Abstract

While thin-film magnetic cores provide significant enhancement to integrated passive circuit elements, including inductors and transformers [1]–[10], often their operating frequency limits their applications to mainly power management [1]–[6]. The growing field of mobile electronics, however, demands higher inductances than achievable by air-core inductors [11], [12] and higher operating frequencies than previously achievable by magnetic materials. This work aims to meet the needs of radio frequency mobile applications, by providing high frequency inductors with large inductance enhancement due to a magnetic core, all in a small form factor. As illustrated by pervious works [7]–[10], the trade-off between inductance enhancement and high operating frequency is challenging to overcome. Although magnetic cores contribute high permeability for inductors, at the intrinsic ferromagnetic resonance (FMR) frequency of magnetic materials (typically between 1 and 2 GHz for large blanket films), the relative permeability drops to unity, thereby making the inductance enhancement due to the material negligible. Extending the operating frequency, however, by means of patterning the magnetic core and increasing the FMR frequency through shape anisotropy has the consequence of also decreasing the permeability. In this work, we analyze the most crucial considerations regarding material selection and the fabrication process as well as magnetic-core and device design in order to produce and inductor offering a magnetic enhancement over air core extending beyond 6 GHz, a low frequency inductance of 1 nH, and a peak quality factor of four at approximately 3 GHz [13]. Both air-core and magnetic-core solenoid inductors of the same topology, shown in Fig. 1, were fabricated and measured. First, $\mathrm {a}40 \mu \mathrm {m}$ thick polymer insulating layer (SU-8 2015) was formed on the silicon substrate to provide isolation of the inductor from the parasitic substrate. $3 \mu \mathrm {m}$ thick bottom inductor windings were electroplated starting from a thin Ti/Cu seed layer. Planarization and insulation was completed using two layers of the SU-8 2002 polymer to first fill the gaps between windings for planarization and then provide an additional $1.5 \mu \mathrm {m}$ insulation layer between the windings and the magnetic layer. The magnetic material Co 43 Fe 43 B 14 , selected its excellent high frequency properties [14], [15], was sputter deposited with laminations such that 63 nm CoFeB layers were alternated with 6 nm SiO 2 insulating layers to form a total of $1 \mu \mathrm {m}$ thick magnetic core. The core was then patterned into a narrow shape for high shape anisotropy and high FMR frequency, where laminations further improved broadband performance by reducing high frequency losses due to eddy currents. Finally, the vias and top inductor windings were again formed by an electroplating process. The comparison of measured inductance from the air-core and magnetic-core inductors of the same design provided a direct evaluation of the performance enhancement of the inductor due to the magnetic material. Fig. 2 shows the measured permeability of the blanket film as well as the calculated permeability of the patterned magnetic core used in the inductor. The shape anisotropy is seen to increase the bandwidth of the inductor by approximately 1 GHz. The inductance of the fabricated inductors, shown on the same graph, was approximately 1 nH for the magnetic-core inductor and 0.5 nH for the air-core version. Thus, incorporation of the magnetic core is seen to experimentally double the inductance of the air-core inductor well into the high frequency mobile electronics range. The decreasing inductance at higher frequencies and suppression of the ferromagnetic resonance peak is likely due to a combination of misalignment of the applied magnetic field during deposition of the cores [15], parasitic capacitances, and eddy currents along the length of the core. Nevertheless, these results show that, with continued optimizations, magnetic materials may indeed satisfy the needs of mobile electronics applications. Acknowledgement Research for this project was conducted with government support under FA9550–11-C-0028 and awarded by the Department of Defense, Air Force Office of Scientific Research, National Defense Science and Engineering Graduate (NDSEG) Fellowship, 32 CFR 168a. Part of this work was performed using the Stanford Nanofabrication Facility (SNF) and the Stanford Nano Shared Facilities (SNSF) at Stanford University.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call