Enhancement in magnetic anisotropy in Co_{40}Fe_{40}B_{20}/Fullerene bilayers
Purbasha Sharangi, Esita Pandey, Shaktiranjan Mohanty, Sagarika Nayak, Subhankar Bedanta
EEnhancement in magnetic anisotropy in Co Fe B /Fullerene bilayers † Purbasha Sharangi, a Esita Pandey, a Shaktiranjan Mohanty, a Sagarika Nayak, a and Sub-hankar Bedanta ∗ a , b Organic semiconductor/ferromagnetic bilayer thin films can exhibit novel properties due to the for-mation of the spinterface at the interface. Buckminsterfullerene (C ) has been shown to exhibitferromagnetism at the interface when it is placed next to a ferromagnet (FM) such as Fe or Co.Formation of spinterface occurs due to the orbital hybridization and spin polarised charge transferat the interface. In this work, we have demonstrated that one can tune the magnetic anisotropy ofthe low Gilbert damping alloy CoFeB by introducing a C layer. We have shown that anisotropy isenhanced by increasing the thickness of C which might be a result of the formation of spinter-face. However, the magnetic domain structure remains same in the bilayer samples as comparedto the reference CoFeB film. Organic spintronics has drawn immense research interest in thelast few decades due to its applications in spin valve, magnetictunnel junctions etc. In organic spintronics, organic semiconduc-tors (OSCs) (e.g. C , Alq , ruberene etc.) are used to trans-port or control spin polarized signals . The main advantageof OSCs are their low production cost, light weight, flexible andchemically interactive nature. Usually the spin orbit coupling issmall in organic materials (e.g. C ) as they consist of low Z(atomic number) materials (in particular carbon (C)). Moreover,the zero hyperfine interaction in C results in a longer spinrelaxation time . As a consequence, spin of a carrier weaklyinteracts in organic environment and spin information is main-tained for a long time. There are several reports on organic spinvalves, organic light emitting diodes (OLED) using C as a spacerlayer . It has been shown that C ( ∼ . d − p hybridization at the interface of FM/C modifies the densityof states (DOS) and exhibits room temperature ferromagnetism.Such kind of interface is known as spinterface . It has beenshown that the fundamental magnetic properties like magneticmoment, domain structure and magnetic anisotropy can be tunedby depositing C on top of a Fe or Co layer . It has beenfound that ∼ exhibits magnetic moment ∼ µ B /cage a Laboratory for Nanomagnetism and Magnetic Materials (LNMM), School of PhysicalSciences, National Institute of Science Education and Research (NISER), HBNI, P.O.-Bhimpur Padanpur, Via Jatni, 752050, India. Email: [email protected] b Center for Interdisciplinary Sciences (CIS), National Institute of Science Educationand Research (NISER), HBNI, Jatni, 752050 India at the Fe/C interface . However, to the best of our knowl-edge no such basic study has been performed on CoFeB system.For spintronic application a low damping material is always de-sired as it directly affects the speed of a device. The main ad-vantage of taking CoFeB as a ferromagnet is that it exhibits lowGilbert damping parameter and it is amorphous in nature . It isvery important to explore the effect of interface of such a system(CoFeB/OSC) to enrich our fundamental knowledge of spinter-face.In this regard, we have prepared a CoFeB/C bilayer film andcompared the magnetic properties to its parent CoFeB single layerfilm. Also, we have varied the thickness of C layer to qualita-tively define the extent of spinterface and study any changes inthe basic magnetic properties. To study the qualitative nature ofthe interface, we have performed Kerr microscopy and ferromag-netic resonance (FMR) measurements. Single layer (CoFeB) and bilayer (CoFeB/C ) samples have beendeposited on Si (100) substrate in a multi-deposition high vac-uum chamber manufactured by Mantis Deposition Ltd., UK. Thecomposition of CoFeB considered here is 40:40:20. The base pres-sure in the chamber was × − mbar. CoFeB, C and MgOlayers have been deposited using DC sputtering, thermal evapo-ration and e-beam evaporation techniques, respectively. The sam-ples are named as S1, S2, S3, S4 for thickness of C ( t C ) takenas 0, 1.1, 5, 15 nm, respectively. The schematic of the samplestructure is shown in figure 1(a) (thicknesses not to be scaled).All the layers were deposited without breaking the vacuum toavoid oxidation and surface contamination. The deposition pres- Journal Name, [year], [vol.] ,,
EEnhancement in magnetic anisotropy in Co Fe B /Fullerene bilayers † Purbasha Sharangi, a Esita Pandey, a Shaktiranjan Mohanty, a Sagarika Nayak, a and Sub-hankar Bedanta ∗ a , b Organic semiconductor/ferromagnetic bilayer thin films can exhibit novel properties due to the for-mation of the spinterface at the interface. Buckminsterfullerene (C ) has been shown to exhibitferromagnetism at the interface when it is placed next to a ferromagnet (FM) such as Fe or Co.Formation of spinterface occurs due to the orbital hybridization and spin polarised charge transferat the interface. In this work, we have demonstrated that one can tune the magnetic anisotropy ofthe low Gilbert damping alloy CoFeB by introducing a C layer. We have shown that anisotropy isenhanced by increasing the thickness of C which might be a result of the formation of spinter-face. However, the magnetic domain structure remains same in the bilayer samples as comparedto the reference CoFeB film. Organic spintronics has drawn immense research interest in thelast few decades due to its applications in spin valve, magnetictunnel junctions etc. In organic spintronics, organic semiconduc-tors (OSCs) (e.g. C , Alq , ruberene etc.) are used to trans-port or control spin polarized signals . The main advantageof OSCs are their low production cost, light weight, flexible andchemically interactive nature. Usually the spin orbit coupling issmall in organic materials (e.g. C ) as they consist of low Z(atomic number) materials (in particular carbon (C)). Moreover,the zero hyperfine interaction in C results in a longer spinrelaxation time . As a consequence, spin of a carrier weaklyinteracts in organic environment and spin information is main-tained for a long time. There are several reports on organic spinvalves, organic light emitting diodes (OLED) using C as a spacerlayer . It has been shown that C ( ∼ . d − p hybridization at the interface of FM/C modifies the densityof states (DOS) and exhibits room temperature ferromagnetism.Such kind of interface is known as spinterface . It has beenshown that the fundamental magnetic properties like magneticmoment, domain structure and magnetic anisotropy can be tunedby depositing C on top of a Fe or Co layer . It has beenfound that ∼ exhibits magnetic moment ∼ µ B /cage a Laboratory for Nanomagnetism and Magnetic Materials (LNMM), School of PhysicalSciences, National Institute of Science Education and Research (NISER), HBNI, P.O.-Bhimpur Padanpur, Via Jatni, 752050, India. Email: [email protected] b Center for Interdisciplinary Sciences (CIS), National Institute of Science Educationand Research (NISER), HBNI, Jatni, 752050 India at the Fe/C interface . However, to the best of our knowl-edge no such basic study has been performed on CoFeB system.For spintronic application a low damping material is always de-sired as it directly affects the speed of a device. The main ad-vantage of taking CoFeB as a ferromagnet is that it exhibits lowGilbert damping parameter and it is amorphous in nature . It isvery important to explore the effect of interface of such a system(CoFeB/OSC) to enrich our fundamental knowledge of spinter-face.In this regard, we have prepared a CoFeB/C bilayer film andcompared the magnetic properties to its parent CoFeB single layerfilm. Also, we have varied the thickness of C layer to qualita-tively define the extent of spinterface and study any changes inthe basic magnetic properties. To study the qualitative nature ofthe interface, we have performed Kerr microscopy and ferromag-netic resonance (FMR) measurements. Single layer (CoFeB) and bilayer (CoFeB/C ) samples have beendeposited on Si (100) substrate in a multi-deposition high vac-uum chamber manufactured by Mantis Deposition Ltd., UK. Thecomposition of CoFeB considered here is 40:40:20. The base pres-sure in the chamber was × − mbar. CoFeB, C and MgOlayers have been deposited using DC sputtering, thermal evapo-ration and e-beam evaporation techniques, respectively. The sam-ples are named as S1, S2, S3, S4 for thickness of C ( t C ) takenas 0, 1.1, 5, 15 nm, respectively. The schematic of the samplestructure is shown in figure 1(a) (thicknesses not to be scaled).All the layers were deposited without breaking the vacuum toavoid oxidation and surface contamination. The deposition pres- Journal Name, [year], [vol.] ,, a r X i v : . [ c ond - m a t . m t r l - s c i ] F e b ig. 1 (a)Schematic of the sample structure where, t C = 0, 1.1, 5, 15 nm for samples S1, S2, S3 and S4, respectively. The thicknesses shown inthis schematic is not to be scaled to the thicknesses of the samples. (b) Cross-sectional transmission electron microscopy (TEM) image of S3. (c)Elemental mapping for individual layers. (d)The region of the sample S3 where the STEM-EDX has been performed. (e) EDX line profile for each layerof the sample S3. (f) EDX spectrum of sample S3 showing the presence of different elements. sure was × − mbar for CoFeB and × − mbar for C andMgO evaporation. The deposition rate for CoFeB and C layerswere 0.1 and ∼ layer has been depositednormal to the substrate whereas CoFeB plume was at 30 ◦ w.r.t tothe substrate normal due to chamber’s in-built geometry.To understand the growth of each layer and interfaces, cross-sectional TEM has been performed on sample S3 using a high-resolution transmission electron microscope (HRTEM) (JEOLF200, operating at 200 kV and equipped with a GATAN oneviewCMOS camera). For the compositional analysis we have per-formed scanning transmission electron microscopy - energy dis-persive x-ray spectroscopy (STEM - EDX).We have measured the hysteresis loop and magnetic do-main images at room temperature by magneto-optic Kerr ef-fect (MOKE) based microscopy manufactured by Evico magneticsGmbH, Germany. Longitudinal hysteresis loops are recorded for ± φ ) between the easyaxis (EA) and the applied magnetic field direction.In order to determine the magnetic anisotropy constant andobserve the anisotropy symmetry in the samples, angle depen-dent FMR measurements have been performed at a frequency of7 GHz for each 5 ◦ interval. During the measurement the samplewas kept on the wide coplanar waveguide (CPW) in a flip-chipmanner. Frequency dependent FMR measurements have beenperformed to calculate the Gilbert damping constant( α ). High resolution TEM image is shown in figure 1(b) and all thelayers are marked separately. It shows the amorphous growth ofCoFeB and C . Element specific mapping has been shown in fig-ure 1(c). Figure 1(d) shows the STEM image where, the brighterpart indicates the layer of the element having high atomic num-ber(Z). Presence of Boron(B) is not properly visible as it is alighter atom. Figure 1(e) and (f) represent the EDX line profileand EDX spectra, respectively. The position of the Co and Fe peakat the same place indicates the formation of CoFeB alloy. EDXspectra shows the presence of C, Mg, O, Fe and Co elements inthe sample.Figure 2 shows the in-plane angle ( φ ) dependent hystere-sis loops measured using longitudinal magneto optic Kerr effect(LMOKE) microscopy at room temperature for the samples (a)S1, (b) S2 and (c) S3. φ is defined as the angle between the EAand the applied magnetic field direction. Angle dependent hys-teresis loops show that the magnetic hard axis (HA) of the sam-ples is at 90 ◦ w.r.t the EA, which marks the presence of uniaxialanisotropy in the system. Due to the oblique angle deposition ofCoFeB magnetic layer we have observed uniaxial anisotropy in allthe samples . It should be noticed that there is a change incoercive field ( H C ) in the bilayer samples as compared to the sin-gle layer reference sample. The values of H C are 1.23, 0.70 and1.55 mT for the samples S1, S2 and S3, respectively. Change in Journal Name, [year], [vol.] , ig. 2 Hysteresis loops measured by magneto optic Kerr effect (MOKE) microscopy at room temperature in longitudinal mode by varying the angle ( φ )between the easy axis and the applied magnetic field direction for the samples (a) S1, (b) S2 and (c) S3. Fig. 3
Domain images near H C for samples S1, S2 and S3 are shown in (a) - (e), (f) - (j) and (k) – (o), respectively. The scale bars of the domain imagesfor all the samples are same and shown in image (a). The applied field (H) direction shown in image (a) was kept constant for all the measurementsand the sample was only rotated to capture the domain images at different φ . H C can be attributed to the formation of a spinterface between theCoFeB and C interface. In our previous study we have shownthat the orbital hybridization at the FM (Fe or Co)/C interfacepromote the change in anisotropy of the system .The square shaped loop along EA indicates the magnetizationreversal via domain wall motion whereas, along HA the reversaloccurs via coherent rotation. The magnetization reversal is stud-ied as a function of φ . By varying the angle ( φ ) w.r.t easy axis(0 ◦ ), we have recorded the domain images near the H C at φ ◦ ,30 ◦ , 45 ◦ , 60 ◦ and 90 ◦ . Figure 3(a)-(e), (f)-(j) and (k)–(o) showsthe magnetic domain images near H C for the samples S1, S2 andS3, respectively. Branched domains have been observed in all thesamples due to the amorphous growth of CoFeB. Although the H C is different in all the samples but the change in domain structuresis not significant between the single layer CoFeB and the bilayer CoFeB/C samples.We have further invesigated the magnetization dynamics byperforming the frequency dependent FMR measurement. The ex-perimental data has been fitted using a Lorentzian function (Eq.1), where ∆ H , H res , A and A are linewidth, resonance field, anti-symmetric and symmetric component, respectively . FMR signal = A ∆ H ( H − H res )( ( H − H res )) + ( ∆ H ) − A ( ∆ H ) − ( H − H res ) ( ( H − H res )) + ( ∆ H ) + o f f set (1)The plots of f vs H res and ∆ H vs f are shown in figure 4(a) and(b), respectively. The effective damping constant ( α ) has been Journal Name, [year], [vol.] ,,
Domain images near H C for samples S1, S2 and S3 are shown in (a) - (e), (f) - (j) and (k) – (o), respectively. The scale bars of the domain imagesfor all the samples are same and shown in image (a). The applied field (H) direction shown in image (a) was kept constant for all the measurementsand the sample was only rotated to capture the domain images at different φ . H C can be attributed to the formation of a spinterface between theCoFeB and C interface. In our previous study we have shownthat the orbital hybridization at the FM (Fe or Co)/C interfacepromote the change in anisotropy of the system .The square shaped loop along EA indicates the magnetizationreversal via domain wall motion whereas, along HA the reversaloccurs via coherent rotation. The magnetization reversal is stud-ied as a function of φ . By varying the angle ( φ ) w.r.t easy axis(0 ◦ ), we have recorded the domain images near the H C at φ ◦ ,30 ◦ , 45 ◦ , 60 ◦ and 90 ◦ . Figure 3(a)-(e), (f)-(j) and (k)–(o) showsthe magnetic domain images near H C for the samples S1, S2 andS3, respectively. Branched domains have been observed in all thesamples due to the amorphous growth of CoFeB. Although the H C is different in all the samples but the change in domain structuresis not significant between the single layer CoFeB and the bilayer CoFeB/C samples.We have further invesigated the magnetization dynamics byperforming the frequency dependent FMR measurement. The ex-perimental data has been fitted using a Lorentzian function (Eq.1), where ∆ H , H res , A and A are linewidth, resonance field, anti-symmetric and symmetric component, respectively . FMR signal = A ∆ H ( H − H res )( ( H − H res )) + ( ∆ H ) − A ( ∆ H ) − ( H − H res ) ( ( H − H res )) + ( ∆ H ) + o f f set (1)The plots of f vs H res and ∆ H vs f are shown in figure 4(a) and(b), respectively. The effective damping constant ( α ) has been Journal Name, [year], [vol.] ,, ig. 4 f vs H res and ∆ H vs f plots for S1, S2, S3 and S4 are shown in (a) and (b), respectively. Open circles represent the experimental data, whilethe solid lines are the best fits to the Eq. 2 and 3. Fig. 5
Angle dependent resonance field ( H res ) plot for all the four samplesto calculate the anisotropy constants of the system. The measurementwas performed at room temperature at 7 GHz. Open circles representthe experimental data, while the solid lines are the best fits. determined by fitting the equation 2 and 3 : f = γ π (cid:113) ( H K + H res )( H K + H res + π M e f f ) (2)where, γ (gyromagnetic ratio) = g µ B / ¯ h and g , µ B , ¯ h , H K areLande-g factor, Bohr magneton, reduced Planck’s constant andanisotropy field, respectively. ∆ H = ∆ H + πα f γ (3)where, ∆ H is the inhomogeneous line width broadening whichdepends on the magnetic homogeneity of the sample. α val-ues for the samples S1, S2, S3 and S4 are 0.0095 ± ± ± ± α might be due to the interface roughness or Table 1
The value of K for all the samples extracted from the fitting ofLLG equation. Sample K (erg/cc)S1 2.4 × S2 2.9 × S3 4.1 × S4 4.3 × other effects at the interface. It has been observed that α in-creases with higher C thickness.To quantify the change in anisotropy in all the samples wehave performed in-plane angle dependent FMR measurements ata fixed frequency of 7 GHz. Resonance field ( H res ) has been mea-sured by rotating the sample w.r.t the applied magnetic field in 5 ◦ intervals. H res vs φ plots have been shown in figure 5 to calculate theanisotropy constants of the system. The open circles represent theraw data and the solid lines are the best fits. The experimentaldata is fitted using the Landau-Lifshitz-Gilbert (LLG) equation : f = γ π (( H + K M S Cos φ )( H + π M S + K M S Cos φ )) / (4)where, K is the in-plane uniaxial anisotropy constant, φ is thein-plane angle between the easy axis w.r.t the applied mageticfield direction and M S is the saturation magnetization.The K values extracted from the fitting are listed in table 1. Ithas been observed that by introducing a C layer the anisotropyof the system increased. The possible reason behind the enhance-ment in the magnetic anisotropy is the formation of spinterface atthe CoFeB/ C interface. The anisotropy increases from 2.9 × to 4.1 × erg/cc when the C thickness is varied from 1.1 to5 nm. With further increase in C thickness (at 15 nm), there isnegligible change in the anisotropy ( . × to . × erg/cc).After a certain thickness of C layer, the spinterface thickness re- Journal Name, [year], [vol.] , ains almost constant with increasing C thickness. From ourprevious study it has been observed that the thickness of the spin-terface is ∼ C thickness doesnot change the spinterface thickness, thereby keeping the changein anisotropy value negligible. We have studied the effect of C on the magnetization reversaland the magnetic anisotropy of a low damping amorphous CoFeBlayer. In comparison to the single layer CoFeB sample the mag-netic anisotropy constant has increased for the CoFeB/C bilayersamples. The anisotropy further increases for the bilayer sam-ple with C thickness upto 5 nm . However, for bilayer samplewith thicker C layer i.e., 15 nm there is negligible change inanisotropy as compared to the bilayer sample with C thicknessof 5 nm. This behaviour can be understood in terms of the spin-terface whose thickness will be in the range of 1 to 3 nm. Furtherfrom the Kerr microscopy measurements it is observed that thereis negligible change in the branch domain pattern in the samples.The enhancement in the magnetic anisotropy might be the re-sult of d − p hybridization between the CoFeB and C layer. Thisstudy reveals that one can tune the anisotropy of ferromagneticCoFeB by introducing a C layer, which can be suitable for the fu-ture spintronics devices. Further in future, the nature of spinter-face such as thickness, magnetic moment per atom etc. should beinvestigated by experimental methods such as polarized neutronreflectometry. The results presented here might bring interest tostudy similar system theoretically to elucidate the exact nature ofthe spinterface and the origin behind it. Conflicts of interest
There are no conflicts to declare.
Author contributions:
SB has conceived the idea and coordinated the project. PS hasprepared the samples. PS and EP have performed the Kerr mi-croscopy experiments. PS and SN have performed the FMR mea-surements. SM has done TEM sample preparation. PS and SBanalyzed the data and wrote the manuscript. All authors havecontributed to the manuscript corrections.
Acknowledgements
The authors want to thank Dr. Tapas Gosh for helping inTEM imaging. The authors also acknowledge Department ofAtomic Energy, and Department of Science and Technology- Science and Engineering Research Board, Govt. of India(DST/EMR/2016/007725) for the financial support.
Notes and references
Naturematerials , 2009, , 707–716.2 N. Atodiresei, J. Brede, P. Lazi´c, V. Caciuc, G. Hoffmann,R. Wiesendanger and S. Blügel, Physical review letters , 2010, , 066601. 3 C. Barraud, P. Seneor, R. Mattana, S. Fusil, K. Bouzehouane,C. Deranlot, P. Graziosi, L. Hueso, I. Bergenti, V. Dediu et al. , Nature Physics , 2010, , 615–620.4 T. Moorsom, M. Wheeler, T. M. Khan, F. Al Ma’Mari, C. Ki-nane, S. Langridge, A. Bedoya-Pinto, L. Hueso, G. Teobaldi,V. K. Lazarov et al. , Physical Review B , 2014, , 125311.5 T. L. A. Tran, P. K. J. Wong, M. P. de Jong, W. G. van derWiel, Y. Zhan and M. Fahlman, Applied physics letters , 2011, , 222505.6 T. L. A. Tran, D. Cakır, P. J. Wong, A. B. Preobrajenski,G. Brocks, W. G. van der Wiel and M. P. de Jong, ACS appliedmaterials & interfaces , 2013, , 837–841.7 F. Djeghloul, M. Gruber, E. Urbain, D. Xenioti, L. Joly,S. Boukari, J. Arabski, H. Bulou, F. Scheurer, F. Bertran et al. , The journal of physical chemistry letters , 2016, , 2310–2315.8 M. Gobbi, F. Golmar, R. Llopis, F. Casanova and L. E. Hueso, Advanced Materials , 2011, , 1609–1613.9 X. Zhang, S. Mizukami, T. Kubota, Q. Ma, M. Oogane, H. Na-ganuma, Y. Ando and T. Miyazaki, Nature communications ,2013, , 1–7.10 T. D. Nguyen, F. Wang, X.-G. Li, E. Ehrenfreund and Z. V.Vardeny, Physical Review B , 2013, , 075205.11 S. Mallik, S. Mattauch, M. K. Dalai, T. Brückel and S. Bedanta, Scientific reports , 2018, , 1–9.12 S. Mallik, A. S. Mohd, A. Koutsioubas, S. Mattauch, B. Sat-pati, T. Brückel and S. Bedanta, Nanotechnology , 2019, ,435705.13 S. Mallik, P. Sharangi, B. Sahoo, S. Mattauch, T. Brückel andS. Bedanta, Applied physics letters , 2019, , 242405.14 S. Sanvito,
Nature Physics , 2010, , 562.15 B. B. Singh, S. K. Jena, M. Samanta, K. Biswas, B. Satpati andS. Bedanta, physica status solidi (RRL)–Rapid Research Letters ,2019, , 1800492.16 S. Mallik, N. Chowdhury and S. Bedanta, AIP Adv. 4, 097118(2014) .17 S. Mallik, S. Mallick and S. Bedanta,
Journal of Magnetismand Magnetic Materials , 2017, , 50–58.18 S. Mallik and S. Bedanta,
Journal of Magnetism and MagneticMaterials , 2018, , 270–275.19 S. Mallick, S. Mallik, B. B. Singh, N. Chowdhury, R. Gieniusz,A. Maziewski and S. Bedanta,
Journal of Physics D: AppliedPhysics , 2018, , 275003.20 B. B. Singh, S. K. Jena and S. Bedanta, Journal of Physics D:Applied Physics , 2017, , 345001.21 C. Kittel, Physical review , 1948, , 155.22 B. Heinrich, J. Cochran and R. Hasegawa, Journal of AppliedPhysics , 1985, , 3690–3692.23 S. Pan, T. Seki, K. Takanashi and A. Barman, Physical ReviewApplied , 2017, , 064012. Journal Name, [year], [vol.] ,,