GB2493811A - A magneto-optical Kerr effect microscope system - Google Patents

Abstract

A microscope system, and method of microscopy, comprising means operable to generate a magneto-optical Kerr effect (MOKE) on a surface and for detecting longitudinal and polar MOKE from the surface. The microscope may be used for obtaining single-shot measurements of magnetic field-driven magnetisation switching and spin-polarised current-driven magnetic domain motion in nanostructures. This may be achieved using a submicron probe in the form of a laser beam 14, focused to a submicron diameter spot on a sample 28, and also a polarised beam splitter 22, that selectively reflects light of a particular polarisation and passes it to an optical bridge detector 32 (see Fig 5). An iris 34 can be used to cover half the reflected beam for longitudinal MOKE or left open for polar MOKE measurements. The system may also include a spin-polarised current source (see Fig 9) used in detecting the direction of spin-polarised current-induced domain wall displacement.

Description

A MICROSCOPE SYSTEM

FIELD OF THE INVENTION

The present invention relates to microscope systems and particularly to microscope systems operable to provide non-destructive single-shot measurements of longitudinal and poiar components of magnetisation switching innanostructures.

BACKGROUND OF THE INVENTION

Michael Faraday discovered the first magneto-optical effect in 1845. He found that the effect of a magnetic field applied to a glass specimen was to rotate the polarisation plane of transmitted light [M. Faraday, Trans. Roy. Soc. (London) 5(1846)592]. In 1877 John Kerr discovered a similar effect when examining the polarisation of light reflected from a polished electromagnet pole, which is the so called Magneto-Optical Ken effect (MOKE) [J. Kerr, Philos. Mag. 3 (1877) 339]. The application of the MOKE to study surface magnetism was introduced by Moog and Bader in 1985 [E. R. Moog, S. D. Bader, Superlattices Microstruct. 1(1985)543]. Since then, MOKE has emerged as a premier surface magnetism technique of choice and can be found in many laboratories worldwide.

The MOKE technique can generate the universal currency' in magnetism -the hysteresis loop. It has been applied successfully to address various contemporary magnetism issues in thin film analysis [S. D. Bader, J. Magn. Magn. Mater. 100 (1991) 440]. More recently MOKE has been used to study the behaviour of nanodot chains [R. P. Cowburn and M. E. Welland, Science 287 (2000) 1466]; selective regions of individual nanowire structures [D. A. Allwood, G. Xiong, M. D. Cooke, C. C. Faulkner, D. Atkinson, N. Vernier, and R. P. Cowburn, Science 296 (2002) 2003]; stroboscopically on ultrafast timescales [A. V. Kimel, , A. Kirilyuk, P. A. Usachev, R. V. Pisarev, A. M. Balbashov, and Th. Rasing, Nature 435 (2005) 655-657;C. Bunce, J. Wu, G. Ju, B. Lu, D. Hinzke, N. Kazantseva, U. Nowak, and R. W. Chantrell, Phys Rev B 81, (2010) 1744281, and current-induced domain wall motion [S. Lepadatua, J. Wu, andY. B. Xu, Appl. Phys. Left. 91(2007) 062512].

I

SUMMARY OF THE INVENTION

According to the present invention there is provided a microscope system comprising means operable to generate a magneto-optical Kerr effect (MOKE) on a surface and means for detecting longitudinal and polar MOKE from a said surface.

The microscope system may comprisine means for obtaining single-shot measurements of magnetic-field-driven magnetisation switching and spin-polarised, current-driven, magnetic domain wall motion in nanostmctures.

The microscope system may comprise a submicron probe.

The submicron probe is preferably generated by a laser.

The microscope system advantageously comprises at least one of a spin-polarised current

source and a magnetic field source.

The microscope system as claimed advantageously comprises an optical bridge.

The microscope system advantageously comprises a polarised beam splitter.

Also according to the present invention there is provided a method of microscopy comprising generating a magneto-optical Kerr effect (MOKE) on a surface detecting longitudinal and polar MOKE from a said surface.

The method advantageously comprises generating a magnetic field on the surface of a said sample, probing the said surface with a sub-micron probe and detecting the longitudinal and polar MOKE from the said surface. n

The sub-micron probe is preferably generated by a laser.

Also according to the present invention there is provided a computer readable medium carrying a computer program comprising computer readable instructions adapted to cause a computer to carry out the above-mentioned method.

Also according to the present invention there is provided a computer comprising the above-mentioned computer readable medium.

The invention is a combined longitudinal and polar MOKE microscope system which is capable of obtaining single-shot measurements of magnetic-field-driven magnetisation switching and spin-polarised, current-driven, magnetic domain wall motion in nanostructures using a confined submicron probe. The system improves on the standard MOKE with 1) combined Longitudinal and Polar MOKE detection; 2) a focused submieron detection area so as to analyse nanostructures; 3) a spin-polarised current source in addition to using magnetic fields as an excitation method to charaeterise current-induced domain wall motion; 4) capability of single-shot MOKE loop measurements at sub-micron spatial resolution by using a polarised beam-splitter and an optical bridge detector.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1 is a schematic diagram of Longitudinal, Polar MOKE and combined longitudinal and polar MOKE geometry.

FIGURE 2 is a schematic diagram of the single-shot MOKE setup.

FIGURE 3 is to illustrate the operation of a polarising beam-splitter.

FIGURE 4 is a CCD image of a necked nanowire with the focused laser spot.

FIGURE 5 is a schematic diagram of the optical bridge detector setup.

FIGURE 6 is the amplifier circuit for the photo-diode output of the detector.

FIGURE 7 is a screenshot of the single-shot MOKE microscope controlling program.

FIGIRE 8 is an example of the single shot MOKE loops from a necked nanowire showing four distinct types of loops, observed from the same spot with an inset of the SEM image of the necked wire.

FIGURE 9 is an illustration of the current source setup.

FIGURE 10 shows an example measurement of the current-induced domain wall motion in a necked nanowire recorded using the system.

DETAILED DESCRIPTION OF THE APPARATUS

A microscope system according to the present invention will now be described with reference to the above-mentioned accompanying drawings.

Basic single-shot MOKE microscope set-up: In use order to investigate the magnetisation switching properties of sub-niicrometre sized magnetic structures using the Magneto-Optical Kerr effect (MOKE), the laser probe spot must be focused down to a small diameter. This ensures that a large Kerr signal is obtained for small structures by maximising the area of magnetic material covered by the laser spot as a percentage of the total laser spot area. Further, a small laser spot size enables the extraction of isolated magnetisation switching properties from a particular region; increasing the spatial resolution with which a small structure may be characterised. The small laser probe spot is obtained by using a high numerical aperture (NA) objective lens at normal incidence. Using a high NA objective at normal incidence also allows both longitudinal and polar MOKE signals to be obtained without changing the set-up; which means this microscope can detect the in-plane as well as out-of-plane components of magnetisation switching.

The representative geometry for a longitudinal MOKE setup is shown in Fig. la. The incoming laser beam is linearly polarized and incident on the sample at an angle, 0, measured from the normal to the sample plane. The magnetic field is applied along the line of intersection between the sample plane and incident/reflected laser beam plane.

Due to the Kerr effect, the polarisation of the reflected laser beam is rotated by a small angle with respect to the incident laser beam polarisation angle and this polarisation rotation is directly proportional to the magnetisation in the sample. Thus by measuring the polarisation angle change between two different field values, the magnetisation switching properties of the samples under study may be extracted. For a standard polar MOKE which measures the out-of-plane components of the magnetisation, the effect is maximised at normal incidence. For a combined longitudinal and polar MOKE microscope system the same principle applies and the typical geometry is shown in Fig. lb. Here, instead of applying the laser beam at an angle to the sample, the laser beam is expanded and then directed normal to the sample and focused down by a high NA objective. After the beam passes through the objective, due to the focusing of the laser beam, effectively, one side of the beam may be considered as the incident beam and the other the reflected beam, analogous to the geometry of Fig. Ic. By covering half of the reflected beam, the longitudinal Kerr effect can be extracted; whereas using the whole reflected beam, the polar Kerr effect can be extracted. This is explained in more detail using the single-shot MOKE microscope diagram shown in Fig. 2.

Referring to Figure 2, a microscope system 10, according to the present invention, comprises a diode laser 12 which, for example, is a 5mW HeNe ultra-stable diode laser working at a wavelength of 630nm. The laser beam 14 of such a laser is linearly polarised with a diameter of 1mm. A variable neutral density filter 16 operable to control the intensity of the laser beam is suitably positioned in optical alignment with the laser 12.

A beam expander 18 is positioned in optical alignment and is operable to expand the laser beam to Snmi diameter. Expanding the laser beam to 5mm allows the laser beam to frilly cover the optical aperture of the microscope's objective lens, as discussed later in the

description.

The laser beam then exits the beam expander 18 and passes through a Glan-Taylor polariser 20 with a high extinction ratio of io and becomes horizontally (p) polarised.

A polarising beam splitter cube 22 is optically aligned with the Glan-Taylor polariser 20 such as to receive the laser beam. The polarising beam splitter cube 22 has a distinction ratio of i03 and transmits 99.9% of the p-polarised component of the laser beam in a forward direction and reflects 99.9% of the vertically (s) polarised component in a sideways direction, an example of which is shown in Figure 3.

An objective lens 24 is disposed in optical alignment with the splitter cube 22. The objective lens 24 has an NA==0,95 and focuses the laser beam 14 to a sub-micron diameter spot on a surface 26 of a sample 28 disposed on a sample holder 30.

The laser beam 14 is reflected from the sample surface 26 and returns along the same optical path through the objective lens 24 and enters the beam splitter cube 22.

The Kerr rotation signal, which is s-polarised and comprises approximately two percent of the total beam intensity, is then almost totally reflected to the side by the polarising beam splitter cube 22 and directed towards an optical bridge detector 32, along with only 0.1% of the rest of the beam (p-polarised). Therefore, more than 90% of the detected reflected beam intensity is from the Kerr rotation with less than 10% background. The advantage of using a polarising beam splitter cube rather than a normal beam splitter cube is that the polarising beam splitter cube selectively reflects the total Ken rotation signal with little background from the p-polarised component. By comparison, a normal beam-splitter would reflect half of both polarisations equally, resulting in half the amplitude of the total Ken rotation on top of a large background.

Following the beam splitter, an optical bridge detector 32 is used to analyse the Ken rotation and subtract the background. An iris 34 is placed in front of the detector 32 and is used to cover half of the reflected beam for longitudinal MOKE measurements and is set fully open for polar MOKE measurements.

The theoretical limit for the minimum diameter, D, of the focused laser passing through a given numerical aperture, NA, with a working wavelength, 2, is set by the diffraction limit and the minimum diameter is given by (Eq. 1). n,rNA

In practice the highest available NA for a lens working in air is 0.95 and this waJ#ôn for the MOKE setup. As discussed previously, the beam diameter is expanded to 5mm since this is the actual optical aperture of the objective lens. By using (Eq. 1) the minimum, focused laser spot, diameter theoretically achievable using the system described here is calculated to be around 42Orjm.

In order to be able to position the laser spot precisely on a given part of the sample under study, the microscope system is implemented into the focused MOKE as shown in Fig. 2.

This is achieved using a lamp 36 and CCD camera 38 connected to a computer 40. The light from the lamp 36 is emitted in a cone and this is collimated by a lens 42 and coupled into the optical path by means of a first pellicle beam-splitter 44. The first pellicle beam-splitter transmits 90% of the light and 10% of the light is reflected to the side and is chosen to minimise the loss of laser beam intensity. By using a second pellicle beam-splitter 46, placed just before the microscope objective lens 24, 10% of the light reflected from the sample is now reflected to the side of the beam-splitter 46, as shown in Fig. 3. A 13mm focal length lens 48 is disposed after the beam-splitter, approximately 13mm away from the sample 26, which forms an image on the CCD camera 38. The CCD camera is disposed at a distance of 13mm from the lens. The image captured by the CCD camera is transmitted and processed into the computer. A tube 50 is placed in front of the CCD camera 38 to minimise the amount of stray light shining on the CCD camera 38. This setup has the advantage of imaging both the surface 26 of the sample 28 and the focused laser spot simultaneously, allowing for precise positioning of the laser spot 68. An image of a lRm wide wire 70, with a SOnm bowtie constriction 72, obtained using this system is shown in Fig 4. The laser spot is placed on the arm of the wire 70 and the bowtie constriction 72 is clearly visible. Comparing the spot 68 diameter against a structure of known dimensions allows for the focused laser spot diameter to be estimated. Here, it is estimated to be under 1 jim, which is sufficiently small for the samples studied in this An electromagnet 60, having first and second poles, 62 and 64, respectively, is disposed about the sample holder 30 for supplying a magnetic field on the surface 26 of the sample 28.

A vacuum pump 66 acts to provide a vacuum to the sample holder 30 to secure the sample 28 in place on the sample holder.

Optical Bridge Detector: Referring also to Figure 5, the optical bridge detector 32 is operable to split the incoming laser light 14 into two orthogonal polarisations and then to measure the light intensity of each component with a pair of amplified photodiodes. This provides for a system which is very sensitive to small rotations in the polarisation of the incoming laser beam whilst rejecting changes in intensity of the laser beam -which can be caused by drifting of the laser source. The detector 32 comprises of a Wollaston prism 52, a lens 54 such as a UV- silica lens with a focal length of 10mm, a photo-detector 56, such as a Hamamatsu two-segment photo-detector, and a dual low-noise pre-amplifier circuit 57 as schematically illustrated in Fig 6.

The Wollaston prism 52 is used to separate the two polarisations; they emerge, diverging slightly, and are focussed by the UV-silica lens 54 as two separate spots onto the two-segment, A and B, photo-diode 56. The lens 54 focal length and divergence angle of the prism are chosen to match the separation needed for the photo-detector 56. The electronic amplifiers for the two channels are based on OPA124 op-amps as detailed in Fig 6.

The output of the two channels, A and B, of the photo-detector 56, can be obtained from equations 2 and 3. (Eq. 2) A=Kij9cos2(w_.) B=KLQsin2(q,_i) (Eq.3) Where p is the polarisation rotation introduced at the sample in the reflected beam, 4 is the incident light intensity, and K is the magneto-optical constant of the sample. Another very useful method of obtaining the Kerr rotation angle is by using the formula (A-B)/(A+B) as this approximates 2ço since 49 is very close to zero, as shown in (Eq. 4).

A-B K[cos2(c7_f)_sin2(Q_!)] A+B. 2 12 (Eq.4) K-[cos (q---)+sm (c'---)j Eq. 4 is used to obtain the Ken rotation angle by reading the voltage from channels A and B into the computer. The design of the bridge detector 32 makes it insensitive to the laser intensity drift. Combined with the usage of the polarising beam-splitter, the detector provides excellent signal to noise ratio for Kerr rotation detection and is capable of capturing a magnetic hysteresis loop from a nanowire over a single magnetic field sweep as shown in Fig 8.

Controlling Lab View Program: The controlling software for the MOKE measurements is programmed and executed using Labview which is run on computer 40. A sereenshot of the controlling panel is shown in Fig. 7. The main function of the program is to output a given voltage to the 1000W amplifier 58, which is used to set the field in the electromagnet 60, having first and second poles, 62 and 64, respectively, in a user-specified order, and for each field value, measure the voltage from the A and B photodiodes. These voltage values are used to calculate the Ken rotation angle using (Eq. 4) as explained above. The Kerr rotation readings are processed by computer 40 and displayed on a screen as a function of the applied magnetic field and the data is saved at the end of the measurement.

The electromagnet 60 has a maximum magnetic field strength of 2.SkOe and the program uses a magnet calibration file, selected in the Magnet Calibration box, to determine the voltage it needs to output to the 1000W amplifier 58 in order to apply a desired field value. The calibration files are obtained by measuring the field -using a Hall probe -as a flmction of the applied voltage. A magnet calibration loop can be seen in Fig. 7. Thus, if a certain field is required, the program looks up the output voltage in the magnet calibration file. The format of the magnet calibration file consists of three columns. The first column contains the magnetic field values and the second and third columns contain the voltage values which would result in the field value on a given row.

The first colunm contains the voltage values corresponding to the magnetic field values when the field is incremented while the third column contains the voltage values corresponding to the magnetic field values when the field is decremented. This is necessary because the nonzero coercivity of the magnet poles means that different voltage values are required to obtain a given field value when incrementing and decrementing the field. Typically the magnetic field is cycled between two magnetic field values of equal magnitude and to decrease the difference between the desired field and the actual output field, the program takes into account whether the field is currently being increased or decreased by reading the appropriate column of the magnet calibration file.

The accuracy of the actual output field was determined to be within ±1 Oe of the desired

field value under this system.

The magnetic field is cycled according to the data contained in the measurement program file selected in the MOKE Measurement Program box. The format of this file consists of a user-defined number of rows and each row contains, in order, the start field value in Oe, end field value in Oe, number of measurement points between the start and end field values and finally the sealing time after the voltage is output across the electromagnet in milliseconds. Typically this should be set to 1 Oms to allow for the magnetic field to settle. Before the measurement starts, the field is ramped up to the saturation field of the electromagnet, -2.SkOe, and then cycled back to the start field of the measurement. This is implemented to ensure that a consistent field value is reached every time, independent of the state the magnet was left in after the previous measurement which could vary, for example, if the measurement was stopped suddenly before it finished. After the measurement ends, the field value is cycled back to zero.

The complete measurement cycle can be repeated a number of times; specified in the Loops box. The current measurement number is displayed in the Loop Number box and the current field value is displayed in the Applied Field box. The measurement loops are saved in a file specified in the Save Measurement box and this consists of two columns. The first column contains the applied field and the second contains the corresponding Kerr measurement. The measured data is also displayed in the Real-Time MOKE Graph plot while the Averaged MOKE Graph plot contains a running average of all the loops measured from the start. For convenience, the voltage measured from photodetector 56, A or B, may be plotted in the first graph on the upper left in Fig. 7 -selected with a pull-down menu.

Finally, the program has the ability to discard measured loops where voltage drift has affected the measurement. This is done by calculating the voltage difference between the ends of the loop and if this value exceeds the value specified in the Rejection Voltage box then the loop is discarded. In these instances, the Last Loop Rejected indicator lights and the difference voltage is displayed in the Voltage D?[ference box. The program uses a given number of points, specified in the Averaging Points box, from the start and end of the loop to calculate this voltage difference. The user also has a control to manually reject loops by checking the Manual Rejection box; useful if, for example, a large spike occurs in the measurement. A set of single-shot MOKE loops from the nano-constriction 72 of a Permalloy wire 70, acquired using this system, are shown in Fig 8, indicating four different switching behaviours with equal statistics.

Focus MOKE with a spin-polarised current source: In ferromagnetic metals, the interaction between itinerant electron spins and wall spins can excite a wall spin precession, giving rise to domain wall motion due to the intra-atomic s-d exchange torque exerted by the angular momentum of the conduction electrons [L. Berger, Phys. Rev. B, 33, 1572 (1986)1. In order to detect the direction of spin-polarised current-induced domain wall displacement in a nanostructure, a second embodiment of the focus MOKE is presented with the capacity to apply spin-polarised current into the nanostructure. Referring also to Figure 9, a Keithley power supply 61 is used to output a voltage or current and measure a current or voltage value. By controlling the Keithley power supply 61 with the Labview program 40, focused MOKE measurements are performed with the laser spot 68 focused on one of the wire arms 70, either side of the constriction 72. The sample to be measured is mounted in a 44-pin ceramic chip where the nanowires are wire-bonded to the chip pins. This chip is then held in a custom made chip holder between the poles 62 and 64 of the electromagnet 61. A diagram of the setup is shown in Fig. 9.

The measurement procedure is as follows: After nucleation of a domain wall at the constriction 72, detected by monitoring the MOKE signal, the current is applied between the leads at the ends of the nanowire 70 and increased in 1 OLIA steps in either the positive or the negative direction. During this progression, the MOKE signal is measured with the laser spot 68 on either the left or right arm. This procedure is repeated for the four combinations of current direction and laser spot position. Since it is the longitudinal MOKE effect being measured, an increase in the MOKE signal represents magnetisation switching from left to right and the inverse for a decrease in the MOKE signal.

Comparison of the four different MOKE measurements allows the direction of domain wall displacement to be obtained. An example of such a MOKE measurement is shown in Fig. 10, with respect to the laser spot 68 and wire 70, having constriction 72. Figure 10 shows graphs of focussed MOKE versus current measurements at zero field: (a) negative current direction laser spot on left arm; (b) positive current direction laser spot on left arm; (c) positive current direction laser spot on right arm; and (d) negative current direction laser spot on right arm. For current-induced domain wall movement at zero field, the direction of domain wall displacement is always in the direction of the current carriers. n

Claims (1)

1. <claim-text>CLAIMS: 1. A microscope system comprising means operable to generate a magneto-optical Kerr effect (MOKE) on a surface and means for detecting longitudinal and polar MOKE from a said surface.</claim-text> <claim-text>2. A microscope system as claimed in claim 1, comprising means for obtaining single-shot measurements of magnetic-field-driven magnetisation switching and spin-polarised, current-driven, magnetic domain wall motion in nanostructures.</claim-text> <claim-text>3. A microscope system as claimed in claim 1 or 2, comprising a submicron probe.</claim-text> <claim-text>4. A microscope system as claimed in claim 3, wherein the submicron probe is generated by a laser.</claim-text> <claim-text>5. A microscope system as claimed in any of the preceding claims, comprising at least one of a spin-polarised current source and a magnetic field source.</claim-text> <claim-text>6. A microscope system as claimed in any of the preceding claims comprising an optical bridge.</claim-text> <claim-text>7. A microscope system as claimed in any of the preceding claims, comprising a polarised beam splitter.</claim-text> <claim-text>8. A method of microscopy comprising generating a magneto-optical Kerr effect (MORE) on a surface detecting longitudinal and polar MOKE from a said surface.</claim-text> <claim-text>9. A method as claimed in claim 8, comprising generating a magnetic field on the surface of a said sample, probing the said surface with a sub-micron probe and detecting the longitudinal and polar MOKE from the said surface.</claim-text> <claim-text>10. A method as claimed in claim 9, wherein the sub-micron probe is generated from a laser.</claim-text> <claim-text>11. Computer readable medium carrying a computer program comprising computer readable instructions adapted to cause a computer to carry out a method according to claims to 10.</claim-text> <claim-text>12. A computer comprising a computer readable medium as claimed in claim 11.</claim-text>