Modified Surface Electronic and Magnetic ... - ACS Publications

Organic spintronics, a new branch of the field of molecular electronics, has become a hot research topic.(7, 8) However, the mechanisms underlying the...
0 downloads 0 Views 4MB Size
Download PDF
ARTICLE pubs.acs.org/JPCC

Modified Surface Electronic and Magnetic Properties of La0.6Sr0.4MnO3 Thin Films for Spintronics Applications Fenghong Li,*,†,‡ Yiqiang Zhan,‡ Tsung-Hsun Lee,§ Xianjie Liu,‡ Akira Chikamatsu,^ Tzung-Fang Guo,z Hong-Ji Lin,0 J. C. A. Huang,§ and Mats Fahlman*,‡ †

State Key Laboratory for Supramolecular Structure and Materials, Jilin University, Changchun 130012, China Department of Physics Chemistry and Biology, Link€oping University, 58183 Link€oping, Sweden § Institute of Innovation and Advanced Studies, National Cheng Kung University, Tainan, Taiwan 701 ^ Department of Chemistry, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan z Institute of Electro-Optical Science and Engineering, Advanced Optoelectronic Technology Center, National Cheng Kung University, Tainan, Taiwan 701 0 National Synchrotron Radiation Research Center, Hsin-Chu, Taiwan 30076 ‡

ABSTRACT: We present the surface electronic and magnetic properties of half-metal La0.6Sr0.4MnO3 (LSMO) thin film modified by a simple cleaning procedure, the so-called SC1 (5 H2O, 1 NH4OH, 1 H2O2), at 85 C for 10 40 min in ambient atmosphere. In this study, photoemission spectroscopy (XPS/UPS), X-ray absorption spectroscopy (XAS), and X-ray magnetic circular dichroism (XMCD) are used to characterize these properties of the manganites. Thanks to SC1 treatment, the work function of LSMO changes from 4.0 4.1 to 4.8 4.9 eV obtained from UPS measurements, while its surface roughness changes from 0.268 to 0.796 nm in AFM images. Combined O 1s, Mn 2p, Sr 3d, La 4d, and Mn 3s core-level XPS spectroscopy investigations suggest that Mn and Sr contents decrease at the surface and the Mn value becomes 3.7 due to SC1 treament. Mn L-edge XAS spectra of LSMO thin film demonstrate that SC1 treatment results in a removal of Mn2+ and an increase of the Mn4+ concentration. O K-edge XAS spectra further prove an enhancement of hybridization between O 2p orbitals and egV of Mn 3d induced by more Mn4+. XMCD results show that SC1 treatment does not induce any drastic changes of magnetic properties of the LSMO thin film surface.

1. INTRODUCTION Manganite perovskites La1 xSrxMnO3 have attracted considerable attention due to their unusual electronic structure and the strong interplay between magnetic ordering and charge transport properties, such as colossal magnetoresistance (CMR).1 3 The electrical and magnetic properties of the manganites are mainly determined by the Mn valence, which is described as a mixture of Mn3+ and Mn4+, correlated with the ratio between trivalent and divalent cations as well as the oxygen (non)stoichiometry.4,5 Both the Sr doping level and the oxygen content are important parameters that dramatically affect the electrical and magnetic properties of the manganites. It is well known that the control of the oxygen content in manganites is crucial in order to obtain good transport and magnetic properties.6 Generally, oxygen (non)stoichiometry is related to spin, charge, orbit, and lattice in divalent ion-doped manganites. In the case of thin films, the electrical and magnetic properties of La1 xSrxMnO3 are easily affected by changing the oxygen atmosphere during deposition or postannealing.6 In this work, we select half-metal La0.6Sr0.4MnO3 (LSMO) thin film as a model to reveal the influence of r 2011 American Chemical Society

surface oxidation ex situ on the electronic and magnetic properties of the manganites. Organic spintronics, a new branch of the field of molecular electronics, has become a hot research topic.7,8 However, the mechanisms underlying the spin injection into the organic semiconductors still remain one of the main challenges of this new uprising field. Concerning spin injection, LSMO are especially attractive since the conductivity mismatch limitation for direct spin-polarized injection at the ferromagnetic semiconductor interface does not apply to these fully spin-polarized systems. Thereby, it is interesting to treat the LSMO thin film surface to produce an excellent high work function (WF) spin-injecting contact. Findings from our group and Grobosch et al9 demonstrate that the WF of the LSMO surface only cleaned by ultrasonic treatment in organic solvents (acetone and isopropanol) is around 4 4.2 eV. It is substantially smaller than that of a clean Received: December 20, 2010 Revised: July 25, 2011 Published: July 27, 2011 16947

dx.doi.org/10.1021/jp112064y | J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C LSMO surface prepared in situ (∼4.8 eV). When the LSMO substrates are annealed in ultrahigh vacuum and then in an oxygen atmosphere for half an hour at ∼500 C, the WF of such an atomically clean LSMO is 4.8 4.9 eV.10 However, this treatment procedure is not applicable for the practical device fabrication. For many applications, the electrode surfaces are only cleaned by ex situ procedures using ultrasonic treatment in organic solvents prior to the deposition of organic materials in organic spin valves.11,12 To remove organic contaminants and increase oxygen stoichiometry on the LSMO surface, we used a different ex situ cleaning procedure; that is, after ultrasonic treatment in organic solvents, LSMO is heated in the solution, the so-called SC113 or TL114,15 (5 H2O, 1 NH4OH, 1 H2O2), at 85 C for 10 40 min. The WF of LSMO treated by SC1 is 4.8 4.9 eV depending on treatment time. Here, we propose the application of photoemission spectroscopy (XPS/UPS), X-ray absorption spectroscopy (XAS), and X-ray magnetic circular dichroism (XMCD) to investigate the electronic and magnetic properties of LSMO thin films with or without SC1 treatment.

2. EXPERIMENTAL SECTION The LSMO thin films were grown epitaxially on SrTiO3 (STO) substrates by pulsed laser deposition. Organic solvents used in this study are acetone and isopropanol. AFM measurements were performed under ambient conditions using a Digital Instrument Multimode Nanoscope IIIA operating in the tapping mode. Photoemission experiments were carried out using a Scienta ESCA 200 spectrometer in ultrahigh vacuum with a base pressure of 1  10 10 mbar. The measurement chamber is equipped with a monochromatic Al (KR) X-ray source and He discharged lamp providing photons with 1486.6 eV for X-ray photoemission spectroscopy (XPS) and 21.22 eV for ultraviolet photoemission spectroscopy (UPS), respectively. The XPS experimental condition was set so that the full width at halfmaximum of the clean Au 4f7/2 line was 0.65 eV. The total energy resolution of the measurements in UPS, determined by the Fermi edge of clean gold, is about 0.1 eV. All spectra were measured at a photoelectron takeoff angle of 0 (normal emission). The UPS spectra have been corrected for the contributions from He I satellite radiation. The work functions of the films were extracted from the determination of the high binding energy cutoff of the UPS spectra by applying a bias of 3 V to the sample. The polarization-dependent XAS and XMCD experiments were performed at the Dragon beamline of the National Synchrotron Radiation Research Center (NSRRC) in Taiwan with a photon energy resolution of 0.3 eV in a chamber with a base pressure of 2  10 10 mbar. XAS spectra of the Mn L2,3-edge and O K-edge were taken in the total electron yield (TEY) by measuring the sample drain current at room temperature. The degree of linear polarization of the incident light was 99%. XMCD of Mn L2,3edge was recorded at 100, 150, 200, 220, 250, and 302 K in a 1000 G magnetic field with approximately 80% circularly polarized light using the TEY method . The magnetic field makes an angle of 30 with respect to the Poynting vector of the soft X-rays. 3. RESULTS AND DISCUSSION We summarized the WF values of La1 xSrxMnO3 (x = 0.3, 0.4) obtained from UPS measurements in Table 1. The WF is defined by secondary electron cutoff in UPS spectra of samples listed in Table as shown in Figure 1. Obviously, SC1 treatment modifies the WF of La1 xSrxMnO3 up to 4.8 4.9 eV, which is

ARTICLE

Table 1. Work Function Values of La1 xSrxMnO3 Treated by Various Methods work function La1 xSrxMnO3

treatment

(eV)

sample 1 La0.6Sr0.4MnO3

ultrasonic treatment in organic solvents (acetone, isopropanol)

4.0 4.1

sample 2

SC1 for 10 40 min

4.8 4.9

ultrasonic treatment in organic solvents

4.0 4.2

La0.6Sr0.4MnO3 sample 3 La0.7Sr0.3MnO3 sample 4

(acetone, isopropanol) SC1 for 10 40 min

4.8 4.9

O2 (2  10 5 mbar) at 500 C for 30 min

4.8 4.9

La0.7Sr0.3MnO3 sample 5 La0.7Sr0.3MnO3

Figure 1. Secondary electron cutoff in UPS spectra of samples in Table 1, where the black dashed line is for sample 1, red dashed line for sample 2, black line for sample 3, red line for sample 4, and blue line for sample 5.

similar to O2 annealing (2  10 5 mbar) at 500 C for 30 min. It means that SC1 treatment is an effective cleaning procedure to increase the WF of the manganites. Moreover, when LSMO thin film is treated by SC1 in ambient, a change of its surface morphology is clearly visible due to a roughness change from 0.268 nm as prepared to 0.796 nm with 40 min SC1 treatment in Figure 2. SC1 etched the surface of LSMO because of chemical reactions between LSMO and SC1. It is the chemical reaction that took away some components from surface and left such an increased roughness at the surface. It is expected that the LSMO surface was oxidized by SC1 due to both an existence of H2O2 in SC1 and the resulting increase of WF. To explore the nature of the oxidized LSMO surface, we carried out XPS investigations. Figure 3a presents O 1s core-level spectra of LSMO ultrasonically cleaned in organic solvents (acetone and isopropanol) (LSMO 1, black line) and treated by SC1 for 40 min (LSMO 2, red and green lines), where the red line is for LSMO 2 just after half an hour of X-ray radiation, whereas the green line is for LSMO 2 after about 36 h of X-ray radiation due to nonstop XPS measurements. As expected, O 1s core-level spectra of LSMO 1 and LSMO 2 have different line shapes. However, surprisingly, a long-term XPS scan leads to an apparent decrease of peak intensity around 531 eV in the green spectrum relative to the red spectrum. It means that some oxygen species adsorbed at the surface due to SC1 treatment do not form strong chemical bonds 16948

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C

ARTICLE

Figure 2. AFM images of LSMO thin film on STO as prepared and treated by SC1 for 40 min.

Figure 3. O 1s core-level XPS spectra of LSMO thin film ultrasonically cleaned in organic solvents (acetone and isopropanol) (LSMO 1, black line) and treated by SC1 for 40 min (LSMO 2, red and green lines), where the red line is for LSMO 2 after half an hour of X-ray radiation, whereas the green line is for LSMO 2 after about 36 h of X-ray radiation due to nonstop XPS measurements (a). The three spectra in (a) can be fitted as shown in (b) (d), respectively. Height ratios: peak 1/peak 2/peak 3 = 1:0.25:0.18 in (b), peak 1/peak 2 = 1:0.9 in (c), peak 1/peak 2 = 1:0.59 in (d).

and can be easily removed from the surface by X-ray radiation. To go into more quantitative detail, we considered the fitting of the experimental intensity profiles. The high-resolution O 1s core-level spectra were fitted in the same manner with three chemical state components of binding energy for LSMO 1 and two chemical state components of binding energy for LSMO 2 in Figure 3b d (529 eV marked as peak 1, 531 eV marked as peak 2, and 532.8 eV marked as peak 3). The height ratio of the three peaks (peak 1/peak 2/peak 3) is 1:0.25:0.18 in Figure 3b, whereas height ratios of the two peaks (peak 1/peak 2) are 1:0.9 in Figure 3c and 1:0.59 in Figure 3d, respectively. The interpretation of the O 1s peaks at the manganite surface is difficult due to the presence of surface contaminants. The peak 3 at 532.8 eV in the spectra of LSMO 1 is related to the organic contaminants at the surface because it almost disappears in the spectra of LSMO 2. The SC1 treatment gets rid of organic contaminants. The peak 1 at 529 eV can be assigned to Mn O

bonding.16 18 The peak 2 at 531 eV seems related to oxygen associated with lattice defects in the oxide17 and adsorbed oxygen species, such as O or OH , because this peak relative to peak 1 increases from 0.25 for LSMO 1 in Figure 3b to 0.9 for LSMO 2 in Figure 3c after SC1 treatment, and then to 0.59 for LSMO 2 in Figure 3d due to a 36 h X-ray radiation. It is possible that part of the volatile and more loosely bound Sr content at the surface is removed by the applied chemicals. This removal creates a large number of lattice defects, causing the steep increase of the O 1s component corresponding to peak 2. To quantify changes of stoichiometry induced by SC1 treatment, Mn 2p, Sr 3d, and La 4d core-level XPS spectra of LSMO thin film are recorded in Figure 4, in which the black line shape is the spectrum of LSMO ultrasonically cleaned in organic solvents and the red one is that of LSMO thin film treated by SC1 for 40 min. The peaks corresponding to La 4d keep constant before and after SC1 treatment. Therefore, we consider that 16949

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C

ARTICLE

Figure 4. Mn 2p, Sr 3d, and La 4d core-level XPS spectra of LSMO thin film ultrasonically cleaned in organic solvents (black line) and treated by SC1 for 40 min (red line).

Figure 6. Mn L-edge X-ray absorption spectra of LSMO thin film on STO ultrasonically cleaned in organic solvents (black line), and treated by SC1 for 10 min (red line) and 40 min (green line).

Figure 5. Mn 3s core-level XPS spectra of LSMO thin film on STO ultrasonically cleaned in organic solvents (black line) and treated by SC1 for 40 min (red line).

SC1 treatment hardly takes away the La content. However, it is clear that the contents of Sr and Mn at the surface decrease, as shown in Figure 4. Roughly, the peak area of Sr 3d after SC1 treatment is about 40% of that of Sr 3d before SC1 treatment, whereas the peak area of Mn 2p after SC1 treatment is about 20% of that of Mn 2p before SC1 treatment. As a result, we conclude that the component of the surface is not La1 xSrxMnO3 any more due to the removal of both some Sr and lots of Mn by SC1. The 3s core level of the 3d transition metals is known to exhibit an exchange splitting. The exchange splitting of the Mn 3s core-level signal, 4E3s, is a definite indicator for the Mn valence VMn. Therefore, the Mn 3s core-level XPS measurement was carried out to quantitatively determine the Mn valence and clearly analyze the O content at the LSMO surface before or after SC1 treatment. Figure 5 shows Mn 3s core-level XPS spectra of LSMO thin film ultrasonically cleaned in organic solvents (black line) and treated by SC1 for 40 min (red line). 4E3s in the black spectrum for LSMO before SC1 treatment is 5.1 eV, whereas 4E3s in the red spectrum for LSMO treated by SC1 for 40 min is 4.7 eV. On the basis of the measured 4E3s, the VMn can be calculated by the following linear relationship: VMn = 9.67 1.274E3s/eV.16 From the 4E3s value, it can be concluded

that the average VMn is equal to 3.2 for LSMO before SC1 treatment as 4E3s = 5.1, whereas the average VMn is 3.7 for LSMO after SC1 treatment as 4E3s = 4.7. It is then straightforward that the O ratio is 2.9 for La0.6Sr0.4MnO3 y under the assumption of charge neutrality of the whole manganite molecule. It should be noted that an error bar of 0.1 is acceptable because of artificial elements introduced by the poor signal-tonoise ratio of the Mn 3s core-level XPS and the energy resolution of the XPS equipment we are using. It is reasonable that y = 2.9) is recovered to the original La0.6Sr0.4MnO3 y (3 state La0.6Sr0.4MnO3 within the error bar. Therefore, the 4E3s value obtained from Mn 3s XPS spectra is reliable so that it is feasible to use 4E3s to define VMn in this work. Apparently, both the Mn valence and the O content in the manganite are increased after SC1 treatment. Picozzi et al. have pointed out that oxygen vacancies destroy half-metallicity, leading to possible consequences on the LSMO spin injection efficiency.6 Thereby, it is possible that LSMO thin films treated by SC1 feature improved spin injection properties. To explore the fine change of Mn and O atoms in terms of electronic and chemical structures, Mn L-edge and O K-edge XAS spectra of LSMO thin film are presented in Figures 6 and 7, where black spectra are for LSMO thin film ultrasonically cleaned in organic solvents, red ones for LSMO treated by SC1 for 10 min, and green ones for LSMO treated by SC1 for 40 min. Figure 6 shows Mn L2,3-edge XAS spectra probing the unoccupied Mn 3d states via the 2p f 3d dipole transition between 638 and 660 eV for the three kinds of treatment ways. Two main peaks are clearly visible in all of the three spectra. For LSMO thin film ultrasonically cleaned in organic solvents, the main peak located at 643.3 eV is accompanied by two shoulder peaks around 640.9 and 641.7 eV in the region of the L3 transition (2p3/2 hole states), while the other main peak located at 654 eV is accompanied by one shoulder peak around 652.4 eV in the region of the L2 transition (2p1/2 hole states). This indicates that there exist Mn2+ ions corresponding to the peak at 640.9 eV4,15,19 in addition to the mixture of Mn3+ and Mn4+ ions corresponding to the peaks at 641.7 and 643.3 eV19,20 at the surface of LSMO substrates exposed to an ambient atmosphere and cleaned by organic solvents before measurement. However, after SC1 16950

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C

Figure 7. O K-edge X-ray absorption spectra of LSMO thin film on STO ultrasonically cleaned in organic solvents (black line), and treated by SC1 for 10 min (red line) and 40 min (green line).

treatment, the peak at 640.9 eV disappears, as shown in the red and green spectra of Figure 6. It means that SC1 treatment results in a removal of Mn2+ due to an oxidization reaction at the surface of LSMO. On the other hand, it is obvious that the main peak is shifted from 643.3 to 643.6 eV with a shoulder at 641.7 eV in the red spectrum for LSMO treated by SC1 for 10 min. For LSMO treated by SC1 for 40 min, the main peak is further shifted to 643.8 eV with a shoulder at 641.7 eV, as shown in the green spectrum. More importantly, the intensity of the shoulder peak at 641.7 eV gradually becomes weaker with increasing SC1 treatment time from 0 to 10 to 40 min. Because this peak is related to the Mn3+, we here conclude that the Mn4+ concentration increases at the surface with the development of the oxidization due to the extension of SC1 treatment time. As a result, it gives rise to an obvious increase of Mn valence, in agreement with the finding from the Mn 3s core-level XPS. XAS at the O K-edge has proven to be a powerful tool for addressing important questions about the physics of the manganites. The O K-edge absorption process is associated with the O 1s f 2p dipole transitions. A typical O K-edge XAS spectrum of La1 xSrxMnO3 consists of three main structures that have been identified as originating from strong hybridization of the O 2p orbitals with various unoccupied orbitals: Mn 3d at 528 532 eV, La 5d/4f and Sr 4d at 532 537 eV, and Mn 4sp and La 6sp around 543 eV.20,21 The effect of SC1 treatment on the Sr ratio and Mn valence is further supported by the data obtained at the O K-edge XAS measurements. Figure 7 shows O K-edge XAS spectra for the three kinds of treatment ways. In the black spectrum for LSMO ultrasonically cleaned in organic solvents, there are three main features at 543.6, 536.4, and 530 eV. The origin of the broad peak at 543.6 eV is attributed to electron bands of Mn 4sp and La 6sp character, whereas the one at 536.4 eV is related to bands of La 5d/4f and Sr 4d characters. The relevant peaks for the study of Mn valence changes are in the region of 528 532 eV due to dipole transitions from O 1s to O 2p states that are hybridized with the unoccupied Mn 3d sublevels.19,22 Therefore, the intensity of the peaks within the region represents the 2p hole and is also an indirect measure of the Mn 3d level occupancy. Here, we assigned the relevant Mn 3d states, namely, the shoulder at 529 eV, main peak at 530 eV, and faint peak at 532.6 eV, involved in the region of the black spectrum, to egv, t2gV, and egV.21 However, we can find that the

ARTICLE

main peak at 530 eV accompanied by its shoulder at 529 eV drops down, while the peak at 532.6 eV obviously increases when LSMO is treated by SC1 for 10 min. More interestingly, a new peak at 533.9 eV appears in the red spectrum. Moreover, the peak at 532.6 eV shifts to 532.3 eV and its intensity is further increased to be comparable to the intensity of another main peak at 536.6 eV, as shown in the green spectrum when the LSMO is treated by TL1 for 40 min. An increase of intensity of the peak around 532 eV indicates an enhancement of hybridization between O 2p orbitals and egV of Mn 3d induced by more Mn4+. At the same time, the peak at 533.9 eV becomes more apparent in the green spectrum. This peak should have been good evidence that SC1 treatment gives rise to an obvious change at O 1s f 2p dipole transitions. Unfortunately, the assignment of this peak cannot be found in the literature. Here, we assume that it is related to the hybridization of O 2p orbitals with La 5d/4f because the peaks corresponding to strong hybridization of O 2p orbitals with La 5d/4f and Sr 4d are at 532 537 eV and the La/Sr ratio greatly changes before and after SC1 treatment. XMCD is a powerful tool to study electronic and magnetic structures and provide information that is often difficult to obtain by other techniques. The XMCD technique can probe magnetic coupling in mixed valence of the same element provided there is useful chemical shifts between different oxidation states.4,23 25 Because the electronic and magnetic properties of LSMO are mainly determined by the mixed Mn3+ and Mn4+, the influence of SC1 treatment leading to a change of the Mn valence on the magnetic properties of LSMO thin films can be observed in the XMCD curve, which is obtained by taking the difference between XAS spectra with the in-plane magnetization parallel (Mp) and antiparallel (Ma) to the helicity of the circularly polarized light. Figure 8a,b illustrates Mn L2,3-edge XAS and XMCD curves of LSMO ultrasonically cleaned in organic solvents and treated by SC1 for 40 min, which were recorded at 100, 150, 200, 220, 250, and 302 K from bottom to top. All features obtained at room temperature (RT, 302 K) using circularly polarized incident light in Figure 8a,b are the same with the corresponding ones obtained at RT using linearly polarized incident light in Figure 6 in the error range. Compared to Figure 8a, not only the peak at 640.8 eV corresponding to Mn2+ disappears but also the peak intensity of Mn3+ at 641.5 eV becomes weak and the feature for Mn4+ moves from 643.2 to 643.6 eV for the spectra measured in two magnetization directions at RT in Figure 8b. It is in agreement with the Mn L2,3-edge XAS in Figure 6 due to the influence of SC1 treatment on the Mn oxidation states. However, the peak positions of spectra obtained at lower temperature between 100 and 250 K are different from the ones of the spectra at RT even though the line shape of all spectra does not obviously change in both panels a and b in Figure 8. All features at 100 250 K monotonously shift 0.8 0.9 eV toward lower photon energy than features at 302 K in Figure 8a, while all features at 100 250 K shift 0.7 0.8 eV toward lower photon energy than features at 302 K in Figure 8b. In addition, with decreasing the temperature from 250 to 100 K, an obvious change in terms of the line shape and peak postion of the XAS spectra cannot be seen for the two samples in both panels a and b in Figure 8. It indicates that features assigned to Mn2+, Mn3+, and Mn4+ are movable at different temperatures. In our experiments, they are around 640.8, 641.5 and 643.2 643.6 eV at RT, whereas they are around 640.0, 640.8, and 642.4 642.9 eV at 100 250 K. Figure 8c presents XMCD curves zoomed-in from Figure 8a,b. It is clear that all spectral features, including one related to Mn2+ around 639.6 eV, 16951

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C

ARTICLE

Figure 8. Mn L2,3-edge XAS and XMCD curves of LSMO thin film on STO ultrasonically cleaned in organic solvents (a) and treated by SC1 for 40 min (b), where XMCD curves (blue and magenta lines) were obtained by taking the difference between XAS spectra with the in-plane magnetization parallel (Mp, black line) and antiparallel (Ma, red line) at 100, 150, 200, 220, 250, and 302 K from bottom to top, respectively. XMCD curves in (a) and (b) are zoomed-in as shown in (C).

display an XMCD effect, namely, the XAS intensity of spectra depend on the magnetization direction at 100 220 K. However, it is hard to see an XMCD effect in spectra measured above 250 K. In comparison, we see that the SC1 treatment (magenta line) does not induce any drastic changes of magnetic properties of the LSMO thin film surface. The XMCD signal of LSMO treated by SC1 disappears at 220 K, however, the XMCD signal of LSMO without SC1 treatment (blue line) is still there until 250 K, indicating a small loss of the magnetic property of LSMO due to the surface modification.

4. CONCLUSION Modified surface electronic and magnetic properties of halfmetal La0.6Sr0.4MnO3 (LSMO) thin film by SC1 treatment have been investigated using UPS, AFM, XPS, XAS, and XMCD measurements. Thanks to SC1 treatment, the work function of LSMO changes from 4.0 4.1 to 4.8 4.9 eV. Combined O 1s, Mn 2p, Sr 3d, La 4d, and Mn 3s core-level XPS spectroscopy investigations suggest that Sr and Mn at the surface were partially removed and the Mn value was changed into 3.7 by SC1 treatment. Mn L-edge XAS spectra of LSMO thin film demonstrate that SC1 treatment results in a removal of Mn2+ and an increase of the Mn4+ concentration. O K-edge XAS spectra further prove an enhancement of hybridization between O 2p orbitals and egV of Mn 3d induced by more Mn4+. XMCD results show that SC1 treatment does not induce any drastic changes of magnetic properties of the LSMO thin film surface. ’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected] (F.L.), [email protected] (M.F.).

’ ACKNOWLEDGMENT This work was carried out within the EU Integrated Project OFSPIN (EU-FP6-STREP). In general, the Surface Physics and Chemistry division is supported by the Swedish Research

Council (project grant) and the Knut and Alice Wallenberg Foundation.

’ REFERENCES (1) Hueso, L. E.; Pruneda, J. M.; Burnell, V. G.; Valdes-Herrera, J. P.; Simons, B. D.; Littlewood, P. B.; Artacho, E.; Fert, A.; Mathur, N. D. Transformation of spin information into large electrical signals using carbon nanotubes. Nature 2007, 445, 410–3. (2) Park, J.-H.; Vescovo, E.; Kim, H.-J.; Kwon, C.; Ramesh, R.; Venkatesan, T. Direct evidence for a half-metallic ferromagnet. Nature 1998, 392, 794–796. (3) Tokura, Y., Ed. Colossal Magnetoresistive Oxides. Advances in Condensed Matter Science; Gordon and Breach: Amsterdam, 2000; Vol. 2. (4) de Jong, M. P.; Bergenti, I.; Dediu, V. A.; Fahlman, M.; Marsi, M.; Taliani, C. Evidence for Mn2+ ions at surfaces of La0.7Sr0.3MnO3 thin films. Phys. Rev. B 2005, 71, 014434. (5) Sakai, J.; Ito, N.; Imai, S. Oxygen content of La1 xSrxMnO3 y thin films and its relation to electric-magnetic properties. J. Appl. Phys. 2006, 99, 08Q318. (6) Picozzi, S.; Ma, C.; Yang, Z.; Bertacco, R.; Cantoni, M.; Cattoni, A.; Petti, D.; Brivio, S.; Ciccacci, F. Oxygen vacancies and induced changes in the electronic and magnetic structures of La0.66Sr0.33MnO3: A combined ab initio and photoemission study. Phys. Rev. B 2007, 75, 094418. (7) Barraud, C.; Seneor, P.; Mattana, R.; Fusil, S.; Bouzehouane, K.; Deranlot, C.; Graziosi, P.; Hueso, L.; Bergenti, I.; Dediu, V. A.; Petroff, F.; Fert, A. Unravelling the role of the interface for spin injection into organic semiconductors. Nat. Phys. 2010, 6, 615–620. (8) Dediu, V. A.; Hueso, L. E.; Bergenti, I.; Taliani, C. Spin routes in organic semiconductors. Nat. Mater. 2009, 8, 707–716. (9) Grobosch, M.; Dorr, K.; Gangineni, R. B.; Knupfer, M. Energy level alignment and injection barriers at spin injection contacts between La0.7Sr0.3MnO3 and organic semiconductors. Appl. Phys. Lett. 2008, 92, 023302. (10) Zhan, Y. Q.; Bergenti, I.; Hueso, L. E.; Dediu, V. A.; de Jong, M. P.; Li, Z. S. Energy level alignment at Alq3/La0.7Sr0.3MnO3 interface for organic spintronic devices. Phys. Rev. B 2007, 76, 045406. (11) Xiong, Z. H.; Wu, D.; Vardeny, Z. V.; Shi, J. Giant magnetoresistance in organic spin-valves. Nature 2004, 427, 821–824. 16952

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953

The Journal of Physical Chemistry C

ARTICLE

(12) Dediu, V. A.; Hueso, L. E.; Bergenti, I.; Riminucci, A.; Borgatti, F.; Graziosi, P.; Newby, C.; Casoli, F.; de Jong, M. P.; Taliani, C.; Zhan, Y. Room-temperature spintronic effects in Alq3-based hybrid devices. Phys. Rev. B 2008, 78, 115203. (13) Tsunoda, K.; Ohashi, E.; Adachi, S. Spectroscopic characterization of naturally and chemically oxidized silicon surfaces. J. Appl. Phys. 2003, 94, 5613. (14) Andersson, A.; Johansson, N.; Broms, P.; Yu, N.; Lupo, D.; Salaneck, W. R. Fluorine tin oxide as an alternative to indium tin oxide in polymer LEDs. Adv. Mater. 1998, 10, 859–863. (15) Li, F. H.; Graziosi, P.; Tang, Q.; Zhan, Y. Q.; Liu, X. J.; Dediu, V. A.; Fahlman, M. Electronic structure and molecular orientation of pentacene thin films on ferromagnetic La0.7Sr0.3MnO3. Phys. Rev. B 2010, 81, 205415. (16) Beyreuther, E.; Grafstrom, S.; Eng, L. M.; Thiele, C.; Dorr, K. XPS investigation of Mn valence in lanthanum manganite thin films under variation of oxygen content. Phys. Rev. B 2006, 73, 155425. (17) Choi, J.; Dulli, H.; Liou, S. H.; Dowben, P. A.; Langell, M. A. The influence of surface terminal layer and surface defects on the electronic structure of CMR perovskites: La0.65A0.35MnO3 (A d Ca, Sr, Ba). Phys. Status Solidi B 1999, 214, 45–57. (18) Choi, J.; Zhang, J. D.; Liou, S. H.; Dowben, P. A.; Plummer, E. W. The surfaces of the perovskite manganites La1 xCaxMnO3. Phys. Rev. B 1999, 59, 13453. (19) Valencia, S.; Gaupp, A.; Gudat, W.; Abad, L.; Balcells, L.; Cavallaro, A.; Martinez, B.; Palomares, F. J. Mn valence instability in La2/3Ca1/3MnO3 thin films. Phys. Rev. B 2006, 73, 104402. (20) Werner, R.; Raisch, C.; Leca, V.; Ion, V.; Bals, S.; Van Tendeloo, G.; Chasse, T.; Kleiner, R.; Koelle, D. Transport, magnetic, and structural properties of La0.7Ce0.3MnO3 thin films: Evidence for holedoping. Phys. Rev. B 2009, 79, 054416. (21) Mannella, N.; Rosenhahn, A.; Watanabe, M.; Sell, B.; Nambu, A.; Ritchey, S.; Arenholz, E.; Young, A.; Tomioka, Y.; Fadley, C. S. Temperature-dependent X-ray absorption spectroscopy of colossal magnetoresistive perovskites. Phys. Rev. B 2005, 71, 125117. (22) Abad, L.; Laukhin, V.; Valencia, S.; Gaup, A.; Gudat, W.; Balcells, L.; Martínez, B. Interfacial strain: The driving force for selective orbital occupancy in manganite thin films. Adv. Funct. Mater. 2007, 17, 3918–2925. (23) Kavich, J. J.; Warusawithana, M. P.; Freeland, J. W.; Ryan, P.; Zhai, X.; Kodama, R. H.; Eckstein, J. N. Nanoscale suppression of magnetization at atomically assembled manganite interfaces: XMCD and XRMS measurements. Phys. Rev. B 2007, 76, 014410. (24) Parlebas, J. C.; Asakura, K.; Fujiwara, A.; Harada, I.; Kotani, A. X-ray magnetic circular dichroism at rare-earth L23 absorption edges in various compounds and alloys. Phys. Rep. 2006, 431, 1–38. (25) Yokoyama, T.; Nakagawa, T.; Takagi, Y. Magnetic circular dichroism for surface and thin film magnetism: Measurement techniques and surface chemical applications. Int. Rev. Phys. Chem. 2008, 27, 449–505.

16953

dx.doi.org/10.1021/jp112064y |J. Phys. Chem. C 2011, 115, 16947–16953