Spin-based information processing offers a promising alternative to traditional electronics, with magnonics, exploiting spin wave (SW) excitations in magnetic materials—enabling high-speed, non-volatile, and RF-compatible operation. However, practical implementation is hindered by integration challenges and material limitations. In this work, we address both aspects by combining advanced fabrication and hybrid integration strategies. Using soft-UV laser patterning of iron garnets (IG), we demonstrate direct crystallization of magnetic racetracks in non-magnetic matrices, paving the way for arbitrarily shaped magnonic circuits without physical etching. SW propagation over tens of micrometers with damping within one order of magnitude of state-of-theart YIG films is achieved, while Bi-doped YIG further enables laserinduced tailoring of magnetic anisotropy landscapes. To overcome substrate constraints, we integrate YIG with MEMS technology through the European MandMEMS consortium, realizing hybrid systems where MEMS actuation dynamically reconfigures the magnonic response. A proof-of-concept device is presented, with the SW frequency bandwidth tuned by micromagnet position, applied voltage, and external field. Furthermore, we investigate multilayer architectures, exploring the potential of vertical dipolar coupling to enable filtering and multiplexing functionalities. Preliminary results and experiments are reported.

L’elaborazione dell’informazione basata sullo spin rappresenta un’alternativa promettente all’elettronica tradizionale. In questo contesto, la magnonica, che sfrutta le eccitazioni delle onde di spin (SW) nei materiali magnetici, consente operazioni ad alta velocità, non volatili e compatibili con le frequenze RF. Tuttavia, la sua implementazione pratica è limitata da sfide legate all’integrazione e alle proprietà dei materiali. In questa tesi affrontiamo entrambi gli aspetti combinando tecniche avanzate di fabbricazione e strategie di integrazione ibrida.

YIG-based magnonics systems

ANDREA, DEL GIACCO
2025

Abstract

Spin-based information processing offers a promising alternative to traditional electronics, with magnonics, exploiting spin wave (SW) excitations in magnetic materials—enabling high-speed, non-volatile, and RF-compatible operation. However, practical implementation is hindered by integration challenges and material limitations. In this work, we address both aspects by combining advanced fabrication and hybrid integration strategies. Using soft-UV laser patterning of iron garnets (IG), we demonstrate direct crystallization of magnetic racetracks in non-magnetic matrices, paving the way for arbitrarily shaped magnonic circuits without physical etching. SW propagation over tens of micrometers with damping within one order of magnitude of state-of-theart YIG films is achieved, while Bi-doped YIG further enables laserinduced tailoring of magnetic anisotropy landscapes. To overcome substrate constraints, we integrate YIG with MEMS technology through the European MandMEMS consortium, realizing hybrid systems where MEMS actuation dynamically reconfigures the magnonic response. A proof-of-concept device is presented, with the SW frequency bandwidth tuned by micromagnet position, applied voltage, and external field. Furthermore, we investigate multilayer architectures, exploring the potential of vertical dipolar coupling to enable filtering and multiplexing functionalities. Preliminary results and experiments are reported.
YIG-based magnonics systems
24-ott-2025
Inglese
L’elaborazione dell’informazione basata sullo spin rappresenta un’alternativa promettente all’elettronica tradizionale. In questo contesto, la magnonica, che sfrutta le eccitazioni delle onde di spin (SW) nei materiali magnetici, consente operazioni ad alta velocità, non volatili e compatibili con le frequenze RF. Tuttavia, la sua implementazione pratica è limitata da sfide legate all’integrazione e alle proprietà dei materiali. In questa tesi affrontiamo entrambi gli aspetti combinando tecniche avanzate di fabbricazione e strategie di integrazione ibrida.
File in questo prodotto:
File Dimensione Formato  
SpinComputing_FinalUpload.pdf

embargo fino al 21/10/2028

Licenza: Tutti i diritti riservati
Dimensione 13.06 MB
Formato Adobe PDF
13.06 MB Adobe PDF

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14242/356491
Il codice NBN di questa tesi è URN:NBN:IT:POLIMI-356491