IEEE TRANSACTIONS ON MAGNETICS, VOL. 50, NO. 11, NOVEMBER 2014 3303705 Metallic 10 nm Diameter Magnetic Sensors and Large-Scale Ordered Arrays Sang-Yeob Sung 1 , Mazin M. Maqableh 1 , Xiaobo Huang 1 , K. Sai Madhukar Reddy 2 , R. H. Victora 1 , and Bethanie J. H. Stadler 1,2 1 Department of Electrical and Computer Engineering, University of Minnesota, Minneapolis, MN 55455 USA 2 Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455 USA Metallic nanowires with low resistivity were grown inside insulating aluminum oxide matrices that contained very uniform columnar nanopores (10.6 + /1.7 nm diameters). These nanopores can be made with large-scale order (cm 2 ), which is desirable in applications such as hard drive read sensors and random access memories. The nanowires are grown by electrochemical deposition directly inside the alumina to avoid sidewall damage compared to nanostructures that are defined from films by lithographical patterning and etching. Specifically, trilayers of [Co(15 nm)/Cu(5 nm)/Co(10 nm)] were synthesized and measured to have 30  resistance and 19% magnetoresistance. These parameters are desirable for read head sensors, especially because the nanowires described here have 1:1 aspect ratios, and 10× smaller areas and 100× lower resistances than conventional read sensors based on lithographically produced magnetic tunnel junctions. A new nanostamping technique is introduced, in which linear stamps with ordered cm 2 areas are imprinted onto aluminum precursors to produce ordered nanoporous aluminum oxide upon anodization. These stamps are substantially less-time consuming and cheaper to make than dot type stamps, and the order enables closely spaced arrays of CPP-GMR sensors for one-pass 2-D recording and cross recording. Importantly, the GMR sensors are grown directly into aluminum oxide with 20 nm separation. Therefore, a relatively large pattern (30 × 100 nm) can be used to produce three 10 nm-diameter GMR sensors without roughening or redeposition on sidewalls. The sensors are also already embedded in alumina for subsequent device processing. Index Terms— 2-D recording, CPP-GMR, cross recording (CR), nanoimprinting, read sensor arrays. I. I NTRODUCTION L OW-resistivity in interconnects and devices can be chal- lenging at dimensions that are smaller than the mean free path of electrons in metals (λ ∼ 40 nm). The first of the five most critical challenges in the ITRS roadmap for <16 nm is to mitigate the size effects in interconnect structures [1]. It is recognized that film and sidewall roughness will adversely affect electron scattering, causing increased resistivity. This is also an issue in hard drive magnetic read sensors where tun- neling magnetoresistive structures are currently used. As these devices scale smaller, magnetoresistances (MRs) decrease and resistivities increase due to roughness, redeposition, and interlayer mixing at the sidewalls. These effects are detri- mental because resistance-area products need to continuously decrease to improve signal to noise ratios, lower RC time constants, and allow heat dissipation to reduce stochastic shot noise. In addition to requiring low resistivities, future read heads would greatly benefit from array configurations where adjacent tracks can be read concurrently with the track of interest to allow cancellation of intertrack interference (ITI). 2-D magnetic recording (2DMR) [2] and cross recording (CR) [3] have recently been shown to enable low signal-to-noise readout of high density recording media. 2DMR involves the use of signal processing with error correction to allow each bit to be nearly single grained [2]. Manuscript received March 15, 2014; revised April 28, 2014 and May 8, 2014; accepted May 9, 2014. Date of current version November 18, 2014. Corresponding author: B. J. H. Stadler (e-mail: stadler@umn.edu). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TMAG.2014.2325944 CR is a novel paradigm where the transverse field is read as information to achieve higher densities per bit [3]. Interestingly, the same structures proposed for high-density bit patterned media can also be used to produce the read sensor arrays. Specifically, the electrochemical deposition techniques discussed in this paper can be used to fill any insulating, ordered, porous structure with giant magnetoresistive layers. The nanowire sensors reviewed here are Co(15 nm)/ Cu(5 nm)/Co(10 nm) trilayered nanowires with 10 nm diam- eters and 20 nm interwire distances that have 30  resistance and 19% MR, Fig. l [4]. In addition to cross-track arrays, there is also an opportunity to fabricate down-track arrays, which could potentially be used in a manner similar to the proposed adjacent sensors of [5]. The read sensors have low aspect ratios which will be required for reading media that is written by future schemes such as heat- or microwave- assisted magnetic recording, and especially in combination with bit patterned media. Importantly, these all metal sensors have much smaller resistance-area products (RA = 2mum 2 ) than current tunneling MR sensors (0.1–1 um 2 ). Although the resistance-area (RA) product may appear to be too small according to some theoretical work, recall that one order of magnitude is due to the areas being 10× smaller than other recent metallic read sensors [6]. Although read sensors will require very small (5–30 nm) diameter nanowires, larger nanowires will be suitable to other applications (e.g., RAM), and most applications of current per- pendicular to plane giant MR (CPP-GMR) nanoarrays would benefit from perfect long-range order. Here, this order has been achieved in nanopore arrays by a novel line imprinting of the precursor Al prior to the anodization used to produce nanoporous aluminum oxide templates. Smaller nanowires will 0018-9464 © 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.