lator devices is commonly a MO iron garnet material, in particular yttrium iron garnet (YIG, Y 3 Fe 5 O 12 ) with substituents such as Ce or Bi to increase the MO performance. [5][6][7][8][9][10][11][12][13][14][15] However, the integration of garnets on a Si (or other semiconductor) platform is challenging due to the incompatible lattice parameters and the thermal expansion mismatch between garnets and common semiconductor substrates. [6][7][8][9][10][11][12][13][14][15][16] Moreover, crystallization of the garnet phase usually requires a high thermal budget. Garnets formed on semiconductor substrates are polycrystalline and exhibit higher optical absorption than single crystal films. Furthermore, impurity phases such as YFeO 3 , Fe 2 O 3 , and Bi 2 O 3 can form during the crystallization process, [17] which contributes to optical loss. These factors result in inferior optical performance of polycrystalline MO garnet films compared to the bulk garnet material, and a lower figure of merit (FoM), defined as the ratio of the Faraday rotation to the absorption coefficient per length of the material.Considerable work has been done on growth of garnet films on semiconductors to enable demonstrations of isolators and modulators. [6][7][8][9][10][11][12][13][14][15][16] The first monolithically integrated optical isolator [6] used 80 nm thick Ce:YIG which was grown by pulsed laser deposition (PLD) on a pre-annealed 20 nm thick YIG seed layer to induce crystallization of the Ce:YIG. A simplified PLD process was introduced by Sun et al. [11] where the YIG seed layer was placed on top of the MO garnet and both layers were crystallized simultaneously by rapid thermal annealing (RTA). This top-seedlayer process places the MO garnet in direct contact with the underlying Si waveguide, maximizing the coupling of light from the waveguide to the MO cladding, but it has only been applied to Ce:YIG. Recently rare-earth garnets have been developed that crystallize on Si and quartz without a seed layer, including sputter-deposited terbium iron garnet (TIG) and Bi-doped TIG (Bi:TIG). [18,19] Growth of MO materials on the sidewall of the waveguide can enable a wider range of device designs, including isolators for transverse electric (TE) polarization. Integrated semiconductor lasers emit TE-polarized light, and TE mode isolation using NRPS requires placement of the MO material on the sidewall of the waveguide to break left-right symmetry. [18,20,21] However, the NRPS-based integrated optical isolators that have been experimentally demonstrated are made with the MO material on the top or bottom surface of the waveguide, which isolates only the transverse magnetic (TM) polarization. [6][7][8][11][12][13][14] It is therefore essential to establish deposition conditions that yield Thin film magneto-optical (MO) materials are enablers for integrated nonreciprocal photonic devices such as isolators and circulators. Films of polycrystalline bismuth-substituted yttrium iron garnet (Bi:YIG) have been grown on silicon substrates and waveguide d...

ferrimagnetic yttrium iron garnet (YIG, Y 3 Fe 5 O 12 ) offer a desirable combination of a high Faraday rotation and low optical absorption. Cerium-and bismuth-substituted yttrium iron garnet (CeYIG, BiYIG) have an excellent MO figure of merit (FoM, the ratio of Faraday rotation to optical absorption), [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] and bismuth iron garnet and paramagnetic terbium gallium garnet are both used in bulk optical isolators.In order to incorporate these complex oxides into photonic integrated circuits, films with high FoM must be grown onto photonic substrates and devices. Although YIG crystallizes readily on non-garnet substrates, it has a low Faraday rotation (≈+100° cm −1 at 1550 nm) [10] and FoM. Crystallization of CeYIG or BiYIG without secondary phases can be accomplished by depositing then annealing the film directly or by annealing a bilayer consisting of a YIG seedlayer placed either above or below the active MO layer. [12,13] When a seed layer is included, a top seed layer is preferable as it maximizes coupling of the MO garnet with an optical mode propagating through an underlying waveguide.Like YIG, rare earth garnets such as terbium iron garnet (TbIG) or dysprosium iron garnet [19,20] can crystallize directly on non-garnet substrates to form polycrystalline films. TbIG is a thermodynamically stable phase like YIG [21,22] which promotes garnet-phase crystallization without a seed layer. TbIG has a Faraday rotation in 1100-1550 nm wavelength range of ≈ +270 to +1000° cm −1 . [23][24][25] Bi-substituted TbIG has been grown by liquidphase epitaxy and flux methods [23][24][25][26][27] or sol-gel methods, [28] and more recently thin films of Ce-and Bi-substituted TbIG (CeTbIG, BiTbIG) were grown on Si by sputtering. [29][30][31][32][33][34] Sputtered BiTbIG with Bi substitution of 14% of the Tb sites has a Faraday rotation of ≈−500° cm −1 at 1550 nm, [29] the same sign as that of CeYIG, which at 1550 nm wavelength had a Faraday rotation of at least −3700° cm −1 . [29] CeTbIG films with Ce substituting up to 25% of the Tb sites showed a Faraday rotation above −3200 ± 200° cm −1 , [31,32] but a 44 nm thick magnetic dead layer formed between the waveguide and the CeTbIG, hindering the interaction of evanescent light with the MO garnet cladding. [32] However, the optical absorption and FoM of these Bi-and Cesubstituted TbIG materials have not yet been reported.In this paper, we report the growth, magnetic, and optical characteristics of polycrystalline thin films of TbIG, CeTbIG, and BiTbIG synthesized using pulsed laser deposition (PLD) on Films of polycrystalline terbium iron garnet (TbIG), cerium-substituted TbIG (CeTbIG), and bismuth-substituted TbIG (BiTbIG) are grown on Si substrates by pulsed laser deposition. The films grow under tensile strain due to thermal mismatch with the Si substrate, resulting in a dominant magnetoelastic anisotropy which, combined with shape anisotropy, leads to in-plane magnetization. TbIG has a compensation temperature of 253 K which is...

Ceramic-chromium Hall sensors represent a temperature and radiation resistant alternative to Hall sensors based on semiconductors. Demand for these sensors is presently motivated by the ITER and DEMO nuclear fusion projects. The developed ceramic-chromium Hall sensors were tested up to a temperature of 550 °C and a magnetic field of 14 T. The magnitude of the sensitivity of the tested sensor was 6.2 mV/A/T at 20 °C and 4.6 mV/A/T at 500 °C. The sensitivity was observed to be weakly dependent on a temperature above 240 °C with an average temperature coefficient of 0.014%/°C and independent of the magnetic field with a relative average deviation below the measurement accuracy of 0.086%. A simulation of a neutron-induced transmutation was performed to assess changes in the composition of the chromium. After 5.2 operational years of the DEMO fusion reactor, the transmuted fraction of the chromium sensitive layer was found to be 0.27% at the most exposed sensor location behind the divertor cassette with a neutron fluence of 6.08 × 1025 n/m2. The ceramic-chromium Hall sensors show the potential to be suitable magnetic sensors for environments with high temperatures and strong neutron radiation.

No abstract