Abstract:We report an investigation of the effects of variation of composition on the properties of YIG (yttrium iron garnet) films grown on YAG substrates by multi-beam pulsed laser deposition (PLD).Magneto-optic garnets, such as YIG (Y 3 Fe 5 O 12 ), find application in areas including optical communication, where they can be used as optical rotators and isolators [1] via their Faraday effect, and also in the microwave range, where their ferromagnetic properties can be exploited e.g. in magnetostatic wave band pass filters [2].YIG films have been grown by liquid phase epitaxy [3], sputtering [4] or pulsed laser deposition (PLD) [2,5]. Single-beam PLD from a stoichiometric target usually leads to films that are iron (Fe) deficient, as previously reported in the literature [2] and confirmed in our experiments [5].We have studied the composition, analysed by energy-dispersive x-ray spectroscopy, and the effects of stoichiometric variation on the resultant properties of PLD-grown YIG films, whose yttrium (Y) and Fe concentration is varied by co-ablation of two separate targets of polycrystalline YIG with either a Y 2 O 3 or a Fe 2 O 3 target. In this report, we analyse in detail our results with the latter target, for which we have most extensive data.All depositions were performed in our multi-beam PLD system, described in [6]. YIG films were grown on 10 × 10 mm 2 1 mm-thick YAG (100)-oriented substrates, heated to T ≈ 1400 K, the highest temperature possible in our system, lower than the optimum value (T OPT ≈ 1600 K) for YIG/YAG growth [5]. The off-axis configuration of target holders relative to the substrate causes a lower deposition rate, compared to that reported in [5], using our single-beam PLD system; for this reason, we reduced the target-substrate distance to d = 4 cm and, at the same time, increased the oxygen pressure to P O2 = 3.4 Pa, following the general rule of thumb: d•P O2 = const., in order to maintain the same plume dynamics. Multi-PLD experiments of YIG + Fe 2 O 3 were performed with two different laser configurations, as described below.Samples Y45 and Y48-51, whose data are reported in the upper half of Table 1, were deposited by ablating the YIG target with a frequency-quadrupled Nd:YAG laser (wavelength: λ = 266 nm; fluence: F ≈ 1.3 J/cm 2 ; pulse repetition rate: f = 10 Hz) and the Fe 2 O 3 target with a KrF laser (wavelength: λ = 248 nm; fluence: F ≈ 1.3 J/cm 2 ), whose pulse repetition rate was varied between 0 and 4 Hz to increase the Fe concentration in the YIG films.Samples Y52-58, whose data are reported in lower half of Table 1, were deposited by ablating the YIG target with the KrF laser (fluence: F ≈ 2.3 J/cm 2 ; pulse repetition rate: f = 20 Hz, except for Y57, grown at f = 16 Hz) and the Fe 2 O 3 target with the Nd:YAG laser (F ≈ 1.3 J/cm 2 ), whose pulses were gated through a programmable shutter, whose opening and closing times were changed to achieve equivalent repetition rates of 0-4 Hz. The target ablation ratio r is defined as the ratio of the number of laser pulses per...