Searching for Dark Matter with AION (Atom Interferometer Observatory and Network)

Sarah Alam Malik Imperial College London Dark matter Evidence from a variety of scales : – galaxy and galaxy clusters – large scale structure formaon – Cosmic Microwave Background

Dark matter

Dark energy

Nature of dark matter? One of the most fundamental questions in physics 2 Dark Matter : Landscape of candidates

Ultra-light Dark Matter WIMPs Primordial black holes

68 10−22 103 109 1012 10 mDM [eV]

!3 Dark Matter : Landscape of candidates

Ultra-light Dark Matter WIMPs Primordial black holes

68 10−22 103 109 1012 10 mDM [eV]

WIMPs very well motivated theoretically, independently from cosmology and particle physics Direct detection and colliders highly constrained large range of parameter space

!4 Ultra light Dark Matter (ULDM)

An attractive alternative to WIMPs : ULDM covers broad category of theoretical candidates

Ultra-light Dark Matter WIMPs Primordial black holes

68 10−22 103 109 1012 10 mDM [eV]

Axions - Proposed by Peccei and Quinn (PQ) as an elegant solution to the ‘strong CP problem’ : postulated a new global U(1) symmetry that is spontaneously broken at some large energy scale. - Consequence of this mechanism is a new pseudoscalar boson, the axion, which is the Nambu- Goldstone boson of the PQ symmetry. - This PQ symmetry explicitly broken at low energies, so axion acquires a small mass - Properties of axions closely related to those of neutral pions, two-photon interaction plays a key role for most searches

Axion-like particles (ALPs) - String theory predicts existence of axion-like massive scalar ﬁelds.

!5 Ultra light Dark Matter (ULDM)

Experiments searching for these include helioscope searches, Light-Shining-Through-Wall, and haloscope searches Oscillation probability

Helioscope Searches Light-Shining-Through-Wall

● Probability is given by

ggf coupling Form factor to take into account resonant modes

CERN Axion Solar Telescope (CAST) : uses a refurbished LHC test magnet (9T) pointed towards Sun

Nov 7, 2018 S Santpur 12 !6 Future projections Ultra light Dark Matter (ULDM)

These experiments cover a large range of phase space

Nov 7,But, 2018 ULDM candidates with masses in S the Santpur range 10 − 22 – 10 − 14 eV requires new approaches17

!7 Atom Interferometry

One of the emerging technologies that shows promise to probe this mass range is quantum sensors in the form of atom interferometer. • Exploit the wave-like behavior of matter. Typical de Broglie wavelength of atom much smaller than optical wavelength, atom interferometers are remarkably sensitive devices • Analogous to laser interferometers by dividing, reﬂecting and then recombining atomic wavepackets to produce an interference pattern.

beam splitter mirror beam splitter

8 Atom Interferometry

Current-generation interferometers based on two-photon Raman transitions driven by counter propagating beams

beam splitter mirror beam splitter |e> Energy

ω2 ω1

|2>

|1>

Momentum

๏ Atom starts in ground state |1>

๏ A π/2 optical Raman pulse splits atomic wavefunction into an equal superposition of |1⟩ and |2⟩ ๏ Transition driven by counter-propagating light beams with

frequencies ω1 and ω2 ; absorption of photon with

frequency ω1 and stimulated emission of frequency ω2 ๏ Momentum transfer of to the wave function ℏkeff component in state |1> 9 Atom Interferometry

beam splitter mirror beam splitter

๏ Two components propagate freely for a time T ๏ A π pulse swaps the internal states and exchanges momenta

10 Atom Interferometry

beam splitter mirror beam splitter ๏ A π/2 pulse closes interferometer sequence ๏ Produce interference fringes in the populations of |1> and |2> ๏ Accumulation of phase along two paths

Any efect that modiﬁes the energy across the 2 arms of the interferometer appears in this interferometer phase, this can be used to constrain new physics that couples to matter. 11 Atom Interferometer Observatory Network (AION)

A UK-led initiative proposing a series of multi-purpose atom interferometers with progressively increasing baselines to : • Explore well-motivated ultra-light dark matter candidates several orders of magnitude beyond current bounds;

• Explore mid-frequency band GWs from the very early Universe and astrophysical sources • Potential sensitivity to searches for new particles and fields such as fifth forces, dark energy, precision measurements of variations in fundamental constants, basic physical principles such as foundations of quantum mechanics and Lorentz invariance.

!12 Atom Interferometer Observatory Network (AION)

Core team

!13 Atom Interferometer Observatory Network (AION)

Stage 1 : AION-10 Construct 10m atom interferometer following a similar apparatus at Stanford as prototype

1. Strontium atoms are laser-cooled in a 2D magneto-optical trap (MOT) using the 461nm transition 2. Cooled in a 3D MOT using blue 461nm light and red 689nm light 3. Ultracold atoms are transported to interferometry chamber 4. Atoms are launched upwards 5. Atoms follow a parabolic freefall trajectory, during which interferometry sequence is performed. 14 6. Interference fringes are detected using an imaging system. Atom Interferometer Observatory Network (AION)

Stage 2 : AION-100

MAGIS-100

100m atom interferometer is planned that would be sensitive to ULDM - 2 ensembles of atom interferometers along a single vertical baseline. - Signal would be a differential phase shift between the two interferometers. - Potential sensitivity to ULDM in the mass range 10−22 – 10−14

Close collaboration on an international level with the US initiative, MAGIS-100, which pursues a similar goal of an eventual km-scale atom interferometer on a comparable timescale. 15 Ultra light scalar dark matter

Ultra-light spin 0 particles are expected to form a coherently oscillating classical ﬁeld

ϕ(t) = ϕ0cos(Eϕt/ℏ)

m2ϕ2 Energy density of ρ = 0 2

16 ASIMINA ARVANITAKI et al. PHYS. REV. D 97, 075020 (2018)

1 μ 1 2 2 single atomic species by exploiting the time-domain L μϕ ϕ − m ϕ 1 ¼þ2 ∂ ∂ 2 ϕ ð Þ response of a differential atomic interferometer to a scalar DM wave. The search strategy outlined below has a de μν discovery reach for scalar DM couplings that is potentially − 4πG ϕ d m ee¯ − F F ; 2 N m e e 4 μν ð Þ orders of magnitude better than existing constraints and ASIMINA ARVANITAKI et al. PHYS. REV. D 97, 075020 (2018) pffiffiffiffiffiffiffiffiffiffiffiffi proposals in its frequency band and is complementary to where we parametrized the leading interaction with electrons 1 1 the low-frequency, broadband strategies of Refs. [9,13] and μand photons2 2 relative to gravity as insingle Refs. atomic[17,18] species; G is by exploiting the time-domain L μϕ ϕ − m ϕϕ 1 N the high-frequency, resonant searches of Ref. [12]. ¼þ2 ∂ ∂ Newton2 ’s constant, so d dð Þ 1 wouldresponse be of the a couplings differential atomic interferometer to a scalar m e e DM wave. The search strategyAn outlined electronic below transition has a energy ωA depends on the values ASIMINA ARVANITAKI¼ ¼ et al. PHYS. REV. D 97, 075020 (2018) of a scalar graviton.de μν We employ unitsdiscovery in which reachℏ forc scalar1. DMof couplingsm e and thatα and is potentially so will oscillate itself in the presence of a − 4πGNϕ dm m eee¯ − FμνF ; 2 ¼ ¼ The couplingse 4 in Eq. (2) couldð originateÞ orders from of a magnitude Higgs portal better thanscalar existing DM wave: constraints and Ultra light scalar dark matter pffiffiffiffiffiffiffiffiffiffiffifficoupling of the form bϕ H 2, whichproposals is one in its ultraviolet frequency band and is complementary to L ⊃1 1 PHYSICAL REVIEW D 97, 075020 (2018) single atomic species by exploiting the time-domain where we parametrized the leading interaction with electronsj j μ 2 2 ASIMINA ARVANITAKIcompletion intoet a al.L renormalizableμϕ modelϕ −the low-frequency, withm ϕϕ a particularly broadband strategies1 of Refs.PHYS.[9,13] and REV. D 97, 075020 (2018) and photons relative to gravity as in Refs. [17,18]¼þ2; G is 2 ð Þω tresponseω ofω a differentialcos m t ; atomic interferometer6 to a scalar low cutoff [7]. Scalar fields∂ N with∂ quadraticthe high-frequency, couplings resonant to searches of Ref.A [12]≃. A Δ A ϕ ’ ð Þ þ ð Þ ð Þ Newton s constant, so dm e de 1 would be the couplings An electronic transition energy ω depends onDM the values wave. The search strategy outlined below has a matter¼ [19¼–21], e.g. L ⊃ ϕ2 H 2, give rise to analogous A of a scalar graviton.1 We employ units1 in which ℏ c 1. of m and α andde sosingle will oscillateμν atomic itself in species the presencediscovery by of exploiting a reach for the scalar time-domain DM couplings that is potentially L ϕ μϕ m 2 ϕ2 − ¼ 4¼πj Gj ϕ d em ee¯ 1− F F ; 2 The couplings in Eq. (2)signaturesμcould originate− as the fromϕ linear a Higgs couplings portalN butscalar havem e DM drasticallye wave: moreμν ¼þ2 ∂ ∂ 2 ð Þ 4 response of a differentialð Þ ΔωAorders atomicωA of4π magnitude interferometerGNϕ0 dm betterξd toe : athan scalar existing7 constraints and coupling of the form fine-tunedL ⊃ bϕ H 2 masses, which is for one the ultraviolet same physical effect, so we shall ≡ ð e þ Þ ð Þ j j pffiffiffiffiffiffiffiffiffiffiffiffi DM wave. The searchproposals strategy in outlined its frequency below band has and a is complementary to completion into a renormalizablenot consider model them with further. a particularly pffiffiffiffiffiffiffiffiffiffiffiffi Scalar fieldwhere we parametrizedde μν theElectron leadingω interactionA t ≃PhotondiscoveryωA withΔ ωA cos electrons reachm ϕt for; scalarthe DM low-frequency,6 couplings that broadband is potentially strategies of Refs. [9,13] and low cutoff [7]. Scalar− 4 fieldsπBosonicGNϕ withd DMm quadraticm mucheee¯ − couplings lighterFμν thanF to 1 eV; is a highly2 ð classicalÞ þ Above,ð Þ we have neglectedð Þ the vvir-suppressed spatial 2 2ande photons relative to gravitycoupling as incoupling Refs. [17,18]; G is matter [19–21], e.g. stateL ⊃ ϕ becauseH , give of rise high to analogous occupation4 numbersð andÞ canorders be ofvariation magnitudeN and better thethev high-frequency,2 than-suppressed existing temporal resonant constraints incoherence, searches and of Ref. [12]. signatures as the linear couplingsj j Newton but have drastically’s constant, more so d d 1 would be the couplings vir papproximatedffiffiffiffiffiffiffiffiffiffiffiffi by a nonrelativistic planem e wavee solutionω proposalsω to 4πGwhichϕ ind its we frequencyξ willd : restoreAn band electronic in7 our and final is transition complementary results. We energy assumedω toA depends a on the values fine-tuned masses for the same physical effect, so we shall ¼ ¼Δ A A N 0 m e e Bosonicwhere we DM parametrized withEq. mass(1) <<, the 1eVof leading a in scalar highly interaction graviton. classical We withstate, employ electrons approximated units in≡the whichby low-frequency, non-relativisticℏ ðc þ1. broadbandÞ of m ðand strategiesÞ α and so of will Refs. oscillate[9,13] and itself in the presence of a not consider them further. linear¼ ¼ dependence ofeωA on m e, valid to a high degree for and photons relative to gravityThe couplings as in Refs. in Eq.[17,18](2) could; G originateis fromp affiffiffiffiffiffiffiffiffiffiffiffi Higgs portal 21 planeBosonic wave DM solution: much lighter than 1 eV is a highly classical N the high-frequency,all (nonhyperfine) resonantscalar electronic DM searches wave: transitions. of Ref. For[12] the. 5s S ↔ Above, we2 have2 neglected the vvir-suppressed spatial 0 ’ ϕ t; x ϕ cos m t v · x β O v ; 2 3 3 stateNewton becauses constant,of high occupation so dm ecoupling numbersde0 of1 andwould theϕ can form− be theL ⊃ couplingsbϕ H , which isAn one electronic ultraviolet transition energy ω depends on the values ð Þ¼ ½ ð Þþvariationþ andðj j theÞ vvirð -suppressedÞ 5s5p temporalP0 transition incoherence, in Sr I, whichA we will take as a case approximated by a nonrelativisticcompletion plane¼ wave¼ solution into a renormalizable to j j model with a particularly of a scalar graviton. We employ units in which ℏwhichc we will1. restoreof m ine ourandstudy finalα and results. throughout, so Wewill assumed oscillate one has a ξ itself≈ 2.06 in the[22] presence. The dependence of a Eq. (1), with amplitude ϕ ≃ p2ρ =m determined¼ ¼ by the local ωA t ≃ ωA ΔωA cos m ϕt ; 6 The couplings in Eq. (2) couldlow cutoff originate0 [7]DM. from Scalarϕ a Higgs fieldslinear portal dependence with quadratic of ωA on couplingsmofe, valid other to to atomic a high degree transition for energiesð Þ on fundamentalþ ð con-Þ ð Þ 3 scalar DM wave: 21 where amplitude determinedDM energy by local density 2 ρDM ≈ 0.3 GeVall (nonhyperfine)=cm2 .2 The electronic local transitions. For the 5s S ↔ matter [19 21]2 , e.g. L ϕ H , give rise tostants analogous may be found0 in Refs. [23–25]. couplingDMϕ densityt; x ofϕ thecos formm t Lv ⊃· xbϕβH O,– whichffiffiffiffiffiffiffiffiffiffiffiv ; is3 one⊃ ultraviolet3 0 descriptionϕ − of Eq. (3) should be thought0.3 5s5p ofPj astransitionj an incoherent in Sr I, which we will take as a case completionð Þ¼ into½ a renormalizableð signaturesÞþj þj model asðj j theÞ with linearð Þ a couplingsparticularly0 but have drastically more superposition of waves (hence, ϕ0 study∝ pρ throughout,DM) that never- one has ξ ≈ 2.06 [22]. The dependence ω ω 4πG ϕ d ξd : 7 ωA t ≃ ωA ΔωA cosΔm ϕAt ; A N 0 6 m e e withlow amplitude cutoffϕ[7]0 ≃.p Scalar2ρDM=m ϕ fieldsfine-tuneddetermined with by masses quadratic the local for couplings theof same other atomicphysical to transition effect, energies so we shall on fundamental con- ð þ Þ ð Þ theless has a long phase3 coherence time of approximately ð Þ III.þ PHYSICALð EFFECTÞ≡ ð Þ DM energy density ρDM ≈ 0.3notGeV2 consider=cm2 . The them local further. matter [19 21],ffiffiffiffiffiffiffiffiffiffiffi e.g. L2 ϕ H , give−3 rise tostants analogousffiffiffiffiffiffiffiffiffi may be found in Refs. [23–25]. pffiffiffiffiffiffiffiffiffiffiffiffi description of Eq.– (3)2shouldπ=m ϕ bevvir thought⊃where ofv asvir an∼ incoherent10 is the Galactic virial velocity. The light-pulse atom interferometry scheme of Ref. [16], signatures as the linear couplingsBosonicj j but DM have much drastically lighter than more 1 eV is a highly classical Above, we have neglected the vvir-suppressed spatial superposition of wavesThe (hence, coherenceϕ0 ∝ p arisesρDM) that from never- the nonrelativistic nature of depicted in Fig. 1, is like a differential atomic clock, where state because of high occupationIII. numbers PHYSICAL and EFFECT canΔωA be ωAvariation4πGNϕ and0 dm the ξvd2e-suppressed: 7 temporal incoherence, thelessfine-tuned has a long masses phaseDM: coherence for the the angular same time frequency of physical approximately of effect, the wave so we is mostly shall set by the one laser≡ is referenced to twoð spatiallye þ virÞ separatedð atomicÞ 2π=m v2 where v 10−3 is theapproximated Galacticffiffiffiffiffiffiffiffiffi virial velocity. by!17 a nonrelativistic plane wave solution to which we will restore in our final results. We assumed a notϕ considervir themvir rest-mass∼ further. energy m ϕ. It receives smallThe light-pulse kinetic energy atom interferometryensembles. scheme In of thep Ref.ffiffiffiffiffiffiffiffiffiffiffiffi absence[16], of new physics and reducible The coherence arises from theEq. nonrelativistic(1), 2 nature of depicted in Fig. 1, is like a differential atomic clock,linear where dependence of ω on m , valid to a high degree for Bosonic DMcorrections much lighter of O thanm ϕv 1 eV, iswhich a highly do have classical a large spread:Above,backgrounds, we have neglected the phase response the vvir in-suppressed the atomicA ensembles spatiale is DM: the angular frequency of the wave isð mostlyvir setÞ by the one laser is referenced to two spatially separated atomic 21 state because ofv high∼ v occupationand v − v numbers2 ∼ v2 . and can be identical, but2 bothall (nonhyperfine) suffer from laser electronic noise imprinted transitions. onto For the 5s S0 ↔ rest-mass energy m ϕ. It receivesvir small kinetic energyvir ensembles. In thevariation absence of new and2 physics the v andvir-suppressed reducible temporal incoherence, hj ji hðϕ t; xh iÞ iϕ0 cos m ϕ t − v · x β O vthe atoms,; 3 especially3 at frequencies below 1 Hz. However, correctionsapproximated of O m byv2 The a, which nonrelativistic scalar do field haveð a DM large planeÞ¼ oscillations spread: wave½ solutionbackgrounds, ofð Eq. (3) tocombined theÞþ phasewhichþ response weðj j will inÞ the restoreð atomicÞ ensembles5 ins5 ourp P0 isfinaltransition results. in We Sr assumedI, which we a will take as a case ð ϕ virÞ vEq. v(1),and v withv 2 thev couplings2 . to matter of Eq. (2)identical,cause but fundamental both suffer fromthe laser differential noise imprinted atomicstudy onto phasethroughout, response one is insensitive has ξ ≈ 2. to06 laser[22]. The dependence ∼ vir − ∼ virwith amplitude ϕ ≃ p2ρ =m determinedlinear dependence by the local of ωA on m e, valid to a high degree for hj Theji scalar fieldhð DM“hconstantsi oscillationsÞ i ” such of Eq. as the(3) electroncombined0 massthe andDM atoms, the fine-structureϕ especially at frequenciesnoise below when 1 the Hz. atomsof However, other move atomic along transitionfree-fall geodesics energies21 when on fundamental con- all (nonhyperfine)3 electronic transitions. For the 5s S ↔ with the couplings toconstant matter of to Eq.DM oscillate(2) cause energy in fundamental time: density theρDM2 differential≈ 0.3 GeV atomic=cm phase. response Thenot manipulated local is insensitivestants by to laser may the laser. be found This in differential Refs. [230 phase25]. ϕ t; x ϕ0 cos m ϕ t − v · x β O ffiffiffiffiffiffiffiffiffiffiffiv ; 3 3 – “constants” ðsuch asÞ¼ the electron½ massdescriptionð and the fine-structureÞþ of Eq.þ(3) shouldðjnoisej Þ whenbeð thought theÞ atoms5 ofs5 move asp anP0 along incoherentresponsetransition free-fall can geodesics in serve Sr I, as when which a low-background we will take channel as a case to look constant to oscillate in time: not manipulated bystudy the throughout, laser. This differential one has phaseξ 2.06 [22]. The dependence msuperpositiont; x m 1 ofd waves4πG (hence,ϕ t; xϕ0 ∝ pρ4DM) thatfor new never- physics. A GW≈ would modulate the light travel with amplitude ϕ0 ≃ p2ρeðDM=mÞ¼ϕ determinede þ m e byresponse theN ð local canÞ serveof asð other aÞ low-backgroundtime atomic of the transition channel laser pulses to energies look between onIII. fundamental the PHYSICAL atomic ensembles, con- EFFECT theless has a long phase3 coherence time of approximately DMm energyt; x m density1 d ρ 4πG ϕ0.t;3xGeVh =cm4 . Thefor new local physics.i A GW would modulate the light travel e e m e DM ≈N 2 pffiffiffiffiffiffiffiffiffiffiffiffi−3 stantsffiffiffiffiffiffiffiffiffi mayleading be found to a in differential Refs. [23 phase25]. accumulation over the ð Þ¼ þ ffiffiffiffiffiffiffiffiffiffiffi2π=m ϕvðvir whereÞ vð virÞ ∼ time10 ofis the the laser Galactic pulses virial between velocity. the atomic ensembles,The light-pulse– atom interferometry scheme of Ref. [16], description of Eq. (3) shouldα t; x be thoughtα 1 d ofe as4 anπG incoherentNϕ t; x : 5 interferometer sequence [16]. h pTheðffiffiffiffiffiffiffiffiffiffiffiffi coherenceÞ¼ i þ arises fromleadingð theÞ to nonrelativistic a differentialð Þ phase nature accumulation of depicted over the in Fig. 1, is like a differential atomic clock, where superpositionα t; x ofα 1 wavesde 4π (hence,GNϕ t; x ϕ:h0 ∝ pρDM5 ) thatinterferometer never-i sequence [16]. Atomic sensors for GW detection are also intrinsically ð Þ¼ þ DM: theð Þ angular frequencyðpÞffiffiffiffiffiffiffiffiffiffiffiffi of the wave is mostly set byIII. the PHYSICALone laser is EFFECT referenced to two spatially separated atomic theless has a longTemporalh phase variation coherence ofim timee and ofα gives approximately riseAtomic to oscillations sensors for in GW detectionsensitive are to also modulus intrinsically DM waves, without change to the pffiffiffiffiffiffiffiffiffiffiffiffirest-mass energy m ϕ. It receives small kinetic energy ensembles. In the absence of new physics and reducible Temporal variation2 ofenergym e and andα gives length−3 rise to scales oscillations in atoms,ffiffiffiffiffiffiffiffiffi in phenomenasensitive to which modulus were DM waves,experimental without change configuration. to the Given that all phases in the 2π=m ϕvvir where vvir ∼ 10 is the Galactic virial2 velocity. The light-pulse atom interferometry scheme of Ref. [16], energy and length scalesrespectively in atoms,corrections phenomena exploited which of byO DM werem ϕv searchvirexperimental, which proposals do configuration. have using a large Givensequence spread: that allof Fig. phasesbackgrounds,1 cancel in the in absence the phase of response new physics, in the we atomic ensembles is respectivelyThe coherence exploited arises by DM from search the proposals nonrelativistic usingð sequenceÞ nature2 of2 of Fig. 1depictedcancel in absence in Fig. of1, new is like physics,identical, a differential we but bothatomic suffer clock, from where laser noise imprinted onto atomic-clockv pairs∼ v[9]vir andand resonant-massv − v ∼ detectorsvvir. [12]. keep track only of DM-induced phase accumulation Φ of atomic-clockDM: the angular pairs [9]Spatialand frequency resonant-mass variationhj ji of the leads detectors wave to oscillating, ish[12]ð mostly. h keepi chemistry-dependent setÞ i trackby the only ofone DM-inducedlaserthe is phase excited referenced accumulation atomicthe to state atoms,twoΦ of relativespatially especially to that separated at of frequencies the ground atomic state. below 1 Hz. However, Spatial variation leads to oscillating,The chemistry-dependent scalar field DMthe oscillations excited atomic of state Eq. relative(3) combined to that of the ground state. rest-mass energy m ϕ. It receives small kinetic energy ensembles. In the absence of new physics and reducible forces, whichwith can the be couplings looked for to with matter accelerometers of Eq. (2) cause[13] fundamental[Effects on photonthe differential propagation atomic from the phasede responsecoupling of is insensitive to laser forces, which can be looked for2 with accelerometers [13] [Effects on photon propagation from the de coupling of (seecorrections also Ref. [9] offor(seeO a tidal-forcem alsoϕvvir Ref.“,constants effect). which[9] for do” asuch tidal-force have as a the large effect). electronEq. (2) spread:in mass a background andbackgrounds, the scalar fine-structureEq. DM(2) the fieldin phase a can background benoise response ignored when scalar in the the atoms DM atomic field move ensembles can along be ignored free-fall is geodesics when ðWe willÞ show2 that2 the DM-induced temporal variation of since it does not change the speed nor frequency of the laser Wev will∼ showvvir thatand the DM-inducedv − v constant temporal∼ v to. variation oscillate of in time:since it does not changeidentical, the speed but nor both frequency suffernot of the from manipulated laser laser noise by imprinted the laser. onto This differential phase hj ji hð h iÞ i vir atomicThe transition scalar frequencies fieldatomic DM transition can oscillations be searched frequencies for of with Eq. can a (3) belightcombined searched in vacuum.] for with Betweenthe atoms, a timeslight especiallyt0 inand vacuum.]t1, this at amountsresponse frequencies Between to can times below servet0 and as 1 Hz.t a1, low-background this However, amounts to channel to look with the couplings to matter of Eq. (2) cause fundamental the differential atomic phasefor new response physics. is insensitive A GW would to laser modulate the light travel m e t; x m e 1 dm 4πGNϕ t; x 4 “constants” such as the electron massð andÞ¼ the fine-structureþ e noiseð whenÞ the atomsð Þ movetime along of the free-fall laser pulses geodesics between when the atomic ensembles, 075020-2 constant to oscillate in time: h pffiffiffiffiffiffiffiffiffiffiffiffinot075020-2 manipulatedi by theleading laser. to This a differential differential phase phase accumulation over the α t; x α 1 de 4πGNϕresponset; x : can serve5 asinterferometer a low-background sequence channel[16]. to look ð Þ¼ þ ð Þ ð Þ h for newi physics. A GWAtomic would modulate sensors for the GW light detection travel are also intrinsically m e t; x m e 1 dm e 4πGNϕ t; x pffiffiffiffiffiffiffiffiffiffiffiffi4 ð Þ¼ Temporalþ variationð of mÞe and αðgivesÞ risetime to oscillations of the laser in pulsessensitive between to modulus the atomic DM ensembles, waves, without change to the henergy andpffiffiffiffiffiffiffiffiffiffiffiffi length scalesi in atoms, phenomenaleading which to a were differentialexperimental phase accumulation configuration. over Given the that all phases in the α t; x α 1respectivelyde 4πGN exploitedϕ t; x : by DM5 searchinterferometer proposals using sequencesequence[16]. of Fig. 1 cancel in absence of new physics, we ð Þ¼ þ ð Þ ð Þ h atomic-clock pairs [9]i and resonant-massAtomic detectors sensors[12]. for GWkeep detection track only are of alsoDM-induced intrinsically phase accumulation Φ of pffiffiffiffiffiffiffiffiffiffiffiffi the excited atomic state relative to that of the ground state. Temporal variation of m eSpatialand α gives variation rise leads to oscillations to oscillating, in chemistry-dependentsensitive to modulus DM waves, without change to the energy and length scalesforces, in atoms, which phenomena can be looked which for were with accelerometersexperimental configuration.[13] [Effects Given on photon that all propagation phases in from the the de coupling of respectively exploited by(see DM also Ref.search[9] proposalsfor a tidal-force using effect).sequence of Fig. 1 cancelEq. in(2) absencein a background of new physics, scalar DM we field can be ignored atomic-clock pairs [9] andWe resonant-mass will show that detectors the DM-induced[12]. temporalkeep track variation only of of DM-inducedsince it does phase not change accumulation the speedΦ norof frequency of the laser Spatial variation leads toatomic oscillating, transition chemistry-dependent frequencies can be searchedthe excited for atomic with a statelight relative in vacuum.] to that of Between the ground times state.t0 and t1, this amounts to forces, which can be looked for with accelerometers [13] [Effects on photon propagation from the de coupling of (see also Ref. [9] for a tidal-force effect). Eq. (2) in a background scalar DM field can be ignored We will show that the DM-induced temporal variation of since it does not change075020-2 the speed nor frequency of the laser atomic transition frequencies can be searched for with a light in vacuum.] Between times t0 and t1, this amounts to

075020-2 ASIMINA ARVANITAKI et al. PHYS. REV. D 97, 075020 (2018)

!18 Courtesy of Jason Hogan Ultra light scalar dark matter

Based on: Arvanitaki et al., PRD 97, 075020 (2018).

Higgs portal b Coupling to electron Coupling to photon coupling

fϕ [Hz] fϕ [Hz] fϕ [Hz] 10-4 10-2 1 102 104 10-4 10-2 1 102 104 10-4 10-2 1 102 104

-5 1 AION-10m(initial) 1 AION-10m(initial) ) 10 ) AION-10m(initial) (goal (goal -2 Excluded -2 ] -7 10 10 -10m 10 -10m eV | [ Excluded

| Excluded AION-10m(goal) AION -9 AION e me -4 -4 b 10 d d 10 10 | | AION-100m(initial) -11 AION-100m(initial) -6 10-6 AION-100m(initial) 10 10 AION-100m(goal) AION-100m(goal) AION-100m(goal) 10-13 -8 10-8 10 AION-km AION-km AION-km 10-15 -10 10-10 10 coupling portal

- -17

-12 coupling photon -12 10 lcrncoupling electron 10 10 -19 AION-space Higgs 10 AION-space 10-14 10-14 AION-space 10-21 10-16 10-16 10-18 10-16 10-14 10-12 10-10 10-18 10-16 10-14 10-12 10-10 10-18 10-16 10-14 10-12 10-10

Scalar DM mass mϕ [eV] Scalar DM mass mϕ [eV] Scalar DM mass mϕ [eV]

Different conﬁgurations of the atom interferometer also open possibility to search for vector dark matter candidates

!19 Atom Interferometer Observatory Network (AION) Stage 3

Build a km scale atom interferometer for gravitational wave detection in the mid- band frequency range, not covered by LIGO or the future planned LISA detector

Mid-band science • Detect sources BEFORE they reach the high frequency band [LIGO, ET] • Optimal for sky localization: predict when and where events will occur (for multi- messenger astronomy)

Stage 4 : satellite-based (thousands of kilometres scale) detectors

!20 Atom Interferometer Observatory Network (AION)

AION-10: Stage 1 [year 1 to 3] § 1 & 10 m Interferometers & Site Development for 100m Baseline AION-100: Stage 2 [year 3 to 6] § 100m Construction & Commissioning

AION-KM: Stage 3 [ > year 6 ] § Operating AION-100 and planning for 1 km & Beyond AION-SPACE: Stage 4 [ after AION-KM ] § Space based version

Courtesy of Oliver Buchmueller !21 AION-10

Beecroft building, Oxford physics

The Beecroft in Oxford is the proposed site, with a backup at RAL (MICE Hall) in case show-stoppers are encountered.

Courtesy of Oliver Buchmueller !22 Beecroft building, Oxford physics

Ultra-low vibration

Adjacent laser lab reserved for AION use

Vertical space , 12m basement to ground ﬂoor

AION

Courtesy of Oliver Buchmueller !23 Summary

• The AION program is an ambitious multi-staged proposal to build a series of atom interferometers with the capability to probe : •Ultra light dark matter in regions of parameter space that is inaccessible to current experiments •Gravitational waves in the mid-band frequency range •Constrain fundamental constants

• AION foreseen as a staged programme: AION-10, AION-100, AION-KM and AION-SPACE

• Close collaboration with the US initiative, MAGIS-100 and eventual km- scale detectors

• A lot of details to be worked out - sensitivity studies, work on understanding and pushing the design parameters of the experiment underway but preliminary studies look very promising !24 BACKUP

!25 !26 Clock gradiometer

Excited state phase evolution:

Two ways for phase to vary: AION Seminar AION

Dark matter Buchmueller Gravitational wave O. O.

Each interferometer measures the change over time T

Laser noise is common-mode suppressed in the gradiometer

Graham et al., PRL 110, 171102 (2013). Arvanitaki et al., PRD 97, 075020 (2018). 26

!27