A Plasmonic Antenna for Magneto-Optical Imaging at the Deep Nanoscale

Magnetic data storage systems, such as hard disk drives, are constructed from nanoscale magnetic elements. The disk drive industry continually seeks to increase data storage capacity and speed of access. Data is represented in binary format (1s and 0s) by the orientation of a tiny bar magnet (up or down). Today each 'bit' is less than 50 nm long and so the critical features of the read/write transducers must be of comparable size. The time taken to read or write each bit is ~1 ns and becoming shorter. New methods are needed to observe and understand how nanomagnets change their state so that device performance can be improved.

Time resolved scanning Kerr microscopy (TRSKM) is the most powerful tool with which to study magnetization dynamics on fs through to ns timescales. A fs laser beam is focused onto and scanned across the surface of the sample in order to construct time resolved magnetic images. The TRSKM in Exeter has internationally leading performance but its spatial resolution is limited to 3/4 of the optical wavelength by the diffraction-limited focused spot size (300 or 600 nm), which is an inherent property of the wave nature of light. We propose to develop a plasmonic antenna that will be placed between the focusing lens and the sample so as to produce a much smaller near-field optical spot and hence greatly increased spatial resolution.

Light incident upon a metallic surface forces electrons into oscillation. Plasmonics exploits artificial structure to control the electron motion and, in the present case, to enhance the electric field within a small region of space. For example, one antenna design will be reminiscent of the bulls eye in a dart board. A circular grating structure milled into a thin gold film will capture light and channel energy into a hole at its centre. The hole will resonate like an organ pipe, producing an intense electric field at the end opposite to the grating, close to where the sample will be placed. The sample will modify the resonance of the hole and modify the character of the light reradiated by the grating, which will be detected within the TRSKM. For the antenna to be sensitive to the sample magnetization it must possess an additional novel feature: it must absorb and reradiate light of different polarization with equal efficiency. This will be achieved by introducing an appropriate arrangement of slits into the sides of the hole to control its resonant modes.

Focused ion beam milling (FIB) will be used to fabricate antennae and monolithic sample/antenna stacks on planar substrates for optical testing. However, the antenna must be formed on a sharp tip for scanning across the sample surface within the TRSKM. We will fabricate gold tips by depositing gold into a pyramidal-shaped pit in a silicon wafer. FIB milling may be used to define a grating in the gold, before resin is used to fill the remaining volume. The Au and resin will then be peeled off the wafer and FIB milling used to define the hole in the gold at the apex of the pyramid. Finally the tip will be attached to the cantilever arm of an atomic force microscope, which will control the height of the tip above the sample.

The tip antenna will be used in two exemplar studies. Time resolved images will be obtained from the pole pieces of a partially-built hard disk writer structure. New information will be obtained about how magnetic flux propagates within the nanoscale constriction at the pole tip. The magnetization dynamics excited in nanoscale magnetic elements by the spin transfer torque effect will also be explored. Electrons carry both charge and spin angular momentum and the injection of electrons with net angular momentum generates a torque that can change the magnetic state of a suitably designed nanoscale element. We will study novel structures that allow optical access to the element and hence provide new information about both the origin and effect of the torque.

Improved near-field sensing with polarization contrast will benefit all those working on polarizing microscopy, both in academia and industry. Commercial scanning near field optical microscope (SNOM) systems are already available, but robust polarization contrast will greatly enhance these products, and our research will deliver the silicon processing technique by which suitable tips can be fabricated. One large market for a SNOM with proven magnetic sensitivity is wafer level magnetic metrology within the magnetic disk drive industry. For example, Seagate Technology own in excess of fifty $200k scanning magneto-optic systems for wafer metrology, with many more in similar organizations and labs globally.

Both Exeter and Queens University Belfast (QUB) have strong track records of collaboration with industry, and the close relationship between Seagate and Exeter and particularly QUB provides a unique avenue for exposure, industrial impact and benefit for the UK economy. Seagate's Springtown operation in Northern Ireland manufactures up to 400M recording heads per annum, some 25% of the world supply, and contributes some £100M p.a. to the UK economy. In 2010 Seagate made a further £60M investment in R&D operations at Springtown and established the ANSIN centre at QUB, which supports the development of high magnetic moment materials, materials for reader/writer shields, barriers for tunnel magnetoresistance (TMR), and the integration of light delivery in Heat Assisted Magnetic Recording (HAMR).

HAMR is expected to appear as commercial product in the next 5 years. The resonant aperture developed within our project has the potential to provide greater heating power at the sample for the same far field input power, reducing the heating of the antenna itself and solving one of the principal technical problems for HAMR. Industry roadmaps suggest that on longer timescales HAMR will be integrated with bit-patterned media, which will require many-$Bn investment in manufacture. The use of discrete sub-40nm bits presents a major technological challenge in read/write synchronization as the transit time of the head above the bit will be just 10s of ps, and so there is significant incentive to develop alternative faster read/write processes. Magneto-optical sensing was used in magneto-optical recording before being largely overtaken by optical phase change recording (CD/DVD). Given that HAMR may lead to the introduction of hybrid optical hard disk drives in the near future, magneto-optical sensing may find favor again if adequate spatial resolution and signal to noise ratio can be demonstrated. Therefore the sensing capability to be developed within our project significantly extends HAMR related research that Seagate Technology is engaged in either internally or via other projects such as that in ANSIN.

Seagate Technology already fund Exeter to perform time resolved scanning Kerr microscopy (TRSKM) measurements on wafers of partially built writer structures. While this is a very effective tool for testing yoke design, recording track widths and the width of the neck of the writer already lie within the sub-micron regime, and so the neck cannot be imaged with the existing instrument. The enhanced spatial resolution delivered by the plasmonic antenna will allow the magnetic flux distribution within the neck and its rise time to be optimized, leading to higher storage densities and bit rates. The second proposed study is of more basic scientific interest, but aims to deliver improved understanding of spin currents and spin transfer torque that will aid the development of 2nd generation MRAM, spin transfer oscillators, and non-volatile low-power logic.

Finally, the skills and expertise in nanofabrication, optical and high frequency metrology, and modeling gained by the 4 young researchers in the project will make them highly attractive for employment within the magnetic recording, photonics, or other advanced materials industries.