Ultra-High Cooperativity Interactions between Magnons and Resonant Photons in a YIG sphere

Ultra-High Cooperativity Interactions between Magnons and Resonant Photons in a YIG sphere

J. Bourhill jeremy.bourhill@uwa.edu.au    N. Kostylev    M. Goryachev    D. L. Creedon    M.E. Tobar ARC Centre of Excellence for Engineered Quantum Systems, University of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia
July 16, 2019

Resonant photon modes of a 5mm diameter YIG sphere loaded in a cylindrical cavity in the 10-30GHz frequency range are characterised as a function of applied DC magnetic field at millikelvin temperatures. The photon modes are confined mainly to the sphere, and exhibited large mode filling factors in comparison to previous experiments, allowing ultrastrong coupling with the magnon spin wave resonances. The largest observed coupling between photons and magnons is GHz for a 15.5 GHz mode, corresponding to a cooperativity of . Complex modifications beyond a simple multi-oscillator model, of the photon mode frequencies were observed between 0 and 0.1 Tesla. Between 0.4 to 1 Tesla, degenerate resonant photon modes were observed to interact with magnon spin wave resonances with different couplings strengths, indicating time reversal symmetry breaking due to the gyrotropic permeability of YIG. Bare dielectric resonator mode frequencies were determined by detuning magnon modes to significantly higher frequencies with strong magnetic fields. By comparing measured mode frequencies at 7 Tesla with Finite Element modelling, a bare dielectric permittivity of of the YIG crystal has been determined at about .

I Introduction

Hybrid photon-magnon systems in ferromagnetic spheres have recently emerged as a promising approach towards coherent information processing Chumak et al. (2015); Tabuchi et al. (2015, 2014); Zhang et al. (2014); Goryachev et al. (2014a); Bai et al. (2015); Huebl et al. (2013); Zhang et al. (2015a). Due to the large exchange interaction between spins in ferromagnets, they will lock together to form a macrospin that can be utilised for coherent information processing protocols Imamoğlu (2009); Zhang et al. (2015a). The quantised excitation of the collective spin is referred to as a magnon. Yttrium Iron Garnet (YIG) based magnon systems are attractive due to very high spin density, resulting in significant cooperativity as well as relatively narrow linewidths Goryachev et al. (2014a); Zhang et al. (2015b); Osada et al. (2015); Zhang et al. (2014). Furthermore, due to the possibility of coupling magnon modes to photons at optical frequencies Zhang et al. (2015b); Osada et al. (2015); Shen and Bloembergen (1966); Demokritov et al. (2001), magnon systems may be considered as a candidate for coherent conversion of microwave and optical photons Zhang et al. (2015b); Osada et al. (2015). In addition, magnons interact with elastic waves Kittel (1958); Zhang et al. (2015c) opening a window for combining mechanical and magnetic systems. These systems therefore possesses great potential as an information transducer that mediates inter-conversion between information carriers of different physical nature thus establishing a novel approach to hybrid quantum systems Tian et al. (2004); Imamoğlu (2009); Verdú et al. (2009); Xiang et al. (2013).

Among all magnon systems the central role is devoted to YIG, a material that possesses exceptional magnetic and microwave properties and has been used in microwave systems such as tuneable oscillators and filters for many decades Cherepanov et al. (1993); Gurevich (1963). Although, only recently Soykal and Flatté proposed and modelled the photon-magnon interaction based on YIG nano-spheres with application to quantum systems Soykal and Flatté (2010a, b). As predicted by the authors, extremely large coupling rates, , could be achieved in YIG spheres, which is favourable for coherent information exchange and has been demonstrated experimentally later Goryachev et al. (2014a); Kostylev et al. (2015); Zhang et al. (2014). For these experiments, the interaction is observed between photon and magnon resonances created correspondingly by photon cavity boundary conditions and spin precession under external DC magnetic field. A commonly used method is to place a relatively small YIG sphere in a local maxima of the magnetic field inside a much larger microwave cavity. This is done to achieve quasi uniform distribution of the cavity field over the sphere volume to avoid spurious magnon modes. Cavities can take oval Tabuchi et al. (2014); Zhang et al. (2014) or spherical shapes Zare Rameshti et al. (2015); Soykal and Flatté (2010a, b), and even re-entrant cavities with multiple posts have been used in an attempt to focus the microwave energy over the sphere Goryachev et al. (2014a); Kostylev et al. (2015). In this work we investigate a completely different regime in which the magnon and photon wavelengths are comparable, leading to considerably larger coupling strengths, but additional couplings to higher order modes. In general, for this case the strength of the photon-magnon interaction will be determined by an overlap integral of the two respective mode shape functions. Given the magnon mode shape is limited to the sphere’s volume, this integral will be maximised when the photon mode is confined to the same volume. To achieve the latter, we utilise an exceptionally large YIG sphere with diameter mm, matching magnon and microwave photon mode volumes, unlike previous microwave cavity experiments.

In order to investigate this regime we use common microwave spectroscopy techniques Farr et al. (2013); Goryachev et al. (2014b); Bourhill et al. (2013, 2015a); Goryachev et al. (2014a); Kostylev et al. (2015) to directly observe the mode splitting caused by the magnon interaction to determine the coupling values. Similar systems have been extensively utilised not only in the field of spintronics to investigate the interaction between microwave photons and paramagnetic spin ensembles Farr et al. (2013); Bourhill et al. (2015a); Goryachev et al. (2014b); Bourhill et al. (2013); Farr et al. (2014), but also to realise optical comb generation Del/’Haye et al. (2007), ultra-low threshold lasing Kippenberg et al. (2006), cavity-assisted cooling, control and measurement of optomechanical systems Aspelmeyer et al. (2014); Bourhill et al. (2015b), and extremely stable cryogenic sapphire oscillator clock technology Ivanov and Tobar (2009a, b). To date, exciting internal, highly-confined photonic modes in a YIG sphere has only recently been demonstrated in the optical regime using Whispering Gallery Modes (WGM) Zhang et al. (2015b); Osada et al. (2015) but has never before been achieved in the microwave domain. This is due to the typical sub-millimetre diameter of the spheres. As such, interactions with magnons must be observed via Brillouin scattering Serga et al. (2012), which has yielded high quality factors and also demonstrated a pronounced nonreciprocity and asymmetry in the sideband signals generated by the magnon-induced scattering.

Extremely large mode splittings () cause simultaneous coupling to a higher density of modes, with an overlap of avoided level crossings. Therefore, the model proposed by Soykal and Flatté Soykal and Flatté (2010a) becomes no longer applicable, as it assumes the interaction occurs between a single photonic and magnon mode. More recently, a paper by Rameshti et al. Zare Rameshti et al. (2015) simulated a similar scenario of the presented experiment, in which the ferromagnetic sphere is itself the microwave cavity. Our observed results may appear to be in good agreement with this work’s predictions, however, what is apparent is that in this specialised case, one must consider more than just the magnetostatic, uniform Kittel magnon mode, a limitation of Zare Rameshti et al. (2015). Indeed, due to the nonuniformity of both the microwave mode magnetic field energy density across the sphere, which is unique to this experiment, and the nonuniformity of the sphere parameters arising due to cryogenic cooling, the assumption that only the uniform Kittel magnon resonance participates is no longer valid. Despite this, in this paper we use a two mode model to obtain estimations of coupling strengths, and demonstrate how this results in inconsistent susceptibility values.

Figure 1: (Color online) Cross section of the copper cavity that houses the 5 mm YIG sphere. The sphere sits on a sapphire disk, and microwaves are coupled in and out of it via loop probes which produce and detect an component. A variable DC magnetic field is applied along the axis.

Ii Physical Realization

The mm YIG sphere was manufactured by Ferrisphere, Inc. with a quoted room temperature saturation magnetisation of T. It is placed on a small sapphire disk, with a concavity etched out using a diamond tipped ball grinder, to keep the sphere from rolling out of position, and reduce dielectric losses that would arise if the YIG were in direct contact with the conductive copper housing. Sapphire was chosen over teflon as an intermediary between the YIG and copper to improve the thermal conductivity to the sphere.

Together, the sapphire and YIG are housed inside a copper cavity with dimensions specified in figure 1. A loop probe constructed from flexible subminiature version A cable launchers is used to input microwaves and a second is used to make measurements, allowing the determination of parameters. The entire cavity is cooled to about 20 mK by means of a dilution refrigerator (DR) with a cooling power of about 500 W at 100 mK. The cavity is attached to a copper rod bolted to the mixing chamber stage of the DR that places it at the field center of a 7 T superconducting magnet, whose applied field is oriented in the direction of the cavity. The magnet is attached to the 4 K stage of the DR, with the copper cavity mounted within a radiation shield of approximately 100 mK that sits within the bore of the magnet.

Iii Experimental Observations

Figure 2: (Color online) Transmission data as magnetic field is swept.

The transmission spectrum of the YIG was recorded for DC magnetic fields swept from 0 – 7 T using a vector network analyser (VNA), with partial results shown in figure 2. A host of magnon resonances/higher order magnon-polaritons can be observed originating from (0 T, 0 GHz) with an approximate gradient of 28 GHz/T. The more-or-less horizontal lines approaching the magnon resonances from either side correspond to resonant photon modes of the sphere. Importantly, we can observe that in the dispersive regime, far removed from any microwave resonant mode, there still exist multiple magnon modes. We observe that the anticrossing gaps are populated by unperturbed modes, which are remnant “tails” of both “higher” and “lower” mode interactions, as predicted by Rameshti et al. in the ultrastrong coupling regime Zare Rameshti et al. (2015).

For the remainder of this paper, we will focus on the six lowest frequency photon modes, whose resonant frequencies may only be accurately determined at large magnetic fields, when the entire spin ensemble has been detuned, as shown by figure 3.

Figure 3: (Color online) (a) Asymptotic frequency values of the six lowest order photon modes as T. (b) Transmission spectra at T, from which mode linewidths may be measured.

The modes have been categorised into three distinct classes: mode “” is the lowest frequency and lowest factor mode, the two highest factor modes; “” and “”, and the three remaining highest frequency modes; “1”, “2” and “3”. Their asymptotic frequencies as T are summarised in table 1.

The behaviour of these modes as the magnon resonances are tuned via the applied magnetic field is shown in figure 4. It has been shown previously Goryachev et al. (2014a) that a standard model of two interacting harmonic oscillators can accurately determine the coupling values from such avoided crossings. However, we observe strong distortion around 0 T, and also an asymmetry of the mode splittings about the central magnon resonances due to the ultrastrong coupling of the photon modes to the magnon modes, as was observed previously in Ruby Farr et al. (2014). Therefore we fit only the curves to the right of the magnon resonance. These fits are shown as the dashed lines in figure 4. From these fits we can approximate the values of for each mode, as summarised in table 1. The linewidths, and frequencies, of the photon modes are determined from the transmission spectra taken at high field values (figure 3 (b)), whilst the magnon linewidth, can be determined by analysing the transmission spectra in the dispersive regime. We take a frequency sweep at T from 5.75–9 GHz in order to view the magnon resonance peaks far away from any interaction with the dielectric microwave modes, as shown in figure 5. There is a level of variation amongst the magnon linewidths as calculated by fitting the peaks with Fano resonance fits, as shown in figure 6. This variation and the presence of multiple peaks demonstrates the presence of higher order magnon modes. Taking the average and standard deviation of these linewidths gives a final estimate of magnon linewidth as MHz. Cooperativity is calculated as .

(GHz) (MHz) (GHz) () (%)
12.779 11.84 4.79 5.971.85 18.7
15.506 1.029 7.11 15147.0 22.9
15.563 1.197 4.19 45.214.0 13.5
1 15.732 5.355 6.15 21.86.76 19.5
2 15.893 2.965 3.04 9.602.98 9.56
3 15.950 2.965 0.78 0.6320.196 2.45
Table 1: Measured and calculated results for each mode showing couplings and cooperativities, .
Figure 4: (Color online) Two harmonic oscillator model fitting to modes , , 1, 2 and 3. From the curved lineshapes, one can determine the coupling value .
Figure 5: S21 transmission spectra showing a host of magnon resonance peaks at T.
Figure 6: Fitting the magnon resonances with Fano fits.

The cooperativity values in table 1 demonstrate that all modes are strongly coupled to the magnons, and all with the exception of modes 2 and 3 are in the ultrastrong coupling regime (i.e. Zare Rameshti et al. (2015)). The largest cooperativity value obtained is that of mode , which is, to the authors’ knowledge, the largest value ever reported to date in any previously studied spin system.

A transmission spectrum taken at T is shown in figure 7, demonstrating the mode splitting of mode 1, symmetric about the magnon resonance. Overlaid in red is the bare photon resonance at 7 T, i.e. the microwave mode unperturbed by the magnon modes. From this red curve, is determined to be 5.355 MHz, as shown in table 1. When one takes the average of MHz and , one obtains the line width of the resulting hybrid state when the magnon resonance is tuned coincident in frequency with the photon mode, as depicted by the dashed blue curve in figure 7, i.e., 4.4 MHz. This excellent agreement indicates that at this particular field, mode 1 exists as a hybrid magnon-polariton.

Figure 7: (Color online) Transmission spectrum at T. At this applied magnetic field the magnon resonance is tuned coincident in frequency with mode 1, and the strong coupling between the two results in a mode splitting of 6.2 GHz. The high density of resonant peaks in the centre of the figure suggest a large number of higher order magnon modes are present in this system.
Figure 8: (Color online) Behaviour of the photon modes around T, a result of the internal magnetisation of YIG.

Around T, we observe a severe distortion of the cavity mode frequency dependence on magnetic field as demonstrated by Fig. 8. Around 16 GHz, we see there exist five modes, corresponding to modes , and 1–3, on the “left” side of the magnon resonances. These modes have been given a primed nomenclature to indicate their existence at fields lower than that required to tune the magnons to their frequencies. This phenomenon has been previously observed in single crystal YAG Farr et al. (2015) highly doped with rare-earth Erbium ions, and is explained by the influence on the ferromagnetic phase of the impurity ions on degenerate modes. The effect can be explained by the influence of the ensemble of strongly coupled spins on the centre-propagating waves of the near degenerate mode doublet. For large spin-photon interactions, tails of Avoided Level Crossings (ALCs) from the positive half plane () should still exist on the negative half plane () and vice a versa. Although, instead of gradual change of direction, the system demonstrates an abrupt transition to a “no coupling” state. It is worth mentioning that such effect has not been observed in photonic systems interacting with paramagnetic spin ensembles Farr et al. (2013); Goryachev et al. (2014b, c). In the present case, the effect is much more pronounced, with fractional frequency deviations and the magnetic field range of the effect both orders of magnitude larger than observed previously Farr et al. (2015), a result of the magnetic spin density.

Iv Discussion

COMSOL 3.5’s electromagnetic package was used to model the system. A 3D model was used so as to analyse the degeneracies in the axis of the dielectric modes. The internal copper wall of the cavity is modelled as a perfect electrical conductor, which for the purposes of the desired eigenfrequency study, is an appropriate simplification.

The results of the FEM using a value of and mm, where is the radius of curvature of the sapphire support’s concavity, are summarised in figure 10 and in table 2. The measured frequency of the doublet modes has been taken as the average of the two constituent’s frequencies at T.

 Mode (GHz) (GHz)
12.779 12.785 (0,0)
15.534 15.286 (1,1)
1 15.732 15.736 (1,0)
2 3 15.922 15.921 (1,1)
Table 2: Comparison of FEM and measured frequencies and hence mode identification.

From the FEM and the analytical mode shapes of spherical dielectric resonances described by le Floch et al. (2006), we can identify mode as an mode with no degeneracy. Therefore it is present as a singular resonance. The other five modes appear as modes. There should only exist a fold degeneracy for resonant spherical photon modes, which can be broken by internal impurities or by asymmetric boundary conditions set by a cylindrical enclosure, microwave loop probes, and the sapphire substrate, collectively termed “backscatterers”. This degeneracy arises from a Legendre polynomial in the mode’s and field analytical expressions of the form , where . The integers and represent the number of maxima of the mode’s energy density in the direction over 180, and the number in the direction over 180, respectively. This would imply that for we should observe three distinct modes corresponding to a single and two modes, rather than five modes. However, the FEM demonstrates that the use of the sapphire support base introduces a further degeneracy to the modes depending on the amount of field that permeates the sapphire. The modelling predicts four modes, existing as two sets of two, which are separated by approximately 500 MHz. This is in fair agreement with the separation of modes , with modes 1–3. Therefore it is apparent that modes and are a doublet pair with .

Given that modes 2 and 3 approach relatively similar frequencies at high magnetic fields, it is reasonable to assume that these modes correspond to the second doublet pair, which FEM predicts will have a larger proportion of microwave field inside the sapphire support. This means that mode 1 must be the single mode.

From figure 4, we can see that both the doublet pairs demonstrate a gyrotropic response when interacting with the magnon resonances, i.e., one mode interacts more than its doublet pair. This is a common occurrence in spin ensemble systems and has been observed in paramagnetic systems such as Fe in sapphire Benmessai et al. (2009); Goryachev et al. (2014b); Bourhill et al. (2013). This asymmetric interaction strength for doublet pairs has also been observed in ferromagnets by Krupka et al. Krupka (1991); Krupka et al. (1999) and predicted by Rameshti et al. Zare Rameshti et al. (2015), with the latter stating that , where a different notation to that used here is employed, in which . The notations are equivalent as an doublet in Zare Rameshti et al. (2015) corresponds to a doublet pair here.

The gyrotropic response is a result of the anisotropy of a ferromagnet’s permeability tensor; the same reason why these materials are used in circulators. The permeability tensor, containing off diagonal terms appears as


where is the permeability of free space, and is the magnetic susceptibility of the ferromagnet, which is related to the magnetic permeability tensor by .

When any resonant photonic mode exists as a doublet, it is because the degeneracy has been broken by some backscatterer, and the two resulting modes exist as counter propagating travelling waves Benmessai et al. (2009); Goryachev et al. (2014b). The overall effect is that one travelling wave will see an effective permeability of , whilst the other will see , which can be rewritten as , and we can state that , where is the “unperturbed” magnetic susceptibility that a standing wave would observe.

(GHz) (GHz)
12.779 4.79 0.221 0.159
15.506 7.11 0.594 0.0885
15.563 4.19 0.594 0.0305
1 15.732 6.15 0.728 0.0525
2 15.893 3.04 0.493 0.0185
3 15.950 0.78    0.493   0.00121
Table 3: Calculated magnetic filling factors, and effective magnetic susceptibilities, for each of the photon modes.

The effective susceptibility that a mode experiences will determine the interaction strength of that mode with a magnon resonance according to Goryachev et al. (2014a):


where is the total magnetic filling factor of the mode; i.e., the proportion of magnetic field within the ferromagnetic material compared to the entire system. This parameter is used in an attempt to quantify the overlap of the magnon and photon modes and is calculated as


It should be noted that typically it is only the magnetic field energy density perpendicular to the external magnetic field that is considered to interact with the spin system Goryachev et al. (2014a); Soykal and Flatté (2010a). However, the interaction of mode 1 is far larger than its perpendicular filling factor of 0.075 would suggest. So, in an attempt to account for the interaction with nonuniform magnon modes, the total magnetic filling factor has been used. These have been calculated from the FEM and the resulting values of are displayed in table 3.

Given our assumption that mode 1 represents the (1,0) dielectric mode, which will exist as a standing wave given no possible degeneracy, the calculated value for this mode should represent the unperturbed magnetic susceptibility of the YIG. Taking the average of the values for the doublet modes () and () yields a value of ; in reasonable agreement with the value obtained from mode 1.

The FEM predicts that modes , 2 and 3 will each contain a significant proportion of magnetic field energy within the sapphire support, so one would expect these modes to observe a lower effective magnetic susceptibility, which would appear true for the latter two modes (their average susceptibility yields an unperturbed susceptibility of ). However, mode demonstrates a much larger coupling strength than what should be afforded a mode with its filling factor, hence a value approximately three times larger than the unperturbed value obtained from modes , and 1. This suggests that our approximation of using the total magnetic filling factor to quantify the overlap of the magnon and photon modes is not entirely accurate. To accurately explain the origins of the differing interaction strengths of each mode, knowledge of higher order, nonuniform magnon mode shapes are required, in order to replace the filling factor approximation with an overlap value. Unlike Zhang et al.’s Zhang et al. (2014) ultrastrong coupling results with a mm YIG sphere, in which higher order magnon modes mostly couple weakly with the microwave cavity, here we excite internal, nonuniform electromagnetic resonances, so it is more likely than not that these modes will couple more strongly to nonuniform magnon modes if their mode shapes match up well spatially. The derived values of susceptibility in table 3 agree within an order of magnitude to previously measured results Krupka (1991) but have been underestimated due to the use of filling factor as opposed to a mode overlap integral.

Figure 9: Frequency difference between simulated and measured results as the relative permittivity of YIG is varied in the FEM software. The radius of curvature of the sapphire support used here was mm, which for mode 1 is largely irrelevant, but for modes , 2 and 3 gives good agreement.

Finally, we can use the predicted mode frequencies of the FEM to determine the permittivity of the YIG sample, by varying until the frequencies match the asymptotic values measured at high magnetic fields. At these magnetic field values, the matrix in equation (1) becomes the identity matrix Krupka (1991). By measuring the depth of the sapphire concavity and its width at the surface, the radius of curvature was determined to be mm. With this information, an iterative simulation was conducted mapping mode frequencies versus relative permittivity of YIG. It was found that mode 1 is relatively insensitive to the radius of curvature of the sapphire support. This is due to the absence of electric field density outside the YIG for this particular mode. Given that contains a significant amount of uncertainty, this mode is used to match with . A plot of versus permittivity is shown in figure 9. From this result, we can state that . This value agrees well with previous measurements taken using the so-called “Courtney” technique with YIG samples Krupka et al. (1999).

In conclusion, we observe ultrastrong coupling between internal dielectric microwave resonances and magnons inside a mm YIG sphere. The large diameter of the sphere results in not only an increased number of spins, but also the accessibility of the internal electromagnetic resonances due to their existence below K-band frequencies. The use of internal microwave modes instead of an external cavity resonance results in far larger magnetic filling factors than ever before achieved in such an experiment, hence the coupling values and cooperitivity values observed are, to the authour’s knowledge, the largest ever reported, with a maximum GHz, or mode linewidths, and . This implies an extremely high level of coherence in this system. Most importantly however, the numerous resonant magnon peaks in the dispersive regime and the discrepancies in calculated susceptibilities suggest that higher order magnon modes participate in this system. This implies that the previously theoretically analysed models of such systems are incomplete.

Figure 10: Spherical coordinate field components of the six lowest dielectric modes in the YIG/sapphire/air system. Each mode is viewed parallel to the -axis (top row) and in the plane (bottom row), except where no field is present. We can readily identify modes and 1 as (0,0) and (1,0) spherical dielectric modes, respectively. These two modes appear as “pure” dielectric modes containing only three field components. The two doublet modes (, , 2 and 3) appear to contain energy density in all six field components, and are greatly affected by the sapphire support, which is what appears to lead to the additional degeneracy; splitting two modes in to four modes. From the radial components we can identify these modes as being modified (1,1) spherical dielectric modes.


  • Chumak et al. (2015) A. V. Chumak, V. I. Vasyuchka, A. A. Serga,  and B. Hillebrands, “Magnon spintronics,” Nat Phys 11, 453–461 (2015).
  • Tabuchi et al. (2015) Yutaka Tabuchi, Seiichiro Ishino, Atsushi Noguchi, Toyofumi Ishikawa, Rekishu Yamazaki, Koji Usami,  and Yasunobu Nakamura, “Coherent coupling between a ferromagnetic magnon and a superconducting qubit,” Science 349, 405–408 (2015)http://www.sciencemag.org/content/349/6246/405.full.pdf .
  • Tabuchi et al. (2014) Yutaka Tabuchi, Seiichiro Ishino, Toyofumi Ishikawa, Rekishu Yamazaki, Koji Usami,  and Yasunobu Nakamura, “Hybridizing ferromagnetic magnons and microwave photons in the quantum limit,” Phys. Rev. Lett. 113, 083603 (2014).
  • Zhang et al. (2014) Xufeng Zhang, Chang-Ling Zou, Liang Jiang,  and Hong X. Tang, “Strongly coupled magnons and cavity microwave photons,” Phys. Rev. Lett. 113, 156401 (2014).
  • Goryachev et al. (2014a) Maxim Goryachev, Warrick G. Farr, Daniel L. Creedon, Yaohui Fan, Mikhail Kostylev,  and Michael E. Tobar, “High-cooperativity cavity qed with magnons at microwave frequencies,” Phys. Rev. Applied 2, 054002 (2014a).
  • Bai et al. (2015) Lihui Bai, M. Harder, Y. P. Chen, X. Fan, J. Q. Xiao,  and C.-M. Hu, “Spin pumping in electrodynamically coupled magnon-photon systems,” Phys. Rev. Lett. 114, 227201 (2015).
  • Huebl et al. (2013) Hans Huebl, Christoph W. Zollitsch, Johannes Lotze, Fredrik Hocke, Moritz Greifenstein, Achim Marx, Rudolf Gross,  and Sebastian T. B. Goennenwein, “High cooperativity in coupled microwave resonator ferrimagnetic insulator hybrids,” Phys. Rev. Lett. 111, 127003 (2013).
  • Zhang et al. (2015a) Xufeng Zhang, Chang-Ling Zou, Na Zhu, Florian Marquardt, Liang Jiang,  and Hong X. Tang, “Magnon dark modes and gradient memory,” Nat Commun 6 (2015a).
  • Imamoğlu (2009) Atac Imamoğlu, “Cavity qed based on collective magnetic dipole coupling: Spin ensembles as hybrid two-level systems,” Phys. Rev. Lett. 102, 083602 (2009).
  • Zhang et al. (2015b) X. Zhang, N. Zhu, C.-L. Zou,  and H. X. Tang, “Optomagnonic whispering gallery microresonators,” arXiv:1510.03545 (2015b).
  • Osada et al. (2015) A. Osada, R. Hisatomi, A. Noguchi, Y. Tabuchi, R. Yamazaki, K. Usami, M. Sadgrove, R. Yalla, M. Nomura,  and Y. Nakamura, “Cavity optomagnonics with spin-orbit coupled photons,” arXiv:1510.03545 (2015).
  • Shen and Bloembergen (1966) Y. R. Shen and N. Bloembergen, “Interaction between light waves and spin waves,” Phys. Rev. 143, 372–384 (1966).
  • Demokritov et al. (2001) S.O. Demokritov, B. Hillebrands,  and A.N. Slavin, “Brillouin light scattering studies of confined spin waves: linear and nonlinear confinement,” Physics Reports 348, 441 – 489 (2001).
  • Kittel (1958) C. Kittel, “Excitation of spin waves in a ferromagnet by a uniform rf field,” Phys. Rev. 110, 1295–1297 (1958).
  • Zhang et al. (2015c) X. Zhang, C.-L. Zou, L. Jiang,  and H. X. Tang, “Cavity magnomechanics,” arXiv:1511.02680v2 (2015c).
  • Tian et al. (2004) L. Tian, P. Rabl, R. Blatt,  and P. Zoller, “Interfacing quantum-optical and solid-state qubits,” Phys. Rev. Lett. 92, 247902 (2004).
  • Verdú et al. (2009) J. Verdú, H. Zoubi, Ch. Koller, J. Majer, H. Ritsch,  and J. Schmiedmayer, “Strong magnetic coupling of an ultracold gas to a superconducting waveguide cavity,” Phys. Rev. Lett. 103, 043603 (2009).
  • Xiang et al. (2013) Ze-Liang Xiang, Sahel Ashhab, J. Q. You,  and Franco Nori, “Hybrid quantum circuits: Superconducting circuits interacting with other quantum systems,” Reviews of Modern Physics 85, 623–653 (2013).
  • Cherepanov et al. (1993) V. Cherepanov, I. Kolokolov,  and V. Lvov, “The saga of yig: Spectra, thermodynamics, interaction and relaxation of magnons in a complex magnons,” Physics Reports 229, 81–144 (1993).
  • Gurevich (1963) A.G. Gurevich, Ferrites at Microwave Frequencies (Consultants Bureau, New York, 1963).
  • Soykal and Flatté (2010a) Ö. O. Soykal and M. E. Flatté, ‘‘Strong field interactions between a nanomagnet and a photonic cavity,” Phys. Rev. Lett. 104, 077202 (2010a).
  • Soykal and Flatté (2010b) Ö. O. Soykal and M. E. Flatté, “Size dependence of strong coupling between nanomagnets and photonic cavities,” Phys. Rev. B 82, 104413 (2010b).
  • Kostylev et al. (2015) N. Kostylev, M. Goryachev,  and M. E. Tobar, “Superstrong coupling of a microwave cavity to yig magnons,” arXiv:1508.04967 (2015).
  • Zare Rameshti et al. (2015) Babak Zare Rameshti, Yunshan Cao,  and Gerrit E. W. Bauer, “Magnetic spheres in microwave cavities,” Phys. Rev. B 91, 214430 (2015).
  • Farr et al. (2013) Warrick G. Farr, Daniel L. Creedon, Maxim Goryachev, Karim Benmessai,  and Michael E. Tobar, “Ultrasensitive microwave spectroscopy of paramagnetic impurities in sapphire crystals at millikelvin temperatures,” Phys. Rev. B 88, 224426 (2013).
  • Goryachev et al. (2014b) Maxim Goryachev, Warrick G. Farr, Daniel L. Creedon,  and Michael E. Tobar, “Spin-photon interaction in a cavity with time-reversal symmetry breaking,” Phys. Rev. B 89, 224407 (2014b).
  • Bourhill et al. (2013) J. Bourhill, K. Benmessai, M. Goryachev, D. L. Creedon, W. Farr,  and M. E. Tobar, “Spin bath maser in a cryogenically cooled sapphire whispering gallery mode resonator,” Phys. Rev. B 88, 235104 (2013).
  • Bourhill et al. (2015a) Jeremy Bourhill, Maxim Goryachev, Warrick G. Farr,  and Michael E. Tobar, “Collective behavior of ions in ruby revealed by whispering gallery modes,” Phys. Rev. A 92, 023805 (2015a).
  • Farr et al. (2014) Warrick G. Farr, Maxim Goryachev, Daniel L. Creedon,  and Michael E. Tobar, “Strong coupling between whispering gallery modes and chromium ions in ruby,” Phys. Rev. B 90, 054409 (2014).
  • Del/’Haye et al. (2007) P. Del/’Haye, A. Schliesser, O. Arcizet, T. Wilken, R. Holzwarth,  and T. J. Kippenberg, “Optical frequency comb generation from a monolithic microresonator,” Nature 450, 1214–1217 (2007).
  • Kippenberg et al. (2006) T. J. Kippenberg, J. Kalkman, A. Polman,  and K. J. Vahala, “Demonstration of an erbium-doped microdisk laser on a silicon chip,” Phys. Rev. A 74, 051802 (2006).
  • Aspelmeyer et al. (2014) Markus Aspelmeyer, Tobias J. Kippenberg,  and Florian Marquardt, “Cavity optomechanics,” Rev. Mod. Phys. 86, 1391–1452 (2014).
  • Bourhill et al. (2015b) J. Bourhill, E. Ivanov,  and M. E. Tobar, “Precision measurement of a low-loss cylindrical dumbbell-shaped sapphire mechanical oscillator using radiation pressure,” Phys. Rev. A 92, 023817 (2015b).
  • Ivanov and Tobar (2009a) Eugene N Ivanov and Michael E Tobar, “Microwave phase detection at the level of 10(-11) rad.” Rev Sci Instrum 80, 044701 (2009a).
  • Ivanov and Tobar (2009b) E.N. Ivanov and M.E. Tobar, “Low phase-noise sapphire crystal microwave oscillators: current status,” Ultrasonics, Ferroelectrics, and Frequency Control, IEEE Transactions on 56, 263–269 (2009b).
  • Serga et al. (2012) A. A. Serga, C. W. Sandweg, V. I. Vasyuchka, M. B. Jungfleisch, B. Hillebrands, A. Kreisel, P. Kopietz,  and M. P. Kostylev, “Brillouin light scattering spectroscopy of parametrically excited dipole-exchange magnons,” Phys. Rev. B 86, 134403 (2012).
  • Farr et al. (2015) Warrick G. Farr, Maxim Goryachev, Jean-Michel le Floch, Pavel Bushev,  and Michael E. Tobar, “Evidence of dilute ferromagnetism in rare-earth doped yttrium aluminium garnet,” Applied Physics Letters 107, 122401 (2015), http://dx.doi.org/10.1063/1.4931432.
  • Goryachev et al. (2014c) Maxim Goryachev, Warrick G. Farr, Daniel L. Creedon,  and Michael E. Tobar, “Controlling a whispering-gallery-doublet-mode avoided frequency crossing: Strong coupling between photon bosonic and spin degrees of freedom,” Phys. Rev. A 89, 013810 (2014c).
  • le Floch et al. (2006) Jean-Michel le Floch, James David Anstie, Michael Edmund Tobar, John Gideon Hartnett, Pierre-Yves Bourgeois,  and Dominique Cros, “Whispering modes in anisotropic and isotropic dielectric spherical resonators,” Physics Letters A 359, 1 – 7 (2006).
  • Benmessai et al. (2009) Karim Benmessai, Michael Edmund Tobar, Nicholas Bazin, Pierre-Yves Bourgeois, Yann Kersalé,  and Vincent Giordano, “Creating traveling waves from standing waves from the gyrotropic paramagnetic properties of ions in a high- whispering gallery mode sapphire resonator,” Phys. Rev. B 79, 174432 (2009).
  • Krupka (1991) J. Krupka, “Measurements of all complex permeability tensor components and the effective line widths of microwave ferrites using dielectric ring resonators,” Microwave Theory and Techniques, IEEE Transactions on 39, 1148–1157 (1991).
  • Krupka et al. (1999) Jerzy Krupka, Stephen A Gabelich, Krzysztof Derzakowski,  and Brian M Pierce, “Comparison of split post dielectric resonator and ferrite disc resonator techniques for microwave permittivity measurements of polycrystalline yttrium iron garnet,” Measurement Science and Technology 10, 1004 (1999).
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description