A Field Guide to Spin Liquids - arXivnomenology of spin liquids in order to make connection to past and future experiments. Before we embark on this, we would like to motivate this

A Field Guide to Spin Liquids

Johannes KnolleBlackett Laboratory, Imperial College London, London SW7 2AZ, United Kingdom

Roderich MoessnerMax-Planck-Institut fur Physik komplexer Systeme, Nothnitzer Str. 38, 01187 Dresden, Germany

(Dated: April 9, 2018)

Spin liquids are collective phases of quantum matter which have eluded discovery in correlatedmagnetic materials for over half a century. Theoretical models of these enigmatic topological phasesare no longer in short supply. In experiment there also exist plenty of promising candidate materialsfor their realisation. One of the central challenges for the clear diagnosis of a spin liquid has been toconnect the two. From that perspective, this review discusses characteristic features in experiment,resulting from the unusual properties of spin liquids. This takes us to thermodynamic, spectroscopic,transport, and other experiments on a search for traces of emergent gauge fields, spinons, MajoranaFermions and other fractionalised particles.

CONTENTS

I. The quest for spin liquids 1A. What is a spin liquid 2B. How to tell one, as a matter of principle. . . 3C. . . . and in practice 4D. The role of universality 4

II. How to start looking . . . 4A. . . . and where 5

III. Thermodynamics and transport 5A. Thermodynamics 6B. Longitudinal transport: thermal versus

charge 6C. Bulk-boundary correspondence and

quantization of currents 7D. Unconventional phase transitions 7

IV. Spectroscopy 8A. Inelastic neutron scattering and quasi-particle

kinematics 8B. Light scattering 9C. Local probes 9D. Bound states of fractionalized quasiparticles 10E. Statistics 10

V. Disorder and defect physics 11

VI. Discussion and future directions 11

Acknowledgements 12

References 12

I. THE QUEST FOR SPIN LIQUIDS

The search for spin liquids as fundamentally new statesof matter is a long-running quest1–4. Their occurence ininsulating magnets appears to be greatly facilitated by

frustrated interactions for which the corresponding clas-sical spin systems display a large ground state degener-acy because the local energetics cannot be minimized ina unique way5–10. For a long time, the central and defin-ing concept involved was a negative one – a (ground)state without any magnetic order – in contrast to theprevailing phases with spontaneous symmetry breakingcharacterised by local order parameters. The rejuve-nated interest in resonating valence bond (RVB) physicsthrough high-temperature superconductors11 focused at-tention on topological properties12–15.

Initially, theoretical ideas were centred around wave-functions, e.g. of the RVB type, but it took considerabletime until microscopic Hamiltonians realising such statesat isolated points16 or extended bona fide quantum spinliquid (QSL) phases were established2. The advent of ex-actly soluble model Hamiltonians16,17 has allowed to gainan unprecedented understanding of ground- and excitedstate properties of QSLs. Nowadays, the theory commu-nity has developed a remarkable capacity to invent elab-orate schemes with a plethora of different phenomenolo-gies but arguably with little guidance from experiment.At the same time, while the target space of interestingmodels has exploded, the arsenal of methods for theirdetection has not grown commensurably. Nonetheless,there has been a sustained materials physics effort cover-ing a huge number of magnetic compounds, and many apromising candidate systems have been unearthed, someof which have benefitted from an intense research effortfor clarifying their properties in considerable detail.

The aim of this review is to contribute towards redress-ing this balance. In particular, we discuss the rich phe-nomenology of spin liquids in order to make connectionto past and future experiments.

Before we embark on this, we would like to motivatethis programme with a few words about how it fits inwith the broader research landscape of modern condensedmatter physics. The search for spin liquids nowadaysforms part of the grand challenge of understanding exis-tence, scope and nature of physics beyond the standardtheory of Landau and spontaneous symmetry breaking.

arX

iv:1

804.

0203

7v1

[co

nd-m

at.s

tr-e

l] 5

Apr

201

8

2

As such, they fall in the field of topological condensedmatter physics under the headings of long-range entan-gled phases and topological order.

To this date, we have only one outstanding establishedclass of experimental systems (at least in terms of materi-als science and beyond one dimension) exhibiting a topo-logically ordered quantum phase with fractionalized ex-citations: this is the fractional quantum Hall effect. TheLaughlin state and its even more elaborate brethren18,19

have attracted much attention, most recently fuelled bythe dream of realising a quantum computer topologicallyprotected against decoherence20. This is nicely reviewedin Ref.21.

QSLs have the added attraction of accessing the vastspace of possible materials provided by the combinatorialrichness of the periodic table, the presence of sometimeslarge exchange energy scales, as well as a high degree oftuneability and being amenable to experimental probes,e.g. through the application of magnetic fields and theuse of neutron scattering. The central goal for the fore-seeable future therefore is an unambiguous identificationof a QSL phase. For this, new experiments, includingnew probes, may be needed, accompanied by a reliabletheoretical analysis framework.

Our approach here is to support this quest, but notby providing a complete introduction to quantum spinliquids for the expert. Rather, we present a compendiumof ideas to provide a broader overview, which can alsoact as a guide for the newcomer. Many reviews are avail-able with material which we do not cover here, e.g. com-prehensive reviews on spin liquids22 and frustrated mag-netism more generally23; about spin ice24; introductoryreviews to quantum spin liquids4,25–27; an overview overspin ice28 and Kitaev QSLs29 from a fractionalisationperspective, as well as a nice review of classical emer-gent Coulomb gauge fields30; and works focusing on thematerials aspects of pyrochlores31, herbertsmithite32 andKitaev material candidates33,34.

Physics is an experimental science. However, while dis-coveries are mostly driven by experiments, the resultinginsights are naturally preserved in the language of the-ory. The history of the search for spin liquids is thereforenaturally intertwined with what phenomena one ex- andincludes under this heading. Before we move on to thecore of this field guide, we provide a brief review of thebackground taxonomy.

The remaining sections are then devoted to spin liq-uid phenomenology. The ultimate ambition – to providea textbook, not unlike standard solid state physics text-books on conventional phases, on the behaviour of suchtopological phases – is a step too far for us, and we makea selection of phenomena which we feel are particularlyinstructive and/or realistically attainable.

A. What is a spin liquid

What is certainly true is that the meaning of the termhas shifted over the years. This is not an uncommonstate of affairs, driven not only by the human tendencyto adapt a definition to the requirements of the moment,but also by the fact that as the understanding of thesubtleties of the phenomenon advances, refinements tothe concepts follow.

A constitutive concept for a spin liquid is the absence ofmagnetic order of a system of interacting spins at temper-atures smaller than the interaction scale. This encodesthe idea of a phase beyond the Landau paradigm (whichcovers all forms of magnetic order), as well as the intu-ition that a ’liquid’ should be different from a ’solid’. Inthis sense, other forms of ordering of the spin degreesof freedom – such as nematic orders38,39 – also a prioridisqualify a system from being classified as a spin liquid.

This negative definition, about the absence of some-thing, continues to be the most ubiquitous and intuitive.Besides its practical limitation, to which we return be-low, it is nowadays considered to be somewhat too broad.For instance, it includes models – interesting in their ownright – which are considered somewhat too simple. Oneis a quantum paramagnet, such as the kagome latticeIsing model in a transverse field, which is connected con-tinuously to a high-temperature (classical) paramagneticphase. Another is the Shastry-Sutherland model, whosetwo spins per structural unit cell form a dimer at low tem-perature, producing a simple inert state which again isstraightforwardly connected to a high-temperature para-magnet but which can also display protected edge exci-tations of triplons at low temperatures40. One calls thisbroader class of disordered magnets ‘cooperative param-agnets’, to distinguish them from magnets disordered bythermal fluctuations at high temperatures, although thisnomenclature is by no means universally used.

A more ’modern’, positive definition involves listingconditions which a phase should meet to qualify as a spinliquid. This derives from advances in our understandingof what phases beyond the Landau paradigm can looklike, and applies these to spin systems.

Akin to the Mermin-Wagner theorem41, which forbidsspontaneous breaking of continuous symmetry (SSB) atfinite temperatures in dimensions d≤2, there is a rigorousresult for spin systems with short range interactions. Forsystems with half-odd integer spin per unit cell, henceproper Mott insulators, and without symmetry break-ing, the Lieb-Schultz-Mattis theorem states42 that theground state is either unique with gapless excitations ordegenerate with a gap to excitations. It establishes tosome level of mathematical rigour43,44 the possibility ofgapless QSLs or gapped ones with topological order.

Perhaps the crispest is to demand that the magnetshould at low temperatures be described by a topologi-cal field theory, such as Chern-Simons theory45. On theone hand, this is very restrictive, a priori ruling out gap-less spin liquids, and spin liquids with some additional

3

-2 0 2 h,h,0

Dy2Ti2O7

(a) (b)(c)

(d)

FIG. 1. (Color online) (a) The Curie-Weiss temperature for various levels of dilution in SCGOx is much larger than thefeatureless regime of the susceptibility (adapted from Ref.35). The schematic lattice structures are shown in the insets. (b) Theresidual entropy of spin ice and the frustrated pyrochlore lattice (after Ref.25. (c) Broad features in inelastic neutron scatteringof the kagome lattice material herbertsmithite36. (d) The emergent gauge field of 3d spin ice gives rise to pinch points in thestatic structure factor (left panel, neutron scattering experiment; right panel, theory; adapted from Ref.37).

ordered degrees of freedom46. On the other hand, thisexclusion is not arbitrary–trying to braid quasiparticlesin the presence of gapless Goldstone modes of a ferro-magnet, like for SU(2) quantum Hall Skyrmions47, doespresent an obstacle for envisaged quantum computationexperiments21.

However, in practise to qualify as a spin liquid, it maybe enough to bear in mind that ’some subset’ of thedegrees of freedom should look ’essentially topological’.This is more or less the attitude we will take for the re-mainder of this review. For the purposes of a field guide,we will therefore be looking for fractionalised excitationsand emergent gauge fields. The latter two are intimatelylinked together as standard spin flip excitations alwayslead to integer changes of the total spin. Therefore, exci-tations labeled by ’fractions’ of such quantum numbers,e.g. quasiparticles carrying half-integer spin, need to becreated in even numbers and once separated we can thinkof the emergent background gauge field as taking care ofthe global constraint. The connection holds particularlyin higher dimensional spin liquids, which are the focusof our field guide, but in d = 1 fractionalisation may ap-pear much simpler via domain wall excitations of orderedstates.

B. How to tell one, as a matter of principle. . .

The characterisation of a topological state of matterproceeds most easily via its global properties. Given theimportance that numerical simulations have played in ad-vancing the field, some of these – while appearing rathercomplex – are comparatively straightforwardly diagnosednumerically.

The topological order discovered by Wen and Niuposits that topological states have a degeneracy which

depends on the genus of the surface they live on48. Forinstance, the Laughlin state at filling fraction ν = 1/3is non-degenerate on a sphere, and threefold degenerateon a torus. This is intimately connected with the ex-istence of fractionalised quasiparticles. When a pair ofLaughlin quasiparticles of charge ±e/3 is created from aground state, and one member of the pair moves arounda non-contractible loop of the torus before annihilatingthe other, the system moves from one ground state toanother. Only once three such particles, and hence anelectron with unit charge, have made such a trajectorydoes the system return to the original ground state. In-deed, the connections between quantum Hall physics andquantum spin liquids can become remarkably detailed,such as in transfers of wavefunctions between the twosettings, see e.g.49.

Observing the topological ground state degeneracy canbe challenging50,51 but numerical methods like DMRGare well-suited for extracting an equally reliable quan-tifier for gapped QSLs. In analogy to long range orderof conventional phases the long-range entanglement canserve as an order parameter of topological phases52–54.The entanglement entropy of a ground state wave func-tion can be calculated from a reduced density operatorwith one part of the total degrees of freedom of a bi-partitioned system traced out (with a smooth bound-ary of length L separating the two regions). For gappedphases it follows a universal scaling form

S = cL− γ + ... (1)

The first term with a non-universal prefactor c is the’area’ law common to all gapped phases but the secondterm γ quantifies the long-range entanglement (it is inde-pendent of the length of the boundary of the partition).It is only nonzero in a topological phase and is directly re-lated to the emergent gauge structure of the QSL phase,

4

e.g. in a Z2 QSL γ = ln 254. More detailed analyses canthen yield information on the properties of fractionalisedexcitations and edge spectra55.

C. . . . and in practice

Above, we called these diagnostics ’comparativelystraightforward’ because the experimental situation isconsiderably less promising. The entanglement entropy– in particular, any subleading contribution to it – doesnot correspond to any natural measurement on a many-body quantum spin system. Also, putting even a two-dimensional magnet on a manifold of nontrivial topologysounds like a thought experiment par excellence, evenmore so than the idea of diagnosing a spin stiffness fora conventionally ordered magnet by twisting boundaryconditions.

The core aim of this review is to address precisely thequestion how to move forward from here. Lacking a silverbullet or smoking gun (or whatever alternative martialmetaphor one prefers to use) in experimental reality, oneneeds to make do with the probes that actually exist, andthink about how best to employ and combine them foran unambiguous identification of spin liquids.

D. The role of universality

Before we turn to this in more detail, we would liketo raise an additional ’ideological’ point of fundamen-tal importance which may at times be somewhat under-appreciated when making contact between theory andexperiment.

The tension arises from the need in theory to deviseprecise definitions. One of the most successful of these isthe idea of universality, which is intimately related to thesuccess of the developments of the concepts of symmetrybreaking and encoded in the renormalisation group halfa century ago. Phases and phase transitions have prop-erties which are independent of microscopic details of theHamiltonian – these properties are called universal.

The question then is: how much of this universalityis visible – and where/how – in an actual experiment.The fundamentalist answer is: a priori, nothing. A casein point is the existence of Goldstone modes accompany-ing the breaking of a continuous symmetry, in the limitof long wavelengths and low frequencies. Obviously, thelimit of low frequencies will be cut off by a finite energyresolution – not only due to Heisenberg’s uncertainty re-lations – of any conceivable experimental probe. On topof these, innumerable other limitations, many of thembased on nothing less than the second law of thermody-namics, will always be with us. We have to live withthem (but can at times perhaps, see Sec. V, even turnthem to our advantage).

In practise, this observation is not just a complaintaround the edges. As we will argue below, many of the

most striking manifestations of spin liquid behaviour infact are non-universal in the sense that they could be al-tered without leaving the phase; or conversely, that prox-imate phases may exhibit the behaviour we are interestedin essentially just as characteristically as the pristine ver-sion.

As a poster child of this, we would like to adduce thefractionalised Heisenberg chain, see Fig.3 (a). The agree-ment between theory and experiment is striking, up toconsiderable detail of the structure factor at high ener-gies, including subtle intensity variations with wavevec-tor and frequency. However, none of these are universal.Close inspection of the universal part of the spectrum atlow frequencies reveals the opposite56: due to the resid-ual coupling between neighbouring chains, they undergoan ordering transition into a different phase with differ-ent, ’conventional’ universal behaviour of a long rangeordered 3d magnet.

A fundamentalist ‘universalist’ perspective thereforeleaves two non-palatable alternatives: either one has thelow-temperature, conventional ordered phase; or, abovethe ordering transition, a phase continuously connectedto a boring high-temperature paramagnet. Both missthe remarkable, and in our minds convincing, evidencefor fractionalisation ‘in practise’ in this compound. Asis so often the case, while the worldview of the funda-mentalist is deceptively simple, much of what makes reallife interesting lies in the grey areas, the appreciation ofwhich requires an open mind.

II. HOW TO START LOOKING . . .

A time-honoured way of making a first cut at the di-agnosis of spin liquidity in a candidate material is viastudies of thermodynamic properties, Sect. III, in partbecause these are relatively easy to carry out locallyin a laboratory. The first chore is to establish the ab-sence of magnetic ordering in a strongly interacting (low-temperature) regime.

A popular measure for its existence is the frustrationparameter f = |ΘCW |/Tf 57,58, which is defined as theratio of two quantities. One is the Curie(-Weiss) temper-ature extracted from a straightforward fit to the high-temperature susceptibility, which to first order in a high-temperature expansion is given by χ = C/(T − ΘCW ).This expression applies to an insulating magnet the sizeand nature of whose magnetic moments determine theCurie constant C, and whose interactions determine thesize of ΘCW . The second quantity, Tf , is the locationof any non-analyticity (divergence, cusp, . . .) in χ, in-dicating a residual ordering tendency, or spin freezingwhich is commonly encountered in frustrated magnets.The regime in temperature ΘCW T > Tf , the co-operative paramagnetic regime, is then a natural placeto start looking for a spin liquid, and it is well definedprovided f is sufficiently large, see Fig.1 (a) for a classicexample.

5

If appropriate measurements are possible and available(e.g. thanks to the availability of a neutron source), theabsence of ordering can be confirmed by verifying thatno magnetic Bragg peaks appear when cooling down thesystem.

The challenge in practice lies in the need to eliminatethe presence of less obvious ordering tendencies (such asmultipolar or distortive order), which are more elusive inneutron scattering; and also not to miss any features inthe specific heat, which is nonspecific in the type of order-ings it picks up, other than requiring the correspondingnon-analytic features of the phase transition to be dis-cernible above its smooth background temperature de-pendence. Also, it may be polluted by ‘incidental’ phasetransitions, such as those of the lattice which have littlebearing on its magnetism.

Further valuable insights can be gleaned in spectro-scopic experiments, Sec. IV and Fig.3. These can provideconsiderably more detailed information than the purelymacroscopic thermodynamic ones: inelastic neutron scat-tering has made tremendous technological progress in thepast few years, and now routinely provides data in d+ 1(wavevector+frequency) space. Other probes, such asRaman scattering, and NMR lack wavevector dependenceand provide a local response averaged over the full sys-tem. Crucially, alternative probes each couple differentlyto the magnetic degrees of freedom, and therefore, theycome with different selection rules for probing the quasi-particles (or non-quasiparticle excitations) of the mate-rial thereby providing complementary evidence, as wediscuss below.

One central motivation for the use of finite-frequencyprobes lies in the fact that the ground states of topologi-cal systems are at first sight unspectacular. While gaplessspin liquids should generically come with algebraicallydecaying ground state correlations, gapped spin liquids(just like the Laughlin charge liquid) really look feature-less. It is the fractionalised quasiparticles which providea local indication of the topological physics involved: in-stead of the magnons in an ordered magnet, one looksfor spinons, holons, monopoles and the like. (An alter-native to studying such finite-frequency responses lies inenlisting the help of disorder to nucleate these excitationsalready in the ground state, Sec. V.)

For such excitations, kinematic considerations can playan important role: broad responses for a scattering exper-iment involving the creation of several particles, therebyweakening the restrictions imposed by energy and mo-mentum conservation, are taken as a prime indicator ofnovel spin liquid physics. In the case where the excita-tions of the emergent gauge field are very heavy – as inthe case of Kitaev’s QSL, where they do not move atall – they can effectively remove momentum conserva-tion as a kinematic constraint, as their energy is barelymomentum dependent59, Fig.1(c). In other cases, theremay be direct evidence for the emergent gauge field, asis provided by the pinchpoints in spin ice28, see Fig.1(d).

A. . . . and where

The search for spin liquid compounds has been goingon for a long time, and many compounds have yieldedinteresting insights and phenomena. To conclude thisintroduction, we mention some materials which haveshaped our own thinking – this article tries to synthe-size the resulting insights, rather than discuss and modeleach system – or indeed, all systems – in specific detail.

Kagome-based lattices are prime examples for geo-metric frustration and very prominent for quasi two-dimensional compounds. The jarosites60 represent afamily of compounds with varying degrees of further-neighbour interactions and disorder. Beyond this,two particularly well-studied systems are volborthite61,which is now believed to have an important spatialanisotropy62,63 and herbertsmithite64,65. The latter isin some ways an outstanding candidate32. Next in lineare triangular compounds, where a family of organic sys-tems (’dmits’) are particularly prominent66, as discussedin the thermodynamics section below.

In the last few years, candidate materials67,68 for Ki-taev spin liquids have received intense attention29,33,34,with a combination of mapping out the Hamiltonian andunderstanding the resulting consequences keeping a largecommunity busy. These include the ’Kitaev Iridates’A2 IrO3 with A=Na,Li69,70, as well as α-RuCl3

71,72, seeFig.3(b+c).

Historically very important – as the founding mate-rial of highly frustrated magnetism – has been SCGO57,a kagome-triangle-kagome trilayer, which may alterna-tively be viewed as a slab of the pyrochlore lattice ofcorner-sharing tetrahedra. The pyrochlore lattice in turnhosts the Ising magnet known as spin ice24,28, mostprominently or Dy/Ho3Ti2O7, the only generally ac-knowledged fractionalised magnetic material in three di-mensions. Many other compounds exist on this lattice,such as a more ’quantum’ version with Pr ions73, as wellas a large class of spinel compounds, including the well-studied Cr-spinels74 with, like SCGO, isotropic S = 3/2moments. Diluting a quarter of sites of that lattice inturn yields the hyperkagome lattice, with Na4Ir3O3 themost prominent exponent75.

Newcomers are constantly added to this list, mostrecently Ca10Cr702876, Ba3NiSb2O9

77, YbMgGaO478,

Cu2IrO379 and H3LiIr2O6

80.

III. THERMODYNAMICS AND TRANSPORT

A key target for diagnosing spin liquids is the experi-mental identification of fractionalized excitations at lowenergies. Thermodynamic and transport measurementsare complementary in this endeavour, the former prob-ing the necessary low energy density of states (DOS) andthe latter the mobility of the excitations. The absence ofstandard Goldstone modes from conventional symmetrybreaking phases, e.g. spin waves of an ordered magnet,

6

can be deduced from the the absence of non-analyticitiesin thermodynamic observables. In practice many mate-rial candidates have preemptive symmetry breaking in-stabilities leading to non-analyticities from sub-leadinginteractions. It prevents a true low temperature liquidphase for example due to weak interlayer couplings ofquasi-two-dimensional materials, but as long as the frus-tration parameter f is small enough the correlated para-magnetic regime at intermediate temperatures is a goodstarting point in the search for QSL physics.

Candidate spin liquids often exhibit a spectral weightdownshift of the specific heat as part of their refusal toorder. This can in the most extreme cases go as far asapparent violations of the third law of thermodynam-ics. This happens in spin ice, Fig.1 (b), where uponcooling a ’residual entropy’ is measured, which indicatesthat even at the lowest temperatures, the system contin-ues to explore an exponentially large number of states.This in particular sets cooperative paramagnetism apartfrom, say, dimensionality-induced destruction of order-ing. While purely one-dimensional spin systems such asa S = 1/2 Heisenberg antiferromagnetic chain do notorder at any nonzero temperature, they nonetheless areclose to an ordered state and often lose most of theirentropy already upon cooling through ΘCW .

A. Thermodynamics

Even though characteristic correlations of a QSL areonly expected at temperatures well below ΘCW , the hightemperature thermodynamic response of a candidate ma-terial already contains useful information. Details of amicroscopic description are obtained by a comparison ofthe temperature scale at which (and the form of how)the magnetic susceptibility deviates from Curie-Weiss be-haviour to that calculated in a high-temperature expan-sion of a putative spin Hamiltonian86,87. The angularmagnetic field dependence of χ with respect to crystalorientation is in principle able to detect spin-anisotropicinteractions88. The goal is to invert macroscopic mea-surements to microscopic descriptions which however isonly attainable as long as the low energy spin Hamilto-nian and the magnetic g-tensor are sufficiently simple.

A defining feature of QSLs are fractionalized magneticexcitations which, after subtracting other contributionsto the heat capacity (mainly from phonons, but at times,in particular at the lowest energies, nuclear spin), canbe probed via the low temperature dependence of ther-modynamic observables. In a gapless QSL informationabout a low energy power-law DOS, N(ω) ∝ ωα, canbe readily extracted because the specific heat is able todirectly probe the exponent α

CVT

=1

T

∂

∂T

∫dωωN(ω)n(ω, T ) ∝ Tα. (2)

Hence, in conjunction with the dimensionality of the sys-tem detailed information about the low energy dispersion

can be inferred such as the presence of emergent Fermisurfaces, Dirac points or nodal lines, all of which havebeen proposed in model QSLs89–91. Of course, the proce-dure rests on assumptions like the presence of weakly in-teracting quasi-particles whose thermal distribution onlydepends on their individual energies, for example n(ω, T )being the Fermi-Dirac or Bose-Einstein distribution func-tions.

Another caveat is that these power laws are not neces-sarily fixed in all circumstances. For instance. in a hon-eycomb system with slowly varying bond disorder theresulting hopping problem is equivalent to particles in arandom gauge field. Famously, this problem gives rise toa density of states with a power increasing continuouslywith disorder strength92. Such strains may quite conceiv-ably be present in, say, organic systems with relativelylow lattice rigidities.

Another intrinsic complication for a straightforward in-terpretation of thermodynamic data could be the pres-ence of very different energy scales making it hard to es-timate the right scaling regime. For example, the Diracspectrum of the honeycomb Kitaev QSL with N(ω) ∝ ωwould simply predict CV /T ∝ T which turns out to beonly observable at extremely low temperatures but in-stead over a large temperature window a metallic likeCV /T ∝ T 0 appears93,94. The reason is that spin flipexcitations fractionalize into Majorana fermions and fluxexcitations. The latter have a small gap which is onlya fraction of the total magnetic energy scale. At all butthe lowest temperatures the presence of thermally ex-cited fluxes destroys the Dirac spectrum of the Majoranafermions changing the low energy DOS to roughly a con-stant N(ω) ∝ ω0. This particular example shows, on theone hand, the difficulties of drawing reliable conclusionsfrom thermodynamic measurements. On the other hand,it highlights how a close comparison of microscopic calcu-lations and experimental data could be used in principleto extract complementary information about the frac-tionalized excitations of a QSL.

B. Longitudinal transport: thermal versus charge

In the absence of mobile charge carriers in magneticinsulating materials thermal transport experiments canprobe the mobility of elementary excitations. Ideally theheat flow along a temperature gradient contains informa-tion about the velocity, vk, and mean-free path, lMFP, offractionalized quasi-particles of a QSL with dispersionEk, e.g. as obtained for the longitudinal thermal con-ductivity κ in a semi-classical Boltzmann calculation

κ =∂

∂T

∫ddk lMFPEk|vk|n(Ek, T ). (3)

In a number of low dimensional QSL candidate mate-rials such purely magnetic contributions have been ob-served, e.g. in triangular82,95, kagome85 and Kitaev hon-eycomb96,97 systems. However, in general it is very hard

7

(a) (b) (c)

FIG. 2. (Color online) (a) Linear specific heat81 and residual linear in temperature longitudinal heat conductivity82 (inset) intriangular organic compound ’dmit’. (b) Linewidth in NMR experiment of SCGO due to orphan spins and their interactions83,84.(c) Evidence for a thermal Hall effect in the kagome volborthite85.

to get rid of the spurious phonon contribution. Moreover,in the presence of sizeable spin-phonon couplings mag-netic excitations scatter of phonon contributions and viceversa which makes it hard to disentangle the two. Theseproblems could in principle be overcome by studying di-rectly the spin current transport which is measurable viathe inverse spin Hall effect as demonstrated with insu-lating ordered magnets98. This has recently been pro-posed for QSL materials99 but more theoretical work andexperiments are needed to show whether spin transportmeasurements can be turned into a new tool for studyingQSLs.

C. Bulk-boundary correspondence andquantization of currents

A crowning achievement would be the observation ofquantised transport signatures directly related to topo-logical invariants, to rank alongside the famously quan-tised Hall conductivity of the fractional quantum Halleffect. Again, for insulating magnets, these cannot becharge transport, but nevertheless a quantization of thethermal Hall effect100, κxy, has been predicted in certaintypes of QSLs with broken time reversal symmetry101.The origin of the quantization17 can be illustrated for achiral QSL with ν chiral edge modes (their one dimen-sional dispersions labeled by momentum q connectingzero energy with the gapped bulk states) and at tem-peratures below the bulk gap ∆. The current is simplydetermined by the product of energy, thermal occupation(here for fermionic spinons obeying Fermi-Dirac statis-tics) and their velocity

Ixy = ν

∫ ∆

0

ε(q)n(ε)v(q)dq

2π(4)

= ν

∫ ∞0

ε(q)1

1 + eε(q)/Tdε

dq

dq

2π= ν

π

24T 2.

Recent experiments have observed signatures of a ther-mal Hall effect in disordered magnetic insulators, e.g. inkagome Volbortite85, pyrochlore compound Tb2Ti2O7

102

and the Kitaev candidate α-RuCl3103. However, the un-

ambiguous confirmation of the quantised prefactor of thetemperature dependence is missing because it is yet againdifficult to disentangle from, say, more prosaic types ofheat transport. In addition, not every spin liquid comeswith quantised transport coefficient, and at any rate, dif-ferent spin liquids may not be distinguishable in this wayalone. Nevertheless, the resulting problems of uniquenessand completeness of a classification scheme can safely bedeferred until a time that such quantised transport hasunambiguously been detected in at least two compounds.

D. Unconventional phase transitions

Despite the identification of a distinguishing character-istic of a QSL via its topological properties – the long-range entanglement of its ground state – this can be oflimited practical use as even some of the most paradig-matic states – among them classical spin ice, the Z2 gaugetheories or the Kitaev honeycomb model – are only zerotemperature phases and no phase transition occurs whencooling down from the simple high temperature param-agnet. Nevertheless, the behaviour of short length andtime scales [and here short could mean logarithmic insystem size104] can still be governed by the fractional-ized excitations of the zero temperature QSL105.

The phase transitions out of topological phases can beof considerable autonomous importance. Since the firstnon-symmetry breaking phase transition beyond the Lan-dau paradigm identified by Wegner for lattice gauge the-ories106,107, many other interesting proposals have beenmade. We mention these only in passing since pinningthese down in detail is even more challenging than iden-tifying the relevant phase in itself.

8

The basic attraction of such transitions is that they re-flect the exotic nature of the emergent degrees of freedom.For instance, the Kasteleyn transition108 is an asymmet-ric transition on account of the string-like nature of anemergent U(1) gauge field: when a string has a nega-tive free energy per unit length, it is suppressed by aBoltzmann factor in systems size as it needs to span theentire system. Hence, none – not even in the form offluctuations – are present until the sign of its free en-ergy changes, upon which there is a totally conventionalcontinuous onset of string density. In spin ice, wherethe topological spin state is most reliably established, (athermally rounded version of) this transition has indeedbeen observed109–111.

Spin ice also hosts a liquid-gas transition with a criticalendpoint of the emergent magnetic monopoles as zero-dimensional defects in a three-dimensional topologicalphase112. These form a Coulomb liquid which can thenbe treated with methods imported from electrochemistrysuch as Debye-Huckel theory and its extensions113–115.This is an instance of non-trivial collective behaviour ofthe emergent degrees of freedom, which in itself remainsa largely unexplored aspect of the field, Sec. V.

Much beautiful theory has been developed regardingsuch unusual phase transitions, including the identifi-cation of unusual signatures such as anomalously largeanomalous exponents. A particular case in point is thepossibility of deconfined quantum criticality116,117, wherethe critical point with deconfined excitations separatestwo symmetry-breaking confined phases2.

IV. SPECTROSCOPY

The absence of Bragg peaks in zero frequency measure-ments probing static correlations is an alternative indica-tor for ruling out conventional symmetry-breaking phasesin spin liquid candidate materials. The generic situa-tion of the correlations not only lacking a long-range or-dered component, but also being numerically short-rangemeans that, in reciprocal space, all features are broad.

Inelastic scattering experiments at nonzero frequencyhave the big advantage of also probing excited statesbeyond the asymptotic low energy regime of thermody-namic measurements. The apparent disadvantage thatthese are ’non-universal’ is remedied by the prospect ofidentifying concrete spin liquids in actual experiments.Many experimental probes with ever increasing frequencyresolution are available. Each of these has its well-developed strengths and also well-known set of shortcom-ings, to discuss all of which would go beyond the scopeof a simple review such as this, so that we will emphasizethe points specific to spin liquids in the following.

First of all, just like in the thermodynamic probes, onething to fundamentally look out for is an unusual temper-ature dependence of dynamical correlations118, see Fig.3(c) for a recent example. There should again be a co-operative paramagnetic regime in which interactions are

strong but response functions change little as the tem-perature is lowered.

A. Inelastic neutron scattering and quasi-particlekinematics

The method of choice for measuring the basic spin cor-relation functions – both static and dynamic – is inelasticneutron scattering whose cross-section is given by

dσ

dΩdE∝ F (q)

(δαβ −

qαqβ

q2

)× (5)∑

ri,rj

eiq·(ri−rj)∫

dtdt′e−iω(t−t′)〈Sαri(t)Sβrj

(t′)〉.

From the sum rules connecting static correlations to fre-quency integrated dynamical ones it is apparent thatthe absence of static Bragg peaks in spin liquids goesalong with the spectral weight being found elsewhere atnonzero frequencies.

A central role in pursuing spin liquids and their con-comitant fractionalisation is played by selection rules.Most simply put, if scattering involves a two-body pro-cess – e.g. a neutron spin flip creating a magnon – thetwin constraints of energy and momentum conservationcan lead to a sharp response in the form of a single line ofenergy versus wavevector transfer, ω(q) which representsthe magnon dispersion relation. This simple situation isfortuitous in that the selection rules for neutron scatter-ing off magnons permit precisely such a matrix element.The combined facts that neutrons are well-matched tolength/energy scales and quantum numbers of magnonsunderpins some of their phenomenal success in the fieldof magnetism.

A broader response can therefore have a number of dif-ferent origins. Firstly, there may not be such simple scat-tering processes available, such as in Raman scattering,where zero wavevector transfer q = 0 requires the cre-ation of multiple magnetic excitations. Secondly, theremay not be a simple magnon available, but rather onlyfractionalised excitations which have to be created to-gether, thereby rendering the scattering process a many-particle one. Thirdly, there may not be a quasiparticledescription of the low-energy spectrum in the first place,so that no dispersion relation ω(q) exists even as a matterof principle. This latter situation, especially with viewto gapless spin liquids, is perhaps the least understoodat this point in time.

Hence, while the existence of a broad dynamic scatter-ing response is a good indicator of a correlated paramag-netic regime, it is itself not a sufficient piece of evidencefor fractionalized excitations of a spin liquid. Neverthe-less, given the large amount of information available in afull ω(q) map, it opens up the possibility of providing amore microscopic modelling of the non-universal featuresdiscussed above.

9

For the Heisenberg chain, the kinematics of fractional-isation is most beautifully illustrated already in the threeparabolae which denote the possible energy-momentumcombinations of two independent domain walls obtainedfrom flipping a single spin in an ordered chain. The rathernontrivial intensity map – the kinematically allowed pro-cesses take place with very different intensities – followsfrom an actual enumeration of the matrix elements in-volved, see Fig.3(a).

In high dimensions, an analogous picture may e.g. ap-ply to a Z2 spin liquid where a triplet excitation can bedecomposed into a pair of S = 1/2 spinons, which wouldappear as monomers in a quantum dimer model119. Forthis to be straightforwardly visible, it would be necessaryfor the individual spinon to act like a coherent free par-ticle, which is likely the case in the limit of low spinonexcitation energies. However, it is at present unclear –and one of the most interesting questions – over whatenergy range such long-lived well-defined fractionalisedquasiparticles will exist.

This case of symmetric fractionalisation into two parti-cles is a particularly simple scenario. In addition, manyother emergent particles are possible, and it is a prioristraightforward to come up with mean-field parton con-structions with a wide variety of different fractionalisa-tion schemes89. One which is not uncommon is to end upwith a spin flip corresponding to the creation of a gauge-charged particle, and an excitation of the emergent gaugefield itself, e.g. as happens in the Kitaev models120,121.This fractionalisation can be very asymmetrical: the fluxmay be very heavy, that is to say, have a very small band-width. It can thus take up an arbitrary amount of mo-mentum at almost constant energy, and thereby rendermomentum conservation essentially inoperative59. Thisthen leads to features in reciprocal space which are sobroad that it is hard to infer much about the dispersionof the light particle. Nevertheless, key information aboutthe energy of flux excitations and the DOS of the lightfractionalized particles can in principle be inferred fromthe INS response120–122.

B. Light scattering

Light can interact with the purely magnetic low energydegrees of freedom of Mott insulators via the virtual hop-ping processes, which determine the magnetic exchangeconstant themselves. This has been shown for example tolead to a nonzero electric polarisability. It gives rise to anAC optical conductivity in Mott insulators127 signaturesof which have been analysed for a number of QSLs128–131.Moreover, optical absorption can directly couple via mag-netic dipole excitations to spin flips and thus becomessensitive to the zero momentum structure factor. Suchexperiments on a α-RuCl3 have been recently interpretedas indications of a magnetic field induced QSL132–134.

Alternatively, higher order photon process can in-duce virtual electron-hole pairs causing double spin flip

excitations135. Such dynamical Raman response ofkagome136,137 and Kitaev QSLs138,139 has been analysedtheoretically. Interestingly, the difference in matrix ele-ments compared to INS permits a more direct coupling tocertain types of fractionalized quasi-particles105 and thepolarization dependence of this zero momentum probecontains additional information140. Hence, Raman mea-surements on pyrochlores141, herbertsmithite142 and two-126 and three-dimensional143 Kitaev candiate materialshave been interpreted in terms of the spinon DOS ofQSLs but it is difficult to separate the inevitable phononcontribution to Raman scattering – we return to thispoint below, Sec. IV E. Finally, with a further increasein energy resolution resonant inelastic X-ray scattering(RIXS) will be a promising new tool for probing QSL ex-citations including their momentum dependence144,145.

C. Local probes

Nuclear magnetic resonance (NMR) experiments probethe local magnetic fields of the spin degrees of freedomin insulators via the hyperfine interaction with nuclearlevels. They are a powerful tool in the study of spin liq-uids146. For example, an NMR frequency, which remainssharp and does not split when cooling to low temper-ature, rules out the presence of static magnetism withinequivalent magnetic sites or a static disordered state.Even more information is obtained from the relaxationtime 1/T1T directly sensitive to the local magnetic sus-ceptibility (in the zero frequency limit) related to themagnetic DOS. In a gapped QSL an Arrhenius type be-haviour is expected but gapless QSLs would again leadto characteristic power law behaviour as a function oftemperature similar to the specific heat but without theparasitic phonon contributions.

An alternative probe of local magnetic fields is the re-laxation of spin-polarized muons deposited in a candidatematerial. These µSR experiments are able to reliably dis-tinguish between the presence of static moments due toconventional LRO, which leads to long lived oscillationsof the polarization, or dynamical moments of spin liq-uids, which lead to a quick decay without oscillations147.The main advantage of µSR is its high sensitivity but astraightforward interpretation of the data can be compli-cated because the positively charged muon interacts withthe lattice altering the local magnetic environment.

Unfortunately, experiments with both local and con-trolled spatial resolution are missing for magnetic insu-lators. For weakly correlated electronic materials thanksto scanning tunnelling microscopy (STM) and angle re-solved photo emission spectroscopy (ARPES) a hallmarksignature of topological systems – the bulk-boundary cor-respondence – could be confirmed shortly after its pre-diction, e.g. of the surface Dirac cone of three dimen-sional topological insulators148 or of Majorana zero en-ergy modes in superconducting wires149. Of course, fora direct confirmation of topological surface states in spin

10

(a)E

[meV

]

Wavevector

(b)

Ram

an In

tens

ity

E [meV]

Inte

grat

ed In

tens

ity

(c)

FIG. 3. (Color online) Comparison of inelastic scattering experiment and theory. (a) Fractionalised spinons in the S = 1/2Heisenberg chain material KCuF3

123,124. (b) Finite frequency neutron structure factor of the proximate spin liquid in α-RuCl3

125. (c) Fermionic nature of excitations in RuCl3 as evidenced in Raman scattering105,126.

liquid candidates without charged quasi-particles similarmeasurement tools are highly desirable. In that context itwill promising to explore new directions for example spinnoise spectroscopy, scanning SQUID magnetometry150,Raman microscopes151 or inelastic scanning tunnellingmicroscopy152 on spin liquid candidate thin films all withspatial resolution.

D. Bound states of fractionalized quasiparticles

Contrary to the intuition developed from the discus-sion of INS experiments, fractionalised quasiparticlesneed not have a broad, fully continuous spectrum. In-stead, they may form bound or localised states. In-deed, their quantum numbers may look a lot like thatof the unfractionalised spin flip; we remind the readerthat the statement of deconfinement of fractionalised par-ticles refers to the energy cost of their separation beingbounded; this does not preclude the possibility of discretecomposite states with a finite binding energy. Again,in d = 1 there is a celebrated and experimentally es-tablished instance of this153, when at the magnetic fieldtuned critical point the exchange field leads to a discretepart of the spectrum from bound pairs of domain wallexcitations. In higher dimension, analogous phenomenahave been theoretically proposed. In spin ice, like in thed = 1 case, application of a field can lead to bound statesof monopoles with a characteristic spectrum154.

More exotically, the possibility of the gauge field degreeof freedom being involved in a ground state is present innon-Abelian spin liquids. This is analogues to the caseof a px + ipy superconductor155 where vortices host Ma-jorana fermion bound states, e.g. the two Majoranas ofa pair of vortices lead to a fermionic bound state whoseenergy goes exponentially to zero with vortex separation.Of course, the long-term goal is the controlled manipu-lation of the degenerate ground state manifold for braid-

ing in the context of topological quantum computation21.Slightly less ambitious would be the observation of such aflux-Majorana bound state which has been shown to lead,for example, to a sharp contribution in the spin struc-ture factor121,122 in the non-Abelian phase of the Kitaevhoneycomb QSL in which a spin flip introduces a pairof nearest neighbour fluxes binding a pair of Majoranasbelow the gapped continuum response. Alternatively, al-ready in the Abelian but anisotropic Kitaev phases156,157,the leading response can be a sharp delta-function cor-responding to the addition of a pair of emergent gaugefluxes.

E. Statistics

The quantum statistics of quasiparticles is a very fun-damental property – it affects the many-body density ofstates even for non-interacting particles. This is alreadyevidenced by the (conventional) Fermi sphere and its sup-pressed heat capacity, and thus in principle accessiblein thermodynamic measurements already, Sec. III. Thetemperature dependence of dynamical scattering exper-iments contains information about the thermal distribu-tion functions which are qualitatively different for quasi-particles with different quantum statistics, e.g. Fermi-Dirac versus Bose-Einstein. For example, in the contextof Raman experiments on the Kitaev candidate materialα-RuCl3 a close comparison between the T-dependence ofthe high energy Raman response and experimental dataarguably points to the presence of spin fractionalisationin terms fermionic excitations105.

In general, fractionalised quasiparticles in topologicalphases can have unusual exchange statistics of Anyonsdue to the relative phases picked up when emergent par-ticles of different type are interchanged. Such braid-ing operations are particularly of interest given theirmuch-appreciated potential in procuring a framework

11

for fault-tolerant topological quantum computing20. Di-rectly probing the exchange statistics of emergent par-ticles in a quantum spin liquid is a tall order, which atpresent – given the difficulties in doing the same even inthe much more controlled setting of quantum interferom-eters in the quantum Hall effect – seems not too close athand.

However, there are nonetheless qualitative feature tolook out for. For instance, a scattering process whichcreates a set of fractionalised particles via a local inter-action is sensitive to the statistics of the particles gen-erated together, as their relative wavefunction influencesthe matrix elements for the process in question: for apointlike interaction, the creation of two Fermions, say,is inhibited by the vanishing of their wavefunction as thepair moves close together. Hence, under rather generalconditions for gapped QSLs the onset of the INS responseis dominated by the long range statistical interaction be-tween Anyonic quasiparticles leading to a universal powerlaw dependence as a function of frequency158. Similarly,a promising directions will be ambitious experiments fornoise spectroscopy directly probing statistics or measur-ing emergent quantum numbers of fractionalized quasi-particles.

V. DISORDER AND DEFECT PHYSICS

The presence of disorder – defects, vacancies, impuri-ties and the like – is an unavoidable fact of life in con-densed matter systems. Indeed, in many candidate spinliquids, understanding the role of disorder is a crucialstep towards the identification of the physics involved,see e.g.64,65,73,159.

Besides being a nuisance, however, disorder can also beused as a probe. The basic idea is that disorder can makefractionalisation physics visible in the ground state thatwould otherwise require probing excitations. As a simpleillustration, consider the following picture. A vacancy ina two-dimensional system can be thought of as insertinga microscopically tiny hole into the plane – in a sense, itchanges the topology of the plane into that of an annu-lus. This hole can then have effective degrees of freedom.One instance could occur in a non-Abelian spin liquid,where (well-separated) vacancies can host Majorana zeromodes at zero energy, as described in Sec.IV E. Also, notunlike impurities in a semiconductor, a vacancy can hosta localised fractionalised excitation – such as a magneticmonopole – which would otherwise require an activationenergy in the bulk160.

In this sense, disorder physics is closely related to ourdiscussion of dynamical probes of fractionalised degreesof freedom, upon replacing dynamical probes by probesof the disorder sites. Local probes can then be used to re-solve the signal coming from the defects. While in a bulkprobe, a signal from a small density of defects, unless itis singularly large, is easily swamped by the bulk signal,a local probe like NMR can detect the defect response

at a frequency separated from that of the bulk. Again,in one dimension, edge defects have provided a beautifulpicture of the physics of the gapped Haldane chain161.This kind of study has been carried out in some detailfor gauge-charged vacancy degrees of freedom83 follow-ing detailed NMR measurements on SCGO84, see Fig.2(b). These orphan spins35 establish an oscillating spintexture decaying like a Coulomb law; modelling this hasled to the conclusion that the level of disorder in SCGOis likely higher than that determined from the control-lable non-stochiometry only. The spin liquid response todisorder can thus be used to inverse-infer properties ofthe composition of the material itself.

The response to disorder can also be very intuitive.For a spinon Fermi surface, one can develop an analogyto the response of a Fermi liquid to disorder: an impu-rity can induce Friedel oscillations, leading to a distur-bance surrounding the impurity modulated at the Fermiwavevector, which in turn can give rise to RKKY-typeinteractions between impurities162.

As mentioned above, collective defect physics is a hugefield in its own right, so far little studied. A systematicunderstanding of the many-body state of a finite densityof defects embedded in a topological spin liquid seemslike a particularly promising direction for the discoveryof surprising new phenomena159,163–166. This topic is be-yond the scope of the present article, and arguably meritsa stand-alone treatment.

VI. DISCUSSION AND FUTURE DIRECTIONS

After several decades of searching for quantum spinliquids in magnetic materials we are still awaiting theunambiguous sighting of this elusive state of matter. En-couragingly, recent years have seen a flurry of discoveriesof new candidate materials and novel indicative signa-tures of liquidity, which raise the hopes that this longsearch will come to a successful conclusion in the nottoo distant future. One of the great attractions of find-ing a topological phase in a magnetic systems is theirhigh tunability. For instance, magnetic fields of a fewtens of teslas in strength are becoming increasingly rou-tinely available. They potentially push ’proximate’ spinliquid candidates72 in the desired direction167 and theimpressive advances in the energy-wavevector resolutionof neutron experiments, for example, enable a visuali-sation of the field induced melting of conventional longrange magnetism168. A magnetic field can not only adda Zeeman term to the Hamiltonian, but also acts in thepresence of spin-orbit interactions as a versatile probe,even mimicking an effectively staggered field on differentsublattices which can change the effective dimensionalityof the emergent gauge field28.

In the absence of uniquely and individually compellingfeatures of spin liquidity in a particular material, an in-vestigation of how a particular feature (gap size, modedispersion, continuum bandwidth, bound state energy)

12

changes as a function of tuning parameter (externalfields, pressure, composition, strain) will be a crucial in-gredient for assessing the validity of a particular interpre-tation of experimental data in terms of a spin liquid underthe merciless action of Ockham’s razor. In that context,a joint effort of both theory and experiment continuesto be called for pinning down the rich phenomenology ofspin liquids.

Besides such more detailed model-based input,methodological progress is also on the horizon. High onthe wishlist are improved local probes, especially witha resolution approaching the lattice scale, which is cur-rently elusive for, say, SQUID-based devices measuringlocal field distributions. These could then be used toprobe boundary modes or impurity susceptibilities evenin insulating magnets, in the hope of emulating theirhuge success in electronic systems with charge degrees offreedom via STM and ARPES. Another exciting devel-opment for probing fractionalized excitations lies in therealm of non-equilibrium techniques, e.g. pump-probemeasurements169,170 or spin echo and noise spectroscopy.The non-equilibrium physics of topological phases is anascent field that will surely hold numerous surprises forthe patient explorer.

Our field guide is also subject to continual extensionon account of the search for new materials. These mayarise in the form of bulk materials, metal organic frame-works171, metallic Kondo systems172, or in ‘artificial’ set-tings such as nanostructured/thin film samples. In addi-tion, there remains the promise173,174 of one day realis-ing new phases of spin systems in analogue cold atomicquantum simulators.

In conclusion of this little field guide, it is worth re-calling that in physics, the most interesting phenomenaare often the ones which are not anticipated at all, sothat the most basic suggestion remains to produce andanalyse experimental data with an open mind.

ACKNOWLEDGEMENTS

We are grateful to John Chalker, Dima Kovrizhin andShivaji Sondhi, with who our respective journeys intospin liquid territory commenced; and to all the other col-laborators and discussion partners since, too numerousto mention here, who have helped us shape the view ofthe field summarised above.

1 P.W. Anderson. Resonating valence bonds - new kindof insulator. Materials Research Bulletin, 8(2):153–160,1973.

2 R. Moessner and S. L. Sondhi. Resonating valence bondphase in the triangular lattice quantum dimer model.Phys. Rev. Lett., 86:1881–1884, Feb 2001.

3 Xiao-Gang Wen. Quantum Field Theory of Many-BodySystems. Oxford University Press, 2004.

4 Patrick A. Lee. An end to the drought of quantum spinliquids. Science, 321(5894):1306–1307, 2008.

5 G. H. Wannier. Antiferromagnetism. the triangular isingnet. Phys. Rev., 79:357–364, Jul 1950.

6 P. W. Anderson. Ordering and antiferromagnetism in fer-rites. Phys. Rev., 102:1008–1013, May 1956.

7 Jacques Villain. Insulating spin glasses. Zeitschrift furPhysik B Condensed Matter, 33(1):31–42, Mar 1979.

8 P. Chandra, P. Coleman, and A. I. Larkin. Ising transitionin frustrated heisenberg models. Phys. Rev. Lett., 64:88–91, Jan 1990.

9 Andrey Chubukov. Order from disorder in a kagome an-tiferromagnet. Phys. Rev. Lett., 69:832–835, Aug 1992.

10 R. Moessner and J. T. Chalker. Properties of a classicalspin liquid: The heisenberg pyrochlore antiferromagnet.Phys. Rev. Lett., 80:2929–2932, Mar 1998.

11 P. W. Anderson. The resonating valence bond state inla2cuo4 and superconductivity. Science, 235(4793):1196–1198, 1987.

12 V. Kalmeyer and R. B. Laughlin. Equivalence ofthe resonating-valence-bond and fractional quantum hallstates. Phys. Rev. Lett., 59:2095–2098, Nov 1987.

13 Steven A. Kivelson, Daniel S. Rokhsar, and James P.Sethna. Topology of the resonating valence-bond state:Solitons and high-Tc superconductivity. Phys. Rev. B,

35:8865–8868, Jun 1987.14 X. G. Wen. Vacuum degeneracy of chiral spin states

in compactified space. Phys. Rev. B, 40:7387–7390, Oct1989.

15 X. G. Wen. Mean-field theory of spin-liquid states withfinite energy gap and topological orders. Phys. Rev. B,44:2664–2672, Aug 1991.

16 Daniel S. Rokhsar and Steven A. Kivelson. Superconduc-tivity and the quantum hard-core dimer gas. Phys. Rev.Lett., 61:2376–2379, Nov 1988.

17 Alexei Kitaev. Anyons in an exactly solved model andbeyond. Annals of Physics, 321(1):2 – 111, 2006.

18 Gregory Moore and Nicholas Read. Nonabelions inthe fractional quantum hall effect. Nuclear Physics B,360(2):362 – 396, 1991.

19 N. Read and E. Rezayi. Beyond paired quantum hallstates: Parafermions and incompressible states in the firstexcited landau level. Phys. Rev. B, 59:8084–8092, Mar1999.

20 A Yu Kitaev. Fault-tolerant quantum computation byanyons. Annals of Physics, 303(1):2–30, 2003.

21 Chetan Nayak, Steven H. Simon, Ady Stern, MichaelFreedman, and Sankar Das Sarma. Non-abelian anyonsand topological quantum computation. Rev. Mod. Phys.,80:1083–1159, Sep 2008.

22 Yi Zhou, Kazushi Kanoda, and Tai-Kai Ng. Quantumspin liquid states. Rev. Mod. Phys., 89:025003, Apr 2017.

23 Claudine Lacroix, Philippe Mendels, and Frederic Mila,editors. Introduction to Frustrated Magnetism: Materi-als, Experiments, Theory (Springer Series in Solid-StateSciences). Springer, 2011 edition, January 2011.

24 Steven T. Bramwell and Michel J. P. Gingras. Spin icestate in frustrated magnetic pyrochlore materials. Sci-

13

ence, 294(5546):1495–1501, 2001.25 Roderich Moessner and Arthur P Ramirez. Geometrical

frustration. Phys. Today, 59(2):24, 2006.26 Leon Balents. Spin liquids in frustrated magnets. Nature,

464(7286):199–208, 2010.27 Frederic Mila. Quantum spin liquids. European Journal

of Physics, 21(6):499, 2000.28 C. Castelnovo, R. Moessner, and S.L. Sondhi. Spin ice,

fractionalization, and topological order. Annual Reviewof Condensed Matter Physics, 3(1):35–55, 2012.

29 M. Hermanns, I. Kimchi, and J. Knolle. Physics of the ki-taev model: Fractionalization, dynamic correlations, andmaterial connections. Annual Review of Condensed Mat-ter Physics, 9(1):17–33, 2018.

30 Christopher L. Henley. The coulomb phase in frustratedsystems. Annual Review of Condensed Matter Physics,1(1):179–210, 2010.

31 Jason S. Gardner, Michel J. P. Gingras, and John E.Greedan. Magnetic pyrochlore oxides. Rev. Mod. Phys.,82:53–107, Jan 2010.

32 M. R. Norman. Herbertsmithite and the search for thequantum spin liquid. Rev. Mod. Phys., 88:041002, Dec2016.

33 Jeffrey G. Rau, Eric Kin-Ho Lee, and Hae-Young Kee.Spin-orbit physics giving rise to novel phases in correlatedsystems: Iridates and related materials. Annual Reviewof Condensed Matter Physics, 7(1):195–221, 2016.

34 Stephen M Winter, Alexander A Tsirlin, Maria Daghofer,Jeroen van den Brink, Yogesh Singh, Philipp Gegenwart,and Roser Valenti. Models and materials for generalizedkitaev magnetism. Journal of Physics: Condensed Matter,29(49):493002, 2017.

35 P. Schiffer and I. Daruka. Two-population model foranomalous low-temperature magnetism in geometricallyfrustrated magnets. Phys. Rev. B, 56:13712–13715, Dec1997.

36 Tian-Heng Han, Joel S Helton, Shaoyan Chu, Daniel GNocera, Jose A Rodriguez-Rivera, Collin Broholm, andYoung S Lee. Fractionalized excitations in the spin-liquid state of a kagome-lattice antiferromagnet. Nature,492(7429):406, 2012.

37 T. Fennell, P. P. Deen, A. R. Wildes, K. Schmalzl,D. Prabhakaran, A. T. Boothroyd, R. J. Aldus, D. F. Mc-Morrow, and S. T. Bramwell. Magnetic coulomb phase inthe spin ice ho2ti2o7. Science, 326(5951):415–417, 2009.

38 Nic Shannon, Tsutomu Momoi, and Philippe Sindzingre.Nematic order in square lattice frustrated ferromagnets.Phys. Rev. Lett., 96:027213, Jan 2006.

39 Karlo Penc and Andreas M Lauchli. Spin nematic phasesin quantum spin systems. In Introduction to FrustratedMagnetism, pages 331–362. Springer, 2011.

40 PA McClarty, F Kruger, Tatiana Guidi, SF Parker, KeithRefson, AW Parker, Dharmalingam Prabhakaran, andRadu Coldea. Topological triplon modes and boundstates in a shastry–sutherland magnet. Nature Physics,13(8):736, 2017.

41 N. D. Mermin and H. Wagner. Absence of ferromag-netism or antiferromagnetism in one- or two-dimensionalisotropic heisenberg models. Phys. Rev. Lett., 17:1133–1136, Nov 1966.

42 Elliott Lieb, Theodore Schultz, and Daniel Mattis. Twosoluble models of an antiferromagnetic chain. Annals ofPhysics, 16(3):407–466, 1961.

43 Masaki Oshikawa. Commensurability, excitation gap, andtopology in quantum many-particle systems on a periodiclattice. Phys. Rev. Lett., 84:1535–1538, Feb 2000.

44 M. B. Hastings. Lieb-schultz-mattis in higher dimensions.Phys. Rev. B, 69:104431, Mar 2004.

45 S. C. Zhang, T. H. Hansson, and S. Kivelson. Effective-field-theory model for the fractional quantum hall effect.Phys. Rev. Lett., 62:82–85, Jan 1989.

46 M. E. Brooks-Bartlett, S. T. Banks, L. D. C. Jaubert,A. Harman-Clarke, and P. C. W. Holdsworth. Magnetic-moment fragmentation and monopole crystallization.Phys. Rev. X, 4:011007, Jan 2014.

47 S. L. Sondhi, A. Karlhede, S. A. Kivelson, and E. H.Rezayi. Skyrmions and the crossover from the integer tofractional quantum hall effect at small zeeman energies.Phys. Rev. B, 47:16419–16426, Jun 1993.

48 X. G. Wen and Q. Niu. Ground-state degeneracy of thefractional quantum hall states in the presence of a randompotential and on high-genus riemann surfaces. Phys. Rev.B, 41:9377–9396, May 1990.

49 Darrell F. Schroeter, Eliot Kapit, Ronny Thomale, andMartin Greiter. Spin hamiltonian for which the chiralspin liquid is the exact ground state. Phys. Rev. Lett.,99:097202, Aug 2007.

50 Simeng Yan, David A. Huse, and Steven R. White. Spin-liquid ground state of the s = 1/2 kagome heisenberg an-tiferromagnet. Science, 332(6034):1173–1176, 2011.

51 Hong-Chen Jiang, Hong Yao, and Leon Balents. Spinliquid ground state of the spin- 1

2square J1-J2 heisenberg

model. Phys. Rev. B, 86:024424, Jul 2012.52 Alioscia Hamma, Radu Ionicioiu, and Paolo Zanardi. Bi-

partite entanglement and entropic boundary law in latticespin systems. Phys. Rev. A, 71:022315, Feb 2005.

53 Alexei Kitaev and John Preskill. Topological entangle-ment entropy. Phys. Rev. Lett., 96:110404, Mar 2006.

54 Michael Levin and Xiao-Gang Wen. Detecting topologicalorder in a ground state wave function. Phys. Rev. Lett.,96:110405, Mar 2006.

55 L. Cincio and G. Vidal. Characterizing topological or-der by studying the ground states on an infinite cylinder.Phys. Rev. Lett., 110:067208, Feb 2013.

56 Bella Lake, D Alan Tennant, Chris D Frost, andStephen E Nagler. Quantum criticality and universalscaling of a quantum antiferromagnet. Nature materials,4(4):329, 2005.

57 X. Obradors, A. Labarta, A. Isalgu, J. Tejada, J. Ro-driguez, and M. Pernet. Magnetic frustration and latticedimensionality in srcr8ga4o19. Solid State Communica-tions, 65(3):189 – 192, 1988.

58 A P Ramirez. Strongly geometrically frustrated magnets.Annual Review of Materials Science, 24(1):453–480, 1994.

59 Matthias Punk, Debanjan Chowdhury, and SubirSachdev. Topological excitations and the dynamic struc-ture factor of spin liquids on the kagome lattice. NaturePhysics, 10(4):289, 2014.

60 A S Wills. Conventional and unconventional orderingsin the jarosites. Canadian Journal of Physics, 79(11-12):1501–1510, 2001.

61 Zenji Hiroi, Masafumi Hanawa, Naoya Kobayashi, Mi-noru Nohara, Hidenori Takagi, Yoshitomo Kato, andMasashi Takigawa. Spin-1/2 kagom-like lattice in volbor-thite cu3v2o7(oh)22h2o. Journal of the Physical Societyof Japan, 70(11):3377–3384, 2001.

14

62 T. Yavors’kii, W. Apel, and H.-U. Everts. Heisenberg an-tiferromagnet with anisotropic exchange on the kagomelattice: Description of the magnetic properties of volbor-thite. Phys. Rev. B, 76:064430, Aug 2007.

63 G. J. Nilsen, F. C. Coomer, M. A. de Vries, J. R. Stew-art, P. P. Deen, A. Harrison, and H. M. Rønnow. Paircorrelations, short-range order, and dispersive excitationsin the quasi-kagome quantum magnet volborthite. Phys.Rev. B, 84:172401, Nov 2011.

64 J. S. Helton, K. Matan, M. P. Shores, E. A. Nytko, B. M.Bartlett, Y. Yoshida, Y. Takano, A. Suslov, Y. Qiu, J.-H.Chung, D. G. Nocera, and Y. S. Lee. Spin dynamics of thespin-1/2 kagome lattice antiferromagnet zncu3(OH)6cl2.Phys. Rev. Lett., 98:107204, Mar 2007.

65 P. Mendels, F. Bert, M. A. de Vries, A. Olariu, A. Har-rison, F. Duc, J. C. Trombe, J. S. Lord, A. Amato, andC. Baines. Quantum magnetism in the paratacamite fam-ily: Towards an ideal kagome lattice. Phys. Rev. Lett.,98:077204, Feb 2007.

66 Y. Shimizu, K. Miyagawa, K. Kanoda, M. Maesato, andG. Saito. Spin liquid state in an organic mott insulatorwith a triangular lattice. Phys. Rev. Lett., 91:107001, Sep2003.

67 G. Jackeli and G. Khaliullin. Mott insulators in the strongspin-orbit coupling limit: From heisenberg to a quantumcompass and kitaev models. Phys. Rev. Lett., 102:017205,Jan 2009.

68 Ji rı Chaloupka, George Jackeli, and Giniyat Khaliullin.Kitaev-heisenberg model on a honeycomb lattice: Possibleexotic phases in iridium oxides A2iro3. Phys. Rev. Lett.,105:027204, Jul 2010.

69 Yogesh Singh and P. Gegenwart. Antiferromagnetic mottinsulating state in single crystals of the honeycomb latticematerial na2iro3. Phys. Rev. B, 82:064412, Aug 2010.

70 Yogesh Singh, S. Manni, J. Reuther, T. Berlijn,R. Thomale, W. Ku, S. Trebst, and P. Gegenwart. Rele-vance of the heisenberg-kitaev model for the honeycomblattice iridates a2iro3. Phys. Rev. Lett., 108:127203, Mar2012.

71 K. W. Plumb, J. P. Clancy, L. J. Sandilands, V. Vi-jay Shankar, Y. F. Hu, K. S. Burch, Hae-Young Kee,and Young-June Kim. α − RuCl3: A spin-orbit assistedMott insulator on a honeycomb lattice. Phys. Rev. B,90:041112, Jul 2014.

72 A. Banerjee, C. A. Bridges, J.-Q. Yan, A. A. Aczel, L. Li,M. B. Stone, G. E. Granroth, M. D. Lumsden, Y. Yiu,J. Knolle, S. Bhattacharjee, D. L. Kovrizhin, R. Moess-ner, D. A. Tennant, D. G. Mandrus, and S. E. Nagler.Proximate kitaev quantum spin liquid behaviour in a hon-eycomb magnet. Nature Materials, 15:733–740, 2016.

73 J.-J. Wen, S. M. Koohpayeh, K. A. Ross, B. A. Trump,T. M. McQueen, K. Kimura, S. Nakatsuji, Y. Qiu, D. M.Pajerowski, J. R. D. Copley, and C. L. Broholm. Disor-dered route to the coulomb quantum spin liquid: Randomtransverse fields on spin ice in pr2zr2o7. Phys. Rev. Lett.,118:107206, Mar 2017.

74 S.-H. Lee, C. Broholm, T. H. Kim, W. Ratcliff, and S-W.Cheong. Local spin resonance and spin-peierls-like phasetransition in a geometrically frustrated antiferromagnet.Phys. Rev. Lett., 84:3718–3721, Apr 2000.

75 Yoshihiko Okamoto, Minoru Nohara, Hiroko Aruga-Katori, and Hidenori Takagi. Spin-liquid state in thes = 1/2 hyperkagome antiferromagnet na4ir3o8. Phys.Rev. Lett., 99:137207, Sep 2007.

76 Christian Balz, Bella Lake, Johannes Reuther, HubertusLuetkens, Rico Schonemann, Thomas Herrmannsdorfer,Yogesh Singh, ATM Nazmul Islam, Elisa M Wheeler,Jose A Rodriguez-Rivera, et al. Physical realization of aquantum spin liquid based on a complex frustration mech-anism. Nature Physics, 12(10):942, 2016.

77 JG Cheng, G Li, L Balicas, JS Zhou, JB Goodenough,Cenke Xu, and HD Zhou. High-pressure sequence ofba 3 nisb 2 o 9 structural phases: New s= 1 quan-tum spin liquids based on ni 2+. Physical review letters,107(19):197204, 2011.

78 Joseph AM Paddison, Marcus Daum, Zhiling Dun, GeorgEhlers, Yaohua Liu, Matthew B Stone, Haidong Zhou,and Martin Mourigal. Continuous excitations of thetriangular-lattice quantum spin liquid ybmggao4. NaturePhysics, 13(2):117–122, 2017.

79 Mykola Abramchuk, Cigdem Ozsoy-Keskinbora, Jason WKrizan, Kenneth R Metz, David C Bell, and FazelTafti. Cu2iro3: a new magnetically frustrated honey-comb iridate. Journal of the American Chemical Society,139(43):15371–15376, 2017.

80 K Kitagawa, T Takayama, Y Matsumoto, A Kato,R Takano, Y Kishimoto, S Bette, R Dinnebier, G Jackeli,and H Takagi. A spin–orbital-entangled quantum liquidon a honeycomb lattice. Nature, 554(7692):341, 2018.

81 Satoshi Yamashita, Takashi Yamamoto, YasuhiroNakazawa, Masafumi Tamura, and Reizo Kato. Gap-less spin liquid of an organic triangular compoundevidenced by thermodynamic measurements. Naturecommunications, 2:275, 2011.

82 Minoru Yamashita, Norihito Nakata, Yoshinori Senshu,Masaki Nagata, Hiroshi M Yamamoto, Reizo Kato,Takasada Shibauchi, and Yuji Matsuda. Highly mobilegapless excitations in a two-dimensional candidate quan-tum spin liquid. Science, 328(5983):1246–1248, 2010.

83 Arnab Sen, Kedar Damle, and Roderich Moessner.Fractional spin textures in the frustrated magnetsrcr9pga12−9po19. Phys. Rev. Lett., 106:127203, Mar 2011.

84 L. Limot, P. Mendels, G. Collin, C. Mondelli, B. Oulad-diaf, H. Mutka, N. Blanchard, and M. Mekata. Sus-ceptibility and dilution effects of the kagome bilayergeometrically frustrated network: A ga nmr study ofsrcr9pga12−9po19. Phys. Rev. B, 65:144447, Apr 2002.

85 Daiki Watanabe, Kaori Sugii, Masaaki Shimozawa, Yoshi-taka Suzuki, Takeshi Yajima, Hajime Ishikawa, Zenji Hi-roi, Takasada Shibauchi, Yuji Matsuda, and Minoru Ya-mashita. Emergence of nontrivial magnetic excitationsin a spin-liquid state of kagome volborthite. Proceedingsof the National Academy of Sciences, 113(31):8653–8657,2016.

86 Jaan Oitmaa, Chris Hamer, and Weihong Zheng. Seriesexpansion methods for strongly interacting lattice models.Cambridge University Press, 2006.

87 Andre Lohmann, Heinz-Jurgen Schmidt, and JohannesRichter. Tenth-order high-temperature expansion for thesusceptibility and the specific heat of spin-s heisenbergmodels with arbitrary exchange patterns: Application topyrochlore and kagome magnets. Phys. Rev. B, 89:014415,Jan 2014.

88 K. A. Modic, Tess E. Smidt, Itamar Kimchi, Nicholas P.Breznay, Alun Biffin, Sungkyun Choi, Roger D. John-son, Radu Coldea, Pilanda Watkins-Curry, Gregory T.McCandless, Julia Y. Chan, Felipe Gandara, Z. Islam,

15

Ashvin Vishwanath, Arkady Shekhter, Ross D. McDon-ald, and James G. Analytis. Realization of a three-dimensional spin-anisotropic harmonic honeycomb iri-date. Nature Communications, 5:4203, June 2014.

89 Xiao-Gang Wen. Quantum orders and symmetric spinliquids. Phys. Rev. B, 65:165113, Apr 2002.

90 Michael Hermele, T. Senthil, Matthew P. A. Fisher,Patrick A. Lee, Naoto Nagaosa, and Xiao-Gang Wen. Sta-bility of u(1) spin liquids in two dimensions. Phys. Rev.B, 70:214437, Dec 2004.

91 Kevin O’Brien, Maria Hermanns, and Simon Trebst. Clas-sification of gapless Z2 spin liquids in three-dimensionalKitaev models. Phys. Rev. B, 93:085101, Feb 2016.

92 Andreas W. W. Ludwig, Matthew P. A. Fisher,R. Shankar, and G. Grinstein. Integer quantum hall tran-sition: An alternative approach and exact results. Phys.Rev. B, 50:7526–7552, Sep 1994.

93 Joji Nasu, Masafumi Udagawa, and Yukitoshi Motome.Vaporization of kitaev spin liquids. Phys. Rev. Lett.,113:197205, Nov 2014.

94 Joji Nasu, Masafumi Udagawa, and Yukitoshi Motome.Thermal fractionalization of quantum spins in a kitaevmodel: Temperature-linear specific heat and coherenttransport of majorana fermions. Phys. Rev. B, 92:115122,Sep 2015.

95 Minoru Yamashita, Norihito Nakata, Yuichi Kasahara,Takahiko Sasaki, Naoki Yoneyama, Norio Kobayashi,Satoshi Fujimoto, Takasada Shibauchi, and Yuji Mat-suda. Thermal-transport measurements in a quantumspin-liquid state of the frustrated triangular magnet-(bedt-ttf) 2 cu 2 (cn) 3. Nature Physics, 5(1):44, 2009.

96 Ian A. Leahy, Christopher A. Pocs, Peter E. Siegfried,David Graf, S.-H. Do, Kwang-Yong Choi, B. Normand,and Minhyea Lee. Anomalous thermal conductivityand magnetic torque response in the honeycomb magnetα−rucl3. Phys. Rev. Lett., 118:187203, May 2017.

97 Richard Hentrich, Anja UB Wolter, Xenophon Zotos,Wolfram Brenig, Domenic Nowak, Anna Isaeva, ThomasDoert, Arnab Banerjee, Paula Lampen-Kelley, David GMandrus, et al. Large field-induced gap of kitaev-heisenberg paramagnons in α-rucl3. arXiv preprintarXiv:1703.08623, 2017.

98 Y Kajiwara, K Harii, S Takahashi, Jun-ichiro Ohe,K Uchida, M Mizuguchi, H Umezawa, H Kawai, K Ando,K Takanashi, et al. Transmission of electrical signals byspin-wave interconversion in a magnetic insulator. Nature,464(7286):262, 2010.

99 Shubhayu Chatterjee and Subir Sachdev. Probing exci-tations in insulators via injection of spin currents. Phys.Rev. B, 92:165113, Oct 2015.

100 Hosho Katsura, Naoto Nagaosa, and Patrick A. Lee. The-ory of the thermal hall effect in quantum magnets. Phys.Rev. Lett., 104:066403, Feb 2010.

101 Hyunyong Lee, Jung Hoon Han, and Patrick A. Lee. Ther-mal hall effect of spins in a paramagnet. Phys. Rev. B,91:125413, Mar 2015.

102 Max Hirschberger, Jason W. Krizan, R. J. Cava, andN. P. Ong. Large thermal hall conductivity of neutralspin excitations in a frustrated quantum magnet. Science,348(6230):106–109, 2015.

103 Y Kasahara, K Sugii, T Ohnishi, M Shimozawa, M Ya-mashita, N Kurita, H Tanaka, J Nasu, Y Motome,T Shibauchi, et al. Thermal hall effect in a kitaev spin liq-uid: A possible signature of majorana chiral edge current.

arXiv preprint arXiv:1709.10286, 2017.104 Claudio Castelnovo and Claudio Chamon. Entanglement

and topological entropy of the toric code at finite temper-ature. Phys. Rev. B, 76:184442, Nov 2007.

105 Joji Nasu, Johannes Knolle, Dima L. Kovrizhin, YukitoshiMotome, and Roderich Moessner. Fermionic responsefrom fractionalization in an insulating two-dimensionalmagnet. Nature Physics, 12:912–915, 2016.

106 Franz J Wegner. Duality in generalized ising models andphase transitions without local order parameters. Journalof Mathematical Physics, 12(10):2259–2272, 1971.

107 John B. Kogut. An introduction to lattice gauge theoryand spin systems. Rev. Mod. Phys., 51:659–713, Oct 1979.

108 Pieter W Kasteleyn. Dimer statistics and phase tran-sitions. Journal of Mathematical Physics, 4(2):287–293,1963.

109 R. Moessner and S. L. Sondhi. Theory of the [111] mag-netization plateau in spin ice. Phys. Rev. B, 68:064411,Aug 2003.

110 T Fennell, ST Bramwell, DF McMorrow, P Manuel, andAR Wildes. Pinch points and kasteleyn transitions inkagome ice. Nature Physics, 3(8):566, 2007.

111 L. D. C. Jaubert, J. T. Chalker, P. C. W. Holdsworth,and R. Moessner. Three-dimensional kasteleyn transition:Spin ice in a [100] field. Phys. Rev. Lett., 100:067207, Feb2008.

112 Claudio Castelnovo, Roderich Moessner, and Shivaji LSondhi. Magnetic monopoles in spin ice. Nature,451(7174):42, 2008.

113 David A. Huse, Werner Krauth, R. Moessner, and S. L.Sondhi. Coulomb and liquid dimer models in three di-mensions. Phys. Rev. Lett., 91:167004, Oct 2003.

114 C. Castelnovo, R. Moessner, and S. L. Sondhi. Debye-huckel theory for spin ice at low temperature. Phys. Rev.B, 84:144435, Oct 2011.

115 Vojtech Kaiser, Jonathan Bloxsom, Laura Bovo, Steven TBramwell, Peter CW Holdsworth, and Roderich Moess-ner. Emergent electrochemistry in spin ice: Debye-h\”uckel theory and beyond. arXiv preprint arXiv:1803.04668,2018.

116 T Senthil, Ashvin Vishwanath, Leon Balents, SubirSachdev, and Matthew PA Fisher. Deconfined quantumcritical points. Science, 303(5663):1490–1494, 2004.

117 Anders W. Sandvik. Evidence for deconfined quantumcriticality in a two-dimensional heisenberg model withfour-spin interactions. Phys. Rev. Lett., 98:227202, Jun2007.

118 C. Broholm, G. Aeppli, G. P. Espinosa, and A. S. Cooper.Antiferromagnetic fluctuations and short-range order in akagome lattice. Phys. Rev. Lett., 65:3173–3176, Dec 1990.

119 Roderich Moessner and Kumar S Raman. Quantum dimermodels. In Introduction to Frustrated Magnetism, pages437–479. Springer, 2011.

120 J. Knolle, D. L. Kovrizhin, J. T. Chalker, and R. Moess-ner. Dynamics of a two-dimensional quantum spin liquid:Signatures of emergent majorana fermions and fluxes.Phys. Rev. Lett., 112:207203, May 2014.

121 Johannes Knolle. Dynamics of a Quantum Spin Liquid(Springer Theses). Springer International Publishing, 1edition, 2016.

122 J. Knolle, D. L. Kovrizhin, J. T. Chalker, and R. Moess-ner. Dynamics of fractionalization in quantum spin liq-uids. Phys. Rev. B, 92:115127, Sep 2015.

16

123 B. Lake, D. A. Tennant, J.-S. Caux, T. Barthel,U. Schollwock, S. E. Nagler, and C. D. Frost. Multispinoncontinua at zero and finite temperature in a near-idealheisenberg chain. Phys. Rev. Lett., 111:137205, Sep 2013.

124 Martin Mourigal, Mechthild Enderle, Axel Klpperpieper,Jean-Sbastien Caux, Anne Stunault, and Henrik M. Rn-now. Fractional spinon excitations in the quantum heisen-berg antiferromagnetic chain. Nature Physics, 9:435–441,2013.

125 Arnab Banerjee, Jiaqiang Yan, Johannes Knolle, Craig ABridges, Matthew B Stone, Mark D Lumsden, David GMandrus, David A Tennant, Roderich Moessner, andStephen E Nagler. Neutron scattering in the proximatequantum spin liquid α-rucl3. Science, 356(6342):1055–1059, 2017.

126 Luke J. Sandilands, Yao Tian, Kemp W. Plumb, Young-June Kim, and Kenneth S. Burch. Scattering continuumand possible fractionalized excitations in α−rucl3. Phys.Rev. Lett., 114:147201, Apr 2015.

127 L. N. Bulaevskii, C. D. Batista, M. V. Mostovoy, and D. I.Khomskii. Electronic orbital currents and polarization inmott insulators. Phys. Rev. B, 78:024402, Jul 2008.

128 Tai-Kai Ng and Patrick A. Lee. Power-law con-ductivity inside the mott gap: Application toκ−(BEDT−TTF)2cu2(CN)3. Phys. Rev. Lett., 99:156402,Oct 2007.

129 Andrew C. Potter, T. Senthil, and Patrick A. Lee. Mecha-nisms for sub-gap optical conductivity in herbertsmithite.Phys. Rev. B, 87:245106, Jun 2013.

130 Yejin Huh, Matthias Punk, and Subir Sachdev. Opticalconductivity of visons in Z2 spin liquids close to a valencebond solid transition on the kagome lattice. Phys. Rev.B, 87:235108, Jun 2013.

131 Adrien Bolens, Hosho Katsura, Masao Ogata, and SeijiMiyashita. Mechanism for sub-gap optical conductiv-ity in honeycomb kitaev materials. arXiv preprintarXiv:1711.00308, 2017.

132 A. Little, Liang Wu, P. Lampen-Kelley, A. Banerjee,S. Patankar, D. Rees, C. A. Bridges, J.-Q. Yan, D. Man-drus, S. E. Nagler, and J. Orenstein. Antiferromagneticresonance and terahertz continuum in α−rucl3. Phys.Rev. Lett., 119:227201, Nov 2017.

133 Zhe Wang, S. Reschke, D. Huvonen, S.-H. Do, K.-Y.Choi, M. Gensch, U. Nagel, T. Ro om, and A. Loidl.Magnetic excitations and continuum of a possibly field-induced quantum spin liquid in α−rucl3. Phys. Rev. Lett.,119:227202, Nov 2017.

134 C Wellm, J Zeisner, A Alfonsov, AUB Wolter, M Roslova,A Isaeva, T Doert, M Vojta, B Buchner, and V Kataev.Signatures of low-energy fractionalized excitations in α-rucl 3 from field-dependent microwave absorption. arXivpreprint arXiv:1710.00670, 2017.

135 Thomas P. Devereaux and Rudi Hackl. Inelastic lightscattering from correlated electrons. Rev. Mod. Phys.,79:175–233, Jan 2007.

136 O. Cepas, J. O. Haerter, and C. Lhuillier. Detectionof weak emergent broken-symmetries of the kagome an-tiferromagnet by raman spectroscopy. Phys. Rev. B,77:172406, May 2008.

137 Wing-Ho Ko, Zheng-Xin Liu, Tai-Kai Ng, and Patrick A.Lee. Raman signature of the u(1) dirac spin-liquid statein the spin- 1

2kagome system. Phys. Rev. B, 81:024414,

Jan 2010.

138 J. Knolle, Gia-Wei Chern, D. L. Kovrizhin, R. Moessner,and N. B. Perkins. Raman scattering signatures of kitaevspin liquids in A2iro3 iridates with a = Na or li. Phys.Rev. Lett., 113:187201, Oct 2014.

139 Brent Perreault, Johannes Knolle, Natalia B. Perkins,and F. J. Burnell. Theory of raman response in three-dimensional kitaev spin liquids: Application to β- andγ−li2iro3 compounds. Phys. Rev. B, 92:094439, Sep 2015.

140 Brent Perreault, Johannes Knolle, Natalia B. Perkins, andF. J. Burnell. Resonant raman scattering theory for ki-taev models and their majorana fermion boundary modes.Phys. Rev. B, 94:104427, Sep 2016.

141 M. Maczka, M. L. Sanjuan, A. F. Fuentes, K. Hermanow-icz, and J. Hanuza. Temperature-dependent raman studyof the spin-liquid pyrochlore tb2ti2o7. Phys. Rev. B,78:134420, Oct 2008.

142 Dirk Wulferding, Peter Lemmens, Patric Scheib, JensRoder, Philippe Mendels, Shaoyan Chu, Tianheng Han,and Young S. Lee. Interplay of thermal and quantumspin fluctuations in the kagome lattice compound herbert-smithite. Phys. Rev. B, 82:144412, Oct 2010.

143 A. Glamazda, P. Lemmens, S. H. Do, Y. S. Choi, andK. Y. Choi. Raman spectroscopic signature of fraction-alized excitations in the harmonic-honeycomb iridates β-and γ-li2iro3. Nature Communications, 7:12286, 2016.

144 Gabor B. Halasz, Natalia B. Perkins, and Jeroen van denBrink. Resonant inelastic x-ray scattering response of thekitaev honeycomb model. Phys. Rev. Lett., 117:127203,Sep 2016.

145 W. M. H. Natori, M. Daghofer, and R. G. Pereira. Dy-namics of a j = 3

2quantum spin liquid. Phys. Rev. B,

96:125109, Sep 2017.146 Pietro Carretta and Amit Keren. Nmr and µsr in highly

frustrated magnets. In Introduction to Frustrated Mag-netism, pages 79–105. Springer, 2011.

147 Alain Yaouanc and Pierre Dalmas De Reotier. Muonspin rotation, relaxation, and resonance: applications tocondensed matter, volume 147. Oxford University Press,2011.

148 David Hsieh, Dong Qian, Lewis Wray, YuQi Xia, Yew SanHor, Robert Joseph Cava, and M Zahid Hasan. A topolog-ical dirac insulator in a quantum spin hall phase. Nature,452(7190):970, 2008.

149 V. Mourik, K. Zuo, S. M. Frolov, S. R. Plissard, E. P.A. M. Bakkers, and L. P. Kouwenhoven. Signatures of ma-jorana fermions in hybrid superconductor-semiconductornanowire devices. Science, 336(6084):1003–1007, 2012.

150 Denis Vasyukov, Yonathan Anahory, Lior Embon, DorriHalbertal, Jo Cuppens, Lior Neeman, Amit Finkler,Yehonathan Segev, Yuri Myasoedov, Michael L Rappa-port, et al. A scanning superconducting quantum inter-ference device with single electron spin sensitivity. Naturenanotechnology, 8(9):639, 2013.

151 Brent Perreault, Stephan Rachel, F. J. Burnell, andJohannes Knolle. Majorana landau level raman spec-troscopy. arXiv:1612.03951, 2016.

152 J. Fransson, O. Eriksson, and A. V. Balatsky. Theory ofspin-polarized scanning tunneling microscopy applied tolocal spins. Phys. Rev. B, 81:115454, Mar 2010.

153 R. Coldea, D. A. Tennant, E. M. Wheeler, E. Wawrzyn-ska, D. Prabhakaran, M. Telling, K. Habicht, P. Smeibidl,and K. Kiefer. Quantum criticality in an ising chain: Ex-perimental evidence for emergent e8 symmetry. Science,327(5962):177–180, 2010.

17

154 Yuan Wan and Oleg Tchernyshyov. Quantum strings inquantum spin ice. Phys. Rev. Lett., 108:247210, Jun 2012.

155 D. A. Ivanov. Non-abelian statistics of half-quantum vor-tices in p-wave superconductors. Phys. Rev. Lett., 86:268–271, Jan 2001.

156 A. Smith, J. Knolle, D. L. Kovrizhin, J. T. Chalker, andR. Moessner. Neutron scattering signatures of the 3d hy-perhoneycomb kitaev quantum spin liquid. Phys. Rev. B,92:180408, Nov 2015.

157 A. Smith, J. Knolle, D. L. Kovrizhin, J. T. Chalker, andR. Moessner. Majorana spectroscopy of three-dimensionalkitaev spin liquids. Phys. Rev. B, 93:235146, Jun 2016.

158 Siddhardh C. Morampudi, Ari M. Turner, Frank Poll-mann, and Frank Wilczek. Statistics of fractionalizedexcitations through threshold spectroscopy. Phys. Rev.Lett., 118:227201, May 2017.

159 Lucile Savary and Leon Balents. Disorder-induced quan-tum spin liquid in spin ice pyrochlores. Phys. Rev. Lett.,118:087203, Feb 2017.

160 Arnab Sen and R. Moessner. Topological spin glass indiluted spin ice. Phys. Rev. Lett., 114:247207, Jun 2015.

161 M. Hagiwara, K. Katsumata, Ian Affleck, B. I. Halperin,and J. P. Renard. Observation of s=1/2 degrees of free-dom in an s=1 linear-chain heisenberg antiferromagnet.Phys. Rev. Lett., 65:3181–3184, Dec 1990.

162 David F. Mross and T. Senthil. Charge friedel oscillationsin a mott insulator. Phys. Rev. B, 84:041102, Jul 2011.

163 R. N. Bhatt and P. A. Lee. Scaling studies of highly dis-ordered spin-1/2 antiferromagnetic systems. Phys. Rev.Lett., 48:344–347, Feb 1982.

164 Kedar Damle, Olexei Motrunich, and David A. Huse. Dy-namics and transport in random antiferromagnetic spinchains. Phys. Rev. Lett., 84:3434–3437, Apr 2000.

165 A. J. Willans, J. T. Chalker, and R. Moessner. Site di-lution in the kitaev honeycomb model. Phys. Rev. B,84:115146, Sep 2011.

166 Itamar Kimchi, Adam Nahum, and T Senthil. Valencebonds in random quantum magnets: Theory and appli-cation to ybmggao4. arXiv preprint arXiv:1710.06860,2017.

167 Richard Hentrich, Anja U. B. Wolter, Xenophon Zotos,Wolfram Brenig, Domenic Nowak, Anna Isaeva, ThomasDoert, Arnab Banerjee, Paula Lampen-Kelley, David G.Mandrus, Stephen E. Nagler, Jennifer Sears, Young-June Kim, Bernd Buchner, and Christian Hess. Unusualphonon heat transport in α−rucl3: Strong spin-phononscattering and field-induced spin gap. Phys. Rev. Lett.,120:117204, Mar 2018.

168 Arnab Banerjee, Paula Lampen-Kelley, Johannes Knolle,Christian Balz, Adam Anthony Aczel, Barry Winn, Yao-hua Liu, Daniel Pajerowski, Jiaqiang Yan, Craig ABridges, et al. Excitations in the field-induced quantumspin liquid state of α-rucl 3. npj Quantum Materials,3(1):8, 2018.

169 MPM Dean, Yue Cao, X Liu, S Wall, D Zhu, RomanMankowsky, V Thampy, XM Chen, JG Vale, D Casa,et al. Ultrafast energy-and momentum-resolved dynamicsof magnetic correlations in the photo-doped mott insula-tor sr 2 iro 4. Nature materials, 15(6):601, 2016.

170 Zhanybek Alpichshev, Fahad Mahmood, Gang Cao, andNuh Gedik. Confinement-deconfinement transition as anindication of spin-liquid-type behavior in na 2 iro 3. Phys-ical review letters, 114(1):017203, 2015.

171 Masahiko G. Yamada, Hiroyuki Fujita, and Masaki Os-hikawa. Designing kitaev spin liquids in metal-organicframeworks. Phys. Rev. Lett., 119:057202, Aug 2017.

172 S. Nakatsuji, Y. Machida, Y. Maeno, T. Tayama,T. Sakakibara, J. van Duijn, L. Balicas, J. N. Milli-can, R. T. Macaluso, and Julia Y. Chan. Metallic spin-liquid behavior of the geometrically frustrated kondo lat-tice pr2ir2o7. Phys. Rev. Lett., 96:087204, Mar 2006.

173 L.-M. Duan, E. Demler, and M. D. Lukin. Controllingspin exchange interactions of ultracold atoms in opticallattices. Phys. Rev. Lett., 91:090402, Aug 2003.

174 Salvatore R. Manmana, E. M. Stoudenmire, Kaden R. A.Hazzard, Ana Maria Rey, and Alexey V. Gorshkov. Topo-logical phases in ultracold polar-molecule quantum mag-nets. Phys. Rev. B, 87:081106, Feb 2013.