Periodic Reporting for period 1 - Pi4NoRM (Piezomagnetic ferrites for self-biased non-reciprocal millimeter wave devices)
Période du rapport: 2023-11-01 au 2025-04-30
Wireless data transmission has been transforming our World since the end of the 19th century. The internet and mobile communications marked a turning point, leading us into the “Data Age”: increasingly data-centric societies driven by the endless possibilities of wireless communications. Such relentless demand for improved connectivity and higher data transmission rates has accelerated the evolution of communication networks. The 5th generation (5G) is targeting 1Gb/s data rates, while 6G envisions Tb/s rates. The main strategy to reach these objectives is shifting to higher frequencies, eventually in the Sub-THz band with frequencies above 100 GHz. But since at these high frequencies, only direct line transmission is possible, the transition implies a paradigm change in the design, with a denser concentration of base stations with a large number of highly directional miniaturized antennas. This requires miniaturized and low-cost components, but above 100 GHz, the miniaturization and cost reduction are an unmet need for passive non-reciprocal components, such as circulators. These components allow antennas to simultaneously transmit and receive information, speeding up communication.
Ferrites are iron-based insulating magnetic oxides that intrinsically present such non-reciprocal responses, and have been successfully used in circulators operating up to 40 GHz without the application of external magnetic fields. However, to operate at the Sub-THz frequencies, state-of-the-art ferrites require rare-earth permanent magnets that become increasingly bulky with high operating frequencies. The aim of this project is to test the potential of a new family of RF ferrites based on ε-Fe2O3 that could help solve this problem. Project objectives are optimizing the scale-up of ε-Fe2O3 production, the sintering of the ferrite nanoparticles into dense parts geometrically relevant for the circulator operation and the proof-of-concept demonstration of a circulator with a non-reciprocal response above 100 GHz without the use of external magnetic fields.
Ferrites are iron-based insulating magnetic oxides that intrinsically present such non-reciprocal responses, and have been successfully used in circulators operating up to 40 GHz without the application of external magnetic fields. However, to operate at the Sub-THz frequencies, state-of-the-art ferrites require rare-earth permanent magnets that become increasingly bulky with high operating frequencies. The aim of this project is to test the potential of a new family of RF ferrites based on ε-Fe2O3 that could help solve this problem. Project objectives are optimizing the scale-up of ε-Fe2O3 production, the sintering of the ferrite nanoparticles into dense parts geometrically relevant for the circulator operation and the proof-of-concept demonstration of a circulator with a non-reciprocal response above 100 GHz without the use of external magnetic fields.
The work performed within Pi4NoRM pursued the consecution of two scientific and technical objectives:
1-Optimizing the scale-up of ε-Fe2O3 production and sintering for integration on circulators
2-Designing and testing proof-of-concept thin film circulators
For achieving Objective 1, the three following tasks were implemented:
T1.1 Optimising the scale-up of ε-Fe2O3 synthesis
We have optimized the bulk sol-gel synthesis of large high aspect ratio ε-Fe2O3 in amorphous SiO2 matrices, considering the influence of thermal treatment and the addition of Y3+ and other rare earths (La3+, Dy3+) to obtain batches of above 5 g material, avoiding the crystallization of SiO2 and the formation of undesired polymorphs. We have obtained insights into the nanorod growth mechanism, and we have investigated modifications of the synthesis protocol to accelerate the gelation, which is the step that limits the production turnover.
T1.2 Preparation of sintered ε-Fe2O3 parts. The prepared materials were sintered into 8 mm diameter disks of thicknesses ranging between 1 and 2 mm by Spark Plasma Sintering (SPS) under different pressures ranging from 50 to 450 MPa after optimizing the sintering time and temperature. Tests of standard sintering at high temperature in a furnace after isostatic pressing of the powders revealed some degree of transformation to hematite. The sintering of nanocomposites before etching the silica was not successful (friable disks were obtained).
T1.3 Exploiting the piezomagnetic effect in ε-Fe2O3 circulators We further investigated the piezomagnetic effect in ε-Fe2O3 disks sintered by SPS, and it was not possible to systematically reproduce the responses of pressure-induced magnetization previously observed in disks prepared by SPS. Since for most samples a zero magnetization was observed in the previous measurement the magnetization was low, the task was abandoned. Moreover, within the materials preparation objective, we also prepared Al-substituted Sr, Ca hexaferrites of composition Sr0.75Ca0.25Al3Fe9O19 with natural resonances above 100 GHz within the WB8 band (90-140 GHz), which is a complementary material to ε-Fe2O3, with its natural resonance lying within the WB5 band (140-220 GHz).
Objective 2 was implemented through the following tasks:
T2.1 Design and simulation of Y-junction circulators. We simulated the response of the ε-Fe2O3 and hexaferrite materials for Y-junction circulators for the WR8 (90-140 GHz) and WR5 (140-220 GHz) bands. The optimized geometries of the sintered ferrites corresponded to disks of 1 mm diameter and 0.7 mm thick, and 0.7 mm diameter and 0.5 mm thick for the WB8 band.
T2.2 Implementation and testing of waveguide circulators Y-junction circulators were fabricated, based on the designs of T2.1 to operate within the range 140-220 GHz, containing ferrites in the junction. The testing of the circulators revealed moderate non-reciprocal response, without externally applied magnetic fields, for Sr0.75Ca0.25Al3Fe9O19 in the W8 band and ε-Fe2O3, in the WB5 band. Differences of less than 5 dB between S21 and S12 (transmissions in opposite directions), which are still far from the 15 dB required for a commercial product.
T2.3 Implementation and testing of thin-film circulators. We performed tests to implement strip circulators in the WB8 and WB5 bands by depositing gold coplanar waveguides on commercial Rogers substrates, which showed strong losses, which limited the progress in the implementation of thin-film circulators.
1-Optimizing the scale-up of ε-Fe2O3 production and sintering for integration on circulators
2-Designing and testing proof-of-concept thin film circulators
For achieving Objective 1, the three following tasks were implemented:
T1.1 Optimising the scale-up of ε-Fe2O3 synthesis
We have optimized the bulk sol-gel synthesis of large high aspect ratio ε-Fe2O3 in amorphous SiO2 matrices, considering the influence of thermal treatment and the addition of Y3+ and other rare earths (La3+, Dy3+) to obtain batches of above 5 g material, avoiding the crystallization of SiO2 and the formation of undesired polymorphs. We have obtained insights into the nanorod growth mechanism, and we have investigated modifications of the synthesis protocol to accelerate the gelation, which is the step that limits the production turnover.
T1.2 Preparation of sintered ε-Fe2O3 parts. The prepared materials were sintered into 8 mm diameter disks of thicknesses ranging between 1 and 2 mm by Spark Plasma Sintering (SPS) under different pressures ranging from 50 to 450 MPa after optimizing the sintering time and temperature. Tests of standard sintering at high temperature in a furnace after isostatic pressing of the powders revealed some degree of transformation to hematite. The sintering of nanocomposites before etching the silica was not successful (friable disks were obtained).
T1.3 Exploiting the piezomagnetic effect in ε-Fe2O3 circulators We further investigated the piezomagnetic effect in ε-Fe2O3 disks sintered by SPS, and it was not possible to systematically reproduce the responses of pressure-induced magnetization previously observed in disks prepared by SPS. Since for most samples a zero magnetization was observed in the previous measurement the magnetization was low, the task was abandoned. Moreover, within the materials preparation objective, we also prepared Al-substituted Sr, Ca hexaferrites of composition Sr0.75Ca0.25Al3Fe9O19 with natural resonances above 100 GHz within the WB8 band (90-140 GHz), which is a complementary material to ε-Fe2O3, with its natural resonance lying within the WB5 band (140-220 GHz).
Objective 2 was implemented through the following tasks:
T2.1 Design and simulation of Y-junction circulators. We simulated the response of the ε-Fe2O3 and hexaferrite materials for Y-junction circulators for the WR8 (90-140 GHz) and WR5 (140-220 GHz) bands. The optimized geometries of the sintered ferrites corresponded to disks of 1 mm diameter and 0.7 mm thick, and 0.7 mm diameter and 0.5 mm thick for the WB8 band.
T2.2 Implementation and testing of waveguide circulators Y-junction circulators were fabricated, based on the designs of T2.1 to operate within the range 140-220 GHz, containing ferrites in the junction. The testing of the circulators revealed moderate non-reciprocal response, without externally applied magnetic fields, for Sr0.75Ca0.25Al3Fe9O19 in the W8 band and ε-Fe2O3, in the WB5 band. Differences of less than 5 dB between S21 and S12 (transmissions in opposite directions), which are still far from the 15 dB required for a commercial product.
T2.3 Implementation and testing of thin-film circulators. We performed tests to implement strip circulators in the WB8 and WB5 bands by depositing gold coplanar waveguides on commercial Rogers substrates, which showed strong losses, which limited the progress in the implementation of thin-film circulators.
The Pi4NoRM results show that high anisotropy magnetic ferrites displaying such as (Sr, Ca, Al) –hexaferrites and ε-Fe2O3 can be used to implement sub-THz circulators that can operate without external magnetic fields and, therefore, be miniaturized. However, significant developments are needed both in the materials and RF engineering sides. The sintered ferrites tested in the project were not preferentially oriented with their easy magnetic axes along the disk axis and thus presented a low ratio of remanent magnetization over saturation magnetization. With oriented ferrites it is expected that the non-reciprocal response can be significantly increased. On the other hand, progress is also needed in the design of miniaturized microstrip Y-junction circulators prior to the integration and testing of ferrites.