixture acts as the source of OH– ions and this reaction leads to reduction of GO to graphene and formation of Mn3O4 nanosheets simultaneously [20].

2.4 Raman spectra The sample was further characterized by Raman spectroscopy (Fig. 3) using 785 nm (1.58 eV) as excitation wavelength. The D band due to the defects on graphene sheet arises at ~ 1328 cm-1 where as the G band corresponding to the ordered sp2 bonded carbon appears at ~ 1589 cm-1. In addition the 2D band is ascribed to the double-resonant electronic process with peak position at ~ 2619 cm-1 [21,22]. The shape and intensity of this band is sensitive to the number of graphene

Fig. 3 - Raman spectra for Graphene/Mn3O4. Peaks at 1328 and 1589 cm-1 corresponds to the D and G band for graphene where as 2D peak arises at 2619 cm-1. (In the inset) two minor peaks at 311 and 351 cm-1 and one intense peak at 653 cm-1 are for Mn3O4 in the graphene/Mn3O4 sample. layers and we have estimated the number of layers by intensity ratio of G and 2D bands. Here the G and 2D peaks are of comparable intensities and hence the number of layers for this graphene/Mn3O4 sample is ~ 5 [23]. Apart from this, in the inset of Fig. 3, two minor peaks at

311 and 351 cm-1 and one intense peak at 653 cm-1 correspond to the characteristic Raman spectra of Mn3O4 in the graphene/Mn3O4 sample [24].

3. Magnetic studies 3.1. Magnetization The zero-field-cool (ZFC) and field-cool (FC) curves with 500 Oe field is shown in Fig. 4a. Bulk Mn3O4 is a ferrimagnet with Curie temperature at around 42K. But in this case, the ZFC curve shows Curie temperature at 37K which is smaller than the bulk (42K) due to decrease in size of Mn3O4 sheets confirmed by TEM as reported earlier [25]. We have also found the blocking temperature at 29.4K which is smaller than the Curie temperature as observed earlier in case of nanophase Mn3O4 [26,27]. This lowering of blocking temperature is consistent with the wellknown relation between the crystal volume V and blocking temperature TB predicted from the equation KV=25kBTB (K is the anisotropy constant and kB is Boltzmann's constant) [28]. Similar ZFC-FC behavior has also been reported in case of disordered magnetic systems due to onset of glass-like state [29]. Fig. 4b shows the hysteresis loops for the corresponding sample. Bulk Mn3O4 is an established ferrimagnetic material with coercivity beyond 6000 Oe [30,31]. But as Mn3O4 sheets are of 10 nm in size these are expected to form a superparamagnetic assembly. Here, we get the coercivity value at 5K as 744 Oe which is much smaller than the bulk value and the saturation in the magnetization has not been observed till 5T field. The hysteresis loops are retained upto 20K and dies away after the Curie temperature at 37K which is expected in normal magnetic sample. Defective graphene is reported to be ferromagnetic, although, this

ferromagnetism dies away for very high defect density. Therefore, we did not observe any signature of ferromagnetism in the present highly defective graphene sample.

Fig. 4 - (a) ZFC-FC curves with 500 Oe field showing blocking temperature at around 29.4K and Curie temperature at 37K, much lower compared to bulk material (42K). (b) The coercivity at 5K is 744 Oe and the hysteresis loop dies away after the Curie temperature 37K (in the inset).

3.2. Magnetoresistance MR in graphene due to spin polarized charge transport is a controversial and long standing problem. Several contradicting results have been reported in the literature during the last few years [7,32-34]. Among the controversial results, the most important prediction is the creation of half-metallicity in graphene due to the presence of magnetic impurity spin [8,9]. To verify the predicted results, we have investigated in detail the spin polarized charge transport in our graphene/Mn3O4 sample. Fig. 5 shows the MR results where the effect of half-metallicity is remarkable in the as synthesized sample. Here the graphene sheet is highly defective and the resistance goes beyond 1 MΩ even at room temperature. Because of the high resistance, MR measurements could not be performed with our PPMS system. We have measured the MR of the sample by impedance spectroscopy with varying magnetic field. A compact circular pellet of thickness 0.1 cm and diameter 0.8 cm is made by pressing the powder sample by a hydraulic press at 5 ton pressure for 5 minutes. Both sides of the pellet are painted uniformly by silver paste and conducting wires are connected to both sides. The pellet is then put into the middle of two poles of an electromagnet and the wires are connected to the LCR meter probe. Magnetic field is changed by varying the current in the electromagnet, and correspondingly the impedance spectra are taken by LCR meter at different magnetic fields. Fig. 5a shows the Cole-Cole diagram to estimate the dc resistance values from the frequency dependent conductivity at different magnetic fields and the MR with magnetic field is plotted in Fig. 5b. To estimate the DC resistances, the experimental data are fitted with arc of circles as shown in Fig. 5a and the x-intercepts in the real Z΄ axis are the required resistances at different magnetic fields. Remarkable effect of half-metallicity with 90% change in MR is observed in the

sample. This decrease in MR is due to the high spin state of Mn+2 in Mn3O4 phase grown on graphene sheet and acts as a magnetic impurity spin to create half-metallicity (spin polarized charge transport) in defective graphene synthesized by sonication route. These magnetic impurity

Fig. 5 - (a) Cole-Cole diagram to estimate the DC resistances for sonicated sample using impedance spectroscopy at three different magnetic fields. (b) MR results showing remarkable effect of half-metallicity with 90% change in MR. (c) and (d) MR results of graphene and Mn3O4 respectively obtained from similar impedance spectroscopic technique.

spins interact with the up spins and down spins in the defective graphene in a different manner to create impurity levels near the Fermi energy resulting different density of states for two spins [9]. It means one of the spins (say up spin) has larger density of states compared to the other creating half-metallicity which favors the charge transport for up spin than down spin. To check whether the contribution of MR arises due to graphene or Mn3O4 phase we have carried out MR measurements on individual components separately at room temperature using the same procedure applied for the earlier graphene/Mn3O4 sample. Fig. 5c shows the MR results for graphene where ~7% of unsaturated +ve MR is obtained at 1T, quite similar to the results reported in defective graphene [35]. But for Mn3O4, there is no appreciable change in MR observed as shown in Fig. 5d.

4. Conclusion Ultrafine two dimensional Mn3O4 sheets of size ~10 nm are grown on graphene sheets to verify the predicted results of spin polarized charge transport in graphene due to edge modification by magnetic impurity spin (Mn2+) and a remarkable effect of half-metallicity in large graphene sheet with a large change in MR by 90% is observed.

Acknowledgements SM acknowledges DST, New Delhi, Govt. of India, for awarding their fellowship. MB acknowledges Dept. of Chemistry, Burdwan University, for giving permission to carry out this

work. SKS acknowledges DST, New Delhi, Govt. of India for financial support, Project No. SR/NM/NS-1089/2011.

References [1] Wolf SA, Awschalom DD, Buhrman RA, Daughton JM, von Molnar S, Roukes ML, et al. Spintronics: a spin-Based electronics vision for the future. Science 2001;294(5546):1488-95. [2] Zutic I, Fabian J, Das Sarma S. Spintronics: fundamentals and applications. Rev Mod Phys 2004;76(2):323-410. [3] Park JH, Vescovo E, Kim HJ, Kwon C, Ramesh R, Venkatesan T. Direct evidence for a half-metallic ferromagnet. Nature 1998;392:794-6. [4] Kobayashi K, Umetsu RY, Kainuma R, Ishida K, Oyamada T, Fujita A, et al. Phase separation and magnetic properties of half-metal-type Co2Cr1−xFexAl alloys. Appl Phys Lett 2004;85(20):4684-6. [5] Wang L, Cai Z, Wang J, Lu J, Luo G, Lai L, et al. Novel one-dimensional organometallic half metals: vanadium-cyclopentadienyl, vanadium-cyclopentadienyl-benzene, vanadiumanthracene wires. Nano Lett 2008;8(11):3640-4. [6] Kan EJ, Li Z, Yang J, Hou JG. Will zigzag graphene nanoribbon turn to half metal under electric field? Appl Phys Lett 2007;91(24):243116. [7] Son YW, Cohen ML, Louie SG. Half-metallic graphene nanoribbons. Nature 2006;444:3479.

[8] Lisenkov S, Andriotis AN, Menon M. Magnetic anisotropy and engineering of magnetic behavior

of

the

edges

in

Co

embedded

graphene

nanoribbons.

Phys

Rev

Lett

2012;108(18):187208. [9] Rigo VA, Martins TB, da Silva AJR, Fazzio A, Miwa RH. Electronic, structural, and transport properties of Ni-doped graphene nanoribbons. Phys Rev B 2009;79(7):075435. [10] Cho Y, Choi YC, Kim KS. Graphene spin-valve device grown epitaxially on the Ni(111) substrate: a first principles study. J Phys Chem C 2011;115(13):6019-23. [11] Saha SK, Baskey M, Majumdar D. Graphene quantum sheets: a new material for spintronic applications. Adv Mater 2010;22(48):5531-6. [12] Wang Y, Huang Y, Song Y, Zhang X, Ma Y, Liang J, et al. Room-temperature ferromagnetism of graphene. Nano Lett 2009;9(1):220-4. [13] Bai J, Cheng R, Xiu F, Liao L, Wang M, Shailos A, et al. Very large magnetoresistance in graphene nanoribbons. Nat Nanotechnol 2010;5:655-9. [14] Hummers WS, Offeman RE. Preparation of Graphitic Oxide. J Am Chem Soc 1958;80(6):1339. [15] Nawaz K, Khan U, Ul-Haq N, May P, O’Neill A, Coleman JN. Observation of mechanical percolation in functionalized graphene oxide/elastomer composites. Carbon 2012;50(12):448994. [16] Shen J, Li T, Long Y, Shi M, Li N, Ye M. One-step solid state preparation of reduced graphene oxide. Carbon 2012;50(6):2134-40.

[17] Wang D, Li Y, Wang Q, Wang T. Facile synthesis of porous Mn3O4 nanocrystal–graphene nanocomposites for electrochemical supercapacitors. Eur J Inorg Chem 2012;2012(4):628-35. [18] Chen S, Zhu J, Wu X, Han Q, Wang X. Graphene Oxide-MnO2 Nanocomposites for Supercapacitors. ACS Nano 2010;4(5):2822-30. [19] Fan X, Peng W, Li Y, Li X, Wang S, Zhang G, et al. Deoxygenation of exfoliated graphite oxide under alkaline conditions: a green route to graphene preparation. Adv Mater 2008;20(23):4490-3. [20] Gattorno GR, Jacinto PS, Vazquez LR, Nemeth J, Dekany I, Dıaz D. Novel synthesis pathway of ZnO nanoparticles from the spontaneous hydrolysis of zinc carboxylate salts. J Phys Chem B 2003;107(46):12597-604. [21] Yan J, Wei T, Shao B, Fan Z, Qian W, Zhanga M, et al. Preparation of a graphene nanosheet/polyaniline composite with high specific capacitance. Carbon 2010;48(2):487-93. [22] Suzuki S, Hibino H. Characterization of doped single-wall carbon nanotubes by Raman spectroscopy. Carbon 2011;49(7):2264-72. [23] Gupta A, Chen G, Joshi P, Tadigadapa S, Eklund PC. Raman scattering from highfrequency phonons in supported n-graphene layer films. Nano Lett 2006;6(12):2667-73. [24] Julien CM, Massot M, Poinsignon C. Lattice vibrations of manganese oxides: part I. periodic structures. Spectrochim Acta A 2004;60(3):689-700. [25] Seo WS, Jo HH, Lee K, Kim B, Oh SJ, Park JT. Size-dependent magnetic properties of colloidal Mn3O4and MnO nanoparticles. Angew Chem Int Ed 2004;43(9):1115-7.

[26] Wang N, Guo L, He L, Cao X, Chen C, Wang R, et al. Facile synthesis of monodisperse Mn3O4 tetragonal nanoparticles and their large-scale assembly into highly regular walls by a simple solution route. Small 2007;3(4):606-10. [27] Zhao N, Nie W, Liu X, Tian S, Zhang Y, Ji X. Shape- and size-controlled synthesis and dependent magnetic properties of nearly monodisperse Mn3O4 nanocrystals. Small 2008;4(1):7781. [28] Beanand CP, Livingston JD. Superparamagnetism. J Appl Phys 1959;30(4):120S-9S. [29] Tokumoto M, Song YS, Tanaka K, Sato T, Yamabe T. Irreversible and Time-dependent magnetization in TDAE-C60. Solid State Commun 1996;97(5):349-54. [30] Du J, Gao Y, Chai L, Zou G, Li Y, Qian Y. Hausmannite Mn3O4 nanorods: synthesis, characterization and magnetic properties. Nanotechnology 2006;17(19):4923-8. [31] Tan Y, Meng L, Peng Q, Li Y. One-dimensional single-crystalline Mn3O4 nanostructures with

tunable

length

and

magnetic

properties

of

Mn3O4

nanowires.

Chem Commun 2011;47(19):1172-4. [32] Hill EW, Geim AK, Novoselov K, Schedin F, Blake P. Graphene spin valve devices. IEEE Trans Magn 2006;42(10):2694-6. [33] Brey L, Fertig HA. Magnetoresistance of graphene-based spin valves. Phys Rev B 2007;76(20):205435. [34] Kim WY, Kim KS. Prediction of very large values of magnetoresistance in a graphene nanoribbon device. Nat Nanotechnol 2008;3:408-12.

[35] Zhou YB, Han BH, Liao ZM, Wu HC, Yu DP. From positive to negative magnetoresistance in graphene with increasing disorder. Appl Phys Lett 2011;91(22):222502.