We demonstrate the lightning-rod resonance of a lollipop near-field transducer integrated in magnetic writer for heat-assisted magnetic recording by collecting the two-photon excited photoluminescence (TPL) signal when excited by a pulsed femto-second fiber laser tuned to the desired mode resonance. The lollipop transducer consists of a round disk and a protruding peg to take advantage of the lightning-rod effect. It is found that the TPL signal is extremely sensitive to the peg length where even a 3-5 nm deviation from the optimal peg length halves the TPL signal. This method conveniently quantifies the optical performance of an NFT device in situ as a function of geometry with a resolution of better than the light wavelength (λ) divided by 200.
© 2017 Optical Society of America
The imminent commercialization of plasmonic devices signifies considerable advancements in the areas of nanoscale design, fabrication, and characterization. Heat assisted magnetic recording (HAMR)  is one of these emerging technologies enabled by the application of plasmonics. An optical near-field transducer (NFT), such as an optical antenna [1–5] or a sub-wavelength metallic aperture [6–11], is used to locally heat a storage medium above its Curie temperature by delivering optical energy to a spot much smaller than the diffraction limit. By reducing the magneto-crystalline anisotropy Ku at elevated temperatures, the currently available magnetic field can switch the magnetization to form a data bit. High Ku magnetic materials allow stable small magnetic grains against thermal fluctuations at ambient temperature, leading to increased data storage densities.
Functional optical nanoantennas typically exploit the excitations of localized surface plasmon (LSP) modes to concentrate far-field radiation to a local volume in the vicinity of the surface. If the antenna is favorably shaped and includes a shape singularity, such as a gap or a tip, further spatial confinement in the field can be achieved. For the practical application of a NFT to HAMR, there are five key factors to be considered: (1) NFT efficiency, (2) size of the optical spot, (3) NFT heat dissipation, (4) integration of the photonic and magnetic recording subsystems, and (5) process tolerances. Careful selection of design parameters, such as operation wavelength, shape, material, and dimensions of the antenna are necessary for not only optimum impedance matching to a far-field excitation source, but also maximizing recording performance.
Recently, Challener et al.  proposed a lollipop patch antenna integrated into a magnetic writer for HAMR, which demonstrated an areal data density of 1 + Tb/in2 in 2013 . The lollipop transducer is composed of a gold disk and a gold peg of adjustable length, which is used to optimize the LSP resonance given fabricational variations. The resonance of the protruding peg serves as a lightning-rod for confining light and ultimately determines the data track density and thermal gradients for writing sharp magnetic transitions. As is typical for most plasmonic nanostructures, the size is intimately linked to its performance.
While there are methods to accurately measure the size (e.g., electron microscopy), other process variations may affect the optical performance of the NFT, which make direct plasmonic verifications of a LSP resonance most useful within a high-volume manufacturing environment. Previously, cross-polarization in far-field radiation  and photo-thermal reflectance  techniques were developed to detect and optimize the NFT’s surface plasmon resonance. The cross-polarization technique measures the field component orthogonal to the primary incident field in the far-field by placing a polarizer in the transmitted beam path, and it was found that the signal mainly comes from the disk portion of the NFT. The photo-thermal technique measures the temperature-induced change in reflectance by a probe beam due to the relatively large absorption cross-section of the NFT on resonance. These methods perform equally well in characterizing NFT efficiency for the case of an isolated near-field transducer. However, the sensitivity of these techniques is much reduced for fully functional HAMR devices as they integrate numerous components, in particular, the sensitivity to the peg length. In addition to the NFT and its associated photonic subsystem, a tunneling magneto-resistive reader, write and return poles, microheaters, and electrical contacts among other elements must be included for magnetic recording functionality. For the cross-polarization, these extra components further contribute to the signal from the NFT, and for the photo-thermal reflectance, the aggressive heat sinking to the NFT leads to a lower temperature difference.
Since the primary LSP phenomena of the NFT, near-field intensity enhancement, is not readily observable in conventional far-field measurements, further application of known plasmonic responses is needed to realistically characterize these nanostructures. One salient optical property of noble metals is their visible photoluminescence, first reported by Mooradian in 1969 . Upon illumination with femto-second or pico-second laser pulses, a detectable two-photon induced luminescence from gold nanostructures can be measured, and has been studied quite extensively for a variety of sizes and shapes [16–19]. Although there are several exciting descriptions of the physical origins of nonlinear emission from gold nanostructures, existing research recognizes their association with the high field enhancements of LSP resonances [20–22].
Here, we employ two-photon induced luminescence (TPL) to demonstrate the lightning-rod resonance of peg in a lollipop transducer. It is found that the TPL signal is extremely sensitive to peg length and that even a 5-nm deviation from the optimum could drop the TPL signal by half. This is consistent with the observation in the recording performance. To the best of our knowledge, this is the first time to report such sensitivity in lightning-rod resonance
2. Integrated device and experimental setup
Figure 1 shows a schematic of the pertinent optical elements within our HAMR head and a cross-section transmission electron microscopy (TEM) image of the lollipop transducer.
As an example, if the NFT is resonant at a light wavelength λ = 830 nm, the optimal disk size is ~250 nm in diameter, and the peg is 40 nm in width, 30 nm in thickness, and is optimally 20 – 25 nm in length. The NFT is illuminated by a transverse-electric (TE) polarized beam of light propagating in the planar waveguide and focused by a planar solid immersion mirror  (PSIM) etched in the waveguide. There is a π–phase shift in the wavefront between the TE modes on either side of the PSIM. When the beam combines at the PSIM focus, where the lollipop NFT is located, the net optical field has one central spot polarized longitudinally and two sidelobes polarized transversely, which is analogous to the focus of a radially polarized beam. The sidelobes interact with the lollipop’s disk to excite the quadrupole resonance along the circumference of the disk and funnels charges into the peg. The longitudinal field interacts with the lollipop's peg, enhancing and confining the electric field at the end of the peg through the lightning-rod effect. The NFT is placed at 20 nm away from the 125-nm thick Ta2O5 core and is inside the Al2O3 cladding layer. The NFT is connected with a FeCo magnetic pole through a heat sink also composed of gold.
Figure 2(a) shows the system for measuring the TPL signal. The excitation source is a mode-locked femto-second fiber laser (pulse width < 120 fs, repetition rate = 75 MHz, center wavelength λ = 805 nm). The incident power is controlled by a broadband Fresnel beam sampler and an absorptive neutral density filter, and is monitored using a 50:50 non-polarizing beam splitter with a calibrated photodetector. The incident polarization is set using an achromatic zero-order half-wavelength wave plate, and an aspheric lens (NA = 0.25) focuses the beam onto the entrance of the waveguide. It should be noted that these elements are for efficiently coupling light from free space into the waveguide, whereas the light incident on the NFT is further processed as discussed above in Fig. 1. The transmitted light from the device is collected by an objective (100x, NA = 0.90) and is relayed by two lenses in a 4-f configuration such that the light exiting surface is imaged onto several photodetectors. Light of spectral wavelengths below 350 nm is inherently attenuated due to the lens materials. A dichroic beam splitter (685-nm edge) is used to filter the long-wavelength light onto a Si photodetector (PD), and reflect the short-wavelength light onto an EM-CCD camera or a photo-multiplier tube (PMT). Additional short-pass filters (680-nm cutoff) placed on the reflection side of the dichroic further reject the excitation wavelength with an optical density of 10−14. In total, the spectral range of the photoluminescence is limited from 350 nm to 680 nm for the EM-CCD and PMT, while the PD is sensitive to wavelengths above 685 nm. Finally, the PD and PMT signals are measured with a lock-in amplifier, and a mechanical chopper placed in the excitation path serves as the modulation source.
Figure 2(b) shows the photoluminescence signal detected by the PMT as a function of transmission on a log-log scale. The transmitted intensity at the fundamental wavelength (measured by the PD) is used instead because it is more representative of the intensity that is actually incident onto the NFT. It is evident that the relationship is nearly linear, and a simple linear fit of the log-transformed data shows that the signal is proportional to (transmission)2.01 over the measured power range (< 300 µW, which corresponds to a transmission of 38 mV.). The quadratic dependence of the PMT signal on the transmission intensity confirms the nature of the measured photoluminescence signal, which is indeed induced by a two-photon excitation process. The data in the regime below 14 mV seems slightly deviated from the line of fit as there are some fluctuations in the detected signal. Spectroscopic measurements, which are not shown here, revealed that the collected photons consist of the second harmonics and a broadband spectral component.
3. Lightning rod resonance
Figure 3 shows the optical CCD images of the transmitted light. In Fig. 3(a), the short-pass filter is removed from the collection path and the dichroic is replaced with a mirror. Here, the CCD image shows the full transmission, which is dominated by light of the excitation wavelength. The fringe-like pattern seen is due to the high numerical aperture focusing of the PSIM with the π-phase shifted wavefront in the incident beam, and the confinement of the waveguide. In Figs. 3(b)-3(d), the short-pass filters are reinserted and only the photoluminescence is observed. Because the emission region is well below the resolvable limit of the objective, a single bright spot is observed in the images with NFTs as shown in Figs. 3(c) and 3(d). Although the white-light image can't be shown, there are multiple micron-sized magnetic recording elements that are used to easily locate the NFT, and we observe that the bright spot of the TPL emission is in the expected location. There is no detectable emission without an NFT as seen in Fig. 3(b), which confirms that the observed photoluminescence signal is purely from a NFT. The two devices displayed in Figs. 3(c) and 3(d) are different in the peg length. The device shown in Fig. 3(c) presents a much brighter TPL spot.
Figure 4(a) shows the TPL intensity versus peg length for ~2000 devices. The measurements are performed on several row bars from a single wafer, where each row bar consists of over 60 HAMR devices. With a chemical-mechanical polishing (CMP) process widely used in the hard drive industry , the nominally flat row bar is polished at an angle and ideally results in a linearly varying peg length for each NFT across the row bar. Due to the design of the HAMR head, the NFT is entirely surrounded by opaque elements necessary for magnetic recording (see Fig. 1(b)) and effectively obstructs a direct line of sight for easily measuring the true NFT peg length (e.g., by a scanning-electron microscope). For the data shown here, the peg length is determined by calibrating feedback data from the CMP process with a series of cross-section transmission-electron-microscope (TEM) measurements (see Fig. 1(c)). Although the calibration procedure is a source of error for absolute peg length, the relative peg lengths across a row bar are well defined by the CMP process. For clarity in the graph, the calibrated peg lengths for each NFT are rounded to the nearest nanometer, and the average of 10 to 60 single-NFT measurements is shown with error bars representing a 95% confidence interval of the mean. A negative peg length indicates that the peg has been entirely polished away, leaving only the disk portion of the NFT.
For either very long or very short peg lengths, the TPL intensity drops into the background and is only detectable in ~20-nm range of peg length, emphasizing the role of the lightning-rod effect in the lollipop NFT. As discussed earlier, the resonance of a lollipop NFT is determined by its disk size and in particular, the peg length. From Fig. 4(a), it is evident that the TPL signal is critically dependent on the peg length, demonstrating the lightning-rod resonance of the protruding peg, which is substantially enhanced from the resonance of a disc-only device (i.e., peg length = 0 nm). Computer simulation by integrating the square of electric field strength over the NFT shows a similar resonant behavior. Resonance occurs at peg length ≈25 nm, and a small deviation of 3-5 nm from the resonant peg length is seen to reduce the TPL signal by half. Note that the center wavelength of the light (λ) used in our experiment is 805 nm. This means that we obtain a resolution better than λ/200 by the lightning-rod effect.
The PSIM focusing field excites a quadrupole mode on the disk and a longitudinal mode on the peg. The quadrupole mode exhibits significantly different far-field radiation from the longitudinal one. Figure 4(b) shows the TPL far-field radiation pattern in polar plot. The horizontal axis is parallel to the polarization of excitation. The radiation pattern presents two lobes with asymmetry and the two lobes are also rotated from the horizontal axis, which is caused by the interference in the far-field radiation between the quadrupole mode from the disk and dipole mode from the peg . This means that both the disk and peg contribute to the observed TPL signal. The strong sensitivity of TPL signal to peg length indicates that the peg length changes the NFT resonance substantially.
The lightning-rod resonance of a lollipop NFT in a fully integrated device for HAMR has been demonstrated by collecting the TPL signal as a function of the NFT peg size. It is found that the TPL signal mainly comes from the peg and is highly sensitive to the peg length. A 3-5 nm deviation from the optimal peg length halves the TPL signal. This method conveniently quantifies the optical performance of a NFT device and has immediate application in a high-volume manufacturing environment such as HAMR, where the NFT is fabricated early in the work flow and in-line screening is much more economical than end-of-line testing. Once a specification is established, the lightning-rod resonance can very easily distinguish peg lengths at a resolution of better than λ/200, well below the diffraction limit of light.
References and links
1. W. A. Challener, C. Peng, A. V. Itagi, D. Karns, W. Peng, Y. Peng, X. Yang, X. Zhu, N. J. Gokemeijer, Y.-T. Hsia, G. Ju, R. E. Rottmayer, M. A. Seigler, and E. C. Gage, “Heat-assisted magnetic recording by a near-field transducer with efficient optical energy transfer,” Nat. Photonics 3(5), 303 (2009). [CrossRef]
2. P. J. Schuck, D. P. Fromm, A. Sundaramurthy, G. S. Kino, and W. E. Moerner, “Improving the mismatch between light and nanoscale objects with gold bowtie nanoantennas,” Phys. Rev. Lett. 94(1), 017402 (2005). [CrossRef] [PubMed]
3. S. P. Powell, E. J. Black, T. E. Schlesinger, and J. A. Bain, “The influence of media optical properties on the efficiency of optical power delivery for heat assisted magnetic recording,” J. Appl. Phys. 109, 07B775 (2011).
4. T. Matsumoto, F. Akagi, M. Mochizuki, H. Miyamoto, and B. Stipe, “Integrated head design using a nanobeak antenna for thermally assisted magnetic recording,” Opt. Express 20(17), 18946–18954 (2012). [CrossRef] [PubMed]
5. C. Peng, “Efficient excitation of a monopole optical transducer for near-field recording,” J. Appl. Phys. 112(4), 043108 (2012). [CrossRef]
6. E. Betzig, J. K. Trautman, R. Wolfe, E. M. Gyorgy, P. L. Finn, M. H. Kryder, and C. H. Chang, “Near-field magneto-optics and high density data storage,” Appl. Phys. Lett. 61(2), 142–144 (1992). [CrossRef]
7. F. Chen, A. Itagi, J. A. Bain, D. D. Stancil, T. E. Schlesinger, L. Stebounova, G. C. Walker, and B. B. Akhremitchev, “Imaging of optical field confinement in ridge waveguides fabricated on very-small-aperture laser,” Appl. Phys. Lett. 83(16), 3245–3247 (2003). [CrossRef]
8. E. X. Jin and X. Xu, “Obtaining super resolution light spot using surface plasmon assisted sharp ridge nanoaperture,” Appl. Phys. Lett. 86(11), 111106 (2005). [CrossRef]
9. C. Peng, E. X. Jin, T. W. Clinton, and M. A. Seigler, “Cutoff wavelength of ridge waveguide near field transducer for disk data storage,” Opt. Express 16(20), 16043–16051 (2008). [CrossRef] [PubMed]
10. B. C. Stipe, T. C. Strand, C. C. Poon, H. Balamane, T. D. Boone, J. A. Katine, J.-L. Li, V. Rawat, H. Nemoto, A. Hirotsune, O. Hellwig, R. Ruiz, E. Dobisz, D. S. Kercher, N. Robertson, T. R. Albrecht, and B. D. Terris, “Magnetic recording at 1.5 Pb m−2 using an integrated plasmonic antenna,” Nat. Photonics 4(7), 484–488 (2010). [CrossRef]
11. K. Şendur, C. Peng, and W. Challener, “Near-field radiation from a ridge waveguide transducer in the vicinity of a solid immersion lens,” Phys. Rev. Lett. 94(4), 043901 (2005). [CrossRef] [PubMed]
12. A. Q. Wu, Y. Kubota, T. Klemmer, T. Rausch, C. Peng, Y. Peng, D. Karns, X. Zhu, Y. Ding, E. K. C. Chang, Y. Zhao, H. Zhou, K. Gao, J.-U. Thiele, M. Seigler, G. Ju, and E. Gage, “HAMR areal density demonstration of 1+ Tbpsi on spinstand,” IEEE Trans. Magn. 49(2), 779–782 (2013). [CrossRef]
13. C. Peng, “Cross-polarization detecting surface-plasmon resonance of near-field transducer,” Appl. Phys. Lett. 104(6), 061114 (2014). [CrossRef]
14. C. Peng, “Surface-plasmon resonance of a planar lollipop near-field transducer,” Appl. Phys. Lett. 94(17), 10–13 (2009). [CrossRef]
15. A. Mooradian, “Photoluminescence of metals,” Phys. Rev. Lett. 22(5), 185–187 (1969). [CrossRef]
16. K. D. Ko, A. Kumar, K. H. Fung, R. Ambekar, G. L. Liu, N. X. Fang, and K. C. Toussaint Jr., “Nonlinear optical response from arrays of Au bowtie nanoantennas,” Nano Lett. 11(1), 61–65 (2011). [CrossRef] [PubMed]
17. A. Bouhelier, R. Bachelot, G. Lerondel, S. Kostcheev, P. Royer, and G. P. Wiederrecht, “Surface plasmon characteristics of tunable photoluminescence in single gold nanorods,” Phys. Rev. Lett. 95(26), 267405 (2005). [CrossRef] [PubMed]
18. J. Beermann, S. M. Novikov, T. Søndergaard, A. Boltasseva, and S. I. Bozhevolnyi, “Two-photon mapping of localized field enhancements in thin nanostrip antennas,” Opt. Express 16(22), 17302–17309 (2008). [CrossRef] [PubMed]
19. R. A. Farrer, F. L. Butterfield, V. W. Chen, and J. T. Fourkas, “Highly efficient multiphoton-absorption-induced luminescence from gold nanoparticles,” Nano Lett. 5(6), 1139–1142 (2005). [CrossRef] [PubMed]
21. T. Haug, P. Klemm, S. Bange, and J. M. Lupton, “Hot-electron intraband luminescence from single hot spots in noble-metal nanoparticle films,” Phys. Rev. Lett. 115(6), 067403 (2015). [CrossRef] [PubMed]
22. A. Bouhelier, M. R. Beversluis, and L. Novotny, “Characterization of nanoplasmonic structures by locally excited photoluminescence,” Appl. Phys. Lett. 83(24), 5041–5043 (2003). [CrossRef]
23. W. Challener, C. Mihalcea, C. Peng, and K. Pelhos, “Miniature planar solid immersion mirror with focused spot less than a quarter wavelength,” Opt. Express 13(18), 7189–7197 (2005). [CrossRef] [PubMed]
24. M. Jiang, S. Hao, and R. Komanduri, “On the advanced lapping process in the precision finishing of thin-film magnetic recording heads for rigid disc drives,” Appl. Phys., A Mater. Sci. Process. 77(7), 923–932 (2003). [CrossRef]
25. T. D. James, Z. Q. Teo, D. E. Gomez, T. J. Davis, and A. Roberts, “The plasmonic J-pole antenna,” Appl. Phys. Lett. 102(3), 033106 (2013). [CrossRef]