Large 2D Coulomb crystals in a radio frequency surface ion trap B. Szymanski,∗ R. Dubessy,† B. Dubost,‡ S. Guibal, J.-P. Likforman, and L. Guidoni Univ Paris Diderot, Sorbonne Paris Cit´e, Laboratoire Mat´eriaux et Ph´enom`enes Quantiques, UMR 7162, Bat. Condorcet, 75205 Paris Cedex 13, France (Dated: January 13, 2012) We designed and operated a surface ion trap, with an ion-substrate distance of 500 µm, realized withstandardprinted-circuit-boardtechniques. Thetraphasbeenloadedwithuptoafewthousand Sr+ ions in the Coulomb-crystal regime. An analytical model of the pseudo-potential allowed us to determine the parameters that drive the trap into anisotropic regimes in which we obtain large (N >150) purely 2D ion Coulomb crystals. These crystals may open a simple and reliable way to experiments on quantum simulations of large 2D systems. Applications of laser cooled trapped ions are numer- Finally, the surface point Paul trap geometry [20], ide- 2 1 ous and cover several research fields such as quantum ally suited to vary the trapping distance from the sub- 0 information processing [1–3], quantum simulation [4, 5], strate, can be a promising candidate for quantum sim- 2 cold molecule spectroscopy [6] and metrology [7]. In the ulation [21]. Up to ten ions in a 2D cristalline arrange- n frame of quantum information processing, a large-scale menthavebeenloadedinthiskindoftrapsrealizedwith a computer architecture has been proposed based on ion printed circuit board technology and operated in a cryo- J shuttling between interaction and memory zones [8]. A genic environment[22]. 2 very practical way to realize such an architecture relies In this paper we present a linear surface rf ion trap 1 on surface electrode radio-frequency (rf) traps [9, 10] in basedonastandardprintedcircuitboardandwedemon- which a pseudo-potential well is created above the sur- strate the versatility of such a device that allows for the ] h face of a substrate by a set of deposited metallic elec- trapping of large crystallized ion ensembles. Depending p trodes. The vast majority of the surface traps developed on the trap parameters, different crystal shapes can be - so far [11, 12] are devoted to the trapping and shuttling obtained. In particular, we demonstrated the formation t n of short ion strings (1D ion Coulomb crystals). However of single-layer Coulomb crystals containing more than a a peculiar characteristic of the planar geometry, never 150 ions. Previous works have proposed and realized u exploitedtodate,isitsintrinsicanisotropythatcanlead similar devices [21, 23] in which the presence of stray q to the creation of single-layer 2D ion Coulomb crystals. fields [23] or the chosen trap geometry [21] probably [ As suggested by Porras and Cirac [13], such crystals are prevented the formation of large Coulomb crystals. 1 well adapted to simulate quantum phase transitions in v spin systems. In particular, the control of anisotropy is We used a copper FR4 printed circuit board on which 4 ideally suited for the study of zigzag transition instabili- strip-lines, forming the ion trap electrodes, were chemi- 8 5 ties directly related to one-dimensional Ising models in a cally etched and gold-plated (thickness < 1 µm) using 2 transverse field [14, 15]. standard commercial procedures. The board material . In order to create purely 2D laser-cooled ion lattices, hasalreadyproventobeUHVcompatibleandbakableup 1 three different strategies have been considered so far. to150◦C[24]. Thefivewiretrapgeometryispresentedin 0 2 One relies on Penning traps. It allowed for the obser- Fig. 1. The longitudinal confinement is assured by four 1 vation of structural phase transitions [16] and has been ”endcap”electrodes. AnoscillatingpotentialV cos(ωt) rf : morerecentlyusedforquantumcontrolexperiments[17]. with typically V = 125 V and ω/2π = 6.9 MHz is ap- v rf Themaindisadvantageofthisstrategyisthedifficultyof plied to the rf electrodes. Static voltages in the range i X laser-cooling that has to deal with both magnetron and -5V to +5V are typically used to drive the central con- r cyclotron motions [17]. trol electrode (V ), the two lateral control electrodes CC a Radio-frequency trap arrays are also a very promising (V ) and the four endcaps (V ). LC EC scheme [18]. The main advantage of this strategy is the Followingreference25,weperformedananalyticalcal- possibility to design regular lattice structures that are culationofthepseudo-potentialassociatedtothispartic- not imposed by the self-arrangement. However, the ex- ulartrapgeometry. Thecalculationgivesususefulinfor- perimentalinter-iondistancesare,uptonow,quitelarge mationsuchastheionmotionalfrequenciesasafunction and do not allow for sufficient ion-ion interaction [19]. of the trap parameters V , V , V and V , thus rf CC LC EC determining the trap axial and transverse anisotropies. The ion distance from the trap surface (504µm, imposed by the geometry), the trap depth and the stability pa- ∗ [email protected] rameters are also obtained. Using an approach similar † Present address Laboratoire de Physique des Lasers, CNRS- to that described in reference [26], we can also calculate UMR7538, Universit´e Paris 13–Institut Galil´ee, Villetaneuse, France the generalized q and a stability parameters: qx, qy, qz ‡ Also at ICFO-Institut de Ciencies Fotoniques, Mediterranean a , a and a . As an example, Fig. 2 shows three theo- x y z TechnologyPark,08860Castelldefels(Barcelona),Spain retical radial (xy) cross sections of the pseudo-potential 2 Control End cap End cap Control Control (a) (b) (c) End cap End cap Z 1mm FIG. 2. Pseudo-potential cross section calculated for three RF Y X sets of trap parameters. In all cases the ion-surface dis- tance is 504 µm, fixed by the trap geometry and q =0.173, x q =-0.171 and q =-0.002. (a) isotropic potential obtained FIG. 1. Picture of the gold-plated printed circuit board sur- y z for V =125 V, V =-2.25 V, V =-11 V and V =5 V. face trap realized by standard commercial technique on a rf CC LC EC The trap depth is 38 meV and the motional frequencies FR4 card without any further operation. The dimensions are ω =408 kHz, ω =404 kHz, ω =156 kHz. The stabil- are: 500 µm wide rf tracks and 340 µm wide central elec- x y z ity parameters are a =-0.001, a =-0.001 and a =-0.002. (b) trode. The inter-electrode spacing is 200 µm. To minimize x y z anisotropic potential obtained for V =125 V, V =3.73 V, the rf coupling, surface-mount capacitors (10 nF) (visible on rf CC V =5 V and V =5 V. The trap depth is 141 meV thesidesofthepicture)arebondedbetweeneachdcelectrode LC EC and the motional frequencies are ω =266kHz, ω =529kHz, and the ground plane. x y ω =39kHz. The stability parameters are a =-0.009, a =- z x y 0.009 and a =0.0001. (c) anisotropic potential obtained z for V =125 V, V =-6 V, V =5 V and V =-21.05 V. rf CC LC EC obtainedwiththreesetsoftrappingparametersthatpro- The trap depth is 13 meV and the motional frequencies are duce anisotropic or isotropic potentials. ω =474kHz, ω =297kHz, ω =197kHz. The stability param- x y z Sr+ionsarecreatedinthetrappingregion(typicalrate etersarea =0.0039,a =-0.0072anda =0.0033. Isopotential x y z ∼ 20 s−1) out of an atomic vapor using a photoioniza- curves are separated by 10 meV. tion technique based on two-photon absorption of fem- tosecond pulses [27]. The ions are Doppler cooled using the 711 THz 52S1/2 → 52P1/2 optical transition (λ = 52P1/2 state (Γ/2π =21.7 MHz). 422 nm). To avoid optical pumping into the metastable A first test of the pseudo-potential calculation is per- 42D state we use an additional laser adressing the formed by comparing the experimentally measured mo- 3/2 275 THz 42D → 52P transition (λ = 1092 nm). tional frequencies with an harmonic fit of the calculated 3/2 1/2 The laser set-up is very similar to the one described in pseudo-potential at the trap center. The motional fre- reference 28. The trap is placed in a UHV chamber with quencies are obtained by measuring the single-ion flu- an estimated pressure below 10−9 mbar. orescence intensity as function of the frequency of the excitationvoltage(tickle)appliedtoanendcapelectrode In the experiment, voltages slightly different from the (seeinsetofFig.3). Theresults,showninFig.3,present ideal symmetric case have to be applied to the control a very good agreement between calculation and experi- electrodes in order to position a single trapped ion pre- ment. cisely at the node of the rf electric field. This allows for the reduction of the micro-motion of the ion driven by the rf electric field. In order to optimize these voltages, we used the rf correlation technique [29] that measures the arrival time correlations of single fluorescence pho- tons with the rf cycle. By using purely longitudinal (z- propagating) and transverse (propagating at 45◦ in the xzplane)coolingbeamswewereabletocompensatefora residual axial micro-motion and the micro-motion along x. In order to compensate for the micro-motion along y (vertical), we implemented the technique developed by Allcock and co-workers, based on a vertical repumping beam [30]. Weobservestabletrappingandmicro-motioncompen- FIG. 3. Measured (dots) and calculated (lines) motional fre- sation voltages which do not vary significantly over a quencies as function of Vrf. The inset shows an example of an excitation spectrum i.e. single-ion fluorescence intensity 30 min time-scale. However, optimal compensation volt- as a function of excitation frequency at a fixed V . ages are not stable on a day-to-day basis (typical drifts rf of 70 V/m). With optimized compensation voltages, the single ion lifetime is ∼ 20 min, probably limited by the Typical fluorescence images of the trapped ions ob- pressure in the vacuum chamber. Single ion fluorescence tained for different trap parameters are shown in Fig- spectraconfirmalinewidthlimitedbythelifetimeofthe ure 4. In Fig. 4(a) the ions organize themselves as a 3 large three dimensional Coulomb crystal. We observed to the small value of the trap depth (13 meV). The 3D crystals containing up to a few thousand ions, com- trap is actually loaded using parameters close to those parable to the typical numbers obtained in three dimen- used in Fig. 2(c) but with a higher value of V , in CC sional macroscopic linear Paul traps [31]. order to increase the trap depth. Then, V is reduced CC and one can observe the formation of a mono-layer ion crystal perpendicular to the trap surface.With the current trap design we could actually trap up X (a) XX (b()b) XX (c) to sixteen ions in this configuration. As mentioned 500 µm above, these 2D structures may be exploited for the 340 µm YYY quantum simulation of two dimensional systems [32]. In particular, the ”vertical” arrangement could allow YY for an easier ion addressing by lasers since the control Z ZZ Z beamscouldfreelypropagateparalleltothetrapsurface. In this work, we have demonstrated the versatility FIG.4. Fluorescenceimagesofthetrappedions. (a)top-view offered by an inexpensive, easily fabricated rf surface ofalarge3DCoulombcrystalcontaining∼4500ions. (b)2D ion trap based on a printed-circuit board. We have Coulomb crystal containing ∼ 150 ions arranged in a plane shown that large Sr+ ion crystals arranged in a 3D parallel to the trap surface (the inter-ion distance is 11 µm). structure can be obtained, comparable in size to the Thesingle-layercharacterisevidencedbythelateralview. (c) crystals used in recent cavity quantum electrodynamics Single-layercrystalarrangedintheyz planeperpendicularto the printed circuit board (the inter-ion distance is 9 µm). experiments[33]. The same device allowed us to create large ion Coulomb crystals purely bi-dimensional lying parallel to the trap substrate and containing up to 150 Another and particularly interesting configuration ions. In addition, we also demonstrated 2D Coulomb is the single layer Coulomb crystal, demonstrated in crystals standing in a plane perpendicular to the trap Fig. 4(b). Using an imaging system aligned along the surface. This particular geometry allows for an easier x direction, we have actually checked the single layer individual addressing of ions, especially useful in view character of this Coulomb crystal (bottom image). of quantum simulation experiments. This kind of The non-fluorescing ions visible on the top view of the versatile devices will probably become a practical tool crystal (left side) are sympathetically-cooled strontium for quantum simulation experiments. isotopes not addressed by the cooling lasers (only 88Sr+ is laser-cooled in this experiment). Trap potential We thank M. Apfel and P. Lepert for technical sup- calculations allowed us to find a very unusual working port. WewouldliketothankD.T.C.Allcockforfruitful point in which the Coulomb crystal forms a single layer discussions and suggestions. We acknowledge financial perpendicular to the trap surface, as shown in Fig. 4(c). support by R´egion Ile-de-France through the SESAME However, contrary to the two previous cases, ions could project. B. S. gratefully acknowledges the funding from not be directly loaded using the calculated parameters the D´el´egation G´en´erale de l’Armement and the French (see caption of Fig. 2c for this case), propably due Ministry of Education and Research. [1] J.I.CiracandP.Zoller,Phys.Rev.Lett.74,4091(1995). Blakestad,R.J.Epstein,D.B.Hume,W.M.Itano,J.D. [2] D. Leibfried, R. Blatt, C. Monroe, and D. Wineland, Jost,C.Langer,R.Ozeri,N.Shiga, andD.J.Wineland, Rev. Mod. Phys. 75, 281 (2003). Phys. Rev. Lett. 96, 253003 (2006). [3] H.Ha¨ffner,C.Roos, andR.Blatt,Phys.Rep.469,155 [11] J. Britton, D. Leibfried, J. A. Beall, R. B. Blakestad, (2008). J.H.Wesenberg, andD.J.Wineland,Appl.Phys.Lett. [4] D. Porras and J. I. Cirac, Phys. Rev. Lett. 92, 207901 95, 173102 (2009). (2004). [12] D. L. Moehring, C. Highstrete, D. Stick, K. M. Fortier, [5] M.Johanning,A.F.Varo`n, andC.Wunderlich,J.Phys. R.Haltli,C.Tigges, andM.G.Blain,arXiv:1105.1834v1 B: At. Mol. Opt. Phys. 42, 154009 (2009). [quant-ph] (2011). [6] S. Willitsch, Nat. Phys. 6, 240 (2010). [13] D. Porras and J. I. Cirac, Phys. Rev. Lett. 93, 263602 [7] H. S. Margolis, Contemporary Physics 51, 37 (2010). (2004). [8] D. Kielpinski, C. Monroe, and D. J. Wineland, Nature [14] E. Shimshoni, G. Morigi, and S. Fishman, Phys. Rev. 417, 709 (2002). Lett. 106, 010401 (2011). [9] J. Chiaverini, R. B. Blakestad, J. Britton, J. D. Jost, [15] A. Bermudez, J. Almeida, F. Schmidt-Kaler, A. Ret- C. Langer, D. Leibfried, R. Ozeri, and D. J. Wineland, zker, andM.B.Plenio,“arxiv:1108.1024v1[quant-ph],” Quantum Inf. Comput. 5, 419 (2005). (2011). [10] S. Seidelin, J. Chiaverini, R. Reichle, J. J. Bollinger, [16] T. B. Mitchell, J. J. Bollinger, D. H. E. Dubin, X.-P. D. Leibfried, J. Britton, J. H. Wesenberg, R. B. Huang, W. M. Itano, and R. H. Baughman, Science 4 282, 1290 (1998). (2011). [17] M. J. Biercuk, H. Uys, A. P. VanDevender, N. Shiga, [27] S. Removille, R. Dubessy, Q. Glorieux, S. Guibal, W.M.Itano, andJ.J.Bollinger,QuantumInf.Comput. T. Coudreau, L. Guidoni, and J.-P. Likforman, Appl. 9, 920 (2009). Phys. B 97, 47 (2009). [18] R. Schmied, J. H. Wesenberg, and D. Leibfried, Phys. [28] S. Removille, R. Dubessy, B. Dubost, Q. Glorieux, Rev. Lett. 102, 233002 (2009). T. Coudreau, S. Guibal, J.-P. Likforman, and [19] R. J. Clark, T. Lin, K. R. Brown, and I. L. Chuang, J. L.Guidoni,JournalofPhysicsB:Atomic,Molecularand Appl. Phys. 105, 013114 (2009). Optical Physics 42, 154014 (2009). [20] C. E. Pearson, D. R. Leibrandt, W. S. Bakr, W. J. Mal- [29] D. J. Berkeland, J. D. Miller, J. C. Bergquist, W. M. lard, K. R. Brown, and I. L. Chuang, Phys. Rev. A 73, Itano, and D. J. Wineland, Journal of Applied Physics 032307 (2006). 83, 5025 (1998). [21] T. H. Kim, P. F. Herskind, T. Kim, J. Kim, and I. L. [30] D. T. C. Allcock, J. A. Sherman, D. N. Stacey, A. H. Chuang, Phys. Rev. A 82, 043412 (2010). Burrell,M.J.Curtis,G.Imreh,N.M.Linke,D.J.Szwer, [22] R. J. Clark, Z. Lin, K. S. Diab, and I. L. Chuang, J. S. C. Webster, A. M. Steane, and D. M. Lucas, New J. Appl. Phys. 109, 076103 (2011). Phys. 12, 053026 (2010). [23] K. R. Brown, R. J. Clark, J. Labaziewicz, P. Richerme, [31] M. Drewsen, C. Brodersen, L. Hornekær, J. S. Hangst, D. R. Leibrandt, and I. L. Chuang, Phys. Rev. A 75, and J. P. Schifffer, Phys. Rev. Lett. 81, 2878 (1998). 015401 (2007). [32] D. Porras and J. I. Cirac, Phys. Rev. Lett. 96, 250501 [24] C. Rouki, L. Westerberg, and the CHICSi Develop- (2006). ment Group, Physica Scripta 2003, 107 (2003). [33] P.F.Herskind,A.Dantan,J.P.Marler,M.Albert, and [25] M. G. House, Phys. Rev. A 78, 033402 (2008). M. Drewsen, Nat. Phys. 5, 494 (2009). [26] A. Ozakin and F. Shaikh, arXiv:1109.2160v1 [quant-ph]